Synthesis and characterization of ZrO2 as an acid-base catalyst: Dehydration-dehydrogenation of propan-2-ol

Article abstract of DOI:10.1039/a606813h

Two ZrO₂-based catalysts have been synthesized by thermal treatment of a gel prepared from ZrOCl₂·8H₂O. The effect of different preparation parameters on the final properties of the solids was studied. Textural characterization was carried out from nitrogen adsorption-desorption isotherms. XRD and FT-Raman spectroscopy have been used to elucidate the physical state of the catalysts. Their acid-base properties were studied by using temperature-programmed desorption-mass spectrometry (TPD-MS) and diffuse reflectance FTIR techniques (DRIFT), employing pyridine, 2,6-dimethylpyridine and carbon dioxide as probe molecules. Both solids were tested for the dehydration-dehydrogenation of propan-2-ol. The reactivity of the catalysts was studied through isotherm and temperature-programmed surface reactions (TPSR), in a flow reactor. In addition, temperature-programmed oxidation (TPO) and temperature-programmed desorption (TPD) experiments revealed that catalyst deactivation was caused by carbon deposition and irreversible adsorption of propene formed on the catalytic surface.

Full text of DOI:10.1039/a606813h

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis and characterization of ZrO as an acid–base catalyst
2
Dehydration–dehydrogenation of propan-2-ol
M. Angeles Aramend•a, Victoriano Borau, Cesar Jimenez, Jos e M. Marinas, Andres Porras and
Francisco J. Urbano
Department of Organic Chemistry, Faculty of Sciences, University of Cordoba, Avda San Alberto
Magno s/n, E-14004 Cordoba, Spain
Two ZrO -based catalysts have been synthesized by thermal treatment of a gel prepared from ZrOCl É 8H O. The e†ect of
2
2
2
di†erent preparation parameters on the Ðnal properties of the solids was studied. Textural characterization was carried out from
nitrogen adsorptionÈdesorption isotherms. XRD and FT-Raman spectroscopy have been used to elucidate the physical state of
the catalysts. Their acidÈbase properties were studied by using temperature-programmed desorptionÈmass spectrometry (TPDÈ
MS) and di†use reÑectance FTIR techniques (DRIFT), employing pyridine, 2,6-dimethylpyridine and carbon dioxide as probe
molecules. Both solids were tested for the dehydrationÈdehydrogenation of propan-2-ol. The reactivity of the catalysts was
studied through isotherm and temperatureÈprogrammed surface reactions (TPSR), in a Ñow reactor. In addition, temperatureÈ
programmed oxidation (TPO) and temperatureÈprogrammed desorption (TPD) experiments revealed that catalyst deactivation
was caused by carbon deposition and irreversible adsorption of propene formed on the catalytic surface.
AcidÈbase properties are one of the most important types of
surface chemical properties of metal oxide catalysts.1 MgO,
CaO and BeO are typical solid basic catalysts (in particular,
MgO which is frequently used as a reference material for solid
decomposition is the most widely used to characterize the
acidÈbase properties of metal oxides. Many authors correlate
the total acidity with the dehydration kinetic constant, and
the basicity with the ratio of kinetic constants for dehydroge-
nation and dehydration.13
On the other hand, techniques based on temperature-
programmed experiments have been widely used in the last
few years as powerful tools for the study of any process.
Waugh et al.,14 by using TPSR, have performed kinetic and
mechanistic studies of the propan-2-ol decomposition process
over several metallic oxides. Moreover, acidÈbase properties
of catalytic solids have been widely investigated by using TPD
experiments of probe molecules preadsorbed on the cata-
lyst.15,16
basic catalysts2,3), while ZrO has both acidic and basic
2
properties. This amphotheric character, in addition to its high
thermal stability, makes ZrO a promising catalytic material.4
2
At present, it has been applied as a catalyst to many organic
reactions such as esteriÐcations and alkene hydrogenations.5
The textural and acidÈbase properties of ZrO depend
2
greatly on both the synthetic procedure and the calcination
temperature.6 ModiÐcation of these properties by addition of
alkaline-earth metals7 or sulfate ions,8 as well as by mixing
with other metal oxides9 has also been studied extensively.
The normal procedure used to prepare ZrO consists of cal-
In this work, we studied the textural and acidÈbase proper-
ties of two di†erent zirconia catalysts. The e†ect of the calcin-
ation temperature is discussed. AcidÈbase properties were
studied by employing TPDÈMS of preadsorbed probe mol-
ecules such as pyridine (total acidity), 2,6-dimethylpyridine
(BrÔnsted acidity) and carbon dioxide (basicity). Zirconia
catalysts have been compared with representative basic (MgO)
and acidic (AlPO ÈSiO ) catalysts. The dehydration/
2
cination of its hydroxide (hydroxylated gel) which is prepared
by hydrolysis of zirconium salts. This gel presents two di†er-
ent kinds of OH group, terminal and bridged, the ratio of
which depends mainly on the precursor used to form the gel.
Zirconia is known to exist in monoclinic, tetragonal and
cubic phases, and as an amorphous solid. Of these modiÐ-
cations, the monoclinic form is thermodynamically stable
under ambient conditions,10 while the tetragonal and cubic
phases are metastable. Below 420È450 ¡C the solid obtained is
amorphous. The tetragonal form becomes the stable phase
above about 1100 ¡C; however, the transition to the mono-
clinic phase is so rapid that the tetragonal phase is not
obtained even upon rapid cooling from high temperature.
However, the tetragonal phase may be stabilized by using
additives. Lanthanum or yttrium oxides as well as sulfate ions
are frequently used for this purpose.8,9
4
2
dehydrogenation of propan-2-ol has been used as a test reac-
tion to compare the catalysts. In addition to isotherm
measurements. TPSRs were carried out in order to obtain
information on the solids. Finally, the catalyst deactivation
has been studied using techniques such as TPO and TPD of
strongly adsorbed species.
Experimental
In addition, the various crystal forms of ZrO reÑect how
2
Catalyst synthesis
the hydroxide was prepared and treated. In this sense, a gel
with a high proportion of bridged OH groups gives the
tetragonal form, while if terminal OH groups are predominant
in the gel, the monoclinic form is formed.4 However, ageing
for a long period also generally results in the monoclinic form
after calcination, while the tetragonal phase is dominant when
ageing is omitted. Steaming of the hydroxide results in the
formation of the monoclinic form, while a vacuum treatment
results in the tetragonal form.10 The tetragonal form presents
better textural and acidÈbase properties than the monoclinic
and, therefore, is the one most used in catalysis.11
Zirconium hydroxide was prepared from 30
g
of
ZrOCl É 8H O (Merck Art.8917) and 1.5 l of distilled water.
2
2
Next, NaOH (1 M) was added until a pH of 6.8 was achieved.
The gel was aged for 72 h, and then Ðltered and air-dried for
24 h and in an oven at 110 ¡C overnight.
ZrO was obtained by calcination of the dried gel by
2
heating it in vacuo from room temperature to the Ðnal tem-
perature (300 or 600 ¡C) at a rate of 2 ¡C min~1. The Ðnal
temperature was maintained for 5 h and the solids were then
allowed to cool to room temperature, under a vacuum. The
solids were named ZrO -300VAC and ZrO -600VAC, indicat-
AcidÈbase catalysts are usually characterized by measuring
their activity and selectivity for a test reaction.12 Propan-2-ol
2
2
ing their respective calcination temperatures.
J. Chem. Soc., Faraday T rans., 1997, 93(7), 1431È1438
1431

View Article Online
able size (1È10 ll) of a 10~5È10~6 M amine solution in cyclo-
hexane.
The MgO-based catalyst (MgO-600VAC) was obtained by
calcination of the precursor Mg (OH) (CO ) É 4H O (Merck
5
2
3 4
2
Art. 5827) in a vacuum by heating it at 4 ¡C min~1 from room
temperature to 600 ¡C. The solid was kept at the Ðnal tem-
perature for 2 h and allowed to cool to room temperature,
under a vacuum. The AlPO ÈSiO -based catalyst (PM2) was
Carbon dioxide was the probe molecule used to determine
the basic properties of the catalyst. For the CO TPDÈMS
2
experiments the gases used were CO and 5% CO in argon,
2
2
both of which were supplied by Sociedad Espan
8
ola de Oxi

-
4
2
synthesized from silica gel, AlCl É 6H O and H PO (85%
w/w). The Ðnal composition of the solid was 20 : 80 (w/w)
AlPO : SiO . More details about the synthetic procedure
geno S.A. ([99.999%). Carbon dioxide was quantiÐed by the
base (m/z \ 44) and secondary peaks (m/z \ 12, 10%
abundance). Calibration was performed by injecting pulses of
pure CO or 5% CO in argon. In this way, two calibration
3
2
3
4
4
2
have been given elsewhere.17
2
2
graphs were obtained: one encompassing the range (1È
2
0) ] 10~8 by using pulses of 1È10 ml of 5% CO in argon
Physical characterization
2
and the other spanning (15È40) ] 10~8 with pulses of 0.5È1
ml of pure CO . Correlation coefficients over 0.99 were
obtained. Several replicas of each experiment provided a mean
error of 2% in the peak areas.
The textural properties of the solids (speciÐc surface area, pore
volume and mean radius) were determined from nitrogen
adsorptionÈdesorption isotherms at liquid-nitrogen tem-
perature by using a Micromeritics ASAP-2000 instrument.
Surface areas were calculated by the BET method,18 while
pore distributions were determined by the BJH method.19
Prior to measurements, all samples were degassed at 350 ¡C to
2
Prior to the adsorption of any probe molecule, each catalyst
was cleaned by passing an Ar stream over it at 110 ¡C at 50 ml
min~1 for 30 min. The solids were then saturated by placing
them in an amineÈN or CO ÈAr stream (50 ml min~1) at 25
2
2
0.1 Pa.
or 50 ¡C, respectively. Subsequently, a pure N or Ar stream
2
X-Ray investigation of the solids was carried out using a
(50 ml min~1) was passed over the samples at the saturation
temperature for 2 h in order to remove any physisorbed mol-
ecules. Once a stable baseline was obtained, chemisorbed
amine or CO was desorbed by heating the sample from satu-
ration temperature to 600 ¡C in a programmed fashion. The
selected mass peaks were monitored throughout the process.
Each experiment used ca. 100 mg of fresh catalyst. Full details
of the TPDÈMS method and equipment have been given else-
where.15,16
Siemens D-500 di†ractometer provided with an automatic
control and data acquisition system (DACO-MP). The pat-
terns were run with nickel-Ðltered copper radiation
2
(
j \ 1.5405 Ó) at 35 kV and 20 mA; the di†raction angle, 2h,
was scanned at a rate of 2 ¡C min~1.
FT-Raman spectra were obtained on a Perkin-Elmer 2000
NIR FT-Raman system with a diode pumped Nd YAG laser
(
9394.69 cm~1). It was operated at a resolution of 4 cmv1 over
the range 3600È200 cm~1 to gather 64 scans.
Reactivity tests
DRIFT experiments of preadsorbed probe molecules
A quartz reactor (10 mm id, 200 mm long) was located in the
middle of a furnace. The temperature of the furnace was con-
trolled to within ^1 ¡C. The output of the reactor was Ðtted
on line to a VG Sensorlab mass spectrometer (Fisons Instru-
ment, plc/VG quadrupoles) working in MIM mode. The
peaks monitored and relative abundances from ref. 21 were:
propan-2-ol (m/z \ 45, base peak), acetone (m/z \ 58, 33%),
propene (m/z \ 41, base peak), water (m/z \ 18, base peak),
hydrogen (m/z \ 2, base peak), carbon dioxide (m/z \ 44, base
peak) and diisopropyl ether (m/z \ 87, base peak).
DRIFT experiments were conducted on a Bomen MB-100
instrument with an “Environmental ChamberÏ (Spectra-Tech).
It was operated at a resolution of 8 cm~1 over the range
4
000È400 cm~1 to gather 256 scans. In those experiments
with preadsorbed molecules, each catalyst was cleaned by
passing an Ar stream over it at a rate of 50 ml min~1 at
100 ¡C for 30 min. The solids were saturated with the probe
molecule and subsequently Ñushed with a stream of pure Ar
(
50 ml min~1) at the saturation temperature for 2 h in order
Each reaction was performed using a fresh catalytic bed
to avoid physisorption. The solid was then transferred to the
DRIFT environmental chamber and spectra were obtained at
three or four di†erent temperatures (50, 100, 200 and 300 ¡C).
A temperature levelling time of 20 min was used.
(typically 50 mg of catalyst). Nitrogen (120 ml min~1) was
used as the carrier gas. Propan-2-ol (Aldrich Art. 17,598-7)
was introduced at a constant rate of 12 ml min~1 with a
Braun Perfusor. A temperature levelling time of 20 min was
used. Data points were taken for 150 min. Experiments were
performed to check that internal and external di†usion pheno-
mena were absent.
TPD–MS experiments
TPDÈMS experiments were carried out on a Micromeritics
TPD/TPR 2900 instrument on line with a VG Sensorlab
quadrupole mass spectrometer. The optimum TPD conditions
were as follows: heating rate 10 ¡C min~1 and Ar Ñow rate of
Temperature-programmed experiments
After an isotherm run, TPD and TPO experiments were
carried out in order to detect either adsorbed species or
carbon deposits on the catalytic surface.
5
0 ml min~1. The mass spectrometer, which was operated in
the multiple ion monitoring (MIM) mode, was programmed
to perform six scans per minute.
For the TPD experiments, as soon as the isotherm reaction
Ðnished, the nitrogen Ñow was set to 50 ml min~1 and the
catalyst was immediately cooled to room temperature. The
catalyst was kept at room temperature and under a nitrogen
Ñow for at least 4 h to clean the surface from physisorbed
species. Finally, the temperature was raised at a constant rate
(10 ¡C min~1) from room temperature to 600 ¡C. The whole
process was monitored through the mass spectrometer.
TPO experiments were carried out in a similar way to the
The amines used as probe molecules to determine the acidic
properties of the solids were pyridine (pK \ 5.25) and 2,6-
dimethylpyridine (pK \ 7.3). In a previous DRIFT study of
the bands for the two amines in the region 1400È1700 cm~1,
Marinas and co-workers20 found pyridine to be adsorbed at
BrÔnsted and Lewis acid sites, and 2,6-dimethylpyridine to be
adsorbed on the former site type only owing to the steric hin-
drance of its two methyl substituents. The peaks used to quan-
tify pyridine were the base peak (m/z \ 79) and the secondary
peak at m/z \ 52 (80% relative abundance). The MS peaks
chosen for 2,6-dimethylpyridine were the base peak (m/z
TPD ones, but using a 10% O in Ar Ñow (50 ml min~1)
2
during the temperature ramp.
Finally, the TPSR was carried out in two temperature
stages: an isotherm period (100 ¡C for 30 min) and a second
stage in which the temperature was raised from 100 up to
\
107) and the secondary peak at m/z \ 66 (60%
abundance). Calibration was done by injecting pulses of vari-
1432
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
7
00 ¡C at a constant rate (10 ¡C min~1). The feed consisted of
a nitrogen Ñow (50 ml min~1) passed through a saturator with
propan-2-ol at 30 ¡C. The whole process was monitored
through the mass spectrometer.
Results and Discussion
Physical characterization of the catalyst
Thermal analysis of the Zr(OH) gel showed a continuous loss
2
of water from 100 ¡C. Between 300 and 450 ¡C, a second step
takes place, attributed to the conversion of an amorphous
material to a crystalline phase.10 The overall mass loss was
about 15% of the initial gel.
Fig. 1 shows the XRD patterns of the dried gel (a) as well as
those of both ZrO samples studied in this work [(b) and (c)
2
for the solids ZrO -300VAC and ZrO -600VAC, respectively].
2
2
Fig. 1(a) and (b) are typical of amorphous solids, showing
broad bands. In contrast, Fig. 1(c) presents numerous sharp
Fig. 2 FT-Raman spectra of zirconium dried gel (a), as well as ZrO -
2
3
00VAC (b) and ZrO -600VAC (c)
bands, indicating that ZrO -600VAC is a crystalline solid.
2
2
This pattern, corresponding to ZrO -600VAC, was found to
2
be composed of bands attributable to two di†erent crystalline
phases: a minor tetragonal and a major monoclinic phase
closed and well deÐned hysteresis loops. The isotherms were
type IV in the Brunauer, Deming, Deming and Teller (BDDT)
classiÐcation associated with mesoporous solids.24,25
However, the shape of the hysteresis loops in the isotherms of
Fig. 3, (a) and (b) changes from E- to A-type following the
classiÐcation of de Boer26 (H1 and H2 following IUPAC
(
baddeleyite). As discussed above, the tetragonal phase is
metastable and therefore reverts to the monoclinic form,
which is therefore the dominant phase.
FT-Raman spectra conÐrm the results obtained from XRD.
Raman spectroscopic information is particularly valuable for
distinguishing between the di†erent phases of zirconia. In fact,
the unambiguous assignment of tetragonal and cubic phases is
not possible from XRD alone.22,23 Kilo et al.22 reported the
Raman frequencies observed for the modiÐcations of zirconia.
The most striking di†erence between the Raman spectra of
stable monoclinic and metastable tetragonal zirconia is the
presence of a band at ca. 380 cm~1 in the spectrum of the
former, and a characteristic band at 255È265 cm~1 in the
spectrum of the latter. Fig. 2 presents the Raman spectra cor-
responding to the dried gel (a), as well as those of the catalysts
ZrO -300VAC and ZrO -600VAC [(b) and (c), respectively].
nomenclature). E-type hysteresis (ZrO -300VAC) is often
2
associated with “ink-bottleÏ pores, spheroidal cavities or voids
between close-packed spherical particles. On the other hand,
A-type hysteresis (ZrO -600VAC) is characteristic of cylin-
2
drical pores. It seems, therefore, that the zirconia samples pre-
pared are essentially mesoporous and contain spherical pores
which are gradually smoothed into cylindrical pores upon
calcination6
Surface area, pore volume and mean pore radius data are
compiled in Table 1. The e†ect of the calcination temperature
on surface area is dramatic as reported in ref. 4. The BET
surface area decreases from 220 to 28 m2 g~1 when the calcin-
ation temperature goes from 300 to 600 ¡C. Mercera et al.6
identiÐed two possible processes as being responsible for the
changes occurring in the pore structure and surface area upon
increasing the calcination temperature: crystallite growth and
the accompanying phase transformation on the one hand and
sintering between the crystallites (neck formation and growth)
on the other.
2
2
According to the XRD results, the Raman spectra of the dried
gel and the ZrO -300VAC catalyst correspond to an amor-
2
phous solid. On the other hand, the spectra of ZrO -600VAC
2
show peaks characteristic of the monoclinic phase, while those
typical of the tetragonal phase (e.g. that at 255È265 cm~1)
appear only as small shoulders.
Nitrogen adsorptionÈdesorption isotherms obtained for
both zirconium oxides (ZrO -300VAC and ZrO -600VAC)
2
2
are shown in Fig. 3. Both catalysts present isotherms with
Fig. 3 Nitrogen adsorptionÈdesorption isotherms obtained at liquid-
Fig. 1 XRD patterns of ZrOH gel (a) and the catalysts ZrO -
nitrogen temperature for the catalysts ZrO -300VAC (A) and ZrO -
2
2
2
3
00VAC (b) and ZrO -600VAC (c)
600VAC (B); (…) adsorption band, (=) desorption band
2
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
1433

View Article Online
Table 1 Textural properties of the catalysts: speciÐc surface area (S ), pore volume (V ) and average pore radius (r ) obtained for the solids
BET
p
p
studied
Catalyst
Precursor
S
/m2 g~1
V /ml g~1
r /Ó
BET
p
p
ZrO -300VAC
ZrOCl É 8H O
220
28
69
0.15
0.11
0.22
28
155
118
2
2 2
ZrOCl É 8H O
ZrO -600VAC
2
2
2
MgO-600VAC
Mg (OH) (CO )
3 4
AlCl ] H PO
5
2
PM2
402
3
3
4
Acid–base properties
The e†ect of calcination on the basicity is much more pro-
nounced than that on the surface acidity. CO TPDÈMS pro-
Pyridine, 2,6-dimethylpyridine and carbon dioxide were used
as probe molecules to determine the overall acidity, BrÔnsted
acidity and overall basicity of the catalysts from TPD-MS of
preadsorbed probe molecules. Table 2 presents the results
obtained for the solids used in this work. The solid calcined at
2
Ðles (Fig. 6) indicate that the solid calcined at 300 ¡C (ZrO -
2
3
00VAC) has not only a higher number of basic sites but also
a di†erent distribution of strengths compared with ZrO -
2
6
00VAC. ZrO -300VAC thus presents two types of basic site,
2
the main type being of low strength (desorbing CO in the
range 100È200 ¡C). There is also a sharp desorption peak
300 ¡C presents higher acidity and basicity than that calcined
2
at 600 ¡C. It is clear that calcination temperature is a crucial
factor for obtaining ZrO -based solids with good catalytic
properties. The temperature causes severe loss of surface area,
as seen in Table 1, as well as transformation in the structure of
the solid, as proved by XRD and FT-Raman experiments. The
(
maximum at 463 ¡C) due to a small fraction of basic sites of
high strength present in this solid. However, the solid calcined
2
at 600 ¡C (ZrO -600VAC) has lost the weak basic sites,
2
showing only a slight increase of the baseline in the range
3
50È650 ¡C, exhibiting rather a broad population of basic sites
catalyst ZrO -300VAC is an amorphous, high-surface-area
2
solid that exhibits fairly good acidic and basic surface proper-
ties. On the other hand, the solid calcined at 600 ¡C (ZrO -
2
6
00VAC) presented a low surface area and a monoclinic struc-
ture.
Table 2 also shows, for comparison, a typical basic (MgO-
00VAC) and acidic (PM2) solid. The overall basicity of the
6
MgO-based catalyst is far from those of both ZrO samples.
2
However, the acidity is similar to that of ZrO -600VAC, sig-
2
niÐcantly lower than the acidity of ZrO -300VAC. On the
2
other hand, typical acidic catalysts, such as SiO ÈAlPO
2 4
(
PM2), present higher BrÔnsted, Lewis and overall acidity
than ZrO -based catalysts. The PM2 basicity is so low that it
2
was not possible to determine it by this method. It is clear
that we are dealing with ZrO -based, amphoteric catalysts,
2
with a similar number of basic and acidic sites.
This TPDÈMS method allows us to determine, in addition
to the number of basic and acidic sites, the relative strength of
those sites. Fig. 4È6 show the TPDÈMS proÐles obtained for
pyridine, 2,6-dimethylpyridine (DMPY) and CO preadsorbed
2
on both ZrO -based catalysts. As far as overall acidity (Fig. 4)
2
is concerned, both solids present two types of acid site of dif-
ferent strength, desorbing pyridine at 126 and 161 ¡C. The Ðrst
peak coincides with the peak shown in the DMPY TPDÈMS
proÐle (Fig. 5) and therefore is assigned to BrÔnsted acid sites.
The peak at 161 ¡C does not appear in the DMPY TPDÈMS
proÐle (Fig. 5) and therefore is assigned to Lewis acid sites.
From Fig. 4 and 5 we can conclude that, although the acidity
of ZrO -300VAC is higher than that of ZrO -600VAC, the
Fig. 4 TPDÈMS of preadsorbed pyridine over ZrO -600VAC (a)
2
and ZrO -300VAC (b)
2
2
2
acid strength distribution is the same for both solids, BrÔnsted
acid sites being the main type for both catalysts (see Table 2).
This change, upon calcining, in the number of acid sites but
not in their types, can be assigned mainly to the dramatic
reduction in surface area rather than to morphological
changes.
Table 2 AcidÈbase properties determined by TPDÈMS for the solids
studied
acidity/lmol g~1
BrÔnsted
(DMPY) (PYÈDMPY) (PY)
Lewis
total basicity/lmol g~1
CO
catalyst
2
ZrO -300VAC
46
25
27
73
12
5
4
58
30
31
120
22
436
È
2
ZrO -600VAC
2
MgO-600VAC
Fig. 5 TPDÈMS of preadsorbed 2,6-dimethylpyridine over ZrO -
PM2
42
115
2
6
00VAC (a) and ZrO -300VAC (b)
2
1434
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
Fig. 6 TPDÈMS of preadsorbed CO over ZrO -600VAC (a) and
2
2
Fig. 8 DRIFT spectra of preadsorbed carbon dioxide over ZrO -
ZrO -300VAC (b)
2
2
300VAC, at several temperatures: (a) 50, (b) 100, (c) 200 ¡C
of high strength. This change in the acidÈbase properties
more marked on basic sites than on acid sites) may be related
upon CO adsorption on ZrO based catalysts.27h28 Fig. 8
(
2
2
presents those spectra obtained for the catalyst ZrO -
to a morphological change in the catalyst crystal phase (from
amorphous to monoclinic) rather than to the mere reduction
of the surface area. The nature of the acid sites was the same
in both solids, only their numbers di†ered. On the other hand,
basic sites di†ered not only in number and strength, but also
in nature, between the two solids.
2
3
6
00VAC. DRIFT spectra corresponding to the solid ZrO -
2
00VAC are not shown because of their poor quality owing to
the small number of basic sites on this catalyst. The main
species observed on the surface are unidentate carbonate
(
1
bands at 1304 and 1404 cm~1) and bidentate carbonate (at
632 cm~1), the last being the most abundant surface species.
The DRIFT spectra of pyridine and carbon dioxide
A detailed study of the surface species present after carbon
dioxide adsorption on di†erent types of oxide catalyst has
been published.16
adsorbed on both ZrO -based catalysts at room temperature
2
are drawn in Fig. 7 and 8, respectively. Marinas and co-
workers20 found that pyridine is adsorbed on BrÔnsted and
Lewis acid sites showing bands at around 1450, 1490 and 1590
cm~1. The Ðrst and third ones correspond to the adsorption
on BrÔnsted acid sites, while the second is assigned to adsorp-
tion on Lewis acid sites. Both solids exhibit all three bands
Dehydration–dehydrogenation of propan-2-ol
Initially, to check the activity of our catalysts for propan-2-ol
decomposition over a wide range of temperatures, a TPSR
(
Fig. 7) in agreement with the data from PY and DMPY
TPDÈMS experiments which indicate that the catalysts
present both BrÔnsted and Lewis acid sites. Spectra obtained
at temperatures above ambient also conÐrm that the acid
strength of both catalysts is similar, since IR bands disappear
at about the same temperature for both [Fig. 7(A) and (B)].
was carried out from 100 to 500 ¡C for both ZrO -based cata-
2
lysts (Fig. 9). Both catalysts present no reaction below 300 ¡C.
However, in the range 300È500 ¡C, the behaviour of both
catalysts is di†erent not only in activity, but also in selectivity
and in the shape of the traces. ZrO -300VAC is more active
On the other hand, the DRIFT spectra of preadsorbed CO
allow research on the di†erent types of surface species formed
2
than ZrO -600VAC. Although both are more active in dehy-
2
2
dration than in dehydrogenation, the selectivity for dehydro-
genation is better in the ZrO -600VAC catalyst, especially at
2
high temperatures. Of special interest is the behaviour of the
propene trace at high temperatures. While in the ZrO -
2
3
00VAC TPSR proÐle [Fig. 9(A)] the signal corresponding to
propene (m/z \ 41) increases to reach a constant level at 450È
500 ¡C (about 85% conversion), the propene signal in the
ZrO -600VAC proÐle [Fig. 9(B)] peaks at about 470 ¡C and
2
then decreases. Several authors14,30 attributed this behaviour
to a change in the reaction mechanism, a change that can be
related to a drastic drop in the number of acid sites (mainly of
BrÔnsted type) while the basic sites remain since they are
stronger than the acid ones (see Fig. 4È6). Diisopropyl ether
formation is negligible.
Tables 3 and 4 present the percentages of dehydration and
dehydrogenation, respectively, obtained at temperatures in the
range 300È500 ¡C with both ZrO -based catalysts. Results
2
obtained with solids that are predominantly basic (MgO-
600VAC) or acidic (PM2) are also shown (see Table 2).
As far as dehydration is concerned (Table 3), ZrO -300VAC
2
has proved to be fairly active, especially at high temperatures.
At 500 ¡C, the dehydration achieved by this catalyst equals
Fig. 7 DRIFT spectra of preadsorbed pyridine over ZrO -600VAC
2
(
A) and ZrO -300VAC (B), at several temperatures: (a) ambient, (b)
that of typical acidic catalysts (PM2). The ZrO -600VAC
2
2
1
00, (c) 200, (d) 300 ¡C
catalyst is also quite active at high temperatures, although at
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
1435

View Article Online
However, the relationship between dehydrogenation and
surface basicity is not straightforward. The most active cata-
lyst at 500 ¡C (ZrO -600VAC) is not the most basic one.
2
However, its basic sites are stronger than those of ZrO -
3
2
00VAC. At the reaction temperature, most of the basic sites
of ZrO -600VAC and the stronger of ZrO -300VAC are
2
2
active, but not the weaker basic sites of ZrO -300VAC. The
2
strength distribution of the basic sites plays an important role
in the dehydrogenation of propan-2-ol.
The amphoteric character of ZrO
(and similarly for HfO2
2
and TiO ) allows it to be active in both dehydration and
2
dehydrogenation,31 while the MgO-based catalyst (which is
basic) and PM2 (acidic) are active in dehydrogenation or
dehydration exclusively.
The actual reaction mechanism for the conversion of
propan-2-ol is still not totally clear; however, some of the pro-
posed mechanisms account for both the dehydrogenation and
dehydration processes.32h38
The E1 mechanism (IUPAC designation D ] D 39) is a
N
H
two-step process in which the rate-determining step is ioniza-
tion of the substrate to a carbenium ion (by release of a
leaving group such as OH~) that rapidly loses a b proton.
This mechanism normally operates without a base.
In the E2 mechanism (IUPAC designation A D D 39),
n E N
Fig. 9 TPSR proÐles obtained for ZrO -300VAC (A) and ZrO -
both the leaving group and the proton depart simultaneously,
2
2
6
00VAC (B). Reaction conditions: N Ñow rate, 50 ml min~1; heating
2
the proton being pulled o† by a base. The mechanism thus
takes place in a single step.
rate, 10 ¡C min~1; catalyst mass, 50 mg. (m/z \ 45, propan-2-ol;
m/z \ 58, acetone; m/z \ 41, propene; m/z \ 2, hydrogen; m/z \ 87,
diisopropyl ether; m/z \ 18, water.)
There is a third possibility: the proton leaves Ðrst and fol-
lowed by the leaving group. This is a two-step process called
the E1cB mechanism (A D ] D ). The intermediate is a
n E
N
negatively charged species (the conjugate base of the
substrate). Mechanisms E1 and E1cB take place over an acidÈ
base couple.
500 ¡C the percentage of dehydration achieved is half that of
ZrO -300VAC. MgO-600VAC has a very low dehydration
2
activity over the whole temperature range. All these results
The three above mechanisms show more similarities than
di†erences. In each case there is a leaving group that departs
with its electron pair and another (usually hydrogen) that
comes o† without it. The only di†erence is the order in which
they leave. It is now generally accepted that there is a spec-
trum of mechanisms ranging from one extreme, in which the
leaving group departs well before the proton (pure E1), to the
other extreme, in which the proton comes o† Ðrst and then,
after some time, the leaving group follows (pure E1cB). The
pure E2 mechanisms would be somewhere in between, with
both groups leaving simultaneously. However, most E2 reac-
tions do not lie exactly in the middle of the two extremes.
Propene could be formed from any of the three mecha-
nisms, but acetone only with the E1cB. However, Miyata and
agree with the assumption that surface acidity is responsible
for the dehydrating power.
The activity of the catalysts for the dehydrogenation of
propan-2-ol is lower than that for its dehydration (Table 4).
Furthermore, the trends observed are di†erent from those
described above. The most active catalysts are MgO-600VAC,
over the whole working range, and ZrO -600VAC, at high
2
temperatures. On the other hand, the least active solid is
PM2, which is also the one with lowest surface basicity.
Table 3 Activity in the dehydration of propan-2-ol over the cata-
lysts studied (catalyst mass, 50 mg; reaction time, 60 min; propan-2-
ol Ñow rate, 12 ml h~1; carrier (N ) Ñow rate, 120 ml min~1)
co-workers40 demonstrated from kinetic studies on TiO that
2
2
the probability of propene formation following an E1cB
temperature/¡C
mechanism is limited, owing to the low stability of the car-
banionic species formed during the process.
catalyst
300
350
400
450
500
Catalysts with a large number of acid sites (e.g. PM2) lead
predominantly to dehydration (which takes place via an E1
mechanism). However, dehydrogenation takes place to a
limited extent only, via an E1cB mechanism.
ZrO -300VAC
2.3
1.5
11.3
41.0
6.5
4.1
10.8
50.6
44.9
28.2
8.7
73.6
46.3
7.7
78.1
43.2
7.1
2
ZrO -600VAC
2
MgO-600VAC
PM2
61.8
67.5
76.9
In solids with a large number of basic sites (e.g. MgO-
600VAC), both dehydrogenation and dehydration take place
via an E1cB mechanism, with the former predominating over
the latter.
Table 4 Activity in the dehydrogenation of propan-2-ol over the
catalysts studied (catalyst mass, 50 mg; reaction time, 60 min;
Over solids of similar acidÈbase character like zirconia, the
dehydration may take place via a concerted E2 mechanism.
The simultaneous interaction of an acidÈbase couple with the
wOH group and the proton of the carbon in propan-2-ol
would cause both groups to be released and propene to be
produced. However, dehydrogenation takes place via an E1cB
mechanism. The interaction between a basic site and an
alcohol molecule causes a proton to be abstracted from the
alcoholic group, thus producing an adsorbed alkoxide species.
The release, in a subsequent step, of a hydride from the alkox-
ide carbon atom leads to the formation of acetone.
propan-2-ol Ñow rate, 12 ml h~1; carrier (N ) Ñow rate, 120
2
ml min~1)
temperature/¡C
catalyst
300
350
400
450
500
ZrO -300VAC
0.3
0.9
6.4
1.0
0.7
1.1
7.7
0.6
2.6
4.9
12.3
3.4
8.2
14.0
17.5
4.7
11.6
35.3
22.1
5.9
2
ZrO -600VAC
2
MgO-600VAC
PM2
1436
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
Additional tests involving the dehydration of secondary
alcohols such as hexan-2-ol on zirconia revealed that the
dehydration mechanism coexists with the E1cB mechanism
below 450 ¡C. According to Noller et al.,38 the fact that hex-2-
ene is obtained in its cis and trans forms suggests the
occurrence of the E1cB mechanism. A plot of the hex-2-ene/
hex-1-ene ratio as a function of the reaction temperature
revealed such a ratio to be close to unity below 450 ¡C and to
increase with increasing temperature above that level. The
TPSR proÐles in Fig. 9 conÐrm the presence of both dehydra-
tion mechanisms: the propene signal exhibits a change at
about 470 ¡C that has been related to a change in the mecha-
nism by many authors.
However, isotherm reactions carried out in the range speci-
Ðed above show that there is severe deactivation of the cata-
lyst with time-on-stream. Activity proÐles show a continuous
decay even after 150 min on-stream, especially for reactions
carried out at 400È500 ¡C. Thus, the activity for propan-2-ol
decomposition for the catalyst ZrO -600VAC at 450 ¡C falls
Fig. 11 Propene TPD proÐles obtained for ZrO -600VAC, per-
2
2
from 80 to ca. 50% in 150 min. TPD and TPO experiments
formed after keeping the catalysts under reaction conditions (propan-
2
1
-ol decomposition) for 150 min. Reaction conditions: N Ñow rate,
carried out after isothermal reaction have proved that deacti-
vation is caused by strong retention of surface species
2
20 ml min~1; propan-2-ol Ñow rate, 12 ml h~1; catalyst mass, 50
mg; isotherm reaction temperature: 500 (a) and 300 ¡C (b).
(
propene) as well as carbon deposition.
Fig. 10 shows the TPO proÐles obtained for ZrO -600VAC
2
after propan-2-ol decomposition at 300 and 500 ¡C. The
amount of carbon dioxide found in the 500 ¡C proÐle is four
times higher than that found in the proÐle corresponding to
the reaction at 300 ¡C.
Moreover, during a TPD experiment subsequent to an iso-
thermal reaction, only propene (not acetone or propan-2-ol)
was detected. Fig. 11 presents the propene TPD proÐles
tion temperature is crucial in determining the structural, tex-
tural and surface chemical properties of the ZrO obtained.
2
Calcination at 300 ¡C led to an amorphous solid, while calcin-
ation at 600 ¡C gave a mixture of the tetragonal and mono-
clinic phases. A dramatic reduction in the surface area for
obtained for ZrO -600VAC. In the TPD experiments we
2
ZrO -600VAC was obtained. Calcination at 600 ¡C led to a
noted behaviour opposite to that found in the TPO experi-
2
reduction in the number of acid sites without any change in
ments. As the reaction temperature increases the amount of
propene retained by the catalyst decreases. Since the deactiva-
tion increases with temperature, it seems that deactivation at
high temperatures (about 500 ¡C) is due to carbon deposition
while at low temperatures (around 300 ¡C), both carbon depo-
sition and the presence of strongly adsorbed propene appear
to be responsible.
the strength distribution. However, calcination at 600 ¡C
induced changes in the basic strength of the remaining sites:
there were fewer but stronger sites than in the solid calcined at
300 ¡C. We think that the acidity lost is due to surface area
reduction upon sintering of the crystallites, while phase trans-
formation (from amorphous to crystalline) is responsible for
the change in the strength distribution of the basic sites.
(
b) Although both ZrO -based catalysts are more active for
2
Conclusions
dehydration than for dehydrogenation of propan-2-ol, the
solid obtained at 600 ¡C has higher dehydrogenating power
than that obtained at 300 ¡C. This is related to the basic sites
of high strength that appear on this solid.
The results obtained in this work and the above interpretation
allow us to draw the following conclusions: (a) the calcina-
(c) TPSR proÐles showed a change in propene traces that
was related to a change in the mechanism of dehydration. The
analysis of the strength distribution of the acidic and basic
sites of the catalysts seems to indicate that at low tem-
peratures the main way to obtain propene is through the E1
mechanism (over BrÔnsted acid sites). On the other hand, at
high temperatures, most BrÔnsted acid sites disappear and
propene is obtained through the E1cB or, more likely, the E2
mechanism (over an acidÈbase couple).
(d) Isotherm reaction proÐles showed an initial deactivation
period that depended on the reaction temperature. TPD and
TPO experiments performed after an isothermal run indicated
that deactivation at high temperatures (about 500 ¡C) was
related to carbon deposition, while at low temperatures
(around 300 ¡C), both carbon deposition and strongly
adsorbed propene caused catalyst deactivation.
The authors gratefully acknowledge the funding of this
research by the Consejer•
de Andaluc•a and the Direccio
Cient•Ðca y Tecnica (DGICyT) in the framework of Project
PB92-0816. The sta† at the Mass Spectrometry Service of the
University of Cordoba are also acknowledged for their kind

a de Educacio
 n y Ciencia de la Junta


n General de Investigacio
 n
Fig. 10 TPO proÐles obtained for ZrO -600VAC, performed after
2


keeping the catalyst under reaction conditions (propan-2-ol
decomposition) for 150 min. Reaction conditions: N Ñow rate, 120
2
ml min~1; propan-2-ol Ñow rate, 12 ml h~1; catalyst mass, 50 mg;

isotherm reaction temperature: 300 (a), 400 (b) and 500 ¡C (c).
technical assistance in the experiments.
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
1437

View Article Online
References
21 Eight Peak Index of Mass Spectra, The Royal Society of Chem-
istry, Nottingham, 1983.
2
2
2
2
2
3
4
5
M. Kilo, C. Schild, A. Wokaun and A. Baiker, J. Chem. Soc.,
1
K. Tanabe, M. Misono, Y. Ono and H. Hattori, New Solid Acids
Faraday T rans., 1992, 88, 1453.
and Bases, Kodansha/Elsevier, Tokyo/Amsterdam, 1989.
A. Gervasini and A. Auroux, J. Catal., 1991, 131, 190.
R. Sokoll and H. Hobert, J. Catal., 1992, 134, 409.
T. Yamaguchi, Catal. T oday, 1994, 20, 199.
M. Yashima, K. Ohtake, M. Kakihana, H. Arashi and M. Yoshi-
mura, J. Phys. Chem. Solids, 1996, 57, 17.
S. J. Greg and K. S. W. Sing, Adsorption, Surface Area and Poro-
sity, Academic Press, London, 2nd edn., 1982.
K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A.
Pierotti, J. Rouquerol and T. Siemieniewska, Pure Appl. Chem.,
2
3
4
5
6
K. Tanabe and T. Yamaguchi, Catal. T oday, 1994, 20, 185.
P. D. L. Mercera, J. G. Van Ommen, E. B. M. Doesburg, A. J.
Burggraaf and J. R. H. Ross, Appl. Catal., 1990, 57, 127.
T. Matsuda, Y. Sasaki, H. Miura and K. Sugiyama, Bull. Chem.
Soc. Jpn., 1985, 58, 1041.
1
985, 57, 603.
7
8
9
2
6
J. J. de Boer, in T he Structure and Properties of Porous Materials,
ed. D. H. Everett and F. S. Stone, Butterworths, London, 1965, p.
C. J. Norman, P. A. Goulding and I. McAlpine., Catal. T oday,
6
8.
1
994, 20, 313.
27
28
29
D. Bianchi, J-L. Gass, M. Khalfallah and S. J. Teichner, Appl.
P. Turlier, J. A. Dalmon, G. A. Martin and P. Vergnon, Appl.
Catal., 1993, 101, 297.
Catal., 1987, 29, 305.
D. Bianchi, T. ChaÐk, M. Khalfallah and S. J. Teichner, Appl.
Catal., 1993, 105, 223.
1
1
1
0
1
2
Y. Murase and E. Kato, Y ogyo Kyokaishi, 1978, 86, 226.
M. Shibagaki, Fine Chemicals, 1992, 21, 5.
D. Bianchi, T. ChaÐk, M. Khalfallah and S. J. Teichner, Appl.
Catal., 1994, 112, 219.
M. Guisnet, in Catalysis by Acids and Bases, ed. B. Imelik, C.
Naccache, G. Coudurier, V. Bentaarit and J. C. Vedrine., Stud.
Surf. Sci. Catal., Elsevier, Amsterdam, 1985, vol. 20, p. 283; J. R.
Anderson and M. Boudart, in Catalysis, Science and T echnology,
Springer Verlag, Berlin, 1982, vol. 2, p. 232.
3
3
0
1
M. Bowker, R. W. Petts and K. C. Waugh, J. Catal., 1986, 99, 53.
G. A. M. Hussein, N. Sheppard, M. I. Zaki and R. B. Fahim, J.
Chem. Soc., Faraday T rans., 1989, 85, 1723.
M. Bowker, R. W. Petts and K. C. Waugh, J. Chem. Soc.,
Faraday T rans., 1982, 78, 2573.
E. Akiba, M. Sona, T. Onishi and K. Z. Tamaru, Phys. Chem.,
3
3
3
3
2
3
4
5
1
1
1
3
4
5
H. Niiyama and B. Echigoya, Bull. Chem. Soc. Jpn., 1972, 45, 938;
G. L. Kibby and W. Keithhall, J. Catal., 1973, 31, 65.
K. C. Waugh, M. Bowker, R. W. Petts, H. D. Vandervell and J.
OÏMalley, Appl. Catal., 1986, 25, 121.
1
980, 119, 103.
M. Yamasita, F-Y. Day, M. Suzuki and Y. Saito, Bull. Chem. Soc.
Jpn., 1991, 64, 628.
M. A. Aramend•
Porras and F. J. Urbano, Rapid Commun. Mass. Spectrom., 1994,
, 599.
M. A. Aramend•
a, V. Borau, C. Jimenez, J. M. Marinas, A.
C. Williams, M. A. Makarova, L. V. Malysheva, E. A. Paukshtis
and K. I. Zamarev, J. Chem. Soc., Faraday T rans., 1990, 86, 3473.
H. Grisebach and J. B. Mo†tat, J. Catal., 1983, 80, 350.
G. A. M. Hussein and N. Sheppard, J. Chem. Soc., Faraday
T rans., 1989, 85, 1723.
8
1
6
a, V. Borau, C. Jimenez, F. Lafont, J. M.
3
3
6
7
Marinas, A. Porras and F. J. Urbano, Rapid Commun. Mass.
Spectrom., 1995, 9, 193.
1
1
1
2
7
8
9
0
M. A. Aramend • a, V. Borau, C. Jimenez, J. M. Marinas and F.
3
3
4
8
9
0
H. Noller, J. A. Lercher and H. Vinek, Mater. Chem. Phys., 1988,
Rodero, Colloids Surf., 1984, 12, 227.
S. Brunauer, P. H. Emmett and E. Teller, J. Am. Chem. Soc.,
1
8, 577.
J. March, in Advanced Organic Chemistry. Reactions, Mechanisms,
and Structure, Wiley, New York, 4th edn., 1992, p. 290.
1
939, 60, 309.
E. P. Barret, L. S. Joyner and P. P. Halenda, J. Am. Chem. Soc.,
951, 73, 373.
F. M. Bautista, J. M. Campelo, A. Garc•
Marinas and M. R. Urbano, J. Mater. Chem., 1994, 4, 311.
T. Nakajima, H. Miyata and Y. Kubokawa, Bull. Chem. Soc. Jpn.,
1
1
982, 55, 609.

a, D. Luna, J. M.
Paper 6/06813H; Received 4th October, 1996
1438
J. Chem. Soc., Faraday T rans., 1997, V ol. 93