Novel Pd-loaded urchin-like (NH4)xWO3/WO3 as an efficient visible-light-driven photocatalyst for partial conversion of benzyl alcohol

Article abstract of DOI:10.1016/j.jallcom.2020.156225

Photocatalytic oxidation of organic compounds by solar light is a promising strategy for environmentally benign conversion processes. Herein, we first report an urchin-like Pd/(NH₄)_xWO₃/WO₃ with three-dimensional hierarchical microstructure prepared through a thiourea-assisted solvothermal method followed by calcination under the reductive atmosphere. The Pd/(NH₄)_xWO₃/WO₃ exhibits higher selectivity in comparison to Pd/WO₃ nanorods in the partial conversion of aqueous benzyl alcohol into benzaldehyde (e.g., 80percent selectivity for benzaldehyde production with ca. 84percent conversion of benzyl alcohol) under visible light irradiation. The preferential generation of hydroxyl radicals from water on the (NH₄)_xWO₃/WO₃ surface in the initial reaction stage was responsible for the high selectivity for benzaldehyde production because hydroxyl radicals react with benzyl alcohol much more efficiently than benzaldehyde. DFT calculations demonstrate that the energy barrier of the reaction between benzyl alcohol and hydroxyl radicals largely decreased due to the photo-excited triplet states of (NH₄)_xWO₃/WO₃, while such a trend in the energy barrier was not observed for other photocatalysts. On the other hand, with extended irradiation time, a large amount of hydrogen peroxide, which was produced by the multi-electron reduction of oxygen molecules, accumulates on the Pd/(NH₄)_xWO₃/WO₃ due to the low activity of the Pd cocatalyst. The generated hydrogen peroxide preferentially eliminated the photoexcited holes when the concentrations of benzaldehyde increased, thus inhibiting the peroxidation of benzaldehyde by holes. The Pd/(NH₄)_xWO₃/WO₃ may provide a feasible strategy for the photocatalytic oxidation of various other organic compounds due to its unique reactivity shown in this study.

Full text of DOI:10.1016/j.jallcom.2020.156225

Journal of Alloys and Compounds 845 (2020) 156225
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: http://www.elsevier.com/locate/jalcom
Novel Pd-loaded urchin-like (NH4)xWO3/WO3 as an efficient visible-
light-driven photocatalyst for partial conversion of benzyl alcohol
,
,
Ying Lv a, Zhanglian Xu b *, Hisayoshi Kobayashi c **, Koji Nakane d
a College of Materials Science and Engineering, Xi’an Shiyou University, No.18, 2nd East Dianzi Road, Xian, Shaanxi, 710065, China
b School of Energy and Power Engineering, Xi’an Jiaotong University, No. 28, Xianning Road, Xi’an, Shaanxi, 710049, China
c Emeritus professor of Kyoto Institute of Technology, Matsugasaju, Sakyo-ku, Kyoto, 606-8585, Japan
d Frontier Fiber Technology and Science, Graduate School of Engineering, University of Fukui, Bunkyo 3-9-1, Fukui, 910-8507, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Photocatalytic oxidation of organic compounds by solar light is a promising strategy for environmentally
benign conversion processes. Herein, we first report an urchin-like Pd/(NH4)xWO3/WO3 with three-
dimensional hierarchical microstructure prepared through a thiourea-assisted solvothermal method
followed by calcination under the reductive atmosphere. The Pd/(NH4)xWO3/WO3 exhibits higher
selectivity in comparison to Pd/WO3 nanorods in the partial conversion of aqueous benzyl alcohol into
benzaldehyde (e.g., 80% selectivity for benzaldehyde production with ca. 84% conversion of benzyl
alcohol) under visible light irradiation. The preferential generation of hydroxyl radicals from water on the
(NH4)xWO3/WO3 surface in the initial reaction stage was responsible for the high selectivity for benz-
aldehyde production because hydroxyl radicals react with benzyl alcohol much more efficiently than
benzaldehyde. DFT calculations demonstrate that the energy barrier of the reaction between benzyl
alcohol and hydroxyl radicals largely decreased due to the photo-excited triplet states of (NH4)xWO3/
WO3, while such a trend in the energy barrier was not observed for other photocatalysts. On the other
hand, with extended irradiation time, a large amount of hydrogen peroxide, which was produced by the
multi-electron reduction of oxygen molecules, accumulates on the Pd/(NH4)xWO3/WO3 due to the low
activity of the Pd cocatalyst. The generated hydrogen peroxide preferentially eliminated the photoexcited
holes when the concentrations of benzaldehyde increased, thus inhibiting the peroxidation of benzal-
dehyde by holes. The Pd/(NH4)xWO3/WO3 may provide a feasible strategy for the photocatalytic oxida-
tion of various other organic compounds due to its unique reactivity shown in this study.
Received 24 February 2020
Received in revised form
24 June 2020
Accepted 26 June 2020
Available online 30 June 2020
Keywords:
Partial conversion
Pd loading
Pd/(NH4)xWO3/WO3
3D nanostructure
Heterogeneous photocatalysis
© 2020 Elsevier B.V. All rights reserved.
1. Introduction
[8,9]. Therefore, from the perspective of industrial applications, the
development of novel and efficient visible-light-responsive pho-
The conversion of primary and secondary alcohols into their
corresponding aldehydes and ketones with high selectivity using
semiconductor photocatalysts has been extensively studied in
chemical industries. It has been reported that TiO2-based photo-
catalysts were used to partially convert the alkyl-substituted benzyl
alcohols into their corresponding aldehydes in the presence of
molecular O2 as an oxidant [1e7]. However, due to the high pho-
tocatalytic activity of TiO2 under UV light irradiation, the nonse-
lective oxidation to CO2 easily occurred in the TiO2 photocatalyst
system, which led to the low selectivity for the desired products
tocatalysts for high conversion of the primary and secondary
alcohol with high selectivity for the desired products is imperative
but challenging [10,11].
Tungsten trioxide (WO3) is considered a good candidate for
visible-light-driven photocatalysis because it has many advantages,
including the wide absorption range in the solar spectrum, stable
physical and chemical properties, and long hole-diffusion length
[12,13]. It has also been proven that the photocatalytic efficiency is
highly related to structure, morphology, and unique surface char-
acteristics [14,15]. In particular, a three-dimensional (3D) hierar-
chical nanostructure with larger specific surface area and more
surface active sites and better permeability has tremendous po-
tential in electrochemical, optical and catalytic applications [16,17].
Therefore, considerable efforts have been made to synthesize
* Corresponding author.
** Corresponding author.
E-mail address: xuzhanglian@xjtu.edu.cn (Z. Xu).
https://doi.org/10.1016/j.jallcom.2020.156225
0925-8388/© 2020 Elsevier B.V. All rights reserved.

2
Y. Lv et al. / Journal of Alloys and Compounds 845 (2020) 156225
three-dimensional photocatalysts with unique structures to ach-
ieve high activities [18,19]. To date, most of the studies have focused
on the morphological control of WO3 in the construction of one-
dimensional nanostructures, including nanorods, nanotubes and
nanowires [20]. The design of a 3D hierarchical nanostructure on
WO3 is rarely reported.
2.2. Photocatalyst preparation
WCl6 (1 mmol, 0.397 g) and a required amount of PdCl2 (0.5 wt %
as metal) were first dissolved in 40 mL of absolute ethanol in the
beaker at the room temperature, respectively. After completely
dissolved, thiourea with different amounts (0-20 mmol) was added
into the above solution, and ultra-sonicated for 3 min. The mixture
solution was then transferred into a Teflon-lined autoclave, sealed
and maintained at 180 ꢀC for 8 h. After reaction, the precipitate was
collected and washed three times with absolute ethanol and
distilled water through centrifugation, respectively. After that, the
product was then dried at 60 ꢀC for 5 h. As a control samples were
prepared in the presence of different amount of PdCl2 (0.1, 0.5 and
1 wt %) and thiourea (5 mmol) with the other above conditions
unchanged.
The as-prepared dried product was subsequently transferred in
a 30 mL alumina crucible. Then, the 30 mL alumina crucible with
the as-prepared sample was placed into a 100 mL alumina crucible
containing activated carbon (3 g, Wako Pure Chemicals Ind., Ltd.,
Japan). The alumina cover was then put on the 100 mL crucible. This
reaction container was transferred into the electric furnace which
was elevated to a given temperature (450 ꢀC) in 4 h, then main-
tained at 450 ꢀC for 3 h. After that, the sample was naturally cooled
to room temperature and a light grey product was obtained for use.
The schematic diagram of the above reductive calcination experi-
ment is shown in Scheme S1.
On the other hand, the relatively lower conduction band (CB)
edge (approximately þ0.3e0.5 V vs. NHE) than the reduction po-
tentials of O2 severely limits the photocatalytic applications of WO3,
which can largely decrease the photocatalytic efficiency owing to
the rapid recombination of the photogenerated electrons and holes
[21]. Semiconductor coupling [22,23] and noble metal deposition
(Au, Pt and Pd) [24e30] on WO3 were reported to solve the above
problems because the introduction of other semiconductors or
noble metal nanoparticles favor the effective separation of the
photogenerated electrons and holes and thus enhance the photo-
catalytic efficiency. In addition to the above two methods, the
construction of photocatalysts with mixed phases has also been
investigated to promote the separation of electron-hole pairs. For
instance, enhanced photocatalytic performance has been demon-
strated in TiO2 photocatalysts with mixed phases, including
anatase/rutile [31e34], anatase/brookite [35,36], anatase/TiO2(B)
[37], and rutile/brookite [38]. Recently, Bi2O4-x/Bi2O3 [39], In2O3/
In2O3$3H2O [40] and WO3/WO3$H2O [41] photocatalysts with
mixed phases have also been developed. To the best of our current
knowledge, no study has yet been found that describes a facile and
effective method that can in situ synthesize the 3D hierarchical
structural WO3-based photocatalysts with mixed phases.
2.3. Characterization
In this study, we proposed the first report of a facile synthesis of
Pd-loaded urchin-like (NH4)xWO3/WO3 photocatalysts with 3D
hierarchical structure and mixed phases through thiourea-involved
alcoholysis reaction of WCl6 and PdCl2, and subsequent reductive
calcination. The realization of the direct construction of such a
photocatalyst is based on the following inspirations. Choi et al.
synthesized tungsten oxide (W18O49) nanorods through the hy-
drothermal reaction of WCl6 with ethanol in an autoclave [42]. This
synthetic process was considered to obey classical self-aggregation
and Ostwald ripening growth mechanisms. It was also reported by
Custelcean that thiourea could make organic crystals assemble
effectively into nanostructures, including one dimensional chains
and even the spatial structure, because thiourea has strong
hydrogen-bonding properties [43]. It therefore is feasible that
thiourea could be a suitable directing ligand for the construction of
inorganic WO3-based nanostructures in an ethanol/WCl6 system.
We further investigated the application of the as-prepared visible-
light-responsive Pd-loaded (NH4)xWO3/WO3 photocatalyst in the
partial conversion of benzyl alcohol solution in the presence of the
dissolved O2 as the oxidant. DFT calculations for the reaction
mechanism between benzyl alcohol and active radicals were car-
ried out. The results showed the efficient partial oxidation of benzyl
alcohol with high selectivity for benzaldehyde production was
obtained over the Pd/(NH4)xWO3/WO3 photocatalyst under
illumination.
Field-emission scanning electron microscopy (FE-SEM, Zeiss
Oberkochen, Germany) and high-resolution transmission electron
microscopy (HR-TEM, JEM-2010F, JEOL, Japan) were used to observe
the morphology of as-prepared products. X-ray powder diffraction
(XRD, D8 ADVANC, Germany) using a CuKa with a Ni filter (40 kV,
40 mA) was used to confirm the phases of the as-prepared samples.
The chemical states of the as-prepared products were investigated
by X-ray photoelectron spectroscopy (XPS, JEOL, JPS-9010MCY). The
specific surface areas of the as-prepared samples were character-
ized using N2 adsorption at ꢁ196 ꢀC on a specific surface area and
porosity analyser (Micromeritics, TRYSTAR 3000, Japan). The band
energy information of the as-prepared samples was measured by
UV-Vis diffuse reflectance spectroscopy (Shimadzu, UV-2400PC/
2500 PC, Japan).
2.4. Photocatalytic reaction
A photocatalytic reactor was used in this study as shown in
Scheme S2. Xe lamp (500 W, 300 <
l < 600 nm, light intensity
600
m
W∙cmꢁ2) was located in the centre of the photocatalytic
reactor along the axis and protected by a water-cooled quartz
jacket. A magnetic stirrer at the bottom of the reactor was used to
achieve effective dispersion by mechanical agitation. A circular test
tube rack was inserted in the thermostatic bath to hold up the Pyrex
glass tubes. Thus, the visible light (equipped with cutoff filter (L42,
Hoya) between the Xe lamp and Pyrex glass tubes was collected
into the glass tube and ensured that the photocatalytic reaction
occurred uniformly and completely. The reactor was fitted with a
magnetic stirrer for stirring at 700 rpm to keep the catalyst in
suspension.
For the photocatalytic oxidation of benzyl alcohol, the as-
prepared photocatalysts (150 mg) were added in aerated aqueous
benzyl alcohol (100 mL, initial amount 130 mmol). The reaction cell
was tightly sealed with a rubber cap and then was irradiated by the
Xe lamp equipped with a cutoff filter (L42, Hoya). Sample aliquots
were taken out of the reactor cell at given irradiation time intervals
2. Experimental section
2.1. Materials
The raw materials, including tungsten (VI) chloride (WCl6),
palladium chloride (PdCl2) and thiourea (CS(NH2)2) were obtained
from Wako Pure Chemicals Ind., Ltd., Japan. Benzyl alcohol is
commercially available and was used directly without further pu-
rification. The solvents used in this study are distilled water and
absolute ethanol. All other reagents were of analytical grade and
used without further retreatment.

Y. Lv et al. / Journal of Alloys and Compounds 845 (2020) 156225
3
and then centrifuged for analysis. Sample analysis was carried out
by Gas Chromatography (GC7890II, Techcomp) equipped with a
Flame Ionization Detector (FID) for the quantitative identification of
benzyl alcohol and product concentration in solution. The released
CO2 was measured using a gas chromatograph (Align model GC-
6890 N) equipped with a 2 m Porapak-Q column, a methanizer, and
a flame ionization detector, using N2 as the carrier gas. Oxalic
acid(C2H2O4) and isopropanol were added to the photocatalytic
oxidation system in order to capture the active species holes and
ꢂOH in the oxidation process of benzyl alcohol [44,45]. The con-
version efficiency is defined as the molar percentage of consumed
reactant, while the selectivity of the desired product is determined
as the molar percentage of the desired product out of total prod-
ucts. The conversion of benzyl alcohol and the selectivity for
benzaldehyde were defined as follow: Conversion rate (%)¼(C0-
C
BA)/C0 ꢃ 100; Selectivity (%) ¼ CBDA/(C0-CBA) ꢃ 100.Where C0 is the
initial concentration of benzyl alcohol and CBA and CBAD are the
concentration of the substrate benzyl alcohol and the corre-
sponding benzaldehyde, respectively.
2.5. Method of calculation
DFT calculations were carried out using a plane-wave-based
programme, Castep [46,47]. The PerdewꢁBurkeꢁErnzerhof func-
tional [48,49] was used together with the ultrasoft-core potentials
[50]. Semi-empirical dispersion interaction correction was included
based on McNellis et al. [51]. The basis set cut-off energy was set to
300 and 340 eV for geometry optimization and post-energy eval-
uation. Explicitly considered electrons are H: 1s1, C: 2s22p2, N:
2s22p3, O: 2s22p4, and W: 5s25p65d46s2. Among materials con-
sisting of NH4 and WO3 units, a structure with hexagonal symmetry
was adopted (ICSD #23537). The space group is P 63/mcm, and the
lattice constants are a ¼ b ¼ 7.388 Å, c ¼ 7.551 Å,
a
¼
b
¼ 90ꢀ and
Fig. 1. XRD patterns of (a) monoclinic W18O49 nanorod (b) monoclinic WO3 nanorod
after air calcination, (c) PdO-loaded (NH4)xWO3/W18O49 before calcination and (d) Pd-
loaded (NH4)xWO3/WO3 after reductive calcination. (The amounts of thiourea and Pd
in solvethermal reaction are 380 mg and 0.5 wt%, respectively).
g
¼ 60ꢀ. To make slab models, the direction along the c-axis was
cleaved. The lattice constant was set to c ¼ 30 Å, including the
vacuum region. In calculations of the catalytic reaction, all the
atomic coordinates were optimized, and the lattice constants were
fixed as the original values.
formation of the (NH4)xWO3 phase. We further investigated the
effect of the addition amount of thiourea on the phase formation of
the final products after reductive calcination. As shown in Fig. S1,
the products with mixed phases can only form when the added
amount of thiourea is low. With increasing the amount of thiourea
in the solvothermal reaction, the (NH4)xWO3 phase increased first
and then decreased and finally was not generated for a high addi-
tion amount.
3. Results and discussion
3.1. Structure and morphology
Fig. 1 shows the X-ray diffraction (XRD) patterns of the as-
prepared samples obtained from the solvothermal reaction
without adding thiourea and PdCl2 (a, b) as well as in the presence
of thiourea and PdCl2 (c, d). The product obtained in the absence of
thiourea before calcination (Fig. 1a) was identified as monoclinic
X-ray photoelectron spectra (XPS) was carried out to examine
the chemical compositions and their corresponding valence states
for Pd-loaded (NH4)xWO3/WO3. In Fig. 2a, the fully scanned spec-
trum shows the presence of Pd, N, W, O and C in the measured
sample. The peak of C in the sample may originate from the carbon
tape. The XPS peak of N1s (Fig. 2b) located at 402.1 eV comes from
the NH4þ group in (NH4)xWO3. For tungsten, the corresponding W
4f core level XPS spectrum of Pd-loaded (NH4)xWO3/WO3 is shown
in Fig. 2c. After deconvolution, the W 4f core level spectra were well
fitted into two spin-orbit doublets groups which corresponded to
two different oxidation states of W atoms. The main peaks located
W18O49 (JCPDS Card No. 05-0392) with diffraction peaks of the
(010) and (020) planes, respectively. After air calcination, the
monoclinic W18O49 totally converted into monoclinic WO3 (JCPDS
Card No. 43-1035) with the characteristic diffraction peaks of the
(001), (120), (220), (221) and (140) planes, shown in Fig. 1b. While
in the presence of thoiurea and PdCl2, the XRD pattern of the as-
obtained product before calcination (Fig. 1c) shows the coexis-
tence of the monoclinic W18O49 and hexagonal (NH4)xWO3 (JCPDS
Card No. 42-0452) with diffraction peaks of (101), (002), (200),
(202) and (220). After calcination under a reductive atmosphere for
1 h, the XRD pattern in Fig. 1d matched the main mixed phases of
the monoclinic WO3 and (NH4)xWO3, suggesting the monoclinic
at 35.4 eV and 37.5 eV are assigned to W 4f7/2 and W 4f5/2 of W6þ
respectively. The lower binding energies of 34.2 eV and 36.3 eV in
the second doublet could be assigned to W 4f7/2 and W 4f5/2 of W5þ
,
,
respectively. These results regarding the W 4f core level are
consistent with the typical nature of tungsten bronze (MxW6þ
W18O49 and a part of (NH4)xWO3 (based on the intensity of their
1-
xW5þxO3). In contrast, only one group of spin-orbit doublets in a
hexavalent oxidation state was fitted in the W 4f core level spec-
trum from the monoclinic WO3 after air calcination (Fig. S2). The
characteristic peaks) were transformed into monoclinic WO3.
Meanwhile, a small peak appeared at 40.21ꢀ, which was indexed
the face-centered cubic (fcc) phase of Pd (Fig. 1d). Based on the
above results, thiourea was involved in the reaction, leading to the

4
Y. Lv et al. / Journal of Alloys and Compounds 845 (2020) 156225
Fig. 2. XPS spectra of Pd loaded (NH4)xWO3/WO3 (thiourea: 5 mmol, Pd 0.5 wt%): (a) full range XPS spectra; (b) XPS of N1s; (c) deconvolution of W4f core-level spectrum with
peaks corresponding to W5þ and W6þ; (d) XPS spectra of Pd 3d.
high-resolution XPS spectrum (Fig. 2d) of Pd 3d is dominated by
two components. The main peaks located at 335.5 eV and 340.6 eV
are assigned to Pd 3d5/2 and Pd 3d3/2 of the metallic Pd0 state,
respectively. The higher binding energies of 336.7 eV and 341.8 eV
in the second doublet could be assigned to Pd 3d5/2 and Pd 3d3/2 of
PdO, respectively. Therefore, the states of Pd species on the surface
of the as-prepared product under a CO-containing atmosphere
were found to be dominantly the metallic Pd0 state with a relatively
low amount of the oxide state also judging from the peak position
in reference to Pd and PdO (Fig. 2d). Based on the above results, we
can determine the chemical composition of the synthesized prod-
uct as Pd-loaded (NH4)xWO3/WO3.
The scanning electron microscope (SEM) images and trans-
mission electron microscope (TEM) images of Pd-loaded
(NH4)xWO3/WO3 obtained after reductive calcination are shown
in Fig. 3. The Pd-loaded urchin-shaped (NH4)xWO3/WO3 micro-
spheres were observed in Fig. 3a. These microspheres had di-
protons. Meanwhile, WCl6 can generate protons during its alco-
holysis process. These protons may adsorb on the as-prepared
nanoparticle surfaces during the growth process and have an
interaction with thiourea. With their stong and predictable
hydrogen-bonding, N, N0-disubstituted thiourea are robust and
versatile groups that can be rationally employed for the synthesis
of crystalline solids with three-dimensional urchin-like frame-
works. Meanwhile, thiourea used as an ammonium source partic-
ipates in the crystal growth, thereby tuning the unique morphology
and forming the multicomponent nature of the final products. In
contrast, the pure WO3 obtained in the absence of thiourea and
PdCl2 showed the scattered nanorod-like structure in Fig. S3a,
which is the same as the results from previous report of Choi et al.
[42]. These results clearly indicate that the introduction of thiourea
has a significant effect on the morphological evolution.
More details of the nanostructural characteristics of the as-
prepared Pd-loaded urchin-like (NH4)xWO3/WO3 particles were
studied by TEM measurements. Fig. 3d gives the TEM image of an
intercrossed urchin-like (NH4)xWO3/WO3, also showing that
numerous nanorods grew pointing towards the centre of the
sphere. From the TEM image of an individual nanorod in Fig. 3e, we
can see that the (NH4)xWO3/WO3 nanorod has the typical diameter
of ca. 50 nm. Meanwhile, this nanorod was covered with a uniform
distribution of ~10 nm Pd nanoparticles. With clear observation
from a HRTEM image taken from a part of the nanorod shown in
ameters in the range of 1e3 mm and were interconnected to each
other. The high-magnification SEM (Fig. 3b and c) images show that
each urchin-like microsphere was assembled by nanorods with
diameters of ca. 50 nm. These nanorods radiated from the centre of
the sphere, giving them an urchin-shaped structure. In this process,
thioureas typically exist in solution as trans-cis rotamer [52]. It can
function as a hydrogen-bond acceptor through the lone pairs of the
C¼S group or act as a hydrogen-bond donor through its two NH

Y. Lv et al. / Journal of Alloys and Compounds 845 (2020) 156225
5
Fig. 3. FE-SEM images (aec) and TEM images (def) of Pd-loaded (NH4)xWO3/WO3 after reductive calcination. (The amounts of thiourea and Pd loading amount in solvethermal
reaction are 380 mg and 0.5 wt%, respectively).
Fig. 3f, the nanorod was confirmed to be polycrystalline. The lattice
fringes were calculated to be approximately 0.38 nm and 0.64 nm,
corresponding to the (100) and (001) planes of spinel-structured
monoclinic WO3 and (NH4)xWO3, respectively. The lattice fringes
of Pd nanoparticles ca. 0.46 nm corresponding Pd (111) plane was
marked in HRTEM image as shown in Fig. S3b.
The thermal behavior of Pd-loaded (NH4)xWO3/WO3 obtained in
the presence of thiourea (380 mg) and Pd (0.5 wt%) under reductive
conditions is difficult to determine directly. Herein, we tried to
measure the change in the mass of Pd-loaded (NH4)xWO3/WO3 by
TGA to estimate the thermostability in samples. Fig. 4 showed that
the weight loss of (NH4)xWO3/monoclinic W18O49 from room
temperature to 100 ꢀC and up to approximately 230 ꢀC could be
attributed to the loss of surface-adsorbed water and structural
water, respectively. After that, the weight loss in the range of
230e550 ꢀC can be explained by the decomposition of (NH4)xWO3
and W18O49. There is almost no weight loss above 550 ꢀC. In
addition, the final product after the TG measurement was indexed
Fig. 4. (a) TG and DTA spectra of Pd-loaded (NH4)xWO3/WO3.
as a monoclinic WO3 crystal as measured by XRD (data no shown),
indicating the complete decomposition of (NH4)xWO3 into WO3

6
Y. Lv et al. / Journal of Alloys and Compounds 845 (2020) 156225
upon calcination at high temperature.
Fig. 5 shows the diffuse reflectance UV-vis absorption spectra of
the photocatalysts. The Pd-loaded (NH4)xWO3/WO3 showed
increased absorbance above 400 nm compared to the product
before calcination due to the light scattering by the Pd nano-
particles. The PdO/(NH4)xWO3/W18O49 and Pd-loaded (NH4)xWO3/
WO3 showed bathochromic shift compared to the WO3 due to the
mixed chemical valence tungsten ions (W6þ and W5þ) which
exhibited excellent visible absorption properties [53]. The bandgap
of Pd-loaded (NH4)xWO3/WO3 was calculated to be 2.5 eV (inset),
suggesting that the as-prepared product can respond effectively
above 400 nm in the visible light range.
3.2. Photocatalytic activity
Fig. 6 shows the photocatalytic conversion process of benzyl
alcohol over a Pd/(NH4)xWO3/WO3 sample in an aerated aqueous
benzyl alcohol (PhCH2OH) solution under visible light illumination
(420 <
l < 600 nm). At the initial stage of light illumination, the
Fig. 6. Time course of photocatalytic oxidation of benzyl alcohol over Pd/(NH4)xWO3/
WO3 photocatalyst in aerated aqueous solutions of benzyl alcohol under visible light
irradiation (amount of photocatalyst, 150 mg; amount of cocatalyst, 0.5 wt%; irradia-
amount of benzyl alcohol gradually decreased with benzaldehyde
(PhCHO) production in the solution. The benzaldehyde increased
with a relatively steady generation rate and, after illumination for
9 h, the generation of benzaldehyde tended to be saturated. It is
interesting to observe that the concentration of benzaldehyde was
not significantly reduced in a longer period of irradiation (up to
18 h), and the production of CO2 was negligible and out of the range
of detection. On the basis of the amount of benzyl alcohol
consumed in the initial period (3 h), the selectivity for benzalde-
hyde was calculated to be 80% for the Pd (0.5 wt%)/(NH4)xWO3/WO3
photocatalyst when benzyl alcohol was at a low conversion of 27%.
Extending the time of light irradiation up to 18 h, the conversion
percentage of benzyl alcohol reached up to 84%, and the selectivity
for benzaldehyde was still high (ca. 80%). In contrast, the selectivity
for benzaldehyde production (58%) at the conversion (66%) of
benzyl alcohol over Pd (0.5 wt %)/WO3 nanorods after light irradi-
ation of 18 h is shown in Fig. S4a. However, in the absence of Pd
loading, much lower conversion of benzyl alcohol over (NH4)xWO3/
WO3 (Fig. S4b) and WO3 nanorods (Fig. S4a) was observed, sug-
gesting the enhanced effect of Pd cocatalysts on the conversion of
benzyl alcohol.
tion wavelength, 420 <
l < 600 nm).
3.3. Mechanism and discussion
Here, we propose an explanation of the relatively high selec-
tivity for benzaldehyde production from the conversion of benzyl
alcohol in the aqueous solution over Pd/(NH4)xWO3/WO3 under
visible light irradiation. It has been reported that the Pd/WO3 sys-
tem showed high photocatalytic activity towards gaseous organic
pollutants, including aldehydes and ketones, resulting in the
complete decomposition of the gaseous organic molecules into CO2
due to the activity of the Pd cocatalyst [30]. This means that Pd/
(NH4)xWO3/WO3 should have intrinsic oxidation capacity of benz-
aldehyde to CO2, but the activity was largely reduced in aqueous
media. It is thus that one of the possible reasons why Pd/
(NH4)xWO3/WO3 became inert in aqueous solution accompanied
by the high selectivity of benzaldehyde production might be the
shielding of the activity of Pd nanoparticles by the generated
benzaldehyde molecules adsorbed to the Pd surface. To verify this,
we investigated the photocatalytic oxidation of the initial mixed
solution containing benzyl alcohol and benzaldehyde over the Pd/
(NH4)xWO3/WO3 photocatalyst, as shown in Fig. S5. The initial
presence of benzaldehyde did not affect the photocatalytic oxida-
tion of benzyl alcohol, which was almost the same rate as that in
the absence of benzaldehyde. This indicated that Pd cocatalyst-
induced O2 reduction was not suppressed by the presence of
benzaldehyde in the solution. Otherwise, the high selectivity of
benzaldehyde production from the high conversion of benzyl
alcohol could not be obtained because the rapid recombination of
the photogenerated electrons and holes would occur, leading to the
low activity of the Pd/(NH4)xWO3/WO3 photocatalyst. Therefore,
we infer that the relatively high selectivity of benzaldehyde pro-
duction could be attributed to the results of the balance between
the active species produced by the photogenerated holes on the
photocatalyst surface and the active species produced by O2
reduction on the Pd surface.
It has been extensively reported that the photocatalytic oxida-
tion reaction of organic compounds on semiconductor photo-
catalysts proceeds by the following two main ways: the first is the
direct oxidation by the photoexcited holes (hþ) on the photo-
catalyst surface; another is the indirect oxidation by the generated
active oxygen species, including ꢂOH radicals, ꢂO2ꢁ radicals or HO2ꢂ
Fig. 5. UV-vis diffuse reflectance spectra of the samples; the inset is the calculation of
the bandgap of Pd/(NH4)xWO3/WO3 based on the (
thiourea and Pd in solvethermal reaction are 380 mg and 0.5 wt%, respectively).
ahn) n curve. (The amounts of
1/2-h

Y. Lv et al. / Journal of Alloys and Compounds 845 (2020) 156225
7
radicals. These are generated following Eqs. (1)e(3), respectively:
singlet state (left) and in the triplet state (right). The O-O bond of
HOOH is slightly stretched, and HOOH is loosely bonded to the W
atom in the triplet state compared to the singlet state. This suggests
that photo irradiation favours generating ꢂOH species. To gain more
insights into the reaction mechanism, the reaction of benzyl alcohol
and HOOH was further investigated and compared between the
singlet and triplet potential energy surfaces (PES), as shown in Fig. 7.
The optimized structures for the co-adsorption of HOOH and
PhCH2OH on the (NH4)2(WO3)9(OH)2 unit cell, first transition state
(TS1) for PhCH2OH þ HOOH / PhCH2O þ H2O þ ꢂOH reaction, and
the intermediate composed of PhCH2O, H2O and ꢂOH in the singlet
H2O þ hþ/Hþ þ ꢂOH
O2þeꢁ/ꢂO2ꢁ
(1)
(2)
(3)
O2þHþþeꢁ/HO2ꢂ
We noticed that the TiO2 photocatalyst is seldom used in the
conversion of alcohols unless the photocatalyst system is precisely
designed for this purpose. This is because sequential oxidation is
prone to occur by the strong oxidation of the photogenerated holes
and thus reduces the selectivity of the desired products. However,
in the present case, the affinity of the WO3 surface for organic
compounds is much lower than that of TiO2 [54], which reasonably
may reduce the possibility of direct oxidation of the desired prod-
ucts by photoexcited holes. As evidenced in Fig. S5, benzaldehyde
was selectively generated over Pd/(NH4)xWO3/WO3 without
releasing CO2 even when benzaldehyde reached a high concen-
tration, indicating that the direct oxidation by the photoexcited
holes did not occur on the surface of the Pd/(NH4)xWO3/WO3
photocatalyst. It is thus that the initial reaction started mainly by
the indirect oxidation of ꢂOH radicals or other active oxygen species
(ꢂO2ꢁ or HO2ꢂ).
On the other hand, since the minimum value of the WO3 con-
duction band (approximately 0.3e0.5 V vs. NHE) is much lower
than the O2 reduction potential by the single electron reduction
process, the generation of HO2ꢂ and ꢂO2ꢁ can thus be excluded in
this study, which are generally produced through one-electron
reduction processes. Therefore, at the initiation reaction, the pho-
tocatalytic oxidation of benzyl alcohol by ꢂOH radicals produced by
the reaction of water molecules with the photogenerated holes was
dominant. Meanwhile, based on the results of the preferential
oxidation test of the mixed solution containing benzyl alcohol and
benzaldehyde, as shown in Fig. S5, the lower activity of ꢂOH radicals
with respect to benzaldehyde than benzyl alcohol was demon-
strated. This may be responsible for the high selectivity of benzal-
dehyde production from the high conversion of benzyl alcohol in
the initial stage of light irradiation. However, with extended irra-
diation time, when most of benzyl alcohol was consumed, the
generated benzaldehyde may be further oxidized by the formed
ꢂOH radicals due to their strong oxidation ability. Conversely, the
selectivity for benzaldehyde was still high (ca. 80%) at the conver-
sion percentage of benzyl alcohol of 84% after light irradiation for
18 h (Fig. S5). Therefore, the formation of ∙OH radicals by photo-
generated holes may be inhibited when benzyl alcohol achieves
high conversion.
A series of active species trapping experiments were carried out
under the above mentioned photocatalytic experimental condi-
tions by the addition of oxalic acid (OA) as hole scavengers and
isopropyl alcohol (IPA) as ꢂOH scavengers in the solution. As shown
in Fig. S6, when OA and IPA were respectively added to the benzyl
alcohol system to trap the holes and ꢂOH. The conversion of benzyl
alcohol decreased significantly, revealing that the photogenerated
holes and ꢂOH are the major oxidative species for the present
photocatalytic oxidation reaction since photogenerated holes
indirectly oxidized benzyl alcohol through ꢂOH.
Fig. 7. Optimized structures for co-adsorption of HOOH and PhCH2OH on the
(NH4)2(WO3)9(OH)2 unit cell as the reactant in the singlet (a) and triplet states (d), the
first TS (TS1) for PhCH2OH þ HOOH/ PhCH2O þ H2O þ ∙OH reaction in the singlet (b)
and triplet states (e), the intermediate composed of PhCH2O, H2O and ∙OH in the
singlet (c) and triplet states (f). Lower part of (NH4)2(WO3)9(OH)2 unit cell is omitted
for clarity. Energy change along the reaction (g). Medium level optimization energy
(red) and fine level post energy calculation (blue). Plots in reaction coordinate are co-
adsorbed reactant: 1, TS1: 2, intermediate: 3, TS2: 4, and product: 5. The triplet state
calculation was conducted only for the former half of reaction (1e3), and shown by
dot-dash line. (For interpretation of the references to colour in this figure legend, the
reader is referred to the Web version of this article.)
DFT calculations were carried out to study the reaction of benzyl
alcohol and ꢂOH radicals. To accurately simulate the reaction
mechanism of the composite (NH4)xWO3/WO3 photocatalyst, a
(NH4)2(WO3)9(OH)2 model was employed. The reason for choosing
the (NH4)2(WO3)9(OH)2 model is demonstrated in Fig. S7 and Fig. S8.
All the atomic coordinates were optimized, and the lattice constants
were fixed to the original values. Fig. S9 shows the optimized
structures for HOOH adsorption onto (NH4)2(WO3)9(OH)2 in the

8
Y. Lv et al. / Journal of Alloys and Compounds 845 (2020) 156225
Fig. 8. Proposed reaction for benzaldehyde production from benzyl alcohol induced by photoexcited electrons and holes over Pd/(NH4)xWO3/WO3 photocatalysts.
state are shown in Fig. 7a-c, respectively. The acidic proton of
PhCH2O-H was first eliminated by ꢂOH from HOOH, and H2O was
generated. Meanwhile, Fig. 7(def) show the corresponding opti-
mized structures in the triplet state. Bycomparison, the difference in
structures seems small between the singlet and triplet states.
However, the difference in energetics was more significant, as
shown in Fig. 7g. Medium level optimization energy and fine level
energy determined by the post calculation behave in parallel. In the
reaction of PhCH2OH þ HOOH / PhCH2O þ H2O þ ꢂOH, there is an
energy barrier for the singlet PES. However, if excited to the triplet
state by photo irradiation, the energy barrier disappeared, sug-
gesting a favourable (downhill) reaction for PhCH2OH þ HOOH /
PhCH2O þ H2O þ ꢂOH. This change between the singlet and triplet
states, which has not yet been observed in other photocatalyst sys-
tems, clearly demonstrates the advantage of our designed photo-
catalyst. Fig. S10 shows the optimized structures for the second TS
(TS2) for the PHCH2O þ∙OH /PhCHO þ H2O reaction and for the
product (PhCHOand 2H2O). This latterhalfreactionwas investigated
on the singlet PES since quenching seems to occur during the former
half reaction (PhCH2OH þ HOOH /PhCH2O þ H2O þ ꢂOH). DFT
calculations are highly consistent with the experimental results and
clearly demonstrate the reaction mechanism between PhCH2OH
and ꢂOH radicals.
H2O2 þ 2hþ / O2 þ 2Hþ
(5)
Thus, with prolonged reaction time, the photogenerated holes
were preferentially consumed by a portion of the H2O2 produced by
the multielectron O2 reduction on Pd cocatalyst to release O2. This
therefore decreased the rate of ꢂOH production from H2O mole-
cules, consequently keeping high selectivity for benzaldehyde until
the maximum conversion of benzyl alcohol. The effect of Pd loading
amount on the conversion of benzyl alcohol and the selectivity for
benzaldehyde production was also investigated. When the Pd
loading amount on (NH4)xWO3/WO3 was increased from 0.1, 0.5 to
1 wt %, the conversion percentages of benzyl alcohol increased
(63%, 84% and 86%) at the initial irradiation of 9 h. However, the
corresponding selectivity for benzaldehyde production (85%, 80%,
76%) decreased slightly when the Pd loading amount increased.
This demonstrated that the increasing amount of Pd loading
effectively promoted O2 reduction by the multi-electron reaction,
thereby increasing the oxidation rate of benzyl alcohol. However,
with the increase of the Pd loading amount, the selectivity of
benzaldehyde production in the high conversion of benzyl alcohol
decreased, indicating that the rate of photoexcited holes increased,
and a certain degree of undesirable benzaldehyde peroxidation was
induced. Finally, the reusability and stability of the used photo-
catalysts were investigated. The almost unaffected conversion of
benzyl alcohol and selectivity of benzaldehyde production obtained
under visible light irradiation of 18 h after five cycles as shown in
Fig. S12, suggesting the good reusability of Pd/(NH4)xWO3/WO3. In
addition, the morphology of photocatalyst after five reuse cycles
still maintained the urchin-like structure (Fig. S13), showing good
stability of the as-prepared photocatalysts.
Therefore, we would like to make the following conclusions
regarding the photocatalytic oxidation of benzyl alcohol over Pd/
(NH4)xWO3/WO3 in Fig. 8: at the initiation of the reaction, the
photoexcited holes preferentially participated in an oxidation re-
action with water molecules, and ꢂOH radicals were generated. The
generated ꢂOH radicals preferentially attacked benzyl alcohol due
to the inert reactivity of the (NH4)xWO3/WO3 catalyst surface,
leading to a stable rate of benzaldehyde production without
releasing CO2 until the max conversion percentage of benzyl
alcohol. Meanwhile, the photogenerated electrons were consumed
via multielectron O2 reduction on Pd to form H2O2. However, along
with the amount of benzaldehyde reached at the maximum, some
accumulated H2O2 began to be consumed by holes due to the lower
activity of the Pd cocatalyst than Pt. When reaching the max
amount of benzaldehyde, the photogenerated hþ preferentially was
consumed by H2O2 into O2; meanwhile, the photogenerated elec-
trons reacted with O2 to generate H2O2. Thus, such a cycle of re-
action between H2O2 and O2 largely suppressed the peroxidation of
benzaldehyde to CO2, resulting in benzaldehyde generation from
benzyl alcohol with relatively high selectivity. Finally, we believe
It has been previouslyelucidated that the presence of a nanoscale
Pd cocatalyst on WO3 facilitated the multielectron O2 reduction by
photogenerated electrons (O2 þ 2Hþ þ2eꢁ / H2O2, þ0.682 V) as
well as Pt [55]. During the photocatalytic oxidation of benzyl alcohol
over the Pd/(NH4)xWO3/WO3, H2O2 began to form via the reaction
between the accumulated photogenerated electrons on the Pd sur-
face and dissolved O2 molecules. In the Pt/WO3 system, active Pt
metal worked as an electron magnet, as reported by Abe et al. [30]
The produced H2O2 was quickly decomposed according to Eq. (4)
because H2O2 acted as an efficient electron scavenger.
H2O2 þ 2Hþ þ2eꢁ/2H2O, þ1.76V
(4)
This therefore increased the amounts of photoexcited holes,
leading to direct oxidation of organic compounds by holes on the
WO3 surface [53]. However, Pd showed much lower activity than Pt
[56]. Therefore, compared to the H2O2 decomposition in Pt-loaded
WO3, the accumulation of H2O2 generated by multielectron O2
reduction on Pd/(NH4)xWO3/WO3 seemed to be reasonable due to
the low activity of Pd. This can be demonstrated by the initial
photocatalytic reaction of the solution of benzaldehyde and H2O2
over Pd/(NH4)xWO3/WO3 in Fig. S11. H2O2 addition only slightly
affected the reaction on Pd/(NH4)xWO3/WO3. This result demon-
strated that the photocatalyst generated H2O2 in the process and
benzyl alcohol oxidation consumed the H2O2. It is also well known
that H2O2 molecules can react with holes in certain photocatalytic
processes by Eq. (5).

Y. Lv et al. / Journal of Alloys and Compounds 845 (2020) 156225
9
[2] S. Abedi, A. Morsali, ACS Catal. 4 (2014) 1398e1403.
[3] S. Higashimoto, R. Shirai, Y. Osano, M. Azuma, H. Ohue, Y. Sakata,
H. Kobayashi, J. Catal. 311 (2014) 137e143.
[4] P. Zhang, Y. Liu, B.Z. Tian, Y. Luo, J.L. Zhang. Catal. Today 281 (2017) 181e188.
[5] P. Zhang, P. Wu, S. Bao, Z. Wang, B. Tian, J.L. Zhang, Chem. Eng. J. 306 (2016)
1151e1161.
that the merits of the 3D hierarchical nanostructure with a mixed
phase of Pd/(NH4)xWO3/WO3 increased the rate of photocatalytic
conversion of benzyl alcohol compared to that over the Pd/WO3
nanorods, as shown in Fig. S4.
[6] Y. Liu, P. Zhang, B.Z. Tian, J.L. Zhang, ACS Appl. Mater. Interfaces 7 (2015)
13849e13858.
4. Conclusion
[7] Y. Liu, P. Zhang, B.Z. Tian, J.L. Zhang, Catal. Commun. 70 (2015) 30e33.
[8] T.D. Bui, A. Kimura, S. Ikeda, M. Matsumura, Appl. Catal.,
B 107 (2011)
We have successfully prepared Pd-loaded urchin-like (NH4)xWO3/
WO3 with a 3D nanostructure and mixed phases through the
thiourea-assisted alcoholysis of WCl6 and PdCl2 and subsequent
reductive heat treatment. The presence of the thiourea played a
critical role in the morphologic evolution and composition of the final
product. In comparison to Pd/WO3 nanorods, the Pd-loaded
(NH4)xWO3/WO3 exhibited relatively high selectivity for the partial
conversion of benzyl alcohol to the corresponding benzaldehyde in
aqueous solution containing O2 as the oxidant under visible light
irradiation. The high conversion of benzyl alcohol and relatively high
selectivity of benzaldehyde productionwere attributed to the affinity
of Pd/(NH4)xWO3/WO3 towards organic compounds being relatively
low on the surface, leading to the initially selective conversion of the
benzyl alcohol by ꢂOH radicals. Meanwhile, the accumulated H2O2
proceeded through a two-electron O2 reduction by the Pd species due
to its relatively low activity. When the max amount of benzaldehyde
wasachieved, theaccumulatedH2O2 begantoconsumethegenerated
holes on the photocatalysts, thus suppressing the peroxidation of the
benzaldehyde. Moreover, the 3D nanostructure of the Pd-loaded
(NH4)xWO3/WO3 also increased the oxidation reaction rate of
benzyl alcohol compared to that over the Pd/WO3 nanorods.
119e127.
[9] X. Pang, W. Chang, C. Chen, H. Ji, W. Ma, J. Zhao, J. Am. Chem. Soc. 136 (2014)
8714e8721.
[10] K. Maeda, K. Teramura, D. Lu, T. Takata, N. Saito, Y. Inoue, K. Domen, Nature
440 (2006), 295-295.
[11] X. Sun, X. Luo, X. Wu, Y. Xie, J. Am. Chem. Soc. 141 (2019) 3797e3801.
[12] D. Chandra, K. Saito, T. Yui, M. Yagi, ACS Sustain. Chem. Eng. 6 (2018)
16838e16846.
[13] D. Chandra, D. Li, T. Sato, Y. Tanahashi, T. Togashi, M. Ishizaki, M. Kurihara,
E.A. Mohamed, Y. Tsubonouchi, Z.N. Zahran, K. Saito, T. Yuk, M. Yagi, ACS
Sustanible Chem. Eng. 7 (2019) 17896e17906.
[14] M.B. Tahir, G. Nabi, M. Rafique, N.R. Khalid, Int. J. Energy Res. 4 (2018)
4783e4789.
[15] M.B. Tahir, K.N. Riaz, A.M. Asiri, Int. J. Energy Res. 43 (2019) 5747e5758.
[16] L.J. Lauhon, M.S. Gudiksen, D. Wang, C.M. Lieber, Nature 420 (2002) 57e61.
[17] X.L. Yin, J. Liu, W.J. Jiang, X. Zhang, J.S. Hu, L.J. Wan, Chem. Commun. 51 (2015)
13842e13845.
[18] J. Wu, F. Duan, Y. Zheng, Y. Xie, J. Phys. Chem. C 111 (2007) 12866e12871.
[19] L.S. Zhang, W.Z. Wang, Z.G. Chen, L. Zhou, H.L. Xu, W.J. Zhu, Mater. Chem. 17
(2007) 2526e2532.
[20] J.F. Liu, Q.H. Li, T.H. Wang, D.P. Yu, Y.D. Li, Angew. Chem. Int. Ed. 43 (2004)
5048e5052.
[21] Y.P. Xie, G. Liu, L.C. Yin, H.M. Cheng, J. Mater. Chem. 22 (2012) 6746e6751.
[22] L. Yin, D. Chen, M. Hu, H. Shi, D. Yang, B. Fan, G. Shao, R. Zhang, G. Shao,
J. Mater. Chem. A 2 (2014) 18867e18874.
[23] H. He, S.P. Berglund, P. Xiao, W.D. Chemelewski, Y. Zhang, C.B. Mullins,
J. Mater. Chem. A 1 (2013) 12826e12834.
[24] R. Abe, H. Takami, N. Murakami, B. Ontani, J. Am. Chem. Soc. 130 (2008)
7780e7781.
[25] X. Tao, L. Shao, R. Wang, H. Xiang, B. Li, J. Colloid Interface Sci. 541 (2019)
300e311.
CRediT authorship contribution statement
Zhanglian Xu: Conceptualization, Visualization, Supervision.
Hisayoshi Kobayashi: Software, Investigation. Koji Nakane: Re-
sources, Methodology.
[26] B. Li, L. Shao, R. Wang, X. Dong, F. Zhao, P. Gao, Z. Li, J. Mater. Chem. A 6 (2018)
6344e6355.
[27] R. Wang, B. Li, Y. Xiao, X. Tao, X. Su, X. Dong, J. Catal. 364 (2018) 154e165.
[28] B. Li, B. Zhang, S. Nie, L. Shao, L. Hu, J. Catal. 348 (2017) 256e264.
[29] S.S. Ma, K. Maeda, R. Abe, K. Domen, Energy Environ. Sci. 5 (2012) 8390e8397.
[30] O. Tomita, T. Otsubo, M. Higashi, B. Ohtani, R. Abe, ACS Catal. 6 (2016)
1134e1144.
Declaration of competing interest
[31] D.C. Hurum, A.G. Agrios, K.A. Gray, T. Rajh, M.C. Thurnauer, J. Phys. Chem. B
107 (2003) 4545e4549.
[32] S. Bakardjieva, J. Subrt, V. Stengl, M.J. Dianez, M.J. Sayagues, Appl. Catal., B 58
(2005) 193e202.
We declare that we have no financial and personal relationships
with other people or organizations that can inappropriately influ-
ence our work, there is no professional or other personal interest of
any nature or kind in any product service and company that could
be construed as influencing the position presented in the manu-
script entitled “Novel Pd-loaded urchin-like (NH4)xWO3/WO3 as an
efficient visible-light-driven photocatalyst for partial conversion of
benzyl alcohol”.
ꢀ
ꢀ
[33] B. Sun, P.G. Smirniotis, Catal. Today 88 (2003) 49e59.
€
[34] T. van der Meulen, A. Mattson, L. Osterlund, J. Catal. 251 (2007) 131e144.
[35] S. Ardizzone, C.L. Bianchi, G. Cappelletti, S. Gialanella, C. Pirola, V. Ragaini,
J. Phys. Chem. C 111 (2007) 13222e13231.
[36] T. Ozawa, M. Iwasaki, H. Tada, T. Akita, K. Tanaka, S. Ito, J. Colloid Interface Sci.
281 (2005) 510e513.
[37] W. Li, C. Liu, Y. Zhou, Y. Bai, X. Feng, Z. Yang, L. Lu, X. Lu, K.Y. Chan, J. Phys.
Chem. C 112 (2008) 20539e20545.
[38] H. Xu, L. Zhang, J. Phys. Chem. C 113 (2009) 1785e1790.
[39] Z.L. Xu, I. Tabata, K. Hirogaki, K. Hisada, T. Wang, S. Wang, T. Hori, RSC Adv. 2
(2012) 103e106.
Acknowledgement
[40] S. Chen, X. Yu, H. Zhang, W. Liu, J. Hazard Mater. 180 (2010) 735e740.
[41] J. Cao, B. Luo, H. Lin, B. Xu, S. Chen, Appl. Catal., B 111 (2012) 288e296.
[42] H.G. Choi, Y.H. Jung, D.K. Kim, J. Am. Ceram. Soc. 88 (2005) 1684e1686.
[43] R. Custelcean, Chem. Commun. 3 (2008) 295e307.
[44] S. Chen, Y. Hu, S. Meng, X. Fu, Appl. Catal. B Environ. 150 (2014) 564e573.
[45] M. Xie, X. Dai, S. Meng, X. Fu, S. Chen, Chem. Eng. J. 245 (2014) 107e116.
[46] M.C. Payne, M.P. Teter, D.C. Allan, T.A. Arias, J.D. Johnnopoulos, Rev. Mod. Phys.
64 (1992) 1045e1097.
We are grateful to the financial support from Young Talent
Support Plan of Xi’a Jiaotong University (HN6J002)，the program of
Fundamental Research Funds for the Central Universities, and the
financial support by Innovation Team Funding: Design and Appli-
cation of Metal Functional Materials from Xi’an Shiyou University.
Scientific Research Program Funder by Shaanxi Provincial Educa-
tion Department (Program No. 19JK0654)..
[47] V. Milman, B. Winkler, J.A. White, C.J. Pickard, M.C. Payne, E.V. Akhmatskaya,
R.H. Nobes, Int. J. Quant. Chem. 77 (2002) 895e910.
[48] P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 77 (1996) 3865e3868.
[49] P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 78 (1997) 1396.
[50] D. Vanderbilt, Phys. Rev. B 43 (1991) 6796e6799.
Appendix A. Supplementary data
[51] E.R. McNellis, J. Meyer, K. Reuter, Phys. Rev. B 80 (2009) 205414.
[52] R.K. Gosavi, U. Agarwala, C.N.R. Rao, J. Am. Chem. Soc. 89 (1967) 235.
[53] L. Yang, B. Liu, T. Liu, X. Ma, H. Li, S. Yin, T. Sato, Y. Wang, Sci. Rep. 7 (2017)
5075e5080.
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.jallcom.2020.156225.
[54] O. Tomita, B. Ohtani, R. Abe, Catal. Sci. Technol. 4 (2014) 3850e3860.
[55] B. Pal, S. Ikeda, H. Kominami, B. Ohtani, J. Catal. 217 (2003) 152e159.
[56] V.R. Choudhary, S.D. Sansare, A.G. Gaikwad, Catal. Lett. 84 (2002) 81e87.
References
[1] C. Xu, F. Yang, B. Deng, Y. Zhuang, D. Li, B. Liu, W. Yang, Y. Li, J. Catal. 383
(2020) 1e12.