Effect of the addition of Sn to zirconia on the acidic properties of the sulfated mixed oxide

Article abstract of DOI:10.1039/a605291f

Zirconium hydroxide, tin hydroxide and the mixed hydroxide of tin and zirconium (1:9) have been prepared, sulfated with a 0.1 M aqueous H₂SO₄ solution and calcinated at 873 K. The chemical composition of these solids has been characterized by chemical analysis, differential thermal analysis (DTA), X-ray photoelectron spectroscopy (XPS) and energy dispersive X-ray (EDX-) scanning transmission electron microscopy (STEM), and their structural and textural properties have been studied by BET surface area and XRD techniques. The nature of the acid sites (Bronsted or Lewis) was characterized by in situ IR study of pyridine adsorption and desorption and their amount was found to depend on the sulfate content. Catalytic properties have been studied for the isomerization reaction of n-butane to isobutane, in the 423 to 523 K range and for propan-2-ol conversion, in the 373 to 473 K range. Addition of Sn to ZrO₂ (SnO₂:ZrO₂ = 1:9) led to a solid solution and enhanced slightly (ca. 30%) the weak acid features of ZrO₂ and removed the basic properties of pure SnO₂ which was shown to convert propan-2-ol to acetone. Upon sulfation of ZrO₂, SnO₂-ZrO₂[1:9] and SnO₂, the acidity of the oxides was sharply enhanced. SO₄^2-/SnO₂, which was inactive for the isomerization reaction, was only 60% less active than SO₄^2-/ZrO₂ for propan-2-ol dehydration. This shows that sulfation of SnO₂ generates acid sites of moderate strength. The presence of only 10% of SnO₂ in ZrO₂ decreased the reaction rates per SO₄^2- by a factor of six for the initial n-butane isomerization reaction and by a factor of seven for the propan-2-ol dehydration to propene and diisopropyl ether. It is suggested that the presence of Sn decreases the electron acceptor properties of Zr and thus its acidity strength. Since the rate per sulfated species of both reactions decreased in the same proportion, one may consider that on SO₄^2-/ZrO₂ and SO₄^2-/SnO₂-ZrO₂ samples there is no site of moderate strength able to dehydrate propan-2-ol without isomerising n-butane.

Full text of DOI:10.1039/a605291f

View Article Online / Journal Homepage / Table of Contents for this issue
E†ect of the addition of Sn to zirconia on the acidic properties of the
sulfated mixed oxide
Anjali Patel, Gise
Institut de Recherches sur la Catalyse, CNRS, 2 Avenue Albert Einstein, F-69626 V illeurbanne
Cedex, France
` le Coudurier, Nadine Essayem and Jacques C. V e drine*

Zirconium hydroxide, tin hydroxide and the mixed hydroxide of tin and zirconium (1 : 9) have been prepared, sulfated with a 0.1 M
aqueous H SO solution and calcinated at 873 K. The chemical composition of these solids has been characterized by chemical
2
4
analysis, di†erential thermal analysis (DTA), X-ray photoelectron spectroscopy (XPS) and energy dispersive X-ray (EDXÈ) scan-
ning transmission electron microscopy (STEM), and their structural and textural properties have been studied by BET surface
area and XRD techniques. The nature of the acid sites (BrÔnsted or Lewis) was characterized by in situ IR study of pyridine
adsorption and desorption and their amount was found to depend on the sulfate content.
Catalytic properties have been studied for the isomerization reaction of n-butane to isobutane, in the 423 to 523 K range and
for propan-2-ol conversion, in the 373 to 473 K range. Addition of Sn to ZrO (SnO : ZrO \ 1 : 9) led to a solid solution and
2
2
2
enhanced slightly (ca. 30%) the weak acid features of ZrO and removed the basic properties of pure SnO which was shown to
2
2
convert propan-2-ol to acetone. Upon sulfation of ZrO , SnO ÈZrO [1 : 9] and SnO , the acidity of the oxides was sharply
enhanced. SO 2~/SnO , which was inactive for the isomerization reaction, was only 60% less active than SO 2~/ZrO for
propan-2-ol dehydration. This shows that sulfation of SnO generates acid sites of moderate strength. The presence of only 10%
2
2
2
2
4
2
4
2
2
of SnO in ZrO decreased the reaction rates per SO 2~ by a factor of six for the initial n-butane isomerization reaction and by a
2
2
4
factor of seven for the propan-2-ol dehydration to propene and diisopropyl ether. It is suggested that the presence of Sn decreases
the electron acceptor properties of Zr and thus its acidity strength. Since the rate per sulfated species of both reactions decreased
in the same proportion, one may consider that on SO 2~/ZrO and SO 2~/SnO ÈZrO samples there is no site of moderate
4
2
4
2
2
strength able to dehydrate propan-2-ol without isomerising n-butane.
The preparation and use of strong solid acids and superacids
are active areas of research for many reactions such as hydro-
viewpoint of their acidÈbase properties. Detailed studies have
been carried out on TiO ÈSiO ,28,35h39 ZrO ÈSiO ,40,41
2
2
2
2
carbon isomerization, alkylation, cracking and hydrocracking,
esteriÐcations, alcohol dehydration, acylation, methanol con-
version to gasoline etc. Skeletal isomerization of light normal
to C ) is an important reaction in petroleum
TiO ÈSnO 42 and TiO ÈZrO .43h45
2
2
2
2
To our knowledge, no detailed study has been carried out
on SnO ÈZrO . The present work deals with the synthesis,
2
2
alkanes (C4
characterization and study of the catalytic properties of
6
chemistry and is usually performed using liquid acids or Pt on
chlorinated alumina as catalysts. However, this results in
environmental problems related to the recovery of aggressive
and corrosive acids or to the continuous supply of reactants
unsulfated and sulfated ZrO , SnO and SnO ÈZrO
materials. Indeed, sulfated zirconia was extensively studied
and its strong acidic feature is well documented. Sulfation of
2
2
2
2
SnO was shown to considerably enhance its acid strength22
2
such as CCl in the course of the reaction.
and Matsuhashi et al.23 considered that sulfated SnO was
4
2
In order to overcome these problems, di†erent solid super-
almost as acidic as sulfated zirconia. According to TanabeÏs
acids have been proposed. Recent reviews have been
published on this subject,1h5 particularly on sulfated oxides
proposed to replace the Pt/chlorinated alumina catalyst. The
enhancement of catalytic activities for acid catalysed reactions
model for mixed oxides, SnO ÈZrO should present more
2
2
acidic features than its two component oxides but no predic-
tion can be made for its sulfated form. In this work, we have
tried to determine the acidity of these solids by studying their
catalytic activity in two acid-catalysed reactions: isomer-
ization of n-butane and propan-2-ol conversion, and to iden-
tify the nature of the acid sites by spectroscopic studies of the
adsorption of H O or D O (weak bases) and pyridine (strong
by the addition of SO 2~ has been reported for certain metal
4
oxides
such
as
ZrO ,6h17
TiO ,8,9,18
HfO ,19
2
2
2
Fe O ,8,9,18,20,21 SnO 8,9,22,23 etc. The high catalytic activ-
2
3
2
ities are considered to be due to the generation of strong acid
2
2
sites on the addition of SO 2~.
base).
4
Whereas considerable attention has been paid to sulfated
single metal oxides, only some studies have been devoted to
Experimental
sulfated binary oxides such as SO 2~/ZrO ÈAl O ,24
4
2
2 3
SO 2~/ZrO ÈSiO ,25,26 SO 2~/ZrO ÈTiO 27 or SO 2~/
4
2
2
4
2
2
4
Catalyst preparation
TiO ÈSiO .28
2
2
However, it is well known that mixing pure oxides is a good
way either to modulate the properties of the component
oxides or to create new sites. In particular, it has been found
that mixed oxides can develop strong acidity even when the
pure oxide components do not show it29h31 and several theo-
ries have been proposed to justify the enhanced activity of
mixed oxide.29,32h34
Zirconium(IV) hydroxide and tin(IV) hydroxide samples were
prepared by adding a 28 wt.% aqueous ammonia solution to
an aqueous solution of ZrOCl É 8H O (0.31 M) or SnCl (0.33
M), up to pH 8.4. The precipitates were aged at 373 K over a
waterbath for 1 h, washed with deionised water, until
chloride-free washing water was obtained, and dried at 393 K
for 24 h. The mixed hydroxide of zirconium(IV) and tin(IV)
with atomic ratio of Sn to Zr equal to 1 : 9 was prepared by
coprecipitation of a mixed solution of ZrOCl É 8H O and
2
2
4
Among the mixed oxides exhibiting enhanced acidity, the
binary oxides of the Group 4 metals are interesting from the
2
2
J. Chem. Soc., Faraday T rans., 1997, 93(2), 347È353
347

View Article Online
isomerization reaction was carried out over the temperature
range 423È523 K.
SnCl with 28 wt.% aqueous ammonia solution up to pH 8.4,
followed by ageing, washing and drying as in the case of zir-
conium and tin hydroxides.
For sulfation, x g of dry gel were introduced, at room tem-
4
Propan-2-ol conversion (3% in N , total Ñow rate \ 1.8
2
dm3 h~1), in the 393È443 K range, was studied to characterize
perature, into 15x cm3 of a 0.1 M H SO solution for 20 min
acidic and basic properties.
2
4
under stirring and then Ðltered without washing, dried at
Before the two catalytic reactions, the samples, calcined at
873 K, were activated in the reactor at 673 K, for 2 h in air
Ñow.
3
4
93 K for 24 h, and calcinated 873 K with a heating rate of
K min~1 for 4 h. The solids obtained after sulfation and
calcination will be designated below as SO 2~/ZrO ,
SO 2~/SnO ÈZrO and SO 2~/SnO .
4
2
4
2
2
4
2
Results and Discussion
Characterization
Physicochemical characterization
Chemical analyses were performed at the Central Analysis
Service Laboratory of CNRS in Solaise, France. The sulfur
content was determined upon combustion under oxygen at
The chemical composition and the surface area values of the
calcined samples are reported in Table 1. It is obvious, as
already observed for sulfated zirconia,14 that the presence of
sulfated species stabilised a higher surface area. Nevertheless,
this behaviour was less pronounced in the case of the binary
oxide. The sulfur content corresponded to a monolayer of
2
073 K and titration of SO evolved by IR.
XRD studies were performed with a Cu anticathode and a
2
Phillips PW 1710 goniometer.
Thermal gravimetry (TG) and di†erential thermal analyses
SO 2~ species for sulfated tin oxide (assuming that one
4
4
(
DTA) were performed, between 293 and 1223 K, on a
SO 2~ species occupies 0.25 nm2)12 and to a lower coverage
SETARAM TGDTA A92 equipment with 32.4 mg of solid, a
heating rate of 5 K min~1 and an air Ñow rate of 20 cm3
min~1. Platinum crucibles were used as sample holders and
a-Al O was the reference material.
for the other two sulfated oxides, and particularly for the
mixed oxide. The sulfur content being about the same before
calcination, these results were attributed to lower absorption
capacity of the mixed hydroxide.
2
3
Surface area measurements and pore volume determi-
nations were carried out at liquid-nitrogen temperature in an
automated laboratory made equipment on the samples pre-
viously outgassed at 473 K for 2 h.
XRD. It appears that all samples were amorphous (Fig. 1)
before calcination and were transformed by calcining at 873 K
into ZrO monoclinic phase for ZrO and SnO ÈZrO
XPS experiments were carried out in vacuo (10~6 Pa) with
an ESCA LAB VG spectrometer for the di†erent samples as
such, i.e. without any treatment before and after introduction
into the apparatus.
IR studies were made on self-supported pellets placed inside
a cell allowing in situ treatments. The self-supported pellets
2
2
2
2
samples, and into SnO tetragonal phase for SnO and
2
2
SO 2~/SnO samples. On the other hand, sulfated zirconia
and sulfated mixed oxide crystallized mainly in the form of the
ZrO tetragonal phase and no pure SnO phase was detected.
This indicates that a solid solution may probably be formed.
However, there was no variation in the cell parameters as is
generally expected for a solid solution. This may however, be
due to the very close values of the ionic radii (0.72 and 0.69
nm, for Zr4` and Sn4`, respectively).
4
2
2
2
(
23 mg) were prepared from precursor (sulfated hydroxide).
They were calcinated in situ in the IR cell under air at 873 K
for 4 h and cooled to ambient temperature, and left under
ambient atmosphere for rehydration. They were then evac-
uated at 473 K for 2 h at 10~3 Pa. H O (or D O) was
It is also important to note that for all sulfated samples
after calcination at 873 K, the XRD peaks were broader than
for their unsulfated analogues. This can arise either from
smaller particle size in coherence with the enhancement of
surface area or from more defect crystallites.
2
2
adsorbed at 295 K under saturated vapour pressure and then
progressively desorbed at increasing temperatures.
Pyridine was adsorbed under saturation vapour pressure at
295 K, on the samples previously evacuated at 473 K for 2 h
under 10~3 Pa vacuum, and then desorbed for 1 h at ambient
temperature, 423, 523, 623 and 673 K. The same studies were
also performed for the samples which after calcination in air
were evacuated at 673 K for 2 h.
The spectra were recorded with an FTIR spectrophotom-
eter, BRUKER IFS 48, with 500 scans.
TG–DTA analysis. TGÈDTA studies of the uncalcined
hydroxides showed two weight losses. The Ðrst, at 373 K,
accompanied by an endothermal peak, was attributed to the
evolution of adsorbed water. The second occurred contin-
uously between 373 and 873 K and very probably corre-
sponds to the dehydroxylation process. For SnO , no thermal
2
peak was detected, while for ZrO and SnO ÈZrO , an
Catalytic testing
2
2
2
intense and sharp exothermal peak was observed at 701 and
752 K, respectively. Such an exothermal peak (glow exotherm)
has often been described for zirconium hydroxide46 and was
ascribed to the crystallisation of the amorphous phase into
monoclinic or tetragonal phase.
Catalytic properties were studied for n-butane isomerization
using a di†erential microreactor and gas chromatography
detection on line. The Ñow rate of n-butane was 0.2 dm3 h~1
with 200 mg catalyst and N (2 dm3 h~1) as carrier gas. The
2
Table 1 Chemical composition from chemical analysis and surface area values of the unsulfated and sulfated samples calcinated at 873 K
sulfur content
surface area
/m2 g~1
Sn content
(wt.%)
sample
(wt.%)
/lmol m~2
ZrO
27
94
0
1.6
0
5.3
È
È
2
SO 2~/ZrO
4
2
SnO ÈZrO
31
53
0
0.6
0
3.4
9.4
9.5
2
2
SO 2~/SnO ÈZrO
4
2
2
SnO
13
66
0
1.6
0
7.6
78.8
75.0
2
SO 2~/SnO
4
2
348
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
Table 2 Bulk and superÐcial compositions determined by chemical analysis, TG, EDXÈSTEM and XPS
S/(Zr ] Sn) atomic ratio Sn/Zr atomic ratio
sample
chem
TGA
XPS
EDX
chem
XPS
EDX
SO 2~/ZrO
0.065
0
0.072
0
0.12
0
È
È
È
0.1
0.1
È
È
0.13
0.1
È
È
È
4
2
SnO ÈZrO
2
2
SO 2~/SnO ÈZrO
0.025
0.079
0.031
0.071
0.067
0.087
0.039
È
0.09
È
4
2
2
2
SO 2~/SnO
4
Fig. 2 DTA of the pure oxides: (a) ZrO , (b) SnO ÈZrO , (c) SnO
2
2
2
2
2
and of the sulfated oxides: (a@) SO 2~/ZrO , (b@) SO 2~/SnO ÈZrO
4
2
4
2
and (c@) SO 2~/SnO
4
2
water and the second, which started above 973 K correspond-
ed to the evolution of SO . The amounts of S, calculated from
2
these weight losses, allowed us to obtain the S/(Zr ] Sn)
atomic ratios reported in Table 2 which are in good agree-
ment with the chemical analyses data.
EDX–STEM. The mixed oxide was composed of small par-
ticles, homogeneous in size (diameter \ 20 nm). Their com-
position, determined by EDX analysis, (Table 2) for di†erent
grains and at di†erent places of the same grain was very
homogeneous and close to the chemical composition either for
S/(Sn ] Zr) atomic ratio or for Sn/Zr atomic ratio, in agree-
ment with the existence of a solid solution.
Fig. 1 XRD patterns of (a) the pure oxides and (b) the sulfated
oxides. M: monoclinic phase. T: tetragonal phase.
Note that, for the mixed hydroxide, this transition was
shifted to higher temperature, indicating that it is not a simple
physical mixture of ZrO and SnO but a true solid solution,
XPS. In Table 3, the core binding energy values determined
by XPS for all samples are reported. It is seen that the values
of the electron binding energy for Zr 3d, Sn 3d and S 2p
2
2
3@2
in good agreement with the XRD data.
levels were in good agreement with the literature values which
For the uncalcined SO 2~/SnO sample, a third weight
are 182.2, 486.6 and 169.3 eV, respectively, indicating that the
crystallite surfaces of the three sulfated samples contained Zr
and Sn ions only in their 4] oxidation state and S in the 6]
oxidation state.
Note also, that there was a little increase in the binding
energy value of Zr 3d and Sn 3d after sulfation, which may
4
2
loss accompanied by a small endothermal peak occurred
above 923 K up to 1173 K. This corresponds to the evolution
of SO . For
the
uncalcined SO 2~/ZrO
and
2
4
2
SO 2~/SnO ÈZrO samples, SO was evolved in two steps.
4
2
2
2
The Ðrst occurred at 925 K and was accompanied by an endo-
thermal peak which superimposed the right side of an exo-
thermal peak (observed at 845 and 887 K for SO 2~/ZrO
Table 3 Core electron binding energiesa (eV) for di†erent samples
(^0.2 eV)
4
2
and SO 2~/SnO ÈZrO , respectively). The second was con-
4
2
2
tinuous up to 1173 K. The DTA curves for the six samples are
reported in Fig. 2. As already described,7,47 it is noteworthy
that the sulfation reduced the peak amplitude and shifted it to
higher temperatures. As with unsulfated samples, the exother-
mal peak is observed at higher temperature for the mixed
oxide. Hence, it may be concluded that the formation of the
solid solution and the introduction of sulfate ions modify the
phase transition.
sample
O 1s
Zr 3d
Sn 3d
S 2p3@2
ZrO
530.1
530.4
530.0
530.3
529.9
182.3
182.6
182.1
182.5
È
2
SO 2~/ZrO
169.3
4
2
SnO ÈZrO
486.5
486.7
486.8
2
2
SO 2~/SnO ÈZrO
169.3
169.0
4
2
2
2
SO 2~/SnO
The TG curves of the calcined sulfated samples were com-
posed of two weight losses. The Ðrst at 393 K with a small
endothermal peak, corresponded to the evolution of adsorbed
4
a Charging e†ect corrected by adjusting the main C 1s line to 284.7
eV.
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
349

View Article Online
indicate that the metal atom became slightly electron deÐcient,
presumably by transfer of electron density to the sulfated
species. However, the accuracy of the measurement was too
poor (^0.2 eV) to be really signiÐcant. Fig. 3 shows the O 1s
core-level spectra. The main peak at 530.2 ^ 2.2 eV, for the
sulfated and unsulfated oxides, corresponds to the oxide
oxygen. For the sulfated oxides, a broadening of the peak was
observed on the high-energy side as already described for sul-
or ZrO .2,48 This indicates the presence of two
fated TiO2
2
types of O2~ ions, one belonging to the oxides and the other,
to the sulfated species.
For the mixed oxide, sulfated or not, Sn/Zr atomic ratio
values from XPS analyses were very close to those from
chemical analysis. This conÐrms our above statement, that the
mixed oxide is a solid solution. The value of S/(Sn ] Zr)
atomic ratio (0.067) for the sulfated mixed oxide was about
three times higher than that of the chemical analysis values
Fig. 4 IR spectra of (a) the sulfated oxides calcinated at 873 K,
equilibrated in ambient air at room temperature and evacuated in the
IR in situ cell at 473 K, (b) after further adsorption of pyridine at
room temperature
(
0.025) showing that the sulfur concentration was higher on
the surface than in the bulk. The same trend was observed for
the sulfated zirconia.
D O, that there exists an equilibrium on the surface between
2
(
SO ) , (H SO ) , (HSO ~)
and (SO 2~) species.18
3
ads
2
4 ads
4
ads
4
Similar
behaviour
was
observed
presently
for
IR. The spectra in KBr pellets (1%) of all sulfated precur-
sors after drying at 393 K, were similar to those observed for
SO 2~/SnO ÈZrO as demonstrated by the spectra reported
4
2
2
in Fig. 5. When, after outgassing at 473 K [Fig. 5(a)], D O
2
sulfated ZrO 18 and in good agreement with Sohn and Kim10
was added upon saturation, the spectrum shifted to lower
2
who assigned the four bands to the l vibration (1207, 1136
wavenumber and consisted of three main bands at 1212, 1126
and 1032 cm~1 and a shoulder at 992 cm~1. Such a spectrum
was very similar to that observed after calcination and re-
equilibration in ambient atmosphere and is attributed to
3
and 1053 cm~1) and l vibration (997 cm~1) of a bidentate
1
SO 2~ ion in C symmetry. The spectra of the sulfated ZrO
4
2l
2
and the sulfated SnO ÈZrO samples were strictly identical
2
2
while, for SnO , the bands were slightly shifted (1195, 1146,
1
(
SO 2~)
ads .
2
4
050 and 983 cm~1). Intense bands were observed at ca. 3400
and 1640 cm~1, due to the OH groups of the hydroxides and
Progressive desorption, by outgassing at increasing tem-
perature, restored the initial spectrum via two intermediate
species. The Ðrst, observed upon outgassing between ambient
temperature and 323 K, was characterized by a band at 1300,
attributed to the l vibration of (SxO) bonds in (HSO ~)
of adsorbed water.
As already observed,18 the main di†erence, after calcination
at 873 K under air Ñux and re-equilibrium in ambient atmo-
sphere was the drastic intensity decrease of the hydroxyl
group vibration at 3400 cm~1, the sulfated vibrations being
only slightly shifted to higher wavenumbers. Upon evacuation
of self-supported wafers, at 473 K for 2 h, the vibrations of the
sulfated species shifted to higher wavenumbers leading, as
shown in Fig. 4(a), to the formation of a new intense band at
as
4 ads
species and the second, observed after outgassing at 373 K,
exhibited an intense band at 1332 cm~1 due the same vibra-
tion in (H SO ) . These species were very similar to those
2
4 ads
observed with SO 2~/ZrO (bands at 1320 and 1368 cm~1,
4
2
respectively)18 although the vibrations were shifted towards
lower wavenumbers.
1
387,
1360
and
1271
cm~1
for
SO 2~/ZrO ,
Thus, one may conclude that, after outgassing the sample at
4
2
SO 2~/SnO ÈZrO and SO 2~/SnO , respectively, which
4
73 K, the main sulfated species was (SO ) with (HSO ~)
4
2
2
4
2
3
ads
4 ads
was attributed to the l vibration of (SO )
.
and (H SO )
in more or less large amounts, as already
as
3 ads
2
4 ads
It was shown previously, by outgassing sulfated zirconia at
di†erent temperatures and then rehydrating progressively
observed for the SO 2~/ZrO sample.
Thus, the model we have proposed in ref. 18 and 49 which
consists of (H SO ) adsorbed on the zirconia surface can be
4
2
stepwise, by adding known incremental amounts of H O or
2
2
4 ads
extended to the case of the sulfated mixed oxide.
Fig. 5 IR study of the desorption of D O. (a) SO 2~/SnO ÈZrO
2
4
2
2
after evacuation at 473 K, (b) after saturation with 2.3 kPa D O and
2
after desorption for 1 h (c) at ambient temperature, (d) at 373 K, (e) at
Fig. 3 XPS of O 1s for (a) the pure oxides and (b) the sulfated oxides
423 K and (f) at 473 K.
350
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
intensity decreased more slowly upon outgassing than for the
other bands. For the SO 2~/ZrO sample, the L , L , and B
Note that the l vibrations of the species correspond
as
exactly to those of HSO ~ in solution50 and liquid H SO 51
4
2
4
4
2
1
2
for SO 2~/ZrO indicating that the acidic strength is about
bands were about three times more intense than for
4
2
equal to that of 100% H SO , its Ðrst acidity, as schematized
SO 2~/SnO ÈZrO , in good agreement with the S content
2
4
4
2
2
below (Scheme 1), being the strongest one.
determined by chemical analysis. Thus, the amount of acid
sites per sulfated species is the same for the two samples.
Evacuation of the SO 2~/SnO ÈZrO sample at 673 K
For the sulfated mixed oxide these l vibrations shifted
as
toward lower wavenumbers indicating a lower bond order for
4
2
2
the SxO bonds, implying a lower negative charge on the
bridging oxygens and, consequently, a lower positive charge
as compensating charge, could indicate a lower BrÔnsted
acidic strength, as we will discuss below in view of the cata-
lytic reaction results.
[Fig. 6(b)] led to an increase in the Lewis acidity at the
expense of the BrÔnsted acidity.
The curves (b) in Fig. 4 show that, upon pyridine adsorp-
tion, the sulfate bands near 1360 were shifted to 1300 cm~1,
for the SO 2~/SnO ÈZrO sample. Under the same condi-
4
2
2
The adsorption of pyridine at ambient temperature on the
tions, the shifts were 1387 to 1306 cm~1 and 1271 to 1240
cm~1 for SO 2~/ZrO and SO 2~/SnO , respectively. By
SO 2~/SnO ÈZrO
sample calcinated at 873 K, re-
equilibrated under ambient atmosphere and evacuated at
73 K resulted in the appearance of bands due to pyridine
4
2
2
4
2
4
2
desorption, progressive and reversible back shifts occurred for
the three samples. Thus, the following identiÐcation for the
acid sites is proposed:
4
adsorbed on Lewis and BrÔnsted acid sites [Fig. 6(a)]. The
bands at 1542 and 1638 cm~1 were assigned to pyridine
adsorbed on BrÔnsted sites. Two Lewis-acid sites were
BrÔnsted acidity:
(
H SO ) ] Py % PyH` ] (HSO ~)
detected by the presence of bands at 1444 and 1608 cm~1 (L
2
4 ads ads
4
1
sites) and at 1459 and 1630 cm~1 (L sites). The evolution of
2
(HSO ~) ] Py % PyH` ] (SO 2~)
the intensity of the bands at 1445 cm~1 (L species), 1459
4
4
ads
1
cm~1 (L species) and 1540 cm~1 (BrÔnsted species) as a func-
tion of desorption temperature, showed that the species L
Lewis acidity L :
2
2
2
corresponded to the sites which are more acidic, since its
(SO ) ] Py % Py: ] (SO )
3
ads 3 ads
Lewis acidity L :
1
(Sn)Zr4` ] Py % Py: ] (Sn)Zr4`
if some Zr4` or Sn4` ions were accessible.
The formation of pyridinium ion shifts the equilibrium:
(
H SO ) % (HSO ~) ] H` % (SO 2~) ] 2H`
2
4 ads ads ads
4
4
to the right, and one may suggest that only sulfated (SO 2~)
4
ads
and Py: ] (SO ) species exist under pyridine saturation. As
3
ads
the spectrum of (SO 2~) is characterized by bands at 1212,
4
ads
1
1
126 and 1032 cm~1, the bands observed at 1306, 1300 and
240 cm~1 after pyridine adsorption on ZrO , SnO ÈZrO
2
2
2
and SnO , respectively, may be attributed to the l vibration
2
as
in the Py: ] (SO ) species. The electronic transfer from the
3
ads
pyridine molecule to S6` ion should induce a weakening of
the SxO bonds and, in consequence, a red shift of the l
as
vibration as compared to dehydrated samples.
Using the same model as Babou et al.,18,49 the force con-
stants, f, of the SxO bonds were calculated for the sulfated
oxides, before and after adsorption of pyridine. As proposed
by Jin et al.,52 the bond order, n, was obtained using the fol-
lowing empirical relationship:53 n \ 0.111f ] 0.7 and the
charge q borne by the O atom was calculated from: q \ 2
0
0
Scheme 1
[
n.
The data reported in Table 4 show that the bond order of
SxO, i.e. the covalent double-bond character decreased with
the nature of the oxide: ZrO [ SnO ÈZrO [ SnO and
2
2
2
2
decreased upon pyridine adsorption.
Table 4
l
vibration of (SO ) , force constant and bond order of
the SxO bond, charge borne by O atom
as
3 ads
l
f
charge
on O
as
sample
/cm~1
/102 N m~1
n
SO 2~/ZrO
1391
1306
10.3
9.6
1.83
1.77
[0.17
[0.23
4
2
]
Py
SO 2~/SnO ÈZrO
1360
1300
9.9
9.4
1.80
1.74
[0.20
[0.26
4
2
2
]
Py
Fig. 6 IR study of pyridine adsorption on SO 2~/SnO ÈZrO (a)
4
2
2
after evacuation at 473 K, (b) after evacuation at 673 K. 1, saturation
with 2 kPa pyridine, and after desorption for 1 h, (2) at ambient tem-
perature, (3) at 423 K, (4) at 523 K and (5) at 623 K.
SO 2~/SnO
1271
1240
9.1
8.9
1.71
1.69
[0.29
[0.31
4
2
]
Py
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
351

View Article Online
Table 5 IPA conversion at 393 K after 150 min of time on stream
intrinsic rate of IPA conversion
/
10~10 mol s~1 m~2
S
TONa
/10~3 s~1
to acetone
(%)
sample
propene
DIPE
acetone
SO 2~/ZrO
96.6
14
73.8
57.2
14.4
43.6
0
0
0.3
2.9
0.4
1.7
0
0
0.3
4
2
SO 2~/SnO ÈZrO
4
2
2
2
SO 2~/SnO
4
3
% IPA in N , total Ñow rate: 1.8 dm3 h~1, catalyst weight: 100 mg. S: selectivity. a TON for the IPA conversion into propene and DIPE and
2
assuming the whole S on the surface of the crystallites.
Considering a simpliÐed version54 of MortierÏs formula55
For sulfated samples, a turnover number (TON) was calcu-
lated which represents the total rate of IPA dehydration (into
propene and DIPE) per sulfated species as determined by
chemical analysis. Note, that the SO 2~/SnO sample was 1.7
(
derived from the charge equalization principle from
Sanderson),56 we have shown, for a series of sulfated oxides,
that the acidity strength can be correlated to the charge borne
by the cation of the non-sulfated oxide.57 Particularly, it was
4
2
times less active than the SO 2~/ZrO sample while the
4
2
calculated that the charge borne by Zr in ZrO and Sn in
SO 2~/SnO ÈZrO one was seven times less active.
2
4
2
2
SnO equals 0.78 and 0.51, respectively.58 Interaction of SO
Addition of sulfated species to the three oxides strongly
enhanced their total acidic properties. If one considers the
extrapolated rates at 443 K (with an activation energy equal
to 108 kJ mol~1)59 the rates of IPA dehydration were multi-
plied, upon sulfation, by factors of 2000, 320 and 870 for
ZrO , SnO ÈZrO and SnO , respectively, showing the very
2
3
with Zr, which is more electropositive than Sn, may induce a
higher electron withdrawing, leading to a higher bond order
for the SxO bond and, consequently, a lower negative charge
on the oxygens of (SO ) , which is in good agreement with
3
ads
the IR data described here.
2
2
2
2
high activity of the superÐcial sulfated species.
Catalytic Results
n-Butane isomerization. For n-butane isomerization, all non-
sulfated oxides, as well as the sulfated tin oxide, were totally
inactive, whatever the reaction temperature up to 573 K, the
activation temperature and the carrier gas. In Table 7, the
Propan-2-ol conversion. For propan-2-ol (IPA) conversion,
the activities are so di†erent that, in order to preserve a con-
version level between 2 and 10%, the reaction was studied at
results obtained at 473 K, with N as carrier gas, are reported
2
for the solids calcinated at 873 K and activated before reac-
393 K for the sulfated samples and at 443 K for the others.
tion at 673 K for 2 h under air Ñow. A TON was calculated
which represents the isomerization rate per sulfated species.
Note, that the sulfated mixed oxide was ca. six times less
active than the sulfated zirconia.
Three di†erent behaviours were observed as shown in Tables
5
and 6.
Pure ZrO exhibited well known amphoteric properties, as
2
evidenced by the formation of propene, diisopropyl ether
Comparison of these data with those obtained in IPA con-
version allows one to conclude that SO 2~/ZrO contained
(
DIPE) and acetone, and pure SnO had essentially basic fea-
2
tures. As expected from TanabeÏs assumption, the mixed oxide
was more than its two components, but only slightly.
4
2
acid sites of high strength while SO 2~/SnO contains almost
4
2
the same amount of acid sites (only 1.3 times less) but they are
of moderate strength.
The sulfated mixed oxide is ca. 6È7 times less active per
sulfated species than SO 2~/ZrO in the two reactions.
4
2
Table 6 IPA conversion at 443 K after 150 min of time on stream
However, the amount of acid sites per sulfated sites, as deter-
mined by pyridine adsorption, is the same for the two
samples. Hence, one may relate the activity decrease to the
intrinsic rate of formation
/
10~10 mol s~1 m~2
selectivity
to acetone
(%)
acidic strength decrease as evidenced by the shift of the l
as
samples
propene
DIPE
acetone
vibrations of the SxO bonds. Moreover, the activities
decreased in the same proportion for the two reactions, which
may indicate that the distribution of acid strength is narrow
for the two samples, i.e. there is no site of moderate acidic
strength.
ZrO
2.0
3.1
0
1.2
0.5
5.5
0.6
0
82.9
16
0
93.8
2
SnO ÈZrO
2
SnO
2
2
3
mg.
% IPA in N , total Ñow rate: 1.8 dm~3 h~1, catalyst weight: 100
Since we have shown that the mixed oxide was a solid solu-
tion without surface enrichment in Sn, and as the Sn : Zr
2
Table 7 n-Butane isomerization at 473 K after 4 and 84 min of time on stream
intrinsic rate of isobutane formation
selectivity to isobutane
(%)
/
10~10 mol s~1 m~2
TON
10~3 s~1
4 min
/
sample
4 min
84 min
4 min
87
84 min
95
95
È
SO 2~/ZrO
95.4
11.7
0
17.1
5.0
0
1.8
0.3
0
4
2
SO 2~/SnO ÈZrO
88
È
4
2
2
2
SO 2~/SnO
4
9
% n-butane in N , total Ñow rate: 2.2 dm3 h~1, catalyst weight: 200 mg.
2
352
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
drine, J. Catal., 1995, 152,
1
8
F. Babou, G. Coudurier and J. C. Ve

atomic ratio is equal to 1 : 9, it is proposed that each Sn
cation perturbs a large number of neighbouring Zr cations, i.e.
the e†ect is collective.
3
41.
1
2
9
0
K. Arata and M. Hino, React. Kinet. Catal. L ett., 1984, 25, 143.
M. Hino, S. Kobayashi and K. Arata, React. Kinet. Catal. L ett.,
1
981, 18, 491.
Conclusion
2
2
2
1
2
3
A. Kayo, T. Yamaguchi and K. Tanabe, J. Catal., 1983, 83, 99.
G. W. Wang, H. Hattori and K. Tanabe, Chem. L ett., 1983, 277.
H. Matsuhashi, M. Hino and K. Arata, Appl. Catal., 1990, 59,
The coprecipitation of SnO and ZrO in an atomic ratio
2
2
equal to 1 : 9 leads to a solid solution without Sn or Zr super-
Ðcial enrichment as evidenced by XRD, DTA, XPS and EDXÈ
STEM techniques. In good agreement with TanabeÏs
2
05.
24 T. C. Luy, J. C. Yori, A. A. Castro and J. M. Parera, React. Kinet.
Catal. L ett., 1988, 36, 275.
2
5 T. Ishida, T. Yamaguchi and K. Tanabe, Chem. L ett., 1988, 1869.
hypothesis this mixed oxide is slightly more acid than ZrO
2
26 C. Guo, Z. Yu, Z. Qian, J. Huang and Y. Xu, Stud. Surf. Sci.
while SnO is rather basic, as shown by the dehydrogenation
2
Catal., 1994, 90, 543.
of IPA to acetone.
2
7
F. Lo
 nyi, J. Valyon, J. Engelhardt and F. Mizukami, J. Catal.,
Sulfation modiÐes very strongly the properties of the pure
1
996, 160, 279.
and mixed oxides. Sulfated SnO becomes highly active for
28 J. R. Sohn, H. J. Jang, M. Y. Park, E. H. Park and S. E. Park, J.
2
IPA dehydration to propene, but its acidity is not strong
Mol. Catal., 1994, 93, 149.
2
3
3
9
0
1
K. Tanabe, Bull. Chem. Soc. Jpn., 1974, 47, 1064.
H. A. Benesi and B. H. C. Windquist, Adv. Catal., 1978, 27, 97.
K. Tanabe, M. Misono, Y. Ono and H. Hattori, Stud. Surf. Sci.
Catal., 1989, 51, 108.
enough to catalyse n-butane isomerization. The highest acidity
was observed for the sulfated zirconia, in good agreement with
what is known in the literature. The sulfated mixed oxide
exhibited a lower activity in the two test reactions than was
expected for a simple mixture (additivity of the activities).
3
2
T. Seiyama, Metal Oxides and T heir Catalytic Actions,
Kodansha, Tokyo, 1978.
Moreover, the l vibrations of (SO ) , (HSO ~)
as 3 ads
H SO ) species were observed to shift from 1387 to 1360
and
33 T. Kataoka and J. A. Dumesic, J. Catal., 1986, 112, 66.
H. H. Kung, J. Solid State Chem., 1984, 52, 191.
35 M. Itoh, H. Hattori and K. Tanabe, J. Catal., 1974, 35, 225.
4
ads
3
4
(
2
4 ads
cm~1, 1320 to 1300 cm~1 and 1368 to 1332 cm~1, respec-
tively, upon Sn addition. This corresponds to lower SxO
bond order, thus to a less negative charge borne by the bridg-
ing oxygens, which allows us to suggest that the presence of
Sn decreases the electron-acceptor properties of Zr, i.e. its
positive charge and subsequently the BrÔnsted acidity
attached to the sulfated species (HSO ~) and (H SO )
3
3
3
6
7
8
H. Hattori, M. Itoh and K. Tanabe, J. Catal., 1975, 38, 172.
H. Hattori, M. Itoh and K. Tanabe, J. Catal., 1976, 41, 46.
I. C. V. Odenbrand, M. G. J. Brandin and G. Busca, J. Catal.,
1992, 135, 505.
39 Z. Liu, J. Tabora and R. J. Davis, J. Catal., 1994, 149, 117.
4
0
H. J. M. Bosman, E. C. Kruissink, J. van der Spoel and F. van
den Brink, J. Catal., 1994, 148, 660.
41 J. B. Miller, S. E. Rankin and E. I. Ko, J. Catal., 1994, 148, 682.
4
ads
2
4 ads .
4
4
2
3
S. M. Maurer, D. Ng and E. I. Ko, Catal. T oday, 1993, 16, 319.
K. Arata, S. Akutagawa and K. Tanabe, Bull. Chem. Soc. Jpn.,
1976, 49, 390.
References
1
K. Tanabe, M. Misono, Y. Ono and H. Hattori, Stud. Surf. Sci.
Catal., 1989, 51, 199.
44 Y. Hirashima, K. Nishiwaki, A. Miyakoshi, H. Tsuiki, A. Veno
and H. Nakabayashi, Bull. Chem. Soc. Jpn., 1988, 61, 1945.
45 J. A. Navio, M. Macias, F. J. Marchena, J. M. Campelo and J. M.
Marinas, in New Frontiers in Catalysis, ed. L. Guczi, F. Solymosi
2
3
K. Arata, Adv. Catal., 1990, 37, 165.
K. Tanabe, H. Hattori and T. Yamaguchi, Crit. Rev. Surf. Chem.,
1
990, 1, 1.
4
5
6
T. Yamaguchi, Appl. Catal., 1990, 61, 1.
and P. Te te nyl, Akede mia Kiado, Budapest, 1993, p. 2597.
A. Corma and A. Martinez, Catal. Rev. Sci. Eng., 1993, 35, 483.
M. Hino and K. Arata, J. Chem. Soc., Chem. Commun., 1979,
46 P. D. L. Mercera, J. G. van Ommen, E. B. Doesburg, A. J. Bur-
graaf and J. R. H. Ross, Appl. Catal., 1990, 57, 127.
47 R. Srivinasan, R. A. Keogh, D. R. Milburn and B. H. Davis, J.
Catal., 1995, 153, 123.
48 J. P. Chen and R. T. Yang, J. Catal., 1993, 139, 277.
49 F. Babou, B. Bigot and Ph. Sautet, J. Phys. Chem., 1993, 97,
11501.
1
148, 1980, 851.
7
8
9
T. Yamaguchi, K. Tanabe and Yao Chin Kung, Mater. Chem.
Phys., 1986, 7, 67.
T. Yamaguchi, T. Jin and K. Tanabe, J. Phys. Chem., 1986, 90,
3
148.
T. Yamaguchi, T. Jin, T. Ishida and K. Tanabe, Mater. Chem.
Phys., 1987, 17, 3.
50 T. Jin, T. Yamaguchi and K. Tanabe, J. Phys. Chem., 1986, 90,
4794.
1
1
0
1
J. R. Sohn and H. W. Kim, J. Mol. Catal., 1989, 52, 361.
K. I. Mukaida, T. Miyoshi and T. Satoh, in AcidÈbase Catalysis
I, ed. K. Tanabe, H Hattori, T. Yamaguchi and T. Tanaka,
Kodansha, Tokyo, 1989, p. 361.
51 R. J. Gillepsie and E. A. Robinson, Can. J. Chem., 1962, 40, 644.
52 P. A. Gigue` re and R. Savoie, Can. J. Chem., 1963, 41, 287.
53 R. J. Gillespie, E. A. Robinson, Can. J. Chem., 1963, 41, 2074.
54 M. Henry, PhD Thesis, University of Paris VI, 1988.
55 W. J. Mortier, J. Catal., 1978, 55, 138.
1
2
F. R. Chen, G. Coudurier, J. F. Joly and J. C. Ve
Symp. Ser., 1991, 36, 878.

drine, in ACS
56 R. T. Sanderson, Chemical Bonds and Bond Energies, Academic
Press, New York, 2nd edn., 1976.
1
1
3
4
J. M. Parera, Catal. T oday, 1992, 15, 481.
F. R. Chen, G. Coudurier, J. F. Joly and J. C. Ve

drine, J. Catal.,
57 F. Babou, G. Coudurier and J. C. Ve
92, 1457.
58 F. Babou, PhD Thesis, University of Lyon I, 1994, no. 88È94.
59 A. Ouqour, G. Coudurier and J. C. Vedrine, J. Chem. Soc.,
Faraday T rans., 1993, 89, 3151.
 drine, J. Chim. Phys., 1995,
1
993, 143, 616.
1
1
1
5
6
7
F. Babou, B. Bigot, G. Coudurier, Ph. Sautet and J. C. Ve drine,
Stud. Surf. Sci. Catal., 1994, 90, 519.

C. Morterra, G. Cerrato, F. Pinna, M. Signoretto and G. Strukul,
J. Catal., 1994, 149, 181.
Z. Gao, J. M. Chen, W. M. Hua and Y. Tang, Stud. Surf. Sci.
Catal, 1994, 90, 507.
Paper 6/05291F; Received 29th July, 1996
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
353