Facilitating Pt?WOx Species Interaction for Efficient Glycerol Hydrogenolysis to 1,3-Propanediol

Article abstract of DOI:10.1002/cctc.202100773

Designing efficient catalysts for glycerol hydrogenolysis to 1,3-propanediol (1,3-PDO), which involves the selective cleavage of the secondary C?O bond, is a challenging task. Current Pt?WO_x-based catalysts often provide low atom efficiency of W and Pt toward 1,3-PDO production due to undesired catalyst structures. Herein, we fabricate the highly-dispersed substantially uniform WO_x species on inert α-Al₂O₃ support by simple high-temperature heat-treatment, and the amount of Pt?WO_x interface active sites could be adjusted by Pt loading, showing an excellent catalytic performance in glycerol hydrogenolysis at high concentration of glycerol, especially the unprecedented W efficiency (76 g_1,3-PDOg_W^?1 h^?1) toward 1,3-PDO. The high catalytic efficiency is attributed to the strong interaction between the isolated WO₄ species and platinum, which could in-situ generate the Br?nsted acid sites during the reaction as evidenced by IR analysis with NH₃ adsorption.

Full text of DOI:10.1002/cctc.202100773

Full Papers
doi.org/10.1002/cctc.202100773
ChemCatChem
Facilitating PtÀ WO Species Interaction for Efficient
x
Glycerol Hydrogenolysis to 1,3-Propanediol
Binbin Zhao, Yu Liang, Lei Liu,* Qian He,* and Jin-Xiang Dong
[a]
[a]
[a]
[b]
[a, c]
Designing efficient catalysts for glycerol hydrogenolysis to 1,3-
propanediol (1,3-PDO), which involves the selective cleavage of
face active sites could be adjusted by Pt loading, showing an
excellent catalytic performance in glycerol hydrogenolysis at
high concentration of glycerol, especially the unprecedented W
the secondary CÀ O bond, is a challenging task. Current PtÀ WO -
x
À 1 À 1
based catalysts often provide low atom efficiency of W and Pt
toward 1,3-PDO production due to undesired catalyst struc-
tures. Herein, we fabricate the highly-dispersed substantially
uniform WO species on inert α-Al O support by simple high-
efficiency (76 g1,3-PDOgW h ) toward 1,3-PDO. The high catalytic
efficiency is attributed to the strong interaction between the
isolated WO4 species and platinum, which could in-situ
generate the Brønsted acid sites during the reaction as
x
2
3
temperature heat-treatment, and the amount of PtÀ WO inter-
evidenced by IR analysis with NH adsorption.
x
3
Introduction
an appropriate amount of SiO into Pt/WO /ZrO catalyst could
2
x
2
effectively restrain the formation of WO particles and transform
3
22]
[
Glycerol as a versatile platform molecule has been widely
studied for producing high value-added chemicals since it is the
by-product in the biodiesel production.
them into the active polytungstate. García-Fernández et al.
prepared the γ-Al O supported PtÀ WOx catalysts with low
2
3
[1,2]
The hydrogenolysis
polymeric tungsten oxide by decreasing the tungsten surface
density on support. However, this has also reduced the
of glycerol to 1,3-propanediol (1,3-PDO), an important polymer
[3]
precursor, is an attractive value-adding route for the glycerol.
However, achieving a high yield of 1,3-PDO, which requires a
selective cleavage of the secondary CÀ O bond in glycerol, is still
interaction between Pt and WO . The catalyst behaved like a Pt/
x
γ-Al O catalyst and produced 1,2-propanediol (1,2-PDO)
2
3
[23]
instead. Ma and co-workers found that the addition of Mn to
the Pt/WO /ZrO catalyst could result in a medium-degree
[4–12]
a challenging task.
x
2
The PtÀ WO -based materials are among the most promising
polymerization of WO that resulted in a higher yield of 1,3-
x
PDO. Recently, the same group prepared a series of xW4Pt/
x
[24]
glycerol hydrogenolysis catalysts due to their high thermal
[13–16]
stability and high selectivity towards 1,3-PDO.
Previous
SiO catalysts, in which Pt precursor was loaded before the WO
2
x
[21]
studies have shown that the nature of WO species and the
species. This has resulted in a 2W4Pt/SiO catalyst containing
x
2
proximity between WO and Pt are closely associated with
a large amount of low-degree of polymerization of WO which
x
x
[
14,17–20]
catalytic performance.
It is generally accepted that the
gave a high W specific activities toward 1,3-PDO production
À 1 À 1
crystalline WO is not active for the hydrogenolysis of glycerol
(5.786 g1,3-PDOgw h , W efficiency is defined as the amount of
3
to 1,3-PDO, and decreasing the polymerization degree of WOx
species can generally result in an improved yield of 1,3-
1,3-PDO produced per gram W per hour.). Theoretical inves-
tigation on HW O /Pt(111) indicated that the dimeric WO could
provide more Brønsted acid sites. Zong et al. prepared W-
x
y
x
[18,21,22]
[21]
PDO.
For instance, Zhu et al. reported that the addition of
doped SBA-15 supported Pt catalysts (Pt/W-SBA-15), where the
so far highest reported W specific activities (10.556 g1
PDOgw h ) toward 1,3-PDO were observed. However, the low
,3-
À 1 À 1
[
a] B. Zhao, Y. Liang, Prof. L. Liu, Prof. J.-X. Dong
College of Chemistry and Chemical Engineering
Taiyuan University of Technology
Yingze West Street 79
Taiyuan 030024, Shanxi (P. R. China)
E-mail: liulei@tyut.edu.cn
WO loading as well as the high BET surface area for SBA-15
x
result in the very low tungsten surface density, and in turn have
lowered the chance to form Pt/WO interfaces, leading to a
x
À 1 À 1 [18]
decrease in the Pt specific activities (7.339 g1,3-PDOgPt h ).
[
b] Prof. Q. He
Department of Materials Science and Engineering
National University of Singapore
Alumina as the support for Pt and W-species has also been
extensively studied in glycerol hydrogenolysis. Previous reports
show that satisfactory catalytic performance usually requires
9
Engineering Drive 1, Block EA #03-09
[13,14,22,25–29]
Singapore 117575 (Singapore)
E-mail: heqian@nus.edu.sg
c] Prof. J.-X. Dong
School of Chemical Engineering and Light Industry
Guangdong University of Technology
Guangdong University Town, Panyu District
No. 100 Waihuanxi Road
high loadings of Pt and W.
This, in turn, resulted in
coexistence of various (WOx)n species as well as the poor
dispersion of W-species on alumina surface, resulting in
complex catalyst structure and the relatively low specific
activities toward 1,3-PDO. Besides, the reported catalysts usually
used boehmite or γ-Al O , which has intrinsic acidity from
[
2
3
Guangzhou 510006 (P. R. China)
surface hydroxyl groups. This has resulted in additional
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/cctc.202100773
ChemCatChem 2021, 13, 1–12
1
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100773
ChemCatChem
challenges for an in-depth understanding of the role of
tungsten species during the reaction.
2 3
Table 1. Basic physicochemical properties of α-Al O -derived materials.
Sample
Pt loading
wt%]
W loading
[wt%]
S
BET
2
Inspired by these previous studies, we aim to design a
À 1
[
[m g
]
catalyst that can achieve both (i) high dispersion of WO species
with a low degree of polymerization and (ii) intimate interaction
x
α-Al
Pt/α-Al
0.5Pt/WO /α-Al O
Pt/WO
2
O
3
–
–
–
0.40
0.40
0.40
0.40
8.80
8.40
8.20
7.85
7.25
8.01
2
O
x
3
0.86
0.41
0.86
1.73
0.87
between Pt and WO species, the combination of which should
x
2
3
x
2 3
/α-Al O
lead to a highly-efficient catalyst for glycerol hydrogenolysis. To
achieve this, we learned from recent methoddevelopment in
the field of single-atom catalysts. It is well accepted that the
strong interaction between the metal and the support can
2Pt/WO
x
/α-Al
2
O
3
Pt/WO
x
/α-Al
2
O
3
-LT
[24,27,28]
improve the metal dispersion.
Similar to solid-state
reactions, a crucial parameter for preparing supported WO is
Structural properties of α-Al O -derived materials
2 3
x
the reaction temperature. The surface diffusion of atoms or ions
of WO began to occur when the temperature is raised to above
The effect of heat treatment on the dispersion of W on α-Al O3
x
2
the Tammann temperature of WO3 (i.e. 600°C), where the
is demonstrated using transmission electron microscopy (TEM)
bright-field (BF) imaging. As shown in Figure 1b, no cluster or
particles were formed in WO /α-Al O sample that has been
[30]
plastic deformation or creep can be expected. Therefore, it is
possible to fabricate the highly dispersed tungsten oxide
species on the support surface by controlling the calcination
temperature based on the Tammann temperatures of both the
x
2
3
heat-treated at 900°C, indicating that W was highly dispersed.
In contrast, small WO particles can be visualized in the WO /α-
x
x
WO and the support oxide.
Al O -LT sample that has been heat-treated at 500°C (Fig-
3
2
3
Using this strategy, we hereby present a Pt/WO /α-Al O
ure 1a). The Pt was then loaded onto the WO /α-Al O material,
x 2 3
x
2
3
catalyst that achieved high specific activities of glycerol hydro-
genolysis towards 1,3-PDO with the respect to both Pt (17 g
PDOgPt h ) and W (76 g1,3-PDOgW h ), which are higher than
and their representative aberration-corrected scanning trans-
mission electron microscopy (STEM) high angle annular dark-
field (HAADF) imagines are shown in Figure 2. While the lower
magnification image (Figure 2a) demonstrates the absence of
large WOx or Pt particles, the higher magnification images
(Figure 2b and 2c) clearly show clusters of about 1 nm or less. A
comparative Pt/α-Al O catalyst was also characterized (Fig-
1
,3-
À 1 À 1
À 1 À 1
reported data so far. α-Al O was chosen as the support due to
2
3
its moderate-high Tammann temperature (890°C), high thermal
stability, and chemical inertness. The calcination treatment was
applied with a temperature much higher than the Tammann
temperature of WO and slightly higher than that for Al O . In
2
3
ure S1) where small Pt nanoparticles of 2 nm or above can be
easily found. These results unambiguously demonstrate that Pt
and W species are both highly dispersed on the α-Al O surface,
3
2
3
such temperature, the atoms or iron of WO would become
x
more active and be beneficial to the diffusion and migration.
And the activated Al O surface would effectively enhance the
2
3
leading to more interfaces between Pt and WO species on the
2
3
x
interaction between the W-species and support. This allowed us
to achieve a high dispersion of WOx species, which also
alumina surface. These are also consistent with the X-ray
diffraction (XRD) measurements (Figure S2), the characteristic
promoted the formation of PtÀ WO interfaces. The absence of
diffraction peaks corresponding to crystalline WO (JCPDS card
20–1324) and platinum species could not be detected.
x
3
[
31]
acid sites on α-Al O also simplifies the analysis of the role of W-
2
3
species in glycerol hydrogenolysis. With various catalyst charac-
terization techniques, we confirmed that the enhanced catalytic
performance comes from the synergistic effect between Pt and
Diffuse reflectance UV-vis spectroscopy is an effective way
to investigate local molecular coordination and bonding
information for containing transition metal species catalyst. The
UV-vis spectra of Pt/WO /α-Al O catalysts are shown in Fig-
isolated WO species.
x
x
2
3
ure 3a. The band centered at 207 nm emerged for the WO /α-
x
Al O catalysts is attributed to the isolated tungsten species in a
2
3
[18,32]
Results and Discussion
tetrahedral coordination environment (WO4).
Meanwhile,
the shoulder peak band at 237 nm can be assigned to the
In order to achieve highly dispersed WO species onto α-Al O ,
slightly distorted WO species, which implies the presence of
x
2
3
4
with a low degree of polymerization, an impregnation and
centrifugal separation method was used, followed by calcina-
charge transfer between isolated tetrahedral WO species and
x
[33,34]
α-Al O .
Moreover, the tungsten species on the catalyst was
2
3
tion at 900°C in air. This material was designated as WO /α-
further characterized by Raman spectra (Figure 3b). The
x
À 1
Al O . A comparative sample (WO /α-Al O -LT) was prepared
using the same approach, except that a lower calcination
absence of characteristic Raman bands at 713 cm
and
2
3
x
2
3
À 1
807 cm for WO indicates that crystalline nanoparticles are
3
[35]
À 1
temperature (500°C) was used. Pt was then loaded to the
not present in the catalysts. Only a small band at 970 cm
material with another impregnation and calcination step to
associated with the symmetric stretching mode of the terminal
obtain the Pt/WO /α-Al O catalyst. Details for catalyst synthesis
W=O bond corresponding to highly dispersed mono-tungstate
x
2
3
[
14,23]
can be found in the Supplementary Materials. The W and Pt
loadings of different materials were determined by ICP-OES
method, and the basic physicochemical properties of α-Al O
species can be detected on the WO /α-Al O .
This can be
x
2
3
[
36]
assigned to isolated-tetrahedral WO4 species in this work.
Based on STEM, Raman, and UV-vis results, it can be concluded
2
3
before and after loading W and Pt were summarized in Table 1.
that the tungsten species are highly dispersed on the α-Al O3
2
ChemCatChem 2021, 13, 1–12
www.chemcatchem.org
2
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100773
ChemCatChem
x 2 3 x 2 3
Figure 1. Representative TEM bright field images and the corresponding schematic diagram of (a) the WO /α-Al O -LT and (b) the WO /α-Al O sample,
highlighting the difference in the dispersion of WO on α-Al surface.
x
2 3
O
x 2 3
Figure 2. Representative STEM-HAADF images of the Pt/WO /α-Al O catalyst. (a) A relatively lower magnification image, showing the absence of any larger Pt
nanoparticles; (b,c) higher magnification images showing clusters that are similar. The contrast difference shown in (b) and (c) are because some particles are
not in focus due to steep.
surface and are mainly exists in the form of isolated tetrahedral
WO species and WO clusters.
nanoparticles were re-dispersed into smaller WO species as
x
evidenced by TEM images (Figure 1). H -TPR test was also
4
x
2
performed to investigate the interaction between WOx and
support. It is recognized that α-Al O support cannot be
2
3
Strong interaction between PtÀ WO and support
reduced in the temperature range from 40 to 900°C. For the
x
WO /α-Al O sample, the absence of reduction peaks below
830°C suggested that the loaded WO species had a strong
x
2
3
X-ray photoelectron spectroscopy (XPS) measurements are
performed to investigate the chemical state of WO species on
the catalyst. As shown in Figure 4a, the tungsten species in
x
[23,38]
interaction with alumina support (Figure 5a).
The WOx
x
species were highly dispersed on the α-Al O surface in the
2 3
5
+
[18,39,40]
WO /α-Al O show two sets of XPS peaks corresponding to W
form of WÀ OÀ Al linkages,
which could capture hydrogen
x
2
3
6
+
and W species, and the concentration ratio of surface species
species to generate Brønsted acid sites (WÀ OHÀ Al) on the
5
+
6+
5+
[26]
W
and W is 0.81. A large amount of W species revealed
catalyst.
that it existed the charge transfer between the tungsten species
and support, and the W-species were strongly anchored onto
the support as the result of the heat treatment. Especially, as
the calcination temperature used (900°C) is much higher than
For the supported Pt-W catalysts, the interaction between
Pt and W-species is very crucial in the glycerol hydrogenolysis
to 1,3-PDO. H -TPR tests could provide information about the
2
dispersion of Pt species, and Pt and support interactions. For
comparative analysis, various catalysts were investigated by H2-
TPR. As shown in Figure 5a, the pristine α-Al O exhibited two
[
30]
the Tammann temperature of WO (i.e. 600°C). Such treat-
3
ment triggered the tungsten species began to occur plastic
2
3
deformation or creep on the α-Al O surface and then interact
reduction peaks at 82°C and 133°C upon platinum loading,
2
3
with support, which can be referred to as the interfacial
which was ascribed to the reduction of platinum oxides. After
[37]
electronic effect. As a result, the preformed tungsten oxide
predeposition of WO on support, the reduction peak intensity
x
ChemCatChem 2021, 13, 1–12
www.chemcatchem.org
3
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100773
ChemCatChem
FTIR spectra of WO /α-Al O , Pt/α-Al O , and Pt/WO /α-Al O
3
x
2
3
2
3
x
2
samples. The signal of CO adsorption was not observed on
WO /α-Al O , suggesting that the appeared peaks should be
x
2
3
originated from CO adsorbed on two different Pt species. For
À 1
Pt/α-Al O , the dominant band is located at 2082 cm ,
2
3
corresponding to CO linearly absorbed on well-coordinated Pt
[
39,42,43]
nanoparticles.
1
While the weak bands at 2131 and
À 1
835 cm are ascribed to CO linearly absorbed on positively
charged Pt species and the bridge-bonded CO on larger
[42,44,45]
platinum nanoparticles, respectively.
Upon doping of W
(
0.4 wt%) into the alumina, Pt/WO /α-Al O exhibited a strong
x
2
3
À 1
band at 2138 cm , assigning to CO adsorbed on positively
charged Pt species. The increased band intensity illustrated that
the amount of positively charged Pt species are much higher
than that in Pt/α-Al O catalyst. This result was resulted from
2
3
the generation of much more PtÀ OÀ W/Al interface sites,
indicating that the introduction of WO greatly promoted the
x
dispersion of Pt on the α-Al O surface due to the strong
2
3
[27]
interaction between Pt and WOx.
Pt 4d XPS spectra (Fig-
ure 4b) showed that the binding energy of Pt 4d in Pt/WO /α-
x
Al O was negatively shifted by contrast with Pt/α-Al O , mean-
2
3
2
3
while, the binding energy of W 4f (Figure 4c) was positively
shifted (Figure 4a). XPS result revealed that the electronic
structure of Pt was obviously changed owing to the charge
transfer from WOx to Pt after loading WOx species, further
evidencing that the strong electronic interaction between Pt
[31,46,47]
and WO on the surface of α-Al O .
x
2
3
Additionally, the extended X-ray absorption fine structure
EXAFS) and X-ray absorption near edge structure (XANES)
(
spectra (Figure 5, Figure S3 and 4) were collected to further
study the coordination environment and chemical state of
platinum species on the Pt/WO /α-Al O , and the EXAFS fitting
x
2
3
x 2 3
Figure 3. Identification of highly dispersed WO species on α-Al O . (a) UV-vis
spectra and (b) Raman spectra for the selected catalysts and the support.
data were summarized in Table S1. The XANES spectra (Fig-
ure 5d) showed that the white-line intensities for Pt/WO /α-
x
Al O and Pt/α-Al O were both higher than that of Pt foils but
2
3
2
3
lower than that of PtO . The edge energy is 13270.9 eV for Pt
2
[
14]
of platinum oxides before 200°C is weakened, meanwhile, an
additional reduction peak at 540°C is observed. By comparing
H -TPR profiles between Pt/WO /α-Al O and WO /α-Al O , the
foil, 13271.3 eV for Pt/WO /α-Al O , and 13272.6 eV for Pt/α-
x
2
3
Al O , both are greater than that characterizing bulk platinum
2
3
(13270.9 eV), indicating that the Pt species in two catalysts are
2
x
2
3
x
2
3
[48]
reduction peak around 540°C could be ascribed to the highly
both positively charged. However, Pt white-line intensity for
δ+
dispersed Pt species strongly interacting with WO , this case
Pt/WO /α-Al O was lower than that for Pt/α-Al O , suggesting
x
x
2
3
2
3
[23,41]
has been reported in previous studies.
The consumed H2
that the Pt species in Pt/WO /α-Al O hold a higher electron
x 2 3
amount (Table S2) calculated from H -TPR data showed that the
density than that in Pt/α-Al O catalyst. The introduction of W
2 3
2
total consumed H2 amount of Pt/WO /α-Al O is obviously
species induced charge transfer from WO to Pt and form an
x
2
3
x
higher than that of Pt/α-Al O . Owing to the identical Pt loading
electron-enriched surface of Pt clusters in Pt/WO /α-Al O
3
2
3
x
2
[49]
in the two catalysts, the visible higher H consumption of Pt/
catalyst.
XPS, H -TPR, and CO-FTIR characterizations also
2
2
WO /α-Al O over Pt/α-Al O could be well explained by its
illustrated that the introduction of WO enhanced the electronic
x
2
3
2
3
x
notably higher Pt dispersion as evidenced by TEM character-
ization (Figure 2, Figure S1). Besides, the reduction peak for WOx
species moved to 840°C from 870°C after loading Pt owing to
interaction between Pt and WO , herein, we ascribed the Pt-O
x
bond in Pt/WO /α-Al O to the formation of PtÀ OÀ WO inter-
x
2
3
x
face. The possible Pt-Pt interaction according to EXAFS fitting
data is probably due to the formation of ultrasmall sub-
the role of hydrogen spillover on PtÀ WO interface, H -TPR date
x
2
[50]
illustrated that the pre-deposited WO species on α-Al O highly
nanometer sizes of Pt clusters. in good agreement with the
well-uniformly dispersed Pt clusters observed in STEM images.
Arribas et al. reported that the existence of the positively
charged Pt species is closely associated with the interaction
between the highly dispersed Pt nanoparticles and tungsten
x
2
3
interacted with Pt species, and greatly promoted the high
dispersion of Pt.
Furthermore, in-situ diffuse infrared Fourier transform spec-
troscopy (DRIFTS) with CO was also used to understand the
[51]
interaction between Pt and WO . Figure 5b presents the CO-
species on the catalyst. Given that the characteristic results
x
ChemCatChem 2021, 13, 1–12
www.chemcatchem.org
4
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100773
ChemCatChem
Figure 4. Evaluation of chemical state of W and Pt species on α-Al
2 3 x 2 3 2 3
O support. (a) W 4f XPS spectra of WO /α-Al O . (b) Pt 4d XPS spectra of Pt/α-Al O and Pt/
WO /α-Al . W 4f XPS spectra for the fresh-Pt/WO /α-Al (c) and Pt/WO
x
2
O
3
x
2
O
3
x
/α-Al
2
O
3
-spent (d).
above, it could conclude that the positively charged Pt species
were originated from the periphery interface of Pt cluster with
It is generally accepted that Brønsted acid sites are
propitious to the formation of 1,3-PDO, while the Lewis acid
sites will be favorable to the production of 1,2-propanediol (1,2-
PDO) during the glycerol hydrogenolysis. With respect to the
intrinsic acid nature in Pt/WO /α-Al O catalysts, it could not
[18]
WO or carrier by forming the PtÀ OÀ W/Al linkages. The high
x
dispersion of WO and Pt species would facilitate the intimate
x
interaction between them and generated much more PtÀ OÀ W
x
2
3
interfaces working as the active site for glycerol hydrogenolysis
explain why the current catalyst has an excellent performance
in the hydrogenolysis of glycerol to 1,3-PDO. Recently, some
researchers suggested that Brønsted acid sites could be in-situ
[24]
to 1,3-PDO.
generated on the PtÀ WO -based catalysts, which played a vital
x
[18,24,27]
Acid properties
role in hydrogenolysis of glycerol.
Unfortunately, owing to
the acidic nature of the support itself and the acid sites derived
The nature of acid sites in catalysts is also crucial to determine
the catalytic results in glycerol hydrogenolysis. It is well known
that the pristine α-Al O has no Brønsted and Lewis acid sites.
from the high WO loading, it is very difficult to identify which
kind of acid sites (inherent acid sites and in-situ generated acid
sites) on the catalyst is crucial to the reaction, and the synergy
x
2
3
The NH -TPD profiles Figure S5 showed that there is still no
mechanism between Pt and WO for in-situ acid formation is
3
x
obvious desorption peak of NH both in WO /α-Al O and Pt/
still unclear. The in-situ diffuse reflection infrared Fourier
3
x
2
3
WO /α-Al O materials, indicating that the two materials have
transform spectroscopy (DRIFTS) with adsorbed NH has been
x
2
3
3
very low acid amount after loading WO and Pt species. The
proved to be an effective method in probing the transformation
x
pyridine adsorption spectra (PY-IR) also showed that the
Brønsted acid sites corresponding to the signal at 1540 cm
of acid sites on the surface of catalyst upon hydrogen
À 1
[53–55]
exposure.
To understand the acidic nature in Pt/WO /α-
x
are absence in Pt/WO /α-Al O (Figure S6), and only the weak
peaks at ca. 1450 and 1620 cm related to the Lewis acid sites
Al O during the reaction, herein, the relationship between
2 3
x
2
3
À 1
Brønsted and Lewis acid sites was studied by recording the in-
situ DRIFTS with adsorbed NH at 100°C under 10% H /Ar gas
flow. FTIR of NH adsorption on WO /α-Al O (Figure S7) only
3 x 2 3
were detected, which can be assigned to the pyridine adsorbed
3
2
[52]
onto the terminal W=O groups.
À 1
showed a very weak band at 1280–1350 cm corresponding to
ChemCatChem 2021, 13, 1–12
www.chemcatchem.org
5
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100773
ChemCatChem
Figure 5. (a) H
2 2 3
-TPR spectra of different catalysts. (b) In situ FTIR spectra of CO adsorption for different catalyst. (c) Fourier transformation of the Pt L -edge k -
weighted χ(k) data for different catalysts. (d) Normalized XANES spectra at Pt L
2
-edge of different samples.
[55]
À 1
+
[56]
adsorbed NH on Lewis acid sites. In contrast, the new signals
1670 cm were referred to NH4 on Brønsted acid sites. It
3
of NH adsorption were observed on WO /α-Al O containing Pt
can be clearly seen that FTIR spectra over Pt/WO /α-Al O have
3
x
2
3
x
2
3
catalyst, indicating that some acid sites were formed after
platinum loading. As shown in Figure 6, the bands at 1280–
changed dramatically with time on stream after the introduc-
+
tion of H . The characteristic band of NH on Brønsted acid
2
4
À 1
À 1
1
350 cm and 1600 cm corresponded to the adsorbed NH
sites gradually increased, while the bands associated with Lewis
acid sites were decreased under the process, and the finally
generated Brønsted acid sites are dominant over all acid sites.
Additionally, a series of FTIR spectra of Pt/WO /α-Al O was
3
À 1
on Lewis acid sites, while the weak bands at 1430 cm and
x
2
3
recorded under various pretreatment and gas atmosphere to
understand the acid sites transformation, as shown in Figure S8.
These results showed that Pt/WO /α-Al O after H pretreatment
x
2
3
2
did not exhibited the Brønsted acid property under inert gas
atmosphere, and the original Lewis acid sites still existed on the
catalyst surface. However, Brønsted acid sites could be in-situ
formed again upon introducing H into the system. Herein, the
2
in-situ generated Brønsted acid sites may associate with a large
amount of W=O and WÀ OÀ Al bonds in WO /α-Al O , which was
x
2
3
converted to WÀ OH and WÀ (OH)À Al through hydrogen spill-
over owing to the strong interaction between Pt and WO
x
[43,57,58]
species.
We now provide direct evidence to reveal the in-
situ generation of Brønsted acid sites for the supported PtÀ WO
x
catalysts in the presence of H , which might throw light on
2
further understanding the glycerol hydrogenolysis mechanism
over Pt and W-based catalysts.
3 2
Figure 6. FTIR spectra of NH adsorbed during the introduction of H at
1
00°C for 90 min over Pt/WO
x
/α-Al
2
O
3
catalyst.
ChemCatChem 2021, 13, 1–12
www.chemcatchem.org
6
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100773
ChemCatChem
[23]
[13]
Glycerol hydrogenolysis
10WO /Al O
and 0.042 for 1.8Pt/10WO /AlOOH. The low
x
2
3
x
catalytic efficiency has resulted from the failure in fabricating
The catalytic performance of α-Al O -derived catalysts was
more active sites of PtÀ OÀ W interface, which heavily depends
2
3
evaluated in the glycerol hydrogenolysis, the catalytic results
were summarized in Table 2. It has been generally accepted
that glycerol hydrogenolysis to 1,3-PDO proceeded via a
dehydration-hydrogenation mechanism over the bifunctional
on the high dispersion of platinum and active WO species on
x
the support surface. In our study, the prepared PtÀ WO catalysts
x
exhibit an excellent catalytic efficiency with STY as high as
À 1 À 1
0.3 g1,3-PDOgcat
h
for 2Pt/WO /α-Al O , the STY is positively
x 2 3
[6,11,18,19,59]
catalysts.
As expected, the prepared Pt/α-Al O and Pt-
related to the Pt loading in our studied range. Meanwhile, the
production rate of 1,3-PDO based on surface exposed Pt mass
(STY ) for PtÀ WO catalysts (Table 1, entry 3–6) have changed
2
3
free WO /α-Al O catalyst could not catalyze the conversion of
x
2
3
glycerol owing to the absence of the acid sites and Pt species
for them, respectively. After pre-deposition of a small quantity
of W (0.4 wt%) on α-Al O , Pt/WO /α-Al O catalyst exhibits an
Pt
x
very little and maintained in a high level with the increase of Pt
loading. Considering that the good linear relation between STY
and Pt loading as well as the almost unchanged STYPt with
various Pt loading, Pt/α-Al O was confirmed to be inert for
2
3
x
2
3
excellent catalytic performance with high selectivity toward 1,3-
PDO at a high glycerol concentration of 30% (Table 2, entry 4).
For comparison, a control experiment was conducted over a
physical mixture of Pt/α-Al O and WO /α-Al O . The catalytic
2
3
glycerol hydrogenolysis (Table 1, entry 1), therefore, the high
STY value for our prepared PtÀ WO catalysts was contributed to
2
3
x
2
3
x
result (Table 2, entry 2) showed that only the trace of glycerol
was converted, meanwhile, the selectivity of 1,3-PDO is very
low (27.39%) compared to Pt/WO /α-Al O catalyst (50.36%).
the effective connection of platinum and WO species owing to
x
the highly dispersed platinum (Figure 2, Figure S10) and WOx
species (Figure 1) on the support surface.
x
2
3
These results demonstrated the synergy effect between the
The polymerization degree of the pre-deposited WOx
species on support is also closely associated with the catalytic
platinum and WO species in Pt/WO /α-Al O , wherein the close
x
x
2
3
[24,25]
proximities of Pt and W-species contribute to their strong
interaction for generating more PtÀ OÀ W interfaces, which
works as an active center for glycerol hydrogenolysis to 1,3-
efficiency of PtÀ WOx catalysts.
In previous studies, the
polymerization degree of WO species was mainly controlled by
x
changing WO loading, low WO loading is inclined to form low
x
x
[17,18,24,60]
PDO.
Herein, only a small quantity of Lewis acid sites
polymeric WO species on the support surface. Some studies
x
behaved on Pt/WO /α-Al O catalyst (Figure S5), which generally
favor the production toward 1,2-PDO. However, the Pt/WOx/
have shown that decreasing WO polymerization degree would
x
2
3
x
[11]
[22,24,27]
benefit the catalytic performance.
Recently, Ma and co-
α-Al O would in-situ generate Brønsted acid sites companied
workers prepared a series of xW4Pt/SiO catalysts with various
2
3
2
with the disappear of Lewis acid sites in the presence of H2
W/Pt mole ratio, 2W4Pt/SiO catalyst containing a large amount
2
owing to the strong interaction between Pt and WO species, as
of low-degree of polymerization of WO gave the highest W
x
x
evidenced by the in-situ DRIFTS with adsorbed NH (Figure 6).
specific activities towards 1,3-PDO production (5.786 g1,3-
3
À 1 À 1
Therefore, the in-situ generated Brønsted acid resulting from
the PtÀ OÀ W interfaces could dominate the first-step dehydra-
tion for glycerol, then the hydrogenation of intermediate
PDOgW h ), suggesting that the dimeric WO was better than
x
[21]
the isolated WOx in yielding 1,3-PDO. However, Zong et al.
reported W-doped SBA-15 supported Pt catalysts in which the
[19,25]
product proceeds on the highly dispersed Pt species.
W-species existed in the form of the isolated WO , and
4
À 1 À 1
The overall catalytic efficiency, the production rate of 1,3-
PDO based on catalyst mass (STY), can be used to evaluate the
exhibited very high W specific activities (10.56 g1,3-PDOgW h )
towards 1,3-PDO, ascribing to the isolated WO and the small Pt
4
[18]
synergy effect between Pt and WO species for the supported
nanoparticles. In our study, the W loading was fixed, and the
x
PtÀ WOx bi-functional catalysts. In the case of the previous
WO domain size could be controlled by different treatment
x
reported PtÀ WO catalysts, although high loadings of Pt and W
conditions, which would benefit a comparative analysis. For
comparison, glycerol hydrogenolysis was performed over Pt/
WO /α-Al O -LT catalyst containing high polymeric WO clusters
x
were applied for glycerol hydrogenolysis, the values of STY for
these catalysts were generally kept at a very low level with less
x
2
3
x
À 1 À 1 [17,18,24,25,29,61]
than 0.1 g1,3-PDOgcat h ,
for example, 0.039 for 9Pt/
less than 1 nm (Figure 1a, Figure S9). The catalytic efficiency
[
a]
Table 2. The catalytic results of glycerol hydrogenolysis over different catalysts.
[
b]
[c]
Entry
Catalyst
Time
h]
Conversion
[%]
Selectivity [%]
STY
STYPt
À 1 À 1
À 1 À 1
[
1,3-PDO
1,2-PDO
1-Propanol
2-Propanol
[g1,3-PDO
g
cat
h
]
[g1,3-PDO
g
Pt
h
]
1
2
3
4
Pt/α-Al
Pt/α-Al
0.5Pt/WO
2
O
O
3
12
12
12
12
–
2.15
19.05
36.55
59.54
51.66
24.49
–
–
–
–
–
–
0.07
0.15
–
–
16
17
2
3
+WO
x
/α-Al
3
2
O
3
27.39
51.83
50.36
43.90
45.65
47.58
36.14
5.48
4.21
2.77
2.83
5.63
23.76
26.01
27.73
27.54
28.91
22.96
3.99
7.05
4.75
3.39
3.55
4.06
x
/α-Al
2
O
Pt/WO /α-Al O
x
2 3
24
5
6
2Pt/WO
Pt/WO /α-Al
x
/α-Al
2
O
3
12
12
0.30
0.08
17
9
x
2
O
3
-LT
[
a] Reaction conditions: 6 mL 30 wt% glycerol aqueous solution was used as the substrate, 0.2 g of catalyst, temperature 180°C, H
2
pressure of 5.0 MPa,
stirring rate of 600 rpm. The STY and STYPt were acquired at glycerol conversion within 10%; [b] the amount of 1,3-PDO produced per gram catalyst per
hour; [c] the amount of 1,3-PDO produced per gram Pt per hour.
ChemCatChem 2021, 13, 1–12
www.chemcatchem.org
7
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100773
ChemCatChem
both in STY and STYPt for Pt/WO /α-Al O -LT catalyst (Table 2,
entry 6) dropped obviously, the W efficiency toward 1,3-PDO
exhibited a significant high W efficiency with STY value of
W
x
2
3
À 1 À 1
[21]
76 g1,3-PDOgW
h
much higher than the reported data so far,
production (STY ) was also much lower than that of Pt/WO /α-
Al O3 (20 vs 37), confirming that the monodispersed WOx
2
demonstrating the advantage of the high-temperature heat-
treatment method in fabricating the highly dispersed active
W
x
combining Pt species could significantly enhance the catalytic
efficiency towards the formation of 1,3-PDO, which is in good
agreement with that observed by Zong’s group. In previous
WO species on support.
x
In terms of catalyst recycling (Figure S11), the glycerol
conversion had a slight decrease (about 5%) after three
successive runs, while the selectivity of 1,3-PDO remains
unchanged. The leaching of Pt was not detected by ICP analysis
in the reaction liquid. We also detect the W content of the
recycled Pt/WO /α-Al O and Pt/WO /α-Al O -LT to confirm the
[18]
studies, the supported PtÀ WO catalysts containing some
x
mono-dispersed WO species were synthesized by decreasing
x
the W loading to a low level. The W density on the surface is
very low due to the high surface area of the support, leading to
the difficulty in forming PtÀ OÀ W interface which has been
x
2
3
x
2
3
leaching of WO during the reaction. The W content of the
x
[21,62]
proved to be the active sites.
Consequently, the conclusions
recycled Pt/WO /α-Al O after three successive runs decreased
x
2
3
neglect the effect of the low density of active WO on the
to 0.39 wt% from 0.40 wt%, however, The W content of the
x
st
fabrication of PtÀ WO interface, and do not reflect the nature of
recycled Pt/WO /α-Al O -LT only after 1 run had an obvious
x
x
2
3
WO species in the previous synthetic method generally existed
drop from 0.4 wt% to 0.35 wt%. These results further demon-
strated that high-temperature calcination promoted the inter-
action between WO and α-Al O , and consequently inhibited
x
with various domain sizes, which is also not beneficial to truly
study the nature of WO species.
x
x
2
3
For the prepared WO /α-Al O material by high-temperature
the leaching to W-species. XANES (Figure S12) and XPS (Fig-
x
2
3
heat-treatment, the low polymeric/isolated WO species were
ure 4d) results for the spent Pt/WO /α-Al O showed that the
x
x
2
3
highly dispersed on the surface of α-Al O as evidenced by TEM
state of Pt was almost consistent with the fresh one, and a small
2
3
4
+
image (Figure 1), which allows us to exactly calculate the actual
portion of high-valence W-species was reduced to W . The
slight drop in catalytic activity may have resulted from the
2
W surface density (W atom/nm ) on support based on the W
4
+
loading and the support surface area. Although the W loading
is only 0.4wt%, the W surface density can reach a high level of
formation of W -species, because the over-reduction of W-
+
6
5+
species to would hinder the catalytic redox cycle of W ⇋W
[
19]
1
.5 owing to the very low specific surface area for α-Al O .
and, in turn, cause a decrease in the catalytic activity. The
catalyst recycling test also indicated that the fabrication of the
highly-dispersed stable WOx species on support is very
important for enhancing the catalytic efficiency and reusability.
2
3
Consequently, the highly dispersed active WO species with
x
high density on α-Al O enormously facilitated to the con-
2
3
struction of PtÀ OÀ W active site for the supported PtÀ WOx
catalysts. Therefore, it could be expected that the catalytic
efficiency in STY and STYW would linearly increase with the
enhancement of Pt loading. To verify this, STY and STYW values Conclusion
were calculated based on the catalytic results over Pt/WO /α-
x
Al O catalysts with various Pt loading. As expected, the STY
In summary, we have developed a facile strategy to prepare
highly dispersed WO species on inert support of α-Al O with a
very low surface by simple high-temperature treatment, which
facilitate the dispersion of platinum at atom clusters or single-
2
3
value is positively correlated with the Pt loading as discussed
above, the positive correlation is also very similar for STYW and
Pt loading, as shown in Scheme 1. 2Pt/WO /α-Al O catalyst
x
2
3
x
2
3
À 1 À 1
Scheme 1. Models of Pt-W interface correlation with STY
w
value of 1,3-PDO. STY
W
is defined as the amount of 1,3-PDO produced per gram W per hour, g1,3-PDO
g
W
h .
ChemCatChem 2021, 13, 1–12
www.chemcatchem.org
8
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100773
ChemCatChem
atom level, leading to more interfaces between Pt and WOx
species on the alumina surface. The isolated tetrahedral WO4
species combining Pt work as the most effective active sites in
the glycerol hydrogenolysis to 1,3-PDO. The highly dispersed
voltage of 200 kV (resolution of 0.24 nm). All samples were
mounted on 3 mm holey carbon copper grids. High angle annular
dark field scanning transmission electron microscopy (HAADF-
STEM) were carried out using an aberration-corrected JEOL
ARM200CF microscope operated at 200 kV.
W-species facilitate PtÀ WO species interaction, exhibiting a
x
À 1 À 1
XPS spectra of the catalysts were acquired by using an X-ray
photoelectron Spectrometer (AXIS ULTRA DLD, Kratos, UK)
equipped with AlKα resource (hν=1486.6 eV). The contaminated
carbon C 1s signal at 284.8 eV was used to calibrate binding
energies. The spectra were deconvolved with XPS Peak4.1 software
by subtracting the Shirley background and applying the Lorentzian-
Gaussian function.
significantly high W efficiency as high as 76 g1,3-PDOgW h . The
strong interaction between WOx and Pt contributes to the
generation of Brønsted acid sites in-situ on the catalyst during
the reaction in the presence of H . Also, this study provides a
2
new idea to design and synthesize the supported metal oxide
catalysts applied for various reactions.
Temperature-programmed reduction of hydrogen (H -TPR) and
2
temperature-programmed desorption of NH3 (NH -TPD) was con-
3
ducted on a Micromeritics AutoChem II 2920. For H -TPR, typically,
the sample (50 mg) was pre-treated in pure Ar flow (30 mL/min) at
Experimental Section
2
3
00°C for 1 h to remove the moisture and other adsorbed gases on
Catalyst preparation
the catalyst surface. And subsequently cooled to 40°C, the sample
Tungsten supported on α-Al O (Aladdin, 99.99% metals basis) was
prepared by impregnation-centrifugation method. 0.3371 g
was then reduced in a flowing gas of mixed 5% H /Ar (30 mL/min)
2
2
3
increasing the temperature from 40 to 900
°
C, and recorded by a
ammonium metatungstate ((NH ) (H W O )·nH O, Aladdin,�
thermal conductivity detector. For NH
3
-TPD, about 50 mg of the
4
6
2
12 40
2
°
C for
9
9.99%) was dissolved in 50 mL distilled water under stirring at
catalyst was pre-treated in pure He flows (30 mL/min) at 300
°
C and exposed to mixed 5%
/He gas (40 mL/min) for 30 min, followed purged with pure He
(30 mL/min) for 40 min to remove physisorbed NH
600 rpm for 1 h. Then, 5 g of α-Al O was added to the solution
1 h, and subsequently cooled to 100
NH
2
3
under stirring at 600 rpm for 24 h at room temperature, after
centrifuge the solid and dried at 283 K in an oven for 12 hours, the
powder followed by calcining at 773 and 1173 K for 3 h in static air,
marked as WO /α-Al O -LT and WO /α-Al O , respectively. Subse-
3
. Finally, the
3
sample was heated from 100 to 700
and recorded by a thermal conductivity detector.
°
C at a ramp rate of 10°C/min
x
2
3
x
2
3
quently, Pt supported on α-Al O , WO /α-Al O , and WO /α-Al O -LT
2
3
x
2
3
x
2
3
FTIR spectra after pyridine adsorption were recorded on a Bruck
were prepared by impregnation method. Chloroplatinic acid
H2PtCl6
À 1
INVENIO R spectrometer at a resolution of 4 cm . Each spectrum
(
*
6H2O, Aladdin, AR, Pt�37.5%) was added to 40 mL of
was averaged over 16 scans. Before each pyridine adsorption
experiment, the self-supporting wafers of samples were placed in a
vacuum cell with greaseless stopcocks and CaF2 windows. The
distilled water and stirred at room temperature, 2 g of support (α-
Al O , WO /α-Al O , or WO /α-Al O -LT) was added and the solution
2
3
x
2
3
x
2
3
was stirred for 24 h and drying at 383 K in an oven for 12 h. And
then the powder obtained was calcined at 573 K under a static air
for 3 h to obtain Pt/α-Al O , WO /α-Al O -LT, and Pt/ WO /α-Al O ,
À 7
samples were pre-treated under vacuum at 573 K and 10 Torr for
1
h, and then cooling to 40°C. After recording the background,
2
3
x
2
3
x
2
3
followed exposed to pyridine vapors for 3 min and maintained 1 h
at 40°C. Subsequently, the Py-IR spectra were recorded at 573 K
respectively.
after applying a vacuum for 30 min. All spectra were obtained by
subtracting the spectrum of the background.
Catalyst characterization
FTIR spectra of CO adsorption were collected at 40°C with a Bruker
The powder X-ray diffraction (XRD) of all catalysts were performed
on a D8 Bruker employing Cu-Kαradiation monochromatized
radiation (λ=0.1541 nm). The voltage and current intensities were
INVENIO R spectrometer equipped with an LN-MCT Mid detector at
the spectral resolution of 4 cm . Firstly, the samples were pre-
À 1
treated in pure Ar (30 mL/min) flow at 300°C for 1 h. Subsequently,
4
0 kV and 40 mA, respectively. The scanning angle (2θ) ranged
°
C and under pure Ar flow for
À 1
the sample was cooled down to 40
from 20 to 80° at 4° min with a step width of 0.02°.
3
0 min. After recording the background, the gas was changed to
The specific surface areas of the sample were evaluated from N
10% CO/He (30 mL/min) gas and kept for 30 min, then the IR
spectra were recorded till no visible change in the absorption band
intensities under He purging. All spectra were obtained by
subtracting the spectrum of the background.
2
adsorption-desorption isotherms at À 196°C, determined by using
an automatic ASAP 2020 system from Micromeritics. Before
À 4
adsorption, samples were outgassed at 200°C and 10 mbar,
overnight. Brunauer-Emmett-Teller (BET) method was used to
calculate the specific surface areas (SBET).
FTIR spectra of NH3 adsorption were obtained on the same
apparatus above. Firstly, the samples were pre-treated under pure
Ar (30 mL/min) at 300°C for 1 h to clean the surface from moisture,
and then cooling to 100°C under a pure Ar gas flow for 30 min.
The platinum and tungsten content in the catalyst was determined
by inductively coupled plasma-atomic emission spectroscopy (ICP-
AES, Thermo iCAP6300).
After recording the background, the sample was adsorbed with
mixed 5% NH /He gas flow (30 mL/min) for 30 min at 100°C and
3
Raman spectroscopy was performed using a Renishaw inVia Reflex
micro-Raman spectrometer in the back-scattering geometry. The
wavenumber of the laser source is 532 nm. The spectral resolution
followed purged with pure Ar gas flow for 40 min to remove
physisorbed ammonia. Finally, the sample was introduced with
mixed 10% H /Ar gas flow (30 mL/min) at 100°C for 90 min and
2
À 1
was approximately 1 cm . The spectra were recorded at room
recorded the IR spectra of the adsorbed species. All spectra were
obtained by subtracting the spectrum of the background.
À 1
temperature within the 200–1200 cm .
UV-visible diffuse reflectance spectra (DRS) were collected on
XAS measurements for the Pt L2-edge were performed in
fluorescence mode on beamline 20-BM-B with an electron energy
of 7 GeV and an average current of 100 mA which is located in the
Advanced Photon Source at Argonne National Laboratory. The
Perkin-Elmer Lambda 650 spectrometer with BaSO as a reference.
4
Transmission electron microscopy (TEM) images were obtained
using an FEI Tecnai G2 F20 S-Twin microscope at an accelerating
ChemCatChem 2021, 13, 1–12
www.chemcatchem.org
9
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100773
ChemCatChem
radiation was monochromatized by a Si (111) double-crystal
monochromator. XANES and EXAFS data reduction and analysis
were processed by Athena software.
[1] M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi, C. Della Pina, Angew.
Chem. Int. Ed. 2007, 46, 4434–4440; Angew. Chem. 2007, 119, 4516–
4522.
[
[
2] M. Anitha, S. K. Kamarudin, N. T. Kofli, Chem. Eng. J. 2016, 295, 119–130.
3] C. H. Zhou, J. N. Beltramini, Y. X. Fan, G. Q. Lu, Chem. Soc. Rev. 2008, 37,
5
27–549.
Catalytic reaction
[
4] A. D. da Silva Ruy, R. M. de Brito Alves, T. L. Reis Hewer, D. de A-
guiar Pontes, L. S. Gomes Teixeira, L. A. Magalhães Pontes, Catal. Today
The hydrogenation reaction was carried out in a 25 mL-capacity
2
020, DOI 10.1016/j.cattod.2020.06.035.
Teflon-lined stainless-steel autoclave. Reaction conditions: 180°C,
[
[
5] M. R. Nimlos, S. J. Blanksby, X. Qian, M. E. Himmel, D. K. Johnson, J. Phys.
Chem. A 2006, 110, 6145–6156.
6] T. Miyazawa, Y. Kusunoki, K. Kunimori, K. Tomishige, J. Catal. 2006, 240,
213–221.
5
MPa H , 200 mg of catalyst, 6 mL 30 wt% glycerol water solution
2
was added to the Teflon-lined. The autoclave was sealed and
purged with H2 six times to expel air. During the reaction, the
stirring rate was 600 rpm. After the reactions, the liquid samples
were analyzed by an HPLC (Shimadzu, LC-20AD) equipped with a
RID detector and a C18 column (4.6×250 mm, 5 μm, Inertsil ODS-3,
Catalog No.5020-01732). Milli-Q water was used as a mobile phase
with a flow rate of 0.5 ml/min at 35 °C. The liquid product
[7] T. Miyazawa, S. Koso, K. Kunimori, K. Tomishige, Appl. Catal. A 2007, 318,
244–251.
[
8] I. Gandarias, P. L. Arias, J. Requies, M. B. Güemez, J. L. G. Fierro, Appl.
Catal. B 2010, 97, 248–256.
9] J. Oh, S. Dash, H. Lee, Green Chem. 2011, 13, 2004–2007.
[
[
[
[
[
10] D. Coll, F. Delbecq, Y. Aray, P. Sautet, Phys. Chem. Chem. Phys. 2011, 13,
448–1456.
11] S. Zhu, Y. Qiu, Y. Zhu, S. Hao, H. Zheng, Y. Li, Catal. Today 2013, 212,
20–126.
composition was quantified with an external calibration method.
The conversion and selectivity were calculated by Equations (1)–(2).
1
1
(
The error of the carbon balance was within 5% in all the
12] D. D. Falcone, J. H. Hack, A. Y. Klyushin, A. Knop-Gericke, R. Schlögl, R. J.
Davis, ACS Catal. 2015, 5, 5679–5695.
13] R. Arundhathi, T. Mizugaki, T. Mitsudome, K. Jitsukawa, K. Kaneda,
ChemSusChem 2013, 6, 1345–1347.
[14] S. Zhu, X. Gao, Y. Zhu, Y. Li, J. Mol. Catal. A 2015, 398, 391–398.
15] S. García-Fernández, I. Gandarias, Y. Tejido-Núñez, J. Requies, P. L. Arias,
ChemCatChem 2017, 9, 4508–4519.
experiments.)
Conversion ½%ꢀ ¼ ðmol of glycerol convertedÞꢁ
(
1)
2)
1
00=ðinitial mol of glycerolÞ
Selectivity ½%ꢀ ¼ ðcarbon mol of one specific productÞꢁ
00=ðall the carbon mol in glycerol convertedÞ
[
[
[
[
[
[
[
16] J. Dam, K. Djanashvili, F. Kapteijn, U. Hanefeld, ChemCatChem 2013, 5,
(
497–505.
1
17] Y. Fan, S. Cheng, H. Wang, J. Tian, S. Xie, Y. Pei, M. Qiao, B. Zong, Appl.
Catal. B 2017, 217, 331–341.
18] Y. Fan, S. Cheng, H. Wang, D. Ye, S. Xie, Y. Pei, H. Hu, W. Hua, Z. H. Li, M.
Qiao, B. Zong, Green Chem. 2017, 19, 2174–2183.
The turnover frequency (STY) of 1,3-PDO at conversion less than
0% was calculated as follows [Eq. (3)–(5)]:
1
19] J. Wang, X. Zhao, N. Lei, L. Li, L. Zhang, S. Xu, ChemSusChem 2016, 9,
784–790.
STY ¼ ðmol of glycerol converted ꢁ S1,3-PDO ꢁ M1,3-PDOÞ=
ðcatalyst weight ꢁ reaction timeÞ
20] O. Kirichenko, V. Nissenbaum, G. Kapustin, E. Redina, K. Vikanova, N.
Davshan, L. Kustov, J. Therm. Anal. Calorim. 2019, 138, 2205–2218.
21] W. Zhou, Y. Li, X. Wang, D. Yao, Y. Wang, S. Huang, W. Li, Y. Zhao, S.
Wang, X. Ma, J. Catal. 2020, 388, 154–163.
(3)
(4)
(5)
STYPt ¼ ðmol of glycerol converted ꢁ S1,3-PDO ꢁ M1,3-PDOÞ=
ðcatalyst weight ꢁ Pt loading ꢁ reaction timeÞ
[22] S. Zhu, X. Gao, Y. Zhu, J. Cui, H. Zheng, Y. Li, Appl. Catal. B 2014, 158–
159, 391–399.
[
[
[
23] S. García-Fernández, I. Gandarias, J. Requies, M. B. Güemez, S. Bennici, A.
Auroux, P. L. Arias, J. Catal. 2015, 323, 65–75.
24] W. Zhou, J. Luo, Y. Wang, J. Liu, Y. Zhao, S. Wang, X. Ma, Appl. Catal. B
S
T
Y
¼
ð
m
o
l
o
f
g
l
y
c
e
r
o
l
c
o
n
v
e
r
t
e
d
ꢁ
S
W
1
,
3
-
PDO
ꢁ M1,3 PDOÞ=
-
ðcatalyst weight ꢁ W loading ꢁ reaction timeÞ
2
019, 242, 410–421.
25] N. Lei, Z. Miao, F. Liu, H. Wang, X. Pan, A. Wang, T. Zhang, Chin. J. Catal.
020, 41, 1261–1267.
[26] T. Aihara, H. Miura, T. Shishido, Catal. Today. 2020, 352, 73–79.
2
Acknowledgements
[
27] N. Lei, X. Zhao, B. Hou, M. Yang, M. Zhou, F. Liu, A. Wang, T. Zhang,
ChemCatChem 2019, 11, 3903–3912.
28] T. Aihara, H. Miura, T. Shishido, Catal. Sci. Technol. 2019, 9, 5359–5367.
This work was financially supported by the National Natural
Science Foundation (21978194 and U1910202), the Natural
Science Foundation of Shanxi Province (Grant No.
[
[29] S. García-Fernández, I. Gandarias, J. Requies, F. Soulimani, P. L. Arias,
B. M. Weckhuysen, Appl. Catal. B 2017, 204, 260–272.
[
30] J. S. da Cruz, M. A. Fraga, S. Braun, L. G. Appel, Appl. Surf. Sci. 2007, 253,
160–3167.
[31] X. Wu, L. Zhang, D. Weng, S. Liu, Z. Si, J. Fan, J. Hazard. Mater. 2012,
25–226, 146–154.
32] B. Hu, H. Liu, K. Tao, C. Xiong, S. Zhou, J. Phys. Chem. C 2013, 117,
6385–26395.
33] C. Liu, C. Zhang, S. Sun, K. Liu, S. Hao, J. Xu, Y. Zhu, Y. Li, ACS Catal.
015, 5, 4612–4623.
34] L. Dimitrov, R. Palcheva, A. Spojakina, K. Jiratova, J. Porous Mater. 2011,
8, 425–434.
2
01801D121055), and Program for the Shanxi Young Sanjin
3
Scholar. Qian He would like to acknowledge the support by
National Research Foundation (NRF) Singapore, under its NRF
Fellowship (NRF-NRFF11-2019-0002) and the National University
of Singapore Flagship Green Energy Program.
2
[
[
[
2
2
1
Conflict of Interest
[
[
35] X. Chen, G. Clet, K. Thomas, M. Houalla, J. Catal. 2010, 273, 236–244.
36] K. Song, H. Zhang, Y. Zhang, Y. Tang, K. Tang, J. Catal. 2013, 299, 119–
128.
The authors declare no conflict of interest.
[
37] G. Chen, C. Xu, X. Huang, J. Ye, L. Gu, G. Li, Z. Tang, B. Wu, H. Yang, Z.
Zhao, Z. Zhou, G. Fu, N. Zheng, Nat. Mater. 2016, 15, 564–569.
38] N. S. F. V. M. Benitez, Catal. Commun. 2002, 3, 487–492.
39] T. Kitano, T. Hayashi, T. Uesaka, T. Shishido, K. Teramura, T. Tanaka,
ChemCatChem 2014, 6, 2011–2020.
40] D. Kaucký, B. Wichterlová, J. Dedecek, Z. Sobalik, I. Jakubec, Appl. Catal.
A 2011, 397, 82–93.
[
[
Keywords: PtÀ WOx
glycerol · interface
·
1,3-propanediol
·
hydrogenolysis
·
[
ChemCatChem 2021, 13, 1–12
www.chemcatchem.org
10
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cctc.202100773
ChemCatChem
[
41] S. García-Fernández, I. Gandarias, Y. Tejido-Núñez, J. Requies, P. L. Arias,
[54] A. Guntida, K. Suriye, J. Panpranot, P. Praserthdam, Catal. Today. 2020,
358, 354–369.
ChemCatChem 2017, 9, 4508–4519.
[
[
42] M. J. Kale, P. Christopher, ACS Catal. 2016, 6, 5599–5609.
43] K. Ding, A. Gulec, A. M. Johnson, N. M. Schweitzer, G. D. Stucky, L. D.
Marks, P. C. Stair, Science. 2015, 350, 189–192.
[55] S. Boonpai, K. Suriye, B. Jongsomjit, J. Panpranot, P. Praserthdam, Catal.
Today. 2020, 358, 370–386.
[56] S. Boonpai, S. Wannakao, J. Panpranot, B. Jongsomjit, P. Praserthdam, J.
Phys. Chem. C 2020, 124, 15935–15943.
[57] M. Yang, X. Zhao, Y. Ren, J. Wang, N. Lei, A. Wang, T. Zhang, Chin. J.
Catal. 2018, 39, 1027–1037.
[58] J. H. Lee, C. H. Shin, Y. W. Suh, J. Mol. Catal. 2017, 438, 272–279.
[59] J. Chaminand, L. A. Djakovitch, P. Gallezot, P. Marion, C. Pinel, C. Rosier,
Green Chem. 2004, 6, 359–361.
[
[
[
44] Z. Zhang, Y. Zhu, H. Asakura, B. Zhang, J. Zhang, M. Zhou, Y. Han, T.
Tanaka, A. Wang, T. Zhang, N. Yan, Nat. Commun. 2017, 8, 1–10.
45] L. DeRita, S. Dai, K. Lopez-Zepeda, N. Pham, G. W. Graham, X. Pan, P.
Christopher, J. Am. Chem. Soc. 2017, 139, 14150–14165.
46] J. Park, S. Lee, H. E. Kim, A. Cho, S. Kim, Y. Ye, J. W. Han, H. Lee, J. H.
Jang, J. Lee, Angew. Chem. Int. Ed. 2019, 58, 16038–16042; Angew. Chem.
2
019, 131, 16184–16188.
[60] B. Zhao, Y. Liang, L. Liu, Q. He, J. Dong, Green Chem. 2020, 8254–8259.
[61] C. Yang, F. Zhang, N. Lei, M. Yang, F. Liu, Z. Miao, Y. Sun, X. Zhao, A.
Wang, Chin. J. Catal. 2018, 39, 1366–1372.
[62] L. Liu, T. Asano, Y. Nakagawa, M. Gu, C. Li, M. Tamura, K. Tomishige,
Appl. Catal. B 2021, 292, 120164.
[
[
47] X. Zhu, B. Cheng, J. Yu, W. Ho, Appl. Surf. Sci. 2016, 364, 808–814.
48] B. Qiao, A. Wang, X. Yang, L. F. Allard, Z. Jiang, Y. Cui, J. Liu, J. Li, T.
Zhang, Nat. Chem. 2011, 3, 634–641.
49] Q. Sun, B. W. J. Chen, N. Wang, Q. He, A. Chang, C. Yang, H. Asakura, T.
Tanaka, M. J. Hülsey, C. Wang, J. Yu, N. Yan, Angew. Chem. 2020, 132,
[
2
0358–20366; Angew. Chem. Int. Ed. 2020, 59, 20183–20191.
50] J. De Graaf, A. J. Van Dillen, K. P. De Jong, D. C. Koningsberger, J. Catal.
001, 203, 307–321.
51] M. A. Arribas, F. Márquez, A. Martínez, J. Catal. 2000, 190, 309–319.
52] M. Massa, A. Andersson, E. Finocchio, G. Busca, F. Lenrick, L. R.
Wallenberg, J. Catal. 2013, 297, 93–109.
[
2
[
[
Manuscript received: May 27, 2021
Revised manuscript received: June 15, 2021
Accepted manuscript online: June 15, 2021
Version of record online: ■■■, ■■■■
[
53] A. Guntida, K. Suriye, J. Panpranot, P. Praserthdam, Top. Catal. 2018, 61,
1641–1652.
ChemCatChem 2021, 13, 1–12
www.chemcatchem.org
11
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPERS
Hydrogenolysis: A highly-dispersed
B. Zhao, Y. Liang, Prof. L. Liu*, Prof. Q.
He*, Prof. J.-X. Dong
WO species supported on inert α-
x
Al O was successfully fabricated by a
2
3
1
– 12
simple high-temperature heat-
treatment method according to
solid-state reaction principles. The
Facilitating PtÀ WO
Species Interac-
x
tion for Efficient Glycerol Hydroge-
nolysis to 1,3-Propanediol
amount of PtÀ WO interface active
x
sites could be rationally adjusted by
Pt loading, contributing to a remark-
able W and Pt efficiency toward 1,3-
PDO production.