Cesium-substituted dodecatungstophosphoric acid on K-10 clay for benzoylation of anisole with benzoyl chloride

Article abstract of DOI:10.1016/S0021-9517(02)00175-6

Synthesis of cesium-substituted dodecatungstophosphoric acid (Cs _2.5H_0.5PW₁₂O₄₀) supported on K-10 clay is reported. The catalyst is fully characterized. It serves as a promising catalyst for the acylation of aromatic compounds. The acylation of anisole with benzoyl chloride was studied systematically by using 20% w/w Cs _2.5H_0.5PW₁₂O₄₀/K-10 and this catalyst can be reused without any further chemical treatment, eliminating the effluent disposal problem. This acylation reaction is 100% selective toward 4-methoxybenzophenone, which is the desired perfumery intermediate. The effect of various important kinetic parameters was evaluated systematically to establish that the reaction obeys the Eley-Rideal type of mechanism with a very weak adsorption of the reactants. The reaction is free from any external mass transfer as well as intraparticle diffusion limitations and is intrinsically kinetically controlled.

Full text of DOI:10.1016/S0021-9517(02)00175-6

Journal of Catalysis 217 (2003) 88–99
www.elsevier.com/locate/jcat
Cesium-substituted dodecatungstophosphoric acid on K-10 clay
for benzoylation of anisole with benzoyl chloride
∗
1
G.D. Yadav, N.S. Asthana, and V.S. Kamble
Department of Chemical Engineering, University Institute of Chemical Technology (UICT), University of Mumbai, Matunga, Mumbai 400 019, India
Received 29 August 2002; revised 4 November 2002; accepted 18 November 2002
Abstract
Synthesis of cesium-substituted dodecatungstophosphoric acid (Cs
H
.5 0.5
PW O ) supported on K-10 clay is reported. The catalyst is
12 40
2
fully characterized. It serves as a promising catalyst for the acylation of aromatic compounds. The acylation of anisole with benzoyl chloride
was studied systematically by using 20% w/w Cs PW /K-10 and this catalyst can be reused without any further chemical
H
.5 0.5
O
12 40
2
treatment, eliminating the effluent disposal problem. This acylation reaction is 100% selective toward 4-methoxybenzophenone, which is the
desired perfumery intermediate. The effect of various important kinetic parameters was evaluated systematically to establish that the reaction
obeys the Eley–Rideal type of mechanism with a very weak adsorption of the reactants. The reaction is free from any external mass transfer
as well as intraparticle diffusion limitations and is intrinsically kinetically controlled.

2003 Elsevier Science (USA). All rights reserved.
Keywords: Cs-substituted dodecatungstophosphoric acid; Clay; Heterogeneous catalysis; Anisole; Benzoylation; 4-Methoxybenzophenone; Kinetics
1
. Introduction
lites have emerged as an alternative for the traditional pol-
luting catalysts, but unfortunately the slow diffusion of reac-
tants through their microporous structure makes them rel-
atively poor catalysts in liquid phase reactions and their
stability is susceptible in reactions where acids are gener-
ated as co-products. The use of highly reactive carboxylic
acid halogenide compounds as acylating agents, as well as
anhydrous HCl, which is produced during the reaction, may
remove the aluminum species from the zeolite framework.
Of late, several researchers have focused their attention on
Hβ and HY zeolites for acylation of substituted aromatics
including anisole. However, zeolites have been found to de-
activate [4–9]. Thus, there is still a scope for developing bet-
ter catalysts for acylation. The current work is a step in that
direction.
Functionalized aromatic ketones, produced via acyla-
tions, are of great importance in a variety of industries. For
example, acylation of anisole (4-methoxybenzene)with ben-
zoyl chloride or benzoic anhydride gives 4-benzoylanisole
(
4-methoxybenzophenone), which is a valuable perfumery
intermediate and also a precursor for antioxidants used in
cosmetics, PVC, unsaturated polyesters, acrylics, rubber,
alkylated resins, cellulose lacquers, and oil paints. The most
common method of introducing an acyl group into an or-
ganic compound is via the Friedel–Crafts reaction by using
either aluminium chloride in nitrobenzene or boron triflu-
oride in liquid HF as a catalyst [1,2]. Various other Lewis
acids such as FeCl3, ZnCl2, SnCl2, InCl3, SbCl5, metal tri-
flates, and Bronsted acids such as H2SO4 and HCl have been
studied extensively [3] and all these lead to serious efflu-
ent problems on an industrial scale. Demand for a cleaner
environment has attracted considerable attention to replace
these homogeneous, polluting, and corrosive systems with
environmentally benign heterogeneous acid catalysts. Zeo-
Among many other solid acid systems, heteropoly acids
(HPA) having Keggin anion structures have received the
most attention due to their simple preparation and strong
acidity [10,11]. Especially dodecatungstophosphoric acid
(
DTP) is the most extensively studied [12–14] since it pos-
sesses superacidity [15]. However, some of the major prob-
lems associated with HPAs in the bulk form are their low ef-
ficiency due to low surface area, rapid deactivation, and rel-
atively poor stability. Attempts to improve the efficiency and
stability of HPAs have been made by using various supports
*
Corresponding author.
E-mail address: gdyadav@yahoo.com (G.D. Yadav).
Applied Chemistry Division, BARC, Mumbai, India.
1
0021-9517/03/$ – see front matter  2003 Elsevier Science (USA). All rights reserved.
doi:10.1016/S0021-9517(02)00175-6

G.D. Yadav et al. / Journal of Catalysis 217 (2003) 88–99
89
◦
including mesoporous silica [16], mesoporous aluminosili-
cates [17], alumina and carbon [18,19], and zirconia [20].
Because of their basic nature, alumina and zirconia tend
to decompose HPAs, resulting in deformation of the parent
Keggin structure, thereby reducing the overall activity [20].
In this context, we had brought out the novelty of K-10 as a
suitable support for HPAs, in a variety of industrially impor-
tant reactions [21–27]. When HPAs are supported on K-10,
no defects in the primary Keggin structure of HPAs were
observed due to the acidic character of K-10 clay. Another
method that enhances the HPA stability and activity in liquid
phase reactions is to prepare catalysts in the form of alkali
metal salts of HPA. Especially Cs salts of DTP have been re-
ported to be better catalysts than DTP itself, but the particles
of Cs2.5H0.5PW12O40 are very fine and separation of the cat-
alyst from the liquid remains a problem. Direct impregnation
of Cs2.5H0.5PW12O40 is not possible due to its insolubility
in any kind of solvent [28,29].
a methanolic solution, dried at 110 C for 12 h and calcined
◦
at 300 C for 3 h. For comparison, bulk Cs2.5H0.5PW12O40
was prepared by adding the Cs precursor solution dropwise
to the DTP solution while stirring. The resulting precipitate
was dried at 110 C for 12 h, and calcined at 300 C for 3 h.
All catalysts were powdered and dried at 120 C for 3 h prior
◦
◦
◦
to their use.
2.3. Characterization of catalysts
Primary Keggin structures of bulk as well as supported
HPA catalysts were confirmed by FT-IR analysis. FT-IR
studies of the catalysts were conducted by using a Bruker
IFS-66 single-channel Fourier transform spectrophotometer.
Thin wafers were prepared by mixing 3 mg of catalyst with
100 mg of spectroscopic-grade KBr. The wafers were sub-
jected to 200 scans, after which the spectra were recorded.
Crystallinity and textural patterns of the catalysts were
predicted from XRD data, which were recorded using a Phil-
ips PW 1729 powder diffractometer with Cu-Kα(1.54 Å)
radiation.
The primary objective of our current work revolves
around the synthesis of Cs2.5H0.5PW12O40 supported on K-
1
0 clay (20% w/w) with intact Keggin anion. Benzoylation
of anisole with benzoyl chloride was used as a model reac-
tion to evaluate the reactivity and stability of this catalyst,
where corrosive acid such as HCl is evolved as a co-product
along with the desired aromatic ketone. This reaction is it-
self useful for studying the robustness of the catalyst and
its reusability. The activity of the above-synthesized cata-
lyst was also compared with other clay-based catalysts (pos-
sessing Bronsted and Lewis acidity) such as K-10 clay, 20%
w/w dodecatungstophoshoric acid (DTP)/K10 and 20%
w/w ZnCl2/K-10, and sulfated zirconia.
The BET surface area, pore volume, and pore size
distribution were determined from BJH and multipoint BET
methods by using a Micromeritics ASAP-2010 surface area
analyzer.
Particle size distribution of 20% w/w Cs2.5H0.5PW12O40/
K-10 was studied by using a Tracor Northern Image ana-
lyzer.
2.4. Reaction procedure
The reactor consisted of a flat-bottomed cylindrical glass
vessel of 100-ml capacity equipped with four baffles, a tur-
bine stirrer, and a condenser. The assembly was kept in a
thermostatic oil bath at a known temperature and mechani-
cally agitated with an electric motor. A typical reaction mix-
ture consisted of 0.0397 mol of benzoyl chloride and 0.27
mol of anisole with 1.05 g of catalyst. The total volume of
the reaction mixture was 35 ml and thus the catalyst loading
2
. Experimental
2
.1. Chemicals
All the chemicals were A.R. grade and procured from
reputed firms. Hexahydrate dodecatungstophosphoric acid
DTP), anisole, and ZnCl2 were obtained from M/s s.d. Fine
3
(
was 0.03 g/cm of the total volume of the reaction mixture.
◦
Chem. Ltd, Mumbai, India. Benzoyl chloride was procured
from E-Merck India Ltd., Mumbai, India. Montmorillonite–
K-10 was obtained from Fluka, Germany.
The reaction was carried out at 70 C at a speed of agita-
tion of 800 rpm. Samples were withdrawn periodically and
HPLC analyses were performed. 100% conversion of ben-
zoyl chloride was observed when the reaction was carried
◦
2
.2. Preparation of catalysts
out at 110 C for 4 h. Due to very high reaction rate, it was
not feasible to study the kinetics of the reaction at a tempera-
◦
◦
K-10 was dried at 120 C for 3 h prior to its use.
ture of 110 C. Therefore, various kinetic parameters affect-
◦
The following supported catalysts were prepared by well-
developed procedures and characterized in our laboratory:
ing the reaction rate were evaluated at 70 C. All the data
collected were in an error domain of 2%.
2
0% w/w ZnCl2/K-10 [25] and 20% w/w DTP/K-10
[
21–27] and sulphated zirconia [28,29].
2.5. Analysis
Cs2.5H0.5PW12O40/K-10 (20% w/w) was prepared by
◦
the incipient wetness technique and calcined at 300 C for
In periodically withdrawn samples, methanol was added
in excess to convert the remaining benzoyl chloride to
methyl benzoate and analyses were performed on HPLC
(Model: Toscho, UV-8010) by using a Merck 50983 column
3
h. K-10 clay was first impregnated with aqueous solution
of the Cs precursor, dried at 110 C for 12 h and calcined at
00 C for 3 h. Following this, DTP was impregnated using
+
◦
◦
3

90
G.D. Yadav et al. / Journal of Catalysis 217 (2003) 88–99
1
Fig. 1. H NMR of 4-methoxybenzophenone.
with a stationary phase of Lichrospher 100 RP-18, particle
size 5 µm, prepacked on a 250 mm × 4 mm i.d. column with
a Tosoh UV-8010 detector set at 255 nm. The eluent used for
analysis was 70:30 methanol (HPLC grade) and water at a
flow rate of 0.8 ml min . Quantitative results were obtained
by comparing the results with the calibration from synthetic
mixtures.
ported on K-10 clay was done by employing the following
techniques.
−1
3.1.1. X-ray diffraction studies
Bulk DTP hexahydrate (Fig. 2a) is characterized by its
cubic structure as shown in the XRD pattern presented in
diffraction lines. It is found to be crystalline in nature. On the
other hand, the crystallography structure is preserved while
preparing Cs2.5H0.5PW12O40 by exchanging the protons of
2
.6. Identification and isolation of product
+
bulk DTP with Cs (Fig. 2b).
The reaction mixture was filtered to remove the solid cat-
K-10 montmorillonite clay is an acid-treated montmo-
rillonite (refluxed in an acid solution). The acids that are
used to make K-10 by the treatment of its parent clay like
Tonsil-13 (which has a ordered layer structure) are sulphuric
acid, hydrochloric acid, and nitric acid. The parent clay is
generally refluxed with one of these mineral acids at high
temperature to produce K-10 montmorillonite. During the
mineral acid treatment at high temperature, the lamellar
structure (or typical montmorillonite structure) gets progres-
sively destroyed. The XRD pattern in Fig. 2c is for delam-
inated K-10 montmorillonite, showing a peak of very low
intensity at 2θ = 9.8, which may be due to the residual 2:1
(T–O–T) structure. The delamination during the process of
preparation of K-10 results in an increase in surface area
and formation of mesopores. In the case of K-10 there is
observed a sharp peak at 2θ = 26.2, which belongs to the
quartz and there is no effect on the 2θ value of quartz, even
after exchanging the K-10 clay either with DTP (Fig. 2d)
or with Cs2.5H0.5PW12O40 (Fig. 2e). From the background
XRD reflection in the K-10 sample, it is inferred that K-10
is amorphous in nature.
alyst and the product was isolated. Typically, the product was
isolated from the reaction mixture after 3 h. The filtrate was
treated with benzene and with water to remove traces of ben-
zoyl chloride. Benzoic acid formed by the reaction of ben-
zoyl chloride with water was removed by treating the mix-
ture with NaHCO3. The mixture was taken in a separating
funnel and allowed to stand for few minutes, water was sepa-
rated, the organic layer was taken in an evaporator to remove
benzene and excess anisole, and the solid mass was then dis-
solved in methanol and tested with thin layer chromatogra-
phy (TLC). Confirmation of the product was done through
melting point, GC, and TLC by comparison with the stan-
dard product. The reaction was found to be 100% selective
towards 4-methoxybenzophenoneas confirmed by the above
1
methods of analysis and by H NMR (Fig. 1).
3
. Results and discussion
3
.1. Characterization
Complete characterization of K-10 [19], 20% w/w
It was also observed that when DTP was impregnated on
K-10 (Fig. 2d) and Cs2.5H0.5PW12O40 on K-10 (Fig. 2e),
some crystallinity of DTP as well as Cs2.5H0.5PW12O40 was
lost.
DTP/K-10 [21–27], sulphated zirconia [28,29], and 20%
w/w ZnCl2/K-10 [30–33] has been provided earlier in de-
tail. Characterization of 20% w/w Cs2.5H0.5PW12O40 sup-

G.D. Yadav et al. / Journal of Catalysis 217 (2003) 88–99
91
Fig. 2. XRD patterns.
3
.1.2. FT-IR
The primary structure of supported HPAs (Proton form)
Fig. 3. FT-IR patterns.
as well as cesium salt of DTP were identified by com-
paring their FT-IR absorbance bands to those of bulk
DTP, Cs2.5H0.5PW12O40, and K-10 clay (Fig. 3). Bulk
DTP (H3PW12O40) and Cs2.5H0.5PW12O40 show the char-
case of Cs2.5H0.5PW O (Fig. 3b), it indicates the pres-
12
40
+
ence of partial H ion directly attached to the polyanion (–
+
OH) and is present in K-10 as M–OH and possibly as H O
3
−1
acteristic IR bands at ca. 1080 cm
(P–O in central
(terminal W=O), and 897 and
12 cm (W–O–W) associated with the asymmetric vibra-
(Fig. 3c).
−1
Furthermore, 31P NMR analyses of 20% w/w DTP/K-
10 and 20% w/w Cs2.5H0.5PW O /K-10 show that there
tetrahedral), 984 cm
8
−
1
12
40
tions in the Keggin polyanion (Fig. 3a and 3b); however,
Cs2.5H0.5PW12O40 is characterized by a split in the W=O
band, suggesting the existence of direct interaction between
is a peak at a chemical shift of ca. −15 ppm with reference
to 0 ppm response from 85% H PO and it confirms that
3
4
3
−
[PW O ] was the only species present on the support.
1
2
40
+
31
the polyanion and Cs . Similar observations have been ob-
In particular no additional P NMR peak, for example,
served for the case of W=O vibrations in Cu1.5PW12O40
at ca. −13 ppm, a peak generally attributed to a defect
−6
[
34] and Cs2.5H0.5PW12O40 supported on mesoporous sil-
(P W O ) Keggin, was evident [36,37]. These results
2 21 71
ica [35]. FT-IR of 20% w/w DTP/K-10 and 20% w/w
Cs2.5H0.5PW12O40/K-10 (Fig. 2d and 2e) indicate that the
primary Keggin structure is preserved after supporting 20%
w/w DTP and 20% w/w Cs2.5H0.5PW12O40 on K-10. The
12 40
regarding a supported form of Cs2.5H0.5PW O are in
good agreement with those obtained by Choi et al. [35] for
50% Cs2.5H0.5PW O supported on mesoporous silica.
12
40
−1
bands in the region of 1600 to 1700 cm
(at 1631–
3.1.3. Nitrogen adsorption
−1
1
642 cm ) is attributed to an –OH bending frequency of
It can be seen from Table 1 that the BET surface area
water molecules present in catalysts (Fig. 3a–e). In DTP, it
is present as water of crystallization (Fig. 3a), whereas in the
12 40
of 20% w/w Cs2.5H0.5PW O /K-10 is higher than that of
20% w/w DTP/K-10 and Cs2.5H0.5PW12O40 and slightly

92
G.D. Yadav et al. / Journal of Catalysis 217 (2003) 88–99
Table 1
3.1.4. SEM analysis
Surface area, pore volume, and average pore diameter of catalysts used in
the reaction
Figs. 5a, 5b and 5c show the scanning electron mi-
crographs of K-10, 20% w/w DTP/K-10, and 20% w/w
Cs2.5H0.5PW12O40/K-10, respectively. A distinct difference
was visible among K-10, 20% w/w DTP/K-10, and 20%
w/w Cs2.5H0.5PW12O40/K-10. It is clearly seen from the
SEM micrographs that both K-10 and 20% w/w DTP/K-10
samples possess rough and rugged surfaces, whereas 20%
w/w Cs2.5H0.5PW12O40/K-10 shows a smoother surface
due to a layer of (Cs salt of DTP) Cs2.5H0.5PW12O40 over
the external surface of K-10. The catalysts K-10, 20% w/w
DTP/K-10, and 20% w/w Cs2.5H0.5PW12O40/K-10 have a
particle size in the range of 5–10 µm.
Catalyst
Surface
Pore
Pore
diameter
(Å)
area
2
volume
3
(
m /g)
(cm /g)
K-10
230
107
207
100
145
0.36
0.32
0.29
0.115
0.28
64
71
58
28
68
2
2
0% DTP/K-10
0% Cs PW
H
.5 0.5
O /K-10
12 40
2
Sulfated zirconia
0% w/w ZnCl /K-10
2
2
lower than that of K-10. The pore size distribution of
sulphated zirconia suggests that the average pore sizes fall
in the range of 35–50 Å with the maximum number of pores
falling in the range of 38–45 Å [38], whereas for K-10 clay
and K-10 clay based catalyst they are found to fall in the
range of 50–75 Å suggesting that pore sizes of the catalysts
lie in the mesoporous region, i.e., > 20 Å (Fig. 4). The
adsorption–desorption isotherms for K-10, 20% DTP/K-10,
2
0% ZnCl2/K-10, and Cs2.5H0.5PW12O40/K-10 show that
they possess the type IV isotherm with the hysteresis loop of
type H3, which is a characteristic of mesoporous solids.
(a) K-10
(b) 20% DTP/K-10
(c) 20% Cs
H
PW O /K-10
2
.5 0.5 12 40
Fig. 4. Pore size distribution plots.
Fig. 5. SEM photographs

G.D. Yadav et al. / Journal of Catalysis 217 (2003) 88–99
93
Table 2
3.3. Reusability and stability of the catalyst
Efficacies of various catalysts on conversion of benzoyl chloride and
reaction rate
The stability of the active species in solution has been of
concern for solid acids, especially for supported materials.
Since 20% w/w ZnCl2/K-10, 20% w/w DTP/K-10, and
Catalyst
Initial activity
Conversion
(%)
−1
−1
(
mol g-cat
s )
−
4
20% w/w ZnCl /K-10
16.6×10
10 ×10
10×10
2×10
38
37
32
26
15
20% w/w Cs2.5H0.5PW O /K-10 showed comparable
2
12 40
−
−
−
−
4
4
4
4
Cs
H
PW O /K-10 (20%)
2
.5 0.5 12 40
results, to select the best candidate catalyst on its stability,
DTP (20%)/K-10
S-ZrO
2
K-10
their reusability studies were performed. High activity of
ZnCl2 supported on K-10 is due to a high local concentration
1.66 ×10
of Zn2 . However, the reusability of this catalyst was
found to be unfavorable. Clark’s group has reported that
some of the active sites of 20% w/w ZnCl2/K-10 become
poisoned by an oxygenated substrate (aromatic ethers, such
as anisole, poison the catalyst active sites in liquid phase
reactions) or product molecules, resulting in a substantial
loss of efficiency of the catalyst during its reusability. As
mentioned earlier, the potential use of HPAs supported
on K-10 clay has been well explored by Yadav et al. in
a variety of reactions including alkylation, etherification,
and dehydration [21–27]. 20% w/w DTP/K-10 has been
found to work efficiently for acylation reactions (when
acetyl chloride is used as an acylating agent) in the liquid
phase [39], but when the above catalyst is used in the
acylation reactions involving a highly corrosive acylating
agent such as benzoyl chloride [40], efficiency of the
catalyst decreases. To establish the effect of reusability in
the presence of benzoyl chloride, reusability of 20% w/w
DTP/K-10 and 20% w/w Cs2.5H0.5PW12O40/K-10 was
studied three times, including the use of fresh catalyst, and
is shown in Fig. 6. After first use and before every reuse,
the catalysts were washed with methanol and recalcined
+
3
Benzoyl chloride:anisole = 1:7, catalyst loading = 0.03 g/cm , temper-
◦
ature = 70 C, speed of agitation = 800 rpm, time = 2 h. DTP: dode-
catungstophosphoric acid.
3
.1.5. Particle size distribution
Since the newly synthesized catalyst is a K-10 clay based
catalyst, it was decided to study the particle size distribu-
tion of 20% w/w Cs2.5H0.5PW12O40/K-10, although it was
observed through its SEM micrograph that the catalyst par-
ticles are in the range of 5–10 µm. Particle size distribution
analysis of 20% w/w Cs2.5H0.5PW12O40/K-10 shows that
the majority of the catalyst particles lie in the range of 2–
1
0 µm.
3
.2. Effects of catalysts
◦
The efficacies of different catalysts at 70 C are given
in Table 2. The catalysts used were cesium (Cs2.5) salt
of dodecatungstophosphoric acid (DTP) supported on K-10
montmorillonite clay (20% w/w), 20% w/w DTP supported
on K-10, 20% w/w ZnCl2/K-10, sulphated zirconia, and
K-10.
As observed from Table 2, 20% ZnCl2/K-10 showed
higher activity, for benzoylation of anisole followed by 20%
w/w Cs2.5H0.5PW12O40/K-10, 20% w/w DTP/K-10, sul-
phated zirconia, and K-10. All the catalysts used were found
to be 100% selective toward 4-methoxybenzophenone. No
significant differences in final conversion of benzoyl chlo-
ride (after 2 h) were observed for 20% w/w ZnCl2/K-10,
◦
at 300 C for 3 h to remove the adsorbed reactants and
product. When 20% w/w DTP/K-10 was used, a decrease
in conversion was observed with its subsequent reuse, while
on the other hand when 20% w/w Cs2.5H0.5PW12O40/K-
1
0 was repeatedly used, no significant effect was observed,
although a minor decrease in final conversions was observed.
A decrease in catalytic activity of 20% w/w DTP/K-
1
0 due to the deteriorating condition of DTP and its
subsequent leaching from the support into the liquid phase
which was confirmed by the leaching test as well [24].
This was further confirmed by a UV spectrum test of
supernatant portions of 20% w/w DTP/K-10 and 20%
w/w Cs2.5H0.5PW12O40/K-10 catalyzed reaction mixtures.
When a supernatant portion of DTP/K-10 catalyzed reaction
mixture was subjected to the UV spectrum, it exhibited
an adsorption band at around 265 nm assigned to Keggin-
2
0% w/w Cs2.5H0.5PW12O40/K-10, and 20% w/w DTP/
K-10.
It is generally believed that acylation of anisole with
benzoyl chloride is a fairly easy reaction and even can
take place in the absence of a catalyst. Therefore, it was
decided to perform the reaction without catalyst and a
standard reaction between 0.27 mol of anisole and 0.0397
mol of benzoyl chloride (as mentioned earlier) was carried
−3
type PW12O40
ions [41], whereas in the case of 20%
w/w Cs2.5H0.5PW12O40/K-10 no such band was visible at
265 nm. The stability of 20% w/w Cs2.5H0.5PW O /K-
◦
out at 70 and 110 C and reflux temperature for 4 h.
12
40
These temperatures were decided in consonance with the
temperature range selected in the current study. On analysis,
the resultant reaction mixture (after 4 h) showed that
no benzophenone derivative was formed. These results
confirmed that the reaction does not take place in the absence
of catalyst.
10 was also confirmed by FT-IR spectra of fresh and
used 20% w/w Cs2.5H0.5PW O /K-10. No change in
12
40
primary Keggin anion was observed for used 20% w/w
Cs2.5H0.5PW O /K-10, even after its repeated use in the
12
40
presence of benzoyl chloride, which could have led to
deformation in its primary Keggin structure. In addition to

94
G.D. Yadav et al. / Journal of Catalysis 217 (2003) 88–99
not get leached out of the K-10 (support) and reaction does
not take place in the absence of catalyst.
Therefore, in all further reactions 20% w/w Cs2.5H0.5
PW12O40/K-10 was used as a catalyst, which not only
exhibits the high activity but also shows the catalyst is
reusable.
3
.5. Proof of absence of external mass transfer resistance
To assess the role of external mass transfer on the reaction
rate, the effect of the speed of agitation was studied. The
speed of agitation was varied from 800 to 1500 rpm. It
was observed that the conversion of benzoyl chloride was
practically the same in all the cases. Thus, it was ensured
that external mass transfer effects did not influence the
reaction. Hence, all further reactions were carried out at
8
00 rpm. A theoretical analysis of the assessment of external
mass transfer resistance is given to support this observation.
Details of this theory for general slurry reactions are given
elsewhere [29].
This reaction is a typical solid–liquid slurry reaction in-
volving the transfer of benzoyl chloride, the limiting reac-
tant (A), and anisole (B) from the bulk liquid phase to the
catalyst wherein external mass transfer of reactants to the
surface of the catalyst particle, followed by intraparticle dif-
fusion, adsorption, surface reactions, and desorption, take
place. The influence of external solid–liquid mass transfer
resistance must be ascertained before a true kinetic model
could be developed. Depending on the relative magnitudes
of external resistance to mass transfer and reaction rates, dif-
ferent controlling mechanisms have been put forward [42].
The liquid phase diffusivity values of the reactants A (ben-
Fig. 6. Effect of reusability on (a) Cs
H
PW O /K-10 (20%)
2
.5 0.5 12 40
and (b) DTP/K-10 (20% w/w). Benzoyl chloride:anisole = 1:7, catalyst
3
loading = 0.03 g/cm , speed of agitation = 800 rpm, time = 2 h,
◦
temperature = 70 C.
the UV test, when the supernatant portion of 20% w/w
Cs2.5H0.5PW12O40/K-10 catalyzed reaction was subjected
to FT-IR to detect any leaching of active polyoxometalate
species (Cs2.5H0.5PW12O40) from the K-10 clay, there was
no leaching of active species at all as no IR spectra were
zoyl chloride) and B (anisole), denoted by DAB and DBA
,
−1
observed in the range of 812–1080 cm
(characteristic
were calculated by using the Wilke–Chang equation [43]
◦
−5
−5
2
bands of primary Keggin anion) as well as the whole
inorganic range. These results clearly show that 20% w/w
Cs2.5H0.5PW12O40/K-10 is very stable in the presence
of corrosive and strong acylating agents such as benzoyl
chloride and this has been further confirmed by study of
heterogeneity of the reaction system by employing 20%
w/w Cs2.5H0.5PW12O40/K-10 as a catalyst.
at 70 C as 2.57 × 10 and 2.88 × 10 cm /s, respec-
tively. The solid–liquid mass transfer, coefficients for both
A and B were calculated from the limiting value of the Sher-
wood number (e.g.), Sh = k
A
SL–A p AB
d /D of 2. The actual
Sherwood numbers are typically higher by order of magni-
tude in well-agitated systems, but for conservative estima-
tions a value of 2 is taken [29,44]. The solid–liquid mass
transfer coefficients kSL–A and kSL–B values were obtained
−2
−2
3
.4. Heterogeneity of the reaction system
as 5.14 × 10 and 5.76 × 10 cm/s, respectively. The ini-
tial rate of reaction was calculated from the conversion pro-
files.
To further bolster the inferences drawn on the stabil-
ity of catalyst and about the intactness of active species
Cs2.5H0.5PW12O40) with K-10 clay (support), the reaction
(
3.6. Effect of catalyst loading
mixture was filtered after 2 h (reaction was conducted at the
standard condition) and the reaction with filtrate, in the ab-
sence of any external added catalyst, was further continued,
under otherwise similar conditions. No increase in conver-
sion was observed, even after 6 h. An increase in conver-
sion would have taken place had there had been any pres-
ence of leached active species in the filtrate of reaction
mixture. This fact confirms that newly prepared 20% w/w
Cs2.5H0.5PW12O40/K-10 is very stable and active species do
In the absence of external mass transfer resistance, the
rate of reaction is directly proportional to catalyst loading
based on the entire liquid phase volume. The catalyst loading
3
was varied over a range of 0.005–0.03 g/cm on the basis of
total volume of the reaction mixture. Fig. 7 shows the effect
of catalyst loading on the conversion of benzoyl chloride.
The conversion increases with increasing catalyst loading,
which is due to the proportional increase in the number of

G.D. Yadav et al. / Journal of Catalysis 217 (2003) 88–99
95
was done based on the Wiesz–Prater criterion to assess the
influence of intraparticle diffusion resistance [43].
According to the Wiesz–Prater criterion, the dimension-
less parameter Cwp, which represents the ratio of the intrin-
sic reaction rate to intraparticle diffusion rate, can be eval-
uated from the observed rate of reaction, the particle radius
(
Rp), effective diffusivity of the limiting reactant (De), and
concentration of the reactant at the external surface of the
particle.
2
(
i) If Cwp = −robsρpR /De[As] ꢀ 1, then the reaction is
P
limited by severe internal diffusional resistance.
(
ii) If Cwp ꢁ 1, then the reaction is intrinsically kinetically
controlled.
The effective diffusivity of benzoyl chloride (De–A)
inside the pores of the catalyst was obtained from the
bulk diffusivity (DAB), porosity (ε), and tortuosity (τ) as
−
6
2
3
.25 × 10 cm /s where De–A = DAB · ε/τ. In the present
−3
case, the value of Cwp was calculated as 1.818 × 10 for
the initial observed rate, which is much less than 1, and
this signifies the absence of resistance due to intraparticle
diffusion; therefore, the reaction could be considered as an
intrinsically kinetically controlled reaction. Further proof of
the absence of intraparticle diffusion resistance was obtained
through the study of the effect of temperature and it will be
evaluated later.
Fig. 7. Effect of catalyst loading. Benzoyl chloride:anisole = 1:7,
catalyst = Cs2
H
PW O
12 40
/K-10 (20%), speed of agitation = 800 rpm,
.5
0.5
◦
temperature = 70 C, time = 2 h.
active sites. Therefore, all further experiments were carried
3
out at 0.03 g/cm of catalyst loading.
3
.8. Effect of mole ratio
At steady state, the rate of external mass transfer (i.e.,
from the bulk liquid phase in which A and B are located with
concentration [A0] and [B0], respectively) to the exterior
surface of the catalyst is proportional to aP, the exterior
surface area of the catalyst where the concentrations of
A and B are [AS] and [BS], respectively. For a spherical
particle, aP is also proportional to w, the catalyst loading per
unit of liquid volume. It is possible to calculate the values of
The effect of mole ratio was studied at an anisole
to benzoyl chloride ratio of 1:1 to 7:1 (Fig. 8) under
otherwise similar conditions. As the concentration of anisole
is increased w.r.t. the concentration of benzoyl chloride,
an increase in conversion of benzoyl chloride and the rate
of reaction is observed, which is quite evident from the
graphs for a mole ratio of 3:1 to 7:1. This is because at the
lower concentration of benzoyl chloride more active sites of
catalyst are available for its surface adsorption, which results
in quantitatively large formation of active electrophilic
benzoylinium cation to react with anisole. At lower mole
ratios, the concentration of bezoyl chloride w.r.t. anisole
increases and the active sites become saturated with benzoyl
chloride, resulting in quantitatively low formation of active
electrophilic benzoylinium to react with anisole present in
the bulk phase, thereby showing a decrease in conversion
of benzoyl chloride and rate of reaction. Therefore, all the
subsequent reactions were carried out with a mole ratio of
[
AS] and [BS]. For instance,
ꢀ
ꢁ
kSL–AaP [A0] − [AS] = robs at steady state
−5
−3 −1
s .
=
3 × 10 molcm
(1)
From the calculations, it was observed that [AS] ≈ [A0];
similarly [BS] ≈ [B0]. Thus, any further addition of catalyst
is not going to be of any consequence for external mass
transfer.
3
.7. Proof of absence of intraparticle resistance
7
:1 without any solvent, where anisole used in excess acts as
Since the average particle size of 20% w/w Cs2.5H0.5
a solvent.
PW12O40/K-10 was found to be in the range of 2–10 µm
and the catalyst is amorphous in nature, it was not possible
to study the effect of catalyst particle size on the rate of reac-
tion. The average particle diameter of Cs2.5H0.5PW12O40/K-
0 used in the reactions was 0.001 cm (as analysed from par-
ticle size distribution data) and thus a theoretical calculation
3
.9. Effect of temperature
◦
1
The effect of temperature was studied from 50 to 110 C
(Fig. 9). The conversion was found to increase substantially

96
G.D. Yadav et al. / Journal of Catalysis 217 (2003) 88–99
with increasing temperature, which suggested that the reac-
tion was intrinsically kinetically controlled. The value of ac-
tivation energy (Section 3.11) also supported the fact that the
overall rate of reaction is not influenced by either external
mass transfer or intraparticle diffusion resistance and it is an
intrinsically kinetically controlled reaction on active sites.
3
.10. Reaction mechanism and kinetics
From the calculated values of mass transfer rates of A and
B, initial observed rate, it is evident that the rate is indepen-
dent of the external mass transfer effects. It is also seen from
the values of activation energy that the intraparticle diffusion
resistance is absent. Thus, the reaction could be controlled
by one of the following steps, namely, (a) adsorption (b) sur-
face reaction, or (c) desorption. Therefore, further develop-
ment of the actual reaction mechanism was undertaken.
3
.11. Development of mechanistic model
The adsorption studies of anisole and benzoyl chloride
were performed to predict the actual mechanism of reac-
tion. The adsorption constants calculated for benzoyl chlo-
ride (KB) and anisole (KA) were found to be 859 and
3
5
0 cm /mol. Although it seems from the adsorption stud-
Fig. 8. Effect of mole ratio. Benzoyl chloride:anisole, catalyst =
3
ies that adsorption of anisole takes place up to some extent
but it is quite negligible in comparison with that of benzoyl
chloride, indicating that KB ꢀ KA.
Cs
H
2.5 0.5
PW
O
12 40
/K-10 (20%), catalyst loading = 0.03 g/cm , temper-
◦
ature = 70 C, speed of agitation = 800 rpm, time = 2 h.
Based on the preliminary studies, a reaction mecha-
nism involving Cs2.5H0.5PW12O40/K-10 as a catalyst was
proposed (Fig. 10) and validated. Other models such as
Langmuir–Hinshelwood–Hougen–Watson and power law
were also tested. For the catalytic activity of Cs2.5H0.5PW12
O40/K-10, the Eley–Rideal model was found to hold. The
reaction involves two organic phase reactants, A (anisole),
B (benzoyl chloride), the desired product C (4-methoxy-
benzophenone), and D (co-product HCl). The Eley–Rideal
mechanism has already been found to be valid for our ear-
lier work of acylation of mesitylene with acyl chloride [39]
and acylation of 2-methoxynaphthalene with acetic anhy-
dride [45], where adsorption of only acylating agents has
been considered, because the formation of acylinum ion is
very important in the reaction.
The first step involves the formation of carbocation by the
adsorption of benzoyl chloride on a catalytic site (S):
kB
+
+
C6H5COCl +H ꢀ C6H5CO +HCl↑.
(2)
(
S)
ꢂ
k
(
B)
(BS)
(D)
B
The rate of chemisorption of B:
ꢂ
−
rB = kBCBCS − k CBSCD,
(3)
B
−
rB = 0, at equilibrium (quasi-equilibrium)
ꢂ
kBCBCS = k CBSCD,
CBS =
(4)
(5)
Fig. 9. Effect of temperature. Benzoyl chloride:anisole = 1:7, catalyst =
B
3
Cs
H
2.5 0.5
PW
O
12 40
/K-10 (20%), catalyst loading = 0.03 g/cm , speed of
B B S
k C C
.
agitation = 800 rpm, time = 2 h.
ꢂ
k CD
B

G.D. Yadav et al. / Journal of Catalysis 217 (2003) 88–99
97
Fig. 11. Plot of −ln(1−XB)+[M/{KB(M −1)+1}]ln{(M −XB)/M} vs
time for different values of M.
Fig. 10. Reaction mechanism.
where w is the catalyst loading. But CAS = 0,
Since KB = kB/kꢂ
B
w = CBS + C ,
S
ꢂ
ꢃ
CBCS
CD
CB
CD
CBCS
w = KB
+ CS = CS 1 + KB
,
CBS = KB
.
(6)
CD
ꢂ
ꢃ
w
Now the reaction between the adsorbed B as BS and the
liquid phase reactant A takes place as follows,
CS =
,
(10)
KBCB
1 +
CD
ꢂ
ꢃ
dCB
KBCBCS
= (kSRKB)
CD
CBCA
CD
w
−
−
= kSRCA
,
KBCB
CD
dt
1 +
kSR
BS −− − −→
(11)
+
+ S,
dCB
dt
kSRKBwCBCA
(A)
=
= rB,
(CD + KBCB)
CB = CB (1 − XB),
0
(C)
CA = CA − CB XB = CB (M − XB),
where kSR = reaction rate constant.
The rate of reaction of benzoyl chloride per unit liquid
volume is given by
0
0
0
D
C = C X .
B0
B
(12)
Where, CA /CB = M, the initial mole ratio of anisole to
0
0
dCB
benzoyl chloride, substituting in terms of X , the fractional
B
−
= kSRCACBS = rate of reaction of B.
(7)
conversion of B, and integrating, the following equations are
obtained.
dt
After substituting for CBS
Case a (M
ꢃ
dCB
dt
KBCBCS
CD
CBCACS
CD
ꢄ
ꢅ ꢄ
ꢅ
−
= kSRCA
= (kSRKB)
.
(8)
(9)
M
M − X
B
−ln(1 − X ) +
B
ln
M
ꢄ
ꢅ
The total site balance is as follows,
kSRKB(M − 1)wt
KB(M − 1) + 1
=
;
(13)
Ctotal = w = CBS + CAS + CS,

98
G.D. Yadav et al. / Journal of Catalysis 217 (2003) 88–99
Fig. 13. Arrhenius plot. (Plot of ln k vs 1/T ). Benzoyl chloride:aniso-
Fig. 12. Plot of −ln(1−XB)+[M/{KB(M −1)+1}]ln{(M −XB)/M} vs
SR
le = 1:7, catalyst = Cs
H
PW O /K-10 (20%), catalyst loading
time at different catalyst loading.
2.5 0.5 12 40
3
=
0.03 g/cm , speed of agitation = 800 rpm, time = 2 h.
Case b (M = 1).
1
XB
kSRKBwt
(1 − KB)
ln(1 − XB) +
=
.
(14)
(
1 − KB) (1 − XB)
been found to exhibit excellent activity for benzoylation
of anisole. Amongst various catalysts used in this reaction
The various equations were validated against the experi-
mental data to establish the validity of Eq. (13) for M
ꢃ= 1
2
0% w/w Cs2.5H0.5PW12O40/K-10 was found to be a bet-
3
(
Fig. 11) wherein KB = 859 cm /gmol. In the same way,
ter catalyst. Various kinetic parameters were studied for ben-
zoylation of anisole over 20% w/w Cs2.5H0.5PW12O40/K-
Eq. (14) was also validated for M = 1. It would therefore
mean that the reaction mechanism is of the Eley–Rideal type
of mechanisms in the absence of any diffusion resistance.
Fig. 12 shows four straight lines passing through the ori-
gin at catalyst loadings of 0.5, 1, 2, and 3% w/v, respec-
tively. The slopes of these lines are
1
0. The reaction follows the second-order kinetics and it
was found to be intrinsically kinetically controlled. The
catalyst was found to exhibit 100% selectivity toward 4-
methoxybenzophenone. The studies were conducted 3 times
including its fresh use and the catalyst showed excellent
reusability; no change in its primary Keggin anion was ob-
served, even after its repeated use in the reaction.
kSRKB(M − 1)w
k =
,
KB(M − 1) + 1
which are functions of w for the same molar ratio M,
indicating that the rate is proportionalto the number of active
sites present on the surface. Therefore, by putting the value
of KB in the above equation, the values of kSR could be
found and activation energy was calculated from the plot of
lnkSR vs 1/T as 15.6 kcal/mol (Fig. 13). This value also
suggests that the reaction is kinetically controlled.
Acknowledgments
Funding for this research was made available from the
BRNS Grant No. 98/37/4/BRNS Cell/724 under GDY. NSA
thanks BRNS for award of JRF and VSK acknowledges
BRNS. GDY gratefully acknowledges support from the Dar-
bari Seth Professorship Endowment and the hospitality of
Michigan State University East Lansing for the Johansen-
Crosby Visiting Chair Professorship of Chemical Engineer-
ing during 2001–2002, which provided an excellent creative
atmosphere.
4
. Conclusion
The synthesis of Cs2.5H0.5PW12O40 (20% w/w) sup-
ported on K-10 clay has been reported and Cs2.5H0.5
PW12O40 was successfully impregnated onto K-10 clay with
complete intact Keggin anion structure. This catalyst has

G.D. Yadav et al. / Journal of Catalysis 217 (2003) 88–99
99
References
[21] G.D. Yadav, N. Kirthivasan, J. Chem. Soc. Chem. Commun. 203
1995).
(
[
22] G.D. Yadav, N. Kirthivasan, in: J.P. Blitz, C.B. Little (Eds.), Funda-
mental and Applied Aspects of Chemically Modified Surfaces, Royal
Society of Chemistry, UK, 1999, p. 254.
[
1] G.A. Olah, Friedel-Crafts and Related Reactions, Vol. 1, Wiley–
Interscience, New York, 1963.
[
[
2] J.Y. Hyat, P.W. Reynolds, J. Org. Chem. 49 (1984) 384.
3] P. Sommai, O. Kazumi, M. Masahiro, M. Satoni, N. Masakatsu,
J. Chem. Soc. Perkin Trans. 1 (1994) 1703.
[23] G.D. Yadav, N. Kirthivasan, Appl. Catal. A 154 (1997) 29.
[24] G.D. Yadav, V.V. Bokade, Appl. Catal. A 147 (1996) 299.
[25] G.D. Yadav, N.S. Doshi, Catal. Today 60 (2000) 263.
[26] G.D. Yadav, N.S. Doshi, Org. Proc. Res. Dev. 6 (2002) 263.
[
4] A. Cybluski, J. Moulijn, M.M. Sharma, R.A. Sheldon, Fine Chemicals
Manufacture Technology and Engineering Book, 1st ed., Elsevier,
Amsterdam, The Netherlands, 2001.
[27] G.D. Yadav, N.S. Doshi, Appl. Catal. A 236 (2002) 129.
[28] G.D. Yadav, J.J. Nair, Micropor. Mesopor. Mater. 33 (1999) 1.
[29] P.S. Kumbhar, G.D. Yadav, Chem. Eng. Sci. 44 (1989) 2535.
[30] H. Clark, S.R. Cullen, S.J. Barlow, T.W. Bastok, J. Chem. Soc. Chem.
Commun. 2 (1994) 1117.
[5] G. Pérot, M. Guisnet, in: M.P.C. Weijnen, A.A.H. Drinkenburg
(Eds.), International Conference on Precision Process Technology.
Perspective for Pollution Prevention, Kluwer Academic, Delft, 1993,
p. 157.
[
[
6] M. Spagnol, L. Gilbert, D. Alby, Ind. Chem. Libr. 8 (1996) 29.
7] M. Spagnol, L. Gilbert, R. Jacquot, H. Guillot, P.J. Tirel, M.A. Le
Govic, in: Proceedings of the Fourth International Symposium on
Heterogeneous Catalysis and Fine Chemicals (8–12 September 1996),
Basel, 1996, p. 92.
[31] J.H. Clark, A.P. Kybett, D.J. Barlow, P. Landon, J. Chem. Soc. Chem.
Commun. (1989) 1353.
[32] S.J. Barlow, J.H. Clark, M.R. Darby, A.P. Kybett, P. Landon, K.
Martin, J. Chem. Res. 74 (1991).
[33] J.H. Clark, A.P. Kybett, D.J. Macquarrie, Supported Reagents: Prepa-
rations, Analysis and Applications, VCH, New York, 1992.
[34] T. Okuhara, T. Nishimura, T. Hibi, M. Misono, J. Catal. 93 (1985)
224.
[
8] P. Metivier, in: R.A. Sheldon, H. van Bekkum (Eds.), Fine Chemicals
through Heterogeneous Catalysis, Wiley, Weinheim, 2001, p. 161.
9] P. Moreau, A. Finiels, P. Meric, J. Mol. Catal. 154 (2000) 185.
[
[
[
[
[
[
[
10] M. Misono, Catal. Rev.-Sci. Eng. 29 (1987) 269.
11] I.V. Kozhvevnikov, Catal. Rev.-Sci. Eng. 37 (2) (1995) 311.
12] M. Misono, N. Nojiri, Appl. Catal. 64 (1) (1990) 1.
[35] S. Choi, Y. Wang, Z. Nie, J. Liu, C.H.F. Peden, Catal. Today 35 (2000)
117.
[36] M.T. Pope, in: Heterpoly and Isopoly Oxometalates, Springer, Heidel-
berg, 1983, p. 58.
[37] R. Massart, R. Contant, J. Fruchart, J. Ciabrini, M. Fournier, Inorg.
Chem. 16 (1977) 2916.
[38] G.D. Yadav, J.J. Nair, Langmuir 16 (2000) 4072.
[39] G.D. Yadav, N.S. Asthana, Ind. Eng. Chem. Res. 41 (2002) 5565.
[40] G.D. Yadav, N.S. Asthana, V.S. Kamble, Appl. Catal. A 240 (2003)
53.
[41] M. Kimura, T. Nakato, T. Okuhara, Appl. Catal. A 165 (1997)
227.
[42] H.S. Fogler, Elements of Chemical Reaction Engineering, Prentice
Hall, New Delhi, 1995.
[43] R.C. Reid, M.J. Prausnitz, T.K. Sherwood, The Properties of Gases
and Liquids, 3rd ed., McGraw–Hill, New York, 1977.
[44] G.D. Yadav, T.S. Thorat, Ind. Eng. Chem. Res. 35 (1996) 721.
[45] G.D. Yadav, M.S. Krishnan, Chem. Eng. Sci. 54 (1999) 4189.
13] A. Corma, Chem. Rev. 95 (1995) 559.
14] T. Okuhara, N. Mizuno, M. Misono, Adv. Catal. 41 (1996) 113.
15] M. Misono, N. Mizuno, K. Katamura, A. Kasai, Y. Konishi, K. Sakata,
T. Okuahra, Y. Yoneda, Bull. Chem. Soc. Jpn. 55 (1982) 400.
16] I.V. Kozhvevnikov, K.R. Kloestra, A. Sinnema, H.W. Zandbergen, H.
van Bekkum, J. Mol. Catal. A: Chem. 114 (1996) 287.
[
[
[
[
[
17] T. Blasco, A. Corma, A. Martinez, P. Martinez-Escolano, J. Catal. 177
(1998) 306.
18] G.I. Kapustin, T.R. Brueva, A.L. Klyachko, M.N. Tifoweeva, S.M.
Kulikov, I.V. Kozhvevnikov, Kinet. Katal. 31 (1990) 1017.
19] L.R. Pizzio, C.V. Caceres, M.N. Blanco, Appl. Catal. A 167 (1998)
283.
20] K. Saita, M. Nimura, S. Ohgoshi, Y. Akai, in: Symposium on Solid
Superacids: Synthesis Characterization and Catalysis, 214th National
Meeting, American Chemical Society, Nevada, 7–11 September, 1997.