Effect of high-temperature calcination on the generation of Bronsted acid sites on WO3/Al2O3

Article abstract of DOI:10.1002/cctc.201400053

The acid properties of a series of alumina-supported tungsten oxide (WO₃/Al₂O₃) catalysts with loadings of 5-50 wt% WO₃ calcined at various temperatures were investigated by acid-catalyzed reactions (benzylation of anisole and isomerization of α-pinene) and FTIR spectroscopy. The relationships between acid properties, structures, and catalytic performances were evaluated. Both the catalytic activity and amount of Bronsted acid sites depend on the calcination temperature and WO₃ loading. High-temperature calcination (1123 K) generated Bronsted acid properties, and 20 wt% WO ₃/Al₂O₃ calcined at 1123 K exhibited the highest activity among the catalysts tested. The activities for the benzylation of anisole and α-pinene isomerization over WO₃/Al ₂O₃ calcined at 1123 K were proportional to the Bronsted acidity, which indicates that these reactions occurred on the Bronsted acid sites. Tungsten oxide, which has distorted octahedral symmetry, was loaded as 2D monolayer domains below 20 wt%, and these domains covered most of the alumina surface at 20 wt%. If the WO₃ loading was sufficient to form 2D tungsten oxide monolayer sheets (>20 wt%), some of the Bronsted acid sites on WO₃/Al₂O₃ were obscured by monoclinic WO₃ that has no Bronsted acid sites, which resulted in a decrease of the catalytic activity. This suggests that Bronsted acid sites are generated at the boundaries between tungsten oxide monolayer domains.

Full text of DOI:10.1002/cctc.201400053

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201400053
Effect of High-Temperature Calcination on the Generation
of Brønsted Acid Sites on WO3/Al2O3
Tomoyuki Kitano,[a] Tomohiro Hayashi,[a] Toshio Uesaka,[a] Tetsuya Shishido,*[a, b]
Kentaro Teramura,[a, b, c] and Tsunehiro Tanaka*[a, b]
The acid properties of a series of alumina-supported tungsten
oxide (WO3/Al2O3) catalysts with loadings of 5–50 wt% WO3
calcined at various temperatures were investigated by acid-cat-
alyzed reactions (benzylation of anisole and isomerization of a-
pinene) and FTIR spectroscopy. The relationships between acid
properties, structures, and catalytic performances were evaluat-
ed. Both the catalytic activity and amount of Brønsted acid
sites depend on the calcination temperature and WO3 loading.
High-temperature calcination (1123 K) generated Brønsted acid
properties, and 20 wt% WO3/Al2O3 calcined at 1123 K exhibited
the highest activity among the catalysts tested. The activities
for the benzylation of anisole and a-pinene isomerization over
WO3/Al2O3 calcined at 1123 K were proportional to the Brønst-
ed acidity, which indicates that these reactions occurred on
the Brønsted acid sites. Tungsten oxide, which has distorted
octahedral symmetry, was loaded as 2D monolayer domains
below 20 wt%, and these domains covered most of the alumi-
na surface at 20 wt%. If the WO3 loading was sufficient to form
2D tungsten oxide monolayer sheets (>20 wt%), some of the
Brønsted acid sites on WO3/Al2O3 were obscured by monoclinic
WO3 that has no Brønsted acid sites, which resulted in a de-
crease of the catalytic activity. This suggests that Brønsted acid
sites are generated at the boundaries between tungsten oxide
monolayer domains.
Introduction
Tungsten compounds are widely used in various industrial pro-
cesses such as oxidations, acid–base reactions, and photocata-
lytic reactions. Dodecatungstophosphoric acid is a heteropoly
acid that contains 12 W atoms and is an active homogeneous
acid catalyst for the hydration of olefins and esterification.[1–3]
Homogeneous acid catalysts are not environmentally friendly
because of their corrosive nature, difficulty of separation from
solution, and necessity of neutralization, so the use of hetero-
geneous acid catalysts in reactions are preferred. As a result,
the acid properties of water-insoluble heteropoly compounds,
such as heteropoly acids supported on silica (H3PW12O40/SiO2
and H4SiW12O40/SiO2) and salts with large monovalent ions
(Cs2.5H0.5PW12O40), have been studied.[4–10] H3PW12O40/SiO2,
H4SiW12O40/SiO2, and Cs2.5H0.5PW12O40 are effective for various
acid–base reactions such as esterification and hydration. Al-
though these heteropoly acids possess some advantages,
leaching into polar solvents such as water and alcohols occurs
in the case of supported heteropoly acids. Moreover, it is diffi-
cult to regenerate heteropoly acids by heat treatment at tem-
peratures that exceed 773 K because heteropoly acids decom-
pose above 773 K.[11,12] Tungsten oxide supported on zirconia,
titania, and iron oxide (WO3/ZrO2, WO3/TiO2, and WO3/Fe2O3)
are solid superacid catalysts. The acid strength of WO3/ZrO2,
WO3/TiO2, and WO3/Fe2O3 is estimated to be H0 <À14.6, À13.1,
and À12.5, respectively. These catalysts promote acid reactions
that require superacid sites, such as the skeletal isomerization
of butane.[13–18]
[a] T. Kitano, T. Hayashi, T. Uesaka, Dr. T. Shishido,+ Dr. K. Teramura,
Prof. Dr. T. Tanaka
Department of Molecular Engineering
Graduate School of Engineering, Kyoto University
Kyoto Daigaku Katsura, Nishikyo-ku, Kyoto 615-8510 (Japan)
Fax: (+81)75-383-2561
E-mail: tanakat@moleng.kyoto-u.ac.jp
shishido-tetsuya@tmu.ac.jp
Niobic acid (Nb2O5·nH2O, hydrated niobium oxide) is
a useful solid acid catalyst and exhibits water-tolerance and
strong Brønsted acid properties.[8,19–22] However, Nb2O5·nH2O
loses its acid properties if it is calcined at 773 K because of the
phase transformation from amorphous to crystalline phases
(TT- and T-phases).[22,23] In contrast, niobium oxide supported
on alumina (Nb2O5/Al2O3) calcined at a high temperature
(1123 K) exhibits a high activity for acid-catalyzed reactions,
such as the benzylation of anisole, cumene cracking, and the
isomerization of a-pinene, because of its Brønsted acid proper-
ties.[24–26] The calcination temperature and loading of Nb2O5
affect the acid properties and structure of supported Nb2O5
strongly, and 16 wt% Nb2O5/Al2O3 calcined at 1123 K showed
the highest activity. The Brønsted acidity of this catalyst was
[b] Dr. T. Shishido,+ Dr. K. Teramura, Prof. Dr. T. Tanaka
Elements Strategy Initiative for Catalysts and Batteries
Kyoto University
Kyoto Daigaku Katsura, Nishikyo-ku, Kyoto 615-8520 (Japan)
[c] Dr. K. Teramura
Precursory Research for Embryonic Science and Technology (PRESTO)
Japan Science and Technology Agency (JST)
4-1-8 Honcho, Kawaguchi, Saitama 332-0012 (Japan)
[+] Present address:
Department of Applied Chemistry
Graduate School of Urban Environmental Sciences
Tokyo Metropolitan University
1-1 Minami-Osawa, Hachioji, Tokyo 192-0397 (Japan)
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2011 – 2020 2011

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
not changed even after treatment at 1173 K under nitrogen
flow, which indicates that Brønsted acid sites generated on the
Nb2O5/Al2O3 catalyst were thermally quite stable. Tantalum
oxide also loses its acid properties if the catalyst is heated at
1073 K,[22,27] but tantalum oxide supported on alumina (Ta2O5/
Al2O3) calcined at a high temperature (1223 K) exhibits catalytic
activity for the benzylation of anisole and isomerization of a-
pinene and has Brønsted acid properties.[28,29] These examples
demonstrate that the alumina support stabilizes the Brønsted
acid sites during calcination at high temperatures and that
both Nb2O5/Al2O3 and Ta2O5/Al2O3 calcined at high tempera-
tures work as thermally stable solid acid catalysts. High-tem-
perature calcination usually causes solid acid catalysts to lose
their acid properties. Conventional solid acid catalysts, such as
aluminosilicate, zeolite, and binary oxides, lose their acid prop-
erties if pretreated at >1000 K. If a solid acid could be pre-
pared by calcination at the high temperatures required to
form ceramics, it would be useful as an acid catalyst even
under severe conditions because of its high stability. We pro-
posed that supported niobium oxide and tantalum oxide form
2D oxides upon which Brønsted acid sites are generated. Such
2D oxides may be an important factor in the generation of
Brønsted acid sites.
The acid properties of tungsten oxide on alumina (WO3/
Al2O3) have also been investigated. Arata et al. reported that 5
and 10 wt% WO3/Al2O3 exhibited high activity for the cracking
of cumene if the catalyst was calcined at temperatures higher
than 1273 K.[30] Soled et al. suggested that high-temperature
calcination increased the fraction of Brønsted acidity to total
acidity for 10 wt% WO3/Al2O3 calcined at 1173 K.[31] However,
the relationship between the structure and acidic properties of
WO3/Al2O3 is still unclear and the structure of the acid sites has
not been clarified. In the present study, we characterized struc-
tural changes in a series of alumina-supported tungsten oxide
catalysts with various loadings of WO3 calcined at various tem-
peratures to clarify the relationship between their acidic prop-
erties and local structures around the W species. A model
structure of the acid sites in WO3/Al2O3 is proposed.
Figure 1. Activity of 20 wt% WO3/Al2O3 calcined at various temperatures for
a) the Friedel–Crafts alkylation (*: benzyl anisole and *: dibenzyl ether)
and b) the isomerization of a-pinene (&: camphene and &: limonene). Pre-
treatment temperature: 473 K.
cined at 1123 K. The ratio of p-benzyl anisole to o-benzyl ani-
sole was approximately 3:2 regardless of the calcination
temperature.
The isomerization of a-pinene was used as a test reaction to
investigate the acid–base properties of the catalyst. The activi-
ty and selectivity of the isomerization of a-pinene over
20 wt% WO3/Al2O3 calcined at various temperatures are sum-
marized in Table 1. Many products were formed in the isomeri-
zation of a-pinene, which can be divided into three groups,
1) bicyclic compounds such as fenchene and camphene,
2) monocyclic compounds such as limonene and terpinene,
and 3) b-pinene (Scheme 1). b-Pinene is produced on basic
sites, and bicyclic and monocyclic compounds are produced
on acid sites. Ohnishi et al.[39] and Yamamoto et al.[40,41] report-
Results and Discussion
Catalysis
The results of the benzylation of anisole on 20 wt%
WO3/Al2O3 calcined at various temperatures are
Table 1. Results for the isomerization of a-pinene on 20 wt% WO3/Al2O3 calcined at
various temperatures.[a]
shown in Figure 1a. This reaction has been per-
formed using various solid acid catalysts and is re-
ported to occur on Brønsted acid sites.[32–38] The main
products of this reaction were p- and o-benzyl ani-
sole, and dibenzyl ether was a minor product. No
multiply alkylated product was formed. Dibenzyl
ether is produced on Lewis acid sites by the dehydra-
tion of two benzyl alcohol molecules. The yield of di-
benzyl ether was less than 0.18 mmol in all of the re-
actions. The calcination temperature of the catalysts
affected its activity significantly. The maximum yield
of benzyl anisole was obtained using the catalyst cal-
Calcination T
[K]
Conversion
[%]
Selectivity[b] [%]
1
2
3
4
5
6
7
8
773
973
18.7
60.0
74.8
94.5
77.7
36.3
14.9
1
tr
1
2
1
66
68
68
66
61
65
76
3
1
tr
tr
tr
2
2
2
4
tr
2
19
21
21
18
26
23
4
1
1
tr
3
tr
1
tr
4
5
6
7
6
5
tr
5
3
2
1
3
3
tr
1073
1123
1173
1273
1423
1
tr
1
20
tr
[a] Pretreatment temperature: 473 K. [b] 1: a-fenchene, 2: camphene, 3: b-pinene, 4:
a-terpinene, 5: limonene, 6: g-terpinene, 7: terpinolene, 8: others, tr: trace.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2011 – 2020 2012

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Scheme 1. a-Pinene and the isomerized products.
ed that monocyclic and bicyclic compounds were produced on
Brønsted acid sites and their ratio depended on the maximum
acid strength of the catalyst. We also demonstrated that the
yields of camphene and limonene correlate with the Brønsted
acidity of the Nb2O5/Al2O3 catalyst.[26] The main products of the
isomerization of a-pinene on WO3/Al2O3 were camphene and
limonene with a selectivity of approximately 66 and 18%, re-
spectively. This result suggests that the isomerization of a-
pinene to camphene and limonene occurred on Brønsted acid
sites generated on the WO3/Al2O3 catalysts.
Figure 2. Activity of WO3/Al2O3 calcined at 1123 K with various loadings for
a) the Friedel–Crafts alkylation (*: benzyl anisole and *: dibenzyl ether)
and b) the isomerization of a-pinene (&: camphene and &: limonene).
Pretreatment temperature: 473 K.
The yield of camphene and limonene produced over
20 wt% WO3/Al2O3 calcined at various temperatures is shown
in Figure 1b. The yield increased with calcination temperature
up to 1123 K and then decreased above 1123 K. The selectivi-
ties for camphene and limonene were almost constant up to
1273 K. Conversely, above 1273 K, the selectivities for cam-
phene and b-pinene increased and that for limonene de-
creased. These results suggest that the number of acid sites in-
creased with calcination up to 1273 K, although the acid
strength remained constant. In contrast, calcination above
1273 K may affect not only the number of acid sites but also
the acid strength.
Table 2. Results for a-pinene isomerization on WO3/Al2O3 calcined at
1123 K with various loadings of WO3.[a]
Loading amount
of WO3 [wt%]
Conversion
[%]
Selectivity[b] [%]
1
2
3
4
5
6
7
8
0
5
15.9
7.6
tr 75 25 tr
tr tr tr tr
tr 61
tr 60
9
3
1
2
2
4
3
3
2
1
22
25
1
1
4
5
6
7
6
7
5
5
2
4
3
1
2
3
2
2
10
15
20
25
30
40
50
22.1
37.4
94.5
79.7
90.4
50.5
24.2
1
2
1
1
1
1
64
66
66
67
70
72
tr
tr
tr
tr
tr
1
23 tr
18
3
22 tr
19 tr
20 tr
18 tr
The results for the benzylation of anisole over WO3/Al2O3 cal-
cined at 1123 K with various loadings of WO3 are presented in
Figure 2a.
[a] Pretreatment temperature: 473 K. [b] 1: a-fenchene, 2: camphene, 3:
b-pinene, 4: a-terpinene, 5: limonene, 6: g-terpinene, 7: terpinolene, 8:
others, tr: trace.
The loading of WO3 affected the yields of benzyl anisoles
significantly. Alumina and WO3 alone were almost inactive. The
catalytic activity increased suddenly from 10 wt% WO3, and
the maximum yield was obtained for 20 wt% WO3. If the load-
ing exceeded 20 wt%, the catalytic activity gradually de-
creased. The ratio of p-benzyl anisole to o-benzyl anisole was
approximately 3:2 regardless of the WO3 loading. Such con-
stant selectivity shows that the acid strength was independent
of the WO3 loading but the amount of acid sites varied.
The results of the isomerization of a-pinene over WO3/Al2O3
calcined at 1123 K with various WO3 loadings are summarized
in Table 2 and Figure 2b. The yields of camphene and limo-
nene increased with the loading up to 20 wt% WO3. Over
20 wt% WO3, the yield of these compounds decreased. On alu-
mina alone, camphene and b-pinene were mainly produced.
The selectivities for camphene and b-pinene decreased and
that for limonene increased as the loading increased to
10 wt% WO3. No significant change of selectivity was found in
the range of 15–30 wt% WO3, whereas the selectivity for cam-
phene increased from 40 wt% WO3. These results indicate that
the number of acid sites changed, but the acid strength and
acid type did not in the range of 10–30 wt% WO3. Houalla
et al.[42] investigated the effect of the WO3 loading on the acidi-
ty of WO3/Al2O3 calcined at 1073 K and suggested that the
number of Brønsted acid sites increased if the WO3 loading ex-
ceeded 1.4 atoms of W per nm2 (9.65 wt%). In this study, WO3/
Al2O3 calcined at 1123 K showed activity if the loading was
10 wt%, which is consistent with their results.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2011 – 2020 2013

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 4. IR spectra of pyridine adsorbed on a) g-Al2O3 and b) 20 wt% WO3/
Al2O3 calcined at 1123 K.
Figure 3. Activity of 20 wt% WO3/Al2O3 calcined at 1123 K pretreated at vari-
ous temperature for the isomerization of a-pinene (&: camphene and
&: limonene).
The effect of the pretreatment temperature on the isomeri-
zation of a-pinene over 20 wt% WO3/Al2O3 calcined at 1123 K
is presented in Figure 3. The highest yields of camphene and
limonene were obtained if the catalyst was pretreated at
473 K. Although the yield of camphene decreased gradually as
the pretreatment temperature increased, WO3/Al2O3 showed
activity even after evacuation at 1173 K (ꢀ2/3 of the yield of
that pretreated at 473 K), which indicates that Brønsted acid
sites on WO3/Al2O3 are thermally stable. Previously, we report-
ed that the activity of a-pinene isomerization over Nb2O5/Al2O3
pretreated at 1173 K was almost same as the activity if the cat-
alyst was pretreated at 673 K and that Brønsted acid sites on
Nb2O5/Al2O3 calcined at high temperature were thermally
stable. In contrast, the activity over zeolites and silica-alumina
decreased significantly following pretreatment at 1173 K.
Figure 5. Brønsted acidity of 20 wt% WO3/Al2O3 calcined at various
temperatures.
As the benzylation of anisole and the isomerization of a-
pinene are reported to be promoted by Brønsted acid sites
and the catalytic activity of WO3/Al2O3 depended strongly on
the loading of WO3 and calcination temperature, the Brønsted
acidity of WO3/Al2O3 prepared under various conditions was
measured by pyridine adsorption FTIR spectra. The Brønsted
acid properties were estimated from the specific amount of
pyridinium ions adsorbed on each WO3/Al2O3 sample, and the
Lewis acid properties were estimated from the amount of pyri-
dine coordinated to Lewis acid sites. IR spectra of pyridine ad-
sorbed on Al2O3 and 20 wt% WO3/Al2O3 calcined at 1123 K are
shown in Figure 4. Four bands at n˜ =1447, 1490, 1578, and
1609 cmÀ1 were observed in the IR spectrum of alumina.[43–46]
These four bands correspond to pyridine species adsorbed on
Lewis acid sites. In contrast, WO3/Al2O3 exhibited new bands at
n˜ =1543 and 1640 cmÀ1, which correspond to pyridine species
adsorbed on Brønsted acid sites. Therefore, the presence of
WO3 generates Brønsted acid sites. The specific amount of
Brønsted acid sites determined from the adsorbed pyridinium
ions was calculated using an integrated molar adsorption coef-
ficient value of e=1.67 cmmmolÀ1 for the band area at n˜ =
1545 cmÀ1 of protonated pyridinium ions.[45,46] The calculated
Brønsted acidity of 20 wt% WO3/Al2O3 calcined at various tem-
peratures is shown in Figure 5. The Brønsted acidity increased
with calcination temperature up to 1123 K, whereas calcination
at temperatures above 1123 K resulted in a reduction of
Brønsted acidity. This result shows that high-temperature calci-
nation is important for the generation of Brønsted acid sites.
The Brønsted acidity of WO3/Al2O3 calcined at 1123 K with vari-
ous WO3 loadings is presented in Figure 6. No Brønsted acid
sites were present on Al2O3, and Brønsted acidity increased
with WO3 loading up to 20 wt%. Above 10 wt%, the Brønsted
acidity increased sharply, and 20 wt% WO3/Al2O3 exhibited the
highest Brønsted acidity of the catalysts. The Brønsted acidity
decreased gradually for samples with loadings over 20 wt%.
The yield of benzyl anisole, camphene, and limonene are pro-
portional to the Brønsted acidity of the catalyst (Figure 7),
which indicates that the Friedel–Crafts reaction and the iso-
merization of a-pinene occurred at the Brønsted acid sites.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2011 – 2020 2014

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 6. Brønsted acidity of WO3/Al2O3 calcined at 1123 K with various load-
ings of WO3.
Figure 8. XRD patterns of 20 wt% WO3/Al2O3 calcined at various tempera-
tures. a) 773, b) 1073, c) 1123, d) 1173, e) 1273, and f) 1423 K. g=g-Al2O3;
m=m-WO3; A=Al2W3O12
Figure 9. Raman spectra of 20 wt% WO3/Al2O3 calcined at various tempera-
tures. a) 1073, b) 1123, c) 1173, d) 1273, and e) 1423 K.
Figure 7. Correlation between the catalytic activities of WO3/Al2O3 calcined
at 1123 K with various loadings and their Brønsted acidities. a) The benzyla-
tion of anisole and b) the isomerization of a-pinene (&: camphene and
&: limonene).
Al2W3O12 cause the number of Brønsted acid sites on WO3/
Al2O3 calcined at temperatures higher than 1123 K to decrease.
XRD patterns and Raman spectra of WO3/Al2O3 calcined at
1123 K with various WO3 loadings were also measured (Fig-
ures 10 and 11, respectively). The 20 wt% WO3/Al2O3 sample
did not show any diffraction peak or Raman shift consistent
with m-WO3 or Al2W3O12. m-WO3 appeared for 25 wt% WO3/
Al2O3 and its contribution grew as the WO3 loading increased.
If we consider the change in acid properties related to the for-
mation of m-WO3 and Al2W3O12, the Brønsted acidity on WO3/
Al2O3 decreased with WO3 loadings above 20 wt%. These re-
sults suggest that amorphous tungsten oxide is the source of
the Brønsted acid sites.
Characterization
XRD patterns and Raman spectra of 20 wt% WO3/Al2O3 cal-
cined at various temperatures are shown in Figures 8 and 9, re-
spectively. Diffraction lines consistent with g-Al2O3 were ob-
served for all 20 wt% WO3/Al2O3. In contrast, crystalline WO3
was not observed in WO3/Al2O3 calcined below 1173 K. Mono-
clinic WO3 (m-WO3) was found in WO3/Al2O3 calcined at 1173 K,
which then transformed to Al2W3O12 if the catalyst was calcined
at 1273 K.
As m-WO3 and Al2W3O12 have no Brønsted acid sites and did
not promote the benzylation of anisole and isomerization of a-
pinene, it is supposed that the formation of m-WO3 and
The surface W/Al ratio on the catalysts was estimated from
the areas of the W4f and Al2p peaks in the X-ray photoelec-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2011 – 2020 2015

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 10. XRD patterns of WO3/Al2O3 calcined at 1123 K with various load-
ings of WO3. a) 5, b) 15, c) 20, d) 25, e) 30, f) 40, and g) 50 wt%. g=g-Al2O3;
m=m-WO3.
Figure 12. Atomic ratio of WO3/Al2O3 calcined at 1123 K with various load-
ings of WO3.
covered completely. This suggests that a monolayer of tung-
sten oxide was formed on 20 wt% WO3/Al2O3. As the maxi-
mum Brønsted acidity was observed for 20 wt% WO3/Al2O3,
a monolayer of tungsten oxide generated the Brønsted acid
sites.
The local structure of WO3/Al2O3 catalysts was characterized
by W L1-edge and W L3-edge extended X-ray absorption fine
structure (EXAFS) spectra. W L1-edge X-ray absorption near-
edge structure (XANES) spectra of WO3/Al2O3 calcined at
1123 K with various loadings along with reference samples are
shown in Figure 13. The pre-edge peak observed in the W L1-
edge XANES spectra, which is assigned to the transition from
Figure 11. Raman spectra of WO3/Al2O3 calcined at 1123 K with various load-
ings of WO3. a) 10, b) 20, c) 25, d) 30, e) 40, and f) 50 wt%.
tron spectra (XPS). The binding energies of the W4f7/2 and
W4f5/2 peaks of the WO3/Al2O3 catalysts were 35.5 and 37.6 eV,
respectively. These binding energies were constant regardless
of the calcination temperature and loading of WO3. This indi-
cates that the valence of the tungsten cation of WO3/Al2O3 was
W6+ regardless of calcination temperature and WO3 loading.
The surface W/Al ratio for 20 wt% WO3/Al2O3 calcined at vari-
ous temperatures was constant, which indicates that W atoms
did not penetrate into the bulk structure independent of the
calcination temperature. The surface W/Al ratio of WO3/Al2O3
calcined at 1123 K with various WO3 loadings is shown in
Figure 12. The surface W/Al ratio increased linearly with WO3
loading up to 20 wt% and then increased gradually. The cross-
sectional area of a WO6 octahedral unit is 0.22 nmÀ2, so if WO3
is loaded on alumina at 20 wt%, the surface of the catalyst is
Figure 13. W L1-edge XANES spectra of reference samples and WO3/Al2O3
calcined at 1123 K with various loadings of WO3. a) Sr2CaWO6, b) m-WO3, c) 5,
d) 10, e) 15, f) 20, g) 25, h) 30, i) 40 wt%, j) Al2W3O12, and k) Na2WO4.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2011 – 2020 2016

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
2s to d-p orbitals, is sensitive to the symmetry around W[47–49]
and can be used to provide information on local symmetry.
The shapes of the spectra of WO3/Al2O3 resemble that of WO3,
which indicates that the local structure of WO3 supported on
alumina is similar to that of WO3. The tungsten oxide unit in
WO3/Al2O3 samples is distorted from ideal octahedral symme-
try, seen from the intense pre-edge peak in the W L1-edge
XANES spectra. Therefore, the pre-edge peak of Na2WO4 (tetra-
hedral WO4) is much more intense than that of tungsten oxide
(distorted octahedral) and that of SrCaWO3 (octahedral WO6) is
much less intense than that of tungsten oxide. If the WO3 load-
ing is lower than 20 wt%, the pre-edge peak area of WO3/
Al2O3 is larger than that of SrCaWO3 and smaller than that of
Na2WO4, which indicates that the coordination symmetry of W
species on WO3/Al2O3 is distorted octahedral.
The W L3-edge k3-weighted EXAFS oscillations of WO3/Al2O3
catalysts with various loadings and reference samples are
shown in Figure 14. The shape of the EXAFS oscillations of the
WO3/Al2O3 catalysts with WO3 loadings of up to 20 wt% was
Figure 15. Fourier transforms of W L3-edge k3-weighted EXAFS of reference
samples and WO3/Al2O3 calcined at 1123 K with various loadings of WO3.
a) m-WO3, b) 5, c) 10, d) 15, e) 20, f) 25, g) 30, h) 40 wt%, and i) Al2W3O12.
WO3 loadings higher than 20 wt%. This indicates that WÀ(O)À
W linkages of various lengths were formed on 20 wt% WO3/
Al2O3, and that WÀ(O)ÀW linkages with a certain length in WO3
clusters were mixed and their length increased for WO3/Al2O3
with WO3 loadings that exceed 20 wt%. These results imply
that monolayer tungsten oxide domains are formed on the
20 wt% WO3/Al2O3 catalyst. If we consider that the 20 wt%
WO3/Al2O3 catalyst exhibits the highest Brønsted acidity, mono-
layer tungsten oxide domains that have a m-WO3-like structure
may play an important role in the generation of Brønsted acid
sites.
The physical properties of WO3/Al2O3 catalysts also depend-
ed on the calcination temperature and loading of WO3. The
physical properties of WO3/Al2O3 with various loadings are
summarized in Table 3. The catalyst areas include that of the
alumina support. S0 is the surface area [g(Al2O3)À1] of a carrier
and was calculated from the BET specific surface area and WO3
content. Soccupied is the estimated area occupied by WO6 units.
Figure 14. W L3-edge k3-weighted EXAFS of reference samples and WO3/
Al2O3 calcined at 1123 K with various loadings of WO3. a) m-WO3, b) 5, c) 10,
d) 15, e) 20, f) 25, g) 30, h) 40 wt%, and i) Al2W3O12.
constant. With loadings over 20 wt%, the shape of
Table 3. Physical properties of WO3/Al2O3 calcined at 1123 K.
the EXAFS oscillations of the WO3/Al2O3 catalysts in
the range of 3–5 and 8–10 ꢁ resemble that of WO3.
This result indicates that amorphous WO3 trans-
formed gradually to crystalline WO3 on the WO3/Al2O3
catalysts with loadings of WO3 above 20 wt%. Fouri-
er-transformed EXAFS oscillations of WO3/Al2O3 cata-
lysts with various loadings and reference samples are
shown in Figure 15. Two peaks were observed at 1.4
and 3.6 ꢁ in the radial structure functions (RSFs) of
WO3 that correspond to WÀO and WÀ(O)ÀW linkages,
respectively. Up to a WO3 loading of 20 wt%, only
one peak at 1.4 ꢁ was found. A new peak at 3.6 ꢁ ap-
peared and increased in intensity for WO3/Al2O3 with
[b]
[c]
[d]
[d]
W content
[mmolg(Al2O3)À1
SBET
S0
Soccupied
Soccupied
]
[wt%][a]
[m2 gÀ1
]
[m2]
[m2 g(cat)À1
]
[m2 g(Al2O3)À1
]
0
0
5
124
131
120
119
118
101
93
124
138
133
140
148
135
133
142
102
0
28.6
57.1
85.7
114
143
171
229
286
0
30.1
63.4
101
143
191
244
382
572
0.22
0.48
0.76
1.08
1.44
1.85
2.88
4.31
10
15
20
25
30
40
50
85
51
[a] As WO3. [b] BET surface area. [c] Surface area of g(Al2O3)À1 carrier. [d] Area occupied
by WO3 units (0.22 nm2).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2011 – 2020 2017

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
It was assumed that all WO3 was present as WO6 octahedra
and the area of each unit is 0.22 nm2.[50] The increase of the
loading of WO3 caused the surface area of WO3/Al2O3 to de-
crease. We suppose that shrinkage of the Al2O3 support is
a cause of this decrease. The area of the alumina support (S0)
decreases and the area occupied by WO6 units (Soccupied) in-
creases with the loading of WO3, and Soccupied becomes compa-
rable to S0 at 20 wt%. If we consider the XPS results, at just
20 wt%, a WO3 monolayer covered most of the surface of the
alumina support. The surface density of W atoms on 20 wt%
WO3/Al2O3 calcined at 1123 K is estimated to be
4.4 atomsnmÀ2. Wachs[51] reported that the monolayer cover-
age of WO3 on alumina corresponds to 4.0 atomnmÀ2, which is
comparable to our estimated value. These results mean that
WO3 is loaded as a 2D monolayer at loadings less than
20 wt%. The acid properties of supported tungsten oxide have
been investigated by a lot of researchers. WO3/ZrO2 is a solid
superacid catalyst, and many researchers have inspected the
acid properties and the structure of tungsten oxide supported
on zirconia.[13–18,49,52–55] The acid properties of WO3/ZrO2 are af-
fected by calcination and the WO3 loading. It has been report-
ed that WO3/ZrO2 shows the highest activity if the surface den-
sity of tungsten oxide is 7–8 atomnmÀ2, which is a slightly
larger than the loading at which a tungsten oxide monolayer
covers the ZrO2 support.[13,17,54] Moreover, some researchers re-
ported that tungsten oxide is loaded as a monolayer on zirco-
nia, and polymeric tungsten oxide plays a role in the genera-
tion of acid sites.[17,18] These results indicate that 2D tungsten
oxide is important in the generation of acid sites.
Around 20 wt%, tungsten oxide monolayer domains cover
most of the alumina surface, and the number of boundaries in-
creased, which resulted in a significant increase of Brønsted
acidity. If Brønsted acid sites are generated on the tungsten
oxide monolayer domains (WÀOH or WÀ(OH)ÀW) or on the
edges of tungsten oxide monolayer domains and alumina (WÀ
(OH)ÀAl), the activity and Brønsted acidity would increase line-
arly with the loading.
A model for the generation of Bronsted acid sites by high-
temperature calcination is shown in Figure 17. The surface area
Figure 17. Models of Brønsted acid sites on WO3/Al2O3
of alumina decreased if the alumina was calcined at high tem-
peratures because the alumina shrank. On WO3/Al2O3 calcined
at low temperatures such as 773 K, tungsten oxide monolayer
domains are dispersed on the alumina surface. In this case, the
number of Brønsted acid sites is small because the number of
boundaries between the WO3 monolayer domains is small. If
the calcination temperature is increased, the tungsten oxide
monolayer domains assemble and the number of boundaries
increases because the alumina shrinks. As a result, the Brønst-
ed acidity increases with calcination temperature up to 1123 K.
Calcination at temperatures higher than 1123 K causes the
number of Brønsted acid sites to decrease because of the ag-
gregation of the tungsten oxide monolayer domains to form
inert crystalline WO3 and Al2W3O12.
In this study, the changes in the acid properties and struc-
ture of WO3/Al2O3 with WO3 loading may be summarized as
follows. (1) The activity for acid-catalyzed reactions and Brønst-
ed acidity increased with the WO3 loading up to 20 wt%.
(2) The acid strength did not change with the loading amount.
(3) The ratio of Brønsted acidity to W atoms supported on
WO3/Al2O3 was quite small (0.04) for 20 wt% WO3/Al2O3 cal-
cined at 1123 K. (4) The surface was covered with a 2D tung-
sten oxide monolayer at loadings up to 20 wt%. Based on
these results, we conclude that hydroxyl groups formed at the
boundary of the tungsten oxide monolayer domains act as
Brønsted acid sites (Figure 16). First, tungsten monolayer do-
mains are dispersed on alumina and few boundaries between
domains are present. As the WO3 loading increases, the
number of monolayer domains and coverage increase, so
boundaries between tungsten oxide monolayer domains form
and the number of Brønsted acid sites increase simultaneously.
We conclude that the generation of Brønsted acid sites by
calcination at high temperatures is related to (1) the formation
of tungsten oxide monolayer domains on the surface of alumi-
na, (2) the formation of boundaries between domains of the
tungsten oxide monolayer, and (3) the shrinkage of alumina
upon calcination at high temperatures, which reduces its sur-
face area. As a result, the distance between the domains of the
tungsten oxide monolayer decreases, which allows boundaries
between tungsten monolayer domains to generate Brønsted
acid sites.
Conclusions
WO3/Al2O3 calcined at high temperature exhibited catalytic ac-
tivities for the benzylation of anisole and the isomerization of
a-pinene and Brønsted acidity. The calcination temperature
and loading of WO3 affected the acid properties and the struc-
Figure 16. Models of Brønsted acid sites on WO3/Al2O3.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2011 – 2020 2018

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
ture of WO3/Al2O3. 20 wt% WO3/Al2O3 calcined at 1123 K exhib-
ited both the highest reactivity and Brønsted acidity. WO3/
Al2O3 showed activity for the isomerization of a-pinene even
after evacuation at 1173 K, which indicates that the Brønsted
acid sites generated on WO3/Al2O3 were thermally stable. Tung-
sten oxide was loaded as 2D monolayer domains below
20 wt%, and these domains covered most of the alumina sur-
face at 20 wt%. Additional loading of WO3 causes the forma-
tion of inert WO3 crystals and a decrease of Brønsted acidity.
Brønsted acid sites are probably generated at the boundaries
of tungsten oxide monolayer domains. A proposed generation
mechanism of the Brønsted acid sites is as follows: At a low
loading amount, tungsten oxide monolayer domains are dis-
persed on alumina and few boundaries are present between
domains. As the WO3 loading increases, the number of tung-
sten oxide monolayer domains increases and boundaries be-
tween domains form Brønsted acid sites. If tungsten oxide
monolayer domains cover most of the alumina surface
(20 wt% WO3/Al2O3), the amount of boundaries between the
tungsten oxide monolayer domains is maximized, which gener-
ates the largest number of Brønsted acid sites.
Characterization
The BET specific surface areas of catalysts were estimated from N2
isotherms at 77 K. N2 adsorption isotherms were measured at 77 K
by using an adsorption analyzer (BELSORP 28SA, BEL Japan, Osaka,
Japan). Prior to measurement, each sample was evacuated at 573 K
for 2 h. The surface area was determined from six points of data
[adsorption amount versus relative pressures (P/P0) ranging from
0.05 to 0.2] in each isotherm.
XRD patterns were obtained by using a powder X-ray diffractome-
ter (MultiFlex DR, Rigaku, Tokyo Japan) with CuKa radiation (c=
1.5405 ꢁ).
Laser Raman spectra were obtained by using a Raman spectrome-
ter (NRS-2000, JASCO, Tokyo, Japan) using the 514.5 nm line of an
Ar laser. The spectral resolution was 4 cmÀ1
.
FTIR spectra were recorded using an FTIR spectrometer (SPECTRUM
ONE, PerkinElmer, Walthham, MA, USA) with a resolution of 4 cmÀ1
.
Each sample (13 mg) was pressed into a self-supporting wafer with
a diameter of 13 mm. Catalysts were pretreated under 6.7 kPa of
O2 for 1 h at 673 K and then evacuated for 1 h at 673 K. To deter-
mine the number of Brønsted and Lewis acid sites on WO3/Al2O3,
the wafer was exposed to 0.667 kPa of pyridine vapor at 298 K for
10 min and then evacuated at 423 K for 10 min.
XPS spectra of the catalysts were acquired by using an X-ray pho-
toelectron spectrometer (5500MT, ULVAC PHI, Kanagawa, JAPAN)
equipped with a hemispherical energy analyzer. Samples were
mounted on In foil and then transferred to the chamber of the
spectrometer. The residual gas pressure in the chamber during
data acquisition was less than 1.33ꢂ10À6 Pa. Spectra were mea-
sured at RT using MgKa radiation. The electron take-off angle was
set at 458. Binding energies were referenced to the O1s level.
Experimental Section
Preparation
A series of WO3/Al2O3 catalysts were prepared by the impregnation
of g-Al2O3 (JRC-ALO-8) with an aqueous solution of ammonium
tungsten parapentahydrate [(NH4)10W12O42·5H2O)] at 353 K. The
product was dried overnight at 353 K and then calcined at various
temperatures for 3 h in dry air. Aluminum tungstate (Al2W3O12) was
synthesized by heating a mixture of Al2O3 and WO3 at 1373 K for
12 h. WO3 was obtained by the calcination of (NH4)10W12O42·5H2O
at 773 K for 3 h.
X-ray absorption experiments were performed by using the
BL01B1 beamline at SPring-8 (Hyogo, Japan). The ring energy was
8 GeV, and the stored current was 99.5 mA. W L1-edge (12.1 keV)
and W L3-edge (10.2 keV) X-ray absorption spectra were recorded
in transmission mode in air or N2 at RT. Si (311) and Si (111) two-
crystal monochromators were used to obtain a monochromatic X-
ray beam. Data reduction was performed by using the REX2000
Ver.2.5.9 (Rigaku) and FEFF8.40 programs.[56]
Reactions
Acknowledgements
The benzylation of anisole (Friedel–Crafts alkylation of anisole with
benzylalcohol) and the isomerization of a-pinene were used as test
reactions to examine the acid properties of WO3/Al2O3.
This work was supported in part by Grant-in-Aid for Scientific Re-
search (B) (Grant 23360355) and Grant-in-Aid for Young Scientist
(B) (21760627) from the Ministry of Education, Culture, Sports,
Science and Technology and by the Iwatani Naoji Foundation.
The X-ray absorption experiments were performed with the ap-
proval of the Japan Synchrotron Radiation Research Institute
(JASRI) (Proposal No. 2009A1606).
The benzylation of anisole was examined in the liquid phase. The
catalyst (0.5 g) was pretreated in a N2 flow at 473 K for 1 h and
then added to a mixture of benzyl alcohol (6.25 mmol) and anisole
(92.5 mmol) in a 100 mL flask. The reaction was performed at 353 K
for 1 h, and the products were determined by gas-liquid chroma-
tography (GLC; GC-14B with a flame ionization detector, Shimadzu,
Kyoto, Japan) and GC–MS (GC–MS QP-5050, Shimadzu), using
a CBP10 column.
Keywords: acidity · heterogeneous catalysts · isomerization ·
monolayers · tungsten
The isomerization of a-pinene was performed under a dry N2 at-
mosphere by using a stirred batch reactor. Prior to each run, the
sample (0.1 g) was pretreated at various temperatures under
13.3 kPa of O2 for 1 h and then evacuated at the same temperature
for 1 h. a-Pinene (12.5 mmol) was added to the reactor and stirred
at 323 K for 3 h. The products were determined by GLC (GC-2014
with a flame ionization detector, Shimadzu) using a CBP20 column.
[1] T. Okuhara, N. Mizuno, M. Misono in Catalytic Chemistry of Heteropoly
Compounds, Vol. 41, 1996, pp. 113–252.
[2] E. J. L. Lana, K. A. D. Rocha, I. V. Kozhevnikov, E. V. Gusevskaya, J. Mol.
Catal. A 2006, 243, 258–263.
[3] E. J. Le¼o Lana, K. A. da Silva Rocha, I. V. Kozhevnikov, E. V. Gusevskaya,
J. Mol. Catal. A 2006, 259, 99–102.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2011 – 2020 2019

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[4] J. Zhang, M. Kanno, Y. Wang, H. Nishii, Y. Miura, Y. Kamiya, J. Phys. Chem.
C 2011, 115, 14762–14769.
[34] K. Yamashita, M. Hirano, K. Okumura, M. Niwa, Catal. Today 2006, 118,
385–391.
[5] E. Rafiee, S. Rashidzadeh, A. Azad, J. Mol. Catal. A 2007, 261, 49–52.
[6] H. Firouzabadi, N. Iranpoor, A. A. Jafari, M. R. Jafari, J. Mol. Catal. A 2006,
247, 14–18.
[7] T. Nakato, M. Kimura, S. Nakata, T. Okuhara, Langmuir 1998, 14, 319–
325.
[8] T. Okuhara, Chem. Rev. 2002, 102, 3641–3665.
[9] Y. Izumi, R. Hasebe, K. Urabe, J. Catal. 1983, 84, 402–409.
[10] Y. Izumi, M. Ono, M. Kitagawa, M. Yoshida, K. Urabe, Microporous Mater.
1995, 5, 255–262.
[11] I. V. Kozhevnikov, Chem. Rev. 1998, 98, 171–198.
[12] P. A. Jalil, N. Tabet, M. Faiz, N. M. Hamdan, Z. Hussain, Appl. Catal. A
2004, 257, 1–6.
[35] K. Okumura, K. Yamashita, M. Hirano, M. Niwa, Chem. Lett. 2005, 34,
716–717.
[36] K. Okumura, K. Yamashita, M. Hirano, M. Niwa, J. Catal. 2005, 234, 300–
307.
[37] T. Cseri, S. Bekassy, F. Figueras, E. Cseke, L. C. Demenorval, R. Dutartre,
Appl. Catal. A 1995, 132, 141–155.
[38] C. Tagusagawa, A. Takagaki, S. Hayashi, K. Domen, J. Am. Chem. Soc.
2008, 130, 7230–7231.
[39] R. Ohnishi, K. Tanabe, S. Morikawa, T. Nishizaki, Bull. Chem. Soc. Jpn.
1974, 47, 571–574.
[40] T. Yamamoto, T. Tanaka, T. Funabiki, S. Yoshida, J. Phys. Chem. B 1998,
102, 5830–5839.
[13] J. G. Santiesteban, J. C. Vartuli, S. Han, R. D. Bastian, C. D. Chang, J.
Catal. 1997, 168, 431–441.
[41] T. Yamamoto, T. Tanaka, T. Matsuyama, T. Funabiki, S. Yoshida, J. Phys.
Chem. B 2001, 105, 1908–1916.
[14] K. Arata, Appl. Catal. A 1996, 146, 3–32.
[42] X. Chen, G. Clet, K. Thomas, M. Houalla, J. Catal. 2010, 273, 236–244.
[43] C. Morterra, G. Magnacca, Catal. Today 1996, 27, 497–532.
[44] P. A. Jacobs, H. E. Leeman, J. B. Uytterhoeven, J. Catal. 1974, 33, 17–30.
[45] T. Barzetti, E. Selli, D. Moscotti, L. Forni, J. Chem. Soc. Faraday Trans.
1996, 92, 1401–1407.
[15] K. Arata, H. Nakamura, M. Shouji, Appl. Catal. A 2000, 197, 213–219.
[16] K. Arata, H. Matsuhashi, M. Hino, H. Nakamura, Catal. Today 2003, 81,
17–30.
[17] D. G. Barton, S. L. Soled, G. D. Meitzner, G. A. Fuentes, E. Iglesia, J. Catal.
1999, 181, 57–72.
[46] C. A. Emeis, J. Catal. 1993, 141, 347–354.
[18] N. Naito, N. Katada, M. Niwa, J. Phys. Chem. B 1999, 103, 7206–7213.
[19] K. Tanabe, Catal. Today 1990, 8, 1–11.
[47] S. Yoshida, T. Tanaka, T. Hanada, T. Hiraiwa, H. Kanai, T. Funabiki, Catal.
Lett. 1992, 12, 277–286.
[20] K. Tanabe, S. Okazaki, Appl. Catal. A 1995, 133, 191–218.
[21] K. Tanabe, Catal. Today 2003, 78, 65–77.
[48] H. Yoshida, T. Tanaka, T. Yoshida, T. Funabiki, S. Yoshida, Catal. Today
1996, 28, 79–89.
[22] T. Ushikubo, K. Wada, Appl. Catal. 1990, 67, 25–38.
[23] T. Ohuchi, T. Miyatake, Y. Hitomi, T. Tanaka, Catal. Today 2007, 120, 233–
239.
[24] T. Shishido, T. Kitano, K. Teramura, T. Tanaka, Catal. Lett. 2009, 129, 383–
386.
[49] T. Yamamoto, A. Orita, T. Tanaka, X-Ray Spectrom. 2008, 37, 226–231.
[50] C. Pfaff, M. J. P. Zurita, C. Scott, P. Patino, M. R. Goldwasser, J. Goldwass-
er, F. M. Mulcahy, M. Houalla, D. M. Hercules, Catal. Lett. 1997, 49, 13–
16.
[51] I. E. Wachs, Catal. Today 1996, 27, 437–455.
[52] E. I. Ross-Medgaarden, W. V. Knowles, T. Kim, M. S. Wong, W. Zhou, C. J.
Kiely, I. E. Wachs, J. Catal. 2008, 256, 108–125.
[25] T. Shishido, T. Kitano, K. Teramura, T. Tanaka, Top. Catal. 2010, 53, 672–
677.
[26] T. Kitano, T. Shishido, K. Teramura, T. Tanaka, J. Phys. Chem. C 2012, 116,
11615–11625.
[53] W. Zhou, E. I. Ross-Medgaarden, W. V. Knowles, M. S. Wong, I. E. Wachs,
C. J. Kiely, Nat. Chem. 2009, 1, 722–728.
[27] T. Ushikubo, K. Wada, Chem. Lett. 1988, 1573–1574.
[28] T. Kitano, S. Okazaki, T. Shishido, K. Teramura, T. Tanaka, Chem. Lett.
2011, 40, 1332–1334.
[29] T. Kitano, S. Okazaki, T. Shishido, K. Teramura, T. Tanaka, Catal. Today
2012, 192, 189–196.
[30] M. Hino, H. Matsuhashi, K. Arata, Catal. Lett. 2006, 107, 161–163.
[31] S. L. Soled, G. B. McVicker, L. L. Murrell, L. G. Sherman, N. C. Dispenziere,
S. L. Hsu, D. Waldman, J. Catal. 1988, 111, 286–295.
[32] M. H. C. de La Cruz, J. F. C. da Silva, E. R. Lachter, Catal. Today 2006, 118,
379–384.
[54] M. Scheithauer, R. K. Grasselli, H. Knozinger, Langmuir 1998, 14, 3019–
3029.
[55] P. Afanasiev, C. Geantet, M. Breysse, G. Coudurier, J. C. Vedrine, J. Chem.
Soc. Faraday Trans. 1994, 90, 193–202.
[56] A. L. Ankudinov, B. Ravel, J. J. Rehr, S. D. Conradson, Phys. Rev. B 1998,
58, 7565–7576.
Received: January 18, 2014
[33] M. H. C. de La Cruz, M. A. Abdel-Rehim, A. S. Rocha, J. F. C. da Silva, A. D.
Faro, E. R. Lachter, Catal. Commun. 2007, 8, 1650–1654.
Published online on June 6, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2011 – 2020 2020