Effect of Size and Extent of Sulfation of Bulk and Silica-Supported ZrO2 on Catalytic Activity in Gas- and Liquid-Phase Reactions

Article abstract of DOI:10.1021/jp030436b

Sulfated zirconia was prepared according to three different procedures, i.e., conventional impregnation with sulfuric acid and calcinations of two zirconia, reaction of zirconium tetrachloride with sulfuric acid giving bulk anhydrous zirconium sulfate, and deposition-precipitation of highly dispersed zirconia on silica and subsequent reaction with either H₂S or O₂, or SO₂ and O₂, or SO₃. For the thermal pretreatment of sulfated zirconia, it was important that the constituents of sulfuric acid show a considerable vapor pressure at significantly lower temperatures than the temperatures at which zirconium sulfate decomposes. The rate of transport of water out of a catalyst batch being thermally pretreated is decisive for the final activity of the catalyst. The catalysts were employed in the gas-phase trans-alkylation of benzene and diethylbenzene to ethylbenzene at 473 and 673 K and in the solvent-free, liquid-phase hydroxyacyloxy-addition of acetic acid to camphene to isobornyl acetate at 338 K. The volatility of the components of the active species also affected the activity in gas-phase reactions. The catalytic activity of sulfated zirconia catalysts originated from supported sulfuric acid containing water. The deficiencies of water-diluted sulfuric acid, volatility and solubility in liquid, had to be considered in the technical utilization of sulfated zirconia catalysts.

Full text of DOI:10.1021/jp030436b

J. Phys. Chem. B 2003, 107, 13403-13413
13403
Effect of Size and Extent of Sulfation of Bulk and Silica-Supported ZrO2 on Catalytic
Activity in Gas- and Liquid-Phase Reactions
Ivo J. Dijs, John W. Geus, and Leonardus W. Jenneskens*
Debye Institute, Department of Physical Organic Chemistry, Utrecht UniVersity, Padualaan 8,
3
584 CH Utrecht, The Netherlands
ReceiVed: April 11, 2003; In Final Form: September 29, 2003
Sulfated zirconia has been prepared according to three different procedures, viz., (i) conventional impregnation
2
with sulfuric acid and calcination (3 h at 773 K) of two zirconia’s (50 and 217 m /g), (ii) reaction of zirconium
tetrachloride with sulfuric acid giving bulk anhydrous zirconium sulfate, and (iii) deposition-precipitation of
highly dispersed zirconia on silica and subsequent reaction with either H S and O , or SO and O , or SO .
2 2 2 2 3
The latter two procedures lead to essentially water-free catalysts. Thermogravimetry showed that the
2
2
impregnated and calcined zirconia’s loose sulfate above 830 K (50 m /g) and 910 K (217 m /g). In a gas flow
containing water, the sulfated silica-supported zirconia loses sulfate already at 673 K because of the reaction
to more volatile sulfuric acid. The catalysts were employed in the gas-phase trans-alkylation of benzene (1)
and diethylbenzene (2) to ethylbenzene (3) at 473 and 673 K and in the solvent-free, liquid-phase hydro-
acyloxy-addition of acetic acid to camphene (4) to camphene (5) to isobornyl acetate (6) at 338 K. The
water-free catalysts were not active; only after addition of water was catalytic activity exhibited. The catalytic
activity of the differently prepared sulfated zirconia’s is governed by the equilibrium: Zr(SO
4
)
2
+ 4H
2
O h
Zr(SO ‚4H O + nH O h ZrO + 2H SO .aq. Addition of water vapor to the bulk sulfate at 473 K led to
4
)
2
2
2
2
2
4
the reaction to the tetrahydrate, which was not active, whereas the highly dispersed silica-supported zirconium
sulfate reacted to form sulfuric acid. The supported catalyst rapidly released the water at 473 K, which resulted
in a rapid drop in catalytic activity. Transport of water through the porous system dominates the activity of
2
the impregnated zirconia’s. Accordingly, the slight activity of the zirconia of 50 m /g rapidly dropped at 473
2
K, whereas the zirconia of 217 m /g displayed a high and stable activity. At 673 K, the transport is much
more rapid. The activity of the highly porous zirconia was therefore at 673 K much lower than at 473 K.
Whereas the gas-phase reaction is governed by transport of water vapor, the liquid-phase reaction is dominated
2
by transport of the reactants to the active sites. Consequently, the sulfated zirconia of 50 m /g showed a
2
considerably higher activity than that of 217 m /g. Also the silica-supported catalyst exhibited a higher activity.
The consistent results demonstrate that sulfated zirconia needs water to display activity in the gas-phase and
liquid-phase reaction studied.
to control,3
-5,18,19
a procedure resulting in a material of a
Introduction
4
-7,20-24
heterogeneous and complex structure.
As indicated by
Homogeneous, corrosive liquid acid catalysts, such as BF3
complexes and H2SO4, are still frequently employed in the
production of (fine) chemicals.1 Replacement of the acids by
a solid catalyst is desirable to achieve more effective catalyst
handling and product purification and to diminish waste
production.1 In the last two decades, “sulfated metal oxides”
have received considerable attention as potentially benign solid
acid catalysts that possess a very high activity and even
the above equilibria, impregnation with a relatively large amount
of diluted sulfuric acid and drying leads to dissolved zirconium
sulfate, which crystallizes to Zr(SO4)2‚4H2O upon further drying.
The hydrated sulfate may be locally deposited within the porous
zirconia. With pore-volume impregnation reaction to Zr(SO4)2‚
,2
,3
4
H2O is less likely. The water remaining after drying will lead
to the presence of bisulfate (-OSO3H) or sulfuric acid on the
zirconia surface.
1
,3-5
superacidity.
Despite the promising properties, technical
1
Starting with zirconium sulfate of a low surface area, the
applications are still lacking, which is due to the activity being
difficult to control in view of, for example, the sensitivity to
above equilibria will only involve reaction to Zr(SO4)2‚4H2O
hydration,6 the thermal stability, and an interference of
,7
8,9
and Zr(SO ) . The water leading to the reaction to sulfuric acid
4 2
reduction processes.1
0-17
will be rapidly removed at a low temperature, where loss of
sulfate will not significantly proceed. Figure A4 of the Sup-
porting Information, which represents the phase diagram of
sulfuric acid, confirms the conclusion. Up to a temperature of
about 473 K, the gas phase only involves water. Only at higher
temperatures does the gas phase contain increasingly the
constituents of sulfuric acid, viz., H2O, SO3, SO2, and O2.
Because water-free bulk zirconium sulfate loses sulfate only at
In this research, we focused on sulfated zirconia, in which
the following equilibria are dominating: Zr(SO4)2 + 4H2O h
Zr(SO4)2‚4H2O + nH2O h ZrO2 + 2H2SO4‚aq. The usual
impregnation of zirconia with sulfuric acid, drying, and calcina-
tion brings about that the activity of sulfated zirconia is difficult
*
To whom correspondence should be addressed: Tel: +31-302533128.
6
Fax: +31-302534533. E-mail: jennesk@chem.uu.nl.
temperatures above 890 K, a rapid transport of water smoothly
1
0.1021/jp030436b CCC: $25.00 © 2003 American Chemical Society
Published on Web 11/08/2003

13404 J. Phys. Chem. B, Vol. 107, No. 48, 2003
Dijs et al.
CHART 1: Gas-Phase Sulfation of ZrO2 by H2S/O2,
SO2/O2, or SO3
CHART 2: Gas-Phase Trans-Alkylation of Benzene (1)
and Diethylbenzene (2), Giving Two Mol Equivalents of
Ethylbenzene (3) (Reaction 4), and the Liquid-Phase
Hydro-Acyloxy-Addition Reaction of Acetic Acid (4) to
Camphene (5), Giving Isobornyl Acetate (6) (Reaction 5)
leads to Zr(SO4)2. With a highly porous zirconia, on the other
hand, which has been impregnated with diluted sulfuric acid, a
sufficiently rapid transport of water out of the porous system
may call for temperatures thus high, so that the vapor phase
also contains significant amounts of the constituents of sulfuric
acid.
The possible loss of sulfate brings about that the catalytic
activity of sulfated zirconia does not depend on the final
temperature during drying and calcination. The heating rate is
highly important as well as the total amount of sulfated zirconia
thermally treated. Release of water and the constituents of
sulfuric acid from a large batch is much more difficult than
that from a small sample. Application of a gas flow can also
affect the final activity, and with large batches treated in a fixed
bed, an inhomogeneous catalyst may result.
through the pores of the zirconia will determine the loss of water
and sulfate of the impregnated zirconia.
The properties and catalytic behavior of the three differently
prepared catalysts are compared in the gas-phase trans-alkylation
reaction of benzene (1) and diethylbenzene (2) giving ethyl-
benzene (3) as well as in the liquid-phase hydro-acyloxy-
addition reaction of acetic acid (4) with camphene (5) giving
First of all, we established the catalytic activity of water-
free zirconium sulfate. In view of the low surface-to-volume
27-30
isobornyl acetate (6) as major product (Chart 2; reactions 4
2
,27,31-34
6
and 5
).
ratio of bulk zirconium sulfate, we prepared also water-free,
The gas phase reaction has been studied at 473 and 673 K.
At 473 K, volatilization of water will dominate, whereas at 673
K, both water and sulfuric acid will desorb. Because the
reactants and the reaction products of the gas-phase reaction
do not adsorb strongly on the surface of the (sulfated) zirconia,
transport limitations are not likely. Transport within the liquid
phase proceeds much slower than in the gas phase. The
accessibility of the sulfated zirconia species therefore can be
expected to dominate the activity in the liquid phase. The
finely divided silica-supported zirconium sulfate. The effect of
addition of water vapor on the activity of the water-free catalysts
was also investigated. To elucidate the influence of the
conditions during drying and calcination, we compared the
activity of conventionally prepared sulfated zirconia’s with that
of the initially water-free catalysts. Because the rate of transport
of water out of the porous system may severely affect the loss
of sulfate, we prepared two sulfated zirconia’s of a considerably
2
2
different porous structure (50 m /g and 217 m /g) according to
the conventional procedure, viz., impregnation with diluted
sulfuric acid, drying, and calcination.
2
sulfated zirconia of 50 m /g and the silica-supported sulfated
zirconia catalyst are therefore likely to exhibit the most elevated
activity with a liquid-phase reaction.
We recently reported about the preparation of anhydrous
6
Zr(SO4)2. Deposition-precipitation was employed to apply
Experimental Section
2
25
zirconia on a nonporous silica support (50 m /g). Three
different procedures (Chart 1) were investigated to react the
supported zirconia to water-free zirconium sulfate, viz., treat-
ment with H2S and subsequent with O2, reaction with SO2 and
O2, and reaction with SO3 produced by in situ oxidation of SO2
with O2 on a Pt/SiO2 catalyst. Thermodynamics (Ellingham
diagrams: see the Supporting Information Figure A1) indicates
that ∆G is slightly positive for the reaction to ZrS2 (reaction
General. ICP-AES (Thermo Jarrell Ash, Atomscan-16
equipped with a photomultiplier detector) was used to analyze
the elements Zr and Si. Prior to analysis, the sample was mixed
with Li B O and melted in a carbon cup to give a glass bead,
2
4
7
which was subsequently dissolved in concentrated HNO . Sulfur
3
(S) was analyzed as SO using a Carlo Erba NA-1500 (Fisons-
2
CE Instruments) NCS-analyzer equipped with a flash-combus-
1
a). An excess of H2S is thus required to get significant
tion chamber, a Porapack QS 1.83 m × 4 mm column and a
TCD. For flash-combustion GC analyses, the samples were
mixed with V O in a small tin container, sealed, and subse-
sulfidation of ZrO2. Reaction 1b, oxidation to the corresponding
sulfate, is thermodynamically much more favorable. Neverthe-
less, sulfation by treatment with either SO2/O2 or SO3/SO2/O2
2
5
quently combusted in an excess of O (flash-combustion). With
2
(
reactions 2 and 3) is preferred, because they can be performed
a He flow, the evolved SO was passed through a reduced Cu
3
26
in one step without introducing water.
column to reduce SO to SO . Before separation on the GC
3
2
The prepared catalysts were thoroughly characterized by
XRD, XPS, SEM-EDX, and elemental analysis. Thermogravim-
etry can provide important information about the reaction to
bulk of surface zirconium sulfate. Whereas volatilization of
sulfuric acid calls for a temperature above about 500 K, the
decomposition of zirconium sulfate has been indicated above
to proceed at much more elevated temperatures. When impreg-
nation with sulfuric acid, drying, and calcination does not lead
to zirconium sulfate, but due to the presence of water to
essentially supported sulfuric acid, release of the constituents
of sulfuric acid can be expected above about 500 K. Transport
column, the gases were dried over Mg(ClO4)2.
Powder XRD patterns were recorded with an Enraf/Nonius
FR 590 X-ray powder diffractometer equipped with a Co anode
(KR 1.78897 Å) and an Inel CPS120 (Ar/C2H4) curved position-
sensitive detector; a reflectrometry-sample-holder was used.
After subtraction of the scattering background, the volume-
averaged diameter of crystallites was calculated from the fwhm
using the Scherrer formula (see the Supporting Information; eq
3
5
A1).
Specific surface areas of the catalysts were determined by
measuring N2 adsorption-desorption isotherms at 77 K using

Bulk and Silica-Supported ZrO2
J. Phys. Chem. B, Vol. 107, No. 48, 2003 13405
a Micrometrics ASAP 2400 apparatus (equilibration interval 15
s) according to the BET approach.36 Prior to the measurements,
the samples (ca. 1 g) were evacuated at 393 K for 16 h.
XPS analysis was performed on a Vacuum Generators (Fisons
Instruments) MT-500 with a nonmonochromatic Al X-ray source
injected via narrow tubes (i.d. ) 1.0 mm) ending below the
level of the liquid using two Gilson Minipuls III peristaltic
pumps. Whereas the 4.0 M HCl was injected at a rate of
0.25 mL/min, the injection of the 4.0 M NH3 solution was
automatically regulated to maintain a pH of 4.5. When a constant
pH of 4.5 was reached, the 4.0 M HCl was replaced by a 4.0 M
HCl (250 mL) solution containing 3.92 g (12.2 mmol) of
ZrOCl2‚8H2O. After addition of these solutions, the pH was
raised to 6.5 with the 4.0 M NH3 solution. The workup and
calcination procedures for the 10 wt % ZrO2/SiO2 dispersion
were the same as those for the preparation employing the
hydrolysis of urea described above. The 10 wt % ZrO2/SiO2
catalyst precursor was removed from the calcination reactor and
stored under dry air.
(KR 1486.6 keV) and a CLAM-2 hemispherical analyzer for
electron detection. The samples were supported on carbon
adhesive tape. Spectra were corrected for charging using the
Si(2p) peak and scaled on the Si(2s) peak. For the determination
37
of the binding energies, a background correction was applied.
The thermal stability of the catalysts was assessed by
analyzing the relative loss of weight in a dry N2 flow (50 mL/
min) as a function of temperature and time with a (PC-
controlled) Perkin-Elmer TGS-2 TGA apparatus, autobalance
AR-2. Temperature program: 1 h at 323 K, heating rate 10
K/min to 1123 K followed by 15 min at 1123 K. Samples of
ca. 3.5 mg were used.
TEM analysis was performed with a Philips EM420 electron
microscope and a Philips CM200 transmission electron micro-
scope equipped with an EDAX detector for elemental analysis.
Copper grids covered by a thin polymer film on which carbon
was deposited were used as supports for the ground and
ultrasonically dispersed (in dry n-hexane) samples.
A Philips XL 30 FEG scanning electron microscope equipped
with an EDAX detector for elemental analysis provided SEM
images of the catalysts and the local element composition. The
samples were supported on carbon adhesive tape and covered
with a carbon layer by vapor deposition for electrical conduc-
tance.
Gas-Phase Sulfation of 10 wt % ZrO2/SiO2 Catalyst
Precursors. Three different gas-phase sulfation procedures were
used: treatments of 10 wt % ZrO2/SiO2 with (i) H2S followed
by oxidation with O2 (acronym: H2S/O2), (ii) SO2 and O2
(acronym: SO2/O2), and (iii) SO2 and O2, but with a Pt/SiO2
catalyst to generate SO3 in situ (acronym: SO3/SO2/O2). Gas
purities were He 99.995%, H2S 99.6%, SO2 99.9%, O2 99.999%,
and N2 99.999%. The gas-phase sulfation reactions were
performed in a quartz fixed-bed reactor (i.d. 10 mm) equipped
with a K-type thermocouple. Helium (He) was used to maintain
a gas flow of 50 mL/min. Glass beads (d ) 1.0 mm) were placed
on top of the silica-supported metal oxide particles (500-850
µm sieve fraction) to achieve effective preheating of the gas
flow. Via a heat-traced tube (423 K), the gas-flow leaving the
reactor was passed through a stirred suspension of Ca(OH)2 in
water. For a schematic representation of the reactor see the
Supporting Information (Figure A2). The treated 10 wt % ZrO2/
SiO2 catalysts were removed from the reactor and stored under
dry air prior to analysis and application.
Catalyst Preparation. Two homogeneous deposition-pre-
cipitation methods were evaluated for the preparation of 10 wt
%
ZrO2/SiO2 catalyst precursors.
Precipitation of ZrO2 on Silica by Hydrolysis of Urea. A
reaction vessel (2 L) equipped with a pH meter, thermometer,
baffles, and a mechanical stirrer (1000 rpm) was charged with
(
1) H2S/O2: 0.7 g of the 500-850 µm sieve fraction of the
1
0 wt % ZrO2/SiO2 precursors was heated under He to 623 K
at 5 K/min. Subsequently, H2S (2.5 vol %) was passed through
the reactor for 2 h, after which the reactor was purged with He
for 30 min. Next, O2 (20 vol %) was passed through for 2 h,
after which the catalyst was cooled to room temperature at 5
K/min under a He flow (50 mL/min).
1
3.5 g (0.225 mol) of SiO2 [Aerosil OX50 (Degussa-H u¨ ls),
2
surface area 50 m /g], 4.0 M HCl (500 mL), and 3.92 g (12.2
mmol) of ZrOCl2‚8H2O, dissolved in 4.0 M HCl (250 mL).
Subsequently, 185 g (3.08 mol) of urea dissolved in 4.0 M HCl
(250 mL) was added, and the reaction mixture was heated to
(2) SO2/O2: 0.7 g of the 10 wt % ZrO2/SiO2 precursors (500-
3
63 K. At this temperature, hydroxyl ions are homogeneously
38,39
850 µm sieve fraction) was heated under He to 673 K at 5
K/min. Next, SO2 (5 vol %) and O2 (5 vol %) were passed
through the reactor for 5.5 h after which the temperature was
lowered to 573 K. Subsequently, the temperature was lowered
to room temperature at 5 K/min under a He flow (50 mL/min).
(3) SO3/SO2/O2: With the exception of the addition of a layer
of a finely powdered 2 wt % Pt/SiO2 oxidation catalyst at half-
height of the glass bead layer, the SO2/O2 procedure was applied.
The oxidation catalyst converts part of the SO2 into SO3 (50%
based on BaSO4 precipitation and UV-Vis analysis) before
reaching the 10 wt % ZrO2/SiO2.
generated by the hydrolysis of urea.
After 3 days, a constant
pH of 6.0 was reached (measured separately at 293 K) and the
suspension was filtered. Because the resulting dispersion of ZrO2
on the silica support can be affected by calcination (both water
and salts may react with the precipitated material to give volatile
4
0
compounds), the wet residue was re-suspended three times in
water (300 mL) for 1 day followed by filtration in order to
remove remaining NH4Cl impurities. The final residue was dried
at 393 K for 1 day and subsequently sieved; the 500-850 µm
fraction was isolated. Calcination was performed under carefully
controlled conditions using a quartz fixed bed reactor (i.d. 10
mm) equipped with a K-type thermocouple. The 500-850 µm
sieve fraction was calcined at 723 K for 10 h (heating/cooling
rate: 5 K/min) in a dry air flow (50 mL/min). On top of the
sieve fraction, glass beads (d ) 1.0 mm) were placed to achieve
effective preheating of the air flow. The 10 wt % ZrO2/SiO2
catalyst precursor was isolated from the calcination reactor and
stored under dry air.
Calcined H2SO4-Impregnated ZrO2 Catalysts. Calcined
H2SO4-impregnated ZrO2 catalysts (acronym: H2SO4/ZrO2)
1
9
were prepared following a literature procedure; two different
ZrO2 sources were used.
(1) Precipitated ZrO2 [acronym: ZrO2(prec.)] 20.23 g (62.8
mmol) of ZrOCl2‚8H2O was dissolved in water (160 mL), and
under vigorous stirring, 25 wt % NH3 was added dropwise until
pH ) 8. Next, NH4Cl was removed by four times re-suspending
the wet residue in water (300 mL) for 1 day followed by
filtration. The ZrO2(prec.) was dried at 433 K for 16 h.
(2) Commercial ZrO2: Gimex B. V. [acronym: ZrO2-
pH-Static Precipitation of ZrO2 on Silica. In a reaction vessel
(2 L) equipped with a pH-meter, thermometer, baffles, and a
stirrer (1000 rpm), 13.5 g (0.225 mol) of SiO2 [Aerosil OX50
2
(
Degussa-H u¨ ls), 50 m /g] was suspended in 750 mL of water.
2
Under stirring, both 4.0 M HCl and 4.0 M NH3 were separately
(Gimex)], The Netherlands, surface area 60 m /g, monoclinic.

1
3406 J. Phys. Chem. B, Vol. 107, No. 48, 2003
Dijs et al.
Sulfation of both zirconia’s was performed by stirring 4.0 g
Results and Discussion
(
32.46 mmol) of each zirconia sample with 0.5 M H2SO4 (20
Highly Dispersed Silica-Supported ZrO2; Preparation and
Characterization. Because our approach starts with the prepa-
ration of highly dispersed ZrO2 on a nonporous silica support
mL) for 3 h and drying at 433 K for 16 h (no filtration). Sieve
fractions (500-850 µm) were calcined in a flow of dry air (50
mL/min) at 773 K for 3 h (heating/cooling rate: 10 K/min).
The catalysts were stored under dry air.
(ZrO2/SiO2), which will subsequently be subjected to gas-phase
sulfation reactions, two different aqueous precipitation methods
were compared, viz., urea hydrolysis and a pH-static procedure.
With both methods, hydroxyl ions are homogeneously generated
in a suspension of silica in a solution of a zirconium salt.
Deposition-precipitation requires that the silica particles will
function as nuclei for (hydrated) ZrO2 precipitation.41 An
important difference between the urea hydrolysis and the pH-
static procedure is the pH range in which precipitation of
zirconia species proceeds. With urea hydrolysis, ZrO2 precipita-
tion occurs at pH ∼ 2-3, whereas under pH-static conditions,
Catalysis. Gas-Phase trans-Alkylation of Benzene (1) and
Diethylbenzene (2). The gas-phase trans-alkylation reaction was
performed in an automated micro-flow apparatus containing a
quartz fixed-bed reactor (i.d. 10 mm) at 10 Pa [16 vol %
benzene (1 p.a., dried on molsieve), 3.2 vol % diethylbenzene
5
(
2, consisting of 25% ortho, 73% meta, and 2% para isomers,
-1
dried on molsieve), N2 balance (50 mL/min), WHSV ) 1.5 h ]
with 2.0 mL of the tube reactor filled with catalyst particles
(
500-850 µm sieve fraction, typically 1.4 g). Two saturators
42
precipitation necessarily proceeds at pH ) 4.5. The iso-electric
point of silica is at pH ∼ 2-3, which brings about that, at pH
levels above 3, silica is negatively charged. Adsorption of
zirconyl ions and subsequent nucleation of (hydrated) zirconia
species on negatively charged silica present at pH levels above
were connected to the inlet of the reactor for the supply of 1
and 2. The partial vapor pressure of 1 and 2 was controlled by
adjusting the temperature of the saturator-condensers and the
N2 flow rate. After equilibration for 30 min at the applied
reaction temperatures (473 and 673 K, heating rate 10 K/min)
within a dry N2 flow (50 mL/min), benzene (1) and diethyl-
benzene (2) were passed through the reactor. To prevent
condensation of reactants and products prior to GC analysis
4
3
3
is more likely than at lower pH levels. Hence, it can be
expected that the pH-static method will lead to a better
dispersion of ZrO2 on the silica surface.
The bulk composition of the ZrO2/SiO2 catalyst precursors
prepared according to either the urea hydrolysis or the pH-static
method was determined by elemental analyses (ICP-AES Zr
and Si). Independent of the applied precipitation procedure,
similar bulk compositions were found [Zr ) 6.9 wt. %, Si )
[Hewlet Packard 5710 A, column: CP-sil 5CB capillary liquid-
phase siloxane polymer (100% methyl) 25 m × 0.25 mm, 323
K, carrier gas: N2, FID, sample-loop volume: 1.01 µL], tubes
were heat-traced (398 K). FID sensitivity factors and retention
times were determined using ethene (99.5%, dried over mol-
sieve) and standard solutions of 1, 2, and ethylbenzene (3, 99%)
in methanol (p.a.). The conversion of the feed (compounds 1
and 2) was determined using eq A2 (see the Supporting
Information). Notice that the conversion of the feed (compounds
3
7 wt. %, residual wt. % ) O], which are in reasonable
agreement with the calculated values [Zr ) 7.4 wt. %, Si ) 42
wt. %, residual wt. % ) O].
To establish whether a crystalline phase of ZrO2 has formed
within the calcined ZrO2/SiO2 catalyst precursors and to obtain
an estimate of the average crystallite size, X-ray diffraction
1
and 2) is about 6 times lower than the (partial) conversion of
diethylbenzene (2) due to the molar ratio 5:1 of benzene (1)
and diethylbenzene (2).
(XRD) was employed. The ZrO2/SiO2 precursors prepared by
the urea hydrolysis or the pH-static procedure only showed a
background halo due to X-ray scattering of the amorphous silica
support (see the Supporting Information, Figure A3).44 Hence,
the deposited ZrO2 particles are amorphous. The calcined
ZrO2/SiO2 precursors were subsequently studied by trans-
mission electron microscopy (TEM). Despite their similar bulk
compositions, TEM shows that the pH-static method leads to
supported ZrO2 particles that are decisively smaller (better
dispersion on the silica spheres) than those prepared by urea
hydrolysis (Figure 1).
X-ray photoelectron spectroscopy (XPS) was used to assess
global differences in the surface composition of the calcined
ZrO2/SiO2 catalyst precursors. The Zr(3d5/2,3/2)/Si(2s)37 peak-
area ratio is the highest for samples prepared by the pH-
static precipitation method: Zr(3d5/2,3/2)/Si(2s)pH-stat ) 2.5 vs
Zr(3d5/2,3/2)/Si(2s)urea ) 0.7 (Figure 2a). Because the bulk
compositions of the precursors prepared by the pH-static
precipitation method and by urea hydrolysis are identical, the
XPS results also indicate that the pH-static precipitation method
leads to a higher dispersion of ZrO2 over the silica surface. Thus,
TEM and XPS confirm the conclusion that the interaction
between the suspended silica support and the precipitating
zirconia species depends on the pH. The availability of more
nucleation sites during deposition-precipitation will lead to a
metal oxide more highly dispersed on the surface of the silica
Liquid-Phase Hydro-Acyloxy-Addition of Acetic Acid (4) to
Camphene (5). A mixture of glacial acetic acid (4 p.a., 0.70
mol), camphene (5, 95%, 0.70 mol) and acetic anhydride (p.a.,
9
3
was quickly suspended in the reaction mixture. The composition
of the soluble fraction of the reaction mixture was analyzed by
capillary GC as a function of reaction time; samples were
prepared as follows: 1.00 mL of the reaction mixture was added
to water (25.00 mL) followed by an extraction with n-heptane
.05 mmol) was mechanically stirred (1500 rpm) overnight at
28 K under a N2 atmosphere. Subsequently, 2.5 g of catalyst
(25.00 mL). 1.00 mL of the n-heptane fraction was diluted with
n-heptane to 25.00 mL in a volumetric flask. 1.0 µL of the
diluted solution was injected into the GC [Varian 3400,
column: DB-5 capillary liquid-phase siloxane polymer (5%
phenyl, 95% methyl), 30 m × 0.323 mm, temperature pro-
gram: 5 min at 333 K, 10 K/min to 553 K, 10 min, carrier gas:
N2, FID]. In the case of hydro-acyloxy-addition reactions
performed in the presence of H2O, 320 µL (17.78 mmol) H2O
was added instead of acetic anhydride (vide supra). To establish
whether leaching occurs, the insoluble catalyst particles were
removed from the reaction mixture by filtration with a double-
ended glass filter under a dry N2 atmosphere before equilibrium
had established. The composition of the reaction mixture was
further measured as a function of time. Since no solid residue
was found after evaporation of the reaction mixture to dryness
in vacuo, the removal of the solid catalysts particles by filtration
was complete.
support. By adjusting the pH, the dispersion of the ZrO can be
2
consequently controlled.
Because the gas-phase sulfation reactions are expected to be
more effective for smaller ZrO2 particles, the ZrO2/SiO2 catalyst

Bulk and Silica-Supported ZrO2
J. Phys. Chem. B, Vol. 107, No. 48, 2003 13407
2 2
Figure 1. TEM photographs of 10 wt % ZrO /SiO prepared by (a)
urea hydrolysis and (b) the pH-static method. Large light-gray
spheres: silica (Aerosil OX50) support material. Dark dots: metal oxide
clusters.
Figure 2. (a) XPS spectra of 10 wt % ZrO /SiO prepared by the pH-
static method (highest Zr/Si ratio) and urea hydrolysis (lowest Zr/Si
2
2
ratio). (b) XPS spectra of 10 wt % ZrO
2
/SiO
2
(blank) and 10 wt %
, or SO /SO /O gas-
ZrO /SiO subjected to either the H S/O , SO
2
2
2
2
2
/O
2
3
2
2
precursors prepared by the pH-static precipitation procedure
were used for the gas-phase treatments.
phase approach. [The spectra have been corrected for charging, scaled
on the Si(2s) peak and vertically displaced for presentation.]
Characterization of Gas-Phase Sulfated ZrO2/SiO2. XRD,
IR, XPS, and TEM-EDAX were used to evaluate the effective-
ness of the three different gas-phase sulfation procedures (H2S/
O2, SO2/O2, and SO3/SO2/O2, see the Experimental Section).
After gas-phase sulfation, no changes were detected in the XRD
pattern of the ZrO2/SiO2 catalysts. Unfortunately, diffuse-
reflectance FT-IR is unsuitable to verify the formation of surface
sulfides and sulfates due the interference of the intense Si-O
absorption bands. XPS analysis, however, provided more
information. Using XPS atomic sensitivity factors [S(2p) ) 0.54
and Zr(3d5/2,3/2) ) 2.1]45 and the experimental S(2p)/Zr(3d5/2,3/2)
peak-area ratio, surface atom ratios can be calculated. For bulk
the SO3/SO2/O2-treated ZrO2/SiO2 precursor (184.1 eV) in
combination with the observed additional broadening of the
Zr(3d5/2,3/2) peak (0.3 eV ∆FWHM) indicates that both ZrO2
and Zr(SO4)2 are present within the surface layer. This is
corroborated by the S(2p)/Zr(3d5/2,3/2) peak-area ratio, which
was found to be 0.25 (SO3/SO2/O2 treatment) instead of 0.50,
6
which was found for bulk anhydrous Zr(SO4)2.
For the conventional, calcined H2SO4-impregnated ZrO2
catalysts (H2SO4/ZrO2), a S(2p) binding energy of 169.3 eV
has been reported, which deviates significantly from the S(2p)
binding energy of bulk anhydrous Zr(SO4)2 (170.3 eV).6
Interestingly, the S(2p) binding energy in H2SO4/ZrO2 is nearly
identical to that of liquid H2SO4 (169.4 eV, see also the
Supporting Information Table A1). In addition, the Zr(3d5/2,3/2)
binding energy of the H2SO4/ZrO2 catalysts is mainly deter-
mined by the contribution of ZrO2, especially after calcination
,18,51
anhydrous Zr(SO4)2 (prepared from ZrCl4 and fuming H2SO4/
6
1
0% SO3), an S(2p)/Zr(3d5/2,3/2) peak-area ratio of 0.50 was
37
previously found, which leads to an S/Zr atom ratio of 2.0 in
line with the chemical composition. Figure 2b shows the XPS
spectra of the ZrO2/SiO2 and the gas-treated ZrO2/SiO2 catalysts.
Gas-phase treatment using H2S/O2 led to a S(2p) peak of low
intensity positioned at 170.0 eV. For the ZrO2/SiO2 catalysts
that were subjected to either SO2/O2 or SO3/SO2/O2, gas-phase
sulfation S(2p) binding energies of 170.0 eV were found, which
are in good agreement with the S(2p) binding energy of
1
8,24
at high temperatures (923 K).
For H2SO4/ZrO2 catalysts
calcined at lower temperatures (773 K), a broadened Zr(3d5/2,3/2)
peak is found, which was ascribed to the presence of primarily
2
4,43
ZrO2 within the surface layer.
Note that the impregnated
zirconia’s used in the catalytic experiments were calcined at
773 K. Deconvolution of the broadened Zr(3d5/2,3/2) peak
resulted in Zr(3d5/2) and Zr(3d3/2) binding energies still lower
than the Zr(3d5/2) and Zr(3d3/2) binding energies of Zr(SO4)2‚
anhydrous bulk Zr(SO4)2 [170.3 eV, see the Supporting
Information (Table A16
,45-50
)]. These results indicate that
(partial) conversion of the surface layer of ZrO2 into Zr(SO4)2
18,24
has indeed occurred. This is further substantiated by the
concomitant shift of the Zr(3d5/2,3/2) binding energy from 183.3
eV (blank) to 184.1 eV [SO3/ZrO2/SiO2 (Figure 2b)].3 For bulk
anhydrous Zr(SO4)2, however, a Zr(3d5/2,3/2) binding energy of
4H2O (184.5 and 186.5 eV, respectively).
Hence, calcination
at 773 K of conventionally prepared, H2SO4-impregnated/ZrO2
catalysts and subsequent exposure to ambient air apparently does
not result in the presence of either Zr(SO4)2‚4H2O or Zr(SO4)2
within the surface layer.
7
1
85.6 eV was found. The lower Zr(3d5/2,3/2) binding energy for

13408 J. Phys. Chem. B, Vol. 107, No. 48, 2003
Dijs et al.
thermore, the sulfur-to-zirconium ratio is much higher for the
supported catalyst, as has to be expected.
Thermal Stability of SO3/ZrO2/SiO2 and H2SO4/ZrO2. An
important distinction between the SO3/ZrO2/SiO2 catalysts
(prepared via the gas-phase procedure) and the water-free bulk
zirconium sulfate, on one hand, and the usual sulfated zirconia
catalysts (H2SO4/ZrO2; conventionally prepared by impregnation
with H2SO4), on the other hand, is the presence of H2SO4. The
XPS results point to sulfuric acid in the surface layer of the
sulfated zirconia’s. Sulfuric acid can be present due to an
incomplete reaction with the surface of the impregnated zirconia
or hydrolysis of zirconium sulfate by residual water or water
from ambient air. Anhydrous bulk Zr(SO4)2 exhibits an onset
6
temperature of decomposition in a dry N2 flow of 890 K, while
the constituents of sulfuric acid are significantly present in the
gas-phase already above a temperature of 473 K (see the
5
3-56
Supporting Information; Figure A4
that, for compositions of a mol fraction of SO3 below 0.5
diluted H2SO4) at temperatures below 473 K, only water will
). Figure A4 indicates
(
5
7
be released into the gas flow. Hence, heating from room
temperature will cause water to evaporate and, consequently,
lead to an increase in concentration of the H2SO4. Above 473
K, however, also significant amounts of the constituents of
H2SO4 will be present in the gas phase. Sulfuric acid present in
the catalysts, consequently, will be apparent from a loss in
weight at temperatures appreciably lower than 890 K. The SO3/
ZrO2/SiO2 and the sulfated zirconia (H2SO4/ZrO2) catalysts were
therefore subjected to thermogravimetric analysis. The thermal
stability of the sulfated metal oxide catalysts will furthermore
indicate the temperature range in which the catalysts can be
expected to remain catalytically active.
Figure 3. Thermogravimetric analysis of (a) SiO
and (b) ZrO (Gimex) and ZrO (prec.) and corresponding (calcined)
SO -impregnated catalysts. [Temperature program: 30 min at 323
2 3 2 2
and SO /ZrO /SiO
2
2
H
2
4
K, 10 K/min to 1100 K and 10 min at 1100 K.]
TABLE 1: ICP-AES and Flash-Combustion GC Results for
2 2
the 10 wt % ZrO /SiO Catalyst Precursor Materials
Subjected to the Gas-Phase SO
3 2 2
/SO /O Treatment and for
2
a
the Two Conventional H SO /ZrO
2
4
Catalyst Materials
The thermal stability of silica and (nonsulfated) zirconia was
measured by TGA (N2) (Figure 3). Whereas the silica support
and ZrO2(Gimex) only eliminated a minute amount of water,
the water content of the ZrO2(prec.) is appreciably higher as is
evident from the relatively large decrease in weight. Figure 3
also represents the weight loss of the SO3/ZrO2/SiO2 and sulfated
zirconia (H2SO4/ZrO2) catalysts, which was also measured in a
flow of nitrogen. The SO3/ZrO2/SiO2 catalyst exhibited a small
continuous decrease in weight (Figure 3a). Decomposition of
the sulfate smoothly follows the loss of water, which proceeds
at low temperatures. Elimination of water from the SO3/ZrO2/
SiO2 catalysts must originate from the silica support, because
the XPS results did not indicate the presence of H2SO4 (vide
supra). However, the thermal stability of Zr(SO4)2 in the
atom (%)
BET
calc. (100% ICP-AES/flash- area
2
sample
/ZrO /SiO
element sulfated) combustion GC (m /g)
SO
3
2
2
S
Zr
Si
3.0
1.5
1.7
1.3
50
27.6
23.7
H
2
2
SO
SO
a
4
/ZrO
2
(prec.)
S
Zr
S
Zr
7.3
23.9
7.3
23.9
4.0
23.9
1.8
24.2
217
50
H
4
/ZrO
2
(Gimex)
Residual atom (%): oxygen.
With TEM-EDAX, it was established that the initial disper-
6
presence of water vapor is much lower. Hence, for the SO3/
sion of ZrO2 on the silica support (Figure 1) did not change
during the subsequent gas treatments. This result is in line with
the near constant XPS Zr(3d5/2,3/2)/Si(2s) peak intensity ratio
found (Figure 2b). The results furthermore show that the SO3/
SO2/O2 treatment (Chart 1; reaction 3) appears to be the most
effective for the preparation of highly dispersed silica-supported
zirconium sulfates. Therefore, this gas-phase sulfated catalyst
was further characterized and its catalytic properties were
evaluated.
ZrO2/SiO2 catalyst materials, formation of a small amount of
H2SO4 or bisulfate groups (-OSO3H) by reaction with water
from the silica support cannot be excluded.
The conventional sulfated zirconia (H2SO4/ZrO2) catalysts
showed a much larger weight loss (Figure 3b). The TGA curves
for the zirconia’s and the H2SO4/ZrO2 catalysts are identical to
2
4
those previously published. Figure 3 shows that a continuous
decrease in weight sets on below 373 K. The initial weight loss
is attributed to the removal of water. It is interesting that the
curves for precipitated zirconia and the sulfated precipitated
zirconia run parallel up to a temperature of about 530 K. Up to
Table 1 presents the results of the elemental analyses as well
as the specific surface areas (BET method) of the SO3/ZrO2/
SiO2 catalyst and the two sulfated zirconia catalysts. The results
show that the amount of sulfur (at. %) and the specific surface
area are higher for the sulfated zirconia (H2SO4/ZrO2) prepared
from freshly precipitated zirconia than for that prepared from
5
30 K, desorption of water from the sulfated precipitated
zirconia is consequently likely. At higher temperatures the
weight loss of the sulfated zirconia (H2SO4) is slightly higher.
Above 530 K, the sulfated zirconia’s (H SO /ZrO ) can release
2
4
2
52
the Gimex zirconia and the SO3/ZrO2/SiO2 catalyst (Table 1).
water from grafted H SO or bisulfate groups (-OSO H), but
2
4
3
also the constituents of sulfuric acid will be volatilized.6
,8,9,24
However, the amount of sulfur per unit surface area is larger
for the sulfated zirconia prepared from Gimex zirconia. Fur-
The catalyst prepared from the Gimex zirconia showed a steep

Bulk and Silica-Supported ZrO2
J. Phys. Chem. B, Vol. 107, No. 48, 2003 13409
TABLE 2: Elemental Analysis (S/Si Atom Ratio) by
ICP-AES and Flash-Combustion GC (Global Bulk
Composition), XPS (Global Surface Composition), and
elemental analyses (Flash-combustion GC and ICP-AES, XPS,
and SEM-EDAX) of the SO3/ZrO2/SiO2 catalyst before and after
prolonged exposure to steam. Comparison of the S/Si atom ratios
of the SO3/ZrO2/SiO2 catalyst before and after the steam-
treatment consistently indicates that the catalyst previously
calcined in dry air at 723 K looses sulfate upon exposure to
water vapor at 673 K. The larger drop in sulfate content apparent
from the XPS and SEM-EDAX measurements suggests that the
external surface layer of the catalyst looses the sulfate more
rapidly than the walls of the pores within the catalyst grains.
The effect of the rate of transport was also evident in the
thermogravimetric experiments. The data of Table 2 suggest
that sulfated zirconia catalysts can be employed in gas flows
containing steam at temperatures up to about 473 K without a
rapid loss in activity. Immersion of the catalyst in liquid water
of 293 K for 1 h resulted in a large loss of sulfate. In the
presence of liquid water, sulfated zirconia therefore is not a solid
acid catalyst but produces dissolved sulfuric acid.
SEM-EDAX (Local Bulk Composition) of SO
3
/ZrO
2 2
/SiO ,
after Treatment with H O(g) at 473 and 673 K for 10 h
2
Followed by Resuspension in H
2
O
(l) (293 K, 1 h)
S/Si (atom ratio)
SEM-EDAX
flash-
treatment of
SO /ZrO /SiO
combustion
GC & ICP-AES XPS
multiple point analysis,
20-point average
3
2
2
0
.072
0.22
0.18
0.06
0.05
0.024
0.019
0.008
0.006
H
2
H
2
H
2
O
O
O
(g), 473 K
(g), 673 K
(l), 293 K
0.073
0.059
0.029
weight loss setting on at about 800 K and that prepared from
the precipitated zirconia only above about 910 K. The weight
loss at about 890 K for the H2SO4/ZrO2 catalysts (Figure 3b)
and the SO3/ZrO2/SiO2 catalyst (Figure 3a) is also observed for
6
bulk anhydrous Zr(SO4)2 (within a dry N2 flow). The steep
Catalytic Activity of Conventional H2SO4/ZrO2 vs SO3/
ZrO2/SiO2 in the Gas-Phase trans-Alkylation of Benzene (1)
and Diethylbenzene (2). To assess the applicability of our
different sulfated zirconia catalysts in gas-phase reactions, the
trans-alkylation reaction of benzene (1) and diethylbenzene (2,
weight loss indicates that Zr(SO4)2 has been formed in the
sulfated zirconia (H2SO4/ZrO2) catalysts. The zirconium sulfate
can be present at temperatures up to about 530 K underneath a
layer of sulfuric acid. The sulfuric acid has desorbed at the
temperature of the steep weight loss. The sulfuric acid may have
remained after the impregnation, drying, and calcinations for 3
h at 773 K. However, the surface of the sulfated zirconia is
highly hygroscopic. Water taken up during transport and storage
may hydrolyze the zirconium sulfate of the surface layer to
sulfuric acid and zirconia; this has to be prevented.
2
7-30
5
Chart 2; reaction 4)
was studied at 10 Pa in a fixed-bed
reactor with a 16 vol % benzene (1) and 3.2 vol % diethylben-
zene (2) feed. The conversion of the feed [benzene (1) and
diethylbenzene (2, see the Supporting Information, eq A2)] was
determined as a function of time. First, the catalytic performance
of the sulfated metal oxide catalysts in the gas-phase trans-
alkylation reaction of 1 and 2 was benchmarked at 473 and
The TGA curves of Figure 3 also reflect a pronounced
difference in the macroscopic structure (porosity) of the sulfated
zirconia (H2SO4/ZrO2) catalysts. A comparison of the curves
of the two catalysts, viz., H2SO4/ZrO2(prec.) and H2SO4/ZrO2-
6
73 K with an amorphous aluminosilicate (ASA) (Si/Al ) 12,
2
BET specific area ) 320 m /g) (Figure 4a). At 673 K, the
conversion is about eight times higher than the low conversion
at 473 K. The higher conversion at 673 K is not only due to
the effect of the higher temperature on the trans-alkylation
reaction but also to the (more temperature-dependent) formation
of ethene (see the Supporting Information; Figure A5). The
ethylbenzene (3) selectivity is 100% at 473 K but only 50% at
(Gimex), reveals that the elimination of water, the release of
the constituents of H2SO4, and the decomposition of Zr(SO4)2
proceeded at higher temperatures with the H2SO4/ZrO2(prec.)
catalyst. The surface area of the H2SO4/ZrO2(prec.) catalyst (217
2
m /g) is considerably larger than that of the H2SO4/ZrO2(Gimex)
2
catalyst (50 m /g). Transport of gas molecules interacting
6
73 K. The initial rapid deactivation of the catalyst at 673 K
significantly with the walls of the pores will proceed much
slower out of the more highly porous structure of the H2SO4/
ZrO2(prec.) catalyst.
may be due to oligomerization of ethene on the acid sites and
subsequent reaction to carbonaceous deposits.
Figure 4, parts b and c, represents the conversion of 1 and 2
at 473 and 673 K in the presence of the conventionally prepared
sulfated zirconia (H2SO4/ZrO2) catalysts. The H2SO4/ZrO2-
(Gimex) exhibited a negligible activity at 673 K and initially a
significant activity at 473 K (Figure 4c). However, the activity
dropped, and after 300 min, the activity was negligible. The
activity of the H2SO4/ZrO2(prec.) catalyst was at 473 K
considerable and stable (Figure 4b). In contrast to the behavior
of the reference ASA catalyst, the activity at 673 K of the
H2SO4/ZrO2(prec.) catalyst was much lower. When after the
reaction at 673 K the H2SO4/ZrO2(prec.) catalyst was cooled
to 473 K, the activity appeared to be completely lost; the
conversion was not significant.
Stability of SO3/ZrO2/SiO2 toward Water Vapor. Ad-
ditional support for the presence of grafted H2SO4 or bisulfate
groups (-OSO3H) in the conventional sulfated zirconia (H2-
SO4/ZrO2) catalysts comes from FT-IR analyses.6
,7,20-23
The
sulfated zirconia catalysts possess a heterogeneous surface
structure, which changes concomitant with temperature, pres-
sure, and/or the presence of water vapor. Above 473 K,
dehydration is accompanied by elimination of H2SO4 (see the
Supporting Information, Figure A4). This implies that hydration
can lead to an irreversible loss of sulfate at temperatures above
4
73 K under flow conditions. Such irreversible behavior was
6,58-62
observed for anhydrous bulk Zr(SO4)2.
With sulfated
zirconia catalysts being employed at temperatures above 473
K, consequently, elimination of corrosive constituents of sulfuric
acid can be expected above 473 K, especially if water vapor is
present. Because the active component of the catalysts will then
be lost, the activity of the catalysts will drop. Also contact of
sulfated zirconia’s with a flow of liquid water will lead to a
loss of sulfate by hydrolysis and dissolution of the resulting
sulfuric acid.
Figure 4d shows the activity of the SO3/ZrO2/SiO2 catalyst
at 473 K. It can be seen that the initial activity disappears within
80 min. To demonstrate that the initial activity is due to the
uptake of water during transport from ambient air, the catalyst
was exposed for 2 h to a flow of 50 mL/min of nitrogen
containing 2 vol % of water vapor at 473 K. From Figure 4d,
it can be seen that the activity of the catalyst exposed to water
vapor is higher and more stable. The activity now disappears
within about 350 min. The activity of the anhydrous bulk
Zr(SO4)2 is represented in Figure 4e, which indicates that the
To establish the loss of sulfate experimentally, the SO3/ZrO2/
SiO2 catalyst was exposed to a gas flow containing water vapor
for 10 h at 473 and 673 K. Table 2 represents the results of

13410 J. Phys. Chem. B, Vol. 107, No. 48, 2003
Dijs et al.
Figure 4. Conversion in the trans-alkylation of benzene (1) and diethylbenzene (2) with (a) ASA, (b) H
2
SO
Gimex); 0 473 K and ] 673 K. Conversion in the trans-alkylation of benzene (1) and diethylbenzene (2) at 473 K with (d) SO
e) bulk anhydrous Zr(SO ; 0 not hydrated and 4 hydrated (2 h, 2% H O, 50 mL/min). [Conversion of the feed (compounds 1 and 2) gives values
4
/ZrO
2
(prec.), and (c) H
2
SO
4
/ZrO
2
-
(
(
3
/ZrO
2
/SiO
2
and
4
)
2
2
that are about 6 times lower than the (partial) conversion of diethylbenzene (2) because the molar ratio benzene (1):diethylbenzene (2) is 5:1.]
anhydrous sulfate has no any catalytic activity in the gas-phase
trans-alkylation reaction of benzene (1) and diethylbenzene (2).
Accordingly, Lewis acid sites catalyzing the trans-alkylation are
not available in water-free bulk zirconium sulfate. In contrast
to the SO3/ZrO2/SiO2 catalyst, prehydration of anhydrous bulk
Zr(SO4)2 with an sub-stoichiometric quantity of water vapor did
not lead to a significant conversion. That no catalytic activity
is induced with the water-free bulk sulfate is due to the formation
of a rather stable tetrahydrate [Zr(SO4)2‚4H2O], which sup-
because the sulfuric acid has been removed from the sites that
are most readily accessible from the gas phase.
The above results indicate that water is required for H2SO4
within H2SO4/ZrO2 and SO3/ZrO2/SiO2 catalysts to be catalyti-
cally active. H2SO4 is not released below 473 K in a gas flow
and will remain on the surface of the catalyst (see the Supporting
Information Figure A4 and Table 2). Nevertheless, the H2SO4/
ZrO2(Gimex) catalyst lost the activity within 300 min at 473 K
and the SO3/ZrO2/SiO2 catalyst within 80 min. Consequently,
water can only be released in the gas flow, which proves that
water is required to activate sulfuric acid present on the surface
of the catalysts. Presumably, water is required to shift the
equilibrium of the hydrolysis of zirconium sulfate to the
zirconium oxide and sulfuric acid side. When water is removed,
the equilibrium shifts to the inactive zirconium sulfate side.
In liquid-phase reactions, contact with liquid water may lead
to dissolution of sulfuric acid, which becomes subsequently
active as a homogeneous catalyst. The difference in catalytic
activity between the H2SO4/ZrO2(prec.) and the H2SO4/ZrO2-
(Gimex) catalyst in the gas-phase trans-alkylation was attributed
to a difference in the rate of transport through the porous
structure of the catalysts. Because transport within liquids is
much slower than that within the gas phase, a different rate of
transport will also affect the apparent activity in the liquid phase.
The above prompted us to study the H2SO4/ZrO2 and SO3/ZrO2/
SiO2 catalysts in a liquid-phase reaction.
6
presses the hydrolysis reaction of anhydrous bulk Zr(SO4)2.
The fact that water vapor is required to activate the silica-
supported water-free zirconium sulfate catalyst is in line with
H2SO4 to be the catalytically active species within H2SO4/ZrO2
catalysts. After calcinations for 3 h at 773 K, the H2SO4/ZrO2-
(Gimex) catalyst has lost the water and sulfuric acid almost
completely leaving behind a surface layer containing zirconium
sulfate. During transport and loading into the reactor, the catalyst
takes up water, which will react with the surface of the
zirconium sulfate to zirconium bisulfate and sulfuric acid. At
473 K, it takes about 300 min for the water to evaporate, which
leaves an inactive zirconium sulfate surface. The release of water
and the constituents of sulfuric acid from the relatively open
structure of the H2SO4/ZrO2(Gimex) catalyst proceeds so rapidly
at 673 K that no measurable activity is exhibited. Calcination
for 3 h at 773 K is not sufficient to remove the sulfuric acid
and some water completely out of the highly porous structure
of the H2SO4/ZrO2(prec.) catalyst. Accordingly, the catalyst
exhibits a substantial activity at 473 K (Figure 4b). Keeping
the H2SO4/ZrO2(prec.) catalyst in a gas flow at 673 K rapidly
removes the water and the constituents of sulfuric acid from
the easily accessible sites within the catalyst structure. The
remaining activity is consequently low. Subsequent measure-
ment of the activity at 473 K results in a negligible activity,
Stability of SO3/ZrO2/SiO2 and H2SO4/ZrO2 toward
Liquid Water. Table 2 also mentions the effect of the contact
with liquid water on the sulfate content of the SO3/ZrO2/SiO2
catalyst. After stirring the SO3/ZrO2/SiO2 catalyst in water for
1 h, followed by filtration and drying (393 K), elemental analysis
(flash-combustion GC and ICP-AES, XPS, and SEM-EDAX)
revealed that the S/Si atom ratio decreases by a factor of 2.45

Bulk and Silica-Supported ZrO2
J. Phys. Chem. B, Vol. 107, No. 48, 2003 13411
Figure 5. Formation of isobornyl acetate (6) for the hydro-acyloxy-
addition of 0.70 mol glacial acetic acid (4) to 0.70 mol camphene (5)
at 338 K (stirred tank-reactor, N
2
atmosphere); 4 no catalyst, 0 8.75
SO
mmol of BF /HOAc, and ] 1.65 mL of 5.3 M H
3
2
4
.
(Table 2). In addition, the filtrates from aqueous suspensions
of the SO3/ZrO2/SiO2 and H2SO4/ZrO2 were all found to have
a pH of about 1-2. Therefore, it can be concluded that in the
presence of liquid water leaching of H2SO4 from these catalysts
occurs. If H2SO4 is either present on the catalyst or is generated
via hydrolysis, it will thus dissolve into the reaction mixture
provided sulfuric acid is soluble in the liquid and act as a
homogeneous acid catalyst.
Catalytic Activity of H2SO4/ZrO2 and SO3/ZrO2/SiO2 in
Liquid-Phase Reactions. To assess the catalytic performance
of H2SO4/ZrO2 and SO3/ZrO2/SiO2 in the liquid-phase, the
liquid-phase hydro-acyloxy-addition reaction of acetic acid (4)
to camphene (5) giving the industrial important pine-fragrance
Figure 6. Formation of isobornyl acetate (6) by the hydro-acyloxy-
addition of 0.70 mol glacial acetic acid (4) to 0.70 mol camphene (5)
at 338 K (stirred tank-reactor, N atmosphere); (a) 0 2.5 g of H SO /
2,27,31-34
isobornyl acetate (6, Chart 2; reaction 5)
was chosen,
2
2
4
in which water is neither formed nor present. The presence of
a strong acid as catalyst, such as BF3/HOAc or H2SO4, is
required to establish the equilibrium corresponding to the
reaction. Under the employed conditions, 70% of 6 can be
formed (Figure 5, see the Experimental Section).
ZrO
b) ] 2.5 g of SO
and 320 µL (17.78 mmol) H
2
(Gimex) without H
2
O, ] 2.5 g of H
2
SO
4
/ZrO
2
(prec.) without H
2
O;
(
3
/ZrO
2
/SiO
2
without H
O.
2 3 2 2
O, 0 2.5 g of SO /ZrO /SiO
2
6
b). The presence of water was also necessary to obtain
6
isobornyl acetate (6) with anhydrous Zr(SO4)2. Hence, no Lewis
acid type catalytic activity is observed with the SO3/ZrO2/SiO2
catalysts. It is important to note that in the gas-phase reaction
a sub-stoichiometric quantity of water was introduced, whereas
a much larger amount of water was added in the liquid-phase
reaction. After removal of the SO3/ZrO2/SiO2 particles from the
reaction mixture (with extra water present) by filtration, the
formation of 6 did not stop. Nonetheless, the reaction rate
decreased (Figure 6b). In analogy with the sulfated zirconia
Anhydrous Zr(SO4)2 was found to be catalytically inactive
in the hydro-acyloxy-addition reaction. In analogy with the gas-
phase trans-alkylation reaction of benzene (1) and diethylben-
zene (2, Figure 4e), this shows that anhydrous Zr(SO4)2 is not
catalytically active as a Lewis acid. With the sulfated zirconia
6
(
H2SO4/ZrO2) catalysts, however, isobornyl acetate (6) was
immediately formed (Figure 6a). Removal of the solid catalyst
particles from the reaction mixture by filtration (see the
Experimental Section) did not stop the formation of 6, albeit a
decrease in reaction rate was observed. Hence, it appears that
H2SO4 slowly dissolves from the zirconia into the reaction
mixture and acts as a homogeneous catalyst. Although the sulfur
content of the H2SO4/ZrO2(prec.) catalyst is higher than that of
the H2SO4/ZrO2(Gimex) catalyst (Table 1), the rate of formation
of 6 is lower with the H2SO4/ZrO2(prec.) catalyst. The lower
activity of the H2SO4/ZrO2(prec.) catalyst indicates that leaching
of H2SO4 and/or the accessibility of the catalytically active
groups is less for the highly porous H2SO4/ZrO2(prec.) catalyst
than for the less porous H2SO4/ZrO2(Gimex) catalyst. The lower
porosity of the ZrO2(Gimex) brings about that calcination leads
to a loss of H2SO4, which is much more rapid than that of the
much more porous precipitated ZrO2(prec.). Within the liquid
phase, the transport through the highly porous H2SO4/ZrO2-
(H2SO4/ZrO2) catalysts, H2SO4 slowly dissolves from the surface
of the SO3/ZrO2/SiO2 catalysts into the reaction mixture and
subsequently homogeneously catalyzes the reaction. Because
the H2SO4/ZrO2 and SO3/ZrO2/SiO2 catalysts are susceptible
to hydrolysis and leaching of catalytically active, but corrosive
H2SO4, it can be questioned whether these catalysts are that
1
,3-5
benign as frequently proposed.
Conclusions
This research has substantiated that the activity of sulfated
zirconia is governed by the equilibria: Zr(SO ) + 4H O h
4
2
2
Zr(SO ) ‚4H O + nH O h ZrO + 2H SO ‚aq.
4
2
2
2
2
2
4
For the thermal pretreatment of sulfated zirconia, it is
important that the constituents of sulfuric acid exhibit a
considerable vapor pressure at significantly lower temperatures
than the temperatures at which zirconium sulfate decomposes.
The rate of transport of water out of a catalyst batch being
thermally pretreated is therefore decisive for the final activity
of the catalyst. Rapid release of water leads to water-free
(
prec.) catalyst is, however, much slower. Consequently, the
more accessible H2SO4 within the ZrO2(Gimex) is more active
and dissolves faster.
With the SO3/ZrO2/SiO2 catalysts, addition of water to the
reaction mixture was a prerequisite to observe activity (Figure

13412 J. Phys. Chem. B, Vol. 107, No. 48, 2003
Dijs et al.
zirconium sulfate, which is only active as a solid acid catalyst
after exposure to an amount of water vapor sufficient to exceed
the amount required for the reaction to the tetrahydrate. Because
water vapor strongly adsorbs on the surface of zirconia and
sulfated zirconia, the transport of water out of porous zirconia’s
and out of large catalyst batches will proceed slowly. Conse-
quently, the final catalyst will have retained a sufficiently large
fraction of water to provide a highly active catalyst. However,
keeping a catalyst for a long period of time in a gas flow
containing water vapor at a sufficiently elevated temperature
can lead to a substantial loss of sulfate. Consequently, the
material situated at the external edge of a larger catalyst batch
can appreciably loose sulfate.
Supporting Information Available: Schematic representa-
tion of the reactor used for the gas-phase treatments of the silica-
supported metal oxides and additional figures, tables, and
equations (8 pages). This material is available free of charge
via the Internet at http://pubs.acs.org.
References and Notes
(1) Tanabe, K.; H o¨ lderich, W. F. Appl. Catal. A 1999, 181, 399-434.
(2) Bledsoe, J. O. Terpenoids. In Kirk-Othmer Encyclopedia of
Chemical Technology, 4th ed.; Kroschwitz, J. I., Howe-Grant, M., Eds.;
Wiley: New York, 1997; Vol. 23, pp 833-882.
(3) Tanabe, K.; Misono, M.; Ono, Y.; Hattori, H. New Solid Acids
and Bases - Their Catalytic Properties; Elsevier: Amsterdam, 1989.
(4) Yadav, G. D.; Nair, J. J. Microporous Mesoporous Mater. 1999,
3
3, 1-48.
The volatility of the components of the active species also
affects the activity in gas-phase reactions. At temperatures above
about 473 K, both water and sulfuric acid can volatilize. When
water is added to the reactants, the hydrolysis of zirconium
sulfate can continue and the catalyst can slowly deactivate
irreversibly. Utilization of sulfated zirconia catalysts in gas-
phase reactions at temperatures significantly above about 473
K will therefore presumably not lead to a catalyst displaying a
stable performance during low periods of time. At lower
temperatures only, the release of water has to be envisioned,
which also leads to an inactive catalyst. Addition of some steam
to the reactants therefore may be required to maintain the activity
at a constant, high level at temperatures below about 473 K.
(
5) Song, X.; Sayari, A. Catal. ReV.-Sci. Eng. 1996, 38, 329-412.
(6) Dijs, I. J.; de Koning, R.; Geus, J. W.; Jenneskens, L. W. Phys.
Chem. Chem. Phys. 2001, 3, 4423-4429 and references therein.
(
49.
7) Babou, F.; Coudurier, G.; Vedrine, J. J. Catal. 1995, 152, 341-
3
(
8) F aˇ rca s¸ iu, D.; Li, J. Q. Appl. Catal. A 1998, 175, 1-9 and references
therein.
(
9) Fraenkel, D. Ind. Eng. Chem. Res. 1997, 36, 52-59.
(
10) Nagase, Y.; Jin, T.; Hattori, H.; Yamaguchi, T.; Tanabe, K. Bull.
Chem. Soc. Jpn. 1985, 58, 916-918.
(11) Rack Sohn, J.; Won Kim, H. J. Mol. Catal. 1989, 52, 361-374.
(12) Morterra, C.; Cerrato, G.; Di Ciero, S.; Signoretto, M.; Pinna, F.;
Strukul, G. J. Catal. 1997, 165, 172-183.
(
13) Ng, F. T. T.; Horv a´ t, N. Appl. Catal. A 1995, 123, L197-L203.
(14) Ebitani, K.; Konno, H.; Tanaka, T.; Hattori, H. J. Catal. 1992, 135,
60-67.
(
15) Brenner, A. M.; Schrodt, J. T.; Shi, B.; Davis, B. H. Catal. Today
Sulfated zirconia is a particularly interesting catalyst in liquid-
phase reactions. Activity calls for the above equilibria being
situated at the right-hand side. Sufficiently hydrated sulfated
zirconia’s can exhibit excellent catalytic properties in liquid-
phase reactions, provided the solubility of sulfuric acid in the
reaction medium is negligible.
1
998, 44, 235-244.
(16) Jin, T.; Machida, M.; Yamaguchi, T.; Tanabe, K. Inorg. Chem.
1984, 23 4396-4398.
(
17) F aˇ rca s¸ iu, D.; Li, J. Q.; Kogelbauer, A. J. Mol. Catal. A: Chem.
1
997, 124, 67-78 and references therein.
(
18) Arata, K.; Hino, M. Mater. Chem. Phys. 1990, 26, 213-237.
(19) Corma, A.; Mart ´ı nez, A.; Mart ´ı nez, C. Appl. Catal. A. 1996, 144,
49-268.
2
In the gas-phase trans-alkylation reaction, the sulfated metal
oxide catalysts showed a lower conversion of diethylbenzene
(20) Sarzanini, C.; Sacchero, G.; Pinna, F.; Signoretto, M.; Cerrato, G.;
Morterra, C. J. Mater. Chem. 1995, 5, 353-360.
(2) at 673 K than at 473 K, which is in contrast to the catalytic
(21) Bensitel, M.; Saur, O.; Lavalley, J.-C. Mater. Chem. Phys. 1988,
1
9, 147-156.
performance of a conventional acid amorphous aluminum
silicate catalyst. According to elemental analyses and in line
with the temperature-composition diagram of the system H2O-
SO3 (H2SO4), the inverse temperature dependence is a conse-
quence of the loss of catalytically active H2SO4 above 473 K.
Hence, in gas-phase reactions, these catalysts can only be used
below 473 K. In the liquid-phase hydro-acyloxy-addition
reaction, supported H2SO4 slowly leaches in the reaction mixture
and thus provides a homogeneous catalyst. Removal of the solid
catalyst particles from the reaction mixture by filtration
consequently did not stop the production of isobornyl acetate
(
22) Escalona Platero, E.; Pe n˜ arroya Mentruit, M.; Otero Are a´ n, C.;
Zecchina, A. J. Catal. 1996, 162, 268-276.
(23) Morterra, C.; Cerrato, G.; Bolis, V. Catal. Today 1993, 17, 505-
15.
5
(
24) Ardizzone, S.; Bianchi, C. L.; Grassi, E. Colloids. Surf. A 1998,
1
35, 41-51.
(25) The use of an acid-inert and structurally defined support-material
(e.g., Aerosil SiO2) is attractive for the preparation of heterogeneous catalysts
with a high specific surface area. This facilitates their integration in reactor
devices, e.g., monoliths and stirrer blades.
10,11
(26) Some of these reactions have been mentioned in the literature.
See also: Anderson, B. G.; Dang, Z.; Morrow, B. A. J. Phys. Chem. 1995,
9
9, 14444-14449. Bensitel, M.; Saur, O.; Lavalley, J. C. Mater. Chem.
Phys. 1987, 17, 249-258. Jin, T.; Machida, M.; Yamaguchi, T.; Tanabe,
K. Inorg. Chem. 1984, 23, 4396-4398. Yamaguchi, T.; Jin, T.; Tanabe,
K. J. Phys. Chem. 1986, 90, 3148-3152.
(6).
The main conclusion of this research is that the catalytic
(27) March, J. AdVanced Organic Chemistry, Reactions, Mechanisms,
activity of sulfated zirconia catalysts originates from supported
sulfuric acid containing water. The deficiencies of water-diluted
sulfuric acid, volatility and solubility in liquid, have to be
considered in the technical utilization of sulfated zirconia
catalysts.
and Structure, 4th ed.; Wiley: New York, 1992; pp 765, 562 and references
therein.
(28) Tsai, T.-C.; Liu, S.-B.; Wang, I. Appl. Catal. A 1999, 181, 355-
3
1
4
98.
(29) Bandyopadhyay, R.; Sugi, Y.; Kubota, Y.; Rao, B. S. Catal. Today
998, 44, 245-252.
(
30) Olah, G. A.; Kaspi, J.; Bukala, J. J. Org. Chem. 1977, 42, 4187-
191.
31) Dijs, I. J.; van Ochten, H. L. F.; van der Heijden, A. J. M.; Geus,
Acknowledgment. This work was carried out as part of the
Innovation Oriented Research Program on Catalysis (IOP
Katalyse, no. IKA96023b) sponsored by The Netherlands
Ministry of Economic Affairs (financial support I.J.D.). Dr. O.
L. J. Gijzeman and A.J.M. Mens (Utrecht University, Depart-
ment of Inorganic Chemistry and Catalysis) are gratefully
acknowledged for performing the XPS measurements. B.H.
Reesink and R. Ansems (Engelhard De Meern B.V., The
Netherlands) kindly performed the ICP-AES and Flash-combus-
tion GC analyses.
(
J. W.; Jenneskens, L. W. Appl. Catal A: General 2003, 241, 185-203.
(32) Dijs, I. J.; van Ochten, H. L. F.; van Walree, C. A.; Geus, J. W.;
Jenneskens, L. W. J. Mol. Catal. A: Chem. 2002, 188, 209-224.
(33) Ikan, R. Natural Products, A Laboratory Guide; Israel University
Press: Jerusalem, 1969; pp 137-177.
(34) Roberts, R. M. G. J. Chem. Soc., Perkin Trans. 2. 1976, 10, 1183-
1190 and references therein.
(
35) Klug, H. P.; Alexander, L. E. X-ray Diffraction Procedures For
Polycrystalline and Amorphous Materials, 2nd ed.; Wiley: New York, 1974.
36) Lercher, J. A. Adsorption Methods for the Assessment of the
Specific Surface Area, the Pore Size Distribution and the Active Sites of
(

Bulk and Silica-Supported ZrO2
J. Phys. Chem. B, Vol. 107, No. 48, 2003 13413
Heterogeneous Catalysts. In Catalysis: An Integrated Approach, 2nd ed.;
Van Santen, R. A., Van Leeuwen, P. W. N. M., Moulijn, J. A., Averill, B.
A., Eds; Elsevier: Amsterdam, 1999; Chapter 13.
(49) Fairbrother, D. H.; Johnston, H.; Samorjai, G. J. Phys. Chem. 1996,
100, 13696-13700.
(50) Guldan, E. D.; Schindler, L. R.; Roberts, J. T. J. Phys. Chem. 1995,
99, 16059-16066.
(
37) Notice that with nonmonochromatic XPS the Zr(3d5/2) and Zr(3d3/2)
peaks will not be resolved.
38) Christian, G. D. Analytical Chemistry, 4th ed.; Wiley: New York,
986; Chapter 6.
39) Skoog, D. A.; West, D. M.; Holler, F. J. Fundamentals of Analytical
Chemistry, 6th ed.; Saunders College: Philadelphia, PA, 1992; Chapter 4.
40) Sch a¨ fer, H. Chemical Transport Reactions; Academic Press: New
York, 1964; Translated by Frankfort, H.
41) If the hydroxyl ions are introduced heterogeneously, nucleation and
growth within the bulk of the solution can occur due to local super-
(51) Ardizzone, S.; Bianchi, C. L.; Signoretto, M. Appl. Surf. Sci. 1998,
136, 213-220.
(
1
(52) During calcination, sulfates can be eliminated. This will cause an
increase of the specific surface area of the H2SO4/ZrO2. At too high
temperatures, however, the structure of the remaining ZrO2 structure also
collapses, which will lead to a decrease of specific area.
(53) Gmelins Handbuch der Anorganischen Chemie, 8. Auflage; Verlag-
Chemie: Weinheim, Germany, 1960; Sulfur (SN9), Part B2, pp 668, 677.
(54) Miles, F. D.; Niblock, H.; Wilson, G. L. Trans. Faraday. Soc. 1940,
36, 345-356.
(
(
(
3
8,39
saturation.
dissolution of the silica support.
42) Pourbaix, M. J. N. Atlas d’equilibres e´ lectrochimiques; Gauthier-
Villars: Paris, 1963; Zr.
43) de Boer, M.; Leliveld, R. G.; Bruil, H. G.; van Dillen, A. J.; Geus,
J. W. Appl. Catal. A 1993, 102, 35-51.
44) Alexander, L. E. X-ray Diffraction Methods in Polymer Science;
Wiley: New York, 1969; Chapters 1 and 3.
45) Practical Surface Analysis, Auger and X-ray Photoelectron Spec-
troscopy, 2nd ed.; Briggs, D., Seah, M. P., Eds.; Wiley: Chichester, U.K.,
900; Vol. 1.
46) Handbook of X-ray Photoelectron Spectroscopy; Muilenberg, G.
The pH of the suspensions must be kept below 7 to prevent
(55) Kneitsch, R. ReV. G e´ n e´ r. Chim. Pur. Appl. 1902, 5, 48-83.
(56) Nilges, J.; Schrage, J. Fluid Phase Equilib. 1991, 68, 247-261.
(57) The application of the Gibbs phase rule and phase diagrams is only
valid under equilibrium conditions. Hence, Figure A4 (see the Supporting
Information) only provides a qualitative guideline for nonequilibrium (flow-)
conditions. Because the vapor pressure of H2O is much larger than that of
H2SO4, in practice, there will be a negligible amount of H2SO4 in the vapor
phase below 473 K.
(
(
(
(
(58) Bear, I. J. Aust. J. Chem. 1967, 20, 415-428.
(59) Bear, I. J. Aust. J. Chem. 1969, 22, 875-889.
1
(
(60) Gmelins Handbuch der Anorganischen Chemie, 8. Auflage; Springer-
Verlag: Berlin, 1958; Zirconium (SN 42), pp 326-353.
(61) NouVeau Trait e´ de Chimie Min e´ rale; Pascal, P., Ed.; Masson: Paris,
1963; Part 9 (Zr), pp 576-621.
E., Wagner, C. D., Riggs, W. M., Davis, L. E., Moulder, J. F., Eds.;
Perkin-Elmer Corporation, Physical Electronics Division: Minnesota,
1
979.
47) Price, L. S.; Parkin, I. P.; Hardy, A. M. E.; Clark, R. J. H.; Hibbert,
T. G.; Molloy, K. C. Chem. Mater. 1999, 11, 1792-1799.
48) Dartigeas, K.; Benoist, L.; Gonbeau, D.; Pfister-Guillouzo, G.;
Ouvrard, G.; Levasseur, A. J. Electron Spectrosc. Relat. Phenom. 1997,
3, 45-58.
(
(62) In general, both the sulfur content and the specific area of the
catalyst affect catalytic activity.4
,9,19
See also: Katada, N.; Endo, J.; Notsu,
(
K.; Yasunobu, N.; Naito, N.; Niwa, M. J. Phys. Chem. B 2000, 104, 10321-
10328. Corma, A.; Forn e´ s, V.; Juan-Rajadell, M. I.; L o´ pez Nieto, J. M.
Appl. Catal. A 1994, 116, 151-163.
8