Journal of Catalysis 392 (2020) 254–265

Contents lists available at ScienceDirect

Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat

Ceria nanorod supported gold nanoparticles as structured catalysts for

the oxidative steam reforming of methanol: Effect of CTAB concentration
on physiochemical properties and catalyst performance
Srisin Eaimsumang a, Nuwong Chollacoop b, Apanee Luengnaruemitchai a,c,⇑, Stuart H. Taylor d
a
The Petroleum and Petrochemical College, Chulalongkorn University, Pathumwan, Bangkok 10330, Thailand
b
Renewable Energy Laboratory, National Metal and Materials Technology Center (MTEC), 114 Thailand Science Park, Klong Luang, Pathumthani 12120, Thailand
c
Center of Excellence in Catalysis for Bioenergy and Renewable Chemicals (CBRC), Chulalongkorn University, Bangkok 10330, Thailand
d
Cardiff Catalysis Institute, School of Chemistry, Cardiff University, Main Building, Park Place, Cardiff CF10 3AT, UK

a r t i c l e i n f o a b s t r a c t

Article history: Ceria (CeO2) supports were synthesized by hydrothermal treatment with different Cetyltrimethy-
Received 19 May 2020 lammonium bromide (CTAB) concentrations (CeO2-X: 0.135, 0.270, 0.540 and 1.080, where X is the
Revised 21 September 2020 CTAB concentration in mmol). The micro-structures of bundle-like ceria nanorods (NR) were produced
Accepted 17 October 2020
by thermal treatment of the crystalline Ce(OH)CO3 precursor. The exposed crystal planes of (2 0 0) and
Available online 2 November 2020
(1 1 1), corresponding to the d-spacing of 0.27 and 0.31 nm, were confirmed by high resolution-
transmission electron microscopy (HR-TEM). The fundamental characteristics of CeO2-0.135 led to the
Keywords:
highest surface area, surface area normalized reducibility and oxidized gold species (Aun+/Au0) on the
Methanol
Hydrogen
surface, relatively higher oxygen vacancies and a smaller average gold particle size (7.9 nm). The activity
OSRM results show that the Au/CeO2-0.135 has the highest methanol conversion and H2 production rate for
CTAB oxidative steam reforming of methanol (OSRM). The reducibility and gold particle size played a crucial
Gold role in determining the H2/CO2 ratio at the temperature of 200 °C.
Ceria nanorod Ó 2020 Elsevier Inc. All rights reserved.

1. Introduction reaction at 200–400 °C [9]. The CO2 selectivity increase as a result

of significant taking place of POM reaction. Furthermore, OSRM has
Gold nanoparticles supported on CeO2 have been regarded as an the advantage of being able to supply energy, as it is an exothermic
effective catalyst for steam reforming of methanol (SRM) [1], oxidation reaction. However, the product distribution (e.g. H2, CO
water–gas shift (WGS)[2–4] and many oxidation reactions [5–8] and CO2) depends on the feed composition, as shown in Eq (1):
in terms of providing high conversion at lower reaction tempera-
CH3 OH+(1-a)H2 O + 0.5aO2 ! CO2 +(3-a)H2 (1)
ture compared to other catalysts. One of the applications for this
catalyst is steam reforming of methanol for H2 production, due to
the synergistic function of gold on an active support like CeO2. DH0298 ¼ 192:2  49:5kJ=mol; ð0 < a < 1Þ
SRM can provide a high H2 yield, and it can also produce higher
CO formation compared to partial oxidation of methanol (POM). Au supported on ceria nanorods (NR) has provided the best cat-
Moreover, it requires a significant supply of energy, due to it being alytic activity for many reactions, such as preferential oxidation of
an endothermic reaction. Therefore, oxidative steam reforming of CO (CO-PROX) [7], WGS [2], SRM [1], and OSRM [10]. This is
methanol (OSRM) has been introduced as a selective reaction, because of the high surface area to volume ratio and exposure of
but it does have the compromise of being able to produce high the active (1 1 0) and (1 0 0) planes, leading to the highest reducibil-
yields of H2, while producing a lower yield of CO when the oxida- ity and oxygen vacancies. In addition, ceria NR influence the char-
tion reaction predominated. Pojanavaraphan et al. investigated the acteristics of gold nanoparticles, for example, gold particle size,
effect of O2 content on the activity of Au/CeO2 catalysts in OSRM oxidation state and dispersion, including interaction between
metal and support. These properties promote superior catalytic
activity when compared to other ceria nanoshapes. Ceria has been-
⇑ Corresponding author at: The Petroleum and Petrochemical College, Chula- classified as an active support because of its oxygen mobility, as
longkorn University, Pathumwan, Bangkok 10330, Thailand. oxygen can be exchanged from the structure depending on tem-
E-mail address: [email protected] (A. Luengnaruemitchai). perature, resulting in the inverted conversion between Ce4+ and

https://doi.org/10.1016/j.jcat.2020.10.023
0021-9517/Ó 2020 Elsevier Inc. All rights reserved.
S. Eaimsumang, N. Chollacoop, A. Luengnaruemitchai et al. Journal of Catalysis 392 (2020) 254–265

Ce3+, consequently generating oxygen vacancies. This property 2.2. Catalyst characterization
strongly relates to various factors, namely, crystallite size, mor-
phology and exposed crystal plane. Typically, for an fcc structure, The surface area of catalysts was determined by N2-
ceria surfaces consist of three main low-index crystal planes: physisorption at 196 °C, using a Quantachrome Autosorb-1 sur-
(1 1 1), (1 1 0) and (1 0 0), which have different surface energies. face area analyzer. The specific surface area of catalysts was deter-
The active surfaces have the order: (1 0 0) > (1 1 0) > (1 1 1), mined by the Brunauer-Emmett-Teller (BET) equation. Prior to the
regarded as the surface free energy of 2.05, 1.26 and 0.69 eV, measurements, all samples were degassed under vacuum circum-
respectively [11]. However, the energy for formation of oxygen stance at 250 °C for 16 h.
vacancies can be ordered as (1 1 0) < (1 0 0) < (1 1 1) with respect X-ray diffraction (XRD) was used to identify the crystalline
to 1.99, 2.27 and 2.6 eV, respectively [12]. structure of CeO2-X supports and Au/CeO2-X using a Rigaku X-
Typically, single crystalline ceria nanorods (NR) are obtained ray diffractometer system (RINT-2200) equipped with Cu Ka radi-
randomly, formed by particle agglomeration, which is the case ation, operated at an accelerating voltage of 40 kV and 30 mA cur-
without assisted synthesis using a template. However, bundle- rent. The analysis was continuously scanned with a speed of 2h
like structures of ceria NR can be designed by thermal conversion equal to 5°/min and a scan step of 0.02(2h), 2h range was from
of a single crystal of Ce(OH)CO3 [13]. This structure can yield many 20 to 80°. Crystallite size (D) calculations of the Au(1 1 1) and
advantages, because it is composed of a multilayer structure and CeO2(1 1 1) were done using the Scherrer equation: D=(0.94  k)/
active surface, (1 0 0), and has a high aspect ratio, and high surface (b  cosh), where k is the wavelength of the Cu Ka (0.15406 nm),
area to volume ratio. To control size and morphological structure of b is the full width at half the maximum (FWHM) of the X-ray
ceria NR, a hydrothermal synthesis method is a simple way to diffraction peak, and h is the Bragg angle at corresponding peak
obtain the desired morphology of the nanomaterial. Nevertheless, (deg).
particle agglomeration and large size distribution can result under The morphological structure of the CeO2-X supports was inves-
the high temperature and pressure used for hydrothermal synthe- tigated by field emission gun scanning electron microscopy (FEG-
sis, leading to a lower surface area. To solve this problem, the addi- SEM) using a Tescan MAIA3 microscope. The samples were coated
tion of organic molecules during synthesis can prevent particle by Au:Pt at a ratio of 80:20 (%v) with a thickness of 10 nm.
agglomeration, by means of stabilizing adsorption on particular High resolution transmission electron microscopy (HR-TEM)
surfaces, and through beneficial steric hindrance it can direct crys- was utilized for identifying the morphologies and exposed crystal
tal growth. Cetyltrimethylammonium bromide (CTAB) is one of the planes of the catalysts. The analysis was performed by using a JEOL
surfactants that is commonly used for catalyst synthesis. However, JEM-2100 instrument operated at a 200 kV.
the concentration during catalyst synthesis should be optimized to H2-temperature-programmed reduction (TPR) analysis was
obtain the best ceria properties for catalysis. Previously, the effect used to measure the reducibility of the catalysts. Approximately
of CTAB to ceria precursor ratio on the textural properties of ceria 0.1 g of sample was contained in a quartz tube reactor and ana-
has been studied, but the role of using surfactant during synthesis lyzed using a TPDRO 1100 instrument (Thermo Finnigan) equipped
was still not fully explored [14]. with a thermal conductivity detector (TCD) The measurement was
To clarify the effects of morphology on the catalytic perfor- performed with 5.13% H2 in Ar at 30 mL/min. The reduction tem-
mance of ceria NR, we synthesized ceria NR using CTAB as a tem- perature was raised from 30 to 900 °C with a heating rate of
plate, and studied the role of varying CTAB surfactant 10 °C/min.
concentrations on ceria morphology and physiochemical charac- X-ray photoelectron spectroscopy (XPS) was utilized to analyze
teristics. The range of ceria NR synthesized was used as a support the elemental composition on the surface and oxidation state of
to prepare 3 wt% Au/CeO2 catalysts. All catalysts were character- material using a Kratos Axis Ultra DLD photoelectron spectrometer
ized by various techniques and the catalytic performance was eval- equipped with a monochromatic Al Ka radiation operated at 15 kV
uated for OSRM performance. under a pressure of less than 5  10-7 Torr. All XPS spectra were
calibrated by the C 1 s (284.6 eV). The obtained XP spectra was
2. Experimental analyzed by the vision processing software supplied with the
instrument. The Shirley background was applied to normalize
2.1. Catalyst preparation peaks of the XP spectra. The peak position of Au varied
about ± 0.3 eV and maintained the FWHM. The spin orbit splitting
The desired amount of CTAB (Aldrich, 96%), 2.3 mmol of Ce of Au4f was at about 3.7 eV. The peak of O1s and Ce3d were fitted
(NO3)36H2O and 9.2 mmol of urea were dissolved in 10 mL of by a Lorentzian Asymmetric Lineshape with a varied FWHM due to
deionized water at room temperature under agitation for 15 min. the asymmetric peak.
The solution was then put into a 125-mL Teflon-lined stainless- Raman spectroscopy was used to observe the vibrational prop-
steel autoclave and maintained at 120C for 12 h. After cooling, erties of catalysts using a Renishaw InVia Raman spectrometer, at a
the white slurry was collected by filtration, washed thoroughly power of 10 mW and a laser excitation source of wavelength
with excess deionized water to eliminate residual ions and dried 514 nm with a spectral resolution of 1 cm1.
in an oven at 80 °C for 24 h. Finally, the white powder was calcined The quantification of actual gold loading for each catalyst was
in static air at 500 °C for 10 h. The CeO2 support was denoted as measured by an inductively coupled plasma-optical emission spec-
CeO2-X, where X is the CTAB concentration in mmol. troscopy (ICP-OES) technique on a Perkin Elmer Optima 4300 D,
The Au/CeO2-X (3 wt%Au) catalysts were prepared by the using Ar/N2 plasma. Approximately 7.5 mg of Au/CeO2-X samples
deposition–precipitation method. Accordingly, the CeO2-X powder were digested by aqua regia solution (HCl 37% (Aldrich)/HNO3
was added in an aqueous solution of 2 mmol/L hydrogen tetra- 67% (Aldrich) : 3/1) at 70 °C under agitation for 20 min. The solu-
chloroauric (III) acid (HAuCl43H2O; 99.99%, Alfa-Aesar) under con- tion was filtered to eliminate residual particles and the final vol-
tinuous stirring at room temperature for 1 h. Meanwhile, 0.1 mol/L ume was subsequently adjusted to 100 mL with deionized water.
of (NH4)2CO3 was added dropwise to control pH of the solution ~ 8.5
[10]. The precipitate was then washed by deionized water until pH 2.3. Catalytic activity measurement
7 was attained and then collected by filtration. It was finally dried
under vacuum at 80 °C for 16 h and calcined in static air at 400 °C Approximately 100 mg of catalyst was packed into the 6 mm
for 4 h. inner diameter quartz reactor tube, which was performed under
255
S. Eaimsumang, N. Chollacoop, A. Luengnaruemitchai et al. Journal of Catalysis 392 (2020) 254–265

atmospheric pressure over a reaction temperature of 100 and the 2-Theta of about 38.2°for all catalysts [15]. Subsequently, gold
range 200–400 °C, with a constant 0.6/2/1 M ratio of oxygen/wa- crystallite sizes were also calculated using the Sherrer equation.
ter/methanol (MeOH). A constant flow rate of distilled water and They were in the range of 8.3–11.9 nm.
MeOH mixture was controlled by a syringe pump at 1.5 mL/h,
and vaporized by passing through a pre-heater, accompanied by 3.1.2. Textural properties
He (45 mL/min) as a carrier gas, and then mixed with the O2 N2 adsorption–desorption isotherms for the synthesized ceria
(5 mL/min) stream before entering the reaction zone. A gas hourly (Fig. SI1) represent those typical of porous materials. They exhib-
space velocity (GHSV) was preserved at 30000 mL/gcat∙h. The gas- ited type II Isotherms according to the IUPAC system, where mono-
eous product composition (e.g. H2, CO2, CH4 and CO) was analyzed layer adsorption predominantly takes place at a low relative
by gas chromatography (GC) using an Agilent 6890 N GC system pressure, while multilayer adsorption occurs at higher relative
equipped with a TCD and a carbosphere column. The condensates pressure. In addition, the capillary condensation occurred at P/
(e.g. MeOH) were identified by GC using an Agilent 7890A GC sys- P0 = 0.4, which belonged to type H4, representing agglomeration
tem equipped with a flame ionization detector and a DB-wax col- of particles forming slit shaped pores [16]. The textural properties
umn [10]. The methanol conversion (X) and product selectivities of all catalysts were determined. Ceria prepared without CTAB had
were obtained from Eqs. 2–3: a lower BET surface area (56 m2/g), as shown in Table 1, than those
with CTAB added during synthesis, and this is in agreement with
C in out
MeOH  C MeOH previous studies [14]. It is clear that addition of CTAB reduced par-
X¼  100ð%Þ ð2Þ
C in
MeOH ticle agglomeration, as CTAB is adsorbed on the cerium oxide sur-
face by electrostatic interaction between alkylammonium ion and
X
Si ¼ ni = ni  100ð%Þ ð3Þ nucleophilic oxide of the ceria [17]. The BET surface area of parent
ceria increased with increasing CTAB concentration and
where, i represented the gas product, namely, H2, CO, CO2 and CH4 approached the highest value at 90 m2/g for CeO2-0.270, and then
and n denoted as the rate of gas produced in mole. subsequently decreased 1.2-fold for CeO2-0.540. The high excess of
the CTAB molecule adsorbing on the ceria surface might result in
the decrease of dissolution and recrystallization rate. The surface
3. Results and discussion
area of the final catalysts was slightly reduced, by about 8–15%,
after gold deposition.
3.1. Characterization
3.1.3. Ceria morphology (SEM and HR-TEM analysis)
3.1.1. XRD analysis
The morphological structure of the ceria was examined by SEM.
All the calcined ceria catalyst structures were ascribed to the
The results showed that they exhibited a flower-like morphology,
fluorite cubic CeO2 structure (ICDD card number 01-078-5328),
which was composed from a bunch of nanorods. This morphology
due to corresponding diffraction peaks of (1 1 1), (2 0 0), (2 2 0),
was also synthesized by a Ce(OH)CO3 route in the presence of CTAB
(3 1 1), (2 2 2), (4 0 0), (3 3 1) and (4 2 0) planes, respectively, as pre-
at 100 °C for 24 h, as previously reported [13]. To investigate the
sented in Fig. 1. The crystallite size of the parent ceria, as deter-
effect of CTAB on inducing ceria morphology, SEM images of ceria
mined by X-ray line broadening, did not alter with the variation
prepared without CTAB are presented in Fig. 2e. The morphology
of the CTAB concentration, as it was in the range of 11.6–
shows some similarity, with many groups of nanorods adhering
12.4 nm. Therefore, CTAB concentration did not significantly affect
together, but forming more dense-particle agglomerates with a
the crystal growth phenomena of ceria. The crystal size of ceria in
more non-uniform direction of rods. Comparing ceria morpholo-
the presence of Au was slightly increased, probably due to the sec-
gies prepared in the presence of CTAB, the dimension of the
ond heat treatment from calcination. All ceria crystallite sizes were
micro-flower structures was relatively larger for higher concentra-
in the range of 12.0–13.4 nm. In addition, the corresponding
tions of CTAB when compared to lower CTAB concentrations. The
diffraction peak of Au(1 1 1), indicative of Au0, can be detected at
capability of CTAB to act as a stabilizer restricts the growth of ceria
particles, because it reduces the surface tension of ceria particles,
leading to a reduction of dissolution/recrystallization during
Au(111)
Au/CeO2-0.135 hydrothermal synthesis. It was evident that CTAB did not have a
Intensity

Au/CeO2-0.270 significant role in altering ceria morphology, but it did improve

Au/CeO2-0.540 some physical properties. When surfactant concentration was less
(111)

37.0 37.5 38.0 38.5 39.0 39.5 40.0

2
Au/CeO2-1.080 than 10-5 M, electrostatic interaction between metal ions and
organic ions is rather weak. Hence, this is the reason it does not
Au(111)

(220)
(200)

(311)
Intensity (a.u.)

directly manipulate morphologies [18].

(331)
(420)
(222)

(400)

The crystallographic planes of Au/CeO2 catalysts were indexed

from the HR-TEM images (Fig. 3). Principally, the rod-like ceria
mainly consisted of the (1 1 1) and (2 0 0) indexed planes, identified
from the measured d-spacings of 0.31 nm and 0.27 nm, respec-
tively. The tip of the rods exposed the most stable crystal plane
of (1 1 1), visible from Fig. 3g, while the crystal planes of (2 0 0)
were mostly perpendicular to the (1 1 1) planes on the stem of
the rods. Furthermore, the existence of crystalline gold nanoparti-
cles was found for Au/CeO2-0.135, where the crystal planes of Au
(1 1 1) and Au (2 0 0), corresponding to the d-spacings of 0.235
and 0.200 nm, respectively, appeared on the higher surface energy
20 30 40 50 60 70 80
(2 0 0) crystal plane of ceria, as shown in Fig. 3a. Likewise, the crys-
2 tal plane of Au (1 1 1) and Au (2 0 0) were noted for Au/CeO2-0.270.
Only Au (1 1 1) was observed, located on the ceria (1 1 1) crystal
Fig. 1. Representative powder XRD patterns of the Au/CeO2 catalysts. plane, in the case of Au/CeO2-0.540 and Au/CeO2-1.080. For the
256
S. Eaimsumang, N. Chollacoop, A. Luengnaruemitchai et al. Journal of Catalysis 392 (2020) 254–265

Table 1
Physicochemical properties of the CeO2-X supports and Au/CeO2-X catalysts.
a b c
Sample- CTAB (mmol) D(1 1 1) (nm) DAu(1 1 1) (nm) DAu SBET (m2/g) d
Au loding (wt%)
CeO2 w/o CTAB 12.3 N/A N/A 56 N/A
CeO2-0.135 11.6 N/A N/A 89 N/A
CeO2-0.270 12.4 N/A N/A 90 N/A
CeO2-0.540 12.0 N/A N/A 75 N/A
CeO2-1.080 11.8 N/A N/A 78 N/A
Au/CeO2-0.135 12.1 8.8 7.9 81 2.38
Au/CeO2-0.270 13.4 8.3 7.8 76 2.45
Au/CeO2-0.540 12.0 9.4 8.7 82 2.35
Au/CeO2-1.080 12.1 11.9 13.1 66 2.29
a
Crystallite size of ceria determined by Sherrer’s equation from (1 1 1) diffraction peak of ceria.
b
Au crystallite size calculated by Sherrer’s equation from Au (1 1 1).
c
Average gold crystallite size measured by counting 50 particles on HR-TEM images by means of particle size distribution.
d
Actual gold loading determined by ICP-EOS.

FCC structure, the low-index CeO2 crystal plane consisted of (1 1 1), rate and extent of nucleation, consequently leading to a larger par-
(1 0 0) and (1 1 0), which (1 1 0) and (1 0 0) had higher surface ticle size.
energy than (1 1 1). The different surface energy of those planes
resulted from the different exposed atomic arrangement. 3.1.4. Reducibility of catalysts
CeO2(1 1 1) surface had repeating unit of atom on 3 planes in a The reducibility of ceria was determined by H2-TPR. The reduc-
symmetric axis position leading to no net dipole moment perpen- tion of active oxygen from the ceria structure at the surface and the
dicular on the surface. Whereas, (1 1 0) and (1 0 0) had atomic con- bulk, represented by the reduction features at temperatures of
figuration only 2 planes generating dipole moment perpendicular about 520 °C and 850 °C, respectively, as illustrated in Fig. SI2.
on those surfaces [11]. It was postulated that different CTAB con- The reduction temperatures of all the ceria samples were similar,
centrations during synthesis influenced the location of gold depo- however, slightly different H2-uptakes (0.42–0.50 mmol/g) were
sition and the exposed Au crystalline phases on the specific ceria evident (Table 2). These results correlated with the similar physical
crystal plane. In addition to investigating the crystallographic properties, such as crystallite size and BET surface area. Interest-
planes of the catalyst, the actual gold particles sizes were esti- ingly, the surface reduction temperature was decreased to much
mated from particles size distribution data obtained from HR- lower values (145–154 °C) after gold deposition, which resulted
TEM images. It seems that a larger gold particle size was present from weakening of the Ce-O bond in the presence of gold [23].
on ceria catalysts synthesized with higher CTAB concentrations, The magnitude of H2 uptake represents improved redox properties
following the order: Au/CeO2-0.270 < Au/CeO2-0.135 < Au/CeO2- for activating molecular oxygen and the interaction between metal
0.540 < Au/CeO2-1.080. The summary of mean gold particle size and support. The H2 consumption significantly increased by 1.2–
is shown in Table 1, gold particle sizes ranged from 7.8 to 13.1 nm. 1.7-fold for all Au/CeO2-X catalysts when compared to the parent
Typically, the crystallization under hydrothermal treatment ceria, except for Au/CeO2-0.540. Au/CeO2-0.135 exhibited the high-
consisted of a nucleation and crystal growth step, via an Ostwald est H2 uptake (0.82 mmol H2/g), which was 1.7 times greater than
ripening process [19]. When CTAB concentration is lower than for pure ceria.
the CMC value, it does not form a micelle in the aqueous solution.
Hence, below the CMC a single CTAB molecule might adsorb on a 3.1.5. Raman analysis
non-specific surface of ceria and prevent agglomeration by steric The symmetrical vibration of lattice oxygen of the prepared cat-
hindrance of the tail group. The growth habit would still follow alysts was investigated by Raman. The Raman spectra of both ceria
the Ostwald ripening process, in which small crystals are thermo- and Au/CeO2-X catalysts are given in Fig. 4. The small peak at
dynamically unstable and dissolve in the solution and then recrys- 259 cm1 was reported as a second-order transverse acoustic mode
tallize on the larger particles. Nevertheless, the higher of CeO2 [24]. A strong F2g band at about 464 cm1 was assigned to
concentration of surfactant adsorbed on the surface retarded the the symmetric stretching vibration of linked Ce-O bonds, whilst
Ostwald ripening process, due to the fact that the particles were the band around 600 cm1 was due to defect sites or intrinsic oxy-
more stable and difficult to dissolve [20]. Therefore, when the par- gen vacancies in the ceria structure [25]. The weak band positioned
ticles cannot dissolve and then recrystallize, the particles tend to at about 1170 cm1 corresponded to the second-order longitudinal
gather together as a bulky group in order to be more stable ther- optical mode of defects [26]. For ceria catalysts containing gold, the
modynamically. As a result, larger particle sizes and lower surface F2g band located at 464 cm1 became broader, measured by an
areas were observed. increase of the FWHM for all gold catalysts, compared to that of
The crystallite size mainly relates to the nucleation rate, which the parent ceria. Moreover, the bands at around 600 cm1 were
involved the diffusion of ions to the growing crystal surface and hardly visible (Fig. 4b), they became broader and lower in intensity,
incorporation of the nucleus into the structure of the crystal lattice probably because the addition of Au caused a more asymmetric
[21]. The nucleation process of Ce(OH)CO3 was started by reaction vibrational band, due to the weakening of the Ce-O bond. Moreover,
between hydrolyzed Ce3+ ion with water to form the complex [Ce the weak bands at 259 cm1 gradually shifted toward lower
(OH)(H2O)n-1]2+ and carbonate ion (CO2 3 ), decomposed from urea wavenumbers in the case of Au/CeO2-X prepared with lower con-
at the temperature used [22]. Fundamentally, in the aqueous solu- centrations of CTAB. The vibrational bands at about 830–837 cm1
tion the diffusion rate of ions is really fast due to very low solution appeared in the gold-containing catalyst series, corresponding to
viscosity [21]. When the amount of CTAB increased in the solution, the presence of peroxide species (O2
2 ) generated on CeO2-x (1 1 1)
it might suppress the hydrolyzation of Ce3+ ion with a water mole- [27]. Au deposition influenced the properties of the ceria. Compar-
cule to form [Ce(OH)(H2O)n-1]2+, and reduce formation in the aque- ing the concentration of ceria intrinsic oxygen vacancy sites, by
ous medium. Therefore, the higher volume of surfactant might means of comparing the peak area ratios (A600 + A1170/A464),
reduce the diffusion rate of ions, resulting in a slower nucleation between the defect band and F2g band. The results are presented
257
S. Eaimsumang, N. Chollacoop, A. Luengnaruemitchai et al. Journal of Catalysis 392 (2020) 254–265

Fig. 2. Representative SEM images of ceria synthesized at various CTAB concentrations at magnification of 30 k of a) CeO2-0.135, b) CeO2-0.270, c) CeO2-0.540, d) CeO2-1.08
and at 5.0 k of e) CeO2-0.000.

in Fig. 5b. There were no significant differences for CeO2-X catalysts 3.1.6. ICP and XPS analysis
that did not contain gold. However, the concentration of oxygen The actual gold loading was measured by ICP-EOS. The results
vacancy sites was considerably enhanced in the case of Au/CeO2- show that the actual gold loading was slightly less than the
0.135 and Au/CeO2-0.270. Gold deposited onto ceria synthesized nominal gold loading (3 wt%,), and measured gold content was in
with higher CTAB concentrations only had slightly increased num- the range 2.29–2.45 wt%. Au/CeO2-0.270 had the highest gold content
bers of oxygen vacancy sites, about a 6.4–9.6% increase. An increase of 2.45 wt%. Additionally, the surface gold concentration was also
defect concentration resulted from the perturbation of the ceria lat- investigated by XPS. Gold was found on the surface at 0.8–1.3 at.%.
tice oxygen by gold deposition and consequently developed oxygen Au/CeO2-0.135 had the highest amount of gold on the surface.
vacancies in the ceria supports [28,29]. It can be briefly described that the high surface area of the

258
S. Eaimsumang, N. Chollacoop, A. Luengnaruemitchai et al. Journal of Catalysis 392 (2020) 254–265

Fig. 3. Representative HR-TEM images to identify crystal-planes and morphologies of (a,b) Au/CeO2 0.135, (c,d) Au/CeO2-0.270, (e,f) Au/CeO2-0.540 and (g,h) Au/CeO2-1.080
catalysts.

Table 2 at BEs of 83.8 and 87.5 eV, representing metallic Au (Au0).

Summary of catalyst reducibility from Temperature Programmed Reduction studies. Whereas, the latter two peaks at BEs of 84.9 and 88.7 eV, were
Sample- CTAB Peak maximum H2 uptake H2 uptake/SBET assigned to Au+. Au3+ was expected at a higher BE of 86.3 and
(mM) (C)a (mmol/g) (mmol/m2) 89.9 eV [15]. Moreover, the ratio of Aun+/Au0 was also calculated
CeO2-0.135 519 0.48 0.0054 by means of the ratio of appropriate peak areas, as concluded in
CeO2-0.270 519 0.50 0.0056 Table 3. There was a tendency for surface Aun+/Au0 to decrease
CeO2-0.540 520 0.49 0.0065 as the concentration of CTAB in the ceria synthesis procedure
CeO2-1.080 519 0.42 0.0053 increased. Au/CeO2-0.135 had the highest Aun+/Au0 ratio of 0.43.
Au/CeO2-0.135 148 0.82 0.0101
Au/CeO2-0.270 146 0.62 0.0082
The presence of gold oxide indicated the cationic gold locating at
Au/CeO2-0.540 145 0.28 0.0035 the interface between Au and support with chemical bonding
Au/CeO2-1.080 154 0.53 0.0080 Au-O-Ce, where the metallic Au was above this layer as a model,
a
Peak maximum for ceria surface reduction.
was reported in [30], and in this study we consider the same is pre-
sent in the catalysts.
Surface oxygen species were also differentiated by analysis of
the XP spectra of the O1s signal (Fig. 7). Peak fitting identified 3
CeO2-0.135 contributed to an increased nucleation site for gold surface oxygen species classified as, lattice surface oxygen of ceria
deposition. (529.0 eV), whereas peaks existing at 531.0 and 533.0 eV were
The electronic states of elements on the catalyst surface were ascribed to a defect site, and an adsorbed oxygen species and/or
also examined by XPS. The oxidation state of cerium, present as a hydroxyl group, respectively [31]. The results found that the ratio
Ce3+ and Ce4+, was differentiated by peak fitting of the Ce 3d spec- of surface lattice oxygen tentatively declined for Au/CeO2-X cata-
tra as demonstrated in Fig. SI3. The predominant peaks located at lysts, when X > 0.135. This result might correspond to the effect
lower binding energies of 882.0, 889.0 and 898.3, were classified of gold particle size and the nature of the support. The smaller gold
as characteristic of Ce 3d5/2. The higher binding energies of 900.5, particle size on the ceria support has a higher surface area, and
907.7 and 916.5 eV are characteristic of the Ce 3d3/2 spin–orbit thereby the lattice oxygen is more available on the ceria surface.
components. The two spin–orbit coupled doublets of Ce3+ were Moreover, a higher amount of surfactant in the solution led to
indexed with a BE of 880.5,885.7, 899.7 and 902.3 (denoted as v slower diffusion of ions to the crystal surface, reducing nucleus for-
symbol), respectively. Furthermore, the surface percentage Ce3+ mation of oxide, corresponding to a lower intensity of the lattice
content was calculated. The results demonstrated that Ce3+ con- oxygen peak (Fig. 7), when compared to the ones synthesized with
tents was consistent for the series of parent ceria supports, and a lower amount of CTAB.
were in the range 11.4–14.8%. Interestingly, it was noticeably dif-
ferent in the case of the gold catalysts. Au/CeO2-0.135 and
Au/CeO2-1.080 had the greatest surface Ce3+ contents, with 23.5 3.2. Catalyst performance
and 22.1% respectively (Table 3).
The surface oxidation state of Au was also measured by the XPS The parent ceria and Au/CeO2 catalysts series were examined
technique in order to studied the characteristics of gold on the dif- for their catalytic activity for the OSRM reaction between 200
ferent ceria supports. The XP spectra of Au 4f 5/2 and Au 4f 7/2 are and 400 °C. The methanol conversion and H2 production rates were
shown in Fig. 6, and these consist of a main pair of peaks located evaluated and are shown in Fig. 8. The reactivity of ceria only cat-
259
S. Eaimsumang, N. Chollacoop, A. Luengnaruemitchai et al. Journal of Catalysis 392 (2020) 254–265

a) 464 cm-1 25
CeO2-0.135 a)
CeO2-0.270 CeO2
CeO2-0.540 Au/CeO2
CeO2-1.080 20
Intensity (a.u.)

FWHM (cm )
-1
15

259 cm-1 600 cm-1 1170 cm-1 10

b) Au/CeO2-0.135
0.10
Au/CeO2-0.270 b)
0.1352
CeO 0.27 0.54 1.08
Au/CeO2-0.540
Au/CeO2
Au/CeO2-1.080 X Data
0.08
-1
Intensity (a.u)

600 cm-1 837 cm

-1
259 cm (A590+A1170)/A464
0.06

830 cm-1
0.04
248 cm-1

0.02

200 400 600 800 1000 1200 1400

Wavenumber (cm-1) 0.00

0.135 0.27 0.54 1.08
Fig. 4. Representative Raman spectra of the CeO2-X supports and Au/CeO2-X
catalysts. -X mmol CTAB
Fig. 5. Representative comparison of (a) FWHM and (b) defect concentration
between the CeO2-X and Au/CeO2-X catalysts, determined from Raman spectra.
alysts did not demonstrate any significant difference in both
methanol conversion and H2 production rate. The methanol con-
version was very low at 250 °C, and was due to combustion, with 100% conversion from 200 °C but CO selectivity was higher than 6%
only CO2 and CO produced. Oxygen in the feed stream was com- at 300 °C [33]. The bimetallic 3% Ni/Cu/ZrO2 performed only at
pletely consumed at 300 °C, and therefore above this temperature about 60% MeOH conversion at 400 °C and CO selectivity was dom-
methanol conversion resulted in the formation of H2. The methanol inant and was as high as 90% [34]. The methanol conversion was
conversion continuously increased with increasing reaction tem- complete only at 400 °C for 8.3% CuO/CeO2/Al2O3 [35]. 20% CuO/
perature, reaching the highest value (86–93%) at 400 °C. All metha- CeO2/Al2O3 also achieved 100% conversion in oxy-reforming of
nol conversion exceeded the maximum conversion of 68%, which is methanol at 200 °C, as reported by Mierczynski et al.[36]. Based
the theoretical maximum if partial oxidation of methanol (POM) is on these literature data, Cu catalysts were very active for OSRM
the only reaction operating, indicating that there was also the reac- reaction that can reach the complete MeOH conversion at lower
tion of methanol steam reforming or methanol decomposition temperature than the present work. However, the copper loading
competing. The catalytic performance of Au/CeO2-X catalysts low- was much higher than required for Au catalysts in order to obtain
ered the temperature for activity significantly when compared to the maximum conversion.
the activity of the parent ceria supports, as activity was observed The rate of H2 production was also studied, the parent ceria ini-
at temperatures around 200 °C lower when Au was added. The tially generated H2 at 300 °C, correlating to the beginning of the
gold-containing catalysts provided high methanol conversion, in reduction temperature of ceria, as evidenced from TPR results. This
the range 80–90%, from 200 °C. The methanol conversion increased was previously studied, and it was reported that the desorption
with increasing reaction temperature up to 400 °C, and almost temperature of H2 on ceria was about 280–311 °C [37]. In the case
100% methanol conversion was achieved for the gold catalyst ser- of the H2 production rate, the results also follow the same trend as
ies. The performance of Au/CeO2-X catalyst is compared to that of results as rates of methanol conversion. H2 production rates were
non-noble metal catalysts for OSRM reaction. The 6 wt% Cu/CeO2 similar for the bare ceria supports, because of the limitation in
can achieved 100% MeOH conversion at 230 °C but H2 selectivity terms of ceria’s reducibility and methanol decomposition at speci-
was the highest at about 26% [32]. The 70 at% CuO/CeO2 provided fic temperature. It is reported that oxidation reactions on the ceria

260
S. Eaimsumang, N. Chollacoop, A. Luengnaruemitchai et al. Journal of Catalysis 392 (2020) 254–265

a) Au/CeO2-0.135 Au 4f7/2 b) Au/CeO2-0.270 Au 4f7/2

Au0
Au+ Au 4f5/2
Au3+ Au 4f5/2
Intensity (cps)

Au 4f7/2 Au 4f7/2
c) Au/CeO2-0.540 d) Au/CeO2-1.080

Au 4f5/2
Au 4f5/2
Intensity (cps)

92 90 88 86 84 82 92 90 88 86 84 82

Binding energy (eV) Binding energy (eV)

Fig. 6. Representative Au4f XP spectra of the Au/CeO2-X catalysts.

Table 3
Summary of elemental surface composition from X-ray Photoelectron Spectroscopy analysis.

Sample Au (at %) Ce (at%) O (at%) Au0 (at%) Au+ (at%) Au3+ (at%) Aun+/Au0 Ce3+ (%) OI(%)
Au/CeO2-0.135 1.32 26.55 72.13 70 25.9 4.1 0.43 23.5 55.7
Au/CeO2-0.270 0.81 24.82 74.37 76 17.8 6.3 0.32 18.0 55.2
Au/CeO2-0.540 1.12 23.19 75.69 81.6 3.8 14.6 0.23 13.9 51.5
Au/CeO2-1.080 0.84 20.36 78.80 76.3 18.4 5.8 0.31 22.1 44.2

surface proceed via the Mars-van Krevelen mechanism, where an The H2 production rate in the present work is lower than Cu-based
adsorbate reacts with lattice oxygen on the ceria surface, followed catalysts because of different testing conditions in terms of feed
by product desorption. Subsequently, oxygen lattice vacancies at composition and space velocity. The 17.3% Cu/SiO2 provided
the surface are replaced by O2 from the gas phase to close the cat- 0.0102 mol H2/min·gcat for OSRM reaction at 250 °C with the molar
alytic cycle [11,38]. Interestingly, the H2 production rate ratio of H2O/MeOH/O2: 2/1/0.25 and 60,000 mL/h·gcat [39].
substantially improved by about 2-fold for the corresponding ser- Whereas, conventional Cu-ZnO/a-Al2O3 catalysts could produce
ies of catalysts containing gold. Au/CeO2-0.135 had the greatest H2 H2 0.003 mol/min·gcat at 200 °C for the OSRM reaction at space
production rate over the temperature range 200–400 °C. This result velocity of 31500 mL/h·gcat [40].
is in agreement with previous work that the highest methanol con- Product distribution of Au/CeO2 catalysts is demonstrated in
version and hydrogen yield in the OSRM reaction were obtained for Fig. 9. The H2 selectivity of all catalysts were varied as 51.5–
an Au/CeO2 catalyst having a gold particle size of about 8 nm [10]. 59.1%. The higher H2 selectivity of Au/CeO2-0.135 could be

261
S. Eaimsumang, N. Chollacoop, A. Luengnaruemitchai et al. Journal of Catalysis 392 (2020) 254–265

a) Au/CeO2-0.135 b) Au/CeO2-0.270
O 1s O 1s
OI
OII
OIII
Intensity (cps)

c) Au/CeO2-0.540 d) Au/CeO2-1.080
X Data X Data

O 1s O 1s
Intensity (cps)

538 536 534 532 530 528 526 538 536 534 532 530 528 526

Binding energy (eV) Binding energy (eV)

Fig. 7. Representative O 1 s XP spectra of the Au/CeO2-X catalysts.

observed at 200–250 °C in Fig. 9a, which might correlate to the ature for CuO/CeO2 catalysts in the OSRM reaction started at 160 °C
greater reducibility of catalyst. The easier migration of oxygen [33]. Whereas, the CuZnO introduced by Au and ZrO2 had the light-
can effectively activate the oxidation reaction and then induce oxy- off temperature > 100 °C for POM reaction [41].
gen vacancies that might initiate WGS/SRM reaction. A small
amount of CH4 generated less than 1% as a result of the reactivity 3.2.1. Reactivity of Au/CeO2-X catalysts in the OSRM reaction at low
of SRM product between H2 and CO. The trend of CO2 selectivity reaction temperature
reduced along with reaction temperature. Whereas, the CO selec- Typically, oxidation reactions are favored at lower temperature
tivity increased indicating more favorable methanol decomposi- due to their exothermic nature, and the opposite for steam reform-
tion at higher temperature. ing of methanol, because it is an endothermic reaction. Therefore,
The catalytic activity of Au/CeO2 catalysts in the OSRM reaction the capability of the catalyst to exhibit activity for oxidation at
was also tested at low temperature (100 °C). Interestingly, all Au/ lower temperature relies on the oxygen mobility within the cubic
CeO2 catalysts produced high MeOH conversion (46.5–56.3%), as fluorite ceria structure. This key feature of oxygen lattice mobility
given in Table 4. The reaction rate of MeOH consumption of Au/ and redox behavior are important for the oxidation activity.
CeO2-0.135 and Au/CeO2-0.270 is higher than Au/CeO2-0.540 and Different levels of methanol conversion were observed at
Au/CeO2-1.080 catalysts. It implies that our catalysts are more 200 °C. These differences may be correlated with the improved
active at lower reaction temperatures compared to other catalysts. redox properties of catalysts, as these are necessary for the oxida-
The 5–40% Cu/CeO2Al2O3 catalysts were also tested in the OSRM tion reaction, due to exchange of oxygen within the structure and
reaction at 100 °C. Only 4.4–6.6% MeOH conversion was obtained the role of participating active oxygen in the reaction. It has been
and the reaction rate of MeOH consumption was only in the range proposed that the lattice oxygen of ceria is responsible for stabiliz-
of 0.017–0.025 mmol/min·gcat [36]. Moreover, the light-off temper- ing a hydrogen atom. Whereas, the oxygen vacancies act as an
262
S. Eaimsumang, N. Chollacoop, A. Luengnaruemitchai et al. Journal of Catalysis 392 (2020) 254–265

62 1.0
a)
60
0.8
58

CH4 selectivity (%)

H2 selectivity (%)
56
0.6

54

0.4
52

50
0.2
48

46 0.0
200 250 300 350 400
o
Temperature ( C)
60 60
b)
Au/CeO2-0.135
Au/CeO2-0.270
50 50
Au/CeO2-0.540
Au/CeO2-1.080
CO2 selectivity (%)

CO selectivity (%)
40 40

30 30

20 20

10 10

0 0
200 250 300 350 400
o
Temperature ( C)
Fig. 8. Catalytic activity of the CeO2-X supports and Au/CeO2-X catalysts for the
Fig. 9. Product selectivities of the CeO2-X and Au/CeO2-X catalysts for OSRM
OSRM reaction at GSHV = 30,000 mL/g.h as measured in terms of a) methanol
reaction between 200 and 400 °C.
conversion (%) and b) H2 production rate (mol/min·gcat).

perature can come from either methanol decomposition reaction

adsorption site for an intermediate methoxy species [1]. Therefore,
or incomplete combustion. CO formation and increasing H2O ratio
Au/CeO2-0.135 and 0.270 exhibited higher methanol conversion
due to by-product of combustion can sufficiently drive the thermo-
than other catalysts as summarized in Table 4, because they had
dynamic equilibrium for the WGS reaction. It could promote the
the highest H2 uptake when normalized for surface area. These
low temperature WGS reaction, which normally was in the range
unique properties can promote the OSRM activity. Moreover, the
of 180–250 °C for Au and Cu catalysts [42,43]. Therefore, the
catalyst surface area normalized H2 uptake also influenced the
WGS reaction may play a key role in determining H2/CO2 selectiv-
selectivity of H2/CO2 formation (Table 4). The highest H2/CO2 for-
ity at the lowest temperature, due to its slightly exothermic nature.
mation was also shown by Au/CeO2-0.135 (1.43), and it decreased
One of the most important energy barriers for this reaction is dis-
in the following order:
sociation of water into OH and H on the ceria surface [44]. Conse-
quently, the low temperature WGS proceeds via CO reacting with
Au=CeO2  0:135 ð1:43Þ > Au=CeO2  0:270 ð1:39Þ OH to form OHCO as an intermediate species and then desorbs as
> Au=CeO2  1:080 ð1:23Þ CO2 gas. Whereas, 2H atoms combine together and then desorb
as H2 gas. It was postulated that the metal surface can effectively
> Au=CeO2  0:540 ð1:11Þ bind CO near the interface and react with OH located at the oxygen
However, the ratio of H2/CO2 was lower than the theoretical vacancies adjacent to the metal atom [45]. This can support the
value, which is in the range 2–3, for the combination of SRM and reactivity of the WGS reaction.
POM reactions, indicating the reaction pathway did not mainly The selective H2 formation might be significantly influenced by
proceed via POM and subsequently alternative steam reforming the WGS reaction. This aspect of Au/CeO2 catalysts for WGS were
of methanol. The reason for the higher CO2 formation in the pro- emphasized, such as the interface between Au/CeO2, available
duct stream may be that additional side reactions, such as the active surface of ceria and the oxidation state of Au. The cationic
WGS, may also operate. The sources of CO formation at low tem- gold species might be responsible for higher activity of H2 produc-

263
S. Eaimsumang, N. Chollacoop, A. Luengnaruemitchai et al. Journal of Catalysis 392 (2020) 254–265

Table 4
The catalytic activity of Au/CeO2 catalysts in OSRM reaction at 100 and 200 °C.

Sample MeOH conversion (%) rMeOH (mmol/min·gcat) rH2 (mmol/min·gcat) MeOH conversion H2/CO2
at 100 °C at 100 °C at 100 °C at 200 °C at 200 °C
Au/CeO2-0.135 56.3 1.84 1.32 92.3 1.43
Au/CeO2-0.270 54.4 1.78 1.37 91.5 1.39
Au/CeO2-0.540 46.5 1.51 0.98 82.5 1.11
Au/CeO2-1.080 48.4 1.59 1.26 89.4 1.23

tion at the lowest reaction temperature as seen in Fig. 9a. The catalysts, with small gold nanoparticles of less than 5 nm, showed
cationic gold species were previously reported as the active state high catalytic activity for partial oxidation of methanol to produce
for WGS reaction [46]. However, they were an unstable and could hydrogen [15,52]. Oxygen adsorbed on oxide supports was also
be easily reduced to the metallic phase under WGS reaction [47]. more readily available in the case of small Au particles [2]. Haruta
Hence, the H2 selectivity of Au/CeO2-0.135 has a high value at et al. proposed that CO oxidation proceeded at the interface
low temperature, and then dropped at higher reaction tempera- between the gold nanoparticle and the ceria support, with smaller
ture. Whereas, the H2 selectivity of Au/CeO2-0.540 was lower at sized gold nanoparticles less than 5 nm being the most active [49].
the lowest temperature in comparison with others, and then it The small gold particle size might play a key role for providing
increased at higher temperature. This is probably relevant to the active sites for the SRM reaction. The interaction between smaller
dominant Au0 species, which was considered as more stable gold gold particles and ceria support is beneficial compared to larger
state, resulting in a relatively stable catalyst at high temperature. gold particles, resulting in weakening of the Ce-O bond to generate
The deactivation of Au catalysts in the WGS reaction at lower tem- oxygen vacancies adjacent to the gold particle. Surface gold and
perature was due to less available surface area of ceria, and strong oxygen vacancies participate in promoting the catalytic activity
adsorption of CO2 could block the active sites for dissociative water by stabilizing an intermediate species. Furthermore, the more
adsorption leading to a slower rate [48]. While the loss of WGS loosely bound oxygen species on the surface can facilitate the con-
activity at high temperature was affected by the destruction of version of intermediate species in the SRM reaction [1].
Au/CeO2 interactions due to the lost contact at the interface
between Au and ceria (most probably cationic gold). 4. Conclusion

3.2.2. The role of Au in the OSRM reaction Hydrothermal CTAB surfactant assisted synthesis of Bundle-like
In the present study, the surface analysis by XPS showed the ceria nanorods can reduce particle agglomeration, leading to better
distribution of Au species, which include Au0, Au+, and Au 3+ for physical properties and higher surface area. However, higher con-
the Au/CeO2 catalysts. It is observed that the catalyst having the centrations of CTAB (>0.270 mmol) can form ceria with less favor-
highest Aun+/Au0 ratio had a very high catalytic activity compared able properties, due to suppression of ion diffusion during
to the other catalysts. This agrees well with our previous work for synthesis crystallization. The properties of CeO2-0.135 resulted in
preferential CO oxidation [7], OSRM [10] and SRM [2] studies cat- formation of a catalyst that most effectively dispersed gold to give
alyzed by Au on different CeO2 morphologies. Several studies have the best performance for OSRM. The more active (2 0 0) surface of
proposed that the active site for catalytic reactions occurred on the the ceria nanorods can disperse gold to a greater extent, producing
perimeter of the gold particle with the oxide interface [48–50]. As smaller gold particles exposing the (1 1 1) and (2 0 0) crystal planes,
the cationic gold is less stable than metallic gold, there are litera- and a higher concentration of cationic gold species. The active site
ture reports that the formation of H2 on cationic gold caused the of cationic gold, smaller gold particle size and improved redox
reduction of cationic gold to metallic gold for Au/CeOx-TiO2 and properties played a major role in determining better catalytic
Au/Ce(Gd)Ox for the WGS reaction [47,51]. The oxidic gold is activity, characterized by higher methanol conversion and H2 con-
related to the interaction between a gold ion and an oxygen atom. tent for OSRM at the lowest temperature of 200C.
The oxidation state of Au can also be influenced by gold particle
size and how well-dispersed particles are on the supports. Corre- Declaration of Competing Interest
sponding to the TEM images, the smaller gold particles are present
on the active surface (2 0 0) for Au/CeO2-0.135. Whilst, significantly The authors declared that there is no conflict of interest.
larger sized gold particles (13.2 nm) were present on the most
stable (1 1 1) surface for Au/CeO2-1.080. This indicates that the Acknowledgements
more active (2 0 0) ceria surface can disperse gold particles to a
greater extent than the less active (1 1 1) ceria surface. Moreover, The work was supported by the Ratchadaphiseksomphot
this increased dispersion resulted in improving redox properties Endowment Fund, Chulalongkorn University (CU-GES-60-04-63-
for ceria after gold incorporation. Au supported on ceria prepared 03). The authors thank the Thailand Research Fund (TRF) and
with lower CTAB concentration had greater redox-properties, as National Science and Technology Development Agency
evidenced from the higher amount of H2-uptake in Table 2. The (PHD/0237/2558) for the scholarship funding of Ms. Srisin Eaim-
improvement of redox properties related to the dispersion of Au sumang. We are grateful to Dr. Thomas David, the School of Chem-
particles on a ceria support has been reported previously [6]. More- istry, Cardiff University, UK for technical support of HR-TEM.
over, it represents the presence of loosely oxygen bound species on
the ceria surface and oxygen mobility within the structure.
References
However, the optimum gold size for the specific reaction is still
debated. The current study found that the highest methanol [1] N. Yi, R. Si, H. Saltsburg, M. Flytzani-Stephanopoulos, Appl. Catal. B Environ. 95
conversion and hydrogen yield were obtained with small gold (2010) 87.
particle sizes of 7.9 nm, in the same way reported previously for [2] N. Yi, R. Si, H. Saltsburg, M. Flytzani-Stephanopoulos, Energy Environ. Sci. 3
(2010) 831.
3%Au/CeO2 in the OSRM reaction [10]. The gold particle size has [3] D. Andreeva, V. Idakiev, T. Tabakova, L. Ilieva, P. Falaras, A. Bourlinos, A.
an important influence on activity, for example, supported gold Travlos, Catal. Today 72 (2002) 51.

264
S. Eaimsumang, N. Chollacoop, A. Luengnaruemitchai et al. Journal of Catalysis 392 (2020) 254–265

[4] A. Luengnaruemitchai, S. Osuwan, E. Gulari, Catal. Commun. 4 (2003) 215. [29] X.S. Huang, H. Sun, L.C. Wang, Y.M. Liu, K.N. Fan, Y. Cao, Appl. Catal. B Environ.
[5] H.Y. Kim, G. Henkelman, J. Phys. Chem. Lett. 3 (2012) 2194. 90 (2009) 224.
[6] F. Yinga, S. Wang, C.-T. Au, S.-Y. Lai, Gold Bull. 43 (2010) 241. [30] A. Longo, L.F. Liotta, G. Pantaleo, F. Giannici, A.M. Venezia, A. Martorana, J.
[7] M. Carltonbird, S. Eaimsumang, S. Pongstabodee, S. Boonyuen, S.M. Smith, A. Phys. Chem. C 116 (2012) 2960.
Luengnaruemitchai, Chem. Eng. J. 344 (2018) 545. [31] W. Zhang, X. Niu, L. Chen, F. Yuan, Y. Zhu, Sci. Rep. 6 (2016) 29062.
[8] A. Luengnaruemitchai, S. Osuwan, E. Gulari, Int. J. Hydrog. Energy 29 (2004) [32] R. Pérez-Hernández, A. Gutiérrez-Martínez, C.E. Gutiérrez-Wing, Int. J. Hydrog.
429. Energy 32 (2007) 2888.
[9] C. Pojanavaraphan, A. Luengnaruemitchai, E. Gulari, Chem. Eng. J. 192 (2012) [33] P.P.C. Udani, P.V.D.S. Gunawardana, H.C. Lee, D.H. Kim, Int. J. Hydrog. Energy
105. 34 (2009) 7648.
[10] S. Eaimsumang, S. Wongkasemjit, S. Pongstabodee, S. MeejooSmith, S. [34] R. Pérez-Hernández, A. Gutiérrez-Martínez, M.E. Espinosa-Pesqueira, M.L.
Ratanawilai, N. Chollacoop, A. Luengnaruemitchai, J. Rare Earths 37 (2019) Estanislao, J. Palacios, Catal. Today 250 (2015) 166.
819. [35] C. Cammarano, G. Bagnasco, M. Turco, E. Moretti, L. Storaro, A. Talon, M.
[11] D.R. Mullins, Surf. Sci. Rep. 70 (2015) 42. Lenarda, One-step synthesized CuO-CeO2-Al2O3 system as catalyst for OSRM
[12] S. Agarwal, B.L. Mojet, L. Lefferts, A.K. Datye, Chapter 2 - Ceria nanoshapes— process, in, 2008.
structural and catalytic properties, in: Z. Wu, S.H. Overbury (Eds.), Catalysis by [36] P. Mierczynski, M. Mosinska, W. Maniukiewicz, M. Nowosielska, A.
Materials with Well-Defined Structures, Elsevier, Amsterdam, 2015, p. 31. Czylkowska, M.I. Szynkowska, React Kinet Mech Cat. 127 (2019) 857.
[13] Z. Guo, F. Du, Z. Cui, Mater. Lett. 61 (2007) 694. [37] D.G. Araiza, A. Gómez-Cortés, G. Díaz, Catal. Today 282 (2017) 185.
[14] A.A. Tillirou, C.R. Theocharis, Adsorp Sci Technol 26 (2008) 687. [38] C. Doornkamp, V. Ponec, J. Mol. Catal. A: Chem. 162 (2000) 19.
[15] T.C. Ou, F.W. Chang, L.S. Roselin, J. Mol. Catal. A: Chem. 293 (2008) 8. [39] L. Selva Roselin, H.-W. Chiu, J. Saudi Chem. Soc. 22 (2017) 692.
[16] G. Leofanti, M. Padovan, G. Tozzola, B. Venturelli, Catal. Today 41 (1998) 207. [40] J.B. Wang, C.-H. Li, T.-J. Huang, Catal. Lett. 103 (2005) 239.
[17] C.S. Pan, D.S. Zhang, L.Y. Shi, J. Solid State Chem. 181 (2008) 1298. [41] H.-Y. Huang, H.-I. Chen, Y.-J. Huang, Catalysts 8 (2018).
[18] S.R. Ede, U. Nithiyanantham, S. Kundu, PCCP 16 (2014) 22723. [42] A.M. Abdel-Mageed, G. Kučerová, J. Bansmann, R.J. Behm, ACS Catal. 7 (2017)
[19] M. Lin, Z.Y. Fu, H.R. Tan, J.P.Y. Tan, S.C. Ng, E. Teo, Cryst. Growth Des. 12 (2012) 6471.
3296. [43] A. Jha, D.-W. Jeong, J.-O. Shim, W.-J. Jang, Y.-L. Lee, C.V. Rode, H.-S. Roh, Catal.
[20] N. Garti, H. Zour, J. Cryst. Growth 172 (1997) 486. Sci. Technol. 5 (2015) 2752.
[21] W.-J. Li, E.-W. Shi, T. Fukuda, Cryst. Res. Technol. 38 (2003) 847. [44] J.A. Rodriguez, P. Liu, J. Hrbek, J. Evans, M. Pérez, Angew. Chem. Int. Ed. 46
[22] M.Y. Cui, J.X. He, N.P. Lu, Y.Y. Zheng, W.J. Dong, W.H. Tang, B.Y. Chen, C.R. Li, (2007) 1329.
Mater. Chem. Phys. 121 (2010) 314. [45] K.G. Azzam, I.V. Babich, K. Seshan, L. Lefferts, J. Catal. 251 (2007) 153.
[23] S. Scirè, S. Minicò, C. Crisafulli, C. Satriano, A. Pistone, Appl. Catal. B Environ. 40 [46] Q. Fu, W. Deng, H. Saltsburg, M. Flytzani-Stephanopoulos, Appl. Catal. B
(2003) 43. Environ. 56 (2005) 57.
[24] C.T. Nottbohm, C. Hess, Catal. Commun. 22 (2012) 39. [47] R. Si, J. Tao, J. Evans, J.B. Park, L. Barrio, J.C. Hanson, Y.M. Zhu, J. Hrbek, J.A.
[25] S. Chen, L. Li, W. Hu, X. Huang, Q. Li, Y. Xu, Y. Zuo, G. Li, ACS Appl. Mater. Rodriguez, J. Phys. Chem. C 116 (2012) 23547.
Interfaces 7 (2015) 22999. [48] R. Burch, PCCP 8 (2006) 5483.
[26] R.B. Zhang, K. Lu, L.J. Zong, S. Tong, X.W. Wang, J. Zhou, Z.H. Lu, G. Feng, J. Mol. [49] M. Haruta, Catal. Today 36 (1997) 153.
Catal. 442 (2017) 173. [50] L.M. Molina, B. Hammer, Appl. Catal. A 291 (2005) 21.
[27] C. Schilling, A. Hofmann, C. Hess, M.V. Ganduglia-Pirovano, J. Phys. Chem. C [51] Q. Wang, J.C. Hanson, A.I. Frenkel, J. Chem. Phys. 129 (2008) 234502.
121 (2017) 20834. [52] E. Hernández-Ramírez, J.A. Wang, L.F. Chen, M.A. Valenzuela, A.K. Dalai, Appl.
[28] Q. Fu, H. Saltsburg, M. Flytzani-Stephanopoulos, Science 301 (2003) 935. Surf. Sci. 399 (2017) 77.

265