On the promoting effect of Au on CO oxidation kinetics of Au-Pt bimetallic nanoparticles supported on SiO2: An electronic effect?

Article abstract of DOI:10.1016/j.jcat.2011.12.011

Bimetallic AuPt nanoparticles prepared by radiolysis with Au/Pt atomic ratios of 1.2 and 1.8 were deposited onto silica and calcined to remove the stabilizing polymers. During this sequence, they lose their core-shell structure and restructure as alloys. FTIR and XPS provide evidence for further AuPt NP restructuring under CO exposure and CO oxidation with Pt surface enrichment. A promoting effect of Au is found on the CO oxidation kinetics in the AuPt catalyst with the lower Au/Pt ratio. This effect is attributed to an electron enrichment of the Pt surface atoms due to charge transfer from Au to Pt, as indicated by XPS and CO-FTIR.

Full text of DOI:10.1016/j.jcat.2011.12.011

Journal of Catalysis 287 (2012) 102–113
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
On the promoting effect of Au on CO oxidation kinetics of Au–Pt bimetallic
nanoparticles supported on SiO2: An electronic effect?
Rachel P. Doherty a,b,c,d, Jean-Marc Krafft a,b, Christophe Méthivier a,b, Sandra Casale a,b, Hynd Remita c,d
,
Catherine Louis a,b, Cyril Thomas a,b,
⇑
a UPMC, UMR 7197, Laboratoire de Réactivité de Surface, Case 178, 4 Place Jussieu, F-75005 Paris, France
b CNRS, UMR 7197, Laboratoire de Réactivité de Surface, 4 Place Jussieu, Case 178, F-75005 Paris, France
c Université de Paris-Sud, UMR 8000, Laboratoire de Chimie Physique, 15 rue Georges Clemenceau, F-91405 Orsay, France
d CNRS, UMR 8000, Laboratoire de Chimie Physique, 15 rue Georges Clemenceau, F-91405 Orsay, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 27 July 2011
Revised 19 October 2011
Accepted 11 December 2011
Available online 16 January 2012
Bimetallic AuPt nanoparticles prepared by radiolysis with Au/Pt atomic ratios of 1.2 and 1.8 were depos-
ited onto silica and calcined to remove the stabilizing polymers. During this sequence, they lose their
core–shell structure and restructure as alloys. FTIR and XPS provide evidence for further AuPt NP restruc-
turing under CO exposure and CO oxidation with Pt surface enrichment. A promoting effect of Au is found
on the CO oxidation kinetics in the AuPt catalyst with the lower Au/Pt ratio. This effect is attributed to an
electron enrichment of the Pt surface atoms due to charge transfer from Au to Pt, as indicated by XPS and
CO-FTIR.
Keywords:
Radiolysis
PtAu alloys
Ó 2011 Elsevier Inc. All rights reserved.
CO-FTIR
Restructuring of nanoparticles
CO oxidation kinetics
Turnover rates
Electron transfer
1. Introduction
Silica was chosen as the support as it is relatively inert toward
catalytic reactions [20] and does not exhibit strong metal-support
Supported bimetallic nanoparticles have been shown to typi-
cally exhibit enhanced catalytic properties relative to their mono-
metallic counterparts [1–8]. Recently, a significant amount of
progress has been made in the preparation of supported bimetallic
AuPt nanoparticles by dendrimer encapsulation [6,9,10] or from a
single-source organometallic precursor such as Pt2Au4(C„CtBu)8
[11,12], improving on traditional impregnation methods, which
do not allow sufficient control over nanoparticle size, structure
and composition. However, these preparation routes require the
use of complex organic or organometallic molecules. Radiolysis
has been shown to be a valuable tool in the preparation of bimetal-
lic nanoparticles [13–15]. Here, simple polymers are employed as
stabilizing agents allowing solution synthesis of bimetallic nano-
particles of controlled size and structure [15–18]. Radiolysis has
also been shown to overcome the difficulty of preparing AuPt
nanoparticles due to the Au/Pt miscibility gap from 18% to 98%
Pt [19].
interactions, in comparison to other available oxides such as TiO2
[21], which has been shown to influence the catalytic properties
of the metal nanoparticles [22].
The catalytic performance of SiO2-supported AuPt nanoparticles
has been previously studied [6,9,12,23,24]. Bimetallic AuPt nano-
particles have been reported to exhibit enhanced catalytic proper-
ties in comparison to their monometallic counterparts for instance
for CO oxidation [6,9]. Explanations for the improvement in the cat-
alytic behavior of supported AuPt nanoparticles include the pres-
ence of geometric [9] and electronic effects [7,11,12,25,26].
Reports to date of electronic effects on AuPt bimetallic nanoparticles
have been contradictory. Some have reported possible electronic ef-
fects with electron transfer from Pt to Au [26] as in the direction ex-
pected by considering the difference in electronegativity of the two
elements, while others have reported charge transfer in the opposite
direction from Au to Pt [7,11,12,25]. For the Au-5d-transition metal
alloying, calculations have revealed that charge transfer between
the two metals does not follow their respective electronegativity
[27].
⇑
Corresponding author. Address: Laboratoire de Réactivité de Surface, UPMC,
In this article, for the first time, the kinetics of CO oxidation of
SiO2-supported AuPt nanoparticles is measured under stoichiome-
tric conditions (2CO/O2 = 1) to gain further insights into the
UMR CNRS 7197,
4 place Jussieu, Case 178, F-75252 Paris Cedex 05, France.
Fax: + 33 1 44 27 60 33.
E-mail address: cyril.thomas@upmc.fr (C. Thomas).
0021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.12.011

R.P. Doherty et al. / Journal of Catalysis 287 (2012) 102–113
103
potential promoting effect of Au on the CO oxidation reaction cat-
2.2.2. UV–vis spectroscopy
alyzed by Pt surface sites of bimetallic AuPt nanoparticles. The re-
sults are compared with those of SiO2-supported Pt and Au, the full
kinetics of the latter is also measured for the first time under these
particular reaction conditions. Indeed, if CO oxidation kinetics of
SiO2-supported Pt nanoparticles has been documented [28], that
of SiO2-supported Au is to our knowledge not reported [29], prob-
ably due to the poor activity of Au/SiO2 catalysts for this reaction.
Recently, Chandler et al. reported CO oxidation kinetics in oxidiz-
ing mixtures (0.1 < 2CO/O2 < 0.3) on AuNi or Au nanoparticles pre-
pared by dendrimer encapsulation [30]. These authors did not
determine the reaction order with respect to CO, which may, how-
ever, have provided valuable information.
UV–visible spectra in liquid phase were obtained in a 2-mm
quartz cell using a HP diode array HP8453 spectrophotometer.
2.2.3. Thermal gravimetric analysis (TGA-MS)
TGA experiments were performed using a SDT Q600 instrument
(TA Instruments) coupled with a mass spectrometer (MS) (Ther-
mostar GDS 301T3, Pfeiffer). Samples were heated at a rate of
3 °C minꢀ1 under air (100 ml minꢀ1).
2.2.4. X-ray diffraction (XRD)
XRD patterns of the prepared samples were obtained on a SIE-
MENS D500 diffractometer with a Cu Ka monochromatized radia-
tion (0.1548 nm) operated at 30 kV and 300 mA from 10° to 90°
with a scanning rate of 0.013° sꢀ1
.
2. Experimental
2.2.5. H2 and CO chemisorption
2.1. Catalyst preparation
Platinum accessibility was evaluated via the irreversible chemi-
sorption method using H2 and CO as probe molecules. These mea-
surements were performed in a static mode at 25 °C using a
conventional volumetric apparatus (Belsorp max, Bel Japan). Typi-
cally, 0.14 g of catalyst was used. Before the H2 or CO chemisorp-
tion measurements, the catalyst was first reduced in H2 (50 mL/
min) at 500 °C for 2 h (3 °C/min heating rate) with subsequent
evacuation at 400 °C for an additional 2 h. Then, the sample was
cooled down under vacuum to 25 °C. Two H2 or CO adsorption iso-
therms were obtained. After the first isotherm, the catalyst was
evacuated again for 2 h at 25 °C. The amounts of total and revers-
ible H2 or CO uptakes were then estimated by extrapolating the
quasi-linear portions of the isotherm to zero pressure. The differ-
ence between these two values gave the amount of irreversible
H2 or CO uptake, from which the number of accessible Pt atoms
was calculated assuming stoichiometries of 1 hydrogen atom or
1 CO molecule per surface Pt atom.
The monometallic Pt/SiO2 catalyst (1.96 wt.% Pt) was prepared
by incipient wetness impregnation (IWI) of SiO2 (Degussa Aerosil
380) with an aqueous solution of 0.04 mol Lꢀ1 Pt(NH3)4(NO3)2
(99.995%, Sigma–Aldrich), using a volume of solution of 2.5 ml gꢀ1
of SiO2. After aging at room temperature for 6 h, the catalyst was
dried overnight at 120 °C. Au/SiO2 (0.70 wt.% Au) was also pre-
pared by IWI of SiO2 with an aqueous solution of 0.02 mol Lꢀ1
HAuCl4 (HAuCl4.3H2O, 99.99%, Sigma–Aldrich). After aging at room
temperature for 6 h, the sample was washed with a 1 mol Lꢀ1 NH3
aqueous solution adjusted to pH 8.5 with HCl according to the pro-
tocol outlined by Delannoy et al. [31] to remove the chloride ions
and avoid the formation of large gold particles after thermal treat-
ment. This was followed by washing with water several times, un-
til a test by AgNO3 no longer showed the presence of residual Clꢀ,
before drying at 60 °C for 48 h.
Bimetallic AuPt nanoparticles were prepared by radiolytic co-
reduction [13,14] of
a 50-ml aqueous solution of HAuCl4
(5 ꢁ 10ꢀ4 mol Lꢀ1) and H2PtCl6 (5 ꢁ 10ꢀ4 mol Lꢀ1) (in the presence
of the stabilizing polymers, polyacrylic acid (0.05 mol Lꢀ1 PAA,
2000 g molꢀ1, Sigma–Aldrich) and polyvinyl alcohol (0.05 mol Lꢀ1
PVA, 85,000–124,000 g molꢀ1 Sigma–Aldrich) at a ratio of 1:1, for
which nanoparticle deposition onto the SiO2 support was found
to be the most effective among the different ratios investigated.
2-propanol (0.1 mol Lꢀ1 Sigma–Aldrich) was added as a radical
scavenger. Solutions were freshly prepared with a Au:Pt ratio of
1:1 in the absence of light to prevent possible photochemical
reduction of Au and bubbled with N2 prior to irradiation to remove
oxygen. Radiolysis of the solution was carried out using a 60Co pan-
2.2.6. Transmission electron microscopy (TEM)
TEM analysis of the metal particles was performed using a JEOL
100 CX II microscope. The size limit for Pt and Au particles detec-
tion is in principle about 1 nm. Particle size measurements were
performed particle by particle, using ITEM software on digitized
micrographs. The average metal particle sizes, d, were determined
from the measurement of at least 130 metal particles, and d was
calculated using the following formula: d =
Rnidi/Rni, where ni is
the number of particles of diameter di. High-resolution TEM and
EDX were performed using a JEOL 2010 microscope.
2.2.7. X-ray photoelectron spectroscopy (XPS)
XPS spectra were collected on a SPECS (Phoibos MCD 100) X-ray
photoelectron spectrometer, using a Al Ka (hm = 1486.6 eV) mono-
chromated radiation source having a 400 W electron beam power
and a 0.5 ꢁ 2 mm spot size. The emission of photoelectrons from
the sample was analyzed at a takeoff angle of 90° under ultra high
vacuum conditions (1 ꢁ 10ꢀ8 Pa). XP spectra were collected at pass
energy of 10 eV for C1s, Si2p, Pt4f and Au4f core XPS levels. Charge
compensation was applied during acquisition with the use of a
oramic gamma source [13] with
a radiation dose rate of
2.2 kGy hꢀ1. Deposition of the radiolytically prepared nanoparti-
cles was achieved by stirring SiO2 into the nanoparticle solution
for 24 h. The resulting AuPt/SiO2 composites were washed repeat-
edly with water, to remove excess polymer, unwanted reaction by-
products and until a AgNO3 solution no longer tested positive for
the presence of Clꢀ ions. The resulting AuPt/SiO2 samples were
dried at 60 °C for 18 h before being finely ground.
Calcination of the AuPt/SiO2 samples was performed at 400 °C
for 2 h in O2 (20%)/He (100 ml minꢀ1) in a quartz, fixed-bed,
continuous-flow reactor.
flood gun (4 eV, 300 lA emission current). Spectrum processing
was carried out using the Casa XPS software package and Origin
7.1 (Origin Lab Corporation). All materials were analyzed after
being stored in air under ambient conditions following their pre-
treatment. For Pt/SiO2 and AuPt(1.2)/SiO2 initially reduced at
500 °C under H2, XPS analysis was also carried out after the
catalysts were reduced in the pretreatment chamber of the XPS
facility at 200 and 400 °C, respectively, for 10 min in flowing
H2(10%)/Ar, thus avoiding sample exposure to the ambient atmo-
sphere. These particular samples will be denoted as ‘‘reduced
2.2. Catalyst characterization
2.2.1. Chemical analysis
Chemical analysis of the supported catalysts was performed by
inductively coupled plasma atom emission spectroscopy (ICP/AES)
at the CNRS Centre of Chemical Analysis (Vernaison, France).

104
R.P. Doherty et al. / Journal of Catalysis 287 (2012) 102–113
XPS’’ to be distinguished from the same catalysts reduced in H2 and
exposed to ambient air (denoted as ‘‘reduced’’).
The kinetic parameters reported in the present work were ob-
tained at steady-state, while the reactor was operated isothermally
and as close to a differential reactor as possible by limiting the con-
version to <10%. The CO reaction order was determined at 280 °C
under reactive gas mixtures by maintaining the O2 concentration
at 0.5% and varying the CO concentration from 0.5% to 1.5%. The
O2 reaction order was determined at the same temperature by
maintaining the CO concentration at 1.0% and varying the O2 con-
centration from 0.3% to 0.7%. The apparent activation energy was
determined under stoichiometric conditions (1.0% CO and 0.5%
O2 in He) at reaction temperatures of 270–300 °C.
2.2.8. Fourier transform infrared spectroscopy of adsorbed CO (CO-
FTIR)
CO-FTIR spectra of adsorbed CO on the SiO2-supported samples
were collected in transmission on a Bruker Vector 22 FTIR spec-
trometer equipped with a liquid N2-cooled MCT detector and a
data acquisition station. A total of 128 scans were averaged with
a spectral resolution of 2 cmꢀ1. The samples were pressed into
self-supporting wafers of about 8–12 mg cmꢀ2 (16–25 mg for wa-
fers of 16 mm diameter). The wafers were loaded in a moveable
glass sample holder, equipped on top with an iron magnet, and in-
serted in a conventional Pyrex-glass cell (CaF2 windows) connected
to a vacuum system. The iron magnet allowed for the transfer of
the catalyst sample from the oven-heated region to the infrared
light beam. Before CO adsorption, the catalysts were submitted
to a dynamic (50 cm3 minꢀ1) reducing pretreatment (5% H2 in Ar,
Air Liquide, 99.999%) at 500 °C for 2 h at atmospheric pressure.
The samples were then evacuated (7.5 ꢁ 10ꢀ7 Torr) at 500 °C for
30 min. Finally, the temperature was decreased to RT under dy-
namic vacuum. The samples were then exposed to CO (Air Liquide,
99.999%) additionally purified by passing through a liquid nitrogen
trap. The spectrum at RT of the pretreated sample was used as a
reference and subtracted from the spectra of the sample exposed
to CO.
3. Results and discussion
3.1. Catalyst characterization
Details of the prepared catalysts are summarized in Table 1.
Prior to analysis, the samples were either calcined in the case of
the bimetallics or reduced in the case of the monometallics. The
metal weight loadings of the monometallic references were 1.96
and 0.70 wt.% for Pt/SiO2 and Au/SiO2, respectively. The resulting
supported bimetallic catalysts were 0.45 wt.% Au and 0.25 wt.%
Pt on SiO2 (Au/Pt = 1.8) and 0.45 wt.% Au and 0.36 wt.% Pt on
SiO2 (Au/Pt = 1.2). From here on, these will be referred to as
AuPt(1.8)/SiO2 and AuPt(1.2)/SiO2, respectively. Overall, the
amount of Pt in the bimetallic catalysts was always lower than
the amount of Au.
2.3. Catalytic experiments
3.1.1. UV–vis of the AuPt colloids formed by radiolysis
2.3.1. Benzene hydrogenation
UV–vis absorption spectra of the PAA/PVA stabilized AuPt nano-
particles in solution prior to deposition onto SiO2 are displayed in
Fig. 1. A plasmon band at 540 nm is visible in the UV–vis spectrum
of the colloids corresponding to the higher Au/Pt ratio (1.8). This
plasmon resonance, which is well documented as being due to
the plasmonic resonance of nanosized Au particles [32,33], is al-
most, if not completely, non-existent in the spectrum of the nano-
particles with the lower Au/Pt ratio (1.2). The growth of bimetallic
AuPt nanoparticles by this method has previously been reported by
Remita and coworkers [14] who concluded that the nanoparticles
formed under similar conditions to ours were of a Aucore-Ptshell
structure. Indeed, previous experiments have shown that radioly-
sis of a solution containing PtCl24ꢀ and AuCl4ꢀ leads to formation
of Aucore–Ptshell NPs of typical diameter 3–4 nm at low dose rate
Before benzene hydrogenation, the catalyst sample (0.050 g
deposited on a plug of quartz wool inserted inside a quartz reactor)
was heated in flowing H2 (100 cm3 NTP minꢀ1) at atmospheric
pressure with a heating rate of 3 °C minꢀ1 up to 500 °C and held
at this temperature for 2 h. After cooling to 50 or 115 °C under
H2, the reaction was started. The partial pressure of benzene
(C6D6, Aldrich) was 51.8 Torr (1 Torr = 133 Pa), and the total flow
rate was 107 cm3 NTP minꢀ1 with H2 as balance. The composition
of the effluent was analyzed using an online gas chromatograph
(Hewlett Packard 5890, FID) equipped with a PONA (paraffins–ole-
fins–naphthenes–aromatics; HP, 50 m long, 0.20 mm i.d., 0.5 lm
film thickness) capillary column. Cyclohexane was the only prod-
uct detected.
(by
c-irradiation), or to alloyed Au–Pt nanoparticles of typical
2.3.2. CO oxidation kinetics
diameter 2–3 nm at high dose rate (electron beam irradiation)
[13,34]. This difference in the NP structure is due to electron trans-
fer from Pt complexes or nascent Pt0 induced by radiolysis to gold
complexes (AuIII or AuI) occurring only at low dose rate when the
reduction kinetics is slow. At high dose rate by electron beam irra-
diations, the reduction of all ions can be achieved rapidly before
any intermetallic electron transfer can occur, leading to alloyed
nanoparticles. In the case of a perfect Aucore–Ptshell structure, no
gold plasmon is expected, and this is indeed, very nearly, the case
for the AuPt colloids with Au/Pt = 1.2. Though it cannot be defini-
tively ascertained from this almost complete absence whether
the nanoparticles are core–shells or alloys, it strongly suggests that
the Au–Pt (1.2) sample is bimetallic. It is however appreciated that
the bimetallic nature of the nanoparticles can only be implied and
not fully determined by UV–vis spectra alone. Further results will
go onto strengthen and confirm this bimetallic interpretation of
the results. The plasmon resonance of the colloids of AuPt(1.8),
more Au-rich, is weak and implies that these nanoparticles are
either bimetallic with the presence of Pt in the particles diminish-
ing the intensity of the Au resonance at 540 nm, or that there is a
small percent of monometallic Au nanoparticles in an otherwise
Before runs, the catalyst samples (about 15 mg for Au/SiO2,
AuPt(1.2)/SiO2 and AuPt(1.8)/SiO2, and 10 mg of a mechanical mix-
ture of 10 wt% of Pt/SiO2 in SiO2) were submitted to a temperature-
programmed reduction from RT to 500 °C (3 °C minꢀ1) under H2
(Air Liquide, 100 cm3 NTP minꢀ1) and held at this temperature
for 2 h. The catalyst temperature was then decreased to 280 °C un-
der H2. The kinetic study was conducted in a U-type quartz dy-
namic differential microreactor (12 mm i.d.) with a total flow
rate of 50 mL minꢀ1. The synthetic gas mixture was fed from inde-
pendent mass flow controllers (Brooks 5850). The reactor outflow
was analyzed using a
l-GC (Varian, CP4900) equipped with two
channels. The first channel used a 5A molecular sieve column
(80 °C, 150 kPa He, 200 ms injection time, 30 s backflush time)
and was used to quantify O2 and CO. The second channel was
equipped with a poraplot Q column (100 °C, 150 kPa He, 50 ms
injection time, 6 s backflush time) and was used to quantify CO2.
CO conversions were calculated as follows: X CO (%) = [CO2]/
[CO]i ꢁ 100, where [CO2] and [CO]i were the concentration of CO2
measured at the outlet of the reactor and the initial concentration
of CO, respectively.

R.P. Doherty et al. / Journal of Catalysis 287 (2012) 102–113
105
Table 1
Summary of the supported catalysts prepared by IWI (monometallic) or radiolytic
determined by TEM are also shown at different catalyst preparation stages.
c reduction followed by deposition onto the SiO2 support (bimetallic). Metal particle diameters
Catalyst
Preparation
method
Au loading
(wt.%)
Pt loading
(wt.%)
Preparation stage
TEM size (nm) Number of sized
nanoparticles
Pt/SiO2
Impregnation
1.96
0.36
H2/500 °C
1.9
518
257
AuPt(1.2)/
SiO2
c
-irradiation
0.45
0.45
0.70
Solution (before deposition onto SiO2) 3.5
Deposited onto SiO2 and dried at 60 °C 3.7
Calcined at 400 °C/2 h
H2/500 °C followed by CO/O2 reaction 4.6
131
219
165
4.2
AuPt(1.8)/
SiO2
c
-irradiation
0.25
Solution (before deposition onto SiO2) 3.5
320
Deposited onto SiO2 and dried at 60 °C 3.4
Calcined at 400 °C/2 h
H2/500 °C followed by CO/O2 reaction 4.6
273
242
455
4.4
Au/SiO2
Impregnation
H2/500 °C 3.7
391
100
95
a
90
85
b
c
200
300
400
500
600
Temperature / °C
300
400
500
600
700
800
900
Fig. 2. TGA results on AuPt (1.8)/SiO2, (a) weight change (sample with no prior heat
treatment), (b) CO2 trace associated with weight change (a), (c) CO2 trace after
calcination at 400 °C for 2 h.
Wavelength / nm
Fig. 1. UV–visible spectra of the AuPt colloids in water (. . . AuPt(1.2), — AuPt(1.8)).
SiO2-supported bimetallic nanoparticles before and after calcina-
tion and after CO oxidation are shown in Fig. 3c and d. The similar-
ity in the lattice parameters of Au and Pt along with the broadness
of the peaks makes resolving the pattern of AuPt(1.2)/SiO2 difficult.
The pattern of AuPt(1.8)/SiO2 can be seen to contain additional
sharper peaks in the appropriate position for Au. These peaks are
believed to be due to the presence of a small number of larger
Au-only particles (8–12 nm). These larger particles, as hinted on
in the UV–visible section, will be later seen in the electron micros-
copy section and accounted for just 1% of the total number of
particles.
The Scherrer equation was used to calculate the average diam-
eter of the Au (4.4 nm) and Pt (22 nm) nanoparticles in the SiO2-
supported samples after reduction and the AuPt nanoparticles in
the AuPt(1.2)/SiO2 sample before (3.7 nm) and after (3.9 nm) calci-
nation (The Scherrer equation was only performed on the samples
exhibiting peaks with a symmetrical distribution of intensity). Uti-
lization of the Scherrer equation to calculate nanoparticle diame-
ters for AuPt(1.8)/SiO2 was hindered by the presence of the sharp
peak due to a small percentage of larger Au nanoparticles. On Pt/
SiO2, the size of the Pt nanoparticles obtained by this method
(22 nm) is considerably larger than expected, and it is suspected
that the XRD pattern in this case has been distorted and overpow-
ered by the presence of a small percentage of large Pt nanoparti-
cles, this was later confirmed by H2 and CO chemisorption
measurements (Section 3.1.4.) and electron microscopy (Section
3.1.5.).
bimetallic sample. XRD and TEM results will show later that the
last hypothesis is more probable.
3.1.2. Decomposition of stabilizing polymers followed by TGA
Before the catalytic properties of the bimetallic nanoparticles
supported on silica can be investigated, it is necessary to remove
the stabilizing polymers. In order to optimize the calcination tem-
perature and minimize potential nanoparticle sintering, TGA-MS
was used to investigate polymer decomposition as a function of
temperature, thus allowing the organic molecules to be decom-
posed at the lowest temperature possible. TGA-MS of AuPt(1.8)/
SiO2 (Fig. 2, trace b) shows a maximum in the release of CO2 at
400 °C which correlates with the most significant decrease in the
weight of the sample (Fig. 2, trace a). TGA-MS was repeated after
calcination at 400 °C (AuPt(1.8)/SiO2, 400 °C/2 h), verifying that
the polymers had been successfully removed (Fig. 2, trace c), no de-
crease in weight was observed for this sample. Similar results (not
shown) were also found for AuPt(1.2)/SiO2.
3.1.3. XRD
XRD patterns of the samples are shown in Fig. 3. The diffraction
patterns of reduced Pt/SiO2 (a) and Au/SiO2 (b) show diffraction
peaks consistent with those of Pt and Au nanoparticles,
respectively. The broad intense contribution at 23 ° is consistent
with the presence of amorphous SiO2. XRD patterns of the two

106
R.P. Doherty et al. / Journal of Catalysis 287 (2012) 102–113
c
a
b
(i)
(ii)
(iii)
d
(i)
(ii)
(iii)
20
40
60
80
20
40
60
80
Bragg Angle 2θ
Bragg Angle 2θ
Fig. 3. XRD patterns of SiO2-supported catalysts, (a) Pt/SiO2 reduced, (b) Au/SiO2 reduced, (c) AuPt(1.2)/SiO2, and (d) AuPt(1.8)/SiO2, (i) before calcination, (ii) after
calcination, (iii) after reduction and CO/O2 reaction. The peak labels denote the associated 2h values in degrees.
3.1.4. H2 and CO chemisorption
to 4.2–4.4 nm can be seen for both AuPt(1.2)/SiO2 and AuPt(1.8)/
SiO2. The diameters of the AuPt nanoparticles in the AuPt(1.2)/
SiO2 catalyst before (3.7 nm) and after calcination (4.2 nm) are in
good agreement with those obtained by Scherrer analysis of the
XRD data, 3.7 nm and 3.9 nm, respectively. A similar smaller fur-
ther increase to 4.6 nm is observed after the AuPt(1.2)/SiO2 and
AuPt(1.8)/SiO2 catalysts were reduced (500 °C/2 h/H2) and exposed
to CO/O2 at up to 300 °C for 30 h. Calcination and reduction steps
therefore result in a small degree of sintering of the supported
nanoparticles.
The nanoparticle size distribution in the AuPt samples at each
stage is reasonably narrow in solution and after deposition on sil-
ica, but is broadened to a small degree after calcination (Fig. 4). H2
reduction followed by CO/O2 is shown to have no significant effect
on the size distribution.
H2 and CO chemisorption data (not shown) lead to a Pt disper-
sion of 30.9% and 32.5%, respectively, rounded to 31.7 0.8% for the
Pt/SiO2 catalyst, to which corresponds a mean particle size of the Pt
particles of about 3.5 nm [35].
The first and second isotherms of H2 or CO overlap on AuPt(1.2)/
SiO2 (not shown). This prevents the use of H2 and CO chemisorp-
tion as a tool to estimate the number of Pt atoms on the surface
of the bimetallic particles and indicates that the number of Pt sur-
face atoms is extremely low on the bimetallic particles, in agree-
ment with the CO-FTIR and benzene hydrogenation results which
will be shown later (Sections 3.1.7 and 3.2.1).
3.1.5. TEM
TEM images were taken of Pt/SiO2 and Au/SiO2 after reduction,
and of AuPt(1.2)/SiO2 and AuPt(1.8)/SiO2 after different catalyst
preparation stages. The average nanoparticle sizes are listed in Ta-
ble 1. Supported Pt and Au monometallic nanoparticles were found
to be on average 1.9 nm and 3.7 nm in diameter, respectively. The
size of the Au/SiO2 nanoparticles is in reasonable agreement with
that obtained by the Scherrer equation (4.4 nm). However, as sus-
pected, TEM images showed that the Pt/SiO2 sample contained a
small percentage of large (>20 nm) Pt nanoparticles. These ac-
counted for less than 1% of the nanoparticles and are responsible
for the lack of coherence between the Scherrer (22 nm) and TEM
(1.9 nm) size estimations. The fact that the mean Pt particle size
estimated to be 3.5 nm from the chemisorption data (Section
3.1.4) is about twice as big as that estimated from TEM is attrib-
uted to the presence of the large particles, as detected by XRD (Sec-
tion 3.1.3). For the calculation of the benzene hydrogenation and
CO oxidation turnover rates, which will be shown later (Section
3.2), the Pt dispersion estimated from the chemisorption data
(31.7%) was thus considered.
The two bimetallic AuPt/SiO2 samples follow the same trend.
The initial nanoparticles deposited onto the TEM grid from solution
immediately after radiolysis, are, for both AuPt/SiO2 samples,
3.5 nm in diameter on average. After deposition onto the SiO2,
there is a small but not significant change (60.2 nm) in the average
size of both samples, potentially due to the poorer contrast be-
tween SiO2 and the nanoparticles, making visualization of the
nanoparticles, particularly those with the smallest diameters, more
difficult. After calcination (400 °C/2 h/air), an increase in diameter
A representative HRTEM image of AuPt(1.8)/SiO2 after CO oxi-
dation is shown in Fig. 5a. The d-spacings between adjacent lattice
fringes of the AuPt (1.8)/SiO2 catalyst after exposure to CO/O2 were
measured for several nanoparticles and found to be 2.30 Å. This va-
lue lies between bulk fcc-Au (111) at 2.355 Å and bulk fcc-Pt (111)
at 2.265 Å, which is a good indication of alloy structuring. Assum-
ing there to be a linear trend between composition and lattice
spacing, which has been reported for AuPt nanoparticles [36,37],
one can calculate that the Au/Pt atomic ratio is equal to 1.6, which
is close to the ratio given by chemical analysis.
EDX analysis was performed on the isolated AuPt nanoparticle
of the AuPt(1.8)/SiO2 sample shown in Fig. 5a, using a beam spot
size of 10–15 nm. The presence of both Au and Pt in the same
nanoparticle can clearly be seen on the EDX spectrum (Fig. 5b),
confirming its bimetallic nature. The measurement was performed
on five other isolated nanoparticles ranging in size from 5.1 to
7.8 nm within the AuPt(1.8)/SiO2 sample, all showing the presence
of both Au and Pt with an almost constant Au/Pt intensity ratio of
between 1.0 and 1.3 (these values are not quantitative as no cali-
bration was made). Because of the lack of any contrast between
the two elements Au and Pt, the HRTEM images cannot provide
any information on the particle structure.
3.1.6. XPS
XPS was used to investigate changes in the electron density of Pt
and in the nanoparticle composition at different catalyst prepara-
tion stages of the AuPt(1.2)/SiO2 sample. The spectra showing the

R.P. Doherty et al. / Journal of Catalysis 287 (2012) 102–113
107
A
average = 3.7 nmb
average = 3.5 nm a
2
4
6
8
10
12
2
4
6
8
10
12
Diameter / nm
Diameter / nm
average = 4.2 nmc
average = 4.6 nm d
2
4
6
8
10
12
2
4
6
8
10
12
Diameter / nm
Diameter / nm
average = 3.4 nm
average = 3.5 nm
B
b
a
2
4
6
8
10
12
2
4
6
8
10
12
Diameter / nm
Diameter / nm
average = 4.6 nm d
average = 4.4 nm
c
2
4
6
8
10
12
2
4
6
8
10
12
Diameter / nm
Diameter / nm
Fig. 4. Size distributions of AuPt nanoparticles; (A) AuPt(1.2)/SiO2, and (B) AuPt (1.8)/SiO2 at different catalyst preparation stages, (a) in solution, (b) after deposition onto SiO2
and drying at 60 °C, (c) after calcination at 400 °C, (d) after reduction at 500 °C followed by reaction CO/O2.
Au 4f7/2 and Pt 4f7/2 core-level peaks of Pt/SiO2, AuPt(1.2)/SiO2 and
Au/SiO2 are displayed in Fig. 6. The corresponding XPS data (BE,
FWMH and atomic compositions) are listed in Table 2. The spectra
(Fig. 6) clearly show the presence of metallic Au and/or Pt in the
nanoparticles, whose binding energies agree with those reported
previously [38,39], i.e. about 84 and 71 eV for Au 4f7/2 and Pt 4f7/2
core-levels (Table 2). As also reported recently for AuPt particles
[40], it can be seen that the Pt 4f7/2 binding energy is lower for
AuPt(1.2)/SiO2 than for Pt/SiO2, whereas that of Au 4f7/2 for
AuPt(1.2)/SiO2 remains similar to that found for Au/SiO2 (Table 2).
The FWHM of the Au 4f7/2 and Pt 4f7/2 peaks is about 2.3 and
3.0 eV (Table 2), respectively. The FWHM of the Au 4f7/2 peaks is
much smaller than that reported by Radnik et al. on Au/SiO2 [41].
In contrast, the FWHM of the Pt 4f7/2 peaks is greater than that ex-
pected from literature data [42] even for the samples that were re-
duced in H2 at 500 °C (Table 2, reduced). This may be attributed to
the presence of oxidized Pt species, due to Pt reoxidation upon expo-
sure to air prior to transferring in the XPS facility, which would lead
to a broadening of the Pt 4f7/2 peaks. To ascertain such an assump-
tion, the Pt/SiO2 and AuPt(1.2)/SiO2 reduced samples were further

108
R.P. Doherty et al. / Journal of Catalysis 287 (2012) 102–113
together with the relatively narrow FWHM (ꢂ2.3 eV), demonstrates
that Au is present only as metallic Au in the samples investigated in
the present work, regardless of the pretreatments applied. More
interesting from the XPS results summarized in Table 2, the Pt 4f7/2
binding energy of AuPt(1.2)/SiO2 decreases to a significant extent
when further reduced in the pretreatment chamber of the XPS facil-
ity (70.0 eV, Table 2). It must also be emphasized that this binding
energy is about 1 eV lower than that expected for metallic Pt attest-
ing to an electron transfer from Au to Pt in the bimetallic particles, in
agreement with the CO-FTIR results which will be shown later (Sec-
tion 3.1.7).
a
Table 2 also shows that the Au/Pt ratio of the supported nano-
particles is changed significantly by the environmental conditions
to which they were exposed. The Au/Pt ratio of the dried AuPt(1.2)/
SiO2 sample is lower than that given by chemical analysis (Table 1,
Au/Pt ꢂ 1.2). This is consistent with the reduction of the Pt species
onto gold seeds and the formation of Ptcore–Aushell colloids in the
course of the radiolysis process [14]. After oxidizing and reductive
pretreatments, the Au/Pt ratio becomes much larger than that cal-
culated from the metal composition determined by chemical anal-
ysis, attesting to a surface gold enrichment in line with predictions
from thermodynamic data [26,43]. This shows that the nanoparti-
cles are not homogeneous alloys and that Au has a preference over
Pt to reside in the outer shell of the nanoparticles. These results
also attested that particle restructuring had occurred during calci-
nation and reduction steps. Finally, the Au/Pt ratio of the bimetallic
nanoparticles decreases to the value expected for the composition
of the homogeneous alloy after CO–O2 reaction, indicating that CO
favors the migration of Pt toward the surface of the nanoparticles,
as also already noticed earlier for Pd in AuPd [44–47] and Pt in PtCu
[47] alloys. Changes in the structure of the AuPt nanoparticles un-
der CO atmospheres have been further investigated by CO-FTIR.
2 nm
Pt
Au
b
Pt
Au
2
4
6
8
10
12
0
5
10
Energy / keV
15
20
Fig. 5. (a) HRTEM of one metal particle of AuPt (1.8)/SiO2 after reduction and CO/O2
reaction, (b) EDX spectrum of the nanoparticle in image (a).
3.1.7. CO-FTIR
Infrared spectroscopy of adsorbed CO was performed on Pt/
SiO2, Au/SiO2, AuPt(1.2)/SiO2 and AuPt(1.8)/SiO2 reduced in situ
(500 °C/2 h/5% H2 in Ar) prior to measurement. The results are dis-
played in Figs. 7 and 8. Fig. 7a (under 1.1 Torr of CO: black curve)
shows the FTIR-CO absorption spectrum of Pt/SiO2, which contains
absorption bands at 2084 cmꢀ1 and 1849 cmꢀ1. These bands are
attributed to the vibration of CO linearly (noted Pt–CO) and
bridged bound to Pt, respectively, according to literature [48–50].
After evacuation of gas-phase CO (Fig. 7a, gray curve), the Pt–CO
band shifts to lower wavenumbers at 2081 cmꢀ1. This is attributed
to a decrease in the coverage and the dipole–dipole coupling of the
Pt–CO species [50]. The FTIR-CO spectrum of Au/SiO2 displayed in
Fig. 7b contains a single band at 2116 cmꢀ1 due to linearly ad-
sorbed CO on reduced gold sites [44,51–56] (noted Au–CO). As in
the case of former studies, no bridge-bound Aun–CO species were
observed [44,57,58].
Initially, after exposure of AuPt(1.2)/SiO2 to an atmosphere of
CO (1.1 Torr, RT), a strong band assigned to Au–CO is observed at
2111 cmꢀ1 (Fig. 8a, spectrum (i)), which is in close agreement with
the Au–CO band observed for the monometallic Au/SiO2 catalyst
(2116 cmꢀ1, Fig. 7b). Two smaller and broader bands centered
around 2042 and 1870 cmꢀ1, and assigned to linear and bridged
Pt–CO species, are also observed. The band at 2042 cmꢀ1 is, how-
ever, significantly broader and shifted by 39 cmꢀ1 from that ob-
served for the monometallic Pt/SiO2 (2081 cmꢀ1, Fig. 7a). After
acquisition of spectrum (i) in Fig. 8a, the CO atmosphere was evac-
uated (10ꢀ6 mbar, RT), upon which a quasi-instantaneous disap-
pearance of the Au–CO peak, spectrum (i) in Fig. 8b, was
observed. The ease with which CO could be desorbed from Au sites
confirms that CO is weakly adsorbed to Au sites [9,12]. This disap-
pearance of the Au–CO band also coincides with an increase in
intensity, and a change in shape of the linear Pt–CO band, which
Pt/SiO2 reduced
Pt/SiO2 reduced XPS
AuPt(1.2)/SiO2 dried
AuPt(1.2)/SiO2 calcined
AuPt(1.2)/SiO2 reduced
AuPt(1.2)/SiO2 reduced XPS
AuPt(1.2)/SiO2 post CO-O2
Au/SiO2 reduced
90
85
80
75
70
65
60
Binding energy (eV)
Fig. 6. XP Spectra and the corresponding decomposition of Au4f and Pt4f peaks of Pt/
SiO2, AuPt(1.2)/SiO2 and Au/SiO2 after different pretreatments.
reduced in H2(10%)/Ar in the pretreatment chamber of the XPS facil-
ity for 10 min at 200 and 400 °C, respectively, thus avoiding expo-
sure to air before XPS analysis (samples denoted as reduced XPS in
Table 2). In this particular case, the FWHM of the Pt 4f7/2 peaks of
Pt/SiO2 and AuPt(1.2)/SiO2 decreases from 2.9 to 2.6 and 2.4 eV,
respectively. These FWHM values agree better with those reported
earlier for metallic Pt [42], confirming that the broadening of the
Pt 4f7/2 peaks is likely attributable to the presence of oxidized Pt spe-
cies due to exposure to air. It is also noteworthy that the FWHM of
the Au 4f7/2 peak of the AuPt(1.2)/SiO2 further reduced in the pre-
treatment chamber of the XPS facility remains unchanged. This,

R.P. Doherty et al. / Journal of Catalysis 287 (2012) 102–113
109
Table 2
XPS data for each catalyst after different pretreatments.
Catalyst
Pt/SiO2
Pretreatment
Pt 4f7/2 (FWHM) (eV)
Au 4f7/2 (FWHM) (eV)
Pt (at.%)
Au (at.%)
Si (at.%)
C (at.%)
Au/Pt ratio
Au/Si ratio (ꢁ10ꢀ3
–
)
Reduced
71.1 (2.9)
71.1 (2.6)
71.0 (3.1)
70.5 (3.0)
70.7 (2.9)
70.0 (2.4)
70.6 (3.1)
–
–
–
0.055
0.047
0.022
0.017
0.015
0.014
0.018
–
–
–
36.1
33.6
30.9
31.3
36.2
37.5
34.3
34.1
1.3
1.3
13.9
1.8
1.7
1.2
–
–
Reduced XPSa
Dried
–
AuPt(1.2)/SiO2
83.5 (2.2)
83.1 (2.4)
83.2 (2.3)
83.1 (2.3)
83.2 (2.4)
83.5 (2.2)
0.019
0.023
0.025
0.025
0.022
0.028
0.9
1.4
1.7
1.8
1.2
–
0.62
0.73
0.69
0.67
0.65
0.82
Calcined
Reduced
Reduced XPSb
Post CO/O2
Reduced
2.2
6.1
Au/SiO2
a
The reduced sample was further reduced in the pretreatment chamber of the XPS facility at 200 °C for 10 min in flowing H2(10%)/Ar before being re-analyzed.
The reduced sample was further reduced in the pretreatment chamber of the XPS facility at 400 °C for 10 min in flowing H2(10%)/Ar before being re-analyzed.
b
on Pt and therefore a decrease in the dipole–dipole coupling
between sites with the same vibration frequency [9,48] (dipole–
dipole coupling diminishes at a rate of rꢀ3, where r is the distance
between two coupling CO molecules). This indicates that Pt and Au
are intimately mixed within and on the surface of individual parti-
cles as already proved by Chandler and co-workers [6] and Mihut
et al. [12].
(a)
2081
0.05
2084
More interesting though is the spectral intensity redistribution
in the case of bimetallic samples. This has been already observed
by Chandler et al. [6] with AuPt/SiO2 samples obtained by deposi-
tion of AuPt in dendrimers. They attributed the spectral intensity
redistribution, i.e., the increase in intensity of the Pt–CO band as
CO molecules are desorbed from the Au sites, to surface restructur-
ing [6,9] and/or screening effect [6] between CO molecules
adsorbed onto the two different metals. According to Hollins
[48], since the frequency difference between the absorption of
1849
1850
2250
2150
2050
1950
1750
Wavenumber (cm-1)
Pt–CO (2042 cmꢀ1
) ) is smaller than
and Au–CO (2111 cmꢀ1
100 cmꢀ1, absorption intensity could be redistributed from the
lower-frequency peak (Pt–CO) to the higher-frequency peak
(Au–CO). This arises from the screening of electric fields from low-
er-frequency oscillators (Pt–CO), resulting in enhanced absorption
by higher-frequency bands (Au–CO). Therefore, as the higher-
energy bands are removed, the intensity borrowing from Pt–CO
would diminish, increasing the intensity of the Pt–CO band. The
screening effect of the Au–CO band on the Pt–CO band could also
explain the increase in intensity of the Pt–CO band as CO molecules
are desorbed from the Au sites, as observed here (Fig. 8a and b,
spectra i).
(b)
2116
0.002
A shift of 39 cmꢀ1 is, however, considerably larger than that
expected from coupling effects alone, since it is known that the
infrared spectrum of CO adsorbed on platinum typically exhibits
a coverage-dependent frequency shift of 30 cmꢀ1 [49,50]. Addi-
tional experiments (not shown) were performed to determine
2250
2150
2050
1950
1850
1750
Wavenumber (cm-1)
the Pt–CO singleton frequency (m
12CO free of dipole–dipole cou-
Fig. 7. CO-FTIR spectra of (a) Pt/SiO2
evacuation at RT) and (b) Au/SiO2 exposed to 1.1 Torr of CO.
(
exposed to 1.1 Torr of CO and
after
pling effect) of the Pt/SiO2 and AuPt(1.2)/SiO2 samples. To achieve
this goal, the CO coverage was lowered by evacuation of the sam-
ples at increasing temperatures, as reported earlier [50]. The Pt–CO
singleton frequency on Pt/SiO2 was estimated to be 2051 cmꢀ1, in
excellent agreement with that reported earlier by Primet via the
same methodology and via isotopic dilution of 12CO by 13CO
(2052 cmꢀ1) [50], whereas that on AuPt(1.2)/SiO2 was estimated
can now be seen to consist of a main band at 2046 cmꢀ1 and a
shoulder at 2064 cmꢀ1 (spectrum (i) in Fig. 8b). In contrast, the
intensity of the bridged Pt–CO was not significantly affected. The
same spectral intensity redistribution from Au–CO to Pt–CO was
observed for AuPt(1.8)/SiO2 (not shown). This phenomenon was
reversible and reproducible, from a qualitative point of view, upon
the re-introduction of CO (Fig. 8).
The broadness of the Pt–CO band in the bimetallic sample in
Fig. 8b can be explained by the presence of a variety of different
Pt sites available for CO adsorption on the nanoparticle surface.
The shift to lower frequency of 39 cmꢀ1 of the Pt–CO peak position
from the monometallic Pt catalyst (2081 cmꢀ1, Fig. 7a) to that of
bimetallic Pt–CO (2042 cmꢀ1) is certainly in part due to a geomet-
ric effect resulting from the dilution of Pt sites by Au [49]. This
dilution would result in a lower density of adsorbed CO molecules
to be 2041 cmꢀ1
.
The larger shift was therefore attributed to an electronic effect,
a shift in the absorption to lower frequency is indicative of a weak-
ening of the CO bond due to increased back-electron transfer into
the CO 2pꢃ antibonding orbital. This suggests that the presence
of Au has increased the electron density of Pt, as also supported
by the XPS data (Section 3.1.6). This electronically induced shift
is coherent with previously reported observations on AuPt/SiO2
systems [7,11,12,25]. The possibility of electron transfer from Au
to Pt despite the difference in electronegativity has been reported
in former studies, as indicated previously in Section 1.

110
R.P. Doherty et al. / Journal of Catalysis 287 (2012) 102–113
0.005
a
b
0.005
(iii)
(ii)
(i)
(iii)
(ii)
(i)
2300
2200
2100
2000
1900
1800
2300
2200
2100
2000
1900
1800
Wavenumber / cm-1
Wavenumber / cm-1
Fig. 8. CO-FTIR spectra of the comparison of consecutive CO exposure-removal cycles on AuPt(1.2)/SiO2: (a) after CO exposure and (b) after evacuation (10ꢀ6 mbar, RT, 0.5 h),
(i) 1st cycle (1.1 Torr CO, RT, 0.1 h), (ii) 2nd cycle (4 Torr CO, RT, 18 h), (iii) 3rd cycle (10.5 Torr CO, 100 °C, 18 h).
The change in the absorbance bands of AuPt(1.2)/SiO2 was
investigated further (Fig. 8). Two further cycles of CO exposure-
evacuation were performed on AuPt(1.2)/SiO2 (spectra ii and iii
in Fig. 8). After the first cycle of CO adsorption–desorption (spectra
i), the catalyst was again exposed to CO, this time at higher CO
pressure and for a longer duration (4 Torr, 18 h, RT). After confir-
mation that CO was re-adsorbed onto Au atoms (Fig. 8a, spectrum
(ii)), the CO was evacuated (10ꢀ6 mbar, RT) and the spectrum re-
corded again (Fig. 8b, spectrum (ii)) showing as before the in-
creased and narrowed Pt–CO band. The catalyst was again,
exposed to CO, a third time, at even higher CO pressure at 100 °C
(10.5 Torr CO, 18 h, 100 °C) (spectra iii). It is clear regarding these
three CO adsorption–desorption spectra (Fig. 8), that the intensity
of the Au–CO band diminishes and that of the two Pt–CO bands in-
creases with each cycle and prolonged exposure to CO. The in-
crease in intensity of the Pt–CO bands observed after further
exposure to CO at elevated temperatures in combination with
the decrease in the Au–CO peak intensity is good argument/evi-
dence for a restructuring of the nanoparticles, as Pt migrates out-
wards, replacing Au at the nanoparticle surface. The restructuring
can be explained thermodynamically since CO chemisorbs to Pt
more strongly than to Au and provides a driving force for pulling
Pt to the particle surface [19]. The surface energy of the nanoparti-
cle will be reduced by the increased bond formation with CO if Pt is
at the surface.
3.2. Catalytic results
3.2.1. Benzene hydrogenation
Benzene hydrogenation experiments were performed on all cat-
alysts at either 50 °C and/or 115 °C. Pt-based catalysts are known
to efficiently catalyze this reaction, and for this reason, it can be
considered as a test-reaction. Reaction rates are summarized in Ta-
ble 3. As expected, the monometallic Pt/SiO2 catalyst was active at
the lower temperature, and monometallic Au/SiO2 was inactive at
both temperatures. AuPt(1.8)/SiO2 was also found to be inactive at
both temperatures. AuPt(1.2)/SiO2 was inactive at 50 °C, but active
at 115 °C. In order to compare the hydrogenation rate of Pt/SiO2
and AuPt(1.2)/SiO2, it is first necessary to extrapolate the rate of
the Pt/SiO2 to 115 °C. This can be achieved since it is known that
the rate of reaction of benzene hydrogenation can be represented
by [59]:
0
0
Rate ¼ k½benzeneꢄ ½H2ꢄ ; where k ¼ eꢀE
ð1Þ
a=RT
thus allowing extrapolation of the hydrogenation rate to 115 °C, as
shown in the following equation:
RateT
RateT
kT
kT
1
2
1
2
¼
ð2Þ
The activation energy of benzene hydrogenation over a Pt/SiO2 cat-
alyst is known to be 47.4 kJ molꢀ1 [59]. Reaction rates are converted
into turnover rates (TOR), the rate of reaction per Pt surface atom,
taking into account the Pt dispersion. The number of Pt surface
atoms was calculated using the chemisorption data for Pt/SiO2 (Sec-
tion 3.1.4) and the average particle diameters determined by TEM
for AuPt/SiO2 (Section 3.1.5) and the method previously described
in [35]. In the bimetallic case, TOR is calculated assuming the nano-
particles are homogeneous alloys in the bulk and on the surface.
3.1.8. Summary on the characterization of the bimetallic nanoparticles
All these results provide evidence for the formation of bimetal-
lic AuPt particles the structure of which evolves with environment.
When the AuPt particles are still in solution and polymer stabi-
lized, the structure is Aucore–Ptshell according to former interpreta-
tion of the UV–Vis spectra [14]. After deposition onto silica and
drying, they become alloy-type with surface enrichment by Au
according to the XPS data, despite bulk AuPt having a large misci-
bility gap extending from 18% to 98 % Pt [19], these results show
consistently, as with former studies on supported AuPt catalysts
arising from organobimetallic precursors [11,12], preformed parti-
cles in dendrimers [9], vapor deposition [26] and in colloidals [36],
that bimetallic AuPt particles can be obtained. Moreover, as also
reported in [9,26,37], the bimetallic nanoparticles consist of an al-
loy-type structure. Under pure CO, further Pt enrichment occurs,
but Au is still present on the surface, as also shown by Chandler
and co-workers [6,9].
Table 3
Rate of reaction of benzene hydrogenation over each catalyst at two reaction
temperatures (n.d. = not determined, ꢃ = calculated).
Catalyst
Rate (ꢁ10ꢀ8 mol sꢀ1 gꢀ1
)
TOR (sꢀ1
)
50 °C
115 °C
50 °C
115 °C
Pt/SiO2
530
0
0
n.d.
14
0
0.17
3.20ꢃ
0.03ꢃ
0
AuPt(1.2)/SiO2
AuPt(1.8)/SiO2
Au/SiO2
0
0
0
0
0
0

R.P. Doherty et al. / Journal of Catalysis 287 (2012) 102–113
111
The benzene hydrogenation turnover rate of Pt/SiO2 at 50 °C is
0.17 sꢀ1, which is comparable with data reported previously [59]
and extrapolates to 3.20 sꢀ1 at 115 °C (Table 3). The TOR for the
Pt surface atoms of the bimetallic AuPt(1.2)/SiO2 catalyst at
115 °C is found to be 0.03 sꢀ1, which is considerably lower than
that of Pt/SiO2 (3.20 sꢀ1) (Table 3). The activity of the bimetallic
catalyst toward benzene hydrogenation is therefore considerably
lower than expected when the particles are assumed to be homo-
geneous alloys. These results suggest that the presence of gold
inhibits the reactivity of Pt in benzene hydrogenation; this can
be due to an electronic effect resulting from gold proximity that
modifies the electronic properties of Pt as attested by the XPS
and FTIR results (Section 3.1.6 and 3.1.7).
The fact that AuPt(1.8)/SiO2 is as inactive as the monometallic
Au catalyst is consistent with the lower proportion of Pt than in
AuPt(1.2)/SiO2.
Benzene hydrogenation of unsupported AuPt alloys of different
compositions has been investigated previously [60]. It has been
suggested that successful benzene hydrogenation requires ensem-
bles of Pt atoms, if Pt sites became too diluted by Au, activity is no
longer observed, which is coherent with the observations reported
here.
The plots used for the determination of the reaction orders
summarized in Table 4 are shown in Fig. 9. The reaction orders
for Pt/SiO2 were ꢀ0.88 with respect to CO and 1.02 with respect
to O2, which is in good agreement with those reported previously
in the same temperature range [63,64]. For Au/SiO2, the reaction
orders were 0.09 with respect to CO and 0.69 with respect to O2.
What is immediately clear from these results is that the reaction
rate is dependent on the concentration of CO when the reaction
CO/O2 is catalyzed by Pt/SiO2, but is near-independent of the CO
concentration when catalyzed by Au/SiO2. The negative reaction
order with respect to CO found on Pt/SiO2 is a result of the very
strong bonding of CO with Pt, which inhibits CO oxidation [63–
65]. The Au–CO bond is much weaker than the Pt–CO bond, as al-
ready attested by the CO-FTIR results (Figs. 7 and 8), which results
in the lack of any inhibiting effect on the reaction rate of CO oxida-
tion for Au catalysts. However, in the present case, the reactivity of
Au/SiO2 is very low; this is because gold is poorly efficient at acti-
vating oxygen [30] and that when supported on a reducible sup-
port, such as titania, the addition of water is usually necessary to
activate oxygen [66].
From the CO and O2 reaction orders measured on AuPt(1.8)/SiO2
(Table 4), which exhibits the lower proportion of Pt, it can be con-
cluded that CO oxidation proceeds exclusively on the Au surface
atoms, which is also fully coherent if one compares the reactions
orders to those of Au/SiO2 and with the lack of activity in benzene
hydrogenation of this particular sample (Table 3). The bimetallic
catalyst with a priori more Pt at the surface (AuPt(1.2)/SiO2) be-
haves in a manner which is between that of the monometallic Au
and Pt catalysts (Table 4), with CO oxidation occurring predomi-
nantly on the Pt surface atoms. The less negative order with re-
spect to CO for AuPt(1.2)/SiO2 compared to Pt/SiO2 can be
explained as a reduction in the inhibiting nature of Pt–CO when
Au is present.
3.2.2. CO oxidation
Given the much greater Pt content of Pt/SiO2 compared with the
AuPt/SiO2 samples (Table 2), Fig. 9 shows that Pt/SiO2 is very active
in the reaction of CO oxidation at 280 °C, whereas Au/SiO2 is poorly
active (it is well known that gold nanoparticles supported on silica
hardly catalyze CO oxidation [61,62]). The bimetallic catalysts ex-
hibit an intermediate activity, and AuPt(1.2)/SiO2 is about three
times more active than AuPt(1.8)/SiO2, even though the content
of Pt in AuPt(1.2)/SiO2 is only 40% greater than in AuPt(1.8)/SiO2
(Table 1).
The activation energies of CO oxidation by O2 over each catalyst
are shown in Table 4. Interestingly, it can be seen that the activa-
tion energy of AuPt(1.2)/SiO2 is comparable to that of Pt/SiO2 and
in good agreement with those reported on supported Pt catalysts
[63,64] in a temperature domain close to that studied in the pres-
ent work, whereas that of AuPt(1.8)/SiO2 is comparable to that of
Au/SiO2. The activation energy found on Au/SiO2 in this study is
slightly lower than that reported by Chandler and co-workers [6].
In contrast, these authors reported a much lower activation energy
on Pt/SiO2 than that found in the present work. It should be noted,
however, that Chandler and co-workers determined this activation
energy at much lower temperatures than in the present study and
that Cant et al. [28] showed that the activation energy of the CO
oxidation reaction on Pt/SiO2 had a tendency to decrease with
decreasing reaction temperatures.
CO oxidation turnover rates were 6 ꢁ 10ꢀ4 sꢀ1 for Au/SiO2 and
0.57 sꢀ1 for Pt/SiO2 (Table 5). The value for Pt/SiO2 is in good agree-
ment with that obtained by Cant et al. [28] (0.4–0.7 sꢀ1, extrapo-
lated from the given data to obtain the appropriate value at
280 °C using the same methodology as that described for benzene
hydrogenation turnover rates (please refer to Eq. (2))).
Given the difference in the composition of the bimetallic nano-
particles and the difference in metal loadings and dispersions in
each catalyst (Tables 1 and 5), comparison of the CO2 production
rates of the bimetallic catalysts with that determined experimen-
tally on Pt/SiO2 can be achieved by calculating their expected
CO2 turnover rates by making assumptions about the Pt surface
composition of the bimetallic nanoparticles, as the CO2 turnover
rate on Au sites (6 ꢁ 10ꢀ4 sꢀ1) has been shown to be negligible
with respect to that of Pt (0.57 sꢀ1).
a
7
b
7
6
5
4
3
2
6
5
4
3
2
-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2
-0.6 -0.4 -0.2 0.0 0.2 0.4
ln PO
ln PCO
2
2
Fig. 9. Kinetic Studies at 280 °C of supported nanoparticle catalysts (j Pt/SiO2, s AuPt(1.2)/SiO2, N AuPt(1.8)/SiO2, e Au/SiO2).

112
R.P. Doherty et al. / Journal of Catalysis 287 (2012) 102–113
Table 4
In the case of PtAu alloys, it has been reported that CO would be
more strongly bound to Pt in PtAu alloys than on a pure Pt sample
[9,67,68]. Such a conclusion, with a promotional effect of Au on the
CO oxidation reaction in AuPt(1.2)/SiO2, is however, rather contro-
versial as CO should be less reactive if more strongly chemisorbed
to Pt. It must be stressed, nonetheless, that Pedersen et al. also re-
ported the existence of Pt sites in diluted PtAu alloys for which CO
is less strongly bound than in pure Pt [68]. These particular Pt sites
could be responsible for the synergistic effect observed in AuPt al-
loys. The fact that CO would be less strongly bound to these Pt sites
would be coherent with an increased electron density of these Pt
surface atoms due to charge transfer from Au to Pt, as indicated
by the CO-FTIR results. This increased electron density would re-
sult in a decrease of the strength of the Pt–CO bond, as CO is known
to be a weak base [69], and therefore in a decreased CO coverage
favoring O2 dissociative adsorption followed by reaction with CO
adsorbed on Pt and may be also on Au, i.e., favoring the CO oxida-
tion kinetics. This increased electron density of the Pt sites and the
particle surface enriched by gold would also account for the de-
creased activity of the AuPt alloys for benzene hydrogenation, in
agreement with previously reported studies on the hydrogenation
of aromatics on electron-enriched Pt sites [70,71].
Kinetic parameters (activation energies and reaction orders) of the CO oxidation
reaction at a temperature of 280 °C.
Catalyst
Ea (kJ molꢀ1
)
Reaction order
Least square fit, R2
CO
O2
CO
O2
Pt/SiO2
112
102
55
ꢀ0.88
ꢀ0.62
0.10
1.02
0.89
0.54
0.69
0.999
0.996
0.540
0.985
0.982
1.000
0.999
0.996
AuPt(1.2)/SiO2
AuPt(1.8)/SiO2
Au/SiO2
64
0.09
Table 5 lists the CO2 turnover rates calculated for the bimetallic
samples assuming that the SiO2-supported nanoparticles are
homogeneous in terms of size and shape, and that the surface com-
position of the bimetallic nanoparticles can be deduced from (i) the
metal dispersion and chemical analysis (making, here, the assump-
tion that the nanoparticles are homogeneous alloys in the bulk and
on the surface, Table 1) and (ii) the rate of benzene hydrogenation
and the corresponding turnover rates on Pt and Au (Table 3). Dif-
ferences between the calculated and measured rates can be inter-
preted as either promotional or inhibiting effects due to the
presence of a second metal.
For AuPt(1.2)/SiO2, the calculated CO2 turnover rates are greater
than that determined experimentally on Pt/SiO2 (0.57 sꢀ1, Table 5).
The greatest turnover rate was that calculated using the results of
the benzene hydrogenation. Even the CO2 turnover rate calculated
for a homogeneous alloy (1.50 sꢀ1), which overestimates the
amount of Pt at the catalyst surface according to the chemisorption
and CO-FTIR data, is greater than the turnover rate on Pt/SiO2
(0.57 sꢀ1).
In the presence of CO at the reaction temperature of 280 °C, one
may argue however, that the bimetallic nanoparticles of AuPt(1.2)/
SiO2 have rearranged and exposed a number of Pt surface sites
greater than that corresponding to the composition of the homoge-
neous alloy. Even if one considers the extreme situation for which
all of the bimetallic particles of AuPt(1.2)/SiO2 have rearranged in
such a way that they exposed only Pt atoms at their surface, which
is known however to be unlikely from a thermodynamic point of
view and the XPS and FTIR results, the corresponding CO2 turnover
rates would be estimated to be 0.67 sꢀ1, which is still greater than
the turnover rate determined on Pt/SiO2 (0.57 sꢀ1). This is evidence
that AuPt(1.2)/SiO2 is experiencing a promotional effect due to the
combination of Au and Pt, and this confirms the synergistic effect
between Au and Pt announced in the introduction.
To account for the observed changes in the CO and O2 orders
and the synergistic effect between the two metals for the CO oxi-
dation reaction in AuPt(1.2)/SiO2, it is proposed that an electronic
effect in the AuPt bimetallic nanoparticles with an electron transfer
from Au to Pt lowers the Pt–CO bond energy in agreement with the
CO-FTIR results of the present study, which favors O2 activation on
platinum.
The results for AuPt(1.8)/SiO2 are however not consistent with
this interpretation. For this catalyst, the kinetic orders are close
to those of Au/SiO2 and the calculated CO2 turnover rate based
on a surface composition of the homogeneous alloy is much lower
(0.01 sꢀ1) than the measured turnover rate on Pt/SiO2 (0.57 sꢀ1
,
Table 5). This suggests that there may be a ‘compositional limit’
for the promotional effect of Au toward Pt and that there is a point
at which Pt surface atom dilution becomes too great for this to
occur. This could be confirmed by looking at other AuPt
compositions.
4. Conclusion
Bimetallic core–shell AuPt nanoparticles have been prepared by
radiolysis and deposited onto silica. After stabilizer decomposition,
the resulting catalysts have been shown to contain alloy AuPt
nanoparticles with a surface enriched by gold according to XPS,
indicating particle restructuring. The Pt–CO band observed in
FTIR-CO spectroscopy was found to be screened by the Au–CO
band confirming the close proximity of the two metals at the nano-
particle surface. FTIR-CO spectroscopy gave evidence for further
restructuring of the AuPt particles with the migration of Pt toward
the surface in the presence of CO. However, Au seems to remain
predominant on the surface according to XPS and catalytic results.
Gold in the bimetallic particles has been shown to inhibit the
catalysis of benzene hydrogenation and to promote the catalysis
of CO oxidation by O2 at least for the AuPt(1.2)/SiO2 sample.
Table 5
CO oxidation reaction rates for each catalyst at 280 °C under stoichiometric conditions (1% CO–0.5% O2 in He).
e
Catalyst
D (%)a
Measured CO2 rated
TOR (sꢀ1
)
Experimental
Calculated
Homogeneous alloy
Benzene hydrogenation
162
Pt/SiO2
31.7b
24.9c
25.0c
31.4c
18,000
6900
34
0.57
AuPt(1.2)/SiO2
AuPt(1.8)/SiO2
Au/SiO2
1.50
0.01
7
0.0006
a
b
c
Metal dispersion.
From H2 and CO chemisorption data (Section 3.1.4).
From TEM (Section 3.1.5).
d
e
ꢁ10ꢀ9 mol CO2 sꢀ1 gꢀ1
.
Rate of reaction per Pt surface atom.

R.P. Doherty et al. / Journal of Catalysis 287 (2012) 102–113
[27] R.E. Watson, J.W. Davenport, M. Weinert, Phys. Rev. B 35 (1987) 508.
113
AuPt(1.8)/SiO2 was found to inhibit both reactions, suggesting that
there is a compositional limit to the promotional effect which can
occur in bimetallic AuPt/SiO2 toward CO oxidation. The observed
synergistic effect between the two metals and the changes in the
CO and O2 orders of AuPt(1.2)/SiO2 compared to Pt/SiO2 are attrib-
uted to an electronic effect in the AuPt bimetallic nanoparticles
with an electron transfer from Au to Pt, as indicated by XPS and
CO-FTIR.
[28] N.W. Cant, P.C. Hicks, B.S. Lennon, J. Catal. 54 (1978) 372.
[29] V. Aguilar-Guerrero, B.C. Gates, Catal. Lett. 130 (2009) 108.
[30] B.D. Chandler, C.G. Long, J.D. Gilbertson, C.J. Pursell, G. Vijayaraghavan, K.J.
Stevenson, J. Phys. Chem. C 114 (2010) 11498.
[31] L. Delannoy, N. El Hassan, A. Musi, N.N. Le To, J.M. Krafft, C. Louis, J. Phys.
Chem. B 110 (2006) 22471.
[32] M.C. Daniel, D. Astruc, Chem. Rev. 104 (2004) 293.
[33] S. Link, M.A. El-Sayed, J. Phys. Chem. B 103 (1999) 8410.
[34] M. Mirdamadi-Esfahani, M. Mostafavi, B. Keita, L. Nadjo, P. Kooyman, H.
Remita, Gold Bull. 43 (2010) 49.
[35] G. Bergeret, P. Gallezot, in: G. Ertl, H. Knozinger, J. Weitkamp (Eds.), Handbook
of Heterogeneous Catalysis, vol. 2, Wiley-VCH, 1997, p. 439.
[36] J. Luo, M.M. Maye, V. Petkov, N.N. Kariuki, L.Y. Wang, P. Njoki, D. Mott, Y. Lin,
C.J. Zhong, Chem. Mater. 17 (2005) 3086.
[37] G.C. Bond, Platinum Met. Rev. 51 (2007) 63.
[38] C.X. Xu, R.Y. Wang, M.W. Chen, Y. Zhang, Y. Ding, Phys. Chem. Chem. Phys. 12
(2010) 239.
[39] E. Irissou, F. Laplante, S. Garbarino, M. Chaker, D. Guay, J. Phys. Chem. C 114
(2010) 2192.
[40] Y. Xu, Y. Dong, J. Shi, M. Xu, Z. Zhang, X. Yang, Catal. Commun. 13 (2011) 54.
[41] J. Radnik, C. Mohr, P. Claus, Phys. Chem. Chem. Phys. 5 (2003) 172.
[42] V. Pitchon, A. Fritz, J. Catal. 186 (1999) 64.
Acknowledgments
This work was supported by the Region Ile-de-France in the
framework of C’Nano IdF. C’Nano IdF is the centre of competences
in nanosciences for the Paris region, supported by CNRS, CEA, MESR
and the region Ile-de-France. The authors also thank Dr. J. Blan-
chard and Dr. L. Delannoy for fruitful discussions about chemisorp-
tion on Pt, and CO-FTIR on Au and alloy restructuring, respectively.
[43] H.B. Liu, U. Pal, J.A. Ascencio, J. Phys. Chem. C 112 (2008) 19173.
[44] A. Hugon, L. Delannoy, J.M. Krafft, C. Louis, J. Phys. Chem. C 114 (2010) 10823.
[45] F. Gao, Y.L. Wang, D.W. Goodman, J. Phys. Chem. C 113 (2009) 14993.
[46] F. Gao, Y.L. Wang, D.W. Goodman, J. Am. Chem. Soc. 131 (2009) 5734.
[47] S.M. Oxford, P.L. Lee, P.J. Chupas, K.W. Chapman, M.C. Kung, H.H. Kung, J. Phys.
Chem. C 114 (2010) 17085.
References
[1] J.M. Thomas, R. Raja, B.F.G. Johnson, S. Hermans, M.D. Jones, T. Khimyak, Ind.
Eng. Chem. Res. 42 (2003) 1563.
[2] L.M. Bronstein, D.M. Chernyshov, I.O. Volkov, M.G. Ezernitskaya, P.M. Valetsky,
V.G. Matveeva, E.M. Sulman, J. Catal. 196 (2000) 302.
[48] P. Hollins, Surf. Sci. Rep. 16 (1992) 51.
[49] P.T. Fanson, W.N. Delgass, J. Lauterbach, J. Catal. 204 (2001) 35.
[50] M. Primet, J. Catal. 88 (1984) 273.
[51] M.A. Bollinger, M.A. Vannice, Appl. Catal. B: Environ. 8 (1996) 417.
[52] J.D. Grunwaldt, M. Maciejewski, O.S. Becker, P. Fabrizioli, A. Baiker, J. Catal. 186
(1999) 458.
[3] N. Toshima, M. Harada, Y. Yamazaki, K. Asakura, J. Phys. Chem. 96 (1992) 9927.
[4] M. Harada, K. Asakura, Y. Ueki, N. Toshima, J. Phys. Chem. 96 (1992) 9730.
[5] O.S. Alexeev, B.C. Gates, Ind. Eng. Chem. Res. 42 (2002) 1571.
[6] B.J. Auten, H.F. Lang, B.D. Chandler, Appl. Catal. B: Environ. 81 (2008) 225.
[7] J.H. Zeng, J. Yang, J.Y. Lee, W.J. Zhou, J. Phys. Chem. B 110 (2006) 24606.
[8] A. Habrioux, E. Sibert, K. Servat, W. Vogel, K.B. Kokoh, N. Alonso-Vante, J. Phys.
Chem. B 111 (2007) 10329.
[53] S. Minico, S. Scire, C. Crisafulli, A.M. Visco, S. Galvagno, Catal. Lett. 47 (1997)
273.
[54] D.J.C. Yates, J. Colloid Interface Sci. 224 (1969) 36.
[55] R. Meyer, C. Lemire, Sh.K. Shaikhutdinov, H.-J. Freund, Gold Bull. 37 (2004) 72.
[56] H.K.M. Mihaylov, K. Hadjiivanov, B.C. Gates, Chemie Ingenieur Technik 79
(2007) 795.
[57] F. Boccuzzi, A. Chiorino, M. Manzoli, D. Andreeva, T. Tabakova, J. Catal. 188
(1999) 176.
[9] H.G. Lang, S. Maldonado, K.J. Stevenson, B.D. Chandler, J. Am. Chem. Soc. 126
(2004) 12949.
[10] X.H. Peng, Q.M. Pan, G.L. Rempel, Chem. Soc. Rev. 37 (2008) 1619.
[11] E. Bus, J.A. van Bokhoven, J. Phys. Chem. C 111 (2007) 9761.
[12] C. Mihut, C. Descorme, D. Duprez, M.D. Amiridis, J. Catal. 212 (2002) 125.
[13] H. Remita, I. Lampre, M. Mostafavi, E. Balanzat, S. Bouffard, Radiat. Phys. Chem.
72 (2005) 575.
[58] J.Y. Lee, J. Schwank, J. Catal. 102 (1986) 207.
[59] L. Salin, C. Potvin, J.F. Tempere, M. Boudart, G. Djega-Mariadassou, J.M. Bart,
Ind. Eng. Chem. Res. 37 (1998) 4531.
[14] F. Ksar, L. Ramos, B. Keita, L. Nadjo, P. Beaunier, H. Remita, Chem. Mater. 21
(2009) 3677.
[60] S. Puddu, V. Ponec, Rec. Trav. Chim. 95 (1976) 255.
[61] M. Haruta, S. Tsubota, T. Kobayashi, H. Kageyama, M.J. Genet, B. Delmon, J.
Catal. 144 (1993) 175.
[15] J. Belloni, H. Remita, in: M. Spotheim-Maurizot, M. Mostafavi, T. Douki, J.
Belloni, Radiation Chemistry: From Basics to Applications in Material and Life
Sciences, EDP Sciences, 2008, p. 97.
[62] M. Haruta, H. Kageyama, N. Kamijo, T. Kobayashi, F. Delannay, T. Inui, Stud.
Surf. Sci. Catal., Elsevier (33).
[16] J. Belloni, M. Mostafavi, H. Remita, J.L. Marignier, M.O. Delcourt, New J. Chem.
22 (1998) 1239.
[63] Y.F. Yu Yao, J. Catal. 87 (1984) 152.
[17] J. Belloni, Catal. Today 113 (2006) 141.
[64] R.H. Nibbelke, M.A.J. Campman, J.H.B.J. Hoebink, G.B. Marin, J. Catal. 171
(1997) 358.
[65] I. Manuel, K. Chaubet, C. Thomas, H. Colas, N. Matthess, G. Djega-Mariadassou,
J. Catal. 224 (2004) 269.
[18] W. Abidi, H. Remita, Recent Patents Eng. 4 (2010) 170.
[19] R. Bouwman, W.M. Sachtler, J. Catal. 19 (1970) 127.
[20] A.Y. Stakheev, L.M. Kustov, Appl. Catal. A 5 (1983) 85.
[21] S.J. Tauster, Acc. Chem. Res. 20 (1987) 389.
[66] G.C. Bond, C. Louis, D. Thompson, Catalysis by Gold, Imperial College Press,
2006.
[67] X.B. Ge, X.L. Yan, R.Y. Wang, F. Tian, Y. Ding, J. Phys. Chem. C 113 (2009) 7379.
[68] M.O. Pedersen, S. Helveg, A. Ruban, I. Stensgaard, E. Laegsgaard, J.K. Norskov, F.
Besenbacher, Surf. Sci. 426 (1999) 395.
[22] J.D. Bracey, R. Burch, J. Catal. 86 (1984) 384.
[23] P. Hernandez-Fernandez, S. Rojas, P. Ocon, J.L.G. de la Fuente, J.S. Fabian, J.
Sanza, M.A. Pena, F.J. Garcia-Garcia, P. Terreros, J.L.G. Fierro, J. Phys. Chem. C
111 (2007) 2913.
[24] A. Sachdev, J. Schwank, J. Catal. 120 (1989) 353.
[69] K.I. Hadjiivanov, G.N. Vayssilov, Adv. Catal. 47 (2002) 307.
[70] A. de Mallmann, D. Barthomeuf, J. Chim. Phys. 87 (1990) 535.
[71] D.E. Ramaker, J. de Graaf, J.A.R. van Veen, D.C. Koningsberger, J. Catal. 203
(2001) 7.
[25] H. Tada, F. Suzuki, S. Ito, T. Akita, K. Tanaka, T. Kawahara, H. Kobayashi, J. Phys.
Chem. B 106 (2002) 8714.
[26] M.F. Luo, C.C. Wang, G.R. Hu, W.R. Lin, C.Y. Ho, Y.C. Lin, Y.J. Hsu, J. Phys. Chem. C
113 (2009) 21054.