Generation of Br?nsted acid sites on Al2O 3-supported Ta2O5 calcined at high temperatures

Article abstract of DOI:10.1016/j.cattod.2012.02.035

An alumina-supported tantalum oxide (Ta₂O₅/Al ₂O₃) calcined at 1223 K promoted both Friedel-Crafts alkylation of anisole with benzyl alcohol and isomerization of α-pinene, and exhibited Bronsted acidity despite the high temperature calcination. Among the Ta₂O₅/Al₂O₃ catalysts tested, 33 wt% Ta₂O₅/Al₂O₃ calcined at 1223 K showed the highest activity. A monolayer of tantalum acid-like compound, which has distorted octahedral symmetry, was stabilized over 33 wt% Ta ₂O₅/Al₂O₃ catalyst calcined at 1223 K. The two-dimensional Ta-O-Ta network of stabilized tantalum acid-like compound probably accounts for the generation of Bronsted acid.

Full text of DOI:10.1016/j.cattod.2012.02.035

Catalysis Today 192 (2012) 189–196
Contents lists available at SciVerse ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Generation of Brønsted acid sites on Al2O3-supported Ta2O5 calcined at high
temperatures
Tomoyuki Kitanoa, Shota Okazakia, Tetsuya Shishidoa,∗, Kentaro Teramuraa,b, Tsunehiro Tanakaa
a Department of Molecular Engineering, Graduate School of Engineering, Kyoto University, Kyoto 615-8510, Japan
b Precursory Research for Embryonic Science and Technology (PRESTO), Japan Science and Technology Agency (JST), 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
An alumina-supported tantalum oxide (Ta2O5/Al2O3) calcined at 1223 K promoted both Friedel–Crafts
alkylation of anisole with benzyl alcohol and isomerization of ␣-pinene, and exhibited Brønsted acidity
despite the high temperature calcination. Among the Ta2O5/Al2O3 catalysts tested, 33 wt% Ta2O5/Al2O3
calcined at 1223 K showed the highest activity. A monolayer of tantalum acid-like compound, which has
distorted octahedral symmetry, was stabilized over 33 wt% Ta2O5/Al2O3 catalyst calcined at 1223 K. The
two-dimensional Ta–O–Ta network of stabilized tantalum acid-like compound probably accounts for the
generation of Brønsted acid.
Received 30 October 2011
Received in revised form 13 January 2012
Accepted 9 February 2012
Available online 7 March 2012
Keywords:
Brønsted acid
Tantalum acid
© 2012 Elsevier B.V. All rights reserved.
High temperature calcination
Friedel–Crafts alkylation
␣-Pinene isomerization
1. Introduction
We have reported that a solid acid prepared with high load-
ing of Nb2O5 on an Al2O3 support and calcined at 1173 K exhibited
Tantalum acid (Ta2O5·nH2O, hydrated tantalum oxide) shows
mainly Lewis acid properties in the absence of water [1,2], but
Brønsted acid sites are generated by treatment with steam at 373 K.
Compared to niobic acid (Nb2O5·nH2O, hydrated niobium oxide),
which is known as a water-tolerant solid acid catalyst that exhibits
strong Brønsted acidity and is effective for reactions in aqueous
media such as esterification, olefin hydration, and alcohol dehydra-
tion [3–7], tantalum acid retains its strong acidity after calcination
at higher temperatures. In the case of niobic acid, dehydration leads
to lose of its acid property by heating above 773 K [8]. On the other
hand, tantalum acid maintains its acidic property up to 1073 K.
Generally, high temperature calcination usually causes solid acid
catalysts to lose the acid property. Ordinary solid acid catalysts,
such as aluminosilicate, zeolite and binary oxides, pretreated at
>873 K lose their acid property [9–11], except in some cases like
SiO2/Al2O3 prepared with CVD method [12], highly siliceous zeo-
lites and layered mixed oxide [13–15]. High stability is one of the
important factors on useful solid acid catalyst, because the life of
solid acid catalyst strongly depends on the stability. If a solid acid
could be prepared by calcination at high temperatures required
for the formation of ceramics (>1173 K), it would be a useful acid-
catalyst even under strict conditions because of its high stability.
Brønsted acid property despite the high temperature calcination.
This strongly suggests that the alumina support stabilizes the
Brønsted acid sites during calcination at high temperature [16,17].
Amount of Brønsted acid site on Nb2O5/Al2O3 changed depend-
ing on both calcination temperature and loading amount of Nb2O5.
This strongly suggests that the structural change around niobium
is expected to have a major impact on Brønsted acid property. We
have proposed that the stabilized niobic acid-like compound which
has two-dimensional Nb–O–Nb network of probably account for
the generation of Brønsted acid site.
Very recently, we reported that alumina-supported tantalum
oxide (Ta2O5/Al2O3) calcined at high temperature such as 1223 K
with high loading of Ta2O5 exhibited Brønsted acid property [18],
suggesting that Ta2O5/Al2O3 would also act as a stable Brønsted
acid catalyst. However, the relationship between the structural
change around tantalum and generation of Brønsted acid site on
Ta2O5/Al2O3 catalyst is still unclear. Here, we have prepared a series
of Ta2O5/Al2O3 solid acids and the relationship between the local
structure of tantalum species in Ta2O5/Al2O3 solid acids and acid
property is investigated.
2. Experimental
2.1. Preparation of catalysts
∗
A series of Ta2O5/Al2O3 solid acids were prepared by impreg-
nation of ␥-alumina (JRC-ALO-8, Catalysis Society of Japan,
Corresponding author. Tel.: +81 75 383 2559; fax: +81 75 383 2561.
E-mail address: shishido@moleng.kyoto-u.ac.jp (T. Shishido).
0920-5861/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2012.02.035

190
T. Kitano et al. / Catalysis Today 192 (2012) 189–196
Fig. 1. Activity of Ta2O5/Al2O3 for Friedel–Crafts alkylation of benzyl alcohol with anisole. (A) Calcined at various temperatures, Ta2O5 loading: 33 wt%, (B) with various
Ta2O5 loadings. Calcination temperature: 1223 K, reaction temperature: 433 K, reaction time: 1 h.
148 m2 g−1
) with an ethanol solution of tantalum ethoxide
diffractometer (Rigaku MultiFlex DR Powder X-ray diffractome-
ter) using Cu K␣ radiation (ꢀ = 1.5405 A). For the measurement
˚
(Ta(OC2H5)5), followed by calcination at various temperatures
for 3 h in dry air. Ta2O5 loading amount was changed from
10 wt% to 50 wt%, and calcination temperature was changed from
773 K to 1423 K. Aluminum tantalate, AlTaO4, was synthesized
from tantalum ethoxide and aluminum isopropoxide (Al(OC3H7)3)
according to an established procedure [19]. Hydrated tantalum
oxide (Ta2O5·nH2O) was prepared from tantalum ethoxide by
calcination at 573 K according to the literature [1,2]. Crystalline
T-phase Ta2O5 was prepared from Ta2O5·nH2O by calcination at
1223 K.
of XRD patterns, the scanning rate and was 0.067 s−1 and the
resolution was 0.02◦. Laser Raman spectra were taken with the
514.5 nm line of the argon laser (JASCO NRS-2000 Raman Spec-
trometer). The scan number, the acquisition time and the resolution
were 16 scans, 70 s and resolution 3.8 cm−1, respectively. The laser
power was 100 mW. Infrared spectra of adsorbed lutidine (2,6-
dimethylpyridine) were recorded with Perkin-Elmer Spectrum One
Fourier transform infrared spectrometer, and used to elucidate the
generation of Brønsted acid sites over Ta2O5/Al2O3. The sample
(13 mg) was pressed into a self-supporting disk (Ø 13 mm) and
pretreated under 13.3 kPa of O2 for 1 h at 773 K, and then evac-
uated for 1 h at 773 K. Then the sample was exposed to 0.53 kPa
of lutidine vapor at 298 K for 10 min followed by evacuation at
473 K for 10 min. The amount of Brønsted acid sites was calculated
based on the amount of adsorbed lutidine. The amount of adsorbed
lutidine was estimated by using the integrated molar adsorption
coefficient value of ε = 6.8 cm ␮mol−1 for the band at 1628 cm−1 and
1647 cm−1 of protonated lutidine as reported by Onfloy et al. [20].
X-ray photoelectron spectra (XPS) were acquired using an ULVAC
PHI 5500MT. XPS samples were mounted on indium foil and the
spectra were measured using Mg Ka radiation (15 kV, 400 W) in
a chamber with the base pressure of ca. 1 × 10−8 Torr. The take-
off angle was set at 45◦. Binding energies were referenced to Al(2p)
level at 74.0 eV. Ta L3- and L1-edges (9.9 and 11.7 keV, respectively)
XAFS data were collected at a facility of the BL01B1 beam line at
the SPring-8, of the Japan Synchrotron Radiation Research Institute,
Japan in the transmission mode in air at room temperature. Inci-
dent X-ray flux and transmitted flux were measured with ionization
chambers. Si (3 1 1) two crystal monochromator for Ta L1-edge and
Si (1 1 1) two crystal monochromator for Ta L3-edge were used to
obtain monochromatic X-ray beam. Energy was calibrated by Ta
foil, and the energy step of measurement in the XANES region was
0.5 eV. The absorption was normalized to 1.0 at an energy position
of ca. 100 eV higher than the absorption edge. The data reduction
was performed by the REX2000 Ver.2.5.9 (Rigaku) and FEFF 8.40
[21] programs.
2.2. Catalytic activity
Friedel–Crafts alkylation and isomerization of ␣-pinene were
carried out as test reactions to investigate the acidic properties of
Ta2O5/Al2O3. Friedel–Crafts alkylation of anisole with benzyl alco-
hol was examined in the liquid phase. The catalyst (0.15 g) was
pretreated at 473 K for 1 h in N2 flow and added to a mixture of
benzyl alcohol (6.25 mmol) and anisole (92.5 mmol) in a 100 mL
flask. The mixture was allowed to react at 433 K. Isomerization
of ␣-pinene was also examined in the liquid phase. The catalyst
(0.10 g) was pretreated at 573 K under 13.3 kPa of O2 for 1 h and
evacuated at the same temperature for 1 h. Then the sample was
added to ␣-pinene (12.5 mmol) in a 20 mL container. The reac-
tion was carried out at 353 K. Products were analyzed by FID-GC
(Shimadzu GC-14B, Kyoto, Japan) and GC-MS (Shimadzu QP-5050).
Tridecane was used for the internal standard in the both of reac-
tions. The carbon balance was above 90% in the both alkylation and
isomerization.
2.3. Characterization
Specific surface area was evaluated by BET method at 77 K (liq-
uid nitrogen temperature) with BEL Japan BELSORP28 28A. The
surface area was determined from six points of data (adsorption
amount vs relative pressures (P/P0) ranging from 0.05 to 0.2) in
the isotherm. X-ray diffraction (XRD) was performed on an X-ray

T. Kitano et al. / Catalysis Today 192 (2012) 189–196
191
Fig. 2. Activity of Ta2O5/Al2O3 for isomerization of ␣-pinene. (A) Calcined at various temperatures, Ta2O5 loading: 33 wt%, (B) with various Ta2O5 loadings. Calcination
temperature: 1223 K, reaction temperature: 353 K, reaction time: 3 h.
3. Results and discussion
33 wt% Ta2O5/Al2O3 catalyst exhibited the highest activity in both
Friedel–Crafts alkylation and isomerization of ␣-pinene. In the case
of 33 wt% Ta2O5/Al2O3 catalyst calcined at 1223 K, the TOFs were
110 h−1 for Friedel–Crafts alkylation and 106 h−1 for isomerization
of ␣-pinene, respectively.
3.1. Catalytic activity
The catalytic activities of Ta2O5/Al2O3 catalysts have been inves-
tigated as a function of calcination temperature. Loading amount of
tantalum oxide was 33 wt%. Fig. 1(A) shows the effect of the calcina-
tion temperature on the product yields in Friedel–Crafts alkylation
of benzyl alcohol with anisole. The main products were ortho- and
para-benzyl anisole isomers, which were presumably formed on
the Brønsted acid sites, and by-product was dibenzyl ether pro-
duced mostly on the Lewis acid sites [14,22–26]. The ratio of ortho-
to para-benzyl anisole was approximately 48:52 throughout the
reaction. The yield of benzyl anisole increased as the calcination
temperature increased to 1223 K. Increasing the calcination above
1223 K resulted in a reduction of yield of benzyl anisole. The yield
of dibenzyl ether was independent of the calcination temperature.
These results strongly suggest that Brønsted acid sites are gener-
ated on Ta2O5/Al2O3 by calcination at high temperatures such as
1223 K. No benzyl anisole was formed on tantalum acid and trace
of dibenzyl ether (0.02 mmol) was obtained under this condition.
Fig. 2(A) shows the effect of the calcination temperature on the
product yields in isomerization of ␣-pinene. The main products
were camphene and limonene. A small amount of terpinene, ter-
pinolene, fenchene, and ␤-pinene was formed as by-products. The
yield of camphene was almost twice to that of limonene regardless
of the calcination temperature. The change in yields of products has
the same tendency as that in yield of benzyl anisole in Friedel–Crafts
alkylation. The yields of camphene and limonene increased with
increasing calcination temperature up to 1223 K and the calcina-
tion at higher temperatures than 1273 K results in the reduction
of yields of these products. This result strongly suggests that cam-
phene and limonene were formed on Brønsted acid sites as well
as benzyl anisole in Friedel–Crafts alkylation. Tantalum acid gave a
small amount of camphene (0.14 mmol) and limonene (0.05 mmol),
respectively.
3.2. Brønsted acidity
To examine the generation of Brønsted acid sites on Ta2O5/Al2O3
calcined at 1223 K, the acid sites were characterized based on FT-
IR spectra of lutidine adsorbed on the catalysts. Fig. 3 shows the
The results of Friedel–Crafts alkylation over Ta2O5/Al2O3 solid
acids with various Ta2O5 loadings and calcined at 1223 K are shown
in Fig. 1(B). Fig. 2(B) demonstrates the effect of loading amount of
Ta2O5 on the yields of camphene and limonene in isomerization
of ␣-pinene. The calcination temperature was fixed at 1223 K. The
Fig. 3. IR spectra of lutidine adsorbed on (a) alumina, and (b) 33 wt% Ta2O5/Al2O3
calcined at 1223 K.

192
T. Kitano et al. / Catalysis Today 192 (2012) 189–196
Fig. 4. The correlation between Brønsted acidity and the yield of benzyl anisole.
Calcination temperature: 1223 K.
Fig. 5. XRD patterns of Ta2O5/Al2O3 catalysts calcined at various temperatures. (a)
773 K, (b) 973 K, (c) 1173 K, (d) 1223 K, (e) 1273 K, (f) 1323 K, (g) 1373 K, and (h)
1423 K. Ta2O5 loading: 33 wt%, (ꢀ) AlTaO4, (᭹) T-Ta2O5, (ꢁ) ␪-Al2O3 and (ꢀ) ␥-Al2O3.
IR spectra of lutidine adsorbed on 33 wt% Ta2O5/Al2O3 calcined at
1223 K and bare ␥-Al2O3. The bands at 1580 and 1608 cm−1 were
assignable to lutidine adsorbed on Lewis acid sites, and the bands
at 1628 and 1647 cm−1 can be assigned to lutidine adsorbed on
Brønsted acid sites [20,27,28]. In the case of 33 wt% Ta2O5/Al2O3,
in addition to the bands at 1580 and 1608 cm−1, the bands at 1628
and 1647 cm−1 appeared. This result clearly indicates that Brønsted
acid sites are generated on Ta2O5/Al2O3 calcined at such high tem-
perature in addition to Lewis acid sites. On the other hand, in the
caseof Al2O3, the bandsat 1580 cm−1 and 1608 cm−1 due to lutidine
adsorbed on Lewis acid sites appeared. The bands at 1632 cm−1 and
1635 cm−1 are assigned to hydrogen-bonding lutidine. Therefore it
can be though that only Lewis acid sites exist on Al2O3. This makes
a great difference to the evidence that tantalum acid itself calcined
at higher temperatures than 1073 K loses acid property, and that
tantalum acid exhibits mainly Lewis acid properties in the absence
of water [1,2].
Fig. 4 shows the correlation between the yield of benzyl anisole
and the Brønsted acidity, which was evaluated from FT-IR spectra
of adsorbed lutidine, of Ta2O5/Al2O3 calcined at 1223 K with var-
ious Ta2O5 loadings. The yield of benzyl anisole correlated with
the Brønsted acidity. The yield of benzyl anisole increased with
increasing the amount of Brønsted acid site. This indicates that the
Friedel–Crafts reaction is promoted on the Brønsted acid sites. As
shown in Figs. 1 and 2, the activities for Friedel–Crafts alkylation
and isomerization of ␣-pinene changed depending on the change
of calcinations temperature and loading amount of Ta2O5. The ten-
dencyof changein activityfor isomerizationof ␣-pineneagainstthe
calcination temperature or loading amount of Ta2O5 was almost
similar to that for the Friedel–Crafts alkylation. This indicates
that isomerization of ␣-pinene to camphene and limonene also
takes place on Brønsted acid sites. These results strongly suggest
that both calcination temperature and Ta2O5 amount affected the
amount of Brønsted acid site. This suggests that structural changes
around the tantalum species have a major impact on the Brønsted
acidic property.
Ta2O5/Al2O3 calcined at 1223 K exhibited the highest Brønsted
acidity and By contrast, pure Ta2O5 calcined at higher tempera-
tures (>1073 K, T-Ta2O5) does not exhibit both Brønsted and Lewis
acid property at all. This is consistent with the previous report [1,2].
3.3. XRD patterns and Raman spectra
Figs. 5 and 6 show the XRD patterns and Raman spectra of
33 wt% Ta2O5/Al2O3 calcined at various temperatures. Calcination
at up to 1223 K did not cause crystallization of tantalum oxide
(T-Ta2O5) or aluminum tantalate (AlTaO4). By contrast, the XRD
patterns and Raman spectra of Ta2O5/Al2O3 calcined at 1273 K or
above exhibited the formation of T-Ta2O5 and AlTaO4. Ushikubo
et al. reported that amorphous Ta2O5·nH2O crystallized to T-Ta2O5
after calcination at 1073 K [1,2]. Therefore, it appears that Al2O3
support stabilizes amorphous tantalum oxide cluster and prevents
it from crystallizing during calcination even at 1223 K. Corre-
sponding to the crystallization results, the specific surface area of
Ta2O5/Al2O3 very gradually decreased from 150 to 100 m2 g−1 with
an increase in calcination temperature up to 1223 K, and drastically
decreased to 50 m2 g−1 with calcination at higher temperatures
(1273–1423 K). In this region, the formation of AlTaO4 proceeded
and increasing the calcination temperature resulted in drastically
decreasing the yields of benzyl anisole, camphene and limonene.
AlTaO4 was completely inert for both Friedel–Crafts alkylation and
isomerization of ␣-pinene. Moreover, even after the exposure to
lutidine, IR spectrum of AlTaO4 showed no band due to adsorbed
lutidine. This result indicates that both Brønsted and Lewis acid
sites are absent on AlTaO4. These results strongly suggest that the
formation of inert AlTaO4 phase by calcination at high temperatures
(>1273 K) reduces the catalytic activity.
XRD patterns and Raman spectra of with various loadings cal-
cined at 1223 K are shown in Figs. 7 and 8. These results indicated
that T-Ta2O5 was not present in Ta2O5/Al2O3 catalysts with Ta2O5
Interestingly, Brønsted acid sites were still detected for the
Ta2O5/Al2O3 catalysts even after the calcination at 1223 K. 33 wt%

T. Kitano et al. / Catalysis Today 192 (2012) 189–196
193
Fig. 6. Raman spectra of Ta2O5/Al2O3 catalysts calcined at various temperatures.
(a) 773 K, (b) 973 K, (c) 1173 K, (d) 1223 K, (e) 1273 K, (f) 1323 K, (g) 1373 K, and (h)
1423 K. Ta2O5 loading: 33 wt%, (ꢀ) AlTaO4 and (ꢁ) TaOx.
Fig. 8. Raman spectra of Ta2O5/Al2O3 catalysts with various loadings. (a) 10 wt%,
(b) 20 wt%, (c) 30 wt%, (d) 33 wt%, (e) 35 wt%, (f) 40 wt%, and (g) 50 wt%. Calcination
temperature: 1223 K, (ꢁ) TaOx.
mass fraction of less than 33%. Above 35%, the larger the mass
fractions of Ta2O5, the increasing the intensity and the decreas-
ing of the full width at half maximum of the diffraction lines for
T-Ta2O5. These results indicate that the amorphous Ta2O5 sup-
ported on ␥-Al2O3 crystallized to T-Ta2O5 and the crystallite size
increased. When the Ta2O5 mass fraction was >35%, the activities
for Friedel–Crafts alkylation and isomerization of ␣-pinene were
greatly reduced. Friedel–Crafts alkylation and isomerization of
␣-pinene did not proceed over T-Ta2O5. Moreover, no band for
adsorbed lutidine was observed in the IR spectrum of T-Ta2O5
exposed to lutidine, indicating that both Brønsted and Lewis acid
sites are absent on T-Ta2O5. Based on these results, it can be thought
that the formation of inert T-Ta2O5 phase reduced the catalytic
activity by increasing the mass fraction of Ta2O5. Onfloy et al.
reported that Nb2O5/TiO2 and WO3/TiO2 exhibit Brønsted acid
property when crystallites of Nb2O5 and WO3 did not formed [29].
Recently, we also found that Nb2O5/Al2O3 exhibit Brønsted acid
property when crystallite of Nb2O5 did not formed [16,17]. Sim-
ilarly, XRD patterns and Raman spectra of Ta2O5/Al2O3 indicate
that amorphous Ta2O5 supported on Al2O3 exhibit Brønsted acid
property.
3.4. XPS measurements and analysis
The valence of tantalum cation of Ta2O5/Al2O3 catalyst sur-
face was evaluated by using XPS. Binding energy of Ta 4f peaks
of Ta2O5/Al2O3 was independent of Ta2O5 loadings (Fig. S1). The Ta
4f7/2 peak at 26.2 eV is assignable to Ta5+. Fig. 9 shows the surface
Ta/Al ratio estimated from the areas of Ta 4f and Al 2p XPS. The sur-
face Ta/Al ratio shows a linear increase with loading of Ta2O5 up to
33 wt% but then it becomes slightly gradual. This value (33 wt% of
Ta2O5) was slightly larger than the calculated value of formation of
Ta2O5 monolayer (28 wt%) based on the cross-section area of TaO6
octahedral unit and surface area of support. This suggests that the
surface of 33 wt% Ta2O5/Al2O3 was fully covered with monolayer
of tantalum oxide and that a small cluster of Ta2O5 coexists on
the surface. In the case of 33 wt% Ta2O5/Al2O3 catalyst, the sur-
face density of Ta atom was estimated to 5.8 Ta atoms nm−2. Bell
et al. [30] and Iwasawa et al. [16] reported that the surface den-
Fig. 7. XRD patterns of Ta2O5/Al2O3 catalysts with various loadings. (a) 10 wt%, (b)
20 wt%, (c) 30 wt%, (d) 33 wt%, (e) 35 wt%, (f) 40 wt%, and (g) 50 wt%. Calcination
temperature: 1223 K, (ꢀ) ␥-Al2O3 and (᭹) T-Ta2O5.
sity of Nb atom at monolayer coverage was 6.3 Nb atoms nm−2
Because ionic radius of Ta5+ is almost similar to that of Nb5+
.
,

194
T. Kitano et al. / Catalysis Today 192 (2012) 189–196
of tantalum atom is figured by a pre-edge peak area of Ta L1-
edge XANES due to 2s to 5d transition. This transition is the dipole
forbidden transition for regular octahedral site-symmetry such as
Sr2GaTaO6. The dipole forbidden transition is partially allowed by
the distortion of octahedral environment, because p orbitals are
mixed with d orbitals caused by the breakdown of inversion sym-
metry due to structure distortion [31]. Therefore, tantalum atom
with tetrahedral symmetry like YTaO4 exhibits relatively large pre-
edge peak area. In order to examine the local symmetry of tantalum
cation, the pre-edge peak areas at 11678 eV of Ta2O5/Al2O3 cat-
alysts were evaluated and summarized in Fig. 11. As shown in
Fig. 11, pre-edge peak areas of Ta2O5/Al2O3 catalysts were smaller
than YTaO4, which had tetrahedral symmetry, and larger than
Sr2GaTaO6 having regular octahedral symmetry. Furthermore, this
value of Ta2O5/Al2O3 catalyst was slightly larger than T-Ta2O5,
but almost similar to tantalum acid having distorted octahedral
symmetry, suggesting that TaO6 unit in Ta2O5/Al2O3 catalyst had
distorted octahedral symmetry like tantalum acid. This value was
almost constant up to 50 wt% of Ta2O5 loading.
Fig. 9. The change in surface Ta/Al ratio with loading amount. Calcination temper-
ature: 1223 K.
it seems that this surface density of Ta atom is almost similar
to the estimated value of 33 wt% Ta2O5/Al2O3 catalyst strongly
suggest that a monolayer of tantalum acid-like compound was
formed over 33 wt% Ta2O5/Al2O3. Combining this result with the
fact that the catalyst exhibited the highest activity at around 33 wt%
of loading, it appears that Brønsted acid site are generated on
a stabilized two-dimensional tantalum acid-like compound over
Al2O3.
3.6. Ta L3-edge EXAFS spectra and analysis
The Ta L3-edge EXAFS oscillations of Ta2O5/Al2O3 catalysts with
various loadings are shown in Fig. 12. In the high k region (>9 A−1),
˚
the amplitude of the EXAFS oscillations of Ta2O5/Al2O3 catalysts
with loading of Ta2O5 up to 30 wt% was weak. Above 35 wt%, the
amplitude of the EXAFS oscillations in this region gradually grew
with increasing the Ta2O5 loading and the period of the oscillation
was similar to that of T-Ta2O5. This indicates that the local struc-
ture of tantalum cation of Ta2O5/Al2O3 catalysts with loading of
Ta2O5 below 30 wt% was similar to that of tantalum acid, that is,
distorted octahedral TaO6 unit. On the other hand, above 35 wt%,
the amorphous Ta2O5 supported on ␥-Al2O3 crystallized to T-Ta2O5
gradually and the crystallite size increased.
3.5. Ta L1-edge XANES spectra and analysis
The coordination environment of tantalum in Ta2O5/Al2O3
catalysts was investigated by Ta L1 and L3-edges XAFS spectra.
Fig. 10 shows Ta L1-edge XANES spectra of reference samples and
Ta2O5/Al2O3 catalysts with various loadings. The local symmetry
Fig. 10. Ta L1-edge XANES spectra of (A) references and (B) Ta2O5/Al2O3 catalysts with various loadings. (a) Sr2GaTaO6, (b) T-Ta2O5, (c) Ta2O5 nH2O, (d) AlTaO4, (e) YTaO4,
(f) 10 wt%, (g) 20 wt%, (h) 30 wt%, (i) 33 wt%, (j) 35 wt%, (k) 40 wt%, and (l) 50 wt%. Calcination temperature: 1223 K.

T. Kitano et al. / Catalysis Today 192 (2012) 189–196
195
Fig. 11. The change in the pre-edge peak areas of Ta2O5/Al2O3 catalysts with
loadings. (a) Sr2GaTaO6, (b) T-Ta2O5, (c) Ta2O5 nH2O, (d) AlTaO4, and (e) YTaO4.
Calcination temperature: 1223 K.
Fig. 13. Fourier transforms of k3-weighted Ta L3-edge EXAFS of references and
Ta2O5/Al2O3 catalysts with various loadings. (a) Ta2O5 nH2O, (b) 10 wt%, (c) 20 wt%,
(d) 30 wt%, (e) 33 wt%, (f) 35 wt%, (g) 40 wt%, (h) 50 wt%, and (i) T-Ta2O5. Calcination
temperature: 1223 K.
Fourier Transforms of Ta L3-edge EXAFS are shown in Fig. 13. The
˚
˚
peak in the range 1-2 A due to Ta–O shell and the peak around 3 A
due to Ta–O–Ta (or Ta–O–Al) were appear. When the loading was
higher than 35 wt%, the amplitude of the second was stronger than
those of catalysts with loading up to 30 wt%. These results strongly
suggest that T-Ta2O5 over catalysts with higher loading than 35 wt%
has a three-dimensional structure, whereas TaO6 unit over cata-
lysts loading below 33 wt% has mainly a two-dimensional structure.
These results agree with the results of XPS measurement. Accord-
ing to the results mentioned above, it appears that the monolayer of
tantalum acid-like is stabilized over the Ta2O5/Al2O3 catalysts with
around 33 wt% Ta2O5 loading. In this monolayer, TaO6 unit having
distorted octahedral symmetry forms a two-dimensional Ta–O–Ta
network. This Ta–O–Ta network may have an important role in the
generation of Brønsted acid sites.
4. Conclusion
We found that the Ta2O5/Al2O3 catalysts calcined at high
temperatures were effective for Friedel–Crafts alkylation and iso-
merization of ␣-pinene. Among the Ta2O5/Al2O3 catalysts tested,
33 wt% Ta2O5/Al2O3 calcined at 1223 K showed the highest activity
in both Frieadel–Crafts Alkylation and isomerization of ␣-pinene.
The Ta2O5/Al2O3 catalyst exhibits the Brønsted acidity despite cal-
cination at such high temperatures. It is concluded that a monolayer
of tantalum acid-like compound, which have distorted octahedral
symmetry, were stabilized over 33 wt% Ta2O5/Al2O3 catalyst cal-
cined at 1223 K. The stabilized tantalum acid-like compound which
has two-dimensional Ta-O-Ta network probably accounts for the
generation of Brønsted acid. The formation of an inert AlTaO4 and
T-Ta2O5 phases resulted in the decreasing the Brønsted acidity
Fig. 12. k3-weighted Ta L3-edge EXAFS oscillations of references and Ta2O5/Al2O3
catalysts with various loadings. (a) Ta2O5 nH2O, (b) 10 wt%, (c) 20 wt%, (d) 30 wt%, (e)
33 wt%, (f) 35 wt%, (g) 40 wt%, (h) 50 wt%, and (i) T-Ta2O5. Calcination temperature:
1223 K.

196
T. Kitano et al. / Catalysis Today 192 (2012) 189–196
Acknowledgements
[9] A. Corma, Chem. Rev. 97 (1997) 2373–2419.
[10] A. Corma, Chem. Rev. 95 (1995) 559–614.
[11] K. Arata, H. Matsuhashi, M. Hino, H. Nakamura, Catal. Today 81 (2003) 17–30.
[12] N. Katada, H. Ishiguro, K. Muto, M. Niwa, Chem. Vap. Deposition 1 (1995) 54–60.
[13] A. Takagaki, M. Sugiura, D. Lu, J.N. Kondo, M. Hara, K. Domen, S. Hayashi, J. Am.
Chem. Soc. 125 (2003) 5479–5485.
[14] C. Tagusagawa, A. Takagaki, S. Hayashi, K. Domen, J. Am. Chem. Soc. 130 (2008)
7230–7231.
[15] A. Takagki, C. Tagusagawa, S. Hayashi, M. Hara, K. Domen, Energy. Environ. Sci.
3 (2010) 82–93.
[16] T. Shishido, T. Kitano, K. Teramura, T. Tanaka, Catal. Lett. 129 (2009) 383–386.
[17] T. Shishido, T. Kitano, K. Teramura, T. Tanaka, Top. Catal. 53 (2010) 672–677.
[18] T. Kitano, S. Okazaki, T. Shishido, K. Teramura, T. Tanaka, Chem. Lett. 40 (2011)
1332–1334.
[19] O. Yamaguchi, D. Tomihisa, T. Uegaki, K. Shimizu, J. Am. Ceram. Soc. 70 (1987)
C335–C338.
[20] T. Onfroy, G. Clet, M. Houalla, Micropor. Mesopor. Mater. 82 (2005) 99–104.
[21] A.L. Ankudinov, B. Ravel, J.J. Rehr, S.D. Conradson, Phys. Rev. B 58 (1998)
7565–7576.
This work was supported in part by a Grant-in-Aid for Scientific
Research (B) (Grant No. 23360355) and Grant-in-Aid for Young Sci-
entists (B) (Grant No. 21760627) from the Ministry of Education,
Culture, Sports, Science and Technology and by the Iwatani Naoji
Foundation (Grant No. 09-011). The X-ray absorption experiments
have been performed under the approval of the JASRI (Proposal Nos.
2009B1398 and 2010B1184).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.cattod.2012.02.035.
[22] K. Okumura, K. Yamashita, M. Hirano, M. Niwa, Chem. Lett. 34 (2005) 716.
[23] K. Okumura, K. Yamashita, M. Hirano, M. Niwa, J. Catal. 234 (2005) 300.
[24] K. Yamashita, M. Hirano, K. Okumura, M. Niwa, Catal. Today 118 (2006) 385.
[25] M.H.C. de la Cruz, J.F.C. da Silva, E.R. Lachter, Catal. Today 118 (2006) 379.
[26] M.H.C. de la Cruz, M.A. Abdel-Reim, J.F.C. da Silva, A.C. Faro Jr., E.R. Lachter,
Catal. Commun. 8 (2007) 1650.
[27] C. Morterra, G. Cerrato, G. Meligrana, Langmuir 17 (2001) 7053–7060.
[28] T. Onfroy, G. Clet, M. Houalla, J. Phys. Chem. B 109 (2005) 14588–14594.
[29] T. Onfroy, G. Clet, S.B. Bukallah, T. Visser, M. Houalla, Appl. Catal. A. 298 (2006)
80–87.
References
[1] T. Ushikubo, K. Wada, Chem. Lett. (1988) 1573–1574.
[2] T. Ushikubo, K. Wada, Appl. Catal. 67 (1990) 25–38.
[3] T. Iizuka, K. Ogasawara, K. Tanabe, Bull. Chem. Soc. Jpn. 56 (1983) 2927–
2931.
[4] K. Tanabe, Catal. Today 78 (2003) 65–77.
[5] K. Tanabe, S. Okazaki, Appl. Catal. A: Gen. 133 (1995) 191–218.
[6] I. Nowak, M. Ziolek, Chem. Rev. 99 (1999) 3603–3624.
[7] T. Okuhara, Chem. Rev. 102 (2002) 3641–3665.
[8] T. Ohuchi, T. Miyatake, Y. Hitomi, T. Tanaka, Catal. Today 120 (2007)
233–239.
[30] T. Tanaka, T. Yoshida, H. Yoshida, H. Aritani, T. Funabiki, S. Yoshida, J.M. Jehng,
I.E. Wachs, Catal. Today 28 (1996) 71–76.
[31] T. Yamamoto, X-ray Spectrom. 37 (2008) 572–584.