Optimizing Pd:Zn molar ratio in PdZn/CeO2 for CO2 hydrogenation to methanol

Article abstract of DOI:10.1016/j.apcata.2019.117185

We report the compositional optimization of Pd:Zn/CeO₂ catalysts prepared via sol-gel chelatization for the hydrogenation of CO₂ under mild reaction conditions. The formation of a PdZn alloy, which is the main active phase for this reaction, was maximized for the catalyst with a Pd to Zn ratio close to 1. For this catalyst, a maximum conversion of 14%, close to thermodynamic equilibrium, and high selectivity to methanol (95%) were achieved at 220 °C, 20 bar, 2400 h^?1 GHSV and H₂:CO₂ stoichiometric ratio of 3:1. The formation of PdZn alloys was achieved by reducing the catalyst precursor at 550 °C under hydrogen flow and confirmed by XRD. XPS study confirmed the presence of Pd°, being maximum for the optimized catalyst composition. At lower temperature, i.e. 180 °C, 1.0PdZn catalyst showed 100% selectivity to methanol with 8% CO₂ conversion. RWGS reaction is responsible for the production of CO and its selectivity increases with temperature. In situ DRIFTS suggests that CO₂ is activated as adsorbed CO₃- species over CeO₂. Surface micro-kinetics demonstrates that methanol can be formed either via formaldehyde or formic acid surface intermediates.

Full text of DOI:10.1016/j.apcata.2019.117185

Accepted Manuscript
Title: Optimizing Pd:Zn molar Ratio in PdZn/CeO2 for CO2
Hydrogenation to Methanol
Authors: Opeyemi A. Ojelade, Sharif F. Zaman, Muhammad
A. Daous, Abdulrahim A. Al-Zahrani, Ali S. Malik, Hafedh
Driss, Genrikh Shterk, Jorge Gascon
PII:
DOI:
Article Number:
S0926-860X(19)30340-0
https://doi.org/10.1016/j.apcata.2019.117185
117185
Reference:
APCATA 117185
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
13 May 2019
28 July 2019
29 July 2019
Please cite this article as: Ojelade OA, Zaman SF, Daous MA, Al-Zahrani AA,
Malik AS, Driss H, Shterk G, Gascon J, Optimizing Pd:Zn molar Ratio in
PdZn/CeO2 for CO2 Hydrogenation to Methanol, Applied Catalysis A, General (2019),
https://doi.org/10.1016/j.apcata.2019.117185
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Optimizing Pd:Zn molar Ratio in PdZn/CeO2 for CO2 Hydrogenation to Methanol
Opeyemi A. Ojeladea, Sharif F. Zamana*, Muhammad A. Daousa, Abdulrahim A. Al-
Zahrania, Ali S. Malika, Hafedh Drissa, Genrikh Shterkb, Jorge Gasconb
a Chemical and Materials Engineering Department, Faculty of Engineering, King
Abdulaziz University, P.O. Box 80204, Jeddah, 21589, Saudi Arabia
b Advanced Catalytic Materials, KAUST Catalysis Center (KCC), King Abdullah
University of Science and Technology, Thuwal 23955, Saudi Arabia
Corresponding author: Sharif F. Zaman, Associate Professor, Chemical and Materials
Engineering Department, Faculty of Engineering, King Abdulaziz University, P.O. Box
80204, Jeddah, 21589, Saudi Arabia. Email: sfzaman@kau.edu.sa; zfsharif@gmail.com.
Graphical Abstract
18
16
14
12
10
8
50
40
30
20
10
0
CO
CH OH
3
CO
2
100
90
80
70
60
50
180
200
220
240
260
280
Temperature (oC)
1

Highlights
 14% CO2 conversion and 95% methanol selectivity were achieved over
PdZn/CeO2
 1.0PdZn/CeO2 catalysts showed optimum catalytic performance
 Two possible routes for methanol formation via formaldehyde and/or formic acid
hydrogenation are proposed.
Abstract
We report the compositional optimization of Pd:Zn/CeO2 catalysts prepared via sol-gel
chelatization for the hydrogenation of CO2 under mild reaction conditions. The formation
of a PdZn alloy, which is the main active phase for this reaction, was maximized for the
catalyst with a Pd to Zn ratio close to 1. For this catalyst, a maximum conversion of 14%,
close to thermodynamic equilibrium, and high selectivity to methanol (95%) were achieved
at 220 oC, 20 bar, 2400 h-1 GHSV and H2:CO2 stoichiometric ratio of 3:1. The formation
of PdZn alloys was achieved by reducing the catalyst precursor at 550oC under hydrogen
flow and confirmed by XRD. XPS study confirmed the presence of Pdo, being maximum
for the optimized catalyst composition. At lower temperature, i.e. 180 oC, 1.0PdZn catalyst
showed 100% selectivity to methanol with 8% CO2 conversion. RWGS reaction is
responsible for the production of CO and its selectivity increases with temperature. In situ
-
DRIFTS suggests that CO2 is activated as adsorbed CO3 species over CeO2. Surface
micro-kinetics demonstrates that methanol can be formed either via formaldehyde or
formic acid surface intermediates.
Keywords: PdZn alloy, CeO2, Citric acid, Chelating agent, Methanol synthesis, DRIFTS.
1.
Introduction
2

Global energy demand is estimated to reach about 20 billion tonnes of oil equivalent
(BTOE) by 2040 [1], from which approximately a 70% will be derived from fossil fuels,
consequently increasing atmospheric carbon build-up. To objectify the target of limiting
the average global temperature rise below 2oC, most countries of the world have agreed to
cut down their CO2 production [2]. In this spirit, Olah [3] proposed the “methanol
economy”, whereby atmospheric CO2 (0.041%) and those captured from industrial
effluents can be converted to methanol, a convenient liquid fuel, hydrogen storage material
and feedstock for hydrocarbon syntheses and their products. To make this process carbon
neutral, the required hydrogen must be supplied from renewable sources, i.e. hydrogen via
water splitting using energy sources like solar, wind or hydrothermal [4].
Currently, methanol is produced using syngas over a Cu/ZnO/Al2O3 catalyst at 250-300oC
and 50-100bar [5, 6]. Unfortunately, the use of similar Cu-based catalyst for the
hydrogenation of CO2 requires excessive pressures to reach decent selectivity to methanol
[7,8]. Meanwhile, it is crucial to note that, Cu supported on ZnO, ZrO2, and CeO2
represents a promising alternative for the process. Metal-oxide centers formed by doping
CeOx on metallic Cu(111) enhanced the chemical properties of CeOx/Cu(111) catalyst and
consequently helped in improving the reaction pathway for methanol synthesis [9]. Also,
the partial coverage of ZnOx on the surface of reduced Cu nanoparticles can form either
Zn/Cu bimetallic alloy or ZnO-Cu interface, serving as the active sites for methanol
synthesis reaction [10-12].
These challenges have opened research opportunities in the search for alternative catalysts
for CO2 hydrogenation to methanol. Pd-based catalysts have proven to be resistive to
sintering and more stable than the conventional Cu-based catalysts for this process,
3

although they suffer from low activity at low temperature and low selectivity at high
temperature [13,14]. The formation of bimetallic alloys (PdZn, PdGa, PdZr) of Pd
supported on ZnO, Ga2O3 and ZrO2 showed enhanced catalytic activity and selectivity to
methanol compared to Pd/support alone [15-22]. Similarly, for methanol steam reforming
(MSR), PdZn surface alloy and ZnPd intermetallic compounds (IMCs) are generally agreed
upon to modify the adsorption properties of metallic Pd and enhance CO2 selectivity and
methanol conversion [23, 24, 25, 26]. Bahruji et al. studied the influence of PdZn particle
size over CO2 conversion and methanol selectivity [18, 27] and found that smaller particles,
with a strong interaction with the support, resulting in more selective catalysts. Pd/Zr
catalysts prepared via in-situ activation revealed isolation of Pd particles from the bulk and
showed higher methanol yields than catalysts prepared by impregnation [26]. Xu et al. [13]
investigated different preparation methods of PdZn over Al2O3 and reported that co-
precipitation method is superior over deposition precipitation and impregnation methods
in terms of methanol selectivity.
Supports like ZrO2, ZnO, SBA-15, MCM-48, Ga2O3 and CeO2 have been tested as potential
supports for Pd in CO2 hydrogenation to methanol [13, 20, 28, 29, 30]. However, CeO2
promotes CO2 activation at the metal/support interface due to its ability to create surface
and bulk oxygen vacancies [31]. Doping Ca on PdZn/CeO2 catalysts improved surface
basicity and subsequently assisted the reducibility of ceria by creating more oxygen vacant
sites and enhanced catalytic activity [17]. Pd promotion in Cu/CeO2 based catalysts
improves the dispersion of Cu and overall catalytic activity [32, 33]. The high cost of Pd
($40.58/gram) necessitates its optimized loading in catalysts.
4

Our group recently reported 5wt%Pd-5wt%Zn/CeO2 (Pd:Zn molar ratio = 0.67) catalysts
for CO2 hydrogenation to methanol via sol-gel technique [17], where structural and textural
properties of the catalyst can be controlled using this preparation method [34], showing
enhanced activity and selectivity for this process. This research focuses on optimizing
Pd:Zn molar ratio (0.7 to 1.1) over CeO2 support prepared via citric acid (CA) chelatization
method based on our previous finding, having the aim to improve the catalytic activity for
methanol synthesis. The synthesized catalysts were characterized using BET, XRD, XPS,
SEM/EDX, EELS-STEM, CO2 TPD, H2-TPR techniques to determine physiochemical
properties and to correlate intrinsic catalytic performance and surface properties. In situ
DRIFT studies were performed in an attempt to identify the surface reaction mechanism
for methanol synthesis.
2.
Experiment
2.1. Catalyst Preparation Procedure
PdZn/CeO2 catalysts with varying Pd:Zn nominal molar ratio (0.7 to 1.1) were synthesized
via sol-gel using CA as a chelating agent as reported in our previous report (CA:Metal
molar ratio of 3:1) [17]. Throughout this article, the prepared catalysts are named as
0.7PdZn, 0.9PdZn, 1.0PdZn and 1.1PdZn, where the preceding number designates the
corresponding molar ratio. The weight percentage of Zn was kept constant (5wt%) while
varying Pd loading to attain the desired Pd:Zn molar ratio. Briefly, a calculated amount of
nitrate salts (Pd, Zn and Ce) and CA were dissolved in 30ml water contained in separate
beakers and mixed dropwise; Pd(NO3)2.xH2O to Zn(NO3)2.6H2O, Ce(NO3)3.6H2O salt to
CA. Obtained Pd-Zn and Ce-CA solutions were mixed under stirring at 90oC . As the
5

interaction approached end-point, the metal ions began to form a suspension in CA which
finally was evaporated leaving behind a gel of integrated network. The gel was aged for
24hrs at 90oC in a water bath followed by overnight drying in an oven at 110oC. The dried
catalyst was pulverized and calcined in dry air in an open crucible at a temperature of 500oC
for 5hrs (5oC/min ramping rate) leaving behind the oxidic form of catalytic precursors.
2.2. Characterization Methods
N2-physiosorption analysis was conducted at 77K using a Micrometrics NOVA 2200e
analyzer. Catalyst samples were degassed at 300oC for 2hrs prior to analysis to remove
entrapped moisture. The surface area, pore volume (PV) and pore size distribution (PSD)
were obtained utilizing Brunauer-Emmett-Teller (BET) at a relative pressure range of P/Po
= 0.05-0.3 and non-local density functional theory (NLDFT) equilibrium models of slit
pore geometry. Pore volume was estimated from N2 isotherms at a relative pressure of P/Po
= 0.98 by utilizing the bulk liquid N2 density.
Temperature programmed reduction (H2-TPR) analysis of calcined catalysts was
conducted using a Quantachrome Pulsar Analyzer. Typically, 0.1g of a sample was loaded
into a U-type quartz reactor and adsorbed moisture was removed from the surface of the
catalyst at 150oC under Ar flow for about 2h and the temperature was then cooled down to
30oC. TPR experiments were conducted from room temperature to 800oC under 20ml/min
flow of 5%H2/N2 mixture at a ramping rate of 10K/min, during which the H2 consumption
was recorded.
6

Temperature programmed desorption (CO2-TPD) analysis of all calcined samples was
carried out using a micromeritics (AutoChem HP) chemisorption analyzer coupled with a
mass spectrometer. Similar to TPR experiments, 0.1g of a sample was loaded into a u-type
reactor. Prior to CO2 adsorption, catalyst precursors were reduced at 550oC under H2 flow
(20 mL/min) for 1 h and then cooled down to room temperature under 20 mL/min He flow
for 2 h. Then, the in situ reduced catalyst was saturated with a 20 mL/min flow of 15%
CO2/Ar mixture for 1h followed by TPD experiment with a heating rate of 10°C/min under
He flow (50 mL/min).
X-ray Diffraction (XRD) analyses were carried out at room temperature in the diffraction
angle range of 0-110o using an Equinox 1000 diffractometer (Co-kα radiation of 0.1709
nm wavelength).
X-ray photoelectron spectroscopy (XPS) was performed on a SPECS GmbH (Germany)
with non-monochromatic Al-Kα X-ray radiation (1486.6 eV). In a typical analysis, the
sample was stuck to a double-sided carbon adhesive tape and attached to the sample holder.
Then, the sample was introduced into the load lock chamber and evacuated overnight until
ultra-high vacuum (UHV) pressure of 10-8 mbar was reached. This is followed by
transferring the sample into the analysis chamber under UHV pressure of 4.8 x 10-10 mbar
for analysis. XPS data were acquired utilizing a PHOIBOS 150 MCD-9 hemispherical
energy analyzer. Adventitious carbon (C1s) at 284.8 eV corresponding to C-C bond was
used as binding energy (B.E) reference for charge correction [35]. The deconvolution of
the XPS peaks and the evaluation of the spectra were performed using the CasaXPS
software package version 2.3.16.
7

Energy dispersive spectroscopy (EDS) with a scanning electron microscope (SEM) was
performed on a Quanta FEG 450 (FEI Brand) to determine the elemental composition and
its distribution over catalysts grain.
In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was
conducted using a Bruker Tensor II FTIR spectrometer installed with Harrick praying
mantis and ZnSe window. All experimental infrared spectra were recorded in the
wavenumber range of 800-4000cm-1. In the standard procedure, about 0.1g of catalyst was
loaded into the sample holder and the catalyst was reduced for 1hr under 20ml/min H2 flow
at 550oC (atmospheric pressure) and subsequently cooled to desired reaction temperature
(220oC). At this temperature, the catalyst was exposed to 15ml/min Ar flushing for about
2hrs to remove gaseous and weakly adsorbed species before taking the background
spectrum. Then, the catalytic experiments were performed under 20ml/min CO2/H2
mixture (or CO2) flow to determine the intermediate species formed at different reaction
conditions (temperature and pressure). Spectra were obtained continuously by averaging
100 scans with 4cm-1 resolution and 30 IR spectra and were collected at 2mins interval for
1hr.
2.3. Methanol Synthesis Reactor Setup
Catalyst activity was tested in a tubular fixed-bed micro reactor (9.1mm x 300mm),
equipped with Bronkhorst mass flow controllers and temperature sensors. In a typical
experiment, the reactor was packed with 0.05g of quartz wool and 0.5g of calcined catalyst.
The catalyst was exposed to 20ml/min Ar flow while the temperature was raised to 550oC.
At this temperature, the flow was switched to 20ml/min H2 for 1hr to reduce the catalyst
and then subsequently cooled down to the desired reaction temperature (i.e. 220oC).
8

Afterwards, 20ml/min flow of CO2/H2 with stoichiometric feed ratio of 1:3 was fed into
the reactor and pressure was raised to 20bar (1bar/min ramp). Methanol synthesis reaction
was conducted until steady state was achieved (3hrs unless otherwise stated). Online
product stream analyses were performed using gas chromatography (GC) (Agilent 7890B)
equipped with HP-pona capillary column for FID (methanol detection) and Haysep Q-
packed column for TCD (H2, CO and CO2 detection). To avoid product condensation, the
post reaction line was maintained at a temperature of 150oC for all experiments. Product
analysis was performed by taking an average of three independent readings (within the
reproducibility range of 5% error) and the data reported herein for product selectivity,
methanol space-time yield (STY) and CO2 conversion were calculated using Eqs. (1-3).
F
CO
−F
CO
2
2
out
in
( )
%
CO2 Conversion:
XCO
=
∗ 100
∗ 100
(1)
(2)
2
F
CO
2
in
F
MeOH
( )
Methanol Selectivity: SMeOH % =
F
CO
−F
CO
2
2
out
in
Mass of MeOH (g)
Catalyst weight (g) ×Hour
Methanol STY:
STYMeOH
=
(3)
Fi = Molar flow rate of species ‘i’.
3.
Results and Discussion
3.1. Catalysts Characterization
3.1.1. Surface Area and Pore Size Distribution
The textural properties of calcined and reduced (fresh) catalysts were measured by
performing N2 physisorption analysis at 77 K. The adsorption isotherms of all reduced
9

catalysts (Fig. 1A) are typical of type IV. The existence of a hysteresis closing at 0.42 P/Po
is a characteristic which indicates the presence of mesopores smaller than 3.4 nm and
bigger cavities in the catalyst structure [36, 37]. BET surface area of the reduced catalysts
was determined in the range of 36-43 m2/g (Table 1). The pore volume of the fresh catalysts
increased with increasing PdZn molar ratio from 0.05-0.09 cm3/g when PdZn ratio was
increased from 0.7 to 1.0 and then observed a decrease in PV (0.05 cm3/g) for 1.1PdZn
catalyst (Table 1). This can be attributed to the formation of larger pores during
chelatization. Such an increase in pore volume is more evident in the PSD plot shown in
Fig. (1B) with a broad peak at 4.2 nm in the mesopore region. For 0.7, 0.9 and 1.1 PdZn
catalysts, two families of pores confirm the presence of a micro and mesopore bimodal
structure. However, the 1.0PdZn catalyst contains only a mesoporous structure in the range
of 4-15 nm.
3.1.2. CO2- TPD
TPD analysis of adsorbed CO2 was conducted in He flow to understand the basicity of the
in situ reduced catalysts (Fig. 2A). All the spectra exhibited low CO2 desorption
temperature with peaks in the range of 80-91oC, attributable to CO2 physically adsorbed on
the catalyst surface. Fan and Fujimoto [38] reported the possibility of CO2 to decompose
into CO over Pd/CeO2 at low temperature. The amount of CO2 desorbed from the catalyst
surface for all samples ranged from 19.7-75 μmol/gcat (Table 1). It appeared from the TPD
profile of pure CeO2 (19.68 μmol/gcat) that the adsorption of CO2 mainly happened on
CeO2/Ce2O3 interfacial site due to its high basicity and surface defects generated after
reduction in H2. The presence of Zn in the vicinity of CeO2 slightly increased CO2
10

adsorption on Zn/CeO2 surface (21.56 μmol/gcat). More so, in our previous report, the
presence of PdZn alloy notably modified the basicity and oxygen vacancies of ceria leading
to enhanced CO2 adsorption capacity [17]. If we take the CO2 desorbed amount as a
measure of basicity, the catalyst 1.0PdZn showed the highest amount of basicity (75
μmol/gcat) among the investigated catalysts and showed a superior catalytic performance
(explained in the activity section 3.2.1). CO2 peak around 90oC is attributed to the
-
2-
decomposition of b-HCO3 while the shoulder peak around 210oC indicates b-CO3 and
2-
m-CO3 species [39], indicating that the surface oxygen vacancies of ceria can form
2-
carbonates species without any constraint. We have observed the CO3 formation in the
DRIFT study (section 3.3) over CeO2 support. As the catalysts were reduced in situ using
o
hydrogen, H2 is desorbed at 55-75 C (vide infra) and is responsible for the formation of
-
HCO3 .
3.1.3. H2 – TPR
The reducibility of the different catalysts was studied using H2-TPR (Fig.2B). A negative
peak, which corresponds to the decomposition of palladium hydride (PdHx) was observed
at the temperature range of 55-75oC over all the Pd-based catalysts [17, 27]. PdHx is
unstable and usually decomposes during high temperature reduction, releasing hydrogen
and subsequently producing a negative peak [40]. Alloying Pd with another metal can
reduce the solubility of H2 and consequently decrease the peak intensity of PdHx [41]. A
narrow peak observed at 70-110oC and additionally, another peak at 150-220oC can be
attributed to the reduction of PdO species over the surface, which may arise due to the
strong interaction of small Pd particles with CeO2 support [42].
11

The broad peak centered at 370oC (Zn/CeO2 sample) can be attributed to Zn-O-Ce surface
interaction, in agreement with Venkataswamy et al. [43]. At such a temperature (370oC),
the reduction of ZnO is impossible in the absence of Pd. However, it appears that the
incorporation of Pd in Zn structure is responsible for the early partial reduction of Zn which
otherwise is reported to start at 600oC [44]. In agreement with other authors, we observed
the reduction of surface oxygen of ceria at about 400-530oC while bulk ceria reduction only
occurs at about 800oC [17, 45]. Nevertheless, similar reduction peaks which are attributable
to the partial reduction of surface CeO2 are observed at 350-500oC for all the Pd:Zn samples
and are in agreement with previous reports by Zhu et al. [46]. It can be deduced from the
TPR spectra that the presence of Pd and Zn assist in decreasing the surface CeO2 reduction
temperature (also shown in Supplementary Figure SF 1). The influence of Pd in lowering
the reduction of surface oxygen of ceria is well known [42, 46]. Pd has the ability to
dissociate hydrogen at the surface and spillover H-adatoms unto CeO2 surface, facilitating
its reduction.
3.1.4. XRD
The crystallographic structure of calcined and freshly reduced catalysts was analyzed using
XRD (Fig. 3A, B). Hexagonal ZnO (101) phase observed at a diffraction angle of 42.4o in
calcined catalysts is ascribed to thermal decomposition of zinc nitrate during the
calcination step. However, no diffraction peak of PdO in the calcined catalyst was
observed, suggesting a very small crystallite size or the formation of amorphous oxide
phases. The presence of a small peak attributed to Pd2Zn (110) was observed at 49o
(specially for 1.1PdZn) [19]. Freshly reduced catalysts revealed two distinct diffraction
12

peaks at 48o (111) and 51.7o (200) respectively, corresponding to the formation of β-PdZn
tetragonal crystal phase. These diffraction peaks appeared at about 42o and 44o respectively
when Cu-kα radiation (λ = 0.0154nm) was used in previous reports [15, 18, 47].
Scherrer’s equation was employed (Supplementary Information SI-1) to estimate the
crystallite size of different phases observed during XRD analysis and are shown in the
supplementary information (SI). The crystallite size of cubic CeO2 (111) phase for freshly
reduced catalysts is in the range of 9.8-10.1nm and increases with the increasing PdZn
loading, in good agreement with the textural characterization. Crystallite size of β-PdZn phase
is in the range of 11.5-15.2nm (Table 1). The calculated XRD peaks assignment for pure ceria
and PdZn alloy are shown in the Supplementary Fig. (SF2).
3.1.4. XPS
The XPS analysis of freshly reduced catalysts reveals the surface oxidation states of Pd,
Zn, Ce and O species. Fig. 4 displays the XPS spectra of Pd (3d5/2) in the range of 335.5-
335.8. Deconvolution of these B.E. peaks shows that Pd is present as Pd0, Pd2+ and Pd4+
oxidation states (Table 2) in accordance with other reports [17, 35, 48]. The XPS analysis
of Pd/CeO2 (5wt% Pd) revealed the main peak of Pd (3d5/2) at a B.E. of 336.8eV, indicating
the presence of PdO [49] (Supplementary Fig SF3). Deconvolution of this spectra revealed
the presence of Pd0 at a B.E. of 335.2eV, which appears to be in the B.E. range for metallic
Pd (335.0 - 335.2eV) [50, 51]. However, the PdZn/CeO2 samples revealed a positive B.E.
shift of 0.4-0.8eV for Pdo oxidation state (Table 2). The shift in Pdo oxidation state is due
to the electron transfer from Pd(4d) to Zn(4d) and also hybridization of Pd(4d) and
Pd(5s,5p) hence a decrease in electronic population in Pd and increasing the B.E. of Pdo
[52] and also an indication of formation of PdZn alloy [53,54]. This finding is supported
13

by the identification of PdZn crystal phase over CeO2 according to XRD analysis. The B.E.
peak intensities of Pdo (in the form of PdZn alloy), shown in Table (2), increases with
increasing Pd:Zn molar ratio (0.7 to 1.0PdZn samples), probably due to the of disruption
of Pd and Zn interaction on exposure to air [53]. We estimated the relative amount of
surface Pdo present on the catalysts to vary from 83- 85mol% with 1.0PdZn showing the
highest concentration of Pdo species (Table 2). High concentration of surface Pdo in the
1.0PdZn catalyst may be responsible for the best catalytic performance which is explained
in the activity section.
No peak was observed at B.E. 992 and 996eV, typical of metallic Zn [55]. Meanwhile, core
level Zn 2p3/2 peak at 1022eV was identified (Supplementary Fig. SF4). Zn LMM Auger
analysis was employed to investigate the change in the chemical state of Zn since a
significant B.E. shift relative to Zn metal was expected (Supplementary Fig. SF5). The
presence of Zn Auger peak at 998eV revealed that most of the Zn present in the catalysts
exist as Zn2+ most probably as ZnO, indicating surface Zn is oxidized with the exposure in
air and a peak/shoulder at 991eV may attribute for interstitial Zn on the catalyst surface
[27, 56-58]. XPS spectra of O 1s for all catalysts revealed that main B.E. was at 529.0-
529.5eV, corresponding to oxygen bonding to cerium and a positive B.E. shift was
observed with increasing Pd:Zn molar ratio. Also, a shoulder peak at 531.5eV was
deconvoluted into peaks indicating the presence of adsorbed hydroxyl and water on the
catalyst surface respectively, which comes mainly from the atmosphere (Supplementary
Fig. SF6).
The main Ce (3d5/2) B.E. peak identified in the range of 882.6-883.3eV, corresponds to
Ce4+ (v doublet) and agrees with Matsumura et al. [35, 45]. The B.E. of Ce (3d5/2) and
14

(3d3/2) energy levels for all catalysts revealed satellite pairs of u’’’, u’’, u, v’’’, v’’ and v
peaks with a shake-up and shake down characteristics in agreement with Bêche et al. [35,
59] and are shown in Fig. 5. Pairs of Ce3+ doublets (vo, uo, u’ and v’) observed during XPS
analysis confirms the occurrence of partial reduction of ceria (Table 3).
3.1.5. EDS/SEM and STEM-EELS Analysis
EDS coupled with SEM was used to quantify the elemental analysis and distribution of
species in the catalysts and the mappings/spectra are shown in Fig. 6 and in Supplementary
Fig. (SF7- SF10). The nominal weight percentages of Pd and Zn used for catalyst
preparations closely agreed with EDX elemental quantification (Supplementary Table
ST1). The triangular phase-diagram for 1.0PdZn shown in Fig. 6 reveals the formation of
Pd and Zn elements over ceria in a heterogeneous (non-uniform) distribution but Pd and
Zn elements are in close proximity (Pd-Zn region), suggests the formation of PdZn alloy
over CeO2.
The catalyst with Pd:Zn ratio of 1:1 was also studied using a transmission electron
microscope FEI's Titan ST 300kV operated in STEM mode. Electron energy loss
spectroscopy (EELS) was used for elemental mapping of the sample. For EELS data
treatment, Gatan Digital Micrograph v.3.30.2016.0 was used. From the elemental maps for
0.7PdZn, 0.9 PdZn (in Supplementary Figure SF 11 & 12) and 1.0PdZn catalyst (Fig. 7),
it is well evidenced that with the increase of Pd loading, the homogeneity of Pd and Zn
distribution over CeO2 was increased. From the Fig. 7 (1.0 PdZn), we can see the highest
proximity of Pd and Zn among the analyzed samples, which may be caused by the Pd-Zn
15

alloy formation and showed superior activity compared to other catalysts. Also, high
carbon contamination which is observed from microscopy data can be related to the
remaining citric acid from the precursor solution.
Supplementary Fig. SF7- SF8 (d) display the mapping and phase diagram of other catalysts,
confirming the non-uniform distribution of Pd and Zn species on the CeO2 support.
However, we observed the additional presence of Zn nanoparticles (for 0.7 and 0.9 PdZn
catalyst) beyond the Pd-Zn region which indicates the presence of ZnO and agrees with
Matin [60]. Similar ZnO was also observed in XRD and XPS analyses, which may be a
cause of air oxidation of Zn.
3.2. Catalytic Activity Performance
3.2.1. Influence of reaction temperature
The influence of temperature in the range of 220-270oC (keeping the pressure at 20 bar and
GHSV at 2400h-1) on the catalytic performance of the prepared catalysts for CO2
hydrogenation to methanol was investigated and results are depicted in Fig. 8A. All the
catalysts showed a rapid increase in CO2 conversion until 250oC after that increase in
conversion was slowed down, except for 1.0PdZn catalyst for which CO2 conversion is
higher compared to other samples, showing a linear increment of activity with temperature.
o
The CO2 conversion at 250oC and 270 C, all the catalysts showed conversion values in
between 77-88% of thermodynamic equilibrium value (Supplementary Table ST2 and
Supplementary Figure SF13). The 1.0PdZn showed high activity (in all temperature range)
compared to the other three catalysts having the closest value to the thermodynamic
equilibrium conversion. The higher activity can be assigned to enhanced RWGS reaction
16

over this catalyst, showing high selectivity to CO compared to other catalysts (Fig 8B and
o
o
Supplementary Figure SF13) specially at 250 C and 270 C. Methanol selectivity was
observed to decrease with increasing reaction temperature due to the competing RWGS
reaction (Fig. 8C). These observations are convened in a methanol selectivity – CO2
conversion curve (Fig. 8D), which clearly shows that with the increase in reaction
temperature, methanol selectivity decreases with increasing CO2 conversion for all the
tested catalysts. In summary, 220oC is the best temperature to operate and 1.0PdZn catalyst
showed a CO2 conversion of 14% and maximum methanol STY of 114g/kgcat/hr with a
methanol selectivity of approximately 95% revealing a superior performance over other
catalysts reported for this reaction at similar reaction conditions (Table 4). The
performance of the best catalyst (1.0PdZn) was also investigated at low temperature and
interestingly showed 12% CO2 conversion and 98% selectivity at 200oC and 8% CO2
conversion and 100% methanol selectivity at 180oC.
3.2.2. Influence of Pd:Zn Molar Ratio
Fig. 9A shows the influence of Pd:Zn molar ratio on catalytic activity at 220 oC, 20 bar and
2400 h-1 GHSV. Methanol selectivity (Fig. 9B) was decreased with increasing Pd:Zn molar
ratio while selectivity to CO expectedly increased which suggests that increasing amount
of Pd loading enhances the RWGS reaction.
A significant enhancement in CO2 conversion and methanol space time yield (STY) were
observed as the Pd:Zn ratio is increased from 0.7 to 1.0 reaching a maximum for 1.0PdZn.
Further increase in Pd:Zn ratio revealed a notable drop in catalytic activity. Remarkably,
17

optimum STY of methanol over 1.0PdZn catalyst showed about 58% higher than 1.1PdZn
catalyst corresponding to 114 and 72g/kgcat/h respectively. Table 4 reports the activity
results of the investigated catalysts and compared other PdZn based catalysts reported in
the literature. Surface characterization of the catalysts revealed that the highest relative
content of Pdo species, present as PdZn bimetallic alloy (XRD, XPS section) in the 1.0PdZn
catalyst, which is the optimum Pd:Zn ratio over CeO2 surface for this hydrogenation
reaction for methanol synthesis.
The activity data can be well correlated to the surface basicity and relative content of Pdo
(obtained from deconvoluted data of XPS) as presented in Fig. 10. The surface basicity is
related to the CO2 uptake amount (μmol/gm) from the CO2 TPD study. Both the surface
basicity and Pdo content went through a maximum for the 1.0PdZn catalyst which also
showed the best activity of CO2 conversion and high methanol STY.
3.2.3. Stability of 1.0PdZn/CeO2 Catalyst
Time-on-stream (TOS) experiment was conducted to test the stability of the best catalyst
at the optimum reaction conditions (220oC, 20bar, 2400h-1 GHSV) and shown in Fig. 11.
Methanol selectivity reached steady state at approximately 3hrs and within the next 30hrs,
some deactivation was observed (12.5% CO2 conversion and 91.6g/kgcat/h STY) and after
that, the catalyst activity was stable for more than 100hrs. This showed approximately
20% decrease in STY compared to the initial activity. A close comparison of XRD for TOS
spent (Supplementary Fig. SF14) and fresh samples showed no formation of new phases
except an insignificant reduction in the PdZn (111) intensity peak. The crystal growth of
PdZn particles (14.3 to 15.2nm) as evidenced from XRD analysis, reduced catalytic activity
(sintering) during a long exposure in reaction condition. It appears that PdZn/CeO2 catalyst
18

is susceptible for water formation via RWGS and might significantly contribute to the
deactivation through sintering. As the activity reached equilibrium after 30hrs (Fig. 11),
the attained CO2 conversion and methanol selectivity was quite high compared to Cu-based
catalyst at this mild operating condition, which brings in to the conclusion that Pd-based
catalyst can be a potential alternative catalyst for CO2 hydrogenation to methanol.
3.2.4. Effect of Gas Hourly Space Velocity (GHSV)
Catalytic performance of the best catalyst was tested by varying GHSV from 2,400-12,000
h-1 at the optimum operating condition of 220oC and 20bar. At maximum GHSV (12,000h-
1), the STY of methanol reached 304g/kgcat/h with CO2 conversion and methanol selectivity
of 7.4 and 100% respectively (Fig. 12A). In addition, at long residence time (lowest GHSV
of 2400h-1), catalytic performance approaches thermodynamic limit revealing the highest
CO2 conversion of about 14.12% and lowest methanol STY of 114g/Kgcat/h. Although, for
commercial operation, GHSV lower than 2,400h-1 may not be practically applicable from
an economic standpoint since the methanol space time yield will be very low. Generally,
as the GHSV increases, CO2 conversion and CO selectivity decrease while methanol
selectivity and STY increase. The decrease in CO2 conversion can be attributed to a
reduced amount of adsorbed CO2 on the surface of the catalyst at low contact time (high
GHSV).
19

3.2.5. Effect of Total Pressure
The influence of system pressure (1- 40bar) on the performance of 1.0PdZn catalyst were
tested at the temperature and GHSV of 220oC and 2,400h-1 respectively (Fig. 12B). It
showed that as pressure increases from 1-20bar, CO2 conversion and methanol STY
increased and reached a maximum at 20bar. At higher pressure above 20bar, a decrease in
the activity was observed for the catalyst. This may be due to a high occupation of the
surface with H2 (with CO2:H2=1:3 feed) which may inhibit adsorption and reaction of CO2
over the catalyst surface. At 1, 5, 10 and 15bar, we have a considerable amount of
conversion with a very high selectivity to methanol. CO2 conversion and methanol
selectivity of 3% (±0.05) and 85% respectively with methanol STY (19.1g/kgcat/h) at
atmospheric pressure is superior to Iwasa et al. [15] who previously reported a methanol
selectivity of 65% at 1bar over Pd/ZnO catalyst. The activity of this catalyst at atmospheric
pressure suggests that it is feasible to synthesize high performing catalyst for low-pressure
(1-5bar) CO2 hydrogenation to methanol which can be further converted to lower
olefins/dimethyl ether in a single pass reactor.
3.3. DRIFTS Study
3.3.1. Pure CO2 adsorption on 1.0PdZn catalyst
FTIR spectra of pure CO2 adsorbed over the 1.0PdZn catalyst in the temperature range of
50-250oC are displayed in Fig. 13. A high intensity band at 2361 cm-1 was observed for the
gas phase CO2. Bands typical of symmetric and asymmetric bidentate surface carbonate
2-
(b-CO3 ) were observed at 1301-1310 and 1574 cm-1 respectively, at low-temperature
20

o
range (50-100 C) which apparently disappeared forming monodentate carbonate species
2-
(m-CO3 ) at 1078cm-1 at higher temperature [61]. IR bands of surface hydrogen carbonates
-
(HCO3 ) were identified at 1030-1040 and 1397cm-1, probably due to the interaction of
2-
adsorbed CO2 with surface Ce-OH species. It can be inferred that surface m-CO3 species
-
are formed at high temperature via decomposition of HCO3 intermediates, as also observed
in CO2 TPD study.
3.3.2. CO2/H2 adsorption on pure ceria support
The adsorption of CO2/H2 mixture over pure CeO2 support was conducted at the optimal
operating condition (220oC, 20bar) and results are depicted in two sections (Supplementary
Fig. SF15). Bands attributable to linearly adsorbed CO (COL) were identified at 2177 and
2053cm-1. Weak bands of formate species at 2848 and 2939cm-1 were observed, which may
arise due to the interaction of adsorbed CO and surface OH group, in agreement with Li et
al. [62]. In addition, different vibrational modes of formate species were also identified at
1258 and 1357cm-1, while the bands around 1476cm-1 are typical of carbonate species [63].
Band range of 1023-1062cm-1 is attributed to v(CO) stretching of methoxy species on ceria
surface containing different coordination geometry [61]. However, we observed no v(C-H)
vibrational mode of methoxy species over the pure ceria support.
3.3.3. CO2/H2 adsorption over the 1.0PdZn catalyst with temperature and pressure
variation
The effect of temperature on CO2/H2 adsorption over the 1.0PdZn catalyst in the range of
100-220oC at 20 bar pressure was studied and the spectra are depicted in Fig. 14. At a low
21

temperature of 100oC, there was no evidence of molecular adsorption/formation of
methanol. However, weak peaks of linearly adsorbed CO were observed (2076-2179cm-1),
indicating that CO formation via dissociative adsorption of CO2 is quite spontaneous over
the 1.0PdZn catalyst. At higher temperatures (≥ 150oC), bands typical of methoxy species
(2867cm-1), and gaseous methanol (2948-3000cm-1) began to form showing increasing
intensity with the increase of temperature, i.e. the IR peak intensities of methanol and
methoxy species at optimal reaction condition (220oC, 20bar) varied more than five folds
compared to the band intensities at 150oC.
The overlapping peaks in the band range of 2820-3000cm-1 can be assigned to v(C-H)
vibrational mode of methoxy species and methanol [61, 63], which were not observed over
pure CeO2 support. This indicates that the presence of Pd (its H-atom spillover ability)
assisted the formation of intermediate methoxy species needed for methanol formation.
Peaks concentrated in the band range of 1300-1640cm-1 can be ascribed to different
vibrational mode of surface mono and bidentate formate species[64-66]. Weak bands in
the frequency region of 1684-1748cm-1 are typical of molecularly adsorbed formic acid
(HCOOH) and formaldehyde (HCHO) species, in agreement with Li et al. [61].
In the pressure variation experiment (1-20 bar, 220oC), we remarkably observed significant
methanol formation rate at atmospheric pressure over the catalyst surface (Supplementary
Fig. SF16). IR peak intensity of methanol was increased with increasing pressure up to 20
bar, in line with our experimental data. The peak intensity of methanol at 20bar varies more
than three folds compared to 1bar. Also, the increasing peak intensities of linearly adsorbed
CO is an indication that dissociative adsorption of CO2 is a direct function of pressure.
22

-
Conclusively, the formation of surface carbonate (CO3 ) species is the main step for
activating CO2 which occurs at the surface oxygen vacant sites of CeO2, i.e. the formation
of Ce3+ species during catalyst reduction at high temperature. Pd performs a dual role of
activating and supplying hydrogen adspecies (Hs) necessary for CO2 hydrogenation and
accelerating the kinetics of methanol formation. Observation of surface methoxy species,
formaldehyde, and formic acid suggests two possible routes for methanol production,
simultaneously happening over the surface (Fig. 15). In the first route, hydrogenation of
monodentate carbonate to formate occurs followed by subsequent hydrogenation to
methanol via formaldehyde and methoxy intermediates while in the second route, formate
is hydrogenated to methanol via formic acid and methoxy species intermediates.
4.
Conclusion
The effect of Pd:Zn molar ratio (0.7 to 1.1) over CeO2 support for CO2 hydrogenation to
methanol has been studied. The catalyst with a 1:1 Pd:Zn molar ratio showed the optimum
o
performance under reaction conditions at 220 C, 20 bar and 2400 h-1 GHSV, with the
stoichiometric molar ratio 3:1 (H2:CO2), resulting in a methanol STY of 114g/kgcat/h.
Interestingly, this catalyst is also active under atmospheric conditions with 3% CO2
conversion, 85% methanol selectivity and 19.1g/kgcat/h methanol STY. In-depth in situ
DRIFTS analysis suggests the surface mechanism for methanol formation via two
competing surface reaction mechanism over this catalyst i.e. (1) hydrogenation of formate-
formaldehyde-methoxy species (2) hydrogenation of formate-formic acid-methoxy
species.
23

Acknowledgement
This joint project was co-founded by King Abdulaziz University (KAU), Jeddah, and King
Abdullah University of Science and Technology (KAUST), Thuwal, under grant number
“JP-19-001”. The authors, therefore, acknowledge KAU and KAUST for their technical
and financial support.
24

References
1. B.P, Statistical Review of World Energy, in, London, United Kingdom, (2018).
2. I.E. Agency, World Energy Outlook 2018 (2018).
3. G.A. Olah, Angew. Chem. Int. Ed. 44 (2005) 2636-2639.
4. O. Bičáková, P. Straka, Int. J. Hydrogen Energ. 37 (2012) 11563-11578.
5. S. Lee, A. Sardesai, Top Catal. 32 (2005) 197-207.
6. C. Baltes, S. Vukojević, F. Schüth, J. Catal. 258 (2008) 334-344.
7. T. Inui, H. Hara, T. Takeguchi, J. Kim, Catal. Today 36 (1997) 25-32.
8. M. Saito, T. Fujitani, M. Takeuchi, T. Watanabe, Appl. Catal. A Gen. 138 (1996)
311-318.
9. J.A. Rodriguez, P. Liu, D.J. Stacchiola, S.D. Senanayake, M.G. White, J.G. Chen,
ACS Catal. 5(2015) 6696-6706.
10. M. Behrens, F. Studt, I. Kasatkin, S. Kühl, M. Hävecker, F. Abild-Pedersen, S.
Zander, F. Girgsdies, P. Kurr, B.L. Kniep, M. Tovar, Science 336(2012) 893-897.
11. S. Kattel, P.J. Ramírez, J.G. Chen, J.A. Rodriguez, P. Liu, Sciences 355(2017)
1296-1299.
12. S.A. Kondrat, P.J. Smith, P.P. Wells, P.A. Chater, J.H. Carter, D.J. Morgan, E.M.
Fiordaliso, J.B. Wagner, T.E. Davies, L. Lu, J.K. Bartley, Nature 531(2016) 83.
13. J. Xu, X. Su, X. Liu, X. Pan, G. Pei, Y. Huang, X. Wang, T. Zhang, H. Geng,
Appl. Catal. A Gen. 514 (2016) 51-59.
25

14. M. Neumann, D. Teschner, A. Knop-Gericke, W. Reschetilowski, M. Armbrüster,
J. Catal. 340 (2016) 49-59.
15. N. Iwasa, H. Suzuki, M. Terashita, M. Arai, N. Takezawa, Catal. Lett. 96 (2004)
75-78.
16. N. Iwasa, T. Mayanagi, N. Ogawa, K. Sakata, N. Takezawa, Catal. Lett. 54
(1998) 119-123.
17. A.S. Malik, S.F. Zaman, A.A. Al-Zahrani, M.A. Daous, H. Driss, L.A. Petrov,
Appl. Catal. A Gen. 560 (2018) 42-53.
18. H. Bahruji, M. Bowker, G. Hutchings, N. Dimitratos, P. Wells, E. Gibson, W.
Jones, C. Brookes, D. Morgan, G. Lalev, J. Catal. 343 (2016) 133-146.
19. A. Ota, E.L. Kunkes, I. Kasatkin, E. Groppo, D. Ferri, B. Poceiro, R.M.N. Yerga,
M. Behrens, J. Catal. 293 (2012) 27-38.
20. S.E. Collins, J.J. Delgado, C. Mira, J.J. Calvino, S. Bernal, D.L. Chiavassa, M.A.
Baltanás, A. Bonivardi, J. Catal. 292 (2012) 90-98.
21. A.S. García-Trenco, E.R. White, A. Regoutz, D.J. Payne, M.S. Shaffer, C.K.
Williams, ACS Catal. 7 (2017) 1186-1196.
22. C.D. Qin, C. Yu, D. Ng, L. Lim, M. Lai, Mater. Lett. 26 (1996) 17-19.
23. N. Iwasa, N. Takezawa, Top. In Catal. 22 (2003) 215-224.
24. M. Armbrüster, M. Behrens, K. Föttinger, M. Friedrich, É. Gaudry, S.K. Matam,
H.R. Sharma, Catal. Rev. 55(2013) 289-367.
25. M. Friedrich, S. Penner, M. Heggen, M. Armbrüster, Angew. Chem. Int. Ed. 52
(2013) 4389-4392.
26

26. G. Xia, J.D. Holladay, R.A. Dagle, E.O. Jones, Y. Wang, Chemical Engineering
& Technology: Industrial Chemistry‐Plant Equipment‐Process Engineering‐
Biotechnology 28 (2005) 515-519.
27. H. Bahruji, M. Bowker, W. Jones, J. Hayward, J.R. Esquius, D. Morgan, G.
Hutchings, Faraday Discuss. 197 (2017) 309-324.
28. A. Baiker, D. Gasser, J. Chem. Soc. Faraday Trans. 85 (1989) 999-1007.
29. N. Koizumi, X. Jiang, J. Kugai, Catal. Today 194 (2012) 16-24.
30. A.S. Malik, S.F. Zaman, A.A. Al-Zahrani, M.A. Daous, H. Driss, L.A. Petrov,
Catal. Today (2019), Article in press,
https://doi.org/10.1016/j.cattod.2019.05.040.
31. C. de Leitenburg, A. Trovarelli, J. Kašpar, J. Catal. 166 (1997) 98-107.
32. E.J. Choi, Y.H. Lee, D.W. Lee, D.J. Moon, K.-Y. Lee, J. Mol. Chem. 434 (2017)
146-153.
33. J. Kugai, E.B. Fox, C. Song, Appl. Catal. A Gen. 456 (2013) 204-214.
34. M. Cauqui, J. Rodriguez-Izquierdo, J. Non-Cryst. Solids 147 (1992) 724-738.
35. J.F. Moulder et. al., Handbook of X-ray photoelectron spectroscopy. Perkin-
Elmer Corporation, United State of America, 1992.
36. K.S. Sing, R.T. Williams, Adsorpt. Sci. & Technol. 22 (2004) 773-782.
37. M.F. De Lange, T.J. Vlugt, J. Gascon, F. Kapteijn, Microporous
Mesoporous Mater. 200 (2014) 199-215.
38. L. Fan, K. Fujimoto, J. Catal. 172 (1997) 238-242.
39. K. Pokrovski, K.T. Jung, A.T. Bell, Langmuir 17 (2001) 4297-4303.
27

40. N. Iwasa, S. Masuda, N. Ogawa, N. Takezawa, Appl. Catal. A Gen. 125 (1995)
145-157.
41. G. Budroni, S.A. Kondrat, S.H. Taylor, D.J. Morgan, A.F. Carley, P.B. Williams,
G.J. Hutchings, Catal. Sci. & Tech. 3 (2013) 2746-2754.
42. W.J. Shen, Y. Ichihashi, H. Ando, Y. Matsumura, M. Okumura, M. Haruta, Appl.
Catal. A. Gen. 217 (2001) 231-239.
43. P. Venkataswamy, D. Devaiah, D. Mukherjee, M. Vithal, B.M. Reddy, Catalysis
in Green Chemistry and Engineering 1 (2018)
44. W. Yanhua, J. Zhang, X. Hengyong, X. Bai, Chinese J. Catal. 28 (2007) 234-238.
45. Y. Matsumura, W.J. Shen, Y. Ichihashi, M. Okumura, J. Catal. 197 (2001) 267-
272.
46. H. Zhu, Z. Qin, W. Shan, W. Shen, J. Wang, J. Catal. 225 (2004) 267-277.
47. E. Moretti, M. Lenarda, L. Storaro, R. Frattini, P. Patrono, F. Pinzari, J. Colloid
Interphase Sci. 306 (2007) 89-95.
48. X.L. Liang, X. Dong, G.D. Lin, H.B. Zhang, Appl. Catal. B Environ. 88 (2009)
315-322.
49. E.B. Fox, A.F. Lee, K. Wilson, C. Song, Top in Catal. 49 (2008) 89-96.
50. A. Bayer, K. Flechtner, R. Denecke, H.P. Steinruck, K.M. Neyman, N. Rosch,
Surf. Sci. 600 (2006) 78–94.
51. Y. Usami, K. Kagawa, M. Kawazoe, Y. Matsumura, H. Sakurai, M. Haruta, Appl.
Catal. A 171 (1998) 123-130 .
52. J.A. Rodrigues, J. Phys. Chem. 98 (1994) 5758-5764.
53. P.S. Wehner, G.C. Tustin, B.L. Gustafson, J Catal., 88 (1984) 246-248.
28

54. K. Dumbuya, R. Denecke, H.P. Steinrück, Appl. Catal. A Gen 348 (2008) 209-
213.
55. W. Li, L. Fang, G. Qin, H. Ruan, H. Zhang, C. Kong, L. Ye, P. Zhang, F. Wu, J.
Appl. Phy. 117 (2015) 145301.
56. F. Liao, X.P. Wu, J. Zheng, M.M. J. Li, A. Kroner, Z. Zeng, X. Hong, Y. Yuan,
X.Q. Gong, S.C.E. Tsang, Green Chem. 19 (2017) 270-280.
57. H. Bahruji, J.R. Esquius, M. Bowker, G. Hutchings, R.D. Armstrong, W. Jones,
Top. Catal. 61 (2018) 144-153.
58. I.S. Mashkovsky, P.V. Markov, G.O. Bragina, G.N. Baeva, A.V. Rassolov, A.Y.
Stakheev, A.V. Bukhtiyarov, I.P. Prosvirin, V.I. Bukhtiyarov, Mendeleev
Commun., 28 (2018) 152-154.
59. E. Bêche, P. Charvin, D. Perarnau, S. Abanades, G. Flamant, Surf. Interface Anal.
40 (2008) 264-267.
60. M. A. Matin, A. Kumar, R.R. Bhosale, M.S. Saad, F. A. Almomani, M.J. Al-
Marri, RCS Adv. 7 (2017) 42709-42717.
61. C. Li, K. Domen, K. Maruya, T. Onishi, J. Catal. 125 (1990) 445-455.
62. C. Li, Y. Sakata, T. Arai, K. Domen, K. Maruya, T. Onishi,
J. Chem. Soc., Faraday Trans.1 85 (1989) 929-943
63. T. Fujitani, I. Nakamura, Bull. Chem. Soc. Jpn. 75 (2002) 1393-1398.
64. C. Schild, A. Wokaun, A. Baiker, J. Mol. Catal. 63 (1990) 223-242.
65. A.S. Malik, S.F. Zaman, A.A. Al-Zahrani, M.A. Daous, H. Driss, L.A. Petrov,
Bulg. Chem. Commun. 50 (2018) 189-195.
66. S.E. Collins, M.A. Baltanás, A.L. Bonivardi, J. Catal. 226 (2004) 410-421.
29