The acidity and catalytic activity of supported acidic cesium dodecatungstophosphates studied by MAS NMR, FTIR, and catalytic test reactions

Article abstract of DOI:10.1006/jcat.2001.3285

Acidic cesium salts of dodecatungstophosphoric acid (H₃PW) supported on silica and MCM-41 were characterized by multinuclear MAS NMR, FTIR, and catalytic test reactions. ³¹P-NMR data indicate that the supported samples contain salts of various stoichiometries (Cs₃PW and Cs₂HPW) and even H₃PO₄. According to ²⁹Si-NMR data, deposition of the Cs salt does not change the relative amount of various SiOH groups in the case of the silica support. In contrast, when the support is MCM-41, the relative amount of the SiOH groups increases significantly, indicating the breaking of the siloxane bonds resulting from their interaction with H₃PW. The ¹³³Cs MAS-NMR spectra also suggest that the CsPW salt interacts more strongly with the MCM-41 support than with SiO₂. The results of catalytic studies show that cesium salts of dodecatungstophosphoric acid with low Cs content supported on silica or MCM-41 (Cs_1.7H_1.3[PW₁₂O₄₀]-on-SiO ₂ and Cs_1.8H_1.2[PW₁₂O₄₀]-on-MCM41) are active, selective, and recyclable catalysts in Friedel-Crafts alkylations (isopropylation, transalkylation), the aromatic ortho-Claisen rearrangement, and the dehydration of pinacol. The specific activity of these catalyst preparations are comparable with that of neat Cs_2.6H_0.4[PW₁₂O₄₀]. This indicates that the active species are in a well-dispersed form.

Full text of DOI:10.1006/jcat.2001.3285

Journal of Catalysis 202, 379–386 (2001)
doi:10.1006/jcat.2001.3285, available online at http://www.idealibrary.com on
The Acidity and Catalytic Activity of Supported Acidic Cesium
Dodecatungstophosphates Studied by MAS NMR, FTIR,
and Catalytic Test Reactions
,1
´
´
A. Molna´r, T. Beregsza´szi, A. Fudala,† P. Lentz,‡ J. B. Nagy,‡ Z. Ko´nya,† and I. Kiricsi†
Department of Organic Chemistry, University of Szeged, Do´m te´r 8, Szeged, H-6720 Hungary; †Department of Applied and Environmental Chemistry,
University of Szeged, Rerrich B. te´r 1, Szeged, H-6720 Hungary; and ‡Laboratoire de R. M. N., Faculte´s Universitaires
Notre-Dame de la Paix, 61 Rue de Bruxelles, B-5000 Namur, Belgium
Received March 12, 2001; revised May 22, 2001; accepted May 22, 2001; published online August 9, 2001
structural properties of these compounds allow us to use
them as electrophilic catalysts (1–9), and they also promote
oxidation reactions. They work as heterogeneous catalysts
in apolar solvents, but their high solubility in many polar
solvents renders them into their homogeneous phase. In
the latter case, therefore, heterogenization is necessary. The
application of supported HPAs may be a possible solution.
It is known, however, that most common supports can ac-
commodate only a limited amount of HPAs in the range of
7–14 wt% (3). Moreover, supporting HPAs in low loading
often results in the decomposition of the Keggin structure.
One notable exception is activated carbon, which is able
to retain substantial amounts of HPAs without significant
losses even in polar media. The encapsulation of HPAs into
zeolites, carried out successfully with dodecamolybdophos-
phoric acid (H3[PMo12O40]) (10) and dodecatungstophos-
phoric acid (H3[PW12O40], denoted H3PW) (11), is a viable
solution to this problem.
Another possibility pursued recently is the immobiliza-
tion of HPAs into a silica matrix by means of a sol-gel tech-
nique. Promising results have been published with respect
to the application of such catalysts in esterification (12) and
ester hydrolysis (13–15), hydration of isobutylene (12), the
alkylation of phenol (12), the Friedel–Crafts alkylation (16)
and transalkylation (16), and the synthesis of methyl tert-
butyl ether (16).
Acidic cesium salts of dodecatungstophosphoric acid (H3PW)
supported on silica and MCM-41 were characterized by multi-
nuclear MAS NMR, FTIR, and catalytic test reactions. 31P-NMR
data indicate that the supported samples contain salts of various
stoichiometries (Cs3PW and Cs2HPW) and even H3PO4. Accord-
ing to 29Si-NMR data, deposition of the Cs salt does not change
the relative amount of various SiOH groups in the case of the sil-
ica support. In contrast, when the support is MCM-41, the rela-
tive amount of the SiOH groups increases significantly, indicating
the breaking of the siloxane bonds resulting from their interac-
tion with H3PW. The 133Cs MAS-NMR spectra also suggest that
the CsPW salt interacts more strongly with the MCM-41 sup-
port than with SiO2. The results of catalytic studies show that
cesium salts of dodecatungstophosphoric acid with low Cs con-
tent supported on silica or MCM-41 (Cs1.7H1.3[PW12O40]-on-SiO2
and Cs1.8H1.2[PW12O40]-on-MCM41) are active, selective, and re-
cyclable catalysts in Friedel–Crafts alkylations (isopropylation,
transalkylation), the aromatic ortho-Claisen rearrangement, and
the dehydration of pinacol. The specific activity of these catalyst
preparations are comparable with that of neat Cs2.6H0.4[PW12O40].
This indicates that the active species are in a well-dispersed
c
2001 Academic Press
form.
Key Words: heteropoly acids; cesium salts; MCM-41; multinu-
clear 29Si, 31P, and 133Cs MAS NMR; solid acid catalysis.
INTRODUCTION
An important development associated with this field was
the discovery and development of acidic salts of HPAs.
Of these Cs2.5H0.5[PW12O40], a water-tolerant, insoluble,
acidic, nonstoichiometric salt of dodecatungstophospho-
ric acid has been found to exhibit excellent catalytic ac-
tivities in electrophilic transformations of various organic
compounds in liquid-phase heterogeneous systems. Thor-
ough and wide-ranging studies in recent years have re-
vealed that it can be used in skeletal isomerization (17–
20), alkane–alkene alkylation (21, 22), Friedel–Crafts alky-
lation (16, 17, 23–27) and acylation (25, 26), and hydration
of alkenes (28, 29). In addition, Cs2.5H0.5[PW12O40] is also
The heterogenization of homogeneous catalysts has been
an important tendency in catalysis research in recent years.
The ease of handling, easy recovery, reuse, and possible re-
generation are the most significant advantages of heteroge-
neous catalysts as compared to their homogeneous coun-
terparts. Similar research efforts have yielded interesting
and important results in the application of heteropoly acids
(HPAs) with Keggin anion structure as well. The unique
1 To whom correspondence should be addressed. Fax: (36) 62-544-200.
E-mail: amolnar@chem.u-szeged.hu.
379
0021-9517/01 $35.00
c
Copyright
2001 by Academic Press
All rights of reproduction in any form reserved.

´
MOLNAR ET AL.
380
highly active in the transformation of oxygen-containing a purity of at least 99% . The aromatic reagents (benzene,
compounds including alcohol dehydration (19, 24) and the toluene, o-xylene) were stored over sodium wire. Isopropyl
pinacol rearrangement (30, 31), formation of esters (29, 31– mesylate (41), 5-tert-butyl-1,2,3-trimethylbenzene (42),
33), hydrolysis (25, 29, 32, 34, 35) and decomposition (17, and allyl phenyl ether (43) were prepared according to the
24, 31) of esters, and the transformation of methanol and literature.
dimethyl ether to lower hydrocarbons (36).
This outstanding performance of Cs2.5H0.5[PW12O40] is
attributed to its surprisingly high surface area and its micro-
and mesoporous structure (17–19, 24–26, 37). The BET sur-
face area ofcesium dodecatungstophosphates—and, in fact,
that of other salts of heteropoly+acids with a substantially
large cation, such as Rb+ and NH4 —was shown to decrease
with increasing cesium content until the composition of
Cs2H[PW12O40] (17–19, 24–26). Further increase in cesium
content brings about a marked increase in surface area,
Catalyst Preparation
Davisil silica gel (Aldrich, grade 363, 35–60 mesh) and
the mesoporous molecular sieve MCM-41 (synthesized ac-
cording to the method by Kim et al. (44)) were used to
prepare the supported cesium salts applying a literature
method (45). The supports were pretreated at 573 K for
24 h before use. First, CsNO3 (Aldrich) was deposited on
5 g of the support by impregnation with an aqueous so-
lution. After drying (393 K, 12 h) and calcination (773 K,
2 h) H3[PW12O40] (SERVA, 1.6 g) was impregnated fol-
lowed by drying (373 K for 2 h and 393 K for 12 h). The
quantity of CsNO3 was chosen to have less than monolayer
loadings and nominal compositions Cs1H2[PW12O40] and
Cs2.5H0.5[PW12O40].
For comparative studies unsupported salts with nominal
compositions Cs1.5H1.5[PW12O40] and Cs2.5H0.5[PW12O40]
were prepared by adding dropwise an appropriate amount
of aqueous CsNO3 to aqueous H3[PW12O40] under con-
tinuous stirring. The resulting precipitate was aged for
24 h at room temperature, separated by centrifugation, and
washed with distilled water.
1
reaching a value of about 150 m2 g for Cs3[PW12O40].
The highest specific surface acidity and, consequently, the
highest activities are achieved when the stoichiometry of
Cs equals 2.5.
A problem associated with the use of this salt as cata-
lyst material is its easy solubilization in water and in many
organic solvents to form a colloidal solution. However, im-
mobilizing Cs2.5H0.5[PW12O40] into a silica matrix (1, 12–15,
25) may eliminate the difficulty of the separation and work-
up procedures. The resultingmaterialexhibitsstrongacidity
and effectively catalyzes the hydrolysis of ethyl acetate (13)
and is active and selective in the synthesis of methyl tert-
butyl ether (15). We reasoned that a viable alternative could
be the use of high-surface-area support materials that can
accommodate cesium salts. This method may even allow us
to increase the effective surface of other cesium salts of low
surface area, thereby transforming them into more effec-
tive catalysts. This can be of great significance in develop-
ing environmentally friendly practical applications. Similar
approaches have recently been described (38, 39).
A comparative study, therefore, has been carried out us-
ing the well-studied Cs2.5H0.5[PW12O40] and two sets of
newly prepared supported samples containing low and high
amounts of cesium. Following a preliminary report about
the first results of this study (40) herein we disclose our ob-
servations with respect to the MAS-NMR and FTIR char-
acterization of these new supported cesium salts of dode-
catungstophosphoric acid. New examples of catalytic reac-
tions are also presented to widen the scope of applications.
These new data further testify to the excellent performance
of these catalytic materials.
Catalyst Characterization
Elemental analysis of the samples was performed by
IPC and X-ray fluorescence (XRF) analysis. Instrumen-
tal methods of characterization included X-ray diffraction
(XRD; DRON 2, operating under computer control), dif-
ferential scanning calorimetry (DSC; Perkin Elmer DSC 2
equipment), and FTIR spectroscopy (Mattson Genesis 1)
in the framework vibration range using the KBr pellet tech-
nique. Prior to measurements the samples were pretreated
at 573 K for 2 h in vacuo (better than 10 5 torr) or in flowing
helium.
The 29Si, 31P, and 133CsMAS-NMR spectra were recorded
either on a Bruker MSL 400 or on a CXP 200 spectrometer.
For 29Si (79.468 MHz), a 4 s (
=
/6) pulse was used with
a repetition time of 6.0 s. For 31P (161.924 MHz), a 6
s
(
=
/6) pulse was used with a repetition time of 10.0 s.
For 133Cs (52.464 MHz), a 2.0 s (
with a repetition time of 0.5 s.
=
/12) pulse was used
BET measurements were carried out in a conventional
volumetric adsorption apparatus at liquid-N2 temperature
(77 K). IR spectroscopy was applied to determine the acid-
ity of the samples by monitoring the adsorption of pyridine.
EXPERIMENTAL
Materials
Benzene, toluene, o-xylene, and
1,3,5-tri-tert- Self-supporting wafers of the samples were pretreated in an
butylbenzene were purchased from Aldrich whereas IR cell and then cooled to room temperature and treated
2,3-dimethylbutane-2,3-diol was a Fluka product. They had with 10 torr (1 torr = 133 Pa) of pyridine for 1 h. After

ACIDIC CESIUM SALTS
381
evacuation (1 h), spectra were recorded at 293 K, 373 K, (HP 5890 GC coupled with a HP 5970 MSD). None of the
473 K, and 573 K. supports was found to exhibit any activity under the exper-
Chemical methods for characterization included the iso- imental conditions applied.
merization of 1-butene in a recirculatory batch reactor. Af-
ter pretreatment (573 K, 2 h), the samples (50 mg) were
cooled to reaction temperature (423 K), and then 1-butene
(520 torr) was loaded. Sampling was carried out at regular
time intervals and conversions and product distributions
were determined byGC (HP-5710, 4.5-m dimethylsulfolane
column, 293 K).
RESULTS AND DISCUSSION
Preparation and Characterization of Supported
Cesium Salts
Data collected in Table 1 with respect to elemental anal-
ysis of the samples performed with IPC and XRF show that
the actualCs/H ratiosare substantiallyhigher than the nom-
inal values. The loadings calculated considering the actual
Catalytic Studies
The Friedel–Crafts reactions were carried out in a two- compositions of salts are 16 wt% for samples with lower
necked flask (10 ml) with vigorous stirring (magnetic stir- Cs content and 12 wt% for samples with higher Cs con-
rer) under reflux conditions with catalyst samples heat tent. These observations may be explained by the specific
treated before use (573 K, 2 h, flowing He). Isopropyla- method of preparation as opposed to the usual synthesis
tion of benzene was studied by reacting isopropyl mesylate found in the literature. Salts of heteropoly acids are pre-
(0.138 g, 0.001 mol) with 5 ml of benzene in the presence of pared, in general, by titrating the aqueous solution of the
0.1 g of catalyst at 363 K. Transalkylation was carried out acid with Cs2CO3, followed by evaporation of the resulting
by reacting a mixture of 4.24 ml (0.04 mol) of toluene and slurry. In our case, due to the specific requirements of the
0.495 ml (0.001 mol) of 5-tert-butyl-1,2,3-trimethylbenzene method applied, the solution of H3[PW12O40] was added to
in the presence of 0.1 g of catalyst at 383 K. Samples peri- CsNO3 adsorbed on various supports. It appears that such
odically withdrawn were analyzed by gas chromatography modifications result in the formation of cesium salts with
(GC).
The aromatic ortho-Claisen rearrangement was carried
stoichiometries higher than expected.
Further data of the supports and supported cesium salts
out by refluxing a continuously stirred mixture of 2.01 g are also collected in Table 1. It is seen that loading the salts
(0.015 mol) of allyl phenyl ether, 0.05 g of 1,3,5-tri-tert- to the supports brings about a certain decrease in the BET
butylbenzene (internal standard), and 0.1 g of catalyst in surface areas. The resulting supported samples, neverthe-
o-xylene (5 ml) for 3 h.
less, still have high enough surface area to ensure the forma-
The pinacol rearrangement was studied by heating 1 g tion of well-dispersed catalytic materials. Data acquired by
(0.085 mol) of pinacol (2,3-dimethylbutane-2,3-diol) and DSC (not shown) indicate that the supported samples are
0.1 g of catalyst. The product was continuously distilled off stable up to about 800 K and they show full crystallinity sim-
from the reaction flask and collected at 250 K. For further ilar to that of neat Cs2.5PW (XRD). No change in the struc-
details see Table 7.
Product compositions were determined by means of gas
ture of MCM-41 was found after pretreatment at 573 K.
All samples display both Brønsted and Lewis acidity. The
chromatography (Carlo Erba Fractovap 2350, TCD, 1.2-m acidity of the samples was determined by calculating the
1
SE 52 or Reoplex 400 column, and HP-5890, FID, 30-m DB- integral absorption of bands at 1560 cm characteristic of
5 column). Compounds formed were identified via the GC Brønsted acidity and that at 1450 cm 1 assigned as pyridine
retentions of authentic samples and GC-MS spectroscopy bonded to Lewis acid sites. These values were divided by the
TABLE 1
Characteristic Data of the Supports and Cesium Salts
3
Cs content
Acidity (m
Brønsted
)
Surface
acidity
mol g
Sample
denomination
BET
g
1
1
Nominal
Measured
(m2
)
Lewis
(
)
2.5
1.5
2.6
1.7
Cs2.6PW
Cs1.7PW
113
13
30
11.2
SiO2
555
326
312
773
609
544
0.02
0.04
0.27
0.07
0.07
0.22
1.27
2.58
1.95
0.83
1.71
2.34
2.5
1
2.9
1.7
Cs2.9/SiO2
Cs1.7/SiO2
MCM-41
Cs2.8/MCM41
Cs1.8/MCM41
3.5
33.4
2.5
1
2.8
1.8
7.1
30.8

´
MOLNAR ET AL.
382
TABLE 2
31P-NMR Data of the Pure and Supported Cesium Salts without
Cross Polarization
(ppm vs aq H3PO4) (I% )
Sample
Cs2.6PW
Cs1.7PW
Cs2.9/SiO2
Cs1.7/SiO2
Cs2.8/MCM41
Cs1.8/MCM41
Cs3PW
CsH2PW
H3PW
PO34 , H2PO4
14.6 (88)
15.5 (40)
14.5 (77)
14.5 (60)
14.5 (100)
14.5 (75)
10.5 (12)
12.0 (8)
13.6 (16)
13.4 (18)
13.4 (12)
4.9 (33), 0.3 (3)
0.4 (23)
1.6 (4)
10.8 (18)
11.1 (13)
only detected on SiO2-supported samples and the low ce-
sium content initial Cs1.7PW salt.
The pure Cs2.6PW salt contains some 88% Cs3PW and
12% H3PW acid forms. This high content Cs salt, when
deposited on MCM-41, conserves only the Cs3PW form.
When deposited on SiO2, a rather large amount of PO43
impurities (ca. 23% ) is also detected. The low Cs loading
leads to the coexistence of Cs3PW, Cs2HPW, and H3PW
forms in various amounts as shown in Table 2.
FIG. 1. Changes in the composition of the reaction mixture vs time
in the isomerization of 1-butene: (A) MCM-41 (diamonds), Cs1.8/MCM41
(squares), and Cs2.8/MCM41 (circles); (B) SiO2 (diamonds), Cs1.7/SiO2
(squares), and Cs2.9/SiO2 (circles). Black symbols: 1-butene, open symbols:
cis-2-butene, shaded symbols: trans-2-butene.
The supports were characterized by 29Si-NMR both with
and without 1H–29Si cross-polarization. The spectra are
composed of three lines: the 111 ppm line stems from
the Si(OSi)4 groups, the 101 ppm line from the SiOH de-
fect groups, and the 92 ppm line from the Si(OH)2 defect
groups (Table 3) (47). The relative amount of the SiOH
groups does not change when the CsPW salt is deposited
on SiO2. Only a slight decrease of the Si(OH)2 groups can
mass of the samples and the corresponding BET areas and
provide the aciditydata given in columns5and 6. It isclearly
seen that the deposition of cesium salts with low Cs content
(Cs1.7/SiO2 and Cs1.8/MCM41) on the supports results in a
significant increase in Brønsted acidity, which is in harmony
with the NMR results. Significant increases in Lewis acidity,
in contrast, are measured for all supported samples.
Data for the isomerization of 1-butene are shown in
Fig. 1. Of the two supports MCM-41 exhibited an activity
comparable to that of the Cs-containing samples, whereas
SiO2 was completely inactive. Cs salts supported MCM-41
(Cs1.8/MCM41 and Cs2.8/MCM41) showed practically the
same activity. This is in sharp contrast to the behavior of the
silica-supported samples. Here the activity of Cs1.7/SiO2
was much higher than that of Cs2.9/SiO2. The ratio of the
isomeric 2-butenes is approximately unity what is usually
observed in acid-catalyzed double bond migrations.
TABLE 3
29Si-NMR Data of the Pure and Supported Cesium Salts
(A) Without cross polarization
(ppm vs TMS) (I% )
Sample
Si(OSi)4
SiOH
Si(OH)2
SiO2
Cs2.9/SiO2
Cs1.7/SiO2
MCM41
Cs2.8/MCM41
Cs1.8/MCM41
111.0 (64)
111.3 (66)
111.4 (67)
109.0 (70)
110.8 (58)
110.8 (60)
101.3 (30)
101.7 (30)
101.8 (29)
100.2 (25)
101.2 (36)
101.5 (36)
92.0 (6)
91.8 (4)
92.3 (4)
91.8 (5)
91.8 (6)
91.9 (4)
Multinuclear MAS-NMR Characterization
The 31P-NMR chemical shifts of pure Cs2.6PW het-
eropoly acid salt and its supported samples are reported in
Table 2. The spectra were recorded both with and without
1H–31P cross-polarization. The 14.6 ppm line character-
izes the Cs3PW salt, the 13.4 ppm line the Cs2PW salt,
and the 10.6 ppm line the H3PW acid form (19, 23, 46).
Note that the CP spectra only show the 14.6 ppm line. This
underlines unambiguously the fact that in the other forms
the proton exchanges quite quickly, thereby destroying the
cross-polarized signal. The small NMR lines at ca. 0.4 and
1.6 ppm stem from PO43 and H2PO4 impurities. They are
(B) With cross polarization
(ppm vs TMS) (I% )
SiOH
Sample
SiO2
Cs2.9/SiO2
Cs1.7/SiO2
MCM41
Cs2.8/MCM41
Cs1.8/MCM41
Si(OSi)4
Si(OH)2
110.8 (32)
111.1 (31)
111.2 (39)
108.9 (37)
110.8 (17)
110.9 (21)
101.2 (57)
101.4 (61)
101.3 (53)
100.2 (53)
101.0 (70)
101.1 (69)
92.0 (11)
91.8 (8)
92.0 (8)
91.8 (10)
91.7 (13)
91.9 (10)

ACIDIC CESIUM SALTS
383
FIG. 2. 29Si-NMR spectra of MCM-41, Cs1.8/MCM41, and Cs2.8/MCM41.
be found. In contrast, when the CsPW salt is formed on Similar conclusions have been arrived at when organic
MCM-41, the relative amount of SiOH increases from 25% moieties were deposited on MCM-41 (48, 49). Indeed, it
to 36% . The relative amount of Si(OH)2 remains about was shown by both 29Si-NMR and IR results that owing to
the same within experimental errors. Figure 2 illustrates the high strain in the MCM-41 walls this reaction readily
the 29Si-NMR spectra of the initial MCM-41 and the two occurs, while on the relaxed structure of SiO2 this reaction
CsPW/MCM41 samples. The presence of SiOH groups is does not take place. This result also emphasizes the fact that
clearly shown by the increased intensity of the 101 ppm the Keggin units are certainly incorporated in the interior
line in cross-polarized conditions.
The increase in the relative intensities of the SiOH lines
of MCM-41 channels.
The 133Cs MAS-NMR spectra also show a difference be-
on MCM-41 during the formation and deposition of CsPW tween the influence of the two supports (Table 4). SiO2
suggests the reaction depicted in [1], where siloxane bonds does not significantly perturb the environment of the Cs
interacting with dodecatungstophosphoric acid are bro- ion in Cs2.6PW, as both the chemical shift ( 58.7 ppm in
ken to yield the silanol group. At the same time, the 1 M aqueous CsCl solution) and the linewidth (500 Hz)
Keggin anion strongly interacts with the MCM-41 surface. remain the same as for the pure Cs2.6PW salt. In contrast,
the supported MCM-41 shows an influence on the chemi-
cal shift, which becomes equal to ca. 56.5 ppm. This result
also suggests that the Cs2.6PW salt interacts more strongly
with the MCM-41 support, as explained above.
However, the 133Cs MAS-NMR spectra of the Cs1.7PW
sample show greater changes. Indeed, the chemical shift is

´
MOLNAR ET AL.
384
TABLE 4
133Cs-NMR Data of the Pure and Supported
Cesium Salts
Sample
(ppm vs 1 M aq CsCl)a
Cs2.6PW
Cs1.7PW
Cs2.9/SiO2
Cs1.7/SiO2
Cs2.8/MCM41
Cs1.8/MCM41
58.7
63.5b
58.5
59.1
56.2
56.9
a
1H = 500 Hz at
= 2100 Hz.
= 3500 Hz.
rot
b
1H = 1000 Hz at
rot
FIG. 3. Change in catalytic activity vs time in the trans-tert-butylation
equal to 63.5 ppm in the initial sample and the linewidth
is double that of the Cs2.6PW sample. When the Cs1.7PW is
deposited on SiO2, a great change occurs in the chemical
shift, which becomes equal to 59.1 ppm. Finally, as Cs+
ions are adsorbed on MCM-41 the interaction between
them and the support increases, and the chemical shift is
now close to that of the Cs2.8/MCM41 sample.
of toluene. ᭡ Cs1.7/SiO2, 1st run,
᭜ Cs1.8/MCM41.
1
2nd run,
1
3rd run,
Thisdifference observed in the two reactionsbetween the
activities of the salts of high and low Cs contents is easy to
explain. Isopropyl mesylate is a reagent that can quite read-
ily yield the 2-propyl carbocation as the reactive intermedi-
ate participating in the electrophilic aromatic substitution
to give the alkylated product. As a result, even the samples
with high Cs content (Cs2.9/SiO2 and Cs2.8/MCM41), i.e.,
those with low actual surface proton concentration, are able
to generate the carbocation in an appropriate concentration
ensuring certain low activities. Transalkylation, in contrast,
where the reacting tert-butyl carbocation is formed by the
dealkylation of the aromatic compound, is a more demand-
ing reaction. In this case the samples with high Cs content,
i.e., those with low concentration of acid sites, are unable
to generate carbocations.
The catalysts were also employed in repeated experi-
ments to learn about the stability and the possibility of re-
cycling the samples. As data obtained for transalkylation
illustrate (Fig. 3) the Cs1.7/SiO2 sample shows good stabil-
ity in repeated use. Although a certain drop of activity is
seen after each run, the conditions were not optimized and
the possibility of reactivation was not pursued.
In the transformation of allyl phenyl ether (Table 6) the
main product is always 2-methyldihydrobenzofurane (3)
[3]. This compound is thought to be formed by the further
transformation of the primary product o-allylphenol (2),
the product of the aromatic ortho-Claisen rearrangement.
By-products are phenol (the product of de-acetylation)
formed in significant amounts and some oligomers. Simi-
lar selectivities, i.e., the formation of 3 as the major prod-
uct, were observed by Sheldon et al. over H-beta and H-
mordenite zeolites (50) and in our recent study applying
various Nafion-based catalysts (51). The ring-closing trans-
formation of compound 2 to compound 3 was shown to
result from the highly acidic nature of the catalysts used.
In the dehydration of pinacol (5) carried out in the pres-
ence of the supported cesium salts, the main product is
always 3,3-dimethylbutane-2-one (6) formed as a results
CATALYTIC STUDIES
The results of catalytic studies with the supported sam-
ples are shown in Tables 5–7. Also shown are data for com-
parison obtained with the unsupported cesium salts.
Data for the Friedel–Crafts reactions indicate that in the
isopropylation of benzene the supported samples with low
Cs content are more active than the corresponding samples
with high Cs content (Table 5). In the transalkylation of
toluene [2], in turn, the latter samples proved to be com-
pletely inactive (Table 5).
TABLE 5
Activity of the Catalysts in Friedel–Crafts Alkylations
Transalkylation
Isopropylationa
1
Catalyst samples
Conversion
Conversion
TOF (min
1
)
Cs2.6PW
Cs1.7PW
100b (85)c
100b (75)c
54d (70)e
0
Cs2.9/SiO2
Cs1.7/SiO2
Cs2.8/MCM41
Cs1.8/MCM41
47
93
44
0
73d (90)e
0
1.4
1.3
100
69d (81)e
a
Reaction time = 1.5 h.
b
c
d
e
25 mg of catalyst.
20 mg of catalyst.
Reaction time = 3 h.
Reaction time = 7 h.

ACIDIC CESIUM SALTS
385
TABLE 6
highly active, whereas Cs1.7PW shows no activity at all. As
discussed in the introduction, Cs2.5PW is known to have
the highest specific surface acidity of the cesium salts of
dodecatungstophosphoric acid. Although it contains only a
relativelylownumber ofprotons, almost allare available for
catalysis due to the high surface area of this sample. It is not
surprising, therefore, that our Cs2.6PW sample shows good
catalytic activities. Cs1.7PW, in turn, has very low specific
surface acidity although it possesses more protons. Because
of the low surface area, however, most of the protons are
unavailable for participating in a catalytic process. As a
result, this sample is inferior to Cs2.6PW. Isopropylation
and the dehydration of pinacol are more facile processes
and such differences are not observed in these reactions.
A comparison of the results found for Cs1.7PW, and for
the two supported salts with low Cs content (Cs1.7/SiO2 and
Cs1.8/MCM41), shows that the catalytic performance of the
supported samples significantly exceeds that of Cs1.7PW.
This observation clearly indicates that cesium salts of dode-
catungstophosphoric acid with low Cs content and low sur-
face area (and, therefore, low specific surface acidity) can
be transformed into highly active catalytic materials. Activ-
ity data were calculated for transformations when reaction
conditions were identical for the three catalysts samples
using the surface acidity values collected in the last col-
umn in Table 1 (these values were calculated as described
in Ref. 24). The corresponding activity data (TOF values)
for transalkylation (Table 5), aromatic ortho-Claisen rear-
rangement (Table 6), and dehydration of pinacol (Table 7)
clearly show that the specific activities of the two supported
catalysts samples with low Cs content and those found for
the unsupported Cs2.6PW sample are comparable. It fol-
lows that the method described here, i.e., supporting acidic
cesium salts of heteropoly acids in submonolayer loading
on suitable high-surface area carriers, is a viable process
to produce catalytic materials of high efficiency. A further
notable advantage of these samples containing the active
phase in a finely dispersed form is that they can be recycled
as discussed earlier.
Conversions and Selectivities of the ortho-Claisen Rearrangement
of Allyl Phenyl Ether (1)
Selectivity (% )
Catalyst
samples
Conversion
(% )
1
2
3
4
Oligomers TOF (min )
Cs2.6PW
Cs1.7PW
Cs2.9/SiO2
Cs1.7/SiO2
Cs2.8/MCM41
Cs1.8/MCM41
39
0
0
60
0
66
11 59 26
4
10.8
7
8
53 40
62 30
traces
traces
15
17.8
of the pinacol rearrangement (Table 7) [4]. The unsatu-
rated by-products (7 and 8) are formed through stepwise
-eliminations. The appearance of these compounds and
especially that of the unsaturated alcohol (7) are indicative
of decreased acidity as found and discussed earlier (52).
Again, the activity of the samples with high Cs content (low
surface acidity) is significantly and characteristically lower
than that of the salts with low Cs content (high surface acid-
ity) (note the increased reaction temperature necessary to
achieve complete conversion in the case of Cs2.9/SiO2 and
Cs2.8/MCM41).
The activity pattern shown by the unsupported salts is
quite the opposite to what was observed for the supported
samples in the two demanding reactions, in transalkylation
and the transformation of allyl phenyl ether: Cs2.6PW is
TABLE 7
Selectivities in the Dehydration of Pinacol (5)a,b
Selectivity (% )
Catalyst samples
6
7
8
It is also important to discuss the effect of phosphate im-
purities found in some of the samples. No significant differ-
ence is found in the activities of the two supported samples
with low cesium content, although Cs1.8/MCM41 is free of
impurities. In addition, Cs1.7PW with the highest amount of
PO34 and H2PO4 exhibits no activity at all in transalkyla-
tion and the aromatic ortho-Claisen rearrangement, the two
most demanding reactions, just as the pure Cs2.8/MCM41.
These impurities, therefore, do not appear to affect catalyst
performance.
Cs2.6PW
Cs1.7PW
Cs2.9/SiO2
Cs1.7/SiO2
Cs2.8/MCM41c
Cs1.8/MCM41
79
81
85
77
80
80
0
0
3
0
4
0
21
19
12
22
16
20
c
a
Conversion is 100% in each reaction.
Reaction temperature = 438 K.
Reaction temperature = 473 K.
b
c
CONCLUSIONS
According to 31P-NMR data, salts of various stoichiome-
tries (Cs3PW and Cs2HPW) and even H3PO4 exist on the
supported samples. No change in the relative amount of

´
MOLNAR ET AL.
386
various SiOH groups of the silica support is detected by 17. Okuhara, T., Nishimura, T., Watanabe, H., Na, K., and Misono, M.,
29Si-NMR. In contrast, when the support is MCM-41, the
relative amount of the SiOH groups increases significantly,
indicating the breaking of the siloxane bonds resulting from
its interaction with H3PW. A stronger interaction of the
Stud. Surf. Sci. Catal. 90, 419 (1994).
18. Na, K., Okuhara, T., and Misono, M., J. Chem. Soc., Faraday Trans.
91, 367 (1995).
19. Essayem, N., Coudurier, G., Fournier, M., and Ve´drine, J. C., Catal.
Lett. 34, 223 (1995).
CsPW salt with the MCM-41 support than with SiO2 is also 20. Sun, Y., Yue, Y. H., and Gao, Z., React. Kinet. Catal. Lett. 63, 349(1998).
suggested by 133Cs MAS-NMR spectra.
Cesium salts of dodecatungstophosphoric acid with low
Cs content supported on silica or MCM-41 (Cs1.7H1.3
[PW12O40]-on-SiO2 and Cs1.8H1.2[PW12O40]-on-MCM41)
21. Okuhara, T., Yamashita, M., Na, K., and Misono, M., Chem. Lett.
1451 (1994).
22. Corma, V., A. Mart´ınez, A., and Mart´ınez, C., J. Catal. 164, 422 (1996).
23. Okuhara, T., Nishimura, T., Watanabe, H., and Misono, M., J. Mol.
Catal. 74, 247 (1992).
have been found to be active, selective, and recyclable 24. Okuhara, T., Nishimura, T., and Misono, M., Stud. Surf. Sci. Catal.
101, 581 (1996).
catalysts in organic transformations requiring the use
of electrophilic catalysts. The specific activity of these
catalyst preparations are comparable with that of neat
25. Izumi, Y., and Urabe, K., Stud. Surf. Sci. Catal. 90, 1 (1994).
26. Izumi, Y., Ogawa, M., and Urabe, K., Appl. Catal., A 132, 127 (1995).
27. Johnson, K., Viswanathan, B., and Varadarajan, T. K., React. Kinet.
Cs2.6H0.4[PW12O40]. This indicates that the active species
are in a well-dispersed form.
Catal. Lett. 63, 365 (1998).
28. Okuhara, T., Kimura, M., and Nakato, T., Chem. Lett. 839 (1997).
29. Okuhara, T., Kimura, M., Kawai, T., Xu, Z., and Nakato, T., Catal.
Today 45, 73 (1998).
ACKNOWLEDGMENT
30. Toyoshi, Y., Nakato, T., and Okuhara, T., Bull. Chem. Soc. Jpn. 71,
2817 (1998).
Thiswork wassupported bythe Hungarian NationalScience Foundation
(OTKA Grant T30156).
31. Koyano, G., Ueno, K., and Misono, M., Appl. Catal., A 181, 267 (1999).
32. Izumi, Y., Ono, M., Ogawa, M., and Urabe, K., Chem. Lett. 825 (1993).
33. Chen, X., Xu, Z., and Okuhara, T., Appl. Catal., A 180, 261 (1999).
34. Kimura, M., Nakato, T., and Okuhara, T., Appl. Catal., A 165, 227
(1997).
REFERENCES
35. Nakato, T., Kimura, M., Nakata, S., and Okuhara, T., Langmuir 14,
319 (1998).
1. Ono, Y., in “Perspectives in Catalysis” (J. M. Thomas and K. I.
Zamaraev, Eds.), p. 431. Blackwell Scientific Publications, London, 36. Hibi, T., Takahashi, K., Okuhara, T., Misono, M., and Yoneda, Y.,
1992.
Appl. Catal. 24, 69 (1986).
2. Okuhara, T., Mizuno, N., and Misono, M., Adv. Catal. 41, 113 (1996).
3. Mizuno, N., and Misono, M., Chem. Rev. 98, 199 (1998).
37. Bonardet, J. L., Fraissard, J., McGarvey, G. B., and Moffat, J. B.,
J. Catal. 151, 147 (1995).
4. Lee, K.-Y., and Misono, M., in “Handbook of Heterogeneous Cataly- 38. Soled, S., Miseo, S., McVicker, G., Gates, W. E., Gutierrez, A., and
sis” (J. Weitkamp, G. Ertl, and H. Kno¨zinger, Eds.), p. 118. Wiley-VCH,
New York, 1997.
5. Izumi, Y., Urabe, K., and Onaka, M., “Zeolite, Clay and Heteropoly
Acid in Organic Reactions.” Kodansha-New York, Tokyo-New York,
1993.
Paes, J., Catal. Today 36, 441 (1997).
39. Choi, S., Wang, Y., Nie, Z., Liu, J., and Pedersen, C. H. F., Catal.
Today 55, 117 (2000).
´
´
40. A. Molna´r, Beregsza´szi, T., Fudala, A., To¨ro¨k, B., Ro´zsa-Tarja´nyi, M.,
and Kiricsi, I., in “Supported Reagents and Catalysts in Chemistry”
(B. K. Hodnett, A. P. Kybett, J. H. Clark, and K. Smith, Eds.), p. 25.
The Royal Society of Chemistry, Cambridge, 1998.
6. Jansen, R. J. J., van Veldhuizen, H. M., Schwegler, M. A., and van
Bekkum, H., Recl. Trav. Chim. Pays-Bas 113, 115 (1994).
7. Kozhevnikov, I. V., Catal. Rev.—Sci. Eng. 37, 311 (1995); Chem. Rev. 41. Morgan, M. S., and Cretcher, L. H., J. Am. Chem. Soc. 70, 375 (1948).
98, 171 (1998).
42. Schlatter, M. J., and Clark, R. D., J. Am. Chem. Soc. 75, 361 (1953).
43. Entel, J., Rouf, C. H., and Howard, H. C., J. Am. Chem. Soc. 73, 2365
(1951).
8. Bonardet, J. L., Carr, K., McGarvey, G. B., McMonagle, J. B., Seay,
M., and Moffat, J. B., in “Advanced Catalysis and Nanostructured
Materials” (W. R. Moser, Ed.), p. 395. Academic Press, New York, 44. Kim, J. M., Kwak, J. H., Jun, S., and Ryoo, R., J. Phys. Chem. 99,
1996. 16742 (1995).
9. Soled, S., Miseo, S., McVicker, G., Baumgartner, J. E., Gates, W. E., 45. Tatibouet, J.-M., Montalescot, C., and Bru¨ckman, K., Appl. Catal., A
Gutierrez, A., and Paes, J., in “Advanced Catalysisand Nanostructured 138, L1 (1996).
Materials” (W. R. Moser, Ed.), p. 435. Academic Press, New York, 46. Ghambari-Siakkali, A., Philippon, A., Droyer, J., and Anderson,
1996.
M. W., Appl. Catal., A 192, 57 (2000).
10. Sulikowski, B., Haber, J., Kubacka, A., Pamin, K., Olejniczak, Z., and
Ptaszynski, J., Catal. Lett. 39, 27 (1996).
47. Engelhardt, G., and Mitchel, D., “High-Resolution Solid-State NMR
of Silicates and Zeolites.” Wiley, Chichester, 1987.
11. Mukai, S. R., Masuda, T., Ogino, I., and Hashimoto, K., Appl. Catal., 48. Brunel, D., Cauvel, A., Di Renzo, F., Fajula, F., Fubini, B., Onida, B.,
A 165, 219 (1997).
and Garrone, E., New J. Chem. 24, 807 (2000).
12. Izumi, Y., Hisano, K., and Hida, T., Appl. Catal., A 181, 277 (1999).
13. Izumi, Y., Ono, M., Kitagawa, M., Yoshida, M., and Urabe, K., Micro-
porous Mesoporous Mater. 5, 255 (1995).
49. Sutra, P., Fajula, F., Brunel, D., Lentz, P., Daelen, G., and B. Nagy, J.,
Colloids Surf., A: Physicochem. Eng. Aspects 158, 21 (1999).
50. Sheldon, R. A., Elings, J. A., Lee, S. K., Lempers, H. E. B., and
Downing, R. S., J. Mol. Catal. A: Chem. 134, 129 (1999).
14. Izumi, Y., Urabe, K., and Onaka, M., Microporous Mesoporous Mater.
21, 227 (1998).
´
51. To¨ro¨k, B., Kiricsi, I., Molna´r, A., and Olah, G. A., J. Catal. 193, 132
´
15. Molna´r, A., Keresszegi, Cs., and To¨ro¨k, B., Appl. Catal., A 189, 217
(2000).
´
(1999).
52. Molna´r, A., in “Fine Chemicals through Heterogeneous Catalysis”
´
16. Beregsza´szi, T., To¨ro¨k, B., Molna´r, A., Olah, G. A., and Prakash, G.
(R. A. Sheldon and H. van Bekkum, Eds.), p. 232. Wiley–VCH,
Chichester, 2001.
K. S., Catal. Lett. 48, 83 (1997).