Structure-reactivity correlations in sulphated-zirconia catalysts for the isomerisation of α-pinene

Article abstract of DOI:10.1016/S0021-9517(02)00150-1

A range of mesoporous sulphated zirconias with tuneable structural and catalytic properties have been prepared by direct impregnation. The surface sulphate coverage can be readily varied, achieving a maximum value of ～0.2 monolayers. High-temperature calcination induces the crystallisation of tetragonal zirconia while suppressing the monoclinic phase and enhances surface acidity. Superacid sites only appear above a critical threshold SO₄ coverage of 0.08 mL (corresponding to 0.44 wt% total S). Sulphated zirconias show good activity towards α-pinene isomerisation of under mild conditions. Conversion correlates with the number Bronsted acid sites, while the selectivity towards mono- versus polycyclic products depends on the corresponding acid site strength; superacidity promotes limonene formation over camphene.

Full text of DOI:10.1016/S0021-9517(02)00150-1

Journal of Catalysis 215 (2003) 57–65
www.elsevier.com/locate/jcat
Structure–reactivity correlations in sulphated-zirconia catalysts
for the isomerisation of α-pinene
a
b
a,∗
Muriel A. Ecormier, Karen Wilson, and Adam F. Lee
a
Department of Chemistry, University of Hull, Hull HU6 7RX, UK
b
Department of Chemistry, University of York, Heslington, York YO10 5DD, UK
Received 5 August 2002; revised 7 October 2002; accepted 4 November 2002
Abstract
A range of mesoporous sulphated zirconias with tuneable structural and catalytic properties have been prepared by direct impregnation.
The surface sulphate coverage can be readily varied, achieving a maximum value of ∼ 0.2 monolayers. High-temperature calcination induces
the crystallisation of tetragonal zirconia while suppressing the monoclinic phase and enhances surface acidity. Superacid sites only appear
above a critical threshold SO coverage of 0.08 mL (corresponding to 0.44 wt% total S). Sulphated zirconias show good activity towards
4
α-pinene isomerisation of under mild conditions. Conversion correlates with the number Brønsted acid sites, while the selectivity towards
mono- versus polycyclic products depends on the corresponding acid site strength; superacidity promotes limonene formation over camphene.

2003 Elsevier Science (USA). All rights reserved.
Keywords: Solid acid catalysts; Clean technology; Green chemistry; Zirconia; α-pinene
1
. Introduction
20–100 Å and thus new avenues for liquid-phase solid acid
catalysis.
In contrast to liquid acids, which possess well-defined
acid properties, solid acids may contain a variety of acid
sites. Generally they are categorised by their Brønsted and/or
Lewis acidity, the strength and number of these sites, and
the textural properties of the support (surface area and
porosity) [3]. The synthesis of pure Brønsted and pure Lewis
acid catalysts has attracted great academic interest, although
the latter has proven more difficult because Brønsted acidity
often arises from Lewis acid–base complexation [4]. In order
to achieve high selectivity towards the desired products
in a synthesis all these properties must be considered
and if possible controlled. For example, acetal formation
and hydrolysis reactions generally require medium-acid-
strength sites, while electrophilic additions of alcohols or
water to olefins, skeletal rearrangements, and esterification
and alkylation reactions require strong acid sites. The
importance of the type of acid site has likewise been shown
for Friedel Crafts alkylation reactions, where Lewis acid
sites are required for toluene alkylation by benzyl chloride,
while Brønsted sites are preferred for reactions using benzyl
alcohol [5].
The syntheses of many fine and speciality chemicals
often rely on homogeneous mineral acids, bases, or metal
salts, which are frequently used in stoichiometric amounts.
Tightening legislation on the treatment and disposal of
excessive toxic waste, produced during the separation and
neutralisation of products from these reaction media, is
driving industry to consider cleaner technologies, including
the use of heterogeneous catalysis. Of particular concern are
the wide range of liquid-phase industrial reactions which
rely on the use of inorganic or mineral acids. While many
of these processes are catalytic, some require stoichiometric
amounts of acid (e.g., acylation using AlCl3). While zeolites
are widely employed as solid acid catalysts in gas-phase
chemistry, their application in liquid-phase organic synthesis
is limited by their small pore sizes (< 8 Å), which make
them unsuitable for reactions involving bulky substrates.
However, recent developments in materials chemistry have
led to the discovery of the M41S family of mesoporous
molecular sieves [1,2] offering pore sizes in the range
*
There has been much interest in the use of sulphated
metal oxides as strong solid acids, in particular sulphated
Corresponding author.
E-mail address: a.f.lee@chem.hull.ac.uk (A.F. Lee).
0021-9517/03/$ – see front matter  2003 Elsevier Science (USA). All rights reserved.
doi:10.1016/S0021-9517(02)00150-1

58
M.A. Ecormier et al. / Journal of Catalysis 215 (2003) 57–65
Scheme 1.
zirconias, which have been suggested to possess superacidic
properties [6]. While most research has focused on gas-
phase transformations [7,8] for applications in cracking
and isomerisation reactions, their use has been recently
extended to liquid phase processes [9,10]. However, the
nature of the active site(s) remains contentious, with both
sol–gel and impregnation methods reported as effective
preparation routes [11,12]. Sulphated zirconias’ superacid
nature has also been questioned; Hino and Arata first
reported their low-temperature activity towards n-butane
isomerisation [13] and evaluated their acid strength using
Hammett indicators [14], the application of which to solids
has been latterly criticised. A wide range of alternative
techniques including NMR [15], NH3 microcalorimetry
Here we find that direct wet impregnation permits the
preparation of mesoporous sulphated zirconias exhibiting
tuneable surface acidity and concomitant selectivity control
in the isomerisation of α-pinene under mild conditions.
2. Experimental
A series of sulphated zirconias were prepared by wet
impregnation of 5 g zirconia (99%+, Engelhard) with
3
5
0 cm of sulphuric acid (0.01–2.5 M). The resulting slurry
◦
was filtered and dried at 80 C overnight before calcination
◦
at 550 C for 3 h. The calcination temperature was selected
to maximise the degree of surface ordering while minimising
sulphur loss, based on the thermal analysis results presented
in this paper. Samples were subsequently stored in air prior
to characterisation and reactor testing.
[
16,17], and ESR analysis [18] have been employed to
address this issue; however, there is still no consensus.
The super/strong acid sites, which are the active sites
of the catalyst, have also been attributed to either Lewis
Porosity and surface area measurements were performed
[
[
19,20], Brønsted [21,22], or a combination of both acid sites
23,24]. Alternative local zirconia geometries have also been
following the N adsorption on a Micromeritics ASAP 2010
instrument. Surface areas were calculated using the BET
2
postulated [7]. To our knowledge no systematic study of
structure–reactivity relationships within these materials has
been performed wherein acid strength has been correlated
with SO4 content and concomitant selectivity in liquid-phase
chemistry.
To address this issue, we report the effect of controlled
sulphation of a mesoporous ZrO2 support on the structural,
acidic, and catalytic properties of the resulting material. The
rearrangement of α-pinene was selected to probe the effect
of acid strength on catalytic performance [25]. This is a
particularly interesting reaction as the product selectivity is
known to vary with catalyst acid strength, with weak acids
favouring camphene formation, while stronger acids result
in monocyclic products [26,27], as shown in Scheme 1.
The only previous study of this system, by Ponzi et al.
equation over the pressure range P/P = 0.02–0.2, where
0
a linear relationship was maintained, while pore size distrib-
utions were calculated using the BJH model. Sulphur con-
tent was determined by elemental analysis using a Carlo
Erba 1108 CHN/S instrument; quoted percentages refer to
total S levels. XPS measurements were performed using a
Kratos AXIS HSi instrument equipped with a charge neu-
traliser and an Mg-K X-ray source. Spectra were recorded
α
at normal emission using an analyser pass energy of 20 eV
and X-ray power of 225 W. Two-point energy referenc-
ing was employed using adventitous carbon at 285 eV and
the valence band. Ex situ XRD spectra were acquired by
the use of a Siemens D5000 diffractometer and Cu-K
α
◦
◦
radiation. The scan range was from 2θ = 10 to 100 in
◦
increments of 0.02 . Peak-fitting was performed using a
[
28,29], demonstrated that sulphated zirconias were active
pseudo-Voigt function to determine crystallite sizes with
the error estimated as ± 10% for spherical/octahedral parti-
cles. Temperature-programmed reduction was carried out in
a quartz tube using 50 mg of sample under a 5% H2/N2 flow
◦
above 100 C, producing mainly camphene. However, their
catalysts were ill-defined and neither textural properties,
surface sulphate loading, or acid strength were reported.

M.A. Ecormier et al. / Journal of Catalysis 215 (2003) 57–65
59
◦
–1
(
E&W 99%) with a linear ramp rate of 11 C min . DRIFTS
spectra were obtained using a Bruker Equinox FTIR spec-
trometer. Samples were diluted with KBr powder (10 wt% in
KBr), then loaded into an environmental cell and subjected
◦
to additional drying under vacuum at 110 C for 10 min prior
to measurements to remove moisture physisorbed during air
exposure. Ex situ pyridine adsorption was performed by in-
cipient wetness impregnation, with excess pyridine evapo-
rated prior to sample loading and evacuation as described
above. Acidity was determined by Hammett indicators;
3
∼
25 mg of sample was shaken with 1 cm of a solution of
3
Hammett indicator diluted in 10 cm cyclohexane and left to
equilibrate for 2 h. The colour on the catalyst was then noted.
The following Hammett indicators were used: dicinna-
malacetone (pKa = −3.0), benzylacetophenone on chalcone
Fig. 1. Total and surface sulphur loading as a function of impregnating
H SO and molarity for calcined sulphated zirconias.
2
4
(
(
pKa = −5.6), anthraquinone (pKa = −8.3), 4-nitrotoluene
pKa = −11.35), 3-nitrotoluene (pKa = −11.99) and 2,4-
The surface loading mirrors the total sulphur content and ex-
ceeds it in all cases, as expected for an impregnation ap-
proach wherein only a small proportion should be incor-
porated within the bulk zirconia framework. Above > 1 M
H2SO4 the surface loading begins to level off; however, the
total loading continues to rise, approaching the final surface
value of 5 wt% S for the 2.5 M treatment. Complete surface
sulphation thus requires bulk sulphur loadings in excess of
dinitrotoluene (pKa = −13.75). The acid strength is quoted
as being stronger than the weakest indicator which exhibits
a colour change, but weaker than the strongest base that pro-
duces no change.
Reactions were performed in a stirred batch reactor
under nitrogen with samples being taken periodically for
analysis using a Varian CP-3800 gas chromatograph fitted
with a DB1 capillary column (film thickness 0.25 µm,
1
.7 wt%. These values represent the first quantitative deter-
id 0.32 mm, length 30 m) and a CP-8400 autosampler.
◦
mination of surface sulphur within sulphated zirconias.
Prior to calcination the impregnated samples possessed
The isomerisation of α-pinene was conducted at 60
C
using 63 mmol of α-pinene (98%, Aldrich), 100 mg of
2
–1
3
high surface areas, in excess of 250 m g , and showed a
air-exposed catalyst, and 0.2 cm of tetradecane (99%,
◦
slight decrease with sulphur loading. Calcining at 550 C
Lancaster) as an internal standard. Reactions were run for
reduced the surface areas of all samples (although they re-
9
h with initial rates determined from the linear portion
2
–1
main > 50 m g in all cases (Table 1). Pore size distribu-
tion measurements revealed that the resulting calcined mate-
rials were mesoporous (Fig. 2a), with well-defined pore sizes
ranging between 33 and 46 Å reflecting the textural proper-
ties of the calcined ZrO2 precursor (Fig. 2b). There was no
systematic variation in pore diameter with sulphur content.
Fig. 3 shows the effect of calcination and sulphur load-
ing on the thermal stability of impregnated zirconias as stud-
ied via temperature-programmed reduction (TPR). The zir-
of the conversion profile. Catalyst selectivity and overall
mass balances (closure was > 98%) were determined using
appropriate reactant and product response factors derived
from multipoint calibration curves. The systematic errors
in the conversions and selectivities are ± 2% and ± 3%,
respectively.
3
. Results and discussion
conia precursor showed no reduction at temperatures below
◦
7
00 C, in accord with the literature [30,31]. Morterra et
The catalytic performance of sulphated zirconia hinges
al. [32] reported that SO2 is predominantly evolved during
thermal reduction of sulphated zirconias; hence the peaks
upon the successful incorporation of sulphoxy functionali-
ties into the zirconia framework. Fig. 1 shows the total and
surface sulphur content as a function of molarity of impreg-
nating solution. It is evident that the total sulphur loading
shows a rapid increase for low sulphuric acid concentrations
Table 1
Influence of sulphur content and calcination on the surface area of sulphated
zirconias
(
in line with the theoretical loading), demonstrating efficient
◦
Total S content
Uncalcined
550 C Calcined
S inclusion during the impregnation process. The surface
sulphur levels were determined from the S2p XP intensity
and overall surface composition, incorporating appropriate
elemental response factors (RF) according to the equation
2
–1
2 –1
(wt%)
(m g
)
(m g )
0
453
435
376
393
349
246
56
97
141
147
128
125
0
1
.44
.71
Sꢀꢁꢁ
Sꢂ
ꢁ
Oꢂ
1.74
S
S
O
S (wt%) = I × RF
I × RF + I × RF
2
4
.49
.03
ꢁ
ꢂ
ꢁ
Cꢂꢂ.
Zr
Zr
C
×
I
× RF + I × RF

60
M.A. Ecormier et al. / Journal of Catalysis 215 (2003) 57–65
(
a)
(
b)
Fig. 2. (a) N adsorption isotherms and (b) BJH pore-size distributions for calcined sulphated zirconias as a function of sulphur loading.
2

M.A. Ecormier et al. / Journal of Catalysis 215 (2003) 57–65
61
Fig. 3. Temperature-programmed reduction profiles for 1.74 and 4.03 wt%
◦
sulphated zirconias before and after 550 C calcination. Corresponding
differential thermogravimetric loss profile for the calcined 4.03 wt% sample
is also shown.
◦
observed between 400 and 700 C for treated samples can
be associated with the reduction of sulphoxy species. The
calcination step increased the thermal stability of both low-
and high-sulphur-loading materials. The TPR profile for
the 1.74 wt% sample exhibits a single peak-indicating a
relatively homogeneous sulphoxy environment. Although
higher sulphur loadings led to increased thermal stabil-
ity, they also reduced sample homogeneity, resulting in
Fig. 4. X-ray diffractograms of calcined sulphated zirconias as a function of
sulphur loading. The uncalcined zirconia support is shown for comparison.
◦
the emergence of a second, weaker reduction state ∼ 50 C
sis showed that the tetragonal phase comprised small crys-
the main peak. These observations suggest that a single SO4
species predominates up to the saturation coverage deter-
mined by XPS, beyond which sulphur is incorporated within
various subsurface environments. Corresponding thermo-
gravimetric analysis confirmed the principal high tempera-
ture SOx decomposition process and also identified a small
tallites of ∼ 145 Å independent of sulphur loading; high-
◦
temperature calcination (> 650 C) has been shown to elim-
inate differences in sulphated-zirconia crystallite sizes [36].
The monoclinic phase present in the calcined and 0.44 wt%
zirconia samples possessed a similar average crystallite size.
It is interesting to note that sulphated zirconias prepared by
sol–gel routes (with total loadings of 1.69 and 2.73 wt% S,
respectively) gave diffractograms almost identical to those
of the higher loading samples produced by direct impreg-
nation and are similar to those previously observed [37,38].
Neither preparative route gave rise to crystalline, stoichio-
metric zirconium sulphate.
◦
mass loss at 100 C associated with adsorbed water desorp-
tion.
The formation of crystalline sulphated zirconia via the
impregnation process was explored using powder X-ray dif-
fraction, Fig. 4. The uncalcined zirconia precursor con-
tained only a single weak, broad reflection centred around
◦
2
θ = 30 , indicating essentially amorphous character and
The nature of the surface environment within both im-
pregnated and sol–gel materials was explored by XPS.
Zirconium 3d spectra (Fig. 5) of the calcined, unsulphated
zirconia support showed a single poorly resolved doublet
centred around 182.8 eV (3d5/2) with a spin–orbit split-
thus consistent with the porosimetry data. In contrast, the
calcined zirconia support exhibited reflections arising from
both monoclinic (2θ = 24.7 , 28.4 , 31.6 [33]) and tetrag-
◦
◦
◦
◦
◦
◦
◦
◦
◦
onal (2θ = 30.3 , 35.3 , 50.7 , 59.9 , 60.6 and 63.5 [34])
zirconia phases. All the sulphated materials were likewise
highly ordered; however, increasing the sulphur loading
above 0.44 wt% caused a sharp increase in the intensity
of the tetragonal phase and concomitant loss of the mono-
clinic phase. Although sulphation was previously reported
to induce a similar structural transformation in the support
from amorphous to tetragonal zirconia [35], our results show
that there is a threshold sulphur coverage for this evolu-
tion; loadings > 0.44 wt% are necessary to effect a com-
plete monoclinic → tetragonal transition. Peak shape analy-
IV
ting of 2.4 eV. This binding energy is consistent with Zr
sites in ZrO2 [39]. Sulphation resulted in a linear decrease
in signal intensity (Fig. 5, inset) with the highest loading
giving ∼ 30% Zr attenuation, consistent with formation of
a chemisorbed sulphoxy overlayer. This was accompanied
by a slight Zr peak shift around 0.15 eV to higher bind-
ing energy, suggesting the direct coordination of zirconium
atoms to strongly electron-withdrawing SOx centres. Fig. 6
shows that the oxygen 1s spectrum for calcined, unsulphated
zirconia comprises a single dominant peak at ∼ 530.4 eV,

62
M.A. Ecormier et al. / Journal of Catalysis 215 (2003) 57–65
Fig. 5. Zr3d XP spectra of calcined sulphated zirconias as a function
of sulphur loading. Inset shows normalised integrated Zr intensity versus
loading.
Fig. 7. S2p XP spectra of calcined sulphated zirconias as a function of
sulphur loading.
species for all materials. These observations are supported
by the sulphur 2p spectra (Fig. 7), which exhibit a single
broad state at ∼ 169.4 eV, consistent with a unique SO4 en-
vironment [41,42]. A full Zr(SO4)2 monolayer (ML) would
contain ∼ 22.6 wt% S, much greater than the limiting value
of 5 wt%, suggesting that a maximum of only one-fifth of
the zirconia surface (corresponding to ∼ 0.2 mL) undergoes
sulphation.
consistent with literature values [40]. Sulphation resulted
in the gradual attenuation of this principal zirconia state
and concomitant growth of a new high-binding-energy fea-
ture at ∼ 532.3 eV. The separation of these oxygen states
(
∼ 1.9 eV) was independent of sulphur loading, indicating
the formation of a common (higher sulphoxy [41]) surface
Surface functionalisation of the calcined sulphated zirco-
nias was further explored by DRIFTS. Sulphation resulted
–
1
in the appearance of new features at ∼ 1625 cm and in
the region 1400–900 cm–1 (Fig. 8). The former peak arises
from the δHOH bending mode of chemisorbed water coordi-
nated to Lewis acid sites or hydrogen-bonded to polar sul-
phate/hydroxyl groups [43,44], while the latter region cor-
responds to the sulphoxy vibrations [45]. Higher sulphur
loadings enhance the SOx vibrational intensities but degrade
their resolution, resulting in a broad spectral envelope for
the 4.03 wt% sample. The principal SOx components lie at
1
1
240–1265 (S=O stretch), 1134–1152 (νs O=S=O), and
045 cm–1 (S–O stretch). It is important to note the corre-
sponding asymmetric O=S=O stretch, expected at ∼ 1300–
–
1
1
400 cm , was of very low intensity, except for the lowest
loading 0.44 wt% sample (indicative of bidentate sulphate
46,47]). The weakness of this band may reflect a reduced
[
covalent character of the sulphoxy groups, as will be dis-
cussed later [46,48]. Pyridine adsorption was subsequently
used to titrate Brønsted and Lewis acid sites. Fig. 9 shows the
resulting DRIFTS spectra. As anticipated, the calcined zirco-
nia precursor showed no detectable surface acidity, whereas
all sulphated zirconias exhibited sharp peaks characteristic
Fig. 6. O1s XP spectra of calcined sulphated zirconias as a function of
sulphur loading.

M.A. Ecormier et al. / Journal of Catalysis 215 (2003) 57–65
63
Fig. 10. Hammett acidity of fresh and calcined sulphated zirconias as a
function of sulphur loading.
Fig. 8. DRIFTS spectra of calcined sulphated zirconias as a function of
sulphur loading. The principal SOx vibrations are indicated.
the suspended solid with n-butylamine, according to the
Hammett–Bertolacini method [51]. However, determination
of superacidity and acid site distributions of porous solid
acid catalysts by such measurements remains controversial
due to solvent–support interactions [52] and accessibility of
the probe molecule within micropores [53]. Despite this,
provided the polarity of the solvent used to suspend the
solid and indicator is similar to that of the reaction media,
semi-quantitative determination of the maximum solid acid
strength can be obtained. From Fig. 10 uncalcined sulphated
zirconias only displayed a marginal increase in the surface
acidity of all samples, which remained weak solid acids.
In contrast, subsequent calcination dramatically enhanced
the resultant surface acidity, with superacidic properties
emerging for bulk sulphur loadings > 1.71 wt%, which
exhibited a limiting H0 of between −12 and −13. It is
interesting to note that the trend in H0 exhibits a direct
correspondence with the surface SO4 coverage. The S
XP spectra show no evidence for new chemically distinct
SOx species at high loadings; hence superacidity does not
arise from a change in the S oxidation state or significant
restructuring of the sulphate local environment.
–
1
of predominantly Brønsted acid sites (1540 and 1633 cm ),
–
1
although a small contribution from Lewis sites (1480 cm )
cannot be discounted [12,44,49]. A high concentration of
Brønsted versus Lewis acid sites is to be expected as a result
of moisture adsorption; Zhang et al. [50] have shown that
∼
95 % of the total acid sites remain in Brønsted form after
◦
air-exposed sulphated zirconias are dried at 100 C. The in-
tensity of both acid features increases with sulphur loading,
but rapidly plateaus, mirroring the surface sulphur content.
Comparison with the uncalcined 4.03 wt% sample confirms
that thermal processing improves the homogeneity of sur-
face sulphate groups.
Fig. 10 presents the acid strength of the sulphated
zirconias as estimated using Hammett indicators. Acid
strength measurements using Hammett indicators rely on the
use of a series of neutral basic indicators which change to
the colour of their conjugate acid when interacting with acid
sites of a critical pKa. Quantitative acid site distributions
can be determined by UV spectroscopy by titration of
The corresponding catalytic properties of the sulphated
zirconias towards α-pinene isomerisation were subsequently
explored. Fig. 11 shows that the calcined zirconia support
possesses poor activity with a corresponding limiting con-
version of only 5% after 9 h reaction. Sulphation enhances
catalyst performance, with initial rates showing a direct cor-
relation with measured surface sulphate coverage (ranging
from ∼ 0.08 to 0.2 ML SO4) and thus acid site density. The
intrinsic (area-normalised) catalyst activities are almost in-
dependent of sulphate loading and thus acid site strength
–
1
–1 –2
at ∼ 0.35 ± 0.05 mmolh gcat m . Pinene conversion
mirrored the activity trends with a maximum of 66% ob-
tained for the 4.03 wt% catalyst. Sulphation also induces
a significant switchover in selectivity from essentially pure
polycyclic camphene over unsulphated zirconia towards the
monocyclic limonene product, in striking agreement with
the acid strength predictions from the Hammett indicators.
Fig. 9. DRIFTS spectra showing pyridine titration of calcined sulphated
zirconias as a function of sulphur loading.

64
M.A. Ecormier et al. / Journal of Catalysis 215 (2003) 57–65
As the H2SO4 concentration and thus initial bisulphate
loading rise, the degree of reactive surface dehydroxyla-
tion is likewise expected to increase, as indeed evidenced
by the appearance of the tetragonal phase of zirconia (usu-
ally formed via thermally dehydroxylation) for loadings
ꢀ
1.71 wt%. Above a critical bisulphate coverage there
may be insufficient hydroxyl groups for the condensation
reaction to proceed to completion during calcination. This
would result in the stabilisation of permanently bound bisul-
phate Brønsted acid sites as postulated by Clearfield et
al. [54] and Chen et al. [18] and the generation of new
strong acid sites with an H o value similar to that of sul-
phuric acid (Scheme 2b). Both our Hammett indicator mea-
surements and the selectivity switchover from polycylic to
monocylic products during α-pinene isomerisation support
this model. The weaker covalent character of bisulphate ver-
sus sulphate would also account for loss of the asymmetric
O=S=O stretch above 0.44 wt% S. Hence increasing the
sulphate loading, and thus the surface bisulphate:hydroxyl
ratio, seems to induce a transition from ZrO2-derived weak
Lewis/Brønsted acid sites to bisulphate-derived strong Brøn-
sted acid sites.
Fig. 11. Catalytic performance of calcined sulphated zirconias in α-pinene
isomerisation after 9 h reaction as a function of sulphur loading. Treax =
◦
6
0 C. Activities have been normalised to the BET surface areas.
Small amounts of α-terpinene, γ -terpinene, and p-cymene
were also observed as residual secondary by-products, most
notably for the superacidic catalysts.
The correlation between acidity and surface sulphate
loading can be understood by considering the Clearfield
model for acid site formation in sulphated zirconia. Expo-
sure of ZrO2 or Zr(OH)4 surfaces to H2SO4 solutions is
4
. Conclusions
−
proposed to result in surface bisulphate groups (HSO4 )
Direct wet impregnation routes permit the synthesis of
in a bidentate coordination. During subsequent calcination
bisulphate can undergo condensation with adjacent hydroxyl
groups to form a neutral bidentate sulphate species together
with a Lewis acid centre located on the zirconia support
mesoporous sulphated zirconias with tuneable structural and
catalytic properties. The molarity of H SO impregnating
solution controls the surface sulphate coverage, which at-
2
4
tains a maximum of ∼ 0.2 ML SO for acid concentrations
4
(
Scheme 2a). The latter can readily interconvert to yield
> 2.5 M. Sulphation induces a concomitant crystallisation of
weak Brønsted sites on coordination with water. Our struc-
tural measurements reveal that weak sulphuric acid solutions
result in a low surface coverage of bidentate SO4 groups
exhibiting strong covalent character. The corresponding sup-
port shows incomplete crystallisation (Fig. 4, 0.44 wt% sam-
ple) and therefore partial hydration, consistent with a sur-
face in which excess hydroxyl functionality was available
for complete neutralisation of bisulphate.
amorphous zirconia to the tetragonal phase, while suppress-
ing the monoclinic phase, and the formation of superacidic
sites. The emergence of these superacid sites above a criti-
cal threshold SO coverage of 0.08 mL (0.44 wt% total S)
4
appears associated with this bulk structural transformation
and a corresponding change in the surface bisulphate con-
centration. Sulphated zirconias show good activity in the iso-
merisation of α-pinene, which has been successfully used to
Scheme 2.

M.A. Ecormier et al. / Journal of Catalysis 215 (2003) 57–65
65
probe their surface acidity; low S loadings/weak acid sites
favour camphene, while high S loadings/strong acid sites
promote limonene production.
[23] N. Katada, J. Endo, K. Notsu, N. Yasunobu, N. Naito, M. Niwa,
J. Phys. Chem. B 104 (2000) 10321.
[24] B.H. David, R.A. Keogh, R. Srinivasan, Catal. Today 20 (1994) 219.
25] J.C. van der Waal, H. van Bekkum, J.M. Vital, J. Mol. Catal. 105
[
(1996) 185.
[26] A. Stanislaus, L.M. Yeddanapalli, Canad. J. Chem. 50 (1972) 61.
Acknowledgments
[27] A. Severino, A. Esculcas, J. Roca, J. Vital, L.S. Lobo, Appl. Cat. A
Gen. 142 (1996) 255.
[
[
[
28] L. Grzona, N. Comelli, O. Masini, E. Ponzi, M. Ponzi, React. Kinet.
Financial support by the UK Engineering and Physical
Sciences Research Council under Grant GR/M20877 and
by BP-Amoco Chemicals is gratefully acknowledged. MAE
thanks the University of Hull for financial assistance.
Catal. Lett. 69 (2000) 271.
29] N. Comelli, L. Grzona, O. Masini, E. Ponzi, M. Ponzi, React. Kinet.
Catal. Lett. 71 (2000) 27.
30] B.-Q. Xu, W.M.H. Sachtler, J. Catal. 167 (1997) 224.
[
[
31] G. Delahay, E. Enusque, B. Coq, F. Figueras, J. Catal. 175 (1998) 7.
32] C. Morterra, G. Cerrato, S. Di Ciero, M. Signoretto, F. Pinna, G. Stru-
kul, J. Catal. 165 (1997) 172.
33] E.A. Garcia, E.H. Rueda, A.J. Rouco, Appl. Catal. A Gen. 210 (2001)
References
[
3
63.
[
1] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D.
Schmitt, C.T.-W. Chu, D.H. Olsen, E.W. Sheppard, S.B. McCullen,
J.B. Higgins, J.L. Schlenker, J. Am. Chem. Soc. 114 (1992) 10834.
2] U. Ciesla, M. Froba, G. Stucky, F. Schuth, Chem. Mater. 11 (1999)
[
34] T. Yamamoto, T. Tanaka, S. Takenaka, S. Yoshida, T. Onari, Y. Taka-
hashi, T. Kosaka, S. Hasegawa, M. Kudo, J. Phys. Chem. B 103 (1999)
2
385.
35] C. Morterra, G. Cerrato, C. Emanuel, V. Bolis, J. Catal. 142 (1993)
49.
[
[
[
[
227.
3
3] K. Tanabe, M. Misono, Y. Ono, H. Hattori, New Solid Acids and
Bases, in: Studies in Surface Science and Catalysis, Vol. 51, Elsevier,
Amsterdam, 1998.
36] M. Benaissa, J.G. Santiesteban, G. Diaz, C.D. Chang, M. Jose-Yaca-
man, J. Catal. 161 (1996) 694.
37] A.F. Bedilo, K.J. Klabunde, J. Catal. 176 (1998) 448.
38] A. Minesso, F. Genna, T. Finooto, A. Baldan, A. Benedetti, J. Sol Gel
Sci. Technol. 24 (2002) 197.
[
[
[
[
[
[
[
[
4] K. Wilson, J.H. Clark, Pure Appl. Chem. 72 (2000) 1313.
5] A. Corma, H. Garcia, Catal. Today 38 (1997) 257.
6] K. Arata, Appl. Catal. A. 146 (1996) 3.
7] X. Song, A. Sayari, Catal. Rev. 38 (1996) 329.
8] Q.-H. Xia, H. Hidajat, S. Kawi, J. Catal. 205 (2002) 318.
9] G.D. Yadav, J.J. Nair, Micropor. Mesopor. Mater. 33 (1999) 1.
[
[
[
39] A.E. Hughes, B.A. Sexton, J. Electron. Spectrosc. Relat. Phenom. 50
(1990) c15.
40] J.L. Colon, D.S. Thakur, C.-Y. Yang, A. Clearfield, C.R. Martin, J. Ca-
tal. 124 (1990) 148.
41] C.D. Wagner, A.V. Naumkin, A. Kraut-Vass, J.W. Allison, C.J. Powell,
J.R. Rumble Jr., NIST Standard Reference Database 20, Version 3.2,
[
[
10] D.J. McIntosh, R.A. Kydd, Micropor. Mesopor. Mater. 37 (2000) 281.
11] H. Armendariz, B. Coq, D. Tichit, R. Dutartre, F. Figueras, J. Ca-
tal. 173 (1998) 345.
2002.
[
12] V. Parvulescu, S. Coman, P. Grange, V.I. Parvulescu, App. Catal. A
[
42] C.D. Wagner, in: D. Briggs, M.P. Seah (Eds.), Practical Surface
Analysis, Vol. 1, Wiley, Chichester, 1990.
43] B. Li, R.D. Gonzalez, Catal. Today 46 (1998) 55.
44] C. Morterra, G. Cerrato, J. Phys. Chem. 98 (1994) 12373.
45] T. Jin, T. Yamaguchi, K. Tanabe, J. Phys. Chem. 90 (1986) 4794.
46] F. Babou, G. Coudurier, J.C. Vedrine, J. Catal. 152 (1995) 341.
47] T. Yamaguchi, T. Jin, T. Ishiba, K. Tanabe, Mater. Chem. Phys. 17
Gen. 176 (1999) 27.
[
[
[
13] M. Hino, K. Arata, J. Chem. Soc. Chem. Commun. (1980) 851.
14] M. Hino, K. Arata, J. Am. Chem. Soc. 101 (1979) 6439.
15] T. Riemer, D. Spielbauer, M. Hunger, G. Mekhemer, H. Knozinger,
J. Chem. Soc. Chem. Commun. (1994) 1181.
[
[
[
[
[
[
[
[
[
[
16] K.B. Fogash, G. Yaluris, M.R. Gonzalez, P. Ouraipryvan, D.A. Ward,
E.I. Ko, J.A. Dumesic, Catal. Lett. 32 (1995) 241.
(1987) 3.
17] G. Yaluris, R.B. Larson, J.M. Kobe, M.R. Gonzalez, K.B. Fogash, J.A.
Dumesic, J. Catal. 158 (1996) 336.
18] F.G. Chen, G. Coudurier, J.-F. Joly, J.C. Vedrine, J. Catal. 143 (1993)
16.
19] M. Bensitel, O. Saur, J.-C. Lavalley, Mater. Chem. Phys. 17 (1987)
49.
20] F. Pinna, M. Signoretto, G. Strukul, G. Cerrato, C. Morterra, Catal.
Lett. 26 (1994) 339.
[
[
[
[
[
48] E. Escalona Platero, M. Penarroya Mentruit, Catal. Lett. 30 (1995) 31.
49] K.T. Wan, C.B. Khouw, M.E. Davis, J. Catal. 158 (1996) 311.
50] C. Zhang, R. Miranda, B.H. Davis, Catal. Lett. 29 (1994) 349.
51] L. Forni, Catal. Rev. 8 (1974) 65.
52] D. Farcasiu, A. Ghenciu, G. Marino, K.D. Rose, J. Am. Chem.
Soc. 119 (1997) 11826.
6
2
[
[
53] B.C. Gates, T.K. Cheung, Top. Catal. 6 (1998) 41.
54] A. Clearfield, G.P.D. Serrette, A.H. Khazi-Syed, Catal. Today 20
[
[
21] R. Srinivasan, R.A. Keogh, B.H. Davis, Catal. Lett. 36 (1996) 51.
22] F. Babou, B. Bigot, P. Sautet, J. Phys. Chem. 97 (1993) 11501.
(1994) 295.