FTIR and reaction studies of styrene and toluene over silica-zirconia-supported heteropoly acid catalysts

Article abstract of DOI:10.1006/jcat.2002.3787

The interactions of styrene and toluene with a sol-gel-prepared SiO₂-ZrO₂ mixed oxide and with two supported heteropolytungstic acid samples, i.e., a Keggin type (H₃PW₁₂O₄₀) and a Dawson type (H₆PW₁₈O₆₂) were studied by FTIR. Both supported heteropoly acids (HPA) were highly active, leading to styrene conversions of ≤ 94% after 4 hr, while the pure SiO₂-ZrO₂ support was almost inactive, with only 8% conversion after the safe reaction time. The Dawson catalyst showed higher activity than the supported Keggin. The catalysts differed in activity and in the product distribution, as reflected in the levels of conversion after equivalent reaction times. Toluene adsorption on HPA(K)/SiZr and HPA(D)/SiZr gave similar spectra, as observed for adsorption on the support alone. Increased pressures of styrene progressively demonstrated the adsorbed toluene, resulting in a typical spectrum of polymeric species on the catalysts, with maxima at 1601, 1583, 1494, 1478, 1450, and 1378/cm. The presence of zirconium in SiO₂-ZrO₂ mixed oxides led to the generation of Bronsted acid sites, which were otherwise absent in the single-component oxides. HPA addition to the SiO₂-ZrO₂ support led to an increase in the number of sites, which were available for the activation and subsequent reactions involving styrene, as confirmed by the increase activity in catalytic tests in the FTIR study.

Full text of DOI:10.1006/jcat.2002.3787

Journal of Catalysis 212, 231–239 (2002)
doi:10.1006/jcat.2002.3787
FTIR and Reaction Studies of Styrene and Toluene over
Silica–Zirconia-Supported Heteropoly Acid Catalysts
B. Bachiller-Baeza1 and J. A. Anderson
Surface Chemistry and Catalysis Group, Division of Physical and Inorganic Chemistry University of Dundee, Dundee, DD1 4HN, United Kingdom
Received June 6, 2002; revised August 1, 2002; accepted August 13, 2002
rials, the acidic and textural properties can be modulated
(5). This being the case, these SiO2–ZrO2 sol–gels might
prove appropriate as support materials for anchoring dif-
ferent active phases, since the degree of interaction and dis-
tribution over the surface can be controlled to some extent
by adapting the preparation methods and thereby altering,
for example, the chemical and electronic properties of the
support and the pore distribution.
The interactions of styrene and toluene with a sol–gel-prepared
SiO2–ZrO2 mixed oxide and with two supported heteropolytungstic
acid samples have been studied by FTIR. The results of the IR
study of styrene adsorption and toluene/styrene coadsorption indi-
cate that the oligomerization of styrene, which proceeds through
a carbenium intermediate, takes place over the support and the
supported catalysts. The alkylation of toluene by styrene, which
also proceeds via the formation of a carbenium ion from styrene,
is only detected for the supported heteropoly acid catalysts. These
spectroscopic observations, which are in good agreement with the
product distribution obtained for reaction experiments conducted
in the liquid phase, suggest that the reaction of the adsorbed, ac-
tivated styrene with toluene or with a second molecule of styrene
depends on the relative coverage of the two competing reagents. The
incorporation of the HPA to the silica–zirconia support increased
the density and strength of the acid sites favoring the retention
of a higher fraction of toluene and consequently the production
of higher yields of alkylated products. Differences between the
two supported HPA catalysts could be ascribed to the nature of
the supported heteropoly acid species and the modification of the
acid characteristics due to the interaction with the silica–zirconia
c
Heteropolyoxometalates have attracted much attention
in recent years because of their applicability in several
acid-catalyzed reactions (esterification, acylation, isomer-
ization) (6, 7). These compounds with a general formula
for the Keggin structure, X M12O4x0−8, where X is the central
atom (Si4+, P5+), x its oxidation state, and M the metal ion
(Mo6+, W6+), possess characteristic acidic and redox prop-
erties which can be controlled by substitution of the metal
atoms of the anion (central and addenda atoms) and the
countercations and can be related to their catalytic activ-
ity (8). The thermally stable and strongest heteropolyacid in
the Keggin series, the H3PW12O40, has been the most widely
studied. However, due to their low surface area most recent
studies have focused on heteropoly acids (HPA) loaded on
different supports which allow their use in heterogeneous
reactions, replacing the more common solid acid catalysts
such as HY and HX zeolites, SiO2–Al2O3, H3PO4/SiO2.
Basic supports such as alumina or magnesia lead to the de-
composition of the HPA (9), so SiO2, mesoporous molecu-
lar sieves, or activated carbons are favored (10–12).
Recently, the use of heteropoly acids have been sug-
gested as a possible means of supporting transition metals
through the inclusion or attachment of organometallic com-
plexes into the structure of the polyanion (13). By combin-
ing these with an appropriate support, these compounds
could be considered a potential alternative to the typi-
cal homogeneous catalysts currently used in reactions such
as hydrogenation, oxidation, or hydroformylation (14, 15).
However, due to the high activity of the HPA, particularly
in acid-catalyzed reactions, as described above, it is appro-
priate as a first step to study the supported heteropoly acid
under the conditions of the reaction of interest. Addition-
ally, it is well known that loss of the active metal and/or the
support.
ꢀ 2002 Elsevier Science (USA)
Key Words: FTIR; toluene alkylation; styrene; supported het-
eropoly acid; silica–zirconia.
1. INTRODUCTION
Among the metal oxides, mixed oxides with two or more
components are of significant scientific interest due to im-
proved stability or modification of chemical properties with
respect to the component oxides. For example, the incorpo-
ration of ZrO2 into a SiO2 matrix can improve the thermal
stability and the chemical resistance to alkaline attack and
leads to generation of surface acidic properties absent in
the single-component oxides (1–4). These materials may
have application in their own right as solid acids; however,
additionally, it has been shown that by modifying the prepa-
ration parameters during the sol–gel synthesis of the mate-
1 To whom correspondence should be addressed. Fax: +44 1382 345517.
E-mail: b.bachillerbaeza@dundee.ac.uk.
231
0021-9517/02 $35.00
c
ꢀ 2002 Elsevier Science (USA)
All rights reserved.

232
BACHILLER-BAEZA AND ANDERSON
associated ligands to the liquid phase may occur during the purity of the acid was confirmed by 31P NMR and thermo-
hydroformylation reaction over supported metal catalysts gravimetric analysis (TGA).
with chiral modifiers (16). It is therefore appropriate, as
The supported HPA samples (10 wt%) were prepared
a check for degradation of HPA-anchored reagents, that by impregnating the support with an aqueous solution of
the products of direct HPA catalyzed reaction be known. the heteropoly acid previously dispersed in water. The dis-
Patents by the Exxon group for this reaction over zeolite persion was stirred for 2 h and then evaporated to dryness.
HY indicate potential industrial importance of the alkyla- The solids were dried overnight at 343 K and will be re-
tion of toluene by styrene (17). In the present investigation, ferred to as HPA(K)/SiZr for the Keggin precursor and
two catalysts were prepared by supporting two different HPA(D)/SiZr for the Dawson structure.
heteropoly acids, a Keggin type (H3PW12O40) and a Daw-
2.3. FTIR Characterization
son type (H6P2W18O62), on a sol–gel-synthesized SiO2–
ZrO2 support. Results of an IR study of styrene adsorp-
tion and toluene/styrene coadsorption on the two prepared
samples are reported, in addition to results of catalyst test-
ing under standard conditions for the hydroformylation of
styrene.
The samples pressed into 25-mm-diameter self-
supported wafers were placed in a quartz cell connected
to a conventional vacuum-line apparatus and evacuated at
room temperature for 30 min. With this pretreatment, an
amount of adsorbed water was retained by the samples.
However, as reaction studies were also conducted using
samples without pretreatment, removal of all water for
spectroscopically examined samples would have made any
observations less relevant. Infrared spectra for styrene
and toluene/styrene at increasing vapor pressures were
recorded on a Perkin–Elmer 1710 FTIR spectrometer at
4-cm−1 resolution and an average of 25 scans.
2. EXPERIMENTAL
2.1. Preparation of the Support
A silica–zirconia mixed oxide (33% mol Zr) was pre-
pared via the sol–gel technique following a procedure de-
scribed previously (4, 5). Tetraethylorthosilicate (TEOS)
was added to a solution of propanol, water, and nitric acid
and stirred for 2 h under N2. The zirconium tetrapropox-
ide and propanol were then combined, and water was fi-
nally added after 1 h of stirring. The gel obtained was re-
dispersed in boiling ethyl acetate and refluxed for 5 h. The
resultant material was then transferred to a critical-point
drying apparatus, where it was dried under supercritical
CO2 for 2 days, with the CO2 replaced every 3 h. Finally, the
solid was calcined at 873 K for 5 h under flowing air. The
SBET of the support was 252 m2 g−1, the average pore diam-
eter was in the range 25–75 nm, and the solid was stable to
hydrothermal treatment at 873 K. Equivalent samples gave
acid site densities of 0.161 and 0.217 nm−2, for Lewis and
Brønsted sites, respectively, as determined by the amount
of pyridine retained by the samples after evacuation at
423 K (4).
Thin-film liquid-phase spectra of styrene and toluene
were obtained using a cell fitted with CaF2 windows. A
spectrum of a reference polystyrene film was also obtained.
2.4. Activity Measurements
The reaction was carried out under high pressure in a
270-cm3 steel autoclave fitted with an internal glass vessel,
a magnetically coupling stirred head, inlet/outlet ports, and
a pressure measuring gauge. The catalysts (150–200 mg)
were transferred to the reactor flask containing a solution
of styrene (17.3 mmol, Aldrich, 99%) in toluene (18 ml,
˚
Aldrich, 99.8%, HPLC grade) previously dried with a 5-A
molecular sieve. The autoclave was purged in N2, pressur-
ized to 80 bar with the reaction mixture CO/H2 (1 : 1) and
then heated in an oil bath to 333 K. After the reaction had
been carried out for 4.5 h and cooled to room temperature,
the gases were vented off and the composition of the solu-
tion was analyzed by gas chromatography using a 10% SP
2330 on a SUPELCOPORT column and a FID detector. A
GC–MS (HP-6890, MSD HP-5973) fitted with a HP-5MS
column was used to identify the reaction products.
2.2. Preparation of HPA-Supported Samples
Two types of heteropoly acids were used. The Keggin
type or 12-tungstonophosphoric acid (H3PW12O4 · 9H2O)
was obtained commercially (Aldrich). The Dawson struc-
ture (H6P2W18O62 · 2H2O) was synthesized from its ammo-
nium salt according to a procedure reported in the literature
(18). A mixture of Na2WO4 · 2H2O (Aldrich) and H3PO4
in water was refluxed for 8 h. After cooling, the ammonium
saltwasprecipitatedbyaddingsolidNH4Cl. Theprecipitate
was filtered, dissolved in warm water, and recrystallized.
3. RESULTS
3.1. Reaction Results
The results of the test reaction are summarized in Table 1.
The acid was obtained as an addition compound with ether Both supported heteropoly acids were highly active, lead-
from an aqueous solution of the ammonium salt treated ing to styrene conversions of up to 94% after 4 h, while
with ether and concentrated HCl. The ether was then evap- the pure SiO2–ZrO2 support was almost inactive, with only
orated and the acid crystallized in water. The identity and 8% conversion after the safe reaction time. The Dawson

STYRENE AND TOLUENE INTERACTION STUDIES
233
TABLE 1
tion reactions (19) and by maintaining an active solid–acid
catalyst which is resistant to deactivation by coking (19, 20).
It is expected that catalysts which had been pretreated at el-
evated temperatures to remove water would have exhibited
significantly different results, in terms of both activities and
selectivities. The catalysts differ not only in activity, as re-
flected in the levels of conversion after equivalent reaction
times, but also in the product distribution. The SiZr sup-
port exhibited low activity and the products 2 and 3, both
derived from the oligomerization of styrene, present much
greater levels than the alkylation product 4. The greatest
yieldofthemonoalkylatedproduct4wasobtainedusingthe
HPA(D)/SiZr. In addition, this catalyst showed a slightly
higher amount of higher molecular weight alkylates, 6, than
the supported Keggin catalyst.
Data of Reaction of Styrene in Toluene at 333 K under
H2/CO (1 : 1)
Composition (%)
Conversion
(%)
Activity
−1
Catalysts
SiZr
(mmol · g · h−1
)
2 + 3
4
5
6
cat
8
7
1.7
1.8
6.3
11.1
17.7
19.6
54.2
25.2
9.2
4.6 56.2 7.5
1.7 77.9 11.2
HPA(K)
HPA(K)/SiZr
HPA(K)/SiZra
HPA(D)/SiZr
HPA(D)/SiZra
HPA(D)/SiZrb
47
48
94
84
71
15.7 10.2 55.6 18.5
17.1 8.9 56.2 17.8
10.2 16.5 46.1 27.2
11.3 16.8 46.2 25.7
14.6 15.6 46.8 23.0
a Experiment carried out under N2 atmosphere.
b Experiment carried out for 1.5 h.
An unsupported Keggin acid (HPA(K)) was also tested
in the reaction and its activity was very low, in fact some-
catalyst showed higher activity than the supported Keggin. what similar to that obtained for the SiO2–ZrO2 support.
Additionally, the former sample exhibited a color change This is probably a consequence of the low surface area that
from white to slightly blue after removal from the reaction the bulk compound shows and the subsequently lower ac-
vessel, suggesting the formation of the heteropoly blues by cessibility of the acid sites. This sample exhibited a lower
reduction of the Dawson units.
selectivity to the alkylated products than the corresponding
No carbonylation products resulting from the hydro- supported Keggin heteropoly acid.
formylation of styrene were obtained. The products ob-
Experiments carried out by replacing the CO/H2 by
tained resulted from the dimerization and oligomerization N2 for both the HPA(D)/SiZr and HPA(K)/SiZr gave
of the olefin, and from the mono- (and poly-) alkylation slightly dissimilar results in terms of the measured activities
of toluene by styrene and these were detected by GC and (Table 1); however, the product distributions were not
identified by MS–GC. Scheme 1 shows the possible reac- affected by the change of conditions. The color change that
tion pathways. Note that catalysts as prepared would be was observed under the CO/H2 mixture for the supported
expected to contain a few percentage water. The role of Dawson catalyst was not detected when the reaction was
this water has been shown to be critical, both in terms of performed under N2. When the HPA(D)/SiZr was tested
modifying catalysts behavior by suppressing oligomeriza- for shorter reaction times, the conversion was reduced
SCHEME 1

234
BACHILLER-BAEZA AND ANDERSON
FIG. 1 Infrared spectra of (A) SiZr, (B) HPA(K)/SiZr, and (C) HPA(D)/SiZr after (a–d) exposure to increasing pressures of styrene and
(e) evacuation at 298 K (10 min) and (f) 353 K (15 min).
(Table 1) but the product distribution was very similar, observed in the IR spectrum of a reference polystyrene film
indicating that selectivity was effectively independent of (Fig. 2b), although bands at 1478 and 1378 cm−1, attributed
conversion.
to vibrations of CH3 end groups of short-chain polymers
with few repeating units (21), were absent for the latter.
A band at 3735 cm−1 (not shown) due to the stretching
vibration of isolated terminal hydroxyl groups of the silica
component of the mixed oxide (5) was observed for the
support prior to exposure to styrene. This band intensity
was significantly reduced after the adsorption of styrene,
consistent with expectation for an interaction of the double
bond of the molecule with the surface hydroxyl group of
3.2. Adsorption of Styrene
The interaction of styrene with the supported HPA sam-
ples and the bulk support was studied by the adsorption
of styrene from the vapor phase. Figures 1A–1C compare
the spectra for SiZr, HPA(K)/SiZr, and HPA(D)/SiZr
catalysts exposed to styrene vapor at increasing pressures.
Bands in the region 1700–1300 cm−1 confirm the presence
of adsorbed species at lower coverage for the SiZr sample,
with the most intense features, at 1450, 1494, 1576, and
1601 cm−1, assigned to the stretching vibrations of the
phenyl ring, and which are consistent with the features
observed in the spectrum of liquid styrene (Fig. 2a). The
bands at 1630 and 1416 cm−1 are attributed to the, νC
=
C
and δ
vibrations of the vinyl group, respectively.
=C–H
However, the relative intensity of the band at 1630 cm−1
with respect to the 1601-cm−1 peak was lower than for
the liquid styrene, suggesting that either an interaction
existed between the double bond of the molecule and the
support surface, or that only a fraction of the detected
species was present as adsorbed styrene and the remainder
was present as the cyclization product 3 of the dimer or
as oligomeric species 5. As the coverage was increased,
additional bands, at 1478 and 1378 cm−1, were observed.
After brief outgassing of the sample at room temperature
the band assigned to νC=C disappeared and only the bands
at 1600, 1584 (sh), 1494, 1478, 1450, and 1378 cm−1 re-
FIG. 2 Infraredspectraof(a)liquidstyrene, (b)solidpolystyrene film,
mained (Fig. 1A). These bands correspond well with those and (c) liquid toluene.

STYRENE AND TOLUENE INTERACTION STUDIES
235
the support. The intensity of the band due to the isolated tion in the intensity of the bands although evacuation at
silanol group was fully recovered following evacuation of higher temperature (353 K) caused slight loss in intensity,
the adsorbate.
especially in the case of the supported Dawson sample.
Both SiZr-supported heteropoly acids showed similar
spectra following exposure to styrene (Figs. 1B and 1C) but
which differ from that of the pure support. The dominant
species present on the surface did not show any apparent
3.3. Coadsorption of Toluene/Styrene
Figure 3A shows the spectra of SiZr which was firstly
coverage dependence and the peaks at 1601, 1585, 1494, exposed to toluene at 298 K and briefly evacuated, followed
1478, 1450, and 1378 cm−1 that correspond to the polymeric by exposure of styrene. The spectrum of adsorbed toluene
species were detected even at the lowest coverage. Unlike contained the expected bands, in accordance with those
the spectra for the SiZr support, the characteristic bands of observed for liquid toluene (Fig. 2c): peaks at 1603 and
the vinyl group at 1630 and 1416 cm−1 were absent at low 1496 due to the aromatic C–C stretching modes, and bands
coverage for the supported heteropoly acids. Examination at 1460 and 1384 cm−1 due to CH3 deformation vibration
of the spectra at higher coverages showed that the inten- (Fig. 3A, spectrum a). However, in the 2000- to 1600-cm−1
sities of the bands at 1478 and 1378 cm−1 were relatively region, which corresponds to the overtones involving the
higher for the Dawson-type heteropoly acid than for the out-of-plane C–H deformation modes, a shift of 15 cm−1
Keggin type, possibly reflecting the higher activity (Table 1) to higher frequencies with respect to the liquid-phase
of the former. However, as the first pulses of styrene under- spectrum was observed. This may result from some kind of
went complete conversion over both supported heteropoly interaction of the π-type electrons of the benzene ring with
acids, as indicated by the absence of the characteristic vinyl the support, implying that the ring was lying parallel to the
bands at 1630 and 1416 cm−1, the greater activity of the surface. In the presence of styrene, the characteristic C C
=
HPA(D) (Table 1) would not explain the differences ob- stretching vibrations of the group at 1630 and 1416 cm−1
served in the 1478- and 1378-cm−1 band intensities and thus were apparent (Fig. 3A, spectra b and c). With increasing
indicates inherent differences in the selectivities of the two styrene pressure, the band at 1384 cm−1 disappeared,
catalysts. Differences could be interpreted in terms of the indicating the displacement of the more weakly adsorbed
formation of oligomers with shorter chain length or a higher toluene molecules. Evacuation of the sample at room tem-
proportion of CH3 end groups over the supported Daw- perature led to a decrease in intensity of the 1600-, 1495-,
son. Davanzo et al. (21) reported that the polymer chain and 1450-cm−1 bands, while bands, assigned to the vinyl
length for styrene reacted over sulfated titania showed a group disappeared, suggesting the desorption of the molec-
dependence on the sulfate density (and hence acid site ularly adsorbed styrene. This was accompanied by an en-
density/strength) with sulfate-rich samples favoring shorter hancement in the bands at 1478 and 1378 cm−1, as observed
chains. Evacuation of the samples at RT caused no varia- for the experiment involving styrene adsorption alone.
FIG. 3 Infrared spectra of (A) SiZr, (B) HPA(K)/SiZr, and (C) HPA(D)/SiZr after (a) exposure to toluene and brief evacuation at 298 K, (b, c)
addition of increasing pressures of styrene, and (d) evacuation at 298 K and (e) 353 K.

236
BACHILLER-BAEZA AND ANDERSON
at 1601, 1583, 1494, 1478, 1450, and 1378 cm−1. However, an
additional band, at 1512 cm−1, was detected for both cata-
lysts that was absent for the support alone. This additional
band may be assigned to a reaction product with an aroma-
tic ring with substitution in the p position (22), implying that
C–C bond formation had taken place at 298 K. The samples
subjected to evacuation at this temperature showed no vari-
ation in the position or intensity of the bands mentioned,
and even after the evacuation temperature was increased
to 353 K, the spectrum remained largely unchanged.
Additional information regarding the nature of the
species formed during the coadsorption experiment over
supported HPA was obtained by comparing spectra (Fig. 4)
in the C–H stretching region (3200–2800 cm−1). When com-
pared with the spectrum of polystyrene (Fig. 4c), although
bands at 3082, 3060, 3025, 3002, 2925, and 2850 cm−1 were
common to both sets of spectra, additional bands, at 2962
and 2869 cm−1, were present (Fig. 4b) which were also ob-
served for styrene adsorption alone. These bands are as-
cribed to the asymmetric and symmetric stretching vibra-
tions, respectively, of CH3 groups, which can be associated
with the formation of oligomers or polymers of short chain
length with terminal methyl groups (21).
FIG. 4 Infrared spectra of (a) coadsorption of toluene/styrene for
HPA(D)/SiZr, (b) adsorption of styrene on HPA(K)/SiZr, and (c) thin
film of polystyrene.
As shown in Figs. 3B and 3C, adsorption of toluene on
HPA(K)/SiZr and on HPA(D)/SiZr gave similar spectra, as
observed for adsorption on the support alone. Bands due
to the out-of-plane C–H deformation modes were again
shifted by 20 cm−1 to higher frequencies, which again indi-
cates a strong interaction of the aromatic ring with some
3.4. FTIR of Postreaction Solution
In order to complete the spectroscopic investigation, IR
electron-withdrawing centers, pulling the molecule flat spectra were recorded of samples extracted from the so-
down on the surface. Increased pressures of styrene progre- lutions after reaction. Drops of the toluene solution were
ssively displaced the adsorbed toluene, resulting in a typical spread between CaF2 plates and the solvent allowed to
spectrum of polymeric species on the catalysts, with maxima evaporate. Figures 5A and 5B show spectra of postreaction
FIG. 5 Infrared spectra in the 3200- to 2800-cm−1 and 1700- to 1300-cm−1 regions of the solutions postreaction for (a) HPA(K)/SiZr, (b) HPA(D)/
SiZr, and (c) a thin film of polystyrene.

STYRENE AND TOLUENE INTERACTION STUDIES
237
solutions for the two supported HPA catalysts in the 3200-
The interaction of styrene with the SiO2–ZrO2 and with
to 2800-cm−1 and 1700- to 1300-cm−1 regions. Absorption two supported heteropoly acid samples has been evidenced
bands show good correlation with the results for spectra by the FTIR study. In the case of the support alone, this ini-
presented for toluene–styrene coadsorption: strong bands tial interaction was indicated by a decrease in the SiO–H
at 2962, 2869, 1478, and 1378 cm−1 attributed to the pres- band intensity either involving the double bond of the
ence of methyl groups in the polymeric species and a band styrene or via the π electrons of the aromatic ring. Al-
at 1512 cm−1 resulting from the p-substituted aromatic ring. though the sample as presented would have water asso-
Moreover, the intensity of the bands at 1583 and 1478 cm−1 ciated with both the HPA and the support surface, pertur-
were less significant for the solution-phase spectra. This bation of the SiO–H by the adsorbate provides evidence
could indicate that polymerization at the solid–liquid in- that the initial stages of adsorption on the support were not
terface led to products with longer chain and consequently hindered by the presence of the water. Although the lo-
the number of terminal CH3 groups was lower. However, cation of water molecules within the structure and their
it is also possible that these species, once formed, remained importance in activation of alkanes and alkenes is well
on the surface of the catalysts in FTIR experiments per- known in the case of bulk heteropoly acids (20), the spe-
formed by vapor adsorption onto pressed powder disks cific interaction between adsorbed water and supported
and thus appear concentrated, whereas the solubility of the isolated HPA units is less clear, as is the role of water in
polymer in toluene led to transfer from the solid surface to creating active Brønsted acid sites. However, as a typical
the liquid phase for experiments conducted in the reactor acid-catalyzed reaction, the oligomerization process is ex-
vessel. The intensity ratio between the bands at 2962 and pected to take place via a carbenium-type mechanism, in
2925cm−1 andthoseat1512and1378cm−1 providedfurther which the initial slow step is the protonation of the double
evidence for the difference in behavior between the two bond by a Brønsted acid site of the catalyst to generate a
supported HPAs in reaction. The ratio between the peaks secondary carbocation (7). An alternative mechanism, in-
assigned to νasCH3 (2962) and νasCH2 (2925) is 0.82 for the volving a surface-adsorbed carbene-type complex, has been
Dawson-type heteropoly acid but only 0.65 for the Keggin proposed for sulfated titania where the Brønsted acid sites
type. The band characteristic of the para-substituted alky- apparently play no role (21), although this mechanism has
lation product (4) at 1512 cm−1 was more intense than the been refuted (24). Adsorption of toluene is thought to in-
band at 1378 cm−1 for the Dawson-supported catalyst (ra- volve hydrogen bonding interactions of the acidic OH sites
tio, 1.5 : 1), while for the supported Keggin heteropoly acid with the π-electrons of the aromatic ring (25, 26), although
the ratio was 1.1 : 1.
studies of the alkylation of benzene and related aromatics
with olefins are thought (7) to follow an Eley–Rideal-type
mechanism, with the carbenium ion being attacked by the
free or weakly adsorbed aromatic. Although Corma (7) ar-
4. DISCUSSION
The reaction test results indicate that under the condi- gues the case that any molecule within a zeolite pore is
tions employed in this study, all the samples were active, subject to interactions such that it must be viewed as an
leading to products of oligomerization of styrene and alky- adsorbed state, the toluene molecule on samples employed
lation of toluene by styrene. Moreover, it is clear that both here is assumed to undergo alkylation reaction without the
HPA-supported samples show significantly higher reactiv- prerequisite that it be in any adsorbed or activated state.
ity compared with the pure heteropoly acid phase or the The SiZr support and the two HPA-supported samples
SiZr support. This observation, coupled with the enhanced displayed different behavior when styrene was introduced
selectivity to the alkylated compounds, for the supported onto a toluene-covered surface. In the case of the support
HPA catalysts suggests that the accessibility of acid sites alone, the styrene seemed to be in direct competition with
has been increased by incorporation of the heteropoly acid the toluene for the same adsorption sites, as suggested by
onto the support and that the acidity that these catalysts displacement of the latter by the styrene. Once the styrene
present is stronger than the sites on the original support is activated, it reacts with itself to form the corresponding
material. Since compounds of low polarity, such as styrene oligomer (2) as the dominant reaction (Table 1). The in-
and toluene, have weak affinity for the heteropoly acid and corporation of the HPA modifies the acidic properties of
cannot diffuse inside the solid, where most of the acid sites the samples, and even though oligomerization of styrene is
are located for the bulk HPA, reactants are limited to the observed, an additional band, at 1512 cm−1, attributed to
interaction with surface protons of the HPA crystallites, in the p-substituted alkylation product, is also detected. The
contrast to the pseudo-liquid catalysis proposed for polar alkylation reaction also proceeds via a mechanism which
molecules, where the majority of the protons may partic- involves the initial formation of a carbenium ion of the
ipate in the catalytic reaction (9, 23). Thus the dispersion alkylating agent; i.e., both the alkylation and oligomeriza-
of the HPA phases over a support increases the former’s tion reactions proceed via a common intermediate which is
activity in surface-catalyzed reactions.
formed in a rate-determining step involving protonation of

238
BACHILLER-BAEZA AND ANDERSON
styrene. Although some of the adsorbed toluene molecules
It is well known that the presence of Zr in SiO2–ZrO2
are displaced by the more strongly interacting styrene in mixed oxides leads to the generation of Brønsted acid sites
the consecutive adsorption experiment (Fig. 3), alkylation which are otherwise absent in the single-component oxides
selectivity is enhanced over the supported HPAs with re- (1, 2, 4, 5). However, the detection in the IR spectra of
spect to the support alone due to the ability of the latter adsorbed styrene molecules and the absence of the band
to retain a significant fraction of the toluene, as confirmed at 1512 cm−1 attributed to the alkylated product combined
by the FTIR results (Fig. 3). As all active catalysts must be with the low conversions obtained in reaction suggest ei-
considered to contain sites which have a Brønsted acid char- ther that there is a low density of surface sites of sufficiently
acter strong enough to protonate styrene (although present strong acid character or that these sites are blocked by coke
at differing densities), the subsequent reaction with either precursors produced from the adsorbed oligomer species
styrene or toluene will depend on the relative concentration that inhibit activation of further styrene molecules. Studies
of the two competing reagents in the vicinity of the carbo- of but-1-ene isomerization over a sulfated form of the SiO2–
cation. In the reaction between benzene and propene over ZrO2 support show that a good correlation exists between
HZSM-5, the polymerization of propene only competes se- activityandthedensityofBrønstedacidsites, asdetermined
riously with the alkylation reaction forming isopropylben- by pyridine adsorption (4, 30). However, an overestimation
zene when the total pressure is significantly increased (27). of sites was apparent for the nonsulfated SiO2–ZrO2, indi-
As toluene is employed as solvent in the study, its large con- cating that although all sites were of sufficient strength to
centration relative to styrene would favor alkylation over protonate the pyridine molecule, only a small proportion
oligomerization should the reaction proceed via a mecha- of sites were of sufficient strength to protonate but-1-ene at
nism involving a nonadsorbed state of styrene/toluene. The 423 K. It can be envisaged that these same small number of
only manner in which current results can be rationalized is sites on the SiO2–ZrO2 support which protonate but-1-ene
where the reagent reacting with the carbocation is in the ad- are also effective in activating the vinyl group of the styrene
sorbed state and thus selectivity becomes dependent on the molecule.
relative surface concentration of the styrene/toluene. In ad-
The addition of heteropoly acids to the SiO2–ZrO2 sup-
dition to possessing few sites capable of activating styrene port resulted in an increase in the number of sites which
(Table 1), the support alone does not possess sites which were available for the activation and subsequent reactions
are of sufficient strength to maintain a high local coverage involving styrene, as confirmed by the increased activity in
of toluene and thus the oligomerization/polymerization re- catalytic tests and in the FTIR study. The HPA are com-
action is favored. At high styrene conversion, the surface pounds that present very strong, predominantly Brønsted
coverage of the monoalkylation product over the supported acid sites (31), which act as the active sites for the oligomer-
HPA also begins to reach significant levels, such that se- ization and alkylation reactions. It has been claimed that
lectivity to polyalkylated products (6) increases relative to depending on the HPA loading, partial decomposition of
the oligomerization reaction (Table 1). The site which is the structure may be observed on supports such as ZrO2,
responsible for maintaining the localized high concentra- SiO2, or MCM-41 and that in general a reduction in the
tions of the competing styrene and toluene molecules (and acid strength of the HPA is detected when it is supported
at higher conversions, the monoalkylation product) most (20, 32, 33). HPA supported on hydrated ZrO2 modi-
likely involve exposed Lewis acid sites, which interact with fied the distribution of acid +sites on the support through
the π electrons of the ring in the case of toluene, leading the formation of ( Zr–OH2)n [H3−nW12PO40]n−3 or ( Zr–
to the observed 15-cm−1 shift of bands due to the out-of O)n–[H3−nW12PO40−n] distorted Keggin units (32). Simi-
plane C–H modes (Fig. 3) or via the π electrons of the vinyl lar species have been described for the interaction of
group of the styrene (Fig. 2). The relatively weak nature of the HPA with SiO2. Under conditions employed here
exposed Lewis sites in the silica–zirconia support (5) rela- for the HPA(K)/SiZr catalysts, the presence of lacunary
tive to the supported HPAs would be consistent with this species such as PW11O739− or unsaturated dimeric species
scheme; however, IR spectra (Fig. 2) also provide evidence (H6P2W18O62, H6P2W21O71) is most likely due to the
for interaction between π electrons of the vinyl group of the method of preparation, the use of a precalcined support
styrene and silanol groups of the support, thus providing an with an aqueous solution at pH > 2, and the low HPA con-
alternative manner in which adsorbate molecules might be tent (10 wt%) (11, 34). These species show a higher activ-
≡
≡
held in the vicinity of the carbocation intermediate.
ity in dealkylation of alkylphenols than the intact Keggin
The alkylation of alkylaromatics generally favor p-/o- structure, as their acid strength is diminished to a lesser
substitution due to the electron-donating effect of the alkyl extent following interaction with the support. The pres-
group (28). The lack of production of the o-isomer with the ence of some amounts of the intact Keggin species in the
catalysts employed results from the large steric hindrance HPA(K)/SiZr catalyst could account for its lower activity
effect of styrene as the alkylating agent, leading to the dom- when compared with the Dawson-type supported catalyst,
inance of p-substitution. Subsequent isomerization of the where the pure structure is expected (35). On the other
p-isomer leads to the formation of the m-isomer (29).
hand, the HPA can modify (and enhance) the acidity of the

STYRENE AND TOLUENE INTERACTION STUDIES
239
support resulting from the established interaction with the 13. Gouzerh, P., and Proust, A., Chem. Rev. 98, 77 (1998).
14. Augustine, R., Tanielyan, S., Anderson, S., and Yang, H., Chem. Com-
mun. 1257 (1999).
15. Johnson, B. I. S., and Stein, A., Inorg. Chem. 40, 801 (2001).
16. Coronado, J. M., Coloma, F., and Anderson, J. A., J. Mol. Catal. A 154,
143 (2000).
hydroxyl groups of the latter, thereby leading to increased
numbers of available acid centers when the HPA is present
in the supported form.
The absence of compounds 2–6 during the hydroformy-
lation reaction following the anchoring (14) of chiral phos- 17. Boucher, H. A., and Cody, I. A., European Patent 160,144 (1985).
18. Wu, H., J. Biol. Chem. 43, 189 (1920).
19. Kozhevnikov, I. V., Holmes, S., and Siddiqui, M. R. H., Appl. Catal. A
phine complexes to the supported heteropoly acids can be
regarded as an indication of the successful attachment of
the chiral complex to the polyanion (36).
214, 47 (2001).
20. Bardin, B. B., and Davis, R. J., Appl. Catal. A 200, 219 (2000).
21. Davanzo, C. V., Sheppard, N., and Al-Mashta, F., Spectrochim. Acta
ACKNOWLEDGMENTS
A 41, 263 (1985).
22. Arram, M., and Mateescu, G. H., “Infrared Spectroscopy.” Wiley–
Interscience, New York, 1972.
23. Misono, M., Chem. Commun. 1141 (2001).
24. Noda, L. K., and Sala, O., Spectrochim. Acta A 56, 145 (1999).
25. Du, Y., Wang, H., and Chen, S., J. Mol. Catal. A 179, 253 (2002).
26. Gutierrez-Alejandre, A., Gonzalez, H., Ram´ırez, J., and Busca, G.,
Ind. Eng. Chem. Res. 40, 3484 (2001).
The research was supported through a European Community Marie
Curie Fellowship (to B.B.B). We thank Dr. C. Rhodes and Ms. F. Donati
(University of Cardiff) for assistance with the preparation of the HPAs
used in this study and D. J. Rosenberg (University of Dundee) for advice
regarding preparation of the SiO2–ZrO2 support.
27. Brenan, J. A., Garwood, W. E., Yurchak, S., and Lee, W., in “Proc. Int.
Sem. on Alternative Fuels” (A. Germain, Ed.), p. 19.1. 1981.
28. Corma, A., Llopis, F., Melo, F., Viruela, P., and Zicovich-Wilson, C.,
J. Am. Chem. Soc. 116, 134 (1994).
29. Yeom, Y. H., and Shim, S. C., J. Mol. Catal. A 180, 133 (2002).
30. Rosenberg, D. J., and Anderson, J. A., unpublished results.
31. Tanabe, K., Misono, M., Ono, Y., and Hattori, H., Stud. Surf. Sci. Catal.
51, 169 (1989).
REFERENCES
1. Miller, J. B., Rankin, S. E., and Ko, E. I., J. Catal. 148, 673 (1994).
2. Anderson, J. A., and Fergusson, C. A., J. Non-Cryst. Solids 246, 177
(1999).
3. Zhan, Z., and Zeng, H. C., J. Mater. Chem. 9, 2647 (1999).
4. Rosenberg, D. J., Coloma, F., and Anderson, J. A., J. Catal. 210, 218
(2002).
5. Anderson, J. A., Fergusson, C., Rodr´ıguez-Ramos, I., and Guerrero-
Ruiz, A., J. Catal. 192, 344 (2000).
6. Kozhevnikov, I. V., Chem. Rev. 98, 171 (1998).
7. Corma, A., Chem. Rev. 95, 559 (1995).
8. Hu, C., Hashimoto, M., Okuhara, T., and Misono, M., J. Catal. 143,
437 (1993).
32. Lo´pez-Salinas, E., Herna´ndez-Corte´z, J. G., Schifter, I., Torres-Garc´ıa,
E., Navarrete, J., Gutie´rrez-Carrillo, A., Lopez, T., Lottici, P. P., and
Bersani, D., Appl. Catal. A 193, 215 (2000).
33. Kapusan, G. J., Brueva, T. R., Klyachku, A. L., Timofeeva, M.
N., Kulikov, S. M., and Kozhevnikov, I. V., Kinet. Katal. 31, 1017
(1990).
9. Misono, M., Catal. Rev.–Sci. Eng. 29, 269 (1987).
10. Izumi, Y., Hasebe, R., and Urabe, K., J. Catal. 84, 402 (1983).
11. Trolliet, C., Coudurier, G., and Vedrine, J. C., Top. Catal. 15(1), 73
(2001).
34. Kozhevnikov, I. V., Kloetstra, K. R., Sinnema, A., Zandbergen, H. W.,
and Van Bekkum, H., J. Mol. Catal. A 114, 287 (1996).
35. Baronetti, G., Thomas, H., and Querini, C. A., Appl. Catal. A 217, 131
(2001).
12. Izumi, Y., and Urabe, K., Chem. Lett. 663 (1981).
36. Bachiller-Baeza, B., and Anderson, J. A., unpublished results.