Effect of perimeter interface length between 2D WO3 monolayer domain and γ-Al2O3 on selective hydrogenolysis of glycerol to 1,3-propanediol

Article abstract of DOI:10.1039/c9cy01385g

The relationship between the structure of W species on Pt/WO₃/Al₂O₃ catalysts and their activity for selective hydrogenolysis of glycerol to 1,3-propanediol was investigated. Structural analysis by spectroscopic techniques including X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS) and X-ray adsorption fine structure (XAFS) revealed the formation of two-dimensional WO₃ monolayer domains on the surface of γ-Al₂O₃ at a WO₃ loading level of below 20 wt%. Evaluation of the reduction properties of W species by H₂ temperature programed reduction (TPR) suggested the presence of two kinds of W species with different reduction properties loaded on γ-Al₂O₃, and W species at the edge of a WO₃ domain was reduced more easily than that inside of a WO₃ domain. Furthermore, the length of the perimeter interface between a two-dimensional WO₃ monolayer domain and γ-Al₂O₃ (W-Al perimeter interface) could be estimated from the difference in their reducibility. The positive correlation between W-Al perimeter interface length and the yield of 1,3-propanediol in hydrogenolysis of glycerol indicated that a W-(OH)-Al site at the W-Al perimeter interface functioned as a main active site.

Full text of DOI:10.1039/c9cy01385g

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Effect of perimeter interface length between 2D
WO3 monolayer domain and γ-Al2O3 on selective
hydrogenolysis of glycerol to 1,3-propanediol†
Cite this: DOI: 10.1039/c9cy01385g
abd
abcd
Takeshi Aihara,a Hiroki Miura
and Tetsuya Shishido
*
The relationship between the structure of W species on Pt/WO3/Al2O3 catalysts and their activity for selec-
tive hydrogenolysis of glycerol to 1,3-propanediol was investigated. Structural analysis by spectroscopic
techniques including X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS) and X-ray adsorption
fine structure (XAFS) revealed the formation of two-dimensional WO3 monolayer domains on the surface
of γ-Al2O3 at a WO3 loading level of below 20 wt%. Evaluation of the reduction properties of W species by
H2 temperature programed reduction (TPR) suggested the presence of two kinds of W species with differ-
ent reduction properties loaded on γ-Al2O3, and W species at the edge of a WO3 domain was reduced
more easily than that inside of a WO3 domain. Furthermore, the length of the perimeter interface between
a two-dimensional WO3 monolayer domain and γ-Al2O3 (W–Al perimeter interface) could be estimated
from the difference in their reducibility. The positive correlation between W–Al perimeter interface length
and the yield of 1,3-propanediol in hydrogenolysis of glycerol indicated that a W–(OH)–Al site at the W–Al
perimeter interface functioned as a main active site.
Received 14th July 2019,
Accepted 4th September 2019
DOI: 10.1039/c9cy01385g
rsc.li/catalysis
strongly affected by the structure of the catalyst.19–25 Previ-
ously, we found that supported Group V–VI metal oxides (Nb,
Introduction
It is important to elucidate the detailed relationship between
the structures of catalysts and their catalytic performance to
establish guidelines for designing highly active catalysts.1–4
Supported group V–VII metal oxides are widely used as cata-
lysts in oil refinery, biomass conversion and selective catalytic
reduction (SCR) of NOx.5–10 The boundaries between different
metal oxides often exhibit unique catalytic properties that are
not observed in the simple metal oxide. Iglesia and coworkers
investigated the relationship between the structures of WOx
species on ZrO2 and their catalytic activities for o-xylene
isomerization,11–13 in which the domain size of WOx domi-
nated the acid properties of the catalyst.14–16 Wachs and co-
workers proposed the importance of a Brønsted acid site gen-
erated at the boundary between a supported metal oxide and
the support,17,18 and insisted that their acidic properties were
Ta, Mo, and W) on γ-Al2O3 exhibited the greatest Brønsted
acidity when the γ-Al2O3 surface was almost fully covered with
these metal oxides as a two-dimensional (2D) monolayer.26–34
On the other hand, hydrogenolysis is one of the most use-
ful reactions for producing value-added products from
biomass-derived compounds.35–38 Particularly, highly effi-
cient and selective hydrogenolysis of the C–O bond at the sec-
ond position of glycerol is of great importance, since this can
provide 1,3-propanediol, which is a useful monomer for poly-
trimethylene terephthalate (PTT) resin.39–43 The efficient ca-
talysis by tungsten oxide with Pt for the hydrogenolysis of
glycerol to 1,3-propanediol has been reported.44–49 Recently,
we reported that the surface coverage of a 2D WO3 monolayer
on a metal–oxide support significantly affected the activity of
supported Pt/WO3 catalysts for the selective hydrogenolysis of
biomass compounds,50,51 and proposed that the perimeter
interface between WO3 and γ-Al2O3 played an important role
in the selective formation of 1,3-propandiol from glycerol.50
However, the detailed role of tungsten species in catalytic
hydrogenolysis is still under discussion due to the difficulty
of evaluating the structure of the perimeter interface between
different metal–oxides. Despite the fact that microscopic
analyses such as transmission electron microscopy
(TEM),52–54 scanning electron microscopy (SEM)55 and scan-
ning tunneling microscopy (STM)56,57 are often used to ob-
serve the interfacial length or size of metal or metal oxides,
a Department of Applied Chemistry for Environment, Graduate School of Urban
Environmental Sciences, Tokyo Metropolitan University, 1-1 Minami-Osawa,
Hachioji, Tokyo 192-0397, Japan. E-mail: shishido-tetsuya@tmu.ac.jp
b Research Center for Hydrogen Energy-based Society, Tokyo Metropolitan Univer-
sity, 1-1 Minami-Osawa, Hachioji, Tokyo 192-0397, Japan
c Research Center for Gold Chemistry, Tokyo Metropolitan University, 1-1 Minami-
Osawa, Hachioji, Tokyo 192-0397, Japan
d Elements Strategy Initiative for Catalysts & Batteries, Kyoto University, Katsura,
Nishikyo-ku, Kyoto 615-8520, Japan
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c9cy01385g
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
these techniques have hardly been used to observe the bound-
ary between a supported metal oxide and the support due to
slight differences in their contrasting density. Hence, the
development of novel techniques to estimate the length of the
perimeter interface between two kinds of metal oxides would
be useful for controlling the structure and catalytic activity.
Herein, we investigated the relationship between the length
of the perimeter interface between WO3 and γ-Al2O3 (W–Al
perimeter interface) and the catalytic activity for hydrogenolysis
of glycerol to 1,3-propanediol over Pt/WO3/Al2O3 catalysts. The
detailed characterization of the structure and reducibility of W
species on Pt/WO3/Al2O3 catalysts enabled us to evaluate the
length of the W–Al perimeter interface, which supports the
notion that the perimeter interface plays an important role in
the hydrogenolysis of glycerol to 1,3-propanediol.
degrees at 10 degrees min−1 and a resolution of 0.01 degrees.
X-ray photoelectron spectroscopy (XPS) analysis of the cata-
lysts was performed using a JEOL JPS-9010 MX instrument.
The spectra were measured using Mg Kα radiation. All spec-
tra were calibrated using C 1s (284.5 eV) as a reference. The
Brunauer–Emmett–Teller (BET) specific surface area was esti-
mated from N2 isotherms obtained using a BELSORP-mini II
(BEL Japan, Osaka, Japan) at 77 K. The analyzed samples
were evacuated at 573 K for 3 h prior to the measurement.
X-ray absorption spectroscopy (XAS) analysis of the catalysts
was performed at the BL01B1 beamline at SPring-8 (Hyogo,
Japan). The ring energy was 8 GeV, and the stored current
was 99.5 mA. W L3-edge (10.2 keV) and W L1-edge (12.1 keV)
X-ray absorption spectra were recorded by using Si(311) and
Si(111) crystal monochromators, respectively. All spectra were
obtained in transmission mode. Data reduction was
performed by using xTunes (Science & Technology Inst.,
Co.).59 The pre-edge peak areas of W L1-edge XANES spectra
were estimated by the deconvolution of spectra with one or
two Lorenz functions and one arctangent function. Tempera-
ture programmed reduction (H2-TPR) was also carried out in
an Okura BP-2 instrument equipped with TCD for determina-
tion of H2 consumption and molecular sieve 4 Å traps to
remove the water formed during reduction. H2-TPR profiles
of each catalyst (30 mg) were recorded from 323 K to 1273 K
at 10 K min−1 under 5% H2/Ar flowing at 30 mL min−1.
Experimental
Materials
H2PtCl6 and (NH4)10W12O42·5H2O were purchased from
FURUYA METAL Co. Ltd and Wako Chemicals, respectively.
γ-Al2O3 (Sumitomo Chemical Co., Ltd, AKP-G015; JRC-ALO-
8 equivalent) was obtained from the Catalysis Society of
Japan. PtCl2 was purchased from Aldrich. PtCl4 and PtO2
were purchased from Wako Chemicals. For the reference
sample for XAFS analysis, Na2WO4 and WO3 were purchased
from Wako Chemicals. Ba2NiWO6 was synthesized by the
solid-state reaction from BaCO3, NiCO3 (Wako Chemicals)
and WO3 as described in a previous report.58
Results and discussion
Glycerol hydrogenolysis was carried out over Pt/WO3/Al2O3
catalysts with various WO3 loadings (Scheme 1 and Fig. S1†).
Although Pt/Al2O3 catalyst was not effective for the reaction,
the loading of WO3 enabled cleavage of the C–O bond at the
second position of glycerol selectively to give 1,3-propanediol
(Fig. 1). The loading amount of WO3 on γ-Al2O3 remarkably
affected the yield of the product, and the yield increased with
increasing WO3 loading; Pt/WO3/Al2O3 catalyst with 6 wt%
WO3 loading showed the highest yield of the product. On the
other hand, excess loading of WO3 drastically decreased the
catalytic activity of Pt/WO3/Al2O3 catalysts, and catalysts with
over 20% WO3 loading showed no activity for the present
reaction. These results suggest that the structure of WO3 on
γ-Al2O3 should be a dominant factor for determining the cata-
lytic performance in the selective hydrogenolysis of glycerol
to 1,3-propanediol. The surface acidity of Pt/WO3/Al2O3 cata-
lysts before and after reaction investigated by the NH3-TPD
and pyridine adsorbed IR (Fig. S2†). No change in the acidity
on Pt/WO3/Al2O3 catalysts before and after the reaction was
observed, indicating that the acidity of catalysts was not
directly collated to the activity of hydrogenolysis of glycerol to
1,3-propanediol. Although the location of Pt was investigated
Preparation of Pt/WO3/Al2O3 catalysts
A series of WO3/Al2O3 catalysts was prepared by impregnation
of γ-Al2O3 with an aqueous solution of (NH4)10W12O42·5H2O at
353 K for 2 h, dried at 353 K for 6 h, and then calcined at
1123 K for 3 h in flowing air.34 Pt/WO3/Al2O3 catalysts (Pt
loading: 1 wt%) were also prepared by impregnating WO3/
Al2O3 with an aqueous solution of H2PtCl6 for 2 h and dried
at 353 K for 6 h. After the impregnation procedure, the pre-
pared catalysts were calcined at 573 K for 3 h in air.
Procedure for hydrogenolysis
The hydrogenolysis of glycerol was carried out in a Teflon ves-
sel placed in a 50 mL stainless steel autoclave. Three mmol of
glycerol and 100 mg of a catalyst together with 9 mL of H2O
were added to the reactor. After the reactors were sealed, their
air was purged by hydrogen and the reactors were pressurized
to 5 MPa. The reaction was carried out at 453 K for 15 h.
Physical and analytical measurement
The products of the catalytic runs were analyzed by gas chro-
matography (Shimadzu GC-2014 equipped with a flame ioni-
zation detector, Stabilwax capillary column, i.d. 0.53 mm,
length 30 m, 80–250 °C). X-ray diffraction (XRD) patterns of
the catalyst were recorded by Rigaku SmartLab with Cu
Kαradiation. The samples were scanned from 2θ = 10−70
Scheme 1 Hydrogenolysis of glycerol to 1,3-propanediol.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
corresponding to Pt were confirmed for any of the catalysts.
These results indicate that amorphous WO3 is loaded on the
surface of γ-Al2O3 of the catalysts with below 20 wt% WO3
loading, and Pt species is highly dispersed on the WO3/Al2O3
surface. No diffraction peaks corresponding to Pt were con-
firmed for any of the catalysts. Moreover, no clear WO3 parti-
cles were observed in a TEM image of WO3/Al2O3 catalyst
with 7 wt% WO3 loading (Fig. S4†). The location of Pt was
also investigated by TEM technique. However, it was quite
difficult to distinguish whether Pt nanoparticles were distrib-
uted on either the WO3 monolayer or γ-Al2O3, because their
atomic numbers of Pt and W are close each other.
Fig. 1 Effect of WO3 loading on glycerol hydrogenolysis over Pt/WO3/
XP spectra of Pt/WO3/Al2O3 catalysts with various WO3
loadings are shown in Fig. S5.† The peaks due to W 4f7/2 and
W 4f5/2 appeared at 36.0 and 38.1 eV, respectively. These peak
positions were constant regardless of the Pt and WO3 loading
of the catalysts, indicating that W6+ species60,61 is present on
all the catalysts. The surface W/Al atomic ratio on the cata-
lysts was also estimated on the basis of the areas of the W 4f
and Al 2p peaks in the XP spectra (Fig. 3). The surface W/Al
ratio of Pt/WO3/Al2O3 catalysts increased linearly with WO3
loading up to 20 wt%, and then increased gradually. The
same tendency was observed in the case of WO3/Al2O3. These
results indicate that the state of the tungsten species loaded
on the surface of the catalysts changes at 20 wt% WO3 load-
ing regardless of the presence of Pt species. The density of
surface tungsten species on the catalyst with 20 wt% WO3
loading was estimated to be 4.5 W atoms nm−2. Previous liter-
ature reported that the tungsten density is around 4.0 atoms
nm−2 when the WO3 monolayer fully covers the surface of
Al2O3.24,62 This value is comparable to our estimated value,
indicating that WO3 is loaded as a 2D monolayer on γ-Al2O3
at a WO3 loading level of less than 20 wt%.
Al2O3 catalysts. Conditions: glycerol (3 mmol), catalyst (100 mg), H2O
(9 mL), PH = 5 MPa, T = 453 K, t = 15 h.
2
by TEM technique, the identification between of Pt and W
species was hard due to closeness their atomic number. The
location of Pt is currently underway in our laboratory.
Dispersity of Pt species on support with various WO3 load-
ings was estimated by CO chemisorption. Loading amount of
WO3 did not affect the Pt dispersion, and no clear correlation
between Pt dispersion and catalytic activity was observed
(Fig. S3†). These results indicate that the location of Pt is not
dominant factor for the hydrogenolysis over Pt/WO3/Al2O3 cat-
alysts. Hence, in the following section, we focus on elucida-
tion of the relationship between catalytic performance and
the structure of W species on the basis of spectroscopic and
kinetic characterization of the catalysts.
XRD patterns of Pt/WO3/Al2O3 catalysts with various WO3
loadings are shown in Fig. 2. Diffraction peaks due to crys-
talline γ-Al2O3 were observed in all catalysts. In contrast,
peaks attributed to crystalline WO3 (monoclinic WO3 (m-
WO3) and Al2IJWO3)4) were found in the catalyst with greater
than 20 wt% WO3 loading. Moreover, no diffraction peaks
Fig. 4 shows the W L3-edge k3-weighted extended X-ray ab-
sorption fine structure (EXAFS) oscillations of Pt/WO3/Al2O3
catalysts with various WO3 loadings as well as reference sam-
ples. No significant differences were noted in the shapes of
the EXAFS oscillations of catalysts with WO3 loading below
20 wt%. In contrast, the shape of the EXAFS oscillations of
Pt/WO3/Al2O3 catalysts with 30 wt% loading in the range of 3–
5 Å−1 clearly resembles that of m-WO3. This result indicates
Fig. 2 XRD patterns of Pt/WO3/Al2O3 catalysts with various WO3
loadings. (a) 0, (b) 2, (c) 6, (d) 10, (e) 15, (f) 20 and (g) 30 wt%. γ:
γ-Al2O3, m: m-WO3, A: Al2IJWO4)3.
Fig. 3 Surface W/Al ratio of Pt/WO3/Al2O3
catalysts with various WO3 loadings.
( ) and WO3/Al2O3 ( )
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
domains were loaded on the catalyst surface at up to 20 wt%
WO3 loading.
To obtain more detailed information about the local struc-
ture of tungsten species, W L1-edge XANES spectra of Pt/WO3/
Al2O3 catalysts with various WO3 loadings and reference sam-
ples were examined (Fig. 6). The pre-edge peak of W L1-edge
XANES is attributed to the electron transition from 2 s to d–p
orbitals63–65 and the local symmetry of tungsten species can
be estimated from the area of the pre-edge peak. W species
with low symmetry exhibit a larger pre-edge peak area than
that with high symmetry.66,67 The pre-edge peak area of Pt/
WO3/Al2O3 catalysts and reference samples was estimated by
deconvolution of the W L1-edge XANES spectra with one or
two Lorenz functions and an arctangent function (Fig. S6†).
The local W symmetry of Na2WO4, m-WO3 and Ba2NiWO6 is
tetrahedral (Td), distorted octahedral (Oh) and octahedral
(D2), respectively.66 The pre-edge peak area of reference sam-
ples decreased with an increase in the local symmetry of
tungsten species (Fig. 7). Pt/WO3/Al2O3 catalyst with 2 wt%
WO3 loading showed the largest pre-edge peak area among
the series of WO3-loaded catalysts. The pre-edge peak area of
Pt/WO3/Al2O3 catalysts decreased with increasing WO3 load-
ing and the area of 30 wt% WO3 loading catalyst was compa-
rable to that of m-WO3 with Oh symmetry. These results im-
plied that isolated Td-symmetric tungsten species was mainly
formed on the catalyst with 2 wt% WO3 loading and the ratio
of tungsten species with distorted Oh symmetry to that with
Td symmetry increased with an increase in WO3 loading. On
the other hand, W species in the Pt/WO3/Al2O3 catalyst with
30 wt% WO3 loading showed high Oh symmetry.
Fig. 4 W L3-edge EXAFS oscillations of Pt/WO3/Al2O3 catalysts with
various WO3 loadings and reference samples. (a) Na2WO4, (b) 2, (c) 6,
(d) 10, (e) 15, (f) 20, (g) 30 wt%, (h) m-WO3 and (i) Ba2NiWO6.
that crystalline WO3 is formed on the catalyst above 20 wt%
WO3 loading. W L3-edge Fourier-transformed EXAFS oscilla-
tions of Pt/WO3/Al2O3 catalysts with various WO3 loadings
and reference samples are shown in Fig. 5. In the spectra of
the catalysts with less than 20 wt% WO3 loading, only a peak
at 1–2 Å corresponding to W–O linkage was observed, which
indicates that two-dimensional WO3 monolayer domains
were formed on the catalyst surface with WO3 loading of less
than 20 wt%. In contrast, the peak at 3–4 Å, which is attribut-
able to W–O–W linkage, was observed only for Pt/WO3/Al2O3
catalyst with 30 wt% loading, implying the generation of crys-
talline WO3. These results suggest that 2D WO3 monolayer
To understand the electronic state of Pt loaded on WO3/
Al2O3 catalysts, XP spectra in Pt 4d5/2 and 4d3/2 were investi-
gated (Fig. 8). The binding energies of PtCl2 were 316.3 and
333.1 eV, respectively. The binding energies of PtCl4 and PtO2
were 317.8 and 334.7 eV, respectively. The binding energies
of Pt/WO3/Al2O3 catalysts were 316.0 and 332.9 eV, respec-
tively. No difference in binding energy was observed in the
spectra of Pt/WO3/Al2O3 catalysts with various WO3 loadings.
These results indicated that the valence of Pt loaded on the
WO3/Al2O3 catalysts was almost +2 regardless of the loading
amount of WO3.
The reduction properties of Pt/WO3/Al2O3 catalysts with
various WO3 loadings were evaluated by H2-TPR (Fig. 9).
While no reduction peak was confirmed for WO3/Al2O3 cata-
lysts, H2 consumption was observed in the profiles of all the
Pt-loaded catalysts. The fact of H2-TPR in which Pt promoted
the reduction of W species suggests that Pt nanoparticles
were loaded on or around the W species (Pt nanoparticles
may interact with W species). The investigation of the loca-
tion of Pt is currently underway in our laboratory. The H2
consumption peak at around 523 K was observed in the pro-
file of Pt/Al2O3 catalyst. Reduction of Pt2+ species to Pt metal
was reported to occur under 573 K.68,69 Therefore, the peak
observed at around 523 K in Pt/Al2O3 catalyst could be attrib-
uted to the reduction of Pt2+ to Pt0 species. Moreover, the H2
consumption peak at around 373–573 K was strongly affected
Fig. 5 Fourier-transformed W L3-edge EXAFS oscillations of Pt/WO3/
Al2O3 catalysts with various WO3 loadings and reference samples. (a)
Na2WO4, (b) 2, (c) 6, (d) 10, (e) 15, (f) 20, (g) 30 wt%, (h) m-WO3 and (i)
Ba2NiWO6.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
Fig. 6 (A) W L1-edge XANES spectra of Pt/WO3/Al2O3 catalysts with various WO3 loadings and reference samples. (B) Pre-edge region of W L1-
edge XANES. (a) Na2WO4, (b) 2, (c) 6, (d) 10, (e) 15, (f) 20, (g) 30 wt%, (h) m-WO3 and (i) Ba2NiWO6.
by the Pt loading. Iglesia and coworkers also reported that
the addition of Pt promoted the reduction of tungsten spe-
cies to consume H2 at around 373–573 K.15,70,71 A small WO3
domain was reported to be reduced at higher temperature
than a large WO3 domain or bulk WO3.15,72 The position of
the H2 consumption peak shifted to a lower temperature with
an increase in WO3 loading. This result indicated that the
size of the WO3 domain on the γ-Al2O3 surface increased with
an increase in WO3 loading. No remarkable peak shift was
observed at the region of W 4f7/2 and 4f5/2 in the XP spectra
of Pt/WO3/Al2O3 catalysts reduced at 453 and 573 K (Fig. S7†).
This is probably due to a small fraction of reduced W species
in WO3 domain.
of Pt2+ to Pt0. These results suggest that the reduction peak
in the TPR profile of Pt/Al2O3 catalyst should be due to the
reduction of Pt2+ to Pt0. In contrast, the amount of H2 con-
sumption of Pt/WO3/Al2O3 catalysts was greater than that of
Pt/Al2O3 catalyst. These results imply that the H2 consump-
tion peak below 673 K includes the reduction of tungsten
oxides in addition to the reduction of Pt species. The amount
of H2 consumption due to WO3 at each catalyst with A wt%
WO3 loading (XIJA)W⋯XIJA)WO ) can be calculated as (1)
3
X(A)WO = X(A)Total − XPt
(1)
3
XIJA)Total represents the total amount of H2 consumption of
the sample with A wt% WO3 loading and XPt indicates the
The amounts of H2 consumption estimated from the peak
area of each TPR profile of the catalysts are summarized in
Fig. 10(A). The estimated H2 consumption of Pt/Al2O3 catalyst
was 45 μmol g−1, which is almost the same as the theoretical
value of H2 consumption (51 μmol g−1) during the reduction
Fig.
8 XP spectra of Pt/WO3/Al2O3 catalysts with various WO3
Fig. 7 Pre-edge area of W L1-edge XANES of Pt/WO3/Al2O3 catalysts
with various WO3 loadings.
loadings and reference samples in the Pt 4d region. (a) 0, (b) 2, (c) 6,
(d) 10, (e) 15, (f) 20, (g) 30 wt%, (h) PtCl2, (i) PtCl4 and (j) PtO2.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Fig. 11 H2/W ratio of Pt/WO3/Al2O3 catalysts with various WO3
loadings.
Fig. 9 H2-TPR profiles of Pt/WO3/Al2O3 catalysts with various WO3
loadings. (a) 6 wt% WO3/Al2O3 (b) 0, (c) 2, (d) 4, (e) 6, (f) 8, (g) 10, (h)
15, (i) 20 and (j) 30 wt%.
to benzaldehyde.73 In our case, tungsten species at the edge
of a domain should be exposed to different environmental
conditions than that inside of a WO3 domain. W species at
the edge of a domain was deduced to be reduced at lower
temperature than that at the inside of a WO3 domain.
Therefore, we can assume that H2 consumption in the
region of 373 K to 573 K is due to the reduction of W spe-
cies at the perimeter interface between a 2D WO3 monolayer
domain and γ-Al2O3.
amount of H2 consumption due to the reduction of Pt species
at Pt/Al2O3 (45 μmol g−1). Fig. 10(B) shows XIJA)WO ; the value
3
decreased with an increase in WO3 loading. This result sug-
gests that WO3 loading should affect the reduction property
of WO3. Furthermore, to obtain the degree of WO3 reduction,
the ratio of the H2 consumption amount due to supported
tungsten species (H2/W ratio) was also estimated as (2)
If the 2D WO3 monolayer domain possesses a square
shape that consists of “n” pieces of WO6 units on each side,
“n” can be estimated by using (3).
H2/W ratio = X(A)WO /N(A)
(2)
3
Number of WO6 units at perimeter ð4n − 4Þ
H2=W ratio ¼
(3)
NIJA) represents the amount of WO3 loading. The H2/W
ratios were less than 1 regardless of WO3 loading (Fig. 11).
These results indicate that tungsten species were partially
reduced at a temperature region of around 373 K to 573 K
except for the catalyst with 2 wt% WO3 loading. Li and co-
workers reported a difference in reducibility between the
edge and interior of FeO islands on Pt. FeO at the edge of
islands is reduced through the oxidation of benzyl alcohol
Number of total WO6 units ðn2Þ
The cross-sectional area of a WO6 unit that is loaded as a
monolayer on a γ-Al2O3 surface was reported to be 0.47
nm2.72,74 Moreover, the side length of the WO6 unit is esti-
mated to be 0.22 nm from the square root of the cross-
sectional area of the WO6 unit (Fig. 12). Furthermore, eqn (4)
Fig. 10 H2 consumption of Pt/WO3/Al2O3 catalysts with various WO3 loadings for (A) total amount and (B) the amount due to the tungsten
species.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
Fig. 12 Structural models of a WO3 monolayer domain on Pt/WO3/
Al2O3 catalysts.
gives the length of the perimeter interface between a 2D WO3
monolayer domain and γ-Al2O3:
Length of perimeter interface = Side length of a WO6 unit
(0.22 nm) × Number of
Fig. 14 Models of W species on Pt/WO3/Al2O3 catalysts.
WO6 units at the perimeter
(4n − 4)
(4)
An increase in WO3 loading decreased the length of the
perimeter interface between WO3 and γ-Al2O3 (Fig. 13). These
results suggest that small WO3 domains are highly dispersed
on a γ-Al2O3 surface in catalysts with a low loading level of
WO3. In contrast, an increase in WO3 loading causes an
increase in size and aggregation of each domain, which leads
to a decrease in the number of WO6 units at the domain
edge. Structural analyses of the catalysts based on X-ray
spectroscopy revealed that the γ-Al2O3 surface was almost
fully covered with a 2D WO3 monolayer domain in the cata-
lyst with 20 wt% WO3 loading. Therefore, no perimeter inter-
face between WO3 and γ-Al2O3 exists and the perimeter inter-
face length of Pt/WO3/Al2O3 catalysts is constant above 20
wt% WO3 loading (Fig. 14).
Fig. 15 Effect of the perimeter interface length between WO3 and
γ-Al2O3 on hydrogenolysis over Pt/WO3/Al2O3 catalysts.
Fig. 15 shows the relationship between the perimeter
interface length between a 2D WO3 monolayer domain and
γ-Al2O3 and the yield of 1,3-propanediol in the hydrogenolysis
of glycerol over Pt/WO3/Al2O3 catalysts. The positive correla-
tion between the two different parameters indicates that the
W–Al perimeter interface plays an important role in the selec-
tive hydrogenolysis of glycerol to 1,3-propanediol. A Brønsted
acid site is well known to be generated on the bridging bond
of a supported metal oxide and its support.17,18 The active
site of this reaction should be a W–(OH)–Al site at the perim-
eter interface between the WO3 domain and γ-Al2O3. Proton-
ation of the secondary OH of glycerol is promoted by a
W–(OH)–Al site, followed by dissociation of the C–O bond at
the second position of glycerol and the formation of 1,3-
propanediol. Although Pt/WO3/Al2O3 catalysts with low WO3
loading (2–3 wt%) have a large W–Al perimeter interface, the
catalytic activity is quite low. This result implies that isolated
Td tungsten species on γ-Al2O3 cannot generate the W–(OH)–
Al site which acts as an active site for the present reaction.
Conclusions
The relationship between the structure of W species on Pt/
WO3/Al2O3 catalysts and their activities for the selective
hydrogenolysis of glycerol to 1,3-propanediol was investi-
gated. Structural analysis of the catalysts based on X-ray
spectroscopy indicated that a 2D WO3 monolayer domain
was loaded on the surface of γ-Al2O3. H2-TPR profiles of Pt/
WO3/Al2O3 catalysts suggested the presence of two kinds of
tungsten species with different reducing properties. The
tungsten species at the edge of a WO3 domain was reduced
more easily than that inside of a WO3 domain. Furthermore,
Fig. 13 Effect of WO3 loading on the perimeter interface between
WO3 and γ-Al2O3.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
H2-TPR can estimate the length of the perimeter interface
between a 2D WO3 monolayer domain and γ-Al2O3. This
revealed a positive correlation between the length of the W–Al
perimeter interface and the catalytic activity of Pt/WO3/Al2O3
catalyst for the hydrogenolysis of glycerol to 1,3-propanediol.
This result indicated that the W–(OH)–Al site at the perimeter
interface between a WO3 domain and γ-Al2O3 plays an impor-
tant role for the hydrogenolysis of glycerol to form 1,3-
propanediol.
16 E. Iglesia, D. G. Barton, S. L. Soled, S. Miseo, J. E.
Baumgartner, W. E. Gates, G. A. Fuentes and G. D. Meitzner,
Stud. Surf. Sci. Catal., 1996, 101, 533–542.
17 I. E. Wachs, Colloids Surf., A, 1995, 105, 143–149.
18 M. A. Vuurman, D. J. Stufkens, A. Oskam and I. E. Wachs,
J. Mol. Catal., 1992, 76, 263–289.
19 W. Zhou, N. Soultanidis, H. Xu, M. S. Wong, M. Neurock,
C. J. Kiely and I. E. Wachs, ACS Catal., 2017, 7, 2181–2198.
20 W. Zhou, E. I. Ross-Medgaarden, W. V. Knowles, M. S. Wong
and I. E. Wachs, Nat. Chem., 2009, 1, 722–728.
21 W. Lin, A. A. Herzing, C. J. Kiely and I. E. Wachs, J. Phys.
Chem. C, 2008, 112, 5942–5951.
22 E. I. Ross-Medgaarden, W. V. Knowles, T. Kim, M. S. Wong, W.
Zhou, C. J. Kiely and I. E. Wachs, J. Catal., 2008, 256, 108–125.
23 I. E. Wachs, Y. Chen, J.-M. Jehng, L. E. Briand and T.
Tanaka, Catal. Today, 2003, 78, 13–24.
24 I. E. Wachs, Catal. Today, 1996, 27, 437–455.
25 A. M. Turek and I. E. Wachs, J. Phys. Chem., 1992, 96,
5000–5007.
26 T. Kitano, T. Shishido, K. Teramura and T. Tanaka, Catal.
Today, 2014, 226, 97–102.
27 T. Kitano, T. Shishido, K. Teramura and T. Tanaka,
ChemPhysChem, 2013, 14, 2560–2569.
28 T. Kitano, T. Shishido, K. Teramura and T. Tanaka, J. Phys.
Chem. C, 2012, 116, 11615–11625.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
This study was partially supported by the Program for Ele-
ments Strategy Initiative for Catalysts & Batteries (ESICB).
This work was also supported in part by Grants-in-Aid for Sci-
entific Research (B) (Grant 17H03459) and Scientific Research
on Innovative Areas (Grant 17H06443) commissioned by
MEXT of Japan. The XAFS experiments at SPring-8 were
conducted with the approval (No. 2018A1647) of the Japan
Synchrotron Radiation Research Institute (JASRI).
Notes and references
29 T. Shishido, T. Kitano, K. Teramura and T. Tanaka, Top.
Catal., 2010, 53, 672–677.
30 T. Shishido, T. Kitano, K. Teramura and T. Tanaka, Catal.
Lett., 2009, 129, 383–386.
1 H. Hattori, Appl. Catal., A, 2015, 504, 103–109.
2 H. Nair, M. J. Liszka, J. E. Gatt and C. D. Baertsch, J. Phys.
Chem. C, 2008, 112, 1612–1620.
3 S. L. Soled, G. B. McVicker, L. L. Murrell, L. G. Sherman,
N. C. Dispenziere, S. L. Hsu and D. Waldman, J. Catal.,
1988, 111, 286–295.
4 T. Yamamoto, Y. Tanaka, T. Matsuyama, T. Funabiki and S.
Yoshida, J. Phys. Chem. B, 2001, 105, 1908–1916.
5 M. Hino and K. Arata, J. Chem. Soc., Chem. Commun.,
1988, 1259–1260.
6 A. H. Karim, S. Triwahyono, A. A. Jalil and H. Hattori, Appl.
Catal., A, 2012, 433–434, 49–57.
7 R. W. Gosselink, D. R. Stellwagen and J. H. Bitter, Angew.
Chem., Int. Ed., 2013, 52, 5089–5092.
31 T. Kitano, S. Okazaki, T. Shishido, K. Teramura and T.
Tanaka, Catal. Today, 2012, 192, 189–196.
32 T. Kitano, S. Okazaki, T. Shishido, T. Teramura and T.
Tanaka, Catal. Lett., 2011, 40, 1332–1334.
33 T. Kitano, S. Okazaki, T. Shishido, T. Teramura and T.
Tanaka, J. Mol. Catal. A: Chem., 2013, 371, 21–28.
34 T. Kitano, T. Hayashi, T. Uesaka, T. Shishido, K. Teramura
and T. Tanaka, ChemCatChem, 2014, 6, 2011–2020.
35 R. A. Sheldon, Green Chem., 2014, 16, 950–963.
36 A. M. Ruppert, K. Weinberg and R. Palkovits, Angew. Chem.,
Int. Ed., 2012, 51, 2564–2601.
8 Y. Liu, C. Luo and H. Liu, Angew. Chem., Int. Ed., 2012, 51,
3249–3253.
37 M. Pagliaro and M. Rossi, The Future of Glycerol, RSC
Publishing, Cambridge, 2008.
9 R. Foo, T. Vazhnova, D. B. Lukyanov, P. Millington, J.
Collier, R. Rajaram and S. Golunski, Appl. Catal., B,
2015, 162, 174–179.
10 M. Kantcheva, M. Milanova and S. Mametsheripov, Catal.
Today, 2012, 191, 12–19.
11 S. Kwon, P. Deshlahra and E. Iglesia, J. Catal., 2018, 364,
228–247.
12 J. Macht, T. Carr and E. Iglesia, J. Am. Chem. Soc., 2009, 131,
6554–6565.
38 A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107,
2411–2502.
39 Y. Nakagawa, Y. Tamura and K. Tomishige, Res. Chem.
Intermed., 2018, 44, 3897–3903.
40 K. Tomishige, Y. Nakagawa and M. Tamura, Green Chem.,
2017, 19, 2876–2924.
41 Y. Nakagawa, M. Tamura and K. Tomishige, J. Mater. Chem.
A, 2014, 2, 6688–6702.
42 J. ten Dam and U. Hanefeld, ChemSusChem, 2011, 18,
1017–1034.
43 C. H. Zhou, J. N. Beltramini, Y. X. Fan and G. Q. Lu, Chem.
Soc. Rev., 2008, 37, 527–549.
13 K. Chen, A. T. Bell and E. Iglesia, J. Catal., 2002, 209, 35–42.
14 C. D. Baertsch, S. L. Soled and E. Iglesia, J. Phys. Chem. B,
2001, 105, 1320–1330.
15 D. G. Barton, S. L. Soled, G. D. Meitzner, G. A. Fuentes and
E. Iglesia, J. Catal., 1999, 181, 57–72.
44 J. ten Dam and U. Hanefeld, ChemCatChem, 2013(5),
497–505.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Catalysis Science & Technology
Paper
45 R. Arundhathi, T. Mizugaki, T. Mitsudome, K. Jitsukawa and
K. Kaneda, ChemSusChem, 2013, 6, 1345–1347.
46 T. Mizugaki, T. Yamakawa, R. Arundhathi, T. Mitsudome, K.
Jitsukawa and K. Kaneda, Chem. Lett., 2012, 41, 1720–1722.
47 T. Kurosaka, H. Maruyama, I. Naribayashi and Y. Sasaki,
Catal. Commun., 2008, 9, 1360–1363.
48 S. García-Fernández, I. Gandarias, J. Requies, F. Soulimani
and P. L. Arias, Appl. Catal., B, 2017, 204, 260–271.
49 S. García-Fernández, I. Gandarias, J. Requies, M. B. Güemez,
S. Bennici and P. L. Arias, J. Catal., 2015, 323, 65–75.
50 T. Aihara, H. Kobayashi, S. Feng, H. Miura and T. Shishido,
Chem. Lett., 2017, 46, 1497–1500.
51 S. Feng, A. Nagao, T. Aihara, H. Miura and T. Shishido,
Catal. Today, 2018, 303, 207–212.
52 J. Y. Park, L. R. Baker and G. A. Somorjai, Chem. Rev.,
2015, 115, 2781–2817.
53 M. Cargnello, V. V. T. Doan-Nguyen, T. R. Gordon, R. E.
Diaz, E. A. Stach, R. J. Gorte, P. Fornasiero and C. B. Murray,
Science, 2013, 341, 771–773.
54 H. Yoshida, Y. Kuwauchi, J. R. Jinschek, K. Sun, S. Tanaka,
M. Kohyama, S. Shimada, M. Haruta and S. Takeda, Science,
2012, 335, 317–319.
55 E. I. Papaioannou, C. Bachmann, J. J. Neumeier, D. Frankel,
H. Over, J. Janek and I. S. Metcalfe, ACS Catal., 2016, 6,
5865–5872.
56 H. Ariga, T. Taniike, H. Morikawa, R. Tero, H. Kondoh and
Y. Iwasawa, Chem. Phys. Lett., 2008, 454, 350–354.
57 T. F. Jaramillo, K. P. Jørgensen, J. Bonde, J. H. Nielsen, S.
Horch and I. Chorkendorff, Science, 2007, 317, 100–102.
58 T. Varga, A. P. Wilkinson, A. C. Jupe, C. Lind, W. S. Bassett
and C.-S. Zha, Phys. Rev. B: Condens. Matter Mater. Phys.,
2005, 72, 024117.
59 H. Asakura, S. Yamazoe, T. Misumi, A. Fujita, T. Tsukuda
and T. Tanaka, Radiat. Phys. Chem., 2019, DOI: 10.1016/j.
radphyschem.2019.04.020.
60 V. I. Nefedov, Y. V. Salyn, G. Leonhardt and R. A. Scheibe,
J. Electron Spectrosc. Relat. Phenom., 1977, 10, 121–124.
61 M. C. Peignon, C. Cardinaud and G. A. Turban,
J. Electrochem. Soc., 1993, 140, 505–512.
62 L. Salvati, L. E. Makovsky, J. M. Stencel, F. R. Brown and
D. H. J. Hercules, J. Phys. Chem., 1981, 85, 3700–3707.
63 S. Yoshida, T. Tanaka, T. Hanada, T. Hiraiwa, H. Kanai and
T. Funabiki, Catal. Lett., 1992, 12, 277–286.
64 H. Yoshida, T. Tanaka, T. Yoshida, T. Funabiki and S.
Yoshida, Catal. Today, 1996, 28, 79–89.
65 T. Yamamoto, A. Orita and T. Tanaka, X-Ray Spectrom.,
2008, 37, 226–231.
66 S. Yamazoe, Y. Hitomi, T. Shishido and T. Tanaka, J. Phys.
Chem. C, 2008, 112, 6869–6879.
67 J. A. Horsley, I. E. Wachs, J. M. Brown, G. H. Via and F. D.
Hardcastle, J. Phys. Chem., 1987, 91, 4014–4020.
68 K. Kinoshita, K. Routsus and J. A. S. Bett, Thermochem. Acta,
1974, 10, 109–117.
69 J. Barbier, D. Bahloul and P. Marecot, J. Catal., 1992, 137,
377–384.
70 R. Prins, Chem. Rev., 2012, 112, 2714–2738.
71 S. Khoobiar, J. Phys. Chem., 1964, 68, 411–412.
72 A. Martínez, G. Prieto, M. A. Arribas, P. Concepción and J. F.
Sánchez-Royo, J. Catal., 2007, 248, 288–302.
73 G. Zhao, F. Yang, Z. Chen, Q. Liu, Y. Ji, Y. Zhang, Z. Niu, J.
Mao, X. Bao, P. Hu and Y. Li, Nat. Commun., 2017, 8, 14039.
74 C. Pfaff, M. J. P. Zurita, C. Scott, P. Patiño, M. R.
Goldwasser, J. Goldwasser, F. M. Mulcahy, M. Houalla and
D. M. Hercules, Catal. Lett., 1997, 49, 13–16.
This journal is © The Royal Society of Chemistry 2019
Catal. Sci. Technol.