Fuel 325 (2022) 124833

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Highly efficient propane dehydrogenation promoted by reverse water–gas

shift reaction on Pt-Zn alloy surfaces
Chadatip Rodaum a, Peeranat Chaipornchalerm a, Watinee Nunthakitgoson a,
Anawat Thivasasith a, Thana Maihom b, Thassanant Atithep c, Pinit Kidkhunthod d,
Chayapat Uthayopas e, Sarana Nutanong e, Sutarat Thongratkaew f, Kajornsak Faungnawakij f,
Chularat Wattanakit a, *
a
Department of Chemical and Biomolecular Engineering, School of Energy Science and Engineering, Vidyasirimedhi Institute of Science and Technology, Rayong 21210,
Thailand
b
Department of Chemistry, Faculty of Liberal Arts and Science, Kasetsart University, Kamphaeng Saen Campus, Nakhon Pathom 73140, Thailand
c
Frontier Research Center (FRC), Vidyasirimedhi Institute of Science and Technology, Rayong 21210, Thailand
d
Synchrotron Light Research Institute (Public Organization), 111 University Avenue, Muang District, Nakhon Ratchasima, Thailand
e
School of Information Science and Technology, Vidyasirimedhi Institute of Science and Technology, Rayong 21210, Thailand
f
National Nanotechnology Center (NANOTEC), National Science and Technology Development Agency (NSTDA), Pathum Thani 12120, Thailand

A R T I C L E I N F O A B S T R A C T

Keywords: CO2-assisted propane dehydrogenation has been developed in propylene production technology to deal with
CO2 thermodynamic equilibrium limitations. However, the rational design of catalysts is a crucial challenge in
Propane dehydrogenation achieving high catalytic performances. Herein, we report the successful fabrication of highly dispersed PtZn alloy
PtZn alloy nanoparticles
on hierarchical zeolites for CO2-assisted propane dehydrogenation through the reverse water–gas shift reaction
RWGS
(RWGS). Compared with monometallic Pt, the synergistic effect of PtZn plays a crucial role in both direct pro­
pane dehydrogenation and RWGS, resulting in an outstanding catalytic performance with a high turnover fre­
quency (TOF) of 1.82 × 104h− 1. From the catalytic point of view, Pt-Zn alloy surfaces facilitate the equilibrium
shift in propane conversion by consuming produced H2 through RWGS with weakening CO-metal surface
interaction revealed by operando studies and DFT calculations. These findings illustrate the conceptual design of
alloy catalysts and allow insights into the mechanistic details of CO2-assisted alkane dehydrogenation.

1. Introduction It is well-known that the commercial PDH processes such as Catofin

and Oleflex processesTM developed by CB&I Lummus and UOP, respec­
Propylene is one of the essential building blocks for several potential tively [5,6] provide high propylene selectivity. In principle, it still suf­
applications ranging from polypropylene to acrylic acid and propylene fers from an equilibrium conversion limitation due to its reversible
oxide [1,2]. Typically, it has been mainly produced by steam cracking reaction feature (C3H8 ⇌ C3H6 + H2) [7]. To circumvent this issue, one
and fluid catalytic cracking (FCC) processes [3]. However, there are of the most exciting approaches to enhance the propylene yield has been
often some drawbacks, for example, providing low propylene selectivity investigated by shifting the corresponding equilibrium conversion. The
due to a broad range of product distribution and operating at a high produced hydrogen from the PDH reaction could be consumed, even­
reaction temperature, eventually leading to increased energy con­ tually driving the forward reaction. Indeed, by integrating with CO2
sumption and the release of a massive amount of greenhouse gases. To reduction, the propane conversion could be improved remarkably
further develop the propylene production process, propane dehydroge­ because the produced hydrogen consumption through reverse water gas
nation (PDH) has also attracted much attention because it can directly shift (RWGS; CO2 + H2 ⇌ CO + H2O) facilitates the forward reaction
convert propane into propylene, resulting in high propylene selectivity [8–11]. Most importantly, carbon dioxide utilization during the PDH
when using lower temperatures than the traditional propylene produc­ process enhances propane conversion and is one of the most attractive
tion processes [4]. routes to reduce greenhouse gas emissions. It is therefore of great

* Corresponding author.
E-mail address: Chularat.W@vistec.ac.th (C. Wattanakit).

https://doi.org/10.1016/j.fuel.2022.124833
Received 23 March 2022; Received in revised form 27 May 2022; Accepted 7 June 2022
Available online 14 June 2022
0016-2361/© 2022 Elsevier Ltd. All rights reserved.
C. Rodaum et al. Fuel 325 (2022) 124833

interest to rationally develop a promising catalyst that can be applied propane dehydrogenation via RWGS reaction.
simultaneously in both propane dehydrogenation and RWGS reactions.
To date, a noble metal such as platinum (Pt) is one of the most 2. Materials and methods
effective catalysts for direct propane dehydrogenation [12]; however, it
often suffers from low propylene selectivity owing to the too strong 2.1. Materials
adsorption of propylene on Pt surfaces, enabling side reactions including
cracking, hydrogenolysis, and oligomerization-isomerization [13]. Tetraethyl orthosilicate (TEOS, ≥99.0 %, Sigma-Aldrich); sodium
Indeed, the propylene adsorption on Pt surfaces would be weakened hydroxide (NaOH, Carlo Erba, ≥98.0 %); tetra-n-butylammonium hy­
when combining with a promoter, such as Sn, Cu, and Zn, because the droxide (TBAOH: 40 % in aqueous solution, LEONID, Bengaluru); tet­
adsorbate–metal surface interaction is reduced for the alloy surfaces rapropylammonium hydroxide (TPAOH, 1.0 M in H2O, Sigma-Aldrich);
[14–19]. Although alloy catalysts, especially PtSn and PtZn, have been tetraamineplatinum (ii) nitrate (Sigma-Aldrich, ≥50.0 %); zinc nitrate
extensively applied in the PDH process, the high yield of propylene still hexahydrate (Zn(NO3)2⋅6H2O, 98 %, Sigma-Aldrich) were used as
suffers because the excellent propane conversion has never been suc­ received without further purification.
cessful at the low reaction temperature, while the formed byproducts are
unavoidable at high reaction temperature even though the significantly 2.2. Synthesis of silicalite-1 nanosheet (SiNS)
improved conversion could be obtained [20,21].
Several metallic- and bimetallic catalysts, such as Pd, [9] Pt, [22] Co, The silicalite-1 nanosheet has been synthesized by a simple hydro­
[23] Fe, [24] Ni, [22], and Zn [25], have been developed to improve the thermal process with the molar gel composition of 60SiO2: 18TBAOH:
propylene yield through the above-mentioned CO2-assisted dehydroge­ 0.75NaOH: 600H2O. In the typical procedure, the solution containing
nation of propane [26]. In particular, a Zn-modified-based catalyst has 0.02 g of sodium hydroxide and 2.32 g of deionized water (DI water) was
been reported as an interesting active catalyst in the RWGS reaction dropped into a mixture of 8.11 g of TBAOH and 8.67 g of TEOS. Sub­
because it can enhance the formation of oxygen vacancy defects, which sequently, the obtained mixture was stirred at ambient temperature for
are active sites for CO2 chemisorption [27–30]. Consequently, the 12 h, and the homogeneous gel was then transferred to a Teflon-lined
combination of platinum and zinc as a bimetallic catalyst would be not stainless-steel autoclave and heated to 130 ◦ C for 2 days. Afterward,
only a fascinating way to promote both PDH and RWGS but also increase the solid product was collected and washed with DI water until the
the propylene selectivity by diminishing the interaction between pro­ neutral pH of the filtrate was reached. Subsequently, the as-synthesized
pylene and Pt surfaces. Besides, the development of a robust catalyst sample was dried at 100 ◦ C overnight and then calcined to remove SDAs
against CO poisoning, generated from RWGS during the CO2-assisted at 550 ◦ C for 8 h. It should be noted that the obtained sample was
PDH reaction, is a main important issue. Therefore, the design of denoted as SiNS.
bimetallic or alloy catalysts in CO2-assisted PDH reaction to prevent the
strong interaction of CO adsorbed on metal surfaces providing the high 2.3. Synthesis of conventional silicalite-1 (SiCON)
efficiency for propylene production has been a great challenging task.
Apart from the nature of active metals, the lack of confinement of a The conventional silicalite-1 has been synthesized by a hydrothermal
catalyst framework support is a vital difficulty in achieving high cata­ process with the molar gel composition of 10SiO2: 1TPAOH: 1.03NaOH:
lytic performance, including fast deactivation and metal sintering of 400H2O. In a typical procedure, the sodium hydroxide mixed with DI
catalysts during high-temperature reactions. To deal with this issue, a water was added into the prepared silica solution, containing 7 g of
catalyst support with unique confinement structures and high surface TEOS and 3.42 g of TPAOH. Subsequently, the solution was stirred at
areas such as a zeolite has been applied in the PDH reaction [31–34]. ambient temperature for 2 h to obtain the homogeneous gel, and it was
Recently, a hierarchical zeolite containing both micropores and meso­ transferred to a Teflon-lined stainless-steel autoclave and heated to
pores can significantly promote the monometallic nanoparticle disper­ 180 ◦ C for 3 days. The sample was collected and washed with DI water
sion, eventually improving the PDH catalytic efficiency. For example, until the pH of the filtrate was neutral. Subsequently, the sample was
the recently developed hierarchical silicalite-1 nanosheet not only im­ dried and calcined at 550 ◦ C for 8 h. The final sample was denoted as
proves the dispersion of an active metal but also stabilizes a metal on SiCON.
solid support due to the abundance of silanol surfaces [35,36]. There­
fore, it would also be advantageous to use a hierarchical zeolite as solid 2.4. Preparation of zinc-platinum supported on silicalite-1 nanosheet
support for active metals to catalyze the CO2-assisted dehydrogenation (SiNS), conventional silicalite-1 (SiCON), aluminium oxide (Al2O3), and
of propane. cerium oxide (CeO2)
As mentioned above, the integrated aspects of hierarchical struc­
tures, a synergistic effect of active alloy, and using CO2 as a co-feed The zinc-platinum supported on SiNS, SiCON, Al2O3, and CeO2
promoting the PDH conversion would be the effective way to signifi­ samples were prepared by a traditional impregnation method. Gener­
cantly enhance the catalytic activity because a hierarchical zeolite ally, catalyst supports (0.5 g) were dispersed in the solution containing
would promote the high dispersion of alloy nanoparticles, which stim­ the desired amount of zinc nitrate in 10 mL of DI water, followed by
ulates the increased conversion of propane, while the appropriate pro­ stirring for 6 h. After that, the solid product was collected using a rotary
pylene selectivity is preserved. In this contribution, we demonstrated evaporator and calcined at 550 ◦ C for 4 h. Subsequently, the obtained
the rational design of Pt-Zn alloy nanoparticles supported on silicalite-1 sample was added to the prepared solution of tetraamineplatinum(ii)
nanosheets as promising catalysts for propane dehydrogenation in the nitrate (1 wt% of Pt was fixed), followed by stirring for 24 h. The final
presence of CO2 as co-feed. Outstandingly, the Pt-Zn catalyst shows a sample was obtained using a rotary evaporator for solvent removal and
significant improvement in propylene production in CO2-assisted pro­ then calcined for 4 h. The obtained catalysts were denoted as 1Pt-xZn/
pane dehydrogenation reaction compared with the isolated Pt, in SiNS and 1Pt-xZn/SiCON, 1Pt-xZn/Al2O3, and 1Pt-xZn/CeO2, where x
particular, in the presence of CO2. In addition, several independent represents weight percent (wt%) of zinc.
techniques, including transmission electron microscopy (TEM), X-ray
photoelectron spectroscopy (XPS), X-ray absorption spectroscopy (XAS), 2.5. CO2-assisted propane dehydrogenation via the RWGS reaction
in-situ diffuse reflectance infrared fourier transform spectroscopy (in-situ
DRIFTS), and density functional theory (DFT) calculations have been The CO2-assisted propane dehydrogenation via the RWGS reaction
applied to gain insights into the detail regarding the role of Pt-Zn active was carried out in a fixed-bed reactor at 600 ◦ C under atmospheric
site structures and the mechanistic point of view on the CO2-assisted pressure. A 0.5 g of catalyst (0.2–0.4 mm of particle size) was placed into

2
C. Rodaum et al. Fuel 325 (2022) 124833

the reactor with ½ inch of diameter size. Before the reaction progress, where ECO/slab is the total energy of the adsorbed CO molecule on the
the catalyst was pretreated in hydrogen flow (3 % of H2 in He) of 40 mL metal slab model, ECO is the energy of an isolated CO molecule, and Eslab
min− 1 at 600 ◦ C for 2 h. Then the reactant (20 mL min− 1) containing is the energy of the optimized clean metal surface slab model.
propane and CO2 with a 1:1 vol ratio was introduced into the reactor at
600 ◦ C. The reaction products were analyzed using an on-line gas
chromatograph (Agilent 7890B GC) equipped with two detectors, 2.8. Catalyst characterizations
including Thermal Conductivity Detector (TCD) and Flame Ionization
Detector (FID). In addition, the standard error of the mean (s.e.m) of All X-ray powder diffraction (XRD) patterns were recorded on a
propane conversion determined by Eq. (1) was 0.9 %, and the carbon Bruker D8 ADVANCE model using Cu Kα as the radiation source with
mass balance was in the range of 97.2 ± 1.5 %. Moreover, the turnover 0.02◦ of step sizes and 2θ in the range of 5◦ to 60◦ . X-ray Fluorescence
frequency (TOF) and propylene formation rate were determined using (XRF) instrument was used to determine the catalyst composition per­
Eq. (2) and Eq. (3), respectively. forming on a Bruker AXS S8 Tiger Pioneer using an Rh X-ray tube with a
s 75 µm Be window.
s.e.m = √̅̅̅ (1) The catalyst morphology was observed using the JEOL JSM-7610F
n
microscope with 5 kV of an accelerating voltage applied in scanning
where s and n refer to standard deviation and the number of experi­ electron microscopy (SEM). Moreover, Transmission Electron Micro­
ments, respectively. scopy (TEM), Scanning Transmission Electron Microscopy (STEM), and
Energy Dispersive X-ray Spectroscopy (EDS) elemental mapping images
mole of converted propane
TOFpropane = (2) were obtained using the JEOL JEM-ARM 200F microscope at 200 kV.
(mole of Pt metal) × Dis. × (unit time) To obtain the textural properties of all the synthesized samples, N2
adsorption–desorption experiments were carried out using a Micro­
where Dis. refers to the dispersion of Pt metal obtained from the H2-
meritics 3-Flex instrument at − 196 ◦ C. Typically, the samples were
pulse chemisorption experiment.
degassed at 300 ◦ C for 24 h under vacuum conditions before measure­
mole of propylene formation ment. The Brunauer–Emmett–Teller (BET) equation was applied to
r(C3 H6 ) = (3)
(weight of Pt) × (unit time) calculate the total specific surface area (SBET). The Vtotal, total pore
volume, was determined at the relative pressure of 0.99 from the ni­
2.6. In-situ diffuse reflectance infrared Fourier transform spectroscopy trogen adsorbed volume and the micropore area (Smicro), the external
(DRIFTS) study for CO2-assisted propane dehydrogenation via the RWGS surface area (Sext), and the micropore volume (Vmicro) were computed by
reaction the t-plot method.
An ex-situ X-ray photoelectron spectroscopy (XPS) was performed on
The in-situ DRIFTS study for CO2-assisted propane dehydrogenation a JEOL JPS-9010 equipped with nonmonochromatic Al Kα X-rays
via the RWGS reaction was performed on a Bruker Invenio R spec­ (1486.6 eV). The ex-situ XANE and EXAFS experiments were carried out
trometer using Mercury-Cadmium-Telluride (MCT) detector in the range at SUT-NANOTEC-SLRI (BL5.2) of the Synchrotron Light Research
of 400–4000 cm− 1 with a scanning resolution of 4 cm− 1 and scan Institute (SLRI), Thailand. The platinum L3 edge and zinc K edge spectra
numbers of 64. Initially, the catalyst was placed into a DRIFTS cell. After were recorded in fluorescence mode.
the sample pretreatment at 600 ◦ C under H2 flow (10 mLmin− 1) for 2 h, The acid and base properties were measured by ammonia
the mixture of propane and CO2 with a 1:1 vol ratio was introduced into temperature-programmed desorption (NH3-TPD), and carbon dioxide
the DRIFTS cell at 550 ◦ C for 5 min. Subsequently, the reaction was temperature-programmed desorption (CO2-TPD), respectively, detected
paused, and FTIR spectra were recorded at 40 ◦ C under N2 flow. The by thermal conductivity detector (TCD) performed on a BELCAT II
reaction was then continued with the same procedure, and the spectra analyzer. The pretreatment of samples was carried out before the mea­
were collected at a time on stream (TOS) of 5, 10, 15, 30, and 60 min. surement at 600 ◦ C for 1 h under 30 mLmin− 1 of H2 flow. Subsequently,
The spectra analysis was performed by subtracting the observed spectra ammonia or carbon dioxide gas was adsorbed for 30 min at 100 ◦ C.
with the reference spectrum of a fresh zeolite after pretreatment. Afterward, a system was heated from 100 to 800 ◦ C using a 10 ◦ C min− 1
of ramp rate to generate ammonia or carbon dioxide desorption profiles.
Diffuse reflectance infrared Fourier transform spectra (DRIFTS) of pyr­
2.7. Computational study idine adsorption were investigated to identify the acid types by using
MCT detector on a Bruker Invenio R instrument, and the spectra were
All calculations were performed using the density functional theory obtained at 40 ◦ C. Prior to measurement, the synthesized catalyst was
(DFT) as implemented in the Vienna Ab-initio Simulation Package pretreated under H2 flow for 1 h at 600 ◦ C. After that, pyridine was
(VASP) [37,38]. The projector augmented wave (PAW) method [39] and adsorbed at 40 ◦ C for 1 h under its vapor pressure, and the physisorbed
Perdew − Burke − Ernzerhof (PBE) exchange–correlation functional pyridine on catalyst surfaces was removed under vacuum conditions at
[40] were used for all calculations. To describe the long-range van der 40 ◦ C for 1 h. Then the sample was heated to 150 ◦ C. Subsequently, the
Waals (vdW) interactions, Gimme’s scheme (DFT-D3) [41] correction corresponding FTIR spectra were recorded with a scanning resolution of
method was used. The valence electron wavefunctions were expanded 4 cm− 1 and scan numbers of 64.
by a plane-wave basis set with a kinetic energy cutoff of 400 eV. For In terms of the H2 pulse chemisorption experiment, the ChemStar
geometry optimizations, the convergence thresholds were set to be 10− 5 TPxTM chemisorption analyzer was used for H2 pulse chemisorption
eV and 1 × 10− 4 eV/Å for total energy and ionic force, respectively. The experiments. Prior to pulse chemisorption analysis, all samples were
Brillouin zone was sampled using a 3 × 3 × 1 Monkhorst–Pack k grid. A pretreated with 50 mL min− 1 of pure O2 for 20 min and 50 mL min− 1 of
four-layer of 3 × 3 supercell and a four-layer of 4 × 4 supercell were H2 in Ar (3 % v/v) for 30 min at 600 ◦ C. After that, the samples were
applied for Pt (1 1 1) and Pt-Zn (1 1 0), respectively. Moreover, the first cooled down to 50 ◦ C, and then the H2 pulse chemisorption measure­
layer of the slab was allowed to relax while the rest was fixed. The ment was carried out under 50 mL min− 1 of H2 in Ar (5 % v/v) with
vacuum space of 20 Å was set to avoid interaction along the z-direction 0.015 cm3 of the volume of the injection loop.
between periodic images. The adsorption energy (Eads) was calculated Carbon monoxide temperature-programmed desorption (CO-TPD)
using the following equation: was conducted using the same instrument (BELCAT II analyzer) as the
Eads = ECO/slab − (ECO + Eslab ) (1) above-mentioned NH3-TPD and CO2-TPD measurement to investigate
the interaction between CO and catalyst/support surfaces. The

3
C. Rodaum et al. Fuel 325 (2022) 124833

pretreatment of samples was operated before the measurement at 600 ◦ C 3. Results and discussion
for 1 h under 30 mLmin− 1 of H2 flow. After the sample was cooled to the
temperature of 40 ◦ C, the CO in Ar (10 %v/v) was introduced at 50 mL/ 3.1. Characterization of catalysts
min through the sample surface for 60 min. Finally, it was heated from
100 to 400 ◦ C using a 10 ◦ C min− 1 of heating rate to generate ammonia The crystalline structures and morphologies of the synthesized
or carbon monoxide desorption profiles. silicalite-1 nanosheet (SiNS) are confirmed by XRD patterns as well as
To investigate the DRIFTS spectra of CO adsorption, the samples SEM and TEM images, as illustrated in Figs. S1 and S2, respectively. The
were pretreated under a similar condition to the above-mentioned characteristic diffraction peaks belonging to the MFI framework can be
pyridine adsorption experiment to clean the surface of the sample. observed at 2θ of 7.9, 8.8, 23.1, and 23.9◦ , corresponding to the
Then the CO was adsorbed on catalyst surfaces performed on a Bruker reflection planes of (1 0 1), (1 1 1), (5 0 1), and (3 0 3), respectively [42].
Invenio R instrument. After the adsorption of CO gas, the CO molecules The morphology of the synthesized SiNS (Fig. S2) relates to nanosheet-
in the gas phase and the physisorbed CO were removed for 30 min at like crystals with a uniform particle size of 197 ± 22 nm. After intro­
40 ◦ C. Subsequently, N2 (10 mL min− 1) was introduced through the ducing metals (Pt and Zn) to SiNS, the sharp XRD peaks corresponding to
samples at 100 ◦ C, and the corresponding FTIR spectra were recorded at the MFI structure of all the prepared catalysts were observed, implying
100 ◦ C at 60 s of interval time with a scanning resolution of 4 cm− 1 and that zeolite structures were not destroyed when using a traditional
scan numbers of 64. impregnation method to prepare supported catalysts. Moreover, the
characteristic peak of platinum cannot be observed in the case of the
1Pt/SiNS sample due to the high dispersion of Pt particles over SiNS
with a small uniform size of 2.6 ± 1.2 nm as can be seen in TEM and

Fig. 1. (A, D, G) TEM images, (B, E, H) HAADF-STEM images, and (C, F, I) alloy particle size distribution of 1Pt-0.5Zn/SiNS, 1Pt-1Zn/SiNS and 1Pt-2Zn/SiNS,
respectively.

4
C. Rodaum et al. Fuel 325 (2022) 124833

high-angle annular dark-field scanning transmission electron micro­ spectra of pyridine adsorption experiments were carried out. As can be
scope (HAADF-STEM) images in Fig. S3. seen in IR spectra results (Fig. S7B), infrared bands at 1445 and 1600
As shown in Fig. 1, Pt-Zn alloy nanoparticles have been successfully cm− 1 ascribed to strong pyridine adsorbed on Lewis acid sites [45] are
loaded on the SiNS support with the homogeneous dispersion of ul­ observed on samples containing metal species. In addition, an increase
trasmall particles in the range of 3.5 to 5.3 nm. However, a slightly in Lewis acid band intensity when increasing the zinc content from 0.5
increasing particle size trend when zinc content rises can be observed wt% to 2.0 wt% was confirmed. In strong contrast to this, no IR band
from 3.5 nm to 5.3 nm for 0.5 wt% to 2.0 wt% of zinc, respectively. corresponding to Lewis acid sites is detected in the case of the bare SiNS.
Although metals/metal oxides are located on the external surface of It is, therefore, reasonable to assume that Lewis acid sites are generated
zeolite, as can be seen from their large particle size with respect to the from metal species.
pore size of the MFI framework, using the SiNS as support provides
several benefits compared to the conventional silicalite-1. To investigate 3.2. Structural analysis of Pt-Zn alloy species
the advantage of using silicalite-1 nanosheets as catalyst supports, the
Pt-Zn with 1 wt% of each on conventional silicalite-1 (1Pt-1Zn/SiCON) To further study the chemical state of Pt and Zn in alloy catalysts, X-
was also prepared by a typical impregnation method. It was found that ray photoelectron spectroscopy (XPS), X-ray absorption near edge
the particle size of active metals supported on SiCON (Fig. S4) was structure (XANES), and extended X-ray absorption fine structure
almost 13 times larger than that of the one supported on SiNS. These (EXAFS) spectra of the prepared catalysts were observed. It should be
observations confirm that using SiNS as catalyst support is beneficial for noted that all the prepared catalysts were ex-situ pretreated with H2
metal dispersion, eventually preventing metal aggregation. before measurement. As illustrated in Fig. 2, the peak of the isolated Pt
Regarding textural properties, N2 adsorption–desorption isotherms at 74.10 eV and 71.14 eV assigned to the metallic platinum (Pt0) of 4f5/2
of the SiNS support (Fig. S5A) and all the prepared Pt-Zn alloy supported and 4f7/2, respectively, [46,47] was predominantly observed with the
on SiNS (Fig. S5B) correspond to types I and IV, indicating that they minor combination of Pt2+ appeared at 75.20 eV and 72.30 eV, repre­
contain both characteristics of microporous and mesoporous structures. senting to 4f5/2, and 4f7/2 of Pt2+, respectively [48]. Moreover, the XPS
In contrast, the SiCON support reveals only the microporous structure, spectra of all the prepared samples contain both metallic platinum and
as can be seen as the type I isotherm (Fig. S5A). The textural properties platinum(II) oxide species due to some partial oxidation during the
of all the synthesized catalysts are summarized in Table S1. The total catalyst preparation. Noticeably, after the addition of zinc, the peak
specific surface area (SBET) of all the prepared Pt-Zn alloys supported on position of 4f7/2 of Pt0 in 1Pt-0.5Zn/SiNS, 1Pt-1Zn/SiNS, and 1Pt-2Zn/
SiNS is lower than that of the bare SiNS support due to the deposition of SiNS slightly shifts toward lower binding energy at 70.88, 70.89, and
metal/metal oxides on zeolite surfaces. However, their SBET values are 70.90 eV, respectively compared with the isolated platinum (71.14 eV),
significantly higher than those of the Pt-Zn/SiCON. These behaviors indicating the electron transfer from zinc to the platinum species
again confirm the benefit of using the synthesized silicalite-1 nanosheet forming Pt-Zn alloy [49].
for improving the dispersion of Pt-Zn nanoparticles. Moreover, the Zn LMM spectrum was recorded to reveal the existing
To ensure high metal dispersion on silicalite-1 surfaces, H2 pulse states of Zn species, as illustrated in Fig. 2E. The possible zinc species is
chemisorption experiments were performed. As shown in Table S1, the in the Zn2+ form, confirmed by the peak position of 988.8 eV in the case
1Pt-1Zn supported on conventional silicalite-1(SiCON) exhibits a very of the isolated zinc. These observations relate to the fact that zinc could
low metal dispersion (0.1 %), which is 27-fold less than that of the 1Pt- be partially oxidized by air during the preparation of samples before XPS
1Zn/SiNS. The results are consistent with the TEM images of the measurement. In addition, the zinc species is probably confined to the
1Pt1Zn/SiCON, which show a large particle size caused by a high metal zeolite structure and interacts with SiO2, leading to the formation of the
aggregation with respect to the 1Pt-1Zn/SiNS. These observations Zn2+ species [49]. There is a downward shift of the kinetic energy for Pt-
confirm that using SiNS as catalyst support can prevent metal aggrega­ Zn alloy samples, implying that zinc transfers electrons to Pt. Besides,
tion, eventually improving the metal dispersion compared to using a the partial presence of metallic zinc can be observed at the peak of 990
conventional one as catalyst support. eV [49]. The relative portion (%) of metallic Pt (Pt0) species in Pt/SiNS is
To further verify the reducibility of metal species on different sup­ lower than that of Pt-Zn catalysts, as shown in Table S3.
ports, the H2-TPR experiments were carried out, as illustrated in Fig. S6. To further confirm the structural state of metal species, X-ray ab­
The reduction peak of 1Pt/SiNS at 220 ◦ C is attributed to the reduction sorption spectroscopy (XAS) was also performed. X-ray absorption near
of platinum oxide species to metallic platinum. Another high reduction edge structure (XANES) analysis of Pt L3-edge of all the prepared cata­
temperature peak at 500 ◦ C can be referred to the reduction of Pt– lysts, as shown in Fig. 3A, displays a similar trend with the Pt foil.
(− O–Si≡)y complexes, generated by the Pt ion coordinated with silicon Typically, the intensity of the white line in XANES corresponds to a
hydroxyl or defect sites in silicalite-1 zeolites [43]. After adding zinc on direct measurement of the d-band vacancy. For instance, a high white
Pt surfaces to form Pt-Zn alloy nanoparticles, the temperature of metal line intensity of the oxidized platinum is detected due to a high electron
reduction shifts toward a higher range compared to a monometallic Pt. vacancy in the d-orbital. In contrast, a low white line intensity is
In addition, compared to the 1Pt-1Zn/SiNS, the reduction temperature observed on metallic platinum due to a low electron vacancy [50,51]. As
of 1Pt-1Zn/SiCON is significantly higher. Therefore, it is reasonable to shown in Fig. 3A, the white-line intensity of all the prepared Pt-Zn/SiNS
deduce that the SiNS is advantageous for the metal dispersion, resulting samples is significantly lower than that of the 1Pt/SiNS catalyst. These
in a tiny metal particle size, efficiently facilitating the reducibility of Pt- observations relate to the fact that the electron from zinc transfers to the
Zn metal oxides to Pt-Zn alloy. Conversely, the Pt-Zn supported on platinum due to the formation of Pt-Zn alloy nanoparticles, and it is
SiCON contains large-aggregated Pt-Zn particles, which are difficult to consistent with the above-mentioned XPS results.
be transformed into metallic species compared to the small particle one Besides, the corresponding extended X-ray absorption fine structure
[44]. (EXAFS) spectra (Fig. 3C) of all samples show the integrated aspects of
Indeed, it is expected that the propylene selectivity obtained from platinum oxide and metallic platinum, and it is also consistent with XPS
PDH also depends on the acidity of a catalyst. For example, further side results. Although the EXAFS fitting results, as illustrated in Fig. S8 and
reactions, such as catalytic cracking and oligomerization are unavoid­ Table S4, reveal that the average coordination number (CN) of Pt-O of
able when using a high Brønsted acid catalyst. NH3-TPD profiles and 1Pt/SiNS and 1Pt-1Zn/SiNS is ca. 1.2 and 1.4, respectively, the pre­
DRIFTS-pyridine adsorption experiments were conducted, as shown in dominant species in 1Pt/SiNS and 1Pt-1Zn/SiNS samples are mono­
Fig. S7 and Table S2, confirming that the total acid sites of 1Pt/SiNS and metallic Pt, and Pt-Zn alloy with CNs of Pt-Pt and Pt-Zn is ca. 9.6 and 6.2,
all Pt-Zn/SiNS samples are similar to the bare SiNS. However, to respectively. Moreover, an average bond distance of Pt-Pt in the Pt-Zn
differentiate the type of acid sites (Lewis or Brønsted acid sites), DRIFTS sample (1Pt-1Zn/SiNS) is approximately 2.77 ± 0.04 Å, which is

5
C. Rodaum et al. Fuel 325 (2022) 124833

Fig. 2. XPS (Pt 4f) spectra of (A) 1Pt/SiNS, (B) 1Pt-0.5Zn/SiNS, (C) 1Pt-1Zn/SiNS, (D) 1Pt-2Zn/SiNS, and (E) XPS (Zn LMM core level) spectra of (a) 1Zn/SiNS, (b)
1Pt-0.5Zn/SiNS, (c) 1Pt-1Zn/SiNS, and (d) 1Pt-2Zn/SiNS.

longer than that of monometallic Pt (1Pt/SiNS) (2.70 ± 0.1 Å). These propylene yield, a shifting conversion equilibrium is also possible by
observations relate to the fact that the aggregation of Pt can be pre­ integrating the RWGS with a 1:1 vol ratio of C3H8:CO2, which is close to
vented in the Pt-Zn alloy samples [52]. the ideal mole ratio in the coupling reaction of CO2-assisted propane
In terms of XANES and EXAFS of Zn K-edge spectra (Fig. 3B and D), it dehydrogenation (PDH: C3H8 ⇌ C3H6 + H2 and RWGS: CO2 + H2 ⇌ CO
was found that the majority of zinc species in all the prepared catalysts is + H2O).
in the zinc (II) oxide form. This might be attributed to the fact that zinc Initially, to study the effect of CO2 as a co-feed on propane dehy­
can be oxidized in the air under atmospheric conditions or stabilized by drogenation, the 1Pt/SiNS was applied as a catalyst. As illustrated in
interacting with the zeolite oxygen framework [46]. Fig. S11, the conversion of propane is significantly higher in the pres­
In addition, scanning transmission electron microscopy-energy ence of CO2 (35 %) (C3H8:CO2 = 1:1) with respect to without CO2 as a
dispersive X-ray spectroscopy (STEM-EDS), high-angle annular dark- co-feed (26 %). This behavior relates to the fact that CO2 can improve
field imaging (HAADF), and corresponding inverse fast fourier trans­ the propane conversion by shifting equilibrium conversion through
form (IFFT) images of 1Pt/SiNS, and 1Pt-1Zn/SiNS were used to identify RWGS reaction. To further investigate, the catalytic performance in the
a metal structure, including metal species as well as metal crystal planes. presence of CO2 was carried out over different catalysts (1Pt/SiNS, 1Pt-
As shown in Fig. S9, the 1Pt/SiNS exhibits the lattice spacing of 1Zn/SiNS, and 1Zn/SiNS), as demonstrated in Fig. 5A. Remarkably, the
approximately 0.23 nm, corresponding to the plane index of (1 1 1) 1Pt-1Zn/SiNS was found as an excellent catalyst in CO2-assisted propane
[49,53]. Regarding the STEM-EDS mapping images of the 1Pt-1Zn/SiNS dehydrogenation via RWGS, and it exhibited the highest catalytic ac­
(Fig. 4A-C), it reveals the homogeneous distribution of Pt and Zn species tivity with propane conversion up to 47 %. In contrast, other catalysts,
over the entire area of SiNS surfaces. Moreover, the spacing of the lattice including 1Pt/SiNS and 1Zn/SiNS, display only 36 % and 13 % of pro­
fringes of the Pt-Zn nanoparticle, as illustrated in Fig. 4D-F, is approxi­ pane conversion, respectively. The reason for an improved propane
mately 0.284 nm, which agrees well with a lattice plane distance of PtZn conversion might relate to a synergistic effect between zinc and plat­
along with the (1 1 0) directions [49,54]. Interestingly, several Pt species inum, eventually enhancing the catalytic activity by shifting equilibrium
can be found when introducing Pt-Zn alloy clusters on conventional conversion due to H2 consumption via the RWGS reaction. Most
silicalite-1(SiCON), particularly PtO2 species, as shown in Fig. S10. This importantly, almost 80 % of propylene selectivity can be obtained using
behavior is due to the fact that the size of metals deposited on SiCON is the Pt-1Zn/SiNS alloy catalyst, which is remarkably higher than other
much larger than the one deposited on SiNS, and therefore it is difficult catalysts. These findings imply that the Pt-Zn catalyst enhances propane
to form metallic species. As a result, platinum oxide species can be found conversion and increases the selectivity of propylene.
on 1Pt-1Zn/SiCON surfaces. These results also show the benefit of using As mentioned above, the Pt-Zn catalyst plays a vital role in the RWGS
hierarchical structure as catalyst support. reaction because zinc can enhance the CO2 adsorption on metal sites
[56], leading to an increase in CO2 activation and driving the con­
sumption of the produced H2 from the dehydrogenation of propane. As a
3.3. Catalytic performance for CO2-assisted propane dehydrogenation via
result, it can be shifted to the forward reaction. To further illustrate the
RWGS
influence of the Zn loading in the Pt-Zn catalyst, the zinc was combined
with 1 wt% of platinum in the range of 0.5–2.0 wt% as shown in Fig. 5B
It is well known that the low partial pressure of propane is required
and Fig. S12 when using CO2 as a co-feed. Interestingly, there is an
to achieve high propane conversion [6,55]. However, to improve the

6
C. Rodaum et al. Fuel 325 (2022) 124833

Fig. 3. (A) Pt L3-edge XANES spectra, (B) Zn K-edge XANES spectra, (C) Fourier transformation of the k2-weighted EXAFS spectra at Pt L3-edge, and (D) Fourier
transformation of the k2-weighted EXAFS spectra at Zn K-edge of all prepared samples.

upward trend of propane conversion from 39 % to 47 %, with an in­ the low metal surface area derived from a low metal dispersion
crease in zinc content from 0.5 to 1.0 wt%, whereas when using too high (Table S1). However, to investigate the propylene production perfor­
zinc loading (2 wt%), the lower conversion (41 %) was obtained. As mance among different catalysts, the propylene formation rate (r(C3H6))
shown in Fig. S12, the propane conversion as a function of time of all has been applied, as shown in Table S6. As expected, the highest r(C3H6)
catalysts tended to slightly decrease due to catalyst deactivation from in both the initial period and after the reaction time of 3 h was obtained
the metal sintering during the high reaction temperature. In addition, using the 1Pt-1Zn/SiNS with 1.98 × 103 and 1.21 × 103 mmol g-1 − 1
Pth ,
when using too high zinc contents, a high deactivation rate can be respectively. These observations demonstrate that the 1Pt-1Zn/SiNS
observed. This behavior relates to the fact that at too high contents of contains a suitable amount of platinum and zinc, eventually
zinc, CO2 adsorption can be promoted, eventually inhibiting the improving the propane conversion and enhancing the propylene pro­
adsorption of propane on the metal sites. Therefore, to verify the effect duction in CO2-assisted propane dehydrogenation via RWGS with
of the Pt-Zn catalysts on the catalytic activity, as shown in Fig. S13 and respect to other samples.
Table S5, CO2-TPD profiles reveal that the CO2 adsorption ability can be
enhanced with an increase in zinc loading. Although an increased zinc
3.4. The effect of different catalyst supports on catalytic performance for
content can promote the adsorption of CO2, which is beneficial for
CO2-assisted propane dehydrogenation via the RWGS reaction
RWGS reaction, excessive zinc would also inhibit propane adsorption on
metal surfaces, eventually decreasing propane conversion. Conse­
Apart from the effect of Pt-Zn alloy on CO2-assisted propane dehy­
quently, the moderate zinc content (1Pt-1Zn/SiNS) is appropriate for
drogenation via the RWGS reaction, to illustrate the benefit of using
this reaction in high propane conversion and low catalyst deactivation.
synthesized silicalite-1 nanosheet, the effect of other supports, including
In addition, as illustrated in Table S6, the turnover frequency (TOF)
SiCON, Al2O3, and CeO2 on catalytic performances was investigated.
of propane conversion at the initial reaction period over the 1Pt-1Zn/
Comparing the activity among different supports, as shown in Fig. S14,
SiNS (1.82 × 104h− 1) is higher than that of the isolated Pt catalyst
the silicalite-1 nanosheet (SiNS) as catalyst support for Pt-Zn exhibits the
(1.19 × 104h− 1), implying that the Pt-Zn catalyst plays a key role in CO2-
highest catalytic performance with 46.7 % of propane conversion and
assisted propane dehydrogenation via the RWGS reaction. Regarding the
80 % of C3H6 selectivity. In contrast, the propane conversion over 1Pt-
TOF of catalysts containing different zinc contents, the TOF of 1Pt-2Zn/
1Zn/SiCON, 1Pt-1Zn/Al2O3, and 1Pt-1Zn/CeO2 catalysts is approxi­
SiNS shows a higher value of 4.90 × 104h− 1 than other samples due to
mately 11.3 %, 28.9 %, and 21.5 %, respectively. In addition, the

7
C. Rodaum et al. Fuel 325 (2022) 124833

Fig. 4. (A-C) STEM-EDS mapping of 1Pt-1Zn/SiNS, (D) HAADF-STEM image of 1Pt-1Zn/SiNS, and (E-F) IFFT images of Pt-Zn in (D).

Fig. 5. Catalytic activity in propane dehydrogenation at 600 ◦ C over (A) different metals on supports and (B) different zinc loading: C3H8:CO2 = 1:1, total flow rate
= 20 mLmin− 1, atmospheric pressure, and TOS = 15 min. (bar chart = propane conversion and = propylene selectivity).

propylene selectivity is merely 49 %, 64 %, and 54 % when using SiCON, fraction of Pt-Zn active surface species with respect to SiCON, resulting
Al2O3, and CeO2 as catalyst supports, respectively. Moreover, the for­ in an excellent catalytic activity compared to using a conventional one
mation rate of propylene (r(C3H6)) of the Pt-Zn supported on different as catalyst support. In addition, the 1Pt-1Zn supported on Al2O3 and
types of carriers is reported, as shown in Table S7. Herein, the (r(C3H6)) CeO2 shows a significantly lower r(C3H6) compared to the 1Pt-1Zn/SiNS
derived from equation (3) is used to compare the catalytic activity of even though their metal particle size is smaller than the metal particles
different support materials. Interestingly, the r(C3H6) of the 1Pt-1Zn/ on SiNS (Table S1). This relates to the competitive adsorption between
SiNS is remarkably higher than that of the 1Pt-1Zn/SiCON. This CO2 and C3H8 over Al2O3 and CeO2 supports, resulting in a decrease in
behavior relates to the fact that the SiNS contains a higher surface area catalytic performance in CO2-assisted dehydrogenation [8]. It is,
than the SiCON, eventually preventing metal aggregation and promot­ therefore, reasonable to conclude that the high catalytic activity of
ing the dispersion of Pt-Zn alloy over hierarchical nanosheet surfaces propylene production can be greatly improved by using the 1Pt-1Zn
(Table S1). These observations confirm that the SiNS provides a higher supported on SiNS surfaces.

8
C. Rodaum et al. Fuel 325 (2022) 124833

3.5. Mechanistic study of propane dehydrogenation in the presence of as illustrated in Fig. 6. After introducing propane and CO2 on the surface
CO2 over Pt-Zn catalysts using in-situ DRIFTS experiments of active metals and prolonging the reaction time for 5 min, corre­
sponding FTIR spectra were then recorded at 40 ◦ C under N2 flow.
To gain insights into the effect of Pt-Zn alloy on the mechanistic point Subsequently, the reaction was continued under a similar procedure,
of view, in-situ DRIFTS experiments of CO2-assisted propane dehydro­ and the spectra were collected at a time on stream (TOS) of 10, 15, 30,
genation via the RWGS reaction were performed over different catalysts, and 60 min.

Fig. 6. DRIFTS spectra as a function of time-on-stream (TOS) upon CO2-assisted propane dehydrogenation via the RWGS reaction at 550 ◦ C: DRIFTS spectra in C–H
bending region (1400 to 1800 cm− 1) over (A) 1Pt-1Zn/SiNS and (B) 1Pt/SiNS and DRIFTS spectra in O–H stretching region (3300 to 3700 cm− 1) over (C) 1Pt-1Zn/
SiNS and (D) 1Pt/SiNS. The spectra were measured at 40 ◦ C after TOS of 5, 15, 30, 45, 60, and 90 min.

9
C. Rodaum et al. Fuel 325 (2022) 124833

Indeed, the IR band at 2340 cm− 1 assigned to the CO2 gas phase can However, using the Pt-Zn catalyst for CO2-assisted propane dehydro­
be found in both catalyst surfaces (Fig. 6A-B). Therefore, to differentiate genation via the RWGS reaction can prolong the catalytic lifetime.
the effect of catalytic activity over different active metals, the peak in In addition, carbon monoxide temperature-programmed desorption
the range of 2100–1400 cm− 1 was studied. The obtained IR bands at (CO-TPD) experiments, as shown in Fig. S16, were carried out to eval­
1590 cm− 1, 1465 cm− 1, and 2100–1800 cm− 1 can be assigned to uate the interaction between CO and catalyst surfaces. As expected, the
formate, carbonate, and carbonyl species, respectively, which are one of CO desorption peak of Pt-Zn catalysts shifts toward lower temperature
the most important intermediate species in the RWGS reaction [57,58]. with respect to the monometallic Pt (1Pt/SiNS), indicating the weaker
Although these observations imply that both Pt(1 1 1) and Pt-Zn(1 1 0) interaction between CO molecules and Pt-Zn surfaces. The promoted
can promote the RWGS, the sharp peak at approximately 1622 cm− 1 desorption of CO relates to the fact that the CO chemisorption in Pt(5d)
assigned to alkene species: C– – C bond (1600–1680 cm− 1) [59] can be on Pt-Zn surfaces (Pt(5d)–CO(2π*) bonding interaction) is weakened
seen just in case of the Pt-Zn catalyst. These observations indicate that due to the electronic perturbations on Pt surfaces by Zn species [60].
the Pt-Zn could promote the RWGS reaction, eventually improving Moreover, DRIFTS spectra of CO adsorption were also recorded to
propane conversion limitation. Nevertheless, this region (2100–1400 investigate the effect of CO adsorption on different metal species. After
cm− 1) contains several possible intermediates from propane dehydro­ the adsorption process, CO molecules on catalyst surfaces were desorbed
genation and RWGS. Thus, it is quite complicated and insufficient to by purging under N2 flow at 100 ◦ C, and obtained spectra were collected
conclude the mechanistic point of view and to compare the different as a function of desorption time (Fig. 7). The evolution of CO desorption
catalytic performances between Pt and Pt-Zn. Consequently, the O–H over the 1Pt/SiNS catalyst reveals that CO molecules were strongly
stretching region was also studied, as illustrated in Fig. 6C-D, to inves­ adsorbed on monometallic Pt surfaces, even prolonging N2 purging time
tigate the influence of different metal species on the catalytic RWGS up to 8 min. Conversely, the CO desorption on the 1Pt-1Zn/SiNS was
reaction. Surprisingly, the narrow band at 3600 cm− 1 and broad-band in detected immediately even after 2 min of the desorption process. In
the range of 3300–3570 cm− 1 attributed to the nonbonded hydroxy addition, the peak area ratio of CO adsorption on metal surfaces at a
group and hydroxy group (H-bonded), respectively, [59] can be detec­ certain desorption time to the CO adsorption at the initial time (COi/
ted in the case of the 1Pt-1Zn/SiNS catalyst. In contrast, these IR bands COt=0) was used to investigate the CO desorption rate on metal surfaces.
are difficult to be observed over the 1Pt/SiNS sample. These charac­ Obviously, the COi/COt=0 ratio of the 1Pt-1Zn/SiNS decreases dramat­
teristics belong to one of the most significant intermediates from RWGS ically at the beginning step, whereas a slight decline of this ratio was
obtained from hydroxycarbonyl intermediate decomposition (OCOH ⇌ observed in the 1Pt/SiNS sample, implying that the CO desorption rate
CO + OH), subsequently generating water as an RWGS product from the over the 1Pt-1Zn/SiNS is significantly higher than that of the mono­
adsorbed H and OH species (OH + H ⇌ H2O) [57]. These observations metallic Pt. It is, therefore, reasonable to confirm again that the CO
indicate that the synergistic effect between Pt-Zn (1 1 0) remarkably molecule is easier desorbed on the Pt-Zn surface than that of the
promotes the RWGS reaction compared with the monometallic Pt, monometallic Pt (1Pt/SiNS).
eventually improving H2 consumption and shifting the equilibrium
conversion of propane. As a result, a high propane conversion can be 3.6. Computational study of CO adsorption on Pt(1 1 1) and Pt-Zn(1 1 0)
achieved using the Pt-Zn alloy supported on SiNS surfaces. surfaces
It is well-known that the generated CO molecules as by-products
from the RWGS reaction interact strongly on Pt active surfaces, lead­ To further demonstrate the influence of metal-geometric structures
ing to catalyst deactivation. To circumvent this issue, one of the most on CO adsorption, the DFT calculations were applied. There are two
effective ways to improve catalyst efficiency is by introducing other models representing Pt and Pt-Zn slabs with the facet of (1 1 1) and
metals as promoters on Pt surfaces. Herein, zinc can be used as an (1 1 0), respectively, which are considered as the preferable crystal
attractive promoter. Therefore, the advantages of the designed Pt-Zn planes of monometallic platinum (1Pt/SiNS) and Pt-Zn alloy catalyst
catalyst are not only by shifting equilibrium conversion of propane (1Pt-1Zn/SiNS), respectively. As shown in Fig. S17, four layers of 3 × 3
but also by preventing the strong interaction between metal sites and CO supercell and 4 × 4 supercell were used as models for Pt (1 1 1) and Pt-Zn
molecules. To illustrate the improvement of using the Pt-Zn catalyst for (1 1 0), respectively, and a 20-Å vacuum space was applied among
preventing the strong interaction of CO on metal sites, various experi­ periodically repeated slabs to avoid any artificial interaction. The CO
ments, including CO-TPD, DRIFTS spectra of CO adsorption, and DFT molecule adsorbs on both Pt (1 1 1) and Pt-Zn (1 1 0) surfaces through
studies of CO adsorption on Pt(1 1 1) and Pt-Zn(1 1 0) relating to the interaction between C of CO and Pt active site. The formation of this
catalytic RWGS reaction were carried out. adsorption mode is due to the dipole moment direction in CO from O to
For the catalytic experiment in the RWGS reaction, as illustrated in C, leading to high electron density located on the C [61,62]. From the
Fig. S15 and Table S8, the cumulative TON of RWGS over the 1Pt-1Zn/ DFT study, the C–O bond of CO is elongated from 1.13 Å to 1.16 and
SiNS is remarkably higher than that of the 1Pt/SiNS, implying that the 1.14 Å for the adsorption on Pt (1 1 1) and Pt-Zn (1 1 0), respectively.
Pt-Zn catalyst can greatly enhance the catalytic efficiency in RWGS with Moreover, as shown in Table S9, the adsorption energy of CO mol­
respect to the isolated Pt catalyst. In addition, compared with the 1Pt- ecules over the Pt-Zn(1 1 0) surface is significantly lower than that of the
1Zn/SiNS, a slower increase in cumulative TON of RWGS as a function Pt(1 1 1) surface (-1.22 eV and − 1.93 eV for Pt-Zn(1 1 0) and Pt(1 1 1),
of time after prolonging the reaction period can be observed in the case respectively). The less stable CO adsorbed on Pt-Zn(1 1 0) relates to the
of the isolated Pt. These observations indicate that using the mono­ fact that the introduction of Zn in Pt enhances the electron density on the
metallic Pt catalyst still suffers from the deactivation of a catalyst in the Pt binding site due to the higher electronegativity of Pt compared to Zn.
presence of produced CO obtained from the RWGS reaction compared This behavior leads to the increase of nucleophilicity on the Pt site in the
with the Pt-Zn alloy catalyst, also confirmed by the initial deactivation Pt-Zn system, and therefore the lower binding of CO on the Pt active site
rate (kd) in RWGS over the 1Pt/SiNS and the 1Pt-1Zn/SiNS approxi­ in Pt-Zn surfaces compared to the monometallic Pt is acquired. This
mately 0.12 and 0.04 h− 1, respectively (Table S8). Moreover, TOFCO2 of clarifies that the Pt-Zn catalyst can weaken the interaction between CO
the Pt-Zn catalyst exhibits a value of 5,467 h− 1 at the initial of the re­ and metal surfaces, which would be an advantage for improving the
action, and then slightly drops to 4,858 h− 1 after 180 min of time on catalyst stability and regeneration.
stream, while the 1Pt/SiNS displays a TOF of 3,296 h− 1 and 2,339 h− 1 at
the initial time and after 180 min, respectively. These results also indi­ 3.7. Catalyst deactivation and regeneration process
cate that the isolated Pt exhibits a faster initial deactivation than the /Pt-
Zn catalyst. This behavior relates to the fact that the strongly adsorbed As mentioned above, the interaction between CO and Pt-Zn/SiNS is
CO on Pt surfaces inhibits the adsorption of reactant molecules. weakened with respect to the monometallic Pt. Indeed, CO poisoning

10
C. Rodaum et al. Fuel 325 (2022) 124833

Fig. 7. DRIFTS spectra of remaining CO on metal surfaces during the desorption process at 100 ◦ C as a function of time over (A) 1Pt/SiNS and (B) 1Pt-1Zn/SiNS, and
(C) the area ratio of CO adsorption at a certain time to initial CO adsorption (COi/COt=0) over (a) 1Pt/SiNS and (b) 1Pt-1Zn/SiNS.

affects catalyst deactivation, and the change of structural properties of assisted propane dehydrogenation via RWGS, DRIFTS measurements of
metal and metal sintering are among the most important reasons for CO adsorption, and DFT calculations, have been employed to under­
catalyst deactivation. As can be seen in Table S10 and Fig. S18, the stand the corresponding mechanism as well as the influence of syner­
percentage of Pt and Zn of the spent 1Pt-1Zn/SiNS catalyst is unchanged, gistic Pt-Zn catalysts on the catalytic activity with respect to the
while the relative portion of metallic platinum species of 1Pt-1Zn/SiNS monometallic platinum. Moreover, combining experimental data and
after the reaction slightly decreases. These observations confirm that no DFT results clearly confirms that the Pt-Zn nanoparticles on SiNS can
leaching of Pt-Zn species from the SiNS surface can be obtained after the significantly reduce the interaction between metal surfaces and CO
catalytic reaction at a high temperature. molecules, eventually enhancing the overall catalytic performances.
In addition, the metal aggregation after the reaction was monitored, In conclusion, our findings demonstrate that the highly dispersed Pt-
as illustrated in TEM images (Fig. S19). As shown in Fig. S19A-B, the size Zn surfaces have been successfully fabricated when using a hierarchical
of spent Pt-Zn nanoparticles is slightly larger than that of the fresh zeolite nanosheet as solid support. By fine-tuning the Pt-Zn composition
catalyst due to their aggregation during high reaction temperature. on zeolite support, the synergistic effect of Pt-Zn due to the modified
Although the Pt-Zn nanoparticles size of the spent 1Pt-1Zn/SiNS (6.9 ± electronic properties on its surfaces eventually enhances the catalytic
5.5 nm) is greater than that of the fresh one (3.9 ± 1.7 nm), its size is behavior of propane dehydrogenation. Most importantly, the dual effect
obviously smaller than that of the spent 1Pt-1Zn/SiCON (69 ± 34 nm) or of using Pt-Zn alloy surfaces and dehydrogenation of propane in the
even the fresh 1Pt-1Zn/SiCON catalyst (48 ± 14). This again demon­ presence of CO2 via the RWGS reaction allows the rational improvement
strates the advantage of using silicalite-1 nanosheet as a catalyst support of catalytic activity of propylene production with a high turnover fre­
to highly disperse ultrasmall Pt-Zn nanoparticles on silica surfaces. quency (TOF) of 1.82 × 104h− 1. These findings not only open up the
Moreover, the activity of regenerated catalysts was investigated as perspective of rational design of the Pt-Zn nanoparticles on hierarchical
shown in Fig. S20. The slope of propane conversion can be referred to zeolites for dehydrogenation of propane in the presence of CO2 but also
the deactivation rate of the catalyst. Although the propane conversion provide the vital information generated by various techniques ranging
after the 1st regeneration cycle by air and H2 is slightly lower than that from experimental and calculation/simulation points of view, which are
of the fresh catalyst, a relatively high catalytic activity can be observed very useful for the further development of heterogeneous catalytic
in terms of a slow deactivation rate with the stable high propane con­ systems.
version even after several catalytic cycles.
CRediT authorship contribution statement
4. Conclusions
Chadatip Rodaum: Conceptualization, Formal analysis, Methodol­
In this contribution, the highly dispersed platinum-zinc alloy nano­
ogy, Investigation, Visualization, Writing – original draft. Peeranat
particles on the synthesized silicalite-1 nanosheet have been successfully
Chaipornchalerm: Investigation. Watinee Nunthakitgoson: Investi­
fabricated via a simple impregnation method. Owing to the benefit of
gation. Anawat Thivasasith: Investigation. Thana Maihom: Formal
hierarchical structures of solid support, ultrasmall Pt-Zn alloy nano­
analysis. Thassanant Atithep: Investigation. Pinit Kidkhunthod:
particles can be observed with excellent dispersion. To illustrate the
Investigation. Chayapat Uthayopas: Formal analysis. Sarana Nuta­
advantages of both synergistic effect between Pt and Zn and the hier­
nong: Formal analysis. Sutarat Thongratkaew: Investigation.
archical structures of SiNS, the Pt-Zn(1 1 0) is beneficial for both propane
Kajornsak Faungnawakij: Investigation. Chularat Wattanakit:
dehydrogenation and RWGS reaction in CO2-assisted propane dehy­
Conceptualization, Methodology, Resources, Writing – review & editing,
drogenation, eventually facilitating the equilibrium shift of propane
Supervision.
conversion by consuming the produced H2. This allows reaching an
improved catalytic activity in terms of propane conversion and propyl­
ene selectivity with respect to the metallic platinum (1 1 1). Further­ Declaration of Competing Interest
more, various independent techniques, including in-situ DRIFTS of CO2-
The authors declare that they have no known competing financial

11
C. Rodaum et al. Fuel 325 (2022) 124833

interests or personal relationships that could have appeared to influence [19] Nykänen L, Honkala K. Density functional theory study on propane and propene
adsorption on Pt(111) and PtSn alloy surfaces. J Phys Chem C 2011;115:9578–86.
the work reported in this paper.
https://doi.org/10.1021/jp1121799.
[20] Zhang Y, Zhou Y, Qiu A, Wang Y, Xu Y, Wu P. Propane dehydrogenation on PtSn/
Acknowledgments ZSM-5 catalyst: Effect of tin as a promoter. Catal Commun 2006;7:860–6. https://
doi.org/10.1016/j.catcom.2006.03.016.
[21] Zhang S, Zhou Y, Zhang Y, Huang L. Effect of K addition on catalytic performance
This work is financially supported by the Vidyasirimedhi Institute of of PtSn/ZSM-5 catalyst for propane dehydrogenation. Catal Lett 2010;135:76–82.
Science and Technology (VISTEC) and the National Research Council of https://doi.org/10.1007/s10562-010-0269-4.
Thailand (NRCT5-RSA63025-03). In addition, this research project is [22] Gomez E, Kattel S, Yan B, Yao S, Liu P, Chen JG. Combining CO2 reduction with
propane oxidative dehydrogenation over bimetallic catalysts. Nat Commun 2018;9:
supported by National Research Council of Thailand (NRCT): NRCT5- 1398. https://doi.org/10.1038/s41467-018-03793-w.
RGJ63024-178 through the Royal Golden Jubilee (RGJ). Moreover, [23] Gomez E, Xie Z, Chen JG. The effects of bimetallic interactions for CO2-assisted
this work also received financial support from Thailand Science oxidative dehydrogenation and dry reforming of propane. AIChE J 2019;65:
e16670.
Research and Innovation (TSRI, FRB650023/0457) and Energy Con­ [24] Michorczyk P, Ogonowski J. Dehydrogenation of propane in the presence of carbon
servation and Promotion Fund Office (ENCON Fund, 64-03-0003). dioxide over oxide-based catalysts. React kinet catal lett 2003;78:41–7. https://
doi.org/10.1023/A:1022501613772.
[25] Liu J, He N, Zhang Z, Yang J, Jiang X, Zhang Z, et al. Highly-dispersed zinc species
Appendix A. Supplementary data on zeolites for the continuous and selective dehydrogenation of ethane with CO2 as
a soft oxidant. ACS Catal 2021;11:2819–30. https://doi.org/10.1021/
Supplementary data to this article can be found online at https://doi. acscatal.1c00126.
[26] Jiang X, Sharma L, Fung V, Park SJ, Jones CW, Sumpter BG, et al. Oxidative
org/10.1016/j.fuel.2022.124833. dehydrogenation of propane to propylene with soft oxidants via heterogeneous
catalysis. ACS Catal 2021;11:2182–234. https://doi.org/10.1021/
References acscatal.0c03999.
[27] Park S-W, Joo O-S, Jung K-D, Kim H, Han S-H. Development of ZnO/Al2O3 catalyst
for reverse-water-gas-shift reaction of CAMERE (carbon dioxide hydrogenation to
[1] Lu W-D, Wang D, Zhao Z, Song W, Li W-C, Lu A-H. Supported boron oxide catalysts
form methanol via a reverse-water-gas-shift reaction) process. Appl Catal, A 2001;
for selective and low-temperature oxidative dehydrogenation of propane. ACS
211:81–90. https://doi.org/10.1016/S0926-860X(00)00840-1.
Catal 2019;9:8263–70. https://doi.org/10.1021/acscatal.9b02284.
[28] Su X, Yang X, Zhao B, Huang Y. Designing of highly selective and high-temperature
[2] Xu Z, Yue Y, Bao X, Xie Z, Zhu H. Propane dehydrogenation over Pt clusters
endurable RWGS heterogeneous catalysts: recent advances and the future
localized at the Sn single-site in zeolite framework. ACS Catal 2020;10:818–28.
directions. J Energy Chem 2017;26:854–67. https://doi.org/10.1016/j.
https://doi.org/10.1021/acscatal.9b03527.
jechem.2017.07.006.
[3] Gu Y, Liu H, Yang M, Ma Z, Zhao L, Xing W, et al. Highly stable phosphine modified
[29] Lin F, Delmelle R, Vinodkumar T, Reddy BM, Wokaun A, Alxneit I. Correlation
VOx/Al2O3 catalyst in propane dehydrogenation. Appl Catal B 2020;274:119089.
between the structural characteristics, oxygen storage capacities and catalytic
https://doi.org/10.1016/j.apcatb.2020.119089.
activities of dual-phase Zn-modified ceria nanocrystals. Catal Sci Technol 2015;5:
[4] Zhu J, Yang M-L, Yu Y, Zhu Y-A, Sui Z-J, Zhou X-G, et al. Size-dependent reaction
3556–67. https://doi.org/10.1039/C5CY00351B.
mechanism and kinetics for propane dehydrogenation over Pt catalysts. ACS Catal
[30] Álvarez Galván C, Schumann J, Behrens M, Fierro JLG, Schlögl R, Frei E. Reverse
2015;5:6310–9. https://doi.org/10.1021/acscatal.5b01423.
water-gas shift reaction at the Cu/ZnO interface: influence of the Cu/Zn ratio on
[5] Searles K, Chan KW, Mendes Burak JA, Zemlyanov D, Safonova O, Copéret C.
structure-activity correlations. Appl Catal B 2016;195:104–11. https://doi.org/
Highly productive propane dehydrogenation catalyst using silica-supported Ga–Pt
10.1016/j.apcatb.2016.05.007.
nanoparticles generated from single-sites. J Am Chem Soc 2018;140:11674–9.
[31] Chen C, Sun M, Hu Z, Ren J, Zhang S, Yuan Z-Y. New insight into the enhanced
https://doi.org/10.1021/jacs.8b05378.
catalytic performance of ZnPt/HZSM-5 catalysts for direct dehydrogenation of
[6] Sattler JJHB, Ruiz-Martinez J, Santillan-Jimenez E, Weckhuysen BM. Catalytic
propane to propylene. Catal Sci Technol 2019;9:1979–88. https://doi.org/
dehydrogenation of light alkanes on metals and metal oxides. Chem Rev 2014;114:
10.1039/C9CY00237E.
10613–53. https://doi.org/10.1021/cr5002436.
[32] Zhang Y, Zhou Y, Tang M, Liu X, Duan Y. Effect of La calcination temperature on
[7] Chen S, Chang X, Sun G, Zhang T, Xu Y, Wang Y, et al. Propane dehydrogenation:
catalytic performance of PtSnNaLa/ZSM-5 catalyst for propane dehydrogenation.
catalyst development, new chemistry, and emerging technologies. Chem Soc Rev
Chem Eng J 2012;181–182:530–7. https://doi.org/10.1016/j.cej.2011.11.055.
2021;50:3315–54. https://doi.org/10.1039/D0CS00814A.
[33] Santhosh Kumar M, Holmen A, Chen D. The influence of pore geometry of Pt
[8] Atanga MA, Rezaei F, Jawad A, Fitch M, Rownaghi AA. Oxidative dehydrogenation
containing ZSM-5, Beta and SBA-15 catalysts on dehydrogenation of propane.
of propane to propylene with carbon dioxide. Appl Catal B 2018;220:429–45.
Microporous and Mesoporous Mater 2009;126:152–8. https://doi.org/10.1016/j.
https://doi.org/10.1016/j.apcatb.2017.08.052.
micromeso.2009.05.031.
[9] Nowicka E, Reece C, Althahban SM, Mohammed KMH, Kondrat SA, Morgan DJ,
[34] Zhang Y, Zhou Y, Huang L, Xue M, Zhang S. Sn-modified ZSM-5 as support for
et al. Elucidating the role of CO2 in the soft oxidative dehydrogenation of propane
platinum catalyst in propane dehydrogenation. Ind Eng Chem Res 2011;50:
over ceria-based catalysts. ACS Catal 2018;8:3454–68. https://doi.org/10.1021/
7896–902. https://doi.org/10.1021/ie1024694.
acscatal.7b03805.
[35] Wang N, Sun Q, Zhang T, Mayoral A, Li L, Zhou X, et al. Impregnating
[10] Jin R, Easa J, Tran DT, O’Brien CP. Ru-Promoted CO2 activation for oxidative
subnanometer metallic nanocatalysts into self-pillared zeolite nanosheets. J Am
dehydrogenation of propane over chromium oxide catalyst. Catal Sci Technol
Chem Soc 2021;143:6905–14. https://doi.org/10.1021/jacs.1c00578.
2020;10:1769–77. https://doi.org/10.1039/C9CY01990A.
[36] Ketkaew M, Klinyod S, Saenluang K, Rodaum C, Thivasasith A, Kidkhunthod P,
[11] Michorczyk P, Zeńczak-Tomera K, Michorczyk B, Węgrzyniak A, Basta M, Millot Y,
et al. Fine-tuning the chemical state and acidity of ceria incorporated in
et al. Effect of dealumination on the catalytic performance of Cr-containing Beta
hierarchical zeolites for ethanol dehydration. Chem Commun 2020;56:11394–7.
zeolite in carbon dioxide assisted propane dehydrogenation. J CO2 Util 2020;36:
https://doi.org/10.1039/D0CC04886K.
54–63. https://doi.org/10.1016/j.jcou.2019.09.018.
[37] Kresse G, Furthmüller J. Efficiency of ab-initio total energy calculations for metals
[12] Zhang W, Wang H, Jiang J, Sui Z, Zhu Y, Chen D, et al. Size dependence of Pt
and semiconductors using a plane-wave basis set. J Comput Mater Sci 1996;6:
catalysts for propane dehydrogenation: from atomically dispersed to nanoparticles.
15–50. https://doi.org/10.1016/0927-0256(96)00008-0.
ACS Catal 2020;10:12932–42. https://doi.org/10.1021/acscatal.0c03286.
[38] Kresse G, Furthmüller J. Efficient iterative schemes for ab initio total-energy
[13] Zha S, Sun G, Wu T, Zhao J, Zhao Z-J, Gong J. Identification of Pt-based catalysts
calculations using a plane-wave basis set. Phys Rev B 1996;54:11169–86. https://
for propane dehydrogenation via a probability analysis. Chem Sci 2018;9:3925–31.
doi.org/10.1103/PhysRevB.54.11169.
https://doi.org/10.1039/C8SC00802G.
[39] Perdew JP, Burke K, Ernzerhof M. Generalized gradient approximation made
[14] Wang Y, Hu Z-P, Lv X, Chen L, Yuan Z-Y. Ultrasmall PtZn bimetallic nanoclusters
simple. Phys Rev Lett 1996;77:3865–8. https://doi.org/10.1103/
encapsulated in silicalite-1 zeolite with superior performance for propane
PhysRevLett.77.3865.
dehydrogenation. J Catal 2020;385:61–9. https://doi.org/10.1016/j.
[40] Perdew JP, Ernzerhof M, Burke K. Rationale for mixing exact exchange with
jcat.2020.02.019.
density functional approximations. J Chem Phys 1996;105:9982–5. https://doi.
[15] Yu C, Xu H, Ge Q, Li W. Properties of the metallic phase of zinc-doped platinum
org/10.1063/1.472933.
catalysts for propane dehydrogenation. J Mol Catal A Chem 2007;266:80–7.
[41] Grimme S. Semiempirical GGA-type density functional constructed with a long-
https://doi.org/10.1016/j.molcata.2006.10.025.
range dispersion correction. J Comput Chem 2006;27:1787–99. https://doi.org/
[16] Han Z, Li S, Jiang F, Wang T, Ma X, Gong J. Propane dehydrogenation over Pt–Cu
10.1002/jcc.20495.
bimetallic catalysts: the nature of coke deposition and the role of copper.
[42] Imyen T, Wannapakdee W, Ittisanronnachai S, Witoon T, Wattanakit C. Tailoring
Nanoscale 2014;6:10000–8. https://doi.org/10.1039/C4NR02143F.
hierarchical zeolite composites with two distinct frameworks for fine-tuning the
[17] Yang M-L, Zhu Y-A, Zhou X-G, Sui Z-J, Chen D. First-principles calculations of
product distribution in benzene alkylation with ethanol. Nanoscale Adv 2020;2:
propane dehydrogenation over PtSn catalysts. ACS Catal 2012;2:1247–58. https://
4437–49. https://doi.org/10.1039/D0NA00391C.
doi.org/10.1021/cs300031d.
[43] Niu X, Li X, Yuan G, Feng F, Wang M, Zhang X, et al. Hollow Hierarchical Silicalite-
[18] Han SW, Park H, Han J, Kim J-C, Lee J, Jo C, et al. PtZn intermetallic compound
1 Zeolite Encapsulated PtNi Bimetals for Selective Hydroconversion of Methyl
nanoparticles in mesoporous zeolite exhibiting high catalyst durability for propane
Stearate into Aviation Fuel Range Alkanes. Ind Eng Chem Res 2020;59:8601–11.
dehydrogenation. ACS Catal 2021;11:9233–41. https://doi.org/10.1021/
https://doi.org/10.1021/acs.iecr.0c01275.
acscatal.1c01808.

12
C. Rodaum et al. Fuel 325 (2022) 124833

[44] Ketkaew M, Suttipat D, Kidkhunthod P, Witoon T, Wattanakit C. Nanoceria- ferromagnetism enhancement and optimized surface strain. Adv Energy Mater
modified platinum supported on hierarchical zeolites for selective alcohol 2019;9:1803771. https://doi.org/10.1002/aenm.201803771.
oxidation. RSC Adv 2019;9:36027–33. https://doi.org/10.1039/C9RA07793F. [54] Lugo VR, Mondragón-Galicia G, Gutiérrez-Martínez A, Gutiérrez-Wing C, Rosales
[45] Osman AI, Abu-Dahrieh JK, Abdelkader A, Hassan NM, Laffir F, McLaren M, et al. González O, López P, et al. Pt–Ni/ZnO-rod catalysts for hydrogen production by
Silver-modified η-Al2O3 catalyst for DME production. J Phys Chem C 2017;121: steam reforming of methanol with oxygen. RSC Adv 2020;10:41315–23. https://
25018–32. https://doi.org/10.1021/acs.jpcc.7b04697. doi.org/10.1039/D0RA06181F.
[46] Sun Q, Wang N, Fan Q, Zeng L, Mayoral A, Miao S, et al. Subnanometer bimetallic [55] Du X, Yao B, Gonzalez-Cortes S, Kuznetsov VL, AlMegren H, Xiao T, et al. Catalytic
platinum–zinc clusters in zeolites for propane dehydrogenation. Angew Chem Int dehydrogenation of propane by carbon dioxide: a medium-temperature
Ed 2020;59:19450–9. https://doi.org/10.1002/anie.202003349. thermochemical process for carbon dioxide utilisation. Faraday Discuss 2015;183:
[47] Ingale P, Knemeyer K, Preikschas P, Ye M, Geske M, Naumann d’Alnoncourt R, 161–76. https://doi.org/10.1039/C5FD00062A.
et al. Design of PtZn nanoalloy catalysts for propane dehydrogenation through [56] Zhao Q-N, Song Q-W, Liu P, Zhang Q-X, Gao J-H, Zhang K. Catalytic conversion of
interface tailoring via atomic layer deposition. Catal Sci Technol 2021;11:484–93. CO2 to cyclic carbonates through multifunctional zinc-modified ZSM-5 zeolite.
https://doi.org/10.1039/D0CY01528H. Chin J Chem 2018;36:187–93. https://doi.org/10.1002/cjoc.201700573.
[48] Matin MA, Lee E, Kim H, Yoon W-S, Kwon Y-U. Rational syntheses of core–shell [57] Bobadilla LF, Santos JL, Ivanova S, Odriozola JA, Urakawa A. Unravelling the role
Fe@(PtRu) nanoparticle electrocatalysts for the methanol oxidation reaction with of oxygen vacancies in the mechanism of the reverse water–gas shift reaction by
complete suppression of CO-poisoning and highly enhanced activity. J Mater Chem operando DRIFTS and ultraviolet–visible spectroscopy. ACS Catal 2018;8:7455–67.
A 2015;3:17154–64. https://doi.org/10.1039/C5TA03809J. https://doi.org/10.1021/acscatal.8b02121.
[49] Chen S, Zhao Z-J, Mu R, Chang X, Luo J, Purdy SC, et al. Propane dehydrogenation [58] Choi S, Sang B-I, Hong J, Yoon KJ, Son J-W, Lee J-H, et al. Catalytic behavior of
on single-site [PtZn4] intermetallic catalysts. Chem 2021;7:387–405. https://doi. metal catalysts in high-temperature RWGS reaction: In-situ FT-IR experiments and
org/10.1016/j.chempr.2020.10.008. first-principles calculations. Sci Rep 2017;7:41207. https://doi.org/10.1038/
[50] Hsieh B-J, Tsai M-C, Pan C-J, Su W-N, Rick J, Lee J-F, et al. Platinum loaded on srep41207.
dual-doped TiO2 as an active and durable oxygen reduction reaction catalyst. NPG [59] Coates J. Interpretation of infrared spectra, A practical approach. Encyclop Anal
Asia Mater 2017;9. https://doi.org/10.1038/am.2017.78. e403-e. Chem 2006:1–23.
[51] Yoshida H, Nonoyama S, Hattori YYT. Quantitative determination of platinum [60] Rodriguez JA, Kuhn M. Chemical and electronic properties of Pt in bimetallic
oxidation state by XANES analysis. Phys Scr 2005;813. https://doi.org/10.1238/ surfaces: Photoemission and CO-chemisorption studies for Zn/Pt(111). J Chem
physica.topical.115a00813. Phys 1995;102:4279–89. https://doi.org/10.1063/1.469475.
[52] Cybulskis VJ, Bukowski BC, Tseng H-T, Gallagher JR, Wu Z, Wegener E, et al. Zinc [61] Janthon P, Viñes F, Sirijaraensre J, Limtrakul J, Illas F. Adding pieces to the CO/Pt
promotion of platinum for catalytic light alkane dehydrogenation: insights into (111) puzzle: the role of dispersion. J Phys Chem C 2017;121:3970–7. https://doi.
geometric and electronic effects. ACS Catal 2017;7:4173–81. https://doi.org/ org/10.1021/acs.jpcc.7b00365.
10.1021/acscatal.6b03603. [62] Frenking G, Loschen C, Krapp A, Fau S, Strauss SH. Electronic structure of CO—An
[53] Wang T, Liang J, Zhao Z, Li S, Lu G, Xia Z, et al. Sub-6 nm fully ordered L10- exercise in modern chemical bonding theory. J Comput Chem 2007;28:117–26.
Pt–Ni–Co nanoparticles enhance oxygen reduction via Co doping induced https://doi.org/10.1002/jcc.20477.

13