Insight into the nature of Br?nsted acidity of Pt-(WOx)n-H model catalysts in glycerol hydrogenolysis

Article abstract of DOI:10.1016/j.jcat.2020.05.019

In glycerol hydrogenolysis, the conventional preparation methods for Pt-WO_x catalysts on active supports usually result in the coexistence of various active sites, causing complex reaction network and declining the atom efficiency of W toward 1

Full text of DOI:10.1016/j.jcat.2020.05.019

Journal Pre-proofs
Insight into the Nature of Brönsted Acidity of Pt-(WOx)n-H Model Catalysts
in Glycerol Hydrogenolysis
Wei Zhou, Ying Li, Xiaofei Wang, Dawei Yao, Yue Wang, Shouying Huang,
Wei Li, Yujun Zhao, Shengping Wang, Xinbin Ma
PII:
S0021-9517(20)30191-3
https://doi.org/10.1016/j.jcat.2020.05.019
YJCAT 13749
DOI:
Reference:
To appear in:
Journal of Catalysis
Received Date:
Revised Date:
Accepted Date:
15 February 2020
11 May 2020
13 May 2020
Please cite this article as: W. Zhou, Y. Li, X. Wang, D. Yao, Y. Wang, S. Huang, W. Li, Y. Zhao, S. Wang, X.
Ma, Insight into the Nature of Brönsted Acidity of Pt-(WOx)n-H Model Catalysts in Glycerol Hydrogenolysis,
Journal of Catalysis (2020), doi: https://doi.org/10.1016/j.jcat.2020.05.019
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Elsevier Inc. All rights reserved.

Insight into the Nature of Brönsted Acidity of Pt-(WOx)n-H
Model Catalysts in Glycerol Hydrogenolysis
Wei Zhou1, Ying Li1, Xiaofei Wang1, Dawei Yao1, Yue Wang1,2, Shouying Huang1,2, Wei Li3,
Yujun Zhao1, Shengping Wang1 and Xinbin Ma1,2*
1 Key Laboratory for Green Chemical Technology of Ministry of Education, Collaborative
Innovation Centre of Chemical Science and Engineering, School of Chemical Engineering and
Technology, Tianjin University, Tianjin 300072, China
2 Joint School of National University of Singapore and Tianjin University, International Campus
of Tianjin University, Binhai New City, Fuzhou 350207, China
3 Department of Polymer Science and Engineering, School of Material Science and Chemical
Engineering, Ningbo University, Ningbo 315211, P.R. China
*Corresponding author: Prof. Xinbin Ma
E-mail: xbma@tju.edu.cn
Fax: +86-22-87401818
Tel: +86-22-27409248

Abstract: In glycerol hydrogenolysis, the conventional preparation methods for Pt-WOx catalysts
on active supports usually result in the coexistence of various active sites, causing complex
reaction network and declining the atom efficiency of W toward 1,3-propanediol (1,3-PDO)
production. Herein, we report Pt-(WOx)n-H model catalyst prepared via strong metal–oxide
interaction (SMOI) behavior on inert SiO2 support, which shows both high performance and W
efficiency for 1,3-PDO. The W efficiency as high as 5.786 h-1 was obtained on 2W4Pt/SiO2 at 160
°C under 8.0 MPa. Furthermore, these model catalysts also give the advantage in revealing the
Brönsted acidity of Pt-(WOx)n-H sites. Structure characterizations and DFT calculation confirm
the Brönsted acid strength of Pt-(WOx)n-H is significantly sensitive to both the polymerization and
reduction degree of WOx domains. The high catalytic efficiency of 2W4Pt/SiO2 originates from
the strong Brönsted acidity and its synergism with Pt metal. The correlation between Brönsted
acidity and electronic state of Pt-(WOx)n-H is also discussed by electron density difference analysis
and XPS.
Keywords: 1,3-propanediol, Brönsted acid, WOx domain, Pt, glycerol, hydrogenolysis

1. Introduction
The process for fuel and chemical production using biomass is attractive since it is
environmental-friendly and sustainable. Because of the high oxygen content as feedstock,
hydrodeoxygenation becomes one of the important reactions during the upgrading of biomass,
including the C-O cleavage of ethers [1-3], esters [4,5], aromatic [6,7] and aliphatic [8-10] alcohols,
et al. Glycerol is the simplest biomass platform compound containing both primary –OH and
secondary –OH, and the hydrogenolysis of glycerol into useful C3 chemicals is gaining much
attention [11,12]. Among these processes, selective C-O cleavage to 1,3-propanediol (1,3-PDO) is
still one of the challenging problems. WOx based materials have been widely investigated in this
reaction because of its high selectivity on 1,3-PDO. As mentioned by many groups, Brönsted acid
site is critical for 1,3-PDO formation [13-20]. However, as the lack of detail structural information
on the active site, the nature of Brönsted acidity of Pt-WOx catalysts is still not clear.
Recently, we found that 1,3-PDO formation rate can be adjusted by the size of WOx species,
and proposed the intrinsic active site of supported Pt-WOx catalysts is Pt-(WOx)n-H structure with
medium polymeric WOx domains, which can provide both Pt metal and super strong Brönsted acid
adjacent with each other [21]. Based on this viewpoint, some factors should be considered during
the design of effective catalysts.
Firstly, for the construction of Pt-(WOx)n-H site, ensuring WOx domains on Pt surface is the
prerequisite. However, in most reported methods, loading WOx species on the support is the first
step, and Pt metal is introduced after that. In consequence, only a few amounts of active sites could
be formed during the heat treatment and reduction processes. Meanwhile, most WOx species are
existed on the catalysts without connecting with Pt metal, which may contribute to side reactions.

Secondly, the type of support affects much on the diversity of active sites on the catalyst surface.
Our previous studies found that the independent Pt metal can catalyze 1,2-PDO formation and C-
C cleavage reaction [22], while the independent WOx species can’t provide Brönsted acid with
enough strength and excess WOx domains give rise to deep hydrogenolysis to propanol [21].
Active supports such as ZrO2, Al2O3 and TiO2 can be strongly interacted with both Pt and WOx,
and they also have original acid sites on their surface, leading to various kinds of active structures
and intrinsically reducing the selectivity of 1,3-PDO. On the other hand, inert supports such as
SiO2 have much lower interaction with both Pt and WOx, which improve the possibility of the
interaction between Pt and WOx during the preparation. Additionally, the structural sensitivity of
Brönsted acidity on Pt-(WOx)n-H site should also be emphasized, which is not only related to the
size, but also the reduction degree of WOx. Because of the diversity of WOx states and locations
on the catalysts, results of characterization and catalytic performances usually do not reflect the
natural properties of Pt-(WOx)n-H site.
Up to now, achieving a large amount of highly dispersed WOx on the support seems to be the
guideline of supported Pt-WOx catalysts in glycerol hydrogenolysis, and the active supports are
widely used, with a general loading sequence (WOx before Pt) during catalyst preparation.
Recently, Fan et al. broke the conventional idea from the preparation aspect. Before their
impregnation, ZrO2 was calcined at 600 °C for 5 h, which could reduce its surface –OH and other
anchoring sites. Then Pt species was loaded on the ZrO2 support, and after calcination, the WOx
precursor was subsequently impregnated. The reaction rate for 1,3-PDO over Pt-WOx/t-ZrO2
-1
catalyst reached 0.102 g1,3-PDO ∙ gcat ∙ h-1 [20]. The passivation of ZrO2 support under high
temperature and the distinctive loading sequence both increased the amount of Pt-(WOx)n-H sites,
which may contribute to its excellent catalytic performance. In another work, they synthesized W-

SBA-15 support with a very little W content using a modified sol-gel method, and applied it in
glycerol hydrogenolysis after loading Pt. The selectivity of 1,3-PDO is 70.8%, which is the highest
among Pt-WOx catalysts [19]. It seems that inert SiO2 always has a good effect on 1,3-PDO
selectivity although it is not a popular support in WOx based catalysts. In an earlier work, Gong et
al. investigated different preparation conditions of Pt/WO3/TiO2/SiO2 catalyst and found the SiO2
support can improve 1,3-PDO selectivity than other supports [23]. Priya et al. impregnated WOx
and then Pt species to SBA-15 support and applied it in glycerol hydrogenolysis, the selectivity of
1,3-PDO reached 42% [24]. In addition, Zhu et al. [25] investigated the promoting of Pt/WOx/ZrO2
catalysts by SiO2, and found a suitable amount of SiO2 can improve 1,3-PDO selectivity from 44.3%
to 52.0%. All the reports above demonstrates the diversity of active species on supported Pt-WOx
catalysts can be significantly influenced by both the loading sequence and support types.
In their critical review [26], Tomishige et al. proposed the importance of strong metal–oxide
interaction (SMOI) behavior in preparing various metal-oxophilic metal oxide catalysts, which
have shown super performances in selective hydrogenolysis of biomass. Under this thought, they
have applied a series of inert SiO2 supported metal-ReOx catalysts in glycerol hydrogenolysis [27].
The loading sequence in their preparation was always metal precursor before ReOx species. In this
work, SMOI behavior is also used in preparing Pt-WOx catalysts on SiO2 support. In order to
exclude the independent WOx species and ensure Pt-(WOx)n-H as the only active site, the mole
ratio of W/Pt on the catalysts was kept below 1.0. By this way, we try to build the pure Pt-(WOx)n-
H active sites and investigate the effects of WOx structure and reduction degree on the catalytic
performance. Combined with characterizations and DFT calculation, the nature of Brönsted acidity
on Pt-(WOx)n-H is also revealed.
2. Experimental and Theoretical Methods

2.1. Catalyst Synthesis
The sequential incipient wetness impregnation was applied to synthesis the catalysts. Firstly,
an appropriate amount of Pt(NO3)2 aqueous solution (10% Pt, Beijing HWRK Chem.) was dropped
onto SiO2 support, then they were well mixed. After drying at 100 °C overnight, the solid was
pretreated in pure hydrogen for two hours at 250 °C. Then the (NH4)6(H2W12O40)·nH2O (>99.9%,
J&K Chemical Corp.) aqueous solution was mixed with the dark sample, while the concentration
of the solution was changed to achieve different mole ratios of W/Pt on the catalysts. After that,
the sample was dried at 100 °C and calcined under air for 4 h at 500 °C. The obtained catalysts
were named as xW4Pt/SiO2, in which the Pt content was fixed to 4 wt% while the W/Pt mole ratio
was x/4 (x = 0.5, 1, 2 ,4).
For comparison, the normal impregnation process was also used and the corresponding sample
was labelled as 4PtxW/SiO2 (x = 0.5, 1, 2 ,4). Briefly, SiO2 support was firstly impregnated with
aqueous solution of ammonium metatungstate and dried at 100 °C. After the calcination for 4 h
under air at 500 °C, the Pt(NO3)2 aqueous solution was impregnated to the sample and calcined
again at 350 °C for 2 h.
2.2. Catalyst Characterization
The crystalline phase of catalysts was characterized by a Rigaku D/MAX-2500 diffractometer
(Cu K,λ = 1.5406 Å) using a speed 8°/min. The catalyst powder was reduced at 180 °C under H2
for 2 h before testing. The transmission electron microscopy (TEM) was performed under 200 kV
on a JEM-2100F instrument. Sample was grinded and pretreated in hydrogen for 2 h at 180 °C.
After that, it was dispersed in ethanol under ultrasonic for half an hour. Subsequently, the
suspended sample was dropped onto an ultrathin carbon film supported on copper grid and dried

under the infrared lamp. Raman spectroscopy was used to determine the structure and
polymerization degree of WOx domains on the calcined catalysts. The instrument is a Renishaw
inVia reflex Raman spectrometer and the wavenumber of the laser source is 532 nm. The result
was accumulated by 3 scans, with the speed of 10 seconds/scan. The PHI 1600 ESCA spectrometer
(PE Company) was applied to collect X-ray photoelectron spectroscopy of catalysts. The radiation
source was an Al Kα X-ray with the hν of 1486.6 eV. The dried sample was grinded to powder
and pressed to a self-supporting wafer, and then it was reduced at 180 °C for 2 h and cooled to
room temperature. Then the sample was sealed into centrifuge tube filled with Ar. After introduced
into the sample chamber of the instrument within 5 minutes, the sample was degassed by
evacuation. The C 1s peak (284.6 eV) was applied to calibrate the binding energy, within an error
of ±0.2 eV.
The dispersion of Pt particles and H2 adsorption capacity on catalyst were obtained from CO
and H2 chemisorption experiments over a Micromeritics Autochem 2910 analyzer. In CO
chemisorption, about 50 mg (40–60 mesh) catalyst was pretreated in hydrogen for two hours at
180 °C. After purged by He flow, it was cooling to 50 °C. CO adsorption was carried out by pulses
of 10% CO/He (1.076 mL) until the sample surface was saturated. The CO/Pt ratio of 1.0 was used
for the dispersion calculation. During H2 chemisorption, 10% H2/Ar was used as the adsorption
gas, and Ar is used as the purge/carrier gas. A H2/Pt theoretical ratio of 0.5 was used, and the mole
ratio of consumed H atoms amount to Pt atoms amount on the surface was labelled as H/Pt. H2-
temperature programmed reduction (TPR) was also performed using the above instrument. The
catalyst was degassed at 350 °C for one hour under Ar, then it was cooling to 50 °C. After the
system became stable under 10% H2/Ar, the TPR step began from 50 to 780 °C (10 °C/min). NH3
temperature-programmed desorption (NH3-TPD) was measured using a mass spectrometric

detector. Sample was degassed under He at 350 °C for one hour. Subsequently, the system was
dropped to 50 °C and 10% NH3/He was introduced for half an hour. After arriving a steady baseline
under He, the system temperature was increased to 750 °C. To exclude the interference by H2O,
the signal of NH3 with an m/z of 16 was used [28].
DRIFT spectra of CO adsorption under room temperature were determined on a Nicolet 6700
spectrometer. After reduction at 180 °C in the diffuse reflection cell, the catalyst was flushed under
He flow. After recording the background, the gas was changed to 10% CO/He and kept for half an
hour. Subsequently, He flow was introduced again and the result was obtained when the infrared
signal was not changed. In addition, this characterization was also conducted at -130 °C to
investigate the acid species on catalyst surface. The sample was degassed by N2 at 350 °C for one
hour, followed with an evacuation until the temperature was dropped to -130 °C. Subsequently,
the background was record and CO pulse was introduced to the system until the adsorption
spectrum was stable. Then the system was exposed to vacuum again and the spectrum was
collected until it was stable.
2.3. Catalytic reactions
2.0 g glycerol, 18.0 g H2O and 0.25 g catalyst were mixed in the autoclave with a volume of
100 ml. The hydrogen pressure was raised to 5.5 MPa after purging. Then, the temperature was
increased to 180 C, during which the consumed H2 was replenished by pulse. The system was
kept at 8.0 MPa and the stirring speed was 1000 rpm. After the reaction, the liquid product
composition was analyzed by Agilent Micro GC 4890 with a DB-FFAP capillary column. Before
the analysis, n-Butanol and 1,4-butanediol were added into the liquid as the internal standards. The

gas composition was analyzed using Agilent GC 7890B with a HP-AL/S column. The conversion
was calculated as:
conversion (%) = (the consumed moles of glycerol) / (the initial moles of glycerol)
And the selectivity was calculated as:
selectivity (%) = (moles of specific product) / (the consumed moles of glycerol)
2.4. Theoretical Methods
DFT calculations were performed using the Material Studio 6.1 software with the Cambridge
sequential total energy package (CASTEP), while plane-wave basis sets and ultrasoft
pseudopotential were employed [29]. Generalized gradient approximation (GGA) method was
applied to describe the electrons interaction with atomic cores, and Perdew-Burke-Ernzerhof (PBE)
functional was chosen to describe the exchange and correlation energies [30]. The plane-wave
energy cutoff was set as 400 eV, and the surface Brillouin zone was sampled with a 3×3×1
Monkhorst-Pack k-point mesh [31,32]. The WOx/Pt model was constructed with the HWxOy
cluster supported on a (4 × 4) periodic Pt(111) slab of 4-layer. The vacuum space between two
slabs along z-direction was 20 Å and the bottom two Pt atom layers were restrained to the
optimized bulk Pt state. The convergence tolerance of energy change, maximum displacement and
force in the geometry optimization and energy calculation were 2×10-5 eV/atom, 0.002 Å and 0.05
eV/Å.
The NH3 adsorption energy (Eads) on surface Brönsted acid site of the model was estimated by
the equation: Eads = E(NH3/HWxOy/Pt) – E(NH3) – E(HWxOy/Pt), in which E(NH3/HWxOy/Pt),

E(NH3) and E(HWxOy/Pt) represent the total energy of HWxOy/Pt vacuum slab with adsorbed NH3,
free NH3 molecule and HWxOy/Pt vacuum slab, respectively.
3. Results and discussion
3.1. Advantage of SMOI on the preparation of highly efficient catalyst
Firstly, the performance of xW4Pt/SiO2 prepared via SMOI in glycerol hydrogenolysis is
studied. As shown in Fig. 1a, the conversion of glycerol over 4Pt/SiO2 is 7.9%, and the
corresponding selectivity of 1,3-PDO is only 13.2%. The main product is 1,2-PDO, with its
selectivity of 63.0%, while the product selectivity from C-C cleavage is as high as 13.1%, which
reflect a typical hydrogenolysis property on the independent Pt metal. With the W/Pt mole ratio
increases from 0 to 1/4, the conversion is raised to 20.3%, accompanied by a 1,3-PDO selectivity
development up to 36.5% and a drop of 1,2-PDO selectivity to 38.2%. The increased activity of
C-O bond hydrogenolysis is ascribed to the acidity introduced by WOx. Interestingly, when the
W/Pt mole ratio reaches to 2/4, a sharp improvement of catalytic performance is obtained, with a
64.2% conversion and a 57.2% selectivity of 1,3-PDO. In addition, 1,2-PDO selectivity is dropped
to only 10.5%, while 1-PO selectivity is increased obviously from 8.9% to 18.6%. The activity is
increased for more than three times when W/Pt ratio increased from 1/4 to 2/4, and a much higher
1,3-PDO selectivity is also observed, indicating that not only the amount, but also the acidity of
active site is enhanced on 2W4Pt/SiO2. Further increment of W/Pt mole ratio to 4/4 has little effect
on the product distribution, while the conversion is declined to 37.2%. The apparent activation
energy (Eapp) of 1,3-PDO formation over the three catalysts (W/Pt ratio of 1/4, 2/4 and 4/4) are
also compared. As shown in Fig. 1b, 2W4Pt/SiO2 has the lowest Eapp for this reaction, indicating
a stronger acidity on this catalyst for secondary C-O bond activation.

100
90
80
70
60
50
40
30
20
10
0
8
6
4
2
(b)
Conversion
(a)
E
app = 38  6 kJ/mol
others
2-PO
1-PO
1,2-PDO
1,3-PDO
E
app = 55  2 kJ/mol
E
app = 89  4 kJ/mol
2W4Pt/SiO2
4W4Pt/SiO2
1W4Pt/SiO2
2
2
2
2
2
O
i
2.30
2.35
2.40
2.45
(1000/T) / K-1
2.50
2.55
O
O
O
O
i
i
i
i
S
/
t
S
S
S
S
/
t
P
/
/
t
/
t
t
P
P
P
P
4
100
90
80
70
60
50
40
30
20
10
0
4
4
4
4
W
W
W
W
5
.
1
2
4
0
7
6
5
4
3
2
1
0
(c)
Conversion
(d)
others
2-PO
1-PO
1,2-PDO
1,3-PDO
4PtxW/SiO2
xW4Pt/SiO2
2
2
2
2
O
i
0
1
2
3
4
5
O
O
O
i
i
i
S
/
S
/
S
S
/
W
/
W
W
W
2
5
1
4
t
P
.
0
t
t
t
WOx loading / wt.%
P
P
4
4
4
P
4
Fig. 1. (a) Glycerol hydrogenolysis performances and (b) apparent activation energies (Eapp) of
1,3-PDO formation over xW4Pt/SiO2 catalysts. (c) Glycerol hydrogenolysis performances over
4PtxW/SiO2 catalysts. (d) Trend of STYW with the increase of WOx loading. Reaction conditions:
2.0 g glycerol, 18.0 g water, 0.25 g catalyst, 180 °C, 8 MPa, 1000 rpm, 18 h. The catalyst was not
reduced in H2 before reaction. 1,3-PDO: 1,3-propanediol; 1,2-PDO: 1,2-propanediol; 1-PO: 1-
propanol; 2-PO: 2-propanol; others include ethylene glycol, ethanol, and methanol, et al.
For comparison, reaction results of 4PtxW/SiO2 catalysts prepared by the conventional method
(introducing Pt after WOx/SiO2 is obtained) are given in Fig. 1c. The highest performance is also
obtained on the catalyst with a W/Pt mole ratio of 2/4 (4Pt2W/SiO2), with a 40% conversion and
a 52.7% selectivity of 1,3-PDO. However, the activity and selectivity are both lower than those on
2W4Pt/SiO2. To evaluate W efficiency, the 1,3-PDO production rate based on WOx mass (STYW)

over the two series catalysts were presented in Fig. 1d. A volcano relationship was found between
STYW and WOx loading over both series catalysts, reflecting the 1,3-PDO formation rate is
structurally sensitive on WOx polymerization degree, which is consistent with our previous study
[21]. At WOx loading of 2.4 wt.% (W/Pt = 2/4), STYW over 2W4Pt/SiO2 is about two times of that
over 4Pt2W/SiO2, confirming the advantage of SMOI behavior in the construction of Pt-(WOx)n-
H active site.
To further compare the catalytic performances of reported Pt-WOx catalysts, Table S1 gives
the activity values based on catalyst mass (STY), surface exposed Pt mass (STYPt) and WOx mass
-1
(STYW). Under a very low WOx loading, the STY of 2W4Pt/SiO2 reaches 0.135 g1,3-PDO ∙ gcat
∙
h-1, ranking the top two among reported results. The dependence of STYPt and STYW with STY is
shown in Fig. 2a and b, respectively. The STYPt and STYW of 2W4Pt/SiO2 ranks top four and top
two, respectively. Noted that a random distribution is presented between STYPt and STY,
indicating that the activity based on surface exposed Pt sites is not suitable to evaluate the catalytic
performance. On the other hand, the positive correlation is apparent between STYW and STY,
which means a catalyst with high activity usually shows a high STYW value. The exception of
Pt/W-SBA-15 sample is derived from its extremely low WOx loading, leading to a high STYW
while the STY is low [19]. We can see from Fig. 2b that 2W4PtSiO2 gives both high STYW and
STY values, revealing high W efficiency of this material among the reported catalysts. In addition,
the statistical observation in Fig. 2 also demonstrates that WOx state is more important than Pt
state during 1,3-PDO formation on Pt-WOx bifunctional system.

0.20
0.16
0.12
0.08
0.04
(a)
(b)
2W4PtSiO2
2W4PtSiO2
Pt/W–SBA-15
8 10 12
0
5
10
15
20
25
0
2
4
6
STYPt / h-1
STYW / h-1
Fig. 2. The dependence of catalytic activity on surface Pt (a) or WOx (b) based STY values.
3.2. Catalytic performance of 2W4Pt/SiO2 under various conditions
Wang et al. reported that the over reduction of WOx species can “poison” the catalysts and
decrease the activity [16]. It is predictable that besides the polymerization of WOx domain, the
reduction of WOx can also influence the catalytic performance of Pt-(WOx)n-H. So the effect of
reduction temperature of 2W4Pt/SiO2 in H2 was investigated. As shown in Table 1, with the
increase of reduction temperature, both the conversion and 1,3-PDO selectivity are decreased,
accompanied with a 1,2-PDO selectivity increment from 10.5% to 22.3%. The best performance
of 2W4Pt/SiO2 is achieved when it is directly used in the reaction without reduction pretreatment.
Therefore, the reduction of WOx on Pt-(WOx)n-H site is detrimental to 1,3-PDO production, which
is ascribed to the weakening of Brönsted acidity as discussed below.

Table 1. The effect of reduction degree of 2W4Pt/SiO2 in glycerol hydrogenolysis.
Conversionb
(%)
Selectivity (%)
Reductiona
1,3-PDO 1,2-PDO 1-PO 2-PO Others
None
250 °C
400 °C
64.2
20.5
16.3
57.2
51.0
47.7
10.5
20.0
22.3
18.6
16.2
15.7
12.1
10.3
10.4
1.6
2.5
3.9
a The 2W4Pt/SiO2 catalyst was not reduced, reduced in H2 at 250 °C and 400 °C before reaction,
respectively.
b Reaction conditions are the same as that in Fig. 1.
Considered that the reduction of Pt-(WOx)n-H may also occur during the reaction, the
performance of 2W4Pt/SiO2 was investigated in different reaction temperature and pressure. As
shown in Fig. 3a, with the reaction temperature dropping from 180 °C to 160 °C, the activity drops
from 64.2% to 58.2%, followed with an increment of 1,3-PDO selectivity from 57.2% to 64.9%.
The decrease of reaction temperature results in a increment of 1,3-PDO STY from 0.135 to 0.139
g1,3-PDO ∙ gcat-1 ∙ h-1. Although this is not such significant, it indicates that partial reduction of Pt-
(WOx)n-H occurs under the reaction condition of 180 °C and 8 MPa, leading to a lower 1,3-PDO
formation rate. Further declining the reaction temperature to 120 °C causes a gradual decrease of
glycerol conversion from 58.2% to 19.2%, while the product distribution is not obviously
influenced.

100
80
60
40
20
0
(a)
Conversion
others
2-PO
1-PO
1,2-PDO
1,3-PDO
180
160
140
120
Temperature / C
100
80
60
40
20
0
(b)
Conversion
others
2-PO
1-PO
1,2-PDO
1,3-PDO
8.0
5.2
2.5
0.5
Pressure / MPa
Fig. 3. Effects of (a) reaction temperature and (b) pressure on glycerol hydrogenolysis over
2W4Pt/SiO2 catalyst. The pressure effect was investigated at 140 °C, while other conditions were
the same as that in Fig. 1.
The influence of reaction pressure at 140 °C was given in Fig. 3b. Above 5.2 MPa, the catalytic
performance of 2W4Pt/SiO2 is not affected much by the reaction pressure. When the pressure
drops to 2.5 MPa, the conversion of glycerol significantly decreased from 36.1% to 8.5%, while
1,3-PDO selectivity is not changed. At 0.5 MPa, the conversion drops to only 1.9%, while the
primary C-O bond and C-C bond cleavage become the dominant reaction, with a total selectivity
of 93.5%. 1,3-PDO is not formed at such low pressure. The observations above demonstrate that
H2 is critical for 1,3-PDO production. At 0.5 MPa, the concentration of active hydrogen species
on the catalyst surface is too low, so 1,3-PDO is not formed. At 2.5 MPa, the dissolution of H2 in
the liquid is improved and active hydrogen species emerge on the catalyst surface, which promotes
the generation of 1,3-PDO. With the raise of H2 pressure to 5.2 MPa, the active hydrogen on the

catalyst surface is saturated and the formation of 1,3-PDO is near the top level. Therefore, further
increment of the reaction pressure to 8 MPa does not contribute to the catalytic performance.
In our previous work, the reaction under different pressure was also studied on Pt/ZrW catalysts
at 180 °C [22]. It was found that 1,3-PDO yield was decreased after 4.2 MPa, which was also
ascribed to the destroy of WOx structure under higher H2 pressure. In addition, the conversion was
the highest at low pressure of 0.7 MPa, with the main products of 1,2-PDO, C2 alcohols and
hydroxyacetone. However, Fig. 3b shows that glycerol was almost not converted over 2W4Pt/SiO2
at 140 °C under 0.5 MPa, and dehydration products (hydroxyacetone or 3-hydroxypropanal) were
not detected, either. This confirms that glycerol dehydration route is not catalyzed by Pt-(WOx)n-
H site at this mild condition, and 1,3-PDO is formed through the direct hydrodeoxygenation
mechanism with the help of both strong Brönsted acid site and active hydrogen, consistent with
the in situ FTIR studies [21].
3.3. Characterization of catalysts
Table 2 shows the physicochemical properties of the catalysts. The dispersion of Pt measured
by CO chemisorption is decreased with WOx increment, which can be ascribed to Pt metal
aggregation or WOx coverage over Pt surface. Meanwhile, the H/Pt ratio is increased from 4.5 to
6.6, demonstrating an enhancement of hydrogen spillover with the existence of WOx domains [33].
Table 2. Physicochemical properties of xW4Pt/SiO2 catalysts.
Pt sizeb (nm)
Dispersiona
(%)
Catalyst
H/Ptc
dCO dXRD dTEM

4Pt/SiO2
58.0
39.8
27.9
19.2
1.9
2.8
4.1
5.9
2.6
3.9
3.5
2.9
2.9
3.7
3.5
2.6
4.5
4.9
6.4
6.6
1W4Pt/SiO2
2W4Pt/SiO2
4W4Pt/SiO2
a The dispersion of Pt was estimated from CO chemisorption.
b The particle size of Pt was estimated from CO chemisorption, XRD and TEM characterizations.
c The H/Pt ratio was derived from CO chemisorption and H2 chemisorption results.
The crystalline structure of different catalysts is presented in Fig. S1. The wide diffraction peak
at 2θ = 22° is assigned to amorphous SiO2 support. In addition, the weak peaks of Pt metal are also
observed on all catalysts (2θ = 39.8° and 46.2°). The diffraction signals of WO3 at about 24.0° and
34.0° are not found here, indicating WOx are highly dispersed on these catalysts. The particle size
of Pt metal is calculated by Scherrer equation and shown in Table 2. Different from the results
from CO chemisorption, Pt particle size estimated from XRD ranges between 2.6 and 3.9 nm. To
further determine the metal size, TEM was used to analyze Pt size distribution. As shown in Fig.
4, WOx loading increment trend to slightly reduce Pt metal size. The mean particle size observed
from TEM images is also consistent with XRD result (Table 2). Thus, the much lower dispersion
derived from CO chemisorption is caused by the existence of WOx domains on Pt metal, indicating
SMOI existed between Pt and WOx clusters. This is also confirmed by the different CO adsorption
states over catalysts in DRIFT spectra. As seen in Fig. S2, the bridge bonded CO is clearly shown
at 1830 cm-1 on Pt metal surface for 4Pt/SiO2. However, on the catalyst 2W4Pt/SiO2, this band is
almost disappeared and only the linear CO species on Pt metal exists at 2080 cm-1, which is caused
by the significant loss of adjacent Pt atoms after partially covered by WOx species. Line scanning
curves of 2W4Pt/SiO2 sample in Fig. 4d also confirms the close contact between Pt metal and WOx
domains.

Fig. 4. TEM images with Pt particle size distribution for 1W4Pt/SiO2 (a), 2W4PtSiO2 (b),
4W4PtSiO2 (c), and HAADF-STEM image with line scanning of 2W4Pt/SiO2 (d).
The structure of WOx is characterized by Raman spectroscopy. For comparison, the sample of
WOx supported on SiO2 without Pt element (2W/SiO2) was also measured. As shown in Fig. 5,
Raman bands of SiO2 support are presented at about 450, 606 and 808 cm-1, which are assigned to
the D1 and D2 defect bands of tetra- and tri-cyclosiloxane rings, and Si-O-Si symmetrical
stretching, respectively [34-36]. For 2W/SiO2, a strong Raman band assigned to hydrated W=O

structure is emerged at 975 cm-1, accompanied with a broad band of W-O-W stretching near 832
cm-1 [37,38] which is overlapped with Si-O-Si symmetrical stretching at 808 cm-1. However on
2W4Pt/SiO2, the hydrated W=O band is redshifted from 975 cm-1 to 947 cm-1. This indicates a
decrease of the WOx domain size on Pt containing catalysts, which is caused by the interaction
between WOx and Pt. For 4W4Pt/SiO2, a blueshift of hydrated W=O band is observed from 947
cm-1 to 953 cm-1, demonstrating a increment of WOx polymerization with raising WOx content.
To elucidate the influence of impregnation order on the catalyst structure, XRD and Raman
characterizations of 4Pt2W/SiO2 are shown in Fig. S3. Large Pt particle size (11.8 nm) and bulk
WO3 phase are found on 4Pt2W/SiO2, which results in a less amount of interacted Pt-WOx species
and a lower catalytic performance.
450
606
808
4Pt/SiO2
947
2W4Pt/SiO2
953
4W4Pt/SiO2
975
2W/SiO2
200
400
600
800
1000 1200
Raman shift / cm-1
Fig. 5. Raman spectra of the calcined xW4Pt/SiO2 catalysts.
The reducibility of different catalysts is shown in Fig. S4. Most of Pt species on 4Pt/SiO2 can
be reduced before 100 °C, and the broad signal at about 390 °C is ascribed to the hydrogen
consumption for small amount of PtOx species with large particle size or interaction with SiO2
support. With the introduction of WOx, the intensity of reduction peak before 100 °C is weakened

and the main H2 consumption signal is started after 120 °C, and the broad peak reflects a gradual
reduction of both PtOx and WOx species on catalyst. The center of reduction peak is decreased
from 390 °C to 308 °C as W/Pt mole ratio is increased to 4/4, reflecting an easier reduction of
higher polymeric WOx species [38]. The quantification analysis of TPR (Table S2) indicates the
optimized average W valence of 2W4Pt/SiO2 for this reaction is about W5.48+-W5.92+
.
Fig. 6a gives NH3-TPD curves of different catalysts, with the peak fitting results of strong acid
region shown in Fig. 6b and Table S3. A peak with strong intensity is presented below 200 °C for
all samples, which represents the desorption of weakly adsorbed NH3 on the samples. For 4Pt/SiO2,
a small peak is also found at about 464 °C, which is ascribed to NH3 desorption on Pt metal surface.
With the W/Pt ratio increased to 1/4, the peak shifts from 464 °C to 407 °C, indicating a decrease
of the adsorption strength of NH3. This may be caused by a partial coverage of strong NH3
adsorption sites on Pt after WOx introduction, as the decline percent of peak area (27%) is similar
to that of Pt dispersion (31%). In addition, as the reaction rate and 1,3-PDO selectivity increasing,
the contribution of WOx sites to the strong acid amount is also not excluded here. When the W/Pt
ratio is increased to 2/4, a new type of acid site is emerged, with the corresponding NH3 desorption
peak at about 519 °C. Further increasing W/Pt ratio to 4/4 will weaken the strength of this type of
acid as the NH3 desorption temperature is dropped to 472 °C, meanwhile, the acid amount is also
significantly reduced. For comparison, NH3-TPD result of 2W/SiO2 sample is also given here, and
a desorption peak assigned to weaker acid sites is observed at about 300 °C. Thus, the strong acid
is created only when WOx species is connected to Pt metal, and is greatly affected by WOx loading.
On 2W4Pt/SiO2, WOx domains on Pt surface are medium polymeric and the corresponding acidity
is the strongest, thus presenting the lowest apparent activation energy in the formation of 1,3-PDO
(Fig. 1b). Fig. 6a and b also show that after reduction pretreatment at 250 °C, the strong acid

amount is obviously decreased on 2W4Pt/SiO2, consistent with its catalytic activity decrease in
Table 1.
(b)
(a)
4Pt/SiO2
1W4Pt/SiO2
2W4Pt/SiO2
4W4Pt/SiO2
2W4Pt/SiO2(reduced)
2W/SiO2
4Pt/SiO2
1W4Pt/SiO2
2W4Pt/SiO2
4W4Pt/SiO2
2W4Pt/SiO2(reduced)
100
200
300
400
500
600
700
300
400
500
600
700
Temperature / C
Temperature / C
(c)
2170
(d)
2156
2156
4W4Pt/SiO2
2170
4W4Pt/SiO2
2W4Pt/SiO2
2W4Pt/SiO2
0.5W4Pt/SiO2
4Pt/SiO2
0.5W4Pt/SiO2
4Pt/SiO2
2240
2220
2200
2180
2160
2240
2220
2200
2180
2160
Wavenumber / cm-1
Wavenumber / cm-1
Fig. 6. (a) NH3-TPD curves of xW4Pt/SiO2 catalysts and (b) peak fitting results of the strong acid
region. 2W4Pt/SiO2(reduced) refers to the catalyst reduced at 250 °C before NH3-TPD
measurement. DRIFT spectra of CO adsorption at low temperature: (c) spectra after 8 CO doses
into the sample cell, (d) spectra after vacuum extraction.
Because of the darkness and low WOx content of catalysts, pyridine adsorption FTIR is failed
to characterize the acidity of catalysts. Instead, we carried out DRIFT experiments of CO
adsorption under a low temperature of -130 °C to investigate the surface acids. The adsorbed CO

on Lewis acid sites gives characteristic peaks at above 2177 cm-1, while bands at 2170 cm-1 and
2156 cm-1 are ascribed to CO adsorbed on strong and weak Brönsted acids, respectively [39-42].
As shown in Fig. 6c, all samples mainly exhibit Brönsted acids sites, and the bands correlated to
Lewis acids are only presented as shoulder peaks in the spectra. Weak Brönsted acids are dominant
on 4Pt/SiO2. With the WOx content increasing, the band correlated to strong Brönsted acids at
2170 cm-1 develops and become stronger than the band at 2156 cm-1. After exposed to vacuum,
2W4Pt/SiO2 still remains a few of CO adsorbed on strong Brönsted acids, while signals on the
other three samples almost disappear (Fig. 6d). Combined with NH3-TPD results, this confirms
the introduction of WOx species to the catalysts mainly contributes to the formation Brönsted acids,
and the Brönsted acidity on 2W4Pt/SiO2 is stronger than that on other samples.
3.4. DFT study on the nature of Brönsted acid on WOx/Pt model
Based on the experimental evidence, HWxOy cluster located on Pt(111) surface was built to
simulate the structure of Brönsted acid site on xW4Pt/SiO2 catalysts. The HWxOy cluster is
considered as tetrahedral WO4 units connected through W-O-W bond, and the Brönsted acid is a
bridging –OH bonded with one W atom and one Pt atom simultaneously. Similar models were
used to simulate WOx supported on ZrO2 or Pt surface [31,32]. The discussion of various proton
positions and models with edge-share connection between WO4 units is elucidated in Table S4 and
Fig. S5, respectively. NH3 adsorption energy was calculated to estimate the strength of Brönsted
acid, and the effect of WO4 unit amount on the acid strength is investigated. Fig. 7 gives the
+
optimized structure and the NH3 adsorption results. In all models, NH4 was formed after the
adsorption on Brönsted acid site. For isolated HWO4 cluster, the adsorption energy is -93.5 kJ/mol.
With the cluster grows up to a dimeric HW2O7, NH3 adsorption energy is decreased to -117.4
kJ/mol, indicating the Brönsted acid is stronger than that of HWO4 cluster. For trimeric HW3O10,

two structures are investigated and the NH3 adsorption energy is -113.2 and -111.4 kJ/mol,
respectively, both of which are larger than the adsorption energy on HW2O7. Although the
difference is not such significant, it still reflects that HWxOy cluster with higher polymerization
does not benefit the Brönsted acid strength.
Fig. 7. The optimized structure and NH3 adsorption results of HWxOy/Pt(111) models. (a)
HWO4/Pt(111), (b) HW2O7/Pt(111), (c) linear HW3O10/Pt(111) and (d) angular HW3O10/Pt(111).
NH3 adsorption energies are given in kJ/mol.
To study the effect of reduction degree of HWxOy cluster on the Brönsted acid strength, one
oxygen atom is moved from HW2O7 cluster. As shown in Fig. S6, the corresponding NH3
adsorption energy is increased from -117.4 kJ/mol to -80.0 kJ/mol, and the H atom is still
connected with surface oxygen, indicating that the partial reduction of HWxOy cluster can
significantly weaken the Brönsted acid strength.

Fig. 8. H atom based electron density isosurface slice across surface Pt layer of HWxOy/Pt(111)
model: (a) HWO4/Pt(111) and (b) HW2O7/Pt(111). In the color map, the red color represents the
negative charge and the blue color means the positive charge.
The above theoretic studies of Brönsted acid strength on HWxOy/Pt(111) models are consistent
with the acid characterizations (Fig. 6) and catalytic performance results (Fig. 1 and Table 1). To
further elucidate the different Brönsted acid strength of HWxOy clusters from electronic states, the
electron density difference for HWO4 and HW2O7 clusters on Pt(111) were calculated, in which
the H atom was chosen to be one fragment and the rest of model (WO4/Pt or W2O7/Pt) to be the
other fragment. The top views of density isosurface slice across the surface Pt layer are shown in
Fig. 8. The existence of Brönsted acid on HWO4/Pt leads to the formation of a negative Pt atom
under it, while the electron densities of the other two Pt atoms bonded with HWO4 cluster are
slightly decreased. However on HW2O7/Pt, the Pt atom under Brönsted acid becomes more
negative compared with that on HWO4/Pt, and the electron densities of some Pt atoms on the
surface are also increased, even for those who are not connected with HW2O7 cluster. This
indicates that Pt plays a role of electron acceptor on the formation of Brönsted acid, and WO4 unit
is a kind of electronic conductor. For HWO4/Pt, the electron density is distributed mainly at the

narrow region of the Pt-O-H structure. Compared with that, the two adjacent WO4 units of dimeric
HW2O7 cluster can delocalize more negative charge to the wider Pt surface, leading to a much
stronger Brönsted acid.
XPS of Pt4f over 4Pt/SiO2 and 2W4Pt/SiO2 are shown in Fig. S7. For 4Pt/SiO2, the binding
energy of Pt 4f7/2 is 70.7 eV, indicating metallic Pt on the catalyst. Meanwhile, the binding energy
of Pt 4f7/2 on 2W4Pt/SiO2 is obviously decreased to 69.9 eV, which reflects an increased electron
density of Pt metal after loading of WOx. The binding energy change observed here is consistent
with the above electron density difference analysis of HWxOy/Pt(111) models.
It is not such clear why the Brönsted acid strength of trimer HW3O10/Pt is decreased. But from
the geometric optimization results of the two HW3O10/Pt(111) models in Fig. 7c and d, it can be
seen that instead of bonding to Pt surface, the middle WO4 unit is connected with the other two
WO4 units via W-O-W bonds, which can weaken the delocalization ability of negative charge to
Pt surface. Such hanging structures for trimeric or highly polymerized WO4 units were also
reported by Wang et al. [31]
4. Conclusion
The SMOI behavior was applied for the preparation of Pt-(WOx)n-H model catalysts, which
can decrease the interference of other active sites and benefit the study of Pt-(WOx)n-H structure
in glycerol hydrogenolysis. Pt-(WOx)n-H catalyst with medium polymeric WOx domains shows
the highest 1,3-PDO formation rate, but its activity can be significantly decreased after the WOx
domains partially reduced. Both high performance and high W efficiency toward 1,3-PDO are
obtained over 2W4Pt/SiO2 at 160 °C under 8.0 MPa. In addition, glycerol conversion occurs only

when the strong Brönsted acid site and active hydrogen are coexisted, indicating that 1,3-PDO is
formed via a direct hydrodeoxygenation mechanism.
Characterizations and DFT studies reveal that both the polymerization and reduction degree of
WOx influence the Brönsted acidity of Pt-(WOx)n-H. Medium polymeric WOx domains with no
reduction shows the strongest Brönsted acidity and the lowest apparent activation energy in the
formation of 1,3-PDO. Electron density difference analysis and XPS characterization confirm that
in the formation of Brönsted acid, Pt and WO4 unit play the role of electron acceptor and conductor,
respectively. Compared with isolated HWO4, dimeric HW2O7 cluster can delocalize more negative
charge to the wider Pt surface, leading to a much stronger Brönsted acidity. This work also
demonstrates the advantage of SMOI behavior in construction and investigation of bifunctional
active site than the conventional impregnation method.
Appendix A. Supplementary material
The supplementary data contain state of the art in glycerol hydrogenolysis to 1,3-PDO over Pt-
WOx catalysts, XRD profiles of the reduced xW4Pt/SiO2 catalysts, DRIFT spectra of CO adsorbed
on xW4Pt/SiO2 catalysts, XRD profile of the reduced 4Pt2W/SiO2 catalyst (a) and Raman spectra
of the calcined 4Pt2W/SiO2 catalyst (b), H2-TPR curves of xW4Pt/SiO2 catalysts, the consumed
H2 amount and quantification analysis from H2-TPR, peak fitting results of the strong acid region
in NH3-TPD curves, the relative energy of models with various proton positions, the geometric
optimization of models with edge-share connection, the optimized structure and NH3 adsorption
results of partially reduced HW2O6/Pt(111), and Pt 4f XPS results of the reduced xW4Pt/SiO2
catalysts.
AUTHOR INFORMATION

Corresponding Author
*E-mail: xbma@tju.edu.cn
Declaration of Competing Interest
The authors declare no competing financial interest.
Acknowledgements
This work was supported by the National Natural Science Foundation of China (21325626).
References
[1] A.G. Sergeev, J.F. Hartwig, Selective Nickel-Catalyzed Hydrogenolysis of Aryl Ethers,
Science 332 (2011) 439-443.
[2] A.G. Sergeev, J.D. Webb, J.F. Hartwig, A Heterogeneous Nickel Catalyst for the
Hydrogenolysis of Aryl Ethers without Arene Hydrogenation, J. Am. Chem. Soc. 134
(2012) 20226-20229.
[3] V. Molinari, C. Giordano, M. Antonietti, D. Esposito, Titanium Nitride-Nickel
Nanocomposite as Heterogeneous Catalyst for the Hydrogenolysis of Aryl Ethers, J. Am.
Chem. Soc. 136 (2014) 1758-1761.
[4] T.L. Lohr, Z. Li, R.S. Assary, L.A. Curtiss, T.J. Marks, Mono- and tri-ester hydrogenolysis
using tandem catalysis. Scope and mechanism, Energy Environ. Sci. 9 (2016) 550-564.
[5] T.L. Lohr, Z. Li, R.S. Assary, L.A. Curtiss, T.J. Marks, Thermodynamically Leveraged
Tandem Catalysis for Ester RC(O)O-R ' Bond Hydrogenolysis. Scope and Mechanism,
ACS Catal. 5 (2015) 3675-3679.
[6] Y.B. Huang, L. Yan, M.Y. Chen, Q.X. Guo, Y. Fu, Selective hydrogenolysis of phenols and
phenyl ethers to arenes through direct C-O cleavage over ruthenium-tungsten bifunctional
catalysts, Green Chem. 17 (2015) 3010-3017.
[7] Z.C. Luo, Z.X. Zheng, Y.C. Wang, G. Sun, H. Jiang, C. Zhao, Hydrothermally stable
Ru/HZSM-5-catalyzed selective hydrogenolysis of lignin-derived substituted phenols to
bio-arenes in water, Green Chem. 18 (2016) 5845-5858.
[8] W. Yin, R.H. Venderbosch, G. Bottari, K.K. Krawzcyk, K. Barta, H.J. Heeres, Catalytic
upgrading of sugar fractions from pyrolysis oils in supercritical mono-alcohols over Cu
doped porous metal oxide, Appl. Catal. B: Environ. 166 (2015) 56-65.
[9] A.M. Ruppert, K. Weinberg, R. Palkovits, Hydrogenolysis Goes Bio: From Carbohydrates
and Sugar Alcohols to Platform Chemicals, Angew. Chem. Int. Ed. 51 (2012) 2564-2601.
[10] K.Y. Chen, M. Tamura, Z.L. Yuan, Y. Nakagawa, K. Tomishige, One-Pot Conversion of
Sugar and Sugar Polyols to n-Alkanes without C-C Dissociation over the Ir-ReOx/SiO2
Catalyst Combined with H-ZSM-5, ChemSusChem 6 (2013) 613-621.

[11] D. Sun, Y. Yamada, S. Sato, W. Ueda, Glycerol hydrogenolysis into useful C3 chemicals,
Appl. Catal. B: Environ. 193 (2016) 75-92.
[12] Y. Nakagawa, M. Tamura, K. Tomishige, Perspective on catalyst development for glycerol
reduction to C3 chemicals with molecular hydrogen, Res. Chem. Intermed. 44 (2018) 3879-
3903.
[13] L.Z. Qin, M.J. Song, C.L. Chen, Aqueous-phase deoxygenation of glycerol to 1,3-
propanediol over Pt/WO3/ZrO2 catalysts in a fixed-bed reactor, Green Chem. 12 (2010)
1466-1472.
[14] S.H. Zhu, X.Q. Gao, Y.L. Zhu, Y.W. Li, Promoting effect of WOx on selective
hydrogenolysis of glycerol to 1,3-propanediol over bifunctional Pt-WOx/Al2O3 catalysts, J.
Mol. Catal. A: Chem. 398 (2015) 391-398.
[15] S. Garcia-Fernandez, I. Gandarias, J. Requies, M.B. Guemez, S. Bennici, A. Auroux, P.L.
Arias, New approaches to the Pt/WOx/Al2O3 catalytic system behavior for the selective
glycerol hydrogenolysis to 1,3-propanediol, J. Catal. 323 (2015) 65-75.
[16] J. Wang, X.C. Zhao, N. Lei, L. Li, L.L. Zhang, S.T. Xu, S. Miao, X.L. Pan, A.Q. Wang, T.
Zhang, Hydrogenolysis of Glycerol to 1,3-propanediol under Low Hydrogen Pressure over
WOx-Supported Single/Pseudo-Single Atom Pt Catalyst, ChemSusChem 9 (2016) 784-790.
[17] S. Garcia-Fernandez, I. Gandarias, J. Requies, F. Soulimani, P.L. Arias, B.M. Weckhuysen,
The role of tungsten oxide in the selective hydrogenolysis of glycerol to 1,3-propanediol
over Pt/WOx/Al2O3, Appl. Catal. B: Environ. 204 (2017) 260-272.
[18] X.C. Zhao, J. Wang, M. Yang, N. Lei, L. Li, B.L. Hou, S. Miao, X.L. Pan, A.Q. Wang, T.
Zhang, Selective Hydrogenolysis of Glycerol to 1,3-Propanediol: Manipulating the
Frustrated Lewis Pairs by Introducing Gold to Pt/WOx, ChemSusChem 10 (2017) 819-824.
[19] Y.Q. Fan, S.J. Cheng, H. Wang, D.H. Ye, S.H. Xie, Y. Pei, H.R. Hu, W.M. Hua, Z.H. Li,
M.H. Qiao, B.N. Zong, Nanoparticulate Pt on mesoporous SBA-15 doped with extremely
low amount of Was a highly selective catalyst for glycerol hydrogenolysis to 1,3-
propanediol, Green Chem. 19 (2017) 2174-2183.
[20] Y.Q. Fan, S.J. Cheng, H. Wang, J. Tian, S.H. Xie, Y. Pei, M.H. Qiao, B.N. Zong, Pt-WOx
on monoclinic or tetrahedral ZrO2: Crystal phase effect of zirconia on glycerol
hydrogenolysis to 1,3-propanediol, Appl. Catal. B: Environ. 217 (2017) 331-341.
[21] W. Zhou, J. Luo, Y. Wang, J. Liu, Y. Zhao, S. Wang, X. Ma, WOx domain size, acid
properties and mechanistic aspects of glycerol hydrogenolysis over Pt/WOx/ZrO2, Appl.
Catal. B: Environ. 242 (2019) 410-421.
[22] W. Zhou, Y.J. Zhao, S.P. Wang, X.B. Ma, The effect of metal properties on the reaction
routes of glycerol hydrogenolysis over platinum and ruthenium catalysts, Catal. Today 298
(2017) 2-8.
[23] L.F. Gong, Y.A. Lu, Y.J. Ding, R.H. Lin, J.W. Li, W.D. Dong, T. Wang, W.M. Chen,
Selective hydrogenolysis of glycerol to 1,3-propanediol over a Pt/WO3/TiO2/SiO2 catalyst
in aqueous media, Appl. Catal. A: Gen. 390 (2010) 119-126.
[24] S.S. Priya, V.P. Kumar, M.L. Kantam, S.K. Bhargava, A. Srikanth, K.V.R. Chary, High
Efficiency Conversion of Glycerol to 1,3-Propanediol Using a Novel Platinum-Tungsten
Catalyst Supported on SBA-15, Ind. Eng. Chem. Res. 54 (2015) 9104-9115.
[25] S.H. Zhu, X.Q. Gao, Y.L. Zhu, J.L. Cui, H.Y. Zheng, Y.W. Li, SiO2 promoted
Pt/WOx/ZrO2 catalysts for the selective hydrogenolysis of glycerol to 1,3-propanediol,
Appl. Catal. B: Environ. 158 (2014) 391-399.

[26] K. Tomishige, Y. Nakagawa, M. Tamura, Selective hydrogenolysis and hydrogenation
using metal catalysts directly modified with metal oxide species, Green Chem. 19 (2017)
2876-2924.
[27] Y. Nakagawa, M. Tamura, K. Tomishige, Catalytic materials for the hydrogenolysis of
glycerol to 1,3-propanediol, J. Mater. Chem. A 2 (2014) 6688-6702.
[28] S.-P. Sheu, H.G. Karge, R. Schlögl, Characterization of Activated States of Ruthenium-
Containing Zeolite NaHY, J. Catal. 168 (1997) 278-291.
[29] S.J. Clark, M.D. Segall, C.J. Pickard, P.J. Hasnip, M.J. Probert, K. Refson, M.C. Payne,
First principles methods using CASTEP, Zeitschrift Fur Kristallographie 220 (2005) 567-
570.
[30] J.P. Perdew, K. Burke, M. Ernzerhof, Generalized Gradient Approximation Made Simple,
Phys. Rev. Lett. 77 (1996) 3865-3868.
[31] C. Wang, A.V. Mironenko, A. Raizada, T.Q. Chen, X.Y. Mao, A. Padmanabhan, D.G.
Vlachos, R.J. Gorte, J.M. Vohs, Mechanistic Study of the Direct Hydrodeoxygenation of m-
Cresol over WOx-Decorated Pt/C Catalysts, ACS Catal. 8 (2018) 7749-7759.
[32] W. Zhou, N. Soultanidis, H. Xu, M.S. Wong, M. Neurock, C.J. Kiely, I.E. Wachs, Nature
of Catalytically Active Sites in the Supported WO3/ZrO2 Solid Acid System: A Current
Perspective, ACS Catal. 7 (2017) 2181-2198.
[33] W. Zhou, Y.J. Zhao, Y. Wang, S.P. Wang, X.B. Ma, Glycerol Hydrogenolysis to 1,3-
Propanediol on Tungstate/Zirconia-Supported Platinum: Hydrogen Spillover Facilitated by
Pt(111) Formation, ChemCatChem 8 (2016) 3663-3671.
[34] P. Waleska, S. Rupp, C. Hess, Operando Multiwavelength and Time-Resolved Raman
Spectroscopy: Structural Dynamics of a Supported Vanadia Catalyst at Work, J. Phys.
Chem. C 122 (2018) 3386-3400.
[35] E.L. Lee, I.E. Wachs, In situ spectroscopic investigation of the molecular and electronic
Structures of SiO2 supported surface metal oxides, J. Phys. Chem. C 111 (2007) 14410-
14425.
[36] Z.L. Wu, S. Dai, S.H. Overbury, Multiwavelength Raman Spectroscopic Study of Silica-
Supported Vanadium Oxide Catalysts, J. Phys. Chem. C 114 (2010) 412-422.
[37] M. Scheithauer, R.K. Grasselli, H. Knozinger, Genesis and structure of WOx/ZrO2 solid
acid catalysts, Langmuir 14 (1998) 3019-3029.
[38] A. Martinez, G. Prieto, M.A. Arribas, P. Concepcion, J.F. Sanchez-Royo, Influence of the
preparative route on the properties of WOx-ZrO2 catalysts: A detailed structural,
spectroscopic, and catalytic study, J. Catal. 248 (2007) 288-302.
[39] V. Lebarbier, G. Clet, M. Houalla, Relations between structure, acidity, and activity of
WOx/TiO2: Influence of the initial state of the support, titanium oxyhydroxide, or titanium
oxide, J. Phys. Chem. B 110 (2006) 22608-22617.
[40] G. Ferraris, S. De Rossi, D. Gazzoli, I. Pettiti, M. Valigi, G. Magnacca, C. Morterra,
WOx/ZrO2 catalysts Part 3. Surface coverage as investigated by low temperature CO
adsorption: FT-IR and volumetric studies, Appl. Catal. A: Gen. 240 (2003) 119-128.
[41] M. Kantcheva, C. Koz, Spectroscopic characterization of tungstated zirconia prepared by
equilibrium adsorption from hydrogen peroxide solutions of tungsten(VI) precursors, J.
Mater. Sci. 42 (2007) 6074-6086.
[42] X.Y. Chen, G. Clet, K. Thomas, M. Houalla, Correlation between structure, acidity and
catalytic performance of WOx/Al2O3 catalysts, J. Catal. 273 (2010) 236-244.