Acylation of anisole with carboxylic acids catalyzed by tungsten oxide supported on titanium dioxide

Article abstract of DOI:10.1016/j.mcat.2019.110410

Friedel-Crafts (F-C) acylation of anisole with octanoic acid was carried out on tungsten oxide (WO₃) supported on various types of oxide supports. We have found that the highest activity was obtained when TiO₂ was used as the support. WO₃/TiO₂ was found to be active in the acylation of anisole with carboxylic acids of varying alkyl chain lengths (C6–C10). It was possible to recycle the WO₃/TiO₂ catalyst for up to 5 times without deactivation. The turnover frequency (TOF) of the catalyst was closely correlated with the electronegativity of the cation of the support used for WO₃. When a strong basic oxide such as CeO₂ was used as a support, the acid strength of WO₃ was diminished, while the strong acidity of WO₃ was retained on a weak basic support like TiO₂. This explains why the acid strength and consequently, the activity, were found to be the highest for the WO₃/TiO₂ catalyst. The trend of the catalytic performance was consistent with the order of acid strength of WO₃ on different supports measured by temperature-programmed desorption of NH₃.

Full text of DOI:10.1016/j.mcat.2019.110410

Molecular Catalysis 475 (2019) 110410
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.elsevier.com/locate/mcat
Acylation of anisole with carboxylic acids catalyzed by tungsten oxide
supported on titanium dioxide
⁎
Kazu Okumura , Masaki Iida, Hajime Yamashita
Department of Applied Chemistry, School of Advanced Engineering, Kogakuin University, 2665-1 Nakano-machi Hachioji-city, 192-0015, Tokyo, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
3
Friedel-Crafts (F-C) acylation of anisole with octanoic acid was carried out on tungsten oxide (WO ) supported
Friedel-crafts acylation
Octanoic acid
Tungsten oxide
Titanium dioxide
on various types of oxide supports. We have found that the highest activity was obtained when TiO2 was used as
the support. WO /TiO was found to be active in the acylation of anisole with carboxylic acids of varying alkyl
chain lengths (C6–C10). It was possible to recycle the WO /TiO catalyst for up to 5 times without deactivation.
The turnover frequency (TOF) of the catalyst was closely correlated with the electronegativity of the cation of
the support used for WO . When a strong basic oxide such as CeO was used as a support, the acid strength of
WO was diminished, while the strong acidity of WO was retained on a weak basic support like TiO
explains why the acid strength and consequently, the activity, were found to be the highest for the WO
catalyst. The trend of the catalytic performance was consistent with the order of acid strength of WO
ferent supports measured by temperature-programmed desorption of NH
3
2
3
2
3
2
3
3
2
. This
/TiO
on dif-
3
2
3
3
.
1. Introduction
However, new insight into the origin of catalytic activity should be
obtained by systematically correlating the catalytic performance and
Tungsten oxide (WO
and acid catalyst over the past several decades [1]. WO
3
) has been used as a photocatalyst, gas sensor,
-based catalysts
having acidic properties can be classified into several groups. One
group includes WO crystals and those doped with foreign elements in a
nanowire structure, which are used as catalysts and gas sensors [2]. We
have previously reported that WO doped with Nb, prepared by co-
precipitation and hydrothermal methods, is active in the Friedel-Crafts
FeC) reaction [3,4]. Another group of WO -based catalysts, i.e. over-
layer structures or clusters of WO supported on high-surface-area
the acidic property of WO
this purpose, four types of supports—CeO
TiO —were used in this study. FeC acylation using a carboxylic acid
was carried out as the catalytic reaction over WO -loaded catalysts.
FeC alkylation and acylation have been extensively studied using acid
anhydrides and acid chlorides as acylating agents [13]. The use of these
reagents result in the formation of carboxylic acid or HCl as by-pro-
ducts, which is undesirable from both an environmental and economic
point of view. Unlike these, the use of carboxylic acids as acylating
agents should be ideal, forming ketone, and giving out water as the only
by-product [14]. Therefore, herein we performed a representative FeC
acylation using octanoic acid and anisole (1).
3
loaded on different kinds of supports. For
3
2
,
ZrO Al and
2
,
2 3
O
2
3
3
3
(
3
3
oxides have been found to be useful as photocatalysts [5], catalysts for
alkene isomerization [6], dehydration of alcohols [7], and glucose de-
hydration to 5-hydroxymethylfurfural [8]. In the domain of WO -sup-
3
ported catalysts, it is widely accepted that acid properties emerge by
the coating of the metal oxides on basic oxide supports [9]. However,
the genesis of the catalytic activity of supported WO
pletely understood. Wachs et al. reported that WO clusters supported
on the surface showed activity for partial oxidation of methanol [10].
Iglesia et al. suggested that proton evolution occurred in the WO upper
layer on ZrO in a manner similar to heteropolyanion formation [11].
Tanaka et al. proposed that a Brønsted acid site is generated at the
boundary between the WO upper layer domains [12]. Previous re-
3
is not still com-
x
(1)
In addition, by replacing a homogeneous catalyst like aluminum
chloride with a heterogeneous catalyst, it was possible to reuse the
catalyst and realize an environmentally friendly reaction system. The F-
C acylation have been achieved by sulfonic acid functionalized on
mesoporous materials [15]. Another synthetic route to obtain aromatic
ketones is through the heterogeneous photocatalytic oxidation of
x
2
3
x
search was focused mainly on WO supported on specific supports.
⁎
Corresponding author.
E-mail address: okmr@cc.kogakuin.ac.jp (K. Okumura).
https://doi.org/10.1016/j.mcat.2019.110410
Received 25 March 2019; Received in revised form 14 May 2019; Accepted 17 May 2019
2468-8231/ © 2019 Elsevier B.V. All rights reserved.

K. Okumura, et al.
Molecular Catalysis 475 (2019) 110410
Table 1
Physical properties and Friedel-Crafts acylation of anisole with octanoic acid catalyzed by WO
3
loaded on various kinds of supports.a
Entry Support
WO
3
loading / wt% Acid amount / mol
BET surface area before the loading of WO
3
/
BET surface area after the loading of WO
m g
3
/
Yield / %c
−
1b
2
−1
2 −1
kg
m
g
1
2
3
4
5
6
MgO
Al
5
23
9
0.36
0.26
0.18
0.48
0.26
0.18
311
175
50
105
120
175
202
57
34
59
79
77
0
7
80
86
19
0
2
O
3
TiO
TiO
2
(rutile)
2
(anatase) 18
ZrO
CeO
2
17
19
2
a
The amount of WO
3
charged was adjusted to be 4.5 nm−2 at the time of preparation. The reaction was performed using 10 g of anisole and 2 mmol of octanoic
acid in an oil bath at 413 K for 6 h. The mole ratio of anisole and octanoic acid was 46.2. The amount of catalyst used for reaction was adjusted so that the WO
3
content was 0.018 g.
b
Measured with NH
Calculated based on the octanoic acid (2 mmol) used for reaction.
3
TPD.
c
aromatic alcohols [16]. In this study, we compared the catalytic activity
of WO supported on various kinds of supports, for F-C reactions.
Furthermore, the catalytic performance of WO supported on TiO
which exhibited the highest activity, was examined in detail, and the
supports for WO
3
-loaded catalysts are also included in Tables 1 and 2.
3
Unless otherwise stated, the anatase-type (TIO-16; Tables 1 and 2, entry
4) TiO was used as the support for TiO . Zeolite-β(HSZ-920HOA) and
ZrO (RC-100) were supplied by Tosoh Co. and Daiichi Kigenso Kagaku
3
2
,
2
2
2
origin of the enhanced catalytic activity of WO
tion is discussed.
3
/TiO
2
in the F-C reac-
Kogyo Co., respectively. The ammonium paratungstate solution was
evaporated on a water bath with continuous stirring in the presence of
the oxide supports. The resulting powder was heat treated in a furnace
in air at 773 K for 2 h. The color of the WO
air, changed from white to yellow as the amount of tungsten oxide
3 2
/TiO samples, calcined in
2
. Experimental
−
2
increased. The amount of WO
the time of preparation, unless otherwise stated (Table 1). This corre-
sponds to the stipulated amount to ensure a monolayer of WO ac-
cording to the reported cross section of W on Al (0.22–0.245 nm /
3
charged was adjusted to be 4.5 nm at
2.1. Sample preparation
3
The supported WO
pregnation method. Typically, a solution of ammonium paratungstate,
NH 42·4H O (supplied by Wako Pure Chemical Industries,
Ltd.) dissolved in 60 ml of water was immersed in the oxide support.
The oxides used as support for WO and their physical properties are
listed in Tables 1 and 2. MgO (MGO-4), Al (ALO-7), TiO (TIO-4,
16), and CeO (CEO-2) were obtained from the Catalysis Society of
3
catalyst was prepared by the conventional im-
2
2 3
O
W). [17]. In addition, samples in which the loading amount of tungsten
oxide fixed be to 18% by weight were also prepared (Table 2). In the
(
4
)
10
H
2
W
12
O
2
case where TiO
in the range from 7 to 100 wt% (pure WO
2
was used as a support, the loading of WO
).
3
was varied
3
3
2
O
3
2
-
2
Japan. Physical properties (acid amount and surface area) of these
2.2. Characterization
Table 2
X-ray diffraction (XRD) of the obtained sample was measured using
Physical properties and Friedel-Crafts acylation of anisole with octanoic acid
catalyzed by 18 wt%-WO loaded on various kinds of supports and zeolite β.
a
a MiniFlex (Rigaku) X-ray diffractometer (Cu Kα source; 2θ = 10–70°).
X-ray photoelectron spectroscopy (XPS) data were collected with an Al
anode, using a Quantum 2000 (Ulvac Phi) instrument with a beam
diameter of 100 μm and an accelerating voltage of 15 kV. Synchrotron
radiation experiments were conducted at BL01B1 with the approval of
the Japan Synchrotron Radiation Laboratory (JASRI, Proposal number:
3
2
013B1067). A Si (111) single crystal was used to obtain a mono-
chromatic X-ray beam. Measurements were made in quick mode. In
order to collect W L - and L -edge X-ray absorption fine structure
XAFS) data, ionization chambers filled with N and N (50%)/Ar
and I, respectively. Data were analyzed using the
-edge extended X-ray
Entry Support
Acid amount / BET surface area
Yield / %c
−1b
mol kg
after the loading of
WO / m g
3
2
−1
1
3
(
2
2
1
2
3
4
5
6
7
MgO
Al
0.32
0.31
0.20
0.48
0.14
0.13
0.58
149
118
34
59
79
0
(50%) were used as I
0
2
O
3
10
48
86
34
1
REX 2000 program (Rigaku). For analysis of W L
absorption fine structure (EXAFS), Fourier transform of the k χ (k) data
was performed in the k range from 25 to 150 nm
3
TiO
TiO
2
(rutile)
(anatase)
3
2
−1
.
ZrO
CeO
zeolite β (H-type,
2
2
77
405
The acid properties of the prepared WO /TiO catalyst were mea-
3
2
34
sured by studying the temperature-programmed desorption (TPD) of
ammonia using BelCatII (Microtrac Bel Co.). Prior to measurement, the
sample was evacuated at 773 K under He flow. Ammonia (13.3 kPa),
diluted with He, was equilibrated with the pretreated sample at 373 K.
TPD data was collected at a temperature ramp rate of 10 K/min. A
thermal conductivity detector was used for the quantitative determi-
d
Si/Al
2
= 20)
a
The reaction was performed using 10 g of anisole and 2 mmol of octanoic
acid with 0.1 g of catalysts in an oil bath at 413 K for 6 h. The mole ratio of
anisole and octanoic acid was 46.2.
b
Measured with NH
Calculated based on the octanoic acid (2 mmol) used for reaction.
3
TPD.
c
nation of desorbed NH
recorded using a BELSORP-mini (Microtrac BEL Co.) instrument.
3 3 2 2
from WO /TiO . N adsorption isotherms were
d
WO is not loaded on zeolite β.
3
2

K. Okumura, et al.
Molecular Catalysis 475 (2019) 110410
Fourier-transform infrared (FTIR) spectra were recorded with a re-
−1
solution of 1 cm
Co.). Before pyridine was adsorbed, the sample was heat treated at
73 K in air. The treated sample was immersed in a toluene solution of
using a Spectrum One spectrometer (PerkinElmer
7
pyridine (Wako co.) at room temperature and subsequently rinsed with
hexane. The dried samples were diluted with potassium bromide and
pelleted for determination of FTIR.
2.3. F-C acylation
F-C acylation of anisole, with octanoic acid as acylating agent, was
carried out using 10 g anisole and 2 mmol octanoic acid in an oil bath at
13 K with Ar bubbling (30 mL/min), unless otherwise stated. The re-
4
agents were obtained from Tokyo Chemical Industry Co. When the
temperature of the solution reached 413 K, the catalyst was charged to
the flask. For recycling, spent catalyst was separated by filtration fol-
lowed by washing with toluene. After drying at 343 K, the remaining
solid was heat-treated in a furnace in 773 K air to remove adsorbed
species before recycling. The product was analyzed using a gas chro-
matograph (GC-2010, Shimadzu) equipped with a capillary column
(
InertCap1, GL-Science Co.) and a Flame Ionization detector (FID). In
the analysis, tridecane was used as an internal standard. Formation of
1
4
-methoxyoctanophenone was confirmed by H NMR (Fig. S1). The 1H
NMR spectra were recorded on a JEOL JNM-ECZ400S spectrometer
JEOL, Tokyo, Japan) in CDCl . Upon completion of the reaction, the
(
3
solvent was removed in vacuo and the crude product was purified by
column chromatography on silica gel using hexane and ethyl acetate
(
9:1) to give 4-methoxyoctanophenone. The F-C reaction using hex-
anoic acid, heptanoic acid, and decanoic acid (Tokyo Chemical Industry
co.) was carried out in the same manner as the reaction using octanoic
Fig. 1. XRD patterns of (a) WO
3
loaded on various supports, as mentioned in
acid. The resulting mixture was heated at 413 K in N
2
atmosphere and
Table 1, and (b) WO /TiO with different loadings of WO . ▽ and • indicate the
3
2
3
monitored by thin-layer chromatography (TLC) and GC.
3
diffraction due to the monoclinic and hexagonal phases of the WO crystal,
respectively.
3. Results and discussion
3
.1. XRD patterns and W L - and L -edge XAFS
1
3
Fig. 1(a) shows the X-ray diffraction (XRD) patterns of supported
WO /metal-oxides in comparison with those of pure WO crystals and
bare metal oxide supports. No diffraction assignable to crystalline WO
was observed on CeO and ZrO , suggesting that WO is well-dispersed
over the entire surface of these oxides. For WO /Al and WO /TiO
a small diffraction assignable to monoclinic WO appeared in the range
3–24° [18]. This suggests that slightly aggregated WO is formed on
the surface of Al and TiO . Fig. 1(b) focusses on XRD patterns of
, for different addition amounts of WO . The diffraction of
at a WO loading amount of less than 14% by weight is in
. On increasing the loading
amount to more than 18% by weight, the peak attributable to the
monoclinic WO appears, most likely due to the beginning of WO
aggregation. As shown in Fig. 1(b), the small peak assignable to the
100) plane of hexagonal WO is also observed in the 78 wt%-WO
loaded TiO
W L - and L
3
3
3
2
2
3
3
2
O
3
3
2
,
3
2
3
2
O
3
2
WO
WO
3
/TiO
/TiO
2
2
3
3
3
good agreement with that of anatase TiO
2
Fig. 2. W L
for different loadings of WO3.
3 3 2
-edge EXAFS radial distribution functions of WO loaded on TiO
3
3
(
3
3
2
.
1
3
-edge XAFS have been frequently used to provide in-
formation on local symmetry, coordination, and valency of W oxides
supported on metal oxides [19]. Fig. 2 shows radial distribution func-
3 2 3
tions (Fourier-transformed) of WO /TiO for different loadings of WO ,
3
heat treated at 773 K. The corresponding k χ(k) data are displayed in
Fig. S2. The peak appearing at 0.14 nm in FT of each sample in Fig. 2
could be attributed to the WeO bond (phase shift uncorrected)
straightaway. The peak that appears between 0.24 and 0.30 nm in
WO
3 2 3
/TiO , with WO loading of less than 18 wt%, may be assignable to
the mixture of W-O-Ti and WeOeW bonds, suggesting a strong inter-
action between WO
WO loading higher than 37 wt% was close to that of WO
formation of crystalline monoclinic WO
3
and TiO
2
[20,21]. The EXAFS of WO
3
/TiO
due to the
2
with
Fig. 3. W-L
ferent loadings of WO
1 2 4 3 3 2
-edge XANES of Na WO , monoclinic WO and WO /TiO for dif-
3
3
3
.
3
, as confirmed by the XRD
3

K. Okumura, et al.
Molecular Catalysis 475 (2019) 110410
3 3
Fig. 4. Raw (solid lines) and simulated (dotted lines) NH -TPD curves of WO
loaded on different kinds of supports WO loaded on various supports, as
3
mentioned in Table 1.
Fig. 5. The dependence of the acid content of WO
on TiO measured with NH -TPD.
3 2 3
/TiO on the loading of WO
2
3
data. The shape of X-ray absorption near edge structure (XANES) did
not change relative to the loading amount of WO (Fig. 3). The intensity
of the pre-edge peak appearing at 12,100 eV in W L -edge XANES re-
mains the same for different loading amounts. The shape is closer to
that of monoclinic WO , while the intensity of the pre-edge peak is
much smaller than that of Na WO with tetrahedral symmetry, sug-
gesting that the local symmetry around W is a distorted octahedron
22].
3
1
3
2
4
[
3
.2. NH
3
-TPD and N
2
adsorption isotherms
3
-TPD curves for WO loaded on different
Fig. 4 shows the NH
3
supports corresponding to the samples listed in Table 1. The broad
peaks appear in the temperature range between 400 and 800 K, which
could be deconvoluted into three peaks, as shown by the dotted lines in
the figure. The peak at around 470 K may be attributed to weakly ad-
3 3 2
Fig. 6. BET surface area plotted as a function WO loading in WO /TiO .
sorbed NH
arises from NH
shifted depending on the support used. The suggested order of acid
strength is as follows: WO /CeO < WO /Al < WO
ZrO < WO /TiO . The amount of acid was measured based on deso-
rption of NH at temperature higher than 470 K, corresponding to
3
[23], but the broad peak that appears at higher temperature
3
adsorbed at the acidic site of WO . The broad peaks
3
3
2
3
2
O
3
3
/
2
3
2
3
where the peak appears. The values, listed in Table 1, will be utilized
for calculation of TOF, as will be mentioned later.
Next, NH
different loading amounts of WO
3
-TPD of WO
3
/TiO
2
was measured for samples having
(Fig. S3), and the amount of acid
3
measured by the TPD curve was measured as a function of the loading
of WO on TiO (Fig. 5). A steep increase in the acid amount is observed
for WO loading of less than 14 wt%, probably owing to the even
spreading of WO on the surface of TiO . When the loading amount of
WO is further increased, the acidic part gradually decreases. The
3
2
3
3
2
3
−
1
highest acid content was obtained with 18 wt% (0.48 mol kg ). This
agrees with the trend published in the literature that the highest acid
amount is obtained in the vicinity of the upper layer of WO
on ZrO [24]. The decrease in acid content may be caused by the ag-
gregation of WO , as confirmed by XRD (Fig. 1(b)), resulting in the
decrease in surface area.
Fig. S4 shows the N
unloaded TiO and 18 wt% WO
surface area was calculated based on the isotherms. The BET surface
3
supported
2
3
Fig. 7. FTIR spectra of pyridine adsorbed on WO loaded on different kinds of
supports, as mentioned in Table 1.
3
2
adsorption isotherms, classified as type II, for
/TiO . Brunauer-Emmet-Teller (BET)
3.3. FTIR Spectra of adsorbed pyridine
2
3
2
2
−1
Fig. 7 shows the IR spectrum of pyridine adsorbed on WO
3
sup-
area gradually decreased from 88 to 25 m g with an increase in the
amount of the added WO from 0 (TiO ) to 78 wt% (Fig. 6). This trend
may be due to the formation of aggregated WO , as confirmed by XRD.
ported on the selected supports. The broadband for pyridine species
3
2
−
1
adsorbed at the Brønsted acid site typically appears at 1530 cm
,
3
while that for the corresponding Lewis acid site should appear at ca.
4

K. Okumura, et al.
Molecular Catalysis 475 (2019) 110410
reports the % yield of the F-C alkylation for WO
3
loaded on different
types of supports in which the amount of catalyst used for reaction was
adjusted so that the WO
charged in Table 1 was adjusted to be 4.5 nm
3
content was 0.018 g. The amount of WO
3
−
2
at the time of pre-
paration. A particularly high yield was obtained when TiO
a support. A similar tendency was observed in the catalytic performance
of the 18 wt%-WO loaded samples (Table 2). In the case of WO /MgO,
2
was used as
3
3
the reaction solution became viscous. Analysis by GC revealed that
octanoic acid was completely consumed after the reaction even though
4
-methoxyoctanophenone was not detected in the solution. Probably
octanoic acid reacted with MgO to produce magnesium caprylate. The
yield (86%; Table 1 and 2, entry 4) of WO /TiO was much higher than
3
2
that obtained with zeolite β (32%; Table 2, entry 7), which has been
recognized as a promising catalyst for the reaction [27,28]. There was
no appearance of an induction period in the reaction using WO
Fig. S8). The yield of product obtained at 6 h with WO loaded on TIO-
6 with anatase structure (86%) was close to that of WO on TIO-4 with
3 2
/TiO
(
3
1
3
rutile structure (80%, Table 1, entry 3). It can be concluded that the
crystal structure of the support had little influence on the support
structure, which can also be confirmed from the fact that the curves for
yield vs. time of reaction overlap (Fig. S8). The yield of 4-methox-
3
Fig. 8. The relationship between the loading of WO and the ratio of the W-4d
and Ti-2p peak areas of WO /TiO determined by XPS.
3 2
yoctanophenone for WO
Table 1, entry 5). Since the catalyst has been reported to be active in
the acylation of acetic acid with toluene, this seems rather unexpected
29].
The yield of 4-methoxyoctanophenone, measured with GC, plotted
as a function of the amount of WO loaded onto the TiO support is
shown in Fig. 9. Data was collected at 2 h and 6 h after the start of the
reaction. A sharp increase in yield was observed for an increase in the
3
loading amount from 0 to 20 wt% of WO . On further increasing the
3 2 3 2
/ZrO was much lower than that of WO /TiO
−
1
−1
1
450 cm . The intense peak observed at 1530 cm for the employed
(
−
1
samples, while the missing peak at 1450 cm , regardless of the cata-
lyst used, except for WO /Al , shows that Brønsted acid sites were
predominant in WO -loaded samples. FTIR spectra of pyridine adsorbed
on WO /TiO for different loadings of WO are shown in Fig. S5.
Brønsted acid sites were predominant here as well, regardless of the
loading amounts of WO . The peak assigned to adsorbed pyridine did
not appear on bare TiO , suggesting that the Brønsted acid site was
generated only after the addition of WO
3
2 3
O
[
3
3
2
3
3
2
3
2
3
.
loading, the yield gradually decreased. The highest yield (90%) was
obtained at ca. 20% by weight. On plotting the yield at 2 h as a function
3
.4. XPS
of the amount of WO , the highest value was obtained for 26 wt%. This
3
is slightly higher than the value corresponding to the upper layer
coverage of WO3 (18 wt%), and diffraction attributable to crystalline
To obtain information on the valence state of W exposed to the
surface, XPS was measured using a sample loaded with WO . The
3
WO was clearly observed in XRD (Fig. 1(b)) for loadings above 18 wt
3
photoelectron energy of the peak appearing in XPS of W 4d showed that
the valence of W was 6+, regardless of the type of support (Fig. S6(a)).
%. Slightly crystallized WO is therefore suggested as a catalyst for the
F-C reaction.
3
On the other hand, in the case of WO
amount of WO , the binding energy of Ti 2p peaks shifts to a higher
value (Fig. S6(b)). This could be because of the slight change in the
electronic state of Ti after the addition of WO , although the reason was
not clearly understood at this stage. Fig. 8 shows the peak area ratio of
W 4d and Ti 2p plotted as a function of the amount of WO added. This
ratio increased almost linearly as the content of WO increased to 57%
3
/TiO
2
, with an increase in the
To study the recyclability of the catalyst, the spent catalyst was
collected by filtration and subsequently rinsed with toluene. The ob-
tained solid was then heat-treated in air at 773 K for 3 h to remove
carbonaceous deposits. No deactivation was observed even when used
at least 5 times, as shown in Fig. 10.
3
3
3
Catalytic reactions using anisole and carboxylic acids with different
3
by weight. Unlike the previous report regarding the relationship be-
tween the peak area of Raman spectra and the surface density of W in
WO
WO
3
/ZrO
/TiO
2
[25], no clear reflection point was observed in the case of
, suggesting that the obvious transformation from 2D to 3D
3
2
did not take place, even after the generation of WO
loading of WO . When the loading of WO reached 78% by weight, a
sharp increase in the ratio was observed, probably due to the onset of
vigorous aggregation of WO . This fact agrees with the appearance of
diffraction assignable to the aggregated WO in the XRD pattern of the
samples with a larger loading amount than 78 wt% (Fig. 1(b)).
3
clusters with low
3
3
3
3
3.5. F-C acylation
3
Regardless of the WO supported catalyst used in the reaction be-
tween octanoic acid and anisole, para-substituted ketones were ob-
tained with a selectivity higher than 97%, with small amounts of me-
thyl octanoate and phenyl octanoate as byproducts. The reaction
proceeded similar to that on a self-assembly of Nb-W nanofibers [26].
The reaction was performed with the use of different mol ratios of
Fig. 9. Yield of 4-methoxyoctanophenone plotted as a function of WO
on TiO . Reaction time: 2 h (○) and 6 h (•). The yield was calculated based on
the quantity of octanoic acid (2 mmol) used for reaction.
3
loading
anisole and octanoic acid over WO
3
2
/TiO . The yield decreased mono-
2
tonically accompanied by decrease in the molar ratio (Fig. S7). Table 1
5

K. Okumura, et al.
Molecular Catalysis 475 (2019) 110410
3 2
linear alkyl chain lengths were carried out at 18 wt% WO /TiO to
explore the scope of the catalyst. We see that the yield is not dependent
on the chain length of the carboxylic acid, which indicates the gen-
erality of the application of the WO
.6. Proposed mechanism for the high catalytic activity of WO
Based on the data on catalyst performance and NH -TPD, the TOF
3
/TiO
2
catalyst (Table 3).
3
3
/TiO
2
3
was plotted as a function of the electronegativity of the cation of the
corresponding oxide support (Fig. 11). For the calculation of TOF, the
molar amount of the product (4-methoxyoctanophenone) was divided
3 3
by the desorption amount (mol) of NH measured by NH -TPD. Elec-
tronegativity was recognized as being indicative of acid or base prop-
erties of metal oxides. Values for metal cations are expressed in Pauling
and were derived from the corresponding values for the elements using
eq. 2 [30]:
x
i
0
= (1 + 2i) x (2)
Fig. 10. Yield of 4-methoxyoctanophenone in the repeated reaction between
Here, i and x indicate the valence and electronegativity of the element,
0
octanoic acid and anisole over 18 wt%-WO
cycle use, the catalyst was washed with toluene, dried and calcined in air at
73 K.
3 2
/TiO . Reaction time: 6 h. For re-
respectively [9,31]. In other words, the smaller the electronegativity of
oxides, the higher the basicity, and vice versa. Accordingly, the NH
TPD peak shifted to a higher temperature with an increase in x . TOF
showed a sharp increase with increasing electronegativity of the sup-
port, as given in Fig. 11. As a result, WO /TiO exhibited the highest
TOF. It is proposed that the low activity of WO /CeO was caused by
the neutralization or weakening of the acidic sites of WO , due to the
electron donating effect of CeO , which had the highest basicity among
the employed supports. In contrast, because the base strength of TiO is
the weakest, it resulted in the highest activity of WO /TiO among the
catalysts used, so that the inherent acid properties of WO were re-
tained to a greater extent on a TiO support. According to the density
functional theory calculations in the literature, protons prefer the
bridging oxygen sites of corner-sharing WO , suggesting that these
3
-
7
i
Table 3
3
2
Friedel-Crafts acylation of anisole with carboxylic acids with different linear
3
2
a
3 2
alkyl chain lengths catalyzed by 18 wt%-WO /TiO .
3
2
2
3
2
3
2
R
Yield / %b (anisole / acid = 46.2c)
Yield / %b (anisole / acid = 10.0c)
6
protons acted the active sites for F-C reaction [32].
C
C
C
C
5
H
11
H
13
H
15
H
19
84
91
86
95
63
72
67
62
6
7
9
4. Conclusions
We have found that the catalytic activity of WO
with octanoic acid and anisole varies depending on the support used.
The highest activity was obtained when TiO was used as a support. The
catalyst was also applicable for similar reactions involving carboxylic
acids other than octanoic acid. TOF was found to be dependent on the
electronegativity of the cation of the support, implying that the electron
withdrawing or donating effect of the support determines the Brønsted
3
in the F-C acylation
a
The reaction was performed under the same condition given in the footnote
of Table 2.
b
2
Calculated based on the octanoic acid (2 mmol) used for reaction.
Molar ratio of anisole and octanoic acid.
c
3
acid strength and catalytic activity of the supported WO . We believe
that such a general understanding of the acid-base interaction between
the support and the supported oxide provides new insight into the
origin of the catalytic activity of the supported metal oxide catalyst.
Acknowledgements
This research was supported by the Ministry of Education, Science,
Sports and Culture, Grant-in-Aid for Scientific Research (C), 16K06859,
2016–2018.
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.mcat.2019.110410.
3
Fig. 11. TOF in the F-C reaction over WO loaded on various supports as
mentioned in Table 1 plotted as a function of the electronegativity of support
cation.
References
[
1] H. Zheng, J.Z. Ou, M.S. Strano, R.B. Kaner, A. Mitchell, K. Kalantar-zadeh, Adv.
Funct. Mater. 21 (2011) 2175–2196.
[
2] J. Polleux, A. Gurlo, N. Barsan, U. Weimar, M. Antonietti, M. Niederberger, Angew.
Chem. 118 (2006) 267–271.
6

K. Okumura, et al.
Molecular Catalysis 475 (2019) 110410
[19] M. Fernández-García, A. Martínez-Arias, A. Fuerte, J.C. Conesa, J. Phys. Chem. B
109 (2005) 6075–6083.
[20] A. Burrows, C.J. Kiely, R.W. Joyner, H.K. Knözinger, F. Lange, Catal. Lett. 39 (1996)
219–231.
[21] S. Yamazoe, Y. Hitomi, T. Shishido, T. Tanaka, J. Phys. Chem. C 112 (2008)
6869–6879.
[
[
3] K. Okumura, S. Ishida, T. Tomiyama, M. Niwa, Chem. Lett. 40 (2011) 527–529.
4] K. Okumura, T. Tomiyama, S. Shirakawa, S. Ishida, T. Sanada, M. Arao, M. Niwa, J.
Mater. Chem. 21 (2011) 229–235.
[5] B. Gao, Y. Ma, Y. Cao, W. Yang, J. Yao, J. Phys. Chem. B 110 (2006) 14391–14397.
[6] V. Logie, G. Maire, D. Michel, J.-L. Vignes, J. Catal. 188 (1999) 90–101.
[7] C.D. Baertsch, K.T. Komala, Y.-H. Chua, E. Iglesia, J. Catal. 205 (2002) 44–57.
[8] C. Yue, G. Li, E.A. Pidko, J.J. Wiesfeld, M. Rigutto, E.J.M. Hensen, ChemSusChem 9
[22] J.A. Horsley, I.E. Wachs, M. Brown, G.H. Via, F.D. Hardcastle, J. Phys. Chem. 91
(1987) 4014–4020.
[23] T. Miyamoto, N. Katada, J.-H. Kim, M. Niwa, J. Phys. Chem. B 102 (2002)
(
2016) 2421–2429.
9] K. Okumura, T. Kobayashi, H. Tanaka, M. Niwa, Appl. Catal. B Environ. 44 (2003)
25–331.
[
3
6738–6745.
[
[
10] T. Kim, A. Burrows, C.J. Kiely, I.E. Wachs, J. Catal. 246 (2007) 370–381.
11] D.G. Barton, M. Shtein, R.D. Wilson, S.L. Soled, E. Iglesia, J. Phys. Chem. B 103
[24] N. Naito, N. Katada, M. Niwa, J. Phys, Chem. B 103 (1999) 7206–7213.
[25] W. Zhou, N. Soultanidis, H. Xu, M.S. Wong, M. Neurock, C.J. Kiely, I.E. Wachs, ACS
Catal. 7 (2017) 2181–2198.
(1999) 630–640.
[
12] T. Kitano, T. Hayashi, T. Uesaka, T. Shishido, K. Teramura, T. Tanaka,
[26] K. Okumura, S. Ishida, R. Takahata, N. Katada, Catal. Today 204 (2013) 197–203.
[27] M.L. Kantam, K.V.S. Ranganath, M. Sateesh, K.B.S. Kumar, B.M. Choudary, J, . Mol.
Catal. A Chem. 225 (2005) 15–20.
[28] U. Freese, F. Heinrich, F. Roessner, Catal. Today 49 (1999) 237–244.
[29] K. Arata, H. Nakamura, M. Shouji, Appl. Catal. A Gen. 197 (2000) 213–219.
[30] L. Pauling, J. Am. Chem. Soc. 54 (1932) 3570.
[31] K. Tanaka, M. Ozaki, K. Tamaru, Shokubai 6 (1964) 262–265.
[32] H. Lin, F. Zhou, C.P. Liu, V. Ozoliņš, J. Mater. Chem. A Mater. Energy Sustain. 2
(2014) 12280–12288.
ChemCatChem 6 (2014) 2011–2020.
[
[
[
13] G. Sartori, R. Maggi, Chem. Rev. 106 (2006) 1077–1104.
14] M. Kawamura, D.-M. Cui, S. Shimada, Tetrahedron 62 (2006) 9201–9209.
15] E. Doustkhah, J. Lin, S. Rostamnia, C. Len, R. Luque, X. Luo, Y. Bando, K.C.-W. Wu,
J. Kim, Y. Yamauchi, Y. Ide, Chem. Eur. J. 25 (2019) 1614–1635.
16] J.C. Colmenares, R. Luque, Chem. Soc. Rev. 43 (2014) 765–778.
17] C. Pfaff, M.J. Pérez Zurita, C. Scott, P. Patiño, M.R. Goldwasser, J. Goldwasser,
F.M. Mulcahy, M. Houalla, D.M. Hercules, Catal. Lett. 49 (1997) 13–16.
18] J. Zhu, S. Wang, S. Xie, H. Li, Chem. Commun. (Camb.) 47 (2011) 4403–4405.
[
[
[
7