Catalytic conversion of alcohols over alumina-zirconia mixed oxides: Reactivity and selectivity

Article abstract of DOI:10.1016/j.apcata.2011.07.024

The amorphous and crystalline mixed 25, 50 and 75 wt.% Al₂O ₃-ZrO₂ composites were prepared by means of sol-gel method using aluminum iso-propoxide and zirconyl nitrate precursors. Physicochemical properties were determined using XRD, BET, NH₃-TPD, TGA, SEM, and FT-IR spectroscopy. Catalytic activity and selectivity were investigated for the dehydration of 2-octanol and 1,2-diphenyl-2-propanol (DPP). Al ₇₅Zr₂₅ composite is more selective than other forms for the conversion of 2-octanol. The selectivity of DPP for 1-alkene formation was increased over Al₂₅Zr₇₅ catalyst. The crystalline 50 wt.% catalyst dehydrated tertiary alcohol (DPP) selectively in the presence of 2-octanol. The amorphous 50 wt.% catalyst dehydrated DPP predominately to 1-alkene isomer. The crystalline catalyst favored the formation of E-2-alkene.

Full text of DOI:10.1016/j.apcata.2011.07.024

Applied Catalysis A: General 404 (2011) 141–148
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Catalytic conversion of alcohols over alumina–zirconia mixed oxides: Reactivity
and selectivity
Hossein A. Dabbagh∗, Mehdi Zamani
Catalysis Research Laboratory, Department of Chemistry, Isfahan University of Technology, Isfahan, Iran
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 May 2011
Received in revised form 14 July 2011
Accepted 16 July 2011
Available online 22 July 2011
The amorphous and crystalline mixed 25, 50 and 75 wt.% Al2O3–ZrO2 composites were prepared by means
of sol–gel method using aluminum iso-propoxide and zirconyl nitrate precursors. Physicochemical prop-
erties were determined using XRD, BET, NH3-TPD, TGA, SEM, and FT-IR spectroscopy. Catalytic activity
and selectivity were investigated for the dehydration of 2-octanol and 1,2-diphenyl-2-propanol (DPP).
Al75Zr25 composite is more selective than other forms for the conversion of 2-octanol. The selectivity of
DPP for 1-alkene formation was increased over Al25Zr75 catalyst. The crystalline 50 wt.% catalyst dehy-
drated tertiary alcohol (DPP) selectively in the presence of 2-octanol. The amorphous 50 wt.% catalyst
dehydrated DPP predominately to 1-alkene isomer. The crystalline catalyst favored the formation of
E-2-alkene.
Keywords:
Amorphous
Crystalline
Mixed Al2O3–ZrO2
Catalyst reactivity-selectivity
Dehydration
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
boehmite (AlOOH) and bayerite (Al(OH)3). At 1400 ◦C the t-, m-
zirconia and ␣-alumina were crystallized. Zawadzki et al. [14]
Alumina and zirconia are widely used in many important tech-
nological applications [1–6]. Zirconia exhibits various structural
polymorphs including cubic (c-ZrO2), tetragonal (t-ZrO2) and mon-
oclinic (m-ZrO2) structures which are formed below the melting
point (2680 ◦C), below 2370 ◦C and below 950 ◦C, respectively [7].
Addition of appropriate dopants, such as Ca2+, Mg2+, or Y3+ is known
to stabilize the tetragonal and cubic polymorphs [8–11].
used aluminum iso-propoxide and zirconyl nitrate to prepare
10 wt.% ZrO2 composites. They reported that boehmite crystallized
to ␥-alumina at 600 ◦C while, t-zirconia and ␣-alumina crystal-
lized at 1200 ◦C. Kagawa et al. [15] reported that by increasing
of Al2O3 content in the mixed oxide, the amount of the m-ZrO2
decreased, thus indicating the stabilization of the t-ZrO2 by the
Al2O3.
Preparation of amorphous Al2O3–ZrO2 materials was reported
using sol–gel method [16–19], laser-splatting [16], plasma spraying
[16], electrohydrodynamic atomization [17] and melt extraction
[18]. Soisuwan et al. [19] prepared alumina–zirconia mixed oxides
with 0.5, 1, 25, 40, and 75 mol% of alumina in zirconia using alu-
minum nitrate and zirconyl chloride precursors. In composites
calcined at 600 ◦C with low Al contents (<25%), tetragonal phase zir-
conia was observed, while at higher Al contents, the mixed oxide
exhibited only the amorphous phase. Han et al. [20,21] used the
aluminum iso-propoxide and zirconium butoxide to prepare amor-
phous mixed 50 wt.% alumina/zirconia composite. The calcined
composites powders were amorphous at 400 ◦C. By increasing
the calcination temperature to 980 and 1200 ◦C, the c- and t-
ZrO2 were crystallized, respectively. Beitollahi et al. [22] reported
the preparation of amorphous mixed Al2O3–ZrO2 using aluminum
nitrate and zirconyl chloride at 500 ◦C calcination temperature.
At higher temperatures the various alumina metastable phases
were formed. Most of these works were restricted to the synthe-
sis and morphology of composites. A detailed understanding of
◦
◦
◦
Liquid≈2−6→80 Cc-ZrO2≈2−3→70 Ct-ZrO2
−→
≈950−1150 Cm-ZrO2
Alumina also exhibits a remarkable series of structural poly-
morphs including ␣-Al2O3 and a large variety of metastable forms.
The metastable variants are a series of transition aluminas, which
form by dehydration of aluminum hydroxides and oxyhydroxides
[12].
◦
◦
◦
◦
ꢀ-AlOOH≈−50→0 Cꢀ-Al2O3≈−85→0 Cı-A2O3≈1−0→00 Cꢁ-Al2O3≥1−1→00 C˛-Al2O3
Depending on the method and preparation conditions, differ-
ent phases are generated in mixed Al2O3/ZrO2 oxides [13–19].
Sarkar et al. [13] used the aluminum nitrate and zirconyl chloride
precursors to prepare alumina/zirconia (5–10 wt.% ZrO2) com-
posites. They reported that the catalysts calcined at 200 ◦C were
∗
Corresponding author. Tel.: +98 311 391 3257; fax: +98 311 391 2350.
E-mail address: dabbagh@cc.iut.ac.ir (H.A. Dabbagh).
0926-860X/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2011.07.024

142
H.A. Dabbagh, M. Zamani / Applied Catalysis A: General 404 (2011) 141–148
number range 400–4000 cm−1. Catalysts were diluted in KBr with
the ratio of 1:100 (catalysts: KBr).
their catalytic activity for dehydration reaction of alcohols is still
lacking.
Recently, selectivity and reactivity of alcohols over pure alu-
mina and mixed alumina–thoria was investigated by Dabbagh et al.
[23–25]. We reported a decrease in the reactivity but no change in
the selectivity in the dehydration of secondary and tertiary alcohols
by mixed ␥-alumina–thoria [24]. In this work chemically mixed
amorphous Al2O3–ZrO2 powder (25, 50 and 75 wt.%) were synthe-
sized using the sol–gel method with aluminum iso-propoxide and
zirconyl nitrate. Characterization of the catalysts was performed
by X-ray diffraction (XRD), thermo gravimetric analysis (TGA),
scanning electron microscopy (SEM) and infrared spectroscopy
(IR). Evaluation of catalytic activity and selectivity was investi-
gated for dehydration of 2-octanol and DPP were monitored by gas
chromatography (GC) and gas chromatography mass spectrometry
(GC–MS) techniques.
2.3. Dehydration reaction
The conversion of alcohols was performed in a vertical plug flow
reactor made of Pyrex glass and fitted with a thermal well that
extended to the center of the catalyst bed. The Pyrex glass beads
(20–30 cm3 of the reactor volume above the catalyst bed) were used
to serve as pre-heater. Adducts were collected (with increasing
time intervals) at room temperature. The conversion of 2-octanol
(98 wt.%) and DPP (2 wt.%) mixture at 280 ◦C (reaction temperature)
with flow rate of 18 cm3/h over 2.0 g of catalyst was monitored
by Agilent 6890N gas chromatography using 30 m × 0.32 mm HP-5
column packed with (5% phenyl)-methylpolysiloxane. DPP is solid
which was prepared according to the Grignard method reported
earlier [23]. 2-Octanol and DPP were selected for the following rea-
sons. First, the dehydration of secondary and tertiary alcohols was
investigated in one run under identical conditions. The amounts of
catalyst, reaction temperature, mesh size and flow rate were kept
constant. Second, high concentration of 2-octanol was used as sol-
vent to prevent isomerization. Analysis of the product mixture was
carried out by gas chromatography mass spectrometry Fisons 8060
instrument (QD analyzer and Trio 1000 detector model) equipped
with a 30 m HP-5 capillary column.
2. Experimental
2.1. Catalyst preparation
Aluminum iso-propoxide [Al(OC3H7)3] and zirconyl nitrate
[ZrO(NO3)2·xH2O] were purchased from Aldrich. Five samples of
chemically mixed oxides were prepared with different composition
of aluminum iso-propoxide:zirconyl nitrate wt.% (100:0, 75:25,
50:50, 25:75 and 0:100) using the sol–gel method. The samples are
called Al100Zr0, Al75Zr25, Al50Zr50, Al25Zr75 and Al0Zr100, respec-
tively. Initially, aluminum iso-propoxide and zirconyl nitrate with
different composition (100:0, 75:25, 50:50, 25:75 and 0:100 wt.%)
were mixed well. Then excess amount of water (precursor/water
wt.% ratio = 2 for Al100Zr0 and 4 for Al75Zr25, Al50Zr50, Al25Zr75 and
Al0Zr100) was added to this mixture with continuous stirring for 1 h
at room temperature. Finally the gel solution was heated for 24 h
at 120 ◦C to remove the solvent and to allow maximum interac-
tion of the metal hydroxides. The dried gel was calcined at 600 or
1000 ◦C (heating rate of 2 and 5 ◦C/min, respectively) and remained
at these temperatures for 6 h under air atmosphere. The mechani-
cally mixed Al50Zr50 composite was prepared by physically mixing
and milling of the equal ratios of alumina and zirconia (Al100Zr0
(50%) and Al0Zr100 (50%)) samples. This mixture was prepared to
indicate the physical mixing and should not be compared to the
chemically mixed Al50Zr50 from iso-propoxide and zirconyl nitrate
weight ratios.
3. Results and discussion
3.1. Catalysts characterization
Recently, dehydration of aluminum hydroxide to form pure ␥-
alumina was reported as a polymerization process. The monomer
aluminum hydroxide undergoes polymerization by losing water
molecules with random cross-linkage (in three dimensions)
between the polymer chains [23,24]. The rate of dehydration of
this precursor is too fast and difficult to control. Initially, the
boehmite precipitates and peptidizes to a sol. A transparent gel
was obtained by evaporating the solvent from the sol. The number
of cross-linkages was increased (by loss of more water molecules)
and the number of vacancies decreased leading to a very low
surface area of ␣-alumina (a complete cross-linked polymer) at
higher temperatures (from 200 to 1200 ◦C). The side product of the
hydrolysis of Al(OC3H7)3 is isopropyl alcohol (Eq. (1)). The main
advantage of this precursor is that it does not require any additives.
The co-polymerization process was catalyzed by HNO3 during the
hydrolysis of ZrO(NO3)2·xH2O (Eq. (2)). The amorphous phase was
produced by random cross-linkages during the hydrolysis of the
precursor at 200–600 ◦C. A crystalline phase was produced at higher
temperatures (above 900 ◦C). Fig. 1 shows the photograph of mixed
oxide gels dried at 120 ◦C. The color of Al75Zr25 composite was
orange, while the Al50Zr50 and Al25Zr75 composites were yellow
and creamy, respectively. The color of composites was related to
amounts of nitrous and nitrate compounds formed during reaction
between isopropyl alcohol and HNO3 (Eq. (3)).
Fig. S1 shows SEM pictures of Al75Zr25 and Al50Zr50 compos-
ites dried at 120 ◦C. The large rod (Fig. S1e,g), needle (Fig. S1h) and
bulk (Fig. S1e) shapes were observed for Al50Zr50 composite. The
helix (Fig. S1a,c) and bulk (Fig. S1a,b) structures were existing for
Al75Zr25 composite. The bulk structures of both mixed oxides com-
posed of small agglomerated particles (Fig. S1d,f). The length of
rod shape composites was estimated about 2–3 mm with diame-
ter larger than 120 ␮m. The needle shape composites have nearly
500 ␮m length and their diameters at the end and tip of needle
are almost 6 and 50 ␮m, respectively. The length of helix shape
2.2. Characterization of catalysts
The X-ray diffraction patterns were recorded by employing a
˚
Philips Xpert MPD diffractometer using Cu K␣ radiation (ꢂ = 1.54 A)
in the range of 20–90◦ (2ꢁ) with a step size of 0.03◦ (2ꢁ) and a count-
ing time of 16 s for the samples. The Brunauer–Emmett–Teller (BET)
specific surface areas were measured at the boiling point of nitro-
gen (196 ◦C) after outgassing at 300 ◦C using an ASAP 2000. The
NH3-TPD of the calcined catalysts was assessed through thermal
programmed adsorption/desorption of NH3 in a quartz reactor by
rising the temperature (10 ◦C/min heating rate) from 100 to 800 ◦C
using frontal chromatography technique through Micromeritics
TPD-TPR 2900 system. The TPD of 200 mg for each sample was
performed under helium (gas rate: 40 Ncc/min). Scanning elec-
tron microscopy was performed by Philips XCL microscope. Gold is
vapor-deposited over samples before analysis. The thermal decom-
position behavior of the dried gel was studied by TG–DTA analysis
(Linseis, L70/2171 model) up to 1200 ◦C with heating rate of
10 ◦C/min under atmospheric pressure. The infrared spectra were
obtained using a Jasco-FT-IR-680 plus spectrometer in the wave

H.A. Dabbagh, M. Zamani / Applied Catalysis A: General 404 (2011) 141–148
143
Fig. 1. Morphology of mixed Al2O3–ZrO2 composites with different compositions dried at 120 ◦C (a) Al75Zr25, (b) Al50Zr50 and (c) Al25Zr75
.
composites was estimated several millimeters with diameter larger
The XRD patterns for mixture of m, t-ZrO2 (dashed line) and
than 60 ␮m.
crystalline mixed Al2O3–ZrO2 (50 wt.%) composite calcined at
1000 ◦C (solid line) is depicted in Fig. 4. The diffraction (1 1 1),
(2 2 0), (3 1 1) and (2 2 2) lines of crystalline mixed Al2O3–ZrO2
slightly shifted and drastically broadened from the initial positions
of the pure ZrO2. Also, by doping of ZrO2 with alumina, the m-ZrO2
phase was eliminated and the tetragonal phase was stabilized. This
observation complements the results published by Kagawa et al.
[15]. They reported that the m-ZrO2 phase was decreased with the
increase in alumina content which coincides with stabilization of
the t-ZrO2 phase.
Fig. S2 shows SEM photograph of Al2O3–ZrO2 (50 wt.%) compos-
ites which was formed under different experimental conditions.
The flat plane shape structures (Fig. S2a) was detected in the
chemically mixed Al50Zr50 composite calcined at 600 ◦C, while
mechanically mixed Al50Zr50 composite have porous structure
(Fig. S2b). These results indicated that there are distinct differences
between chemically and mechanically mixed Al2O3–ZrO2 compos-
ites. The composite calcined at 1000 ◦C showed limacone (Fig. S2c)
and agglomerated (Fig. S2d) shapes.
Al(OCH(CH3)2)3 + 2H2O → AlO(OH) + 3(CH3)2CHOH
ZrO(NO3)2 + 2H2O → ZrO(OH)2 + 2HNO3
(1)
(2)
(CH3)2CHOH + 2HNO3 → (CH3)2CHONO
+ (CH3)2CHONO2 + H2O
(3)
Fig.
2 shows the X-ray diffraction patterns of calcinated
Al2O3–ZrO2 mixed oxides (Al100Zr0, Al75Zr25, Al50Zr50, Al25Zr75
,
and Al0Zr100 at 600 ◦C). The XRD pattern of the pure ␥-Al2O3 pre-
dicts four (3 1 1), (4 0 0), (4 4 0) and (5 1 1) diffraction lines at (2ꢁ)
39◦, 46◦, 67◦ and 85◦, which were compatible with the ␥-alumina
spectrum reported in the literature [26]. Pure zirconia was pre-
pared from a mixture of two monoclinic and tetragonal phases.
t-ZrO2 has four well-known diffraction lines at 2ꢁ 30◦ (1 1 1), 50◦
(2 2 0), 60◦ (3 1 1) and 63◦ (2 2 2). The (1 1 0) and (1 1 1) diffrac-
tion lines of m-ZrO2 appeared at 2ꢁ 28◦ and 32◦, respectively. The
related JCPDS files are shown in the supporting information (Figs.
S3–S6). The Al75Zr25, Al50Zr50, and Al25Zr75 composites are approx-
imately amorphous and did not indicate distinct peaks in their
XRD patterns. The XRD patterns of Al50Zr50 composites at various
temperatures are shown in Fig. 3. In the dried gel (at 120 ◦C) and
calcinated sample at 600 ◦C no XRD peaks was detected, while both
tetragonal zirconia and ␣-alumina peaks were detected at 1000 ◦C.
Fig. 3c shows the XRD pattern of mechanically mixed Al50Zr50
composites at 600 ◦C. In this sample all peaks of ␥-alumina and
t, m-zirconia were detected.
The specific BET surface areas for the Al50Zr50 mixed oxides cal-
cined at 600 and 1000 ◦C were 104.2 and 0.3 m2/g, respectively.
The specific BET surface area for the ␥-alumina prepared under the
same condition was reported 150 m2/g [23,25]. The aluminas with
crystalline structure which were synthesized by different methods
showed similar surface properties [27]. The zirconia solids gener-
ally have a low surface area of 20–50 m2/g [20,28].
Figs. S7 and S8 (supporting information) show the NH3-TPD
profile for the calcined Al50Zr50 catalysts at 600 and 1000 ◦C,
Fig. 3. XRD patterns of chemically (a, b, d) and mechanically (c) Al50Zr50 composites
at deferent conditions: (a) dried at 120 ◦C, (b, c) calcined at 600 ◦C and (d) calcined
at 1000 ◦C.
Fig. 2. XRD patterns of mixed Al2O3–ZrO2 composites with different compositions
calcined at 600 ◦C: (a) Al100Zr0, (b) Al75Zr25, (c) Al50Zr50, (d) Al25Zr75 and (e) Al0Zr100
.

144
H.A. Dabbagh, M. Zamani / Applied Catalysis A: General 404 (2011) 141–148
Fig. 4. Portion of the XRD patterns of pure ZrO2 (blue dashed line, —) and crystalline mixed Al2O3–ZrO2 (50 wt.%) oxides (red solid line, —) calcined at 1000 ◦C. (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
respectively. The profile of amorphous catalyst indicates two
strong peaks at 263 and 630 ◦C corresponding to desorbed 3.41
and 6.15 mmol NH3/g, respectively. Crystalline catalyst did not
show a significant peak at 263 ◦C, but a weak peak at 664 ◦C corre-
sponding to desorbed 0.08 mmol NH3/g. Unfortunately, calculation
of the number of acid sites was not possible with the instrument in
use. These desorbed values correspond to both acid sites and the
trapped ammonia inside the bulk. The NH3-TPD data, however,
complements the BET finding.
The DTA–TG curve of Al50Zr50 composite dried gel is shown
in Fig. 5. A weigh loss about 42% was observed from room tem-
perature to about 500 ◦C. The initial endothermic peak at range
60–150 ◦C was attributed to the removal of physically and chem-
ically adsorbed water present in the pores of the gel. The second
endothermic peak at 200–320 ◦C corresponds to the dehydration
of the aluminum hydroxides and oxyhydroxides. This temperature
range encompasses important thermal processes such as transi-
tion alumina creation which starts below 300 ◦C [25]. Removal of
the doped nitrate ions from pores and vacancies occurred at higher
temperatures (320–500 ◦C). The exothermic peak at around 920 ◦C
was related to the crystallization of ␣-Al2O3 and t-ZrO2. This feature
was shown clearly by the XRD measurements (Fig. 3).
Fig. 6 shows the FT-IR spectra of the catalysts calcined at dif-
ferent temperatures (120, 600 and 1000 ◦C). The stretching Al–O
vibrational modes of alumina gel dried at 120 ◦C are responsible
for the appearance of four peaks at 475, 619, 735 and 1072 cm−1
.
It is known that the band characterized of boehmite appear at
1072 cm−1, therefore the disappearance of this peak in the sample
calcinated at 600 ◦C indicates the formation of ␥-alumina phase.
The stretching OH modes of boehmite were observed at 3096 and
3420. These peaks were shifted to 3445 cm−1 for the calcinated
samples. The bands at 420, 574 and 747 cm−1 were assigned to the
stretching vibrations of Zr–O in monoclinic phase. The correspond-
ing value for tetragonal phase appeared at 497 cm−1. The Al–OH
and Zr–OH bending vibrational modes of Al75Zr25, Al50Zr50 and
Al25Zr75 catalysts were observed at 1637 (1635 or 1636 for pure
alumina) and 1384 cm−1 (was observed for pure alumina and zirco-
nia at lower temperatures), respectively. These results are in good
agreement with the reported values for alumina [13,24,25,29–31]
and zirconia [13,31,32] materials.
The FT-IR spectra of composites are different from those of pure
materials. The mixed oxide gel dried at 120 and 600 ◦C exhib-
ited a broad strong stretching OH vibration at 2500–3700 cm−1
,
while the mixed oxide calcined at 1000 ◦C showed strong OH
vibrations. In dried gel of composites (at 120 ◦C) the stretching
nitrate peaks appeared at 1541 and 1384 cm−1. The Al–O and Zr–O
vibrations in the dried gel and calcined composites at 600 ◦C are
inconclusive (because of the amorphous nature of composites with
overlapped modes). The sharp signal was observed at 1000 cm−1
for the calcined (at 1000 ◦C) Al50Zr50 composites, because of the
samples crystallinity. The Al–OH bending vibrations appeared at
1638 and 1385 cm−1 which are in agreement with Sarkar et al. work
[13]. The increment of intensity at 1637 cm−1 and elimination of
1385 cm−1 mode for crystalline catalyst in comparison to amor-
phous composite are attributed to appearance of t-ZrO2 phase. The
amount of free OH groups on the surface of ␣-Al2O3 was reported
very low [12] this complements the disappearance of 1384 cm−1
mode in the crystalline catalyst. Examination of the OH stretch-
ing region should be taken with caution since the use of ex situ
IR with KBr pellets prevents a thorough discussion of the inten-
sity and peak position in the OH region. Bands assigned to Al–OH
and Zr–OH bending modes may well be alternatively assigned to
water and nitrate ions. However, the relative comparison of the
FT-IR spectra shed some light to the complex structure of these
oxides.
Fig. 5. TG (top) and DTA (bottom) curves of Al50Zr50 composite dried gel.

H.A. Dabbagh, M. Zamani / Applied Catalysis A: General 404 (2011) 141–148
145
Fig. 6. FT-IR spectrum of mixed Al2O3–ZrO2 composites with different compositions dried/calcined at 120, 600 and 1000 ◦C: (a) Al100Zr0, (b) Al0Zr100, (c) Al75Zr25, (d) Al50Zr50
and (e) Al25Zr75
.
3.2. Dehydration reaction
dynamically controlled alkenes. For E2 mechanism, reaction occurs
on dual acid–base sites to eliminate a proton and hydroxyl group
which produces the kinetically controlled alkenes, whereas for
E1cb mechanism, strong basic sites are required in order to ini-
tially remove a ␤-hydrogen and then eliminate hydroxyl group to
produced 1-alkene. The selectivity of secondary and tertiary alco-
hols dehydration varies depending on which mechanism is acting.
Dabbagh et al. [24,43] used both experimental including kinetic
and isotopic effect studies and computation analysis to show that
dehydration of 2-butanol and 1,2-diphenyl-2-propanol (DPP) over
␥-alumina exclusively undergoes E2 elimination.
It is known that the dehydration of alcohols to alkenes over ␥-
alumina requires temperatures at 100–320 ◦C, depending on the
type of alcohol and catalyst used [44]. Based on the relative thermo-
dynamic stability, any other product formed on the catalyst surface
would be expected to immediately dissociate at higher tempera-
tures to give alkene and water [23].
Zirconia-based catalysts are known to possess a high selectivity
toward the formation of 1-alkenes in the dehydration of secondary
alcohols [33–38]. Alumina is an excellent catalyst for the selective
dehydration of alcohols [23–25,39–42]. The dehydration studies on
secondary alcohols have always been found to be useful in investi-
gating the properties of dehydration catalysts as well as the modes
of elimination to yield the various olefins. During the dehydration
of alcohol, both intramolecular and intermolecular dehydration
may occur. The relative amounts of these pathways depend on
the reaction conditions as well as the reactant and catalyst used.
The first step of alcohol conversion on metal oxide catalysts is the
adsorption of the reactant on the surface. The adsorbed alcohols
can be converted through a two-step mechanism involving inter-
mediate carbocation formation (E1), a concerted pathway (E2) or
via carbanion formation mechanism (E1cb) [35]. E1 mechanism
needs strong acidic catalyst to form carbenium ions by abstraction
of OH group. The carbenium ion is rearranged via isomerization
followed by abstraction of a ␤-hydrogen resulting in the thermo-
Generally zirconia is known as a basic catalyst. One of the main
objectives of this work was to modify the reactivity and selectiv-
ity of alumina by adding zirconia. Here, the catalytic activity of

146
H.A. Dabbagh, M. Zamani / Applied Catalysis A: General 404 (2011) 141–148
Table 1
Table 3
Conversion (%) of 2-octanol and DPP over pure and mixed Al2O3–ZrO2 composites
Products distribution of dehydration reaction of DPP over pure and mixed
at 280 ◦C.
Al2O3–ZrO2 composites at 280 ◦C.
Catalyst
2-Octanol%
DPP%
(DPP%)/(2-octanol%)
Catalyst
1-Alkene
E-2-alkene
Z-2-alkene
1-Ene/2-ene
Z/E
a
a
a
b
a
a
a
a
a
b
a
a
Al100Zr0
Al75Zr25
Al50Zr50
Al50Zr50
Al25Zr75
Al0Zr100
44
73
25
1.2
18
13
98
58
88
87
74
98
2.23
0.79
3.52
72.5
4.11
7.54
Al100Zr0
Al75Zr25
Al50Zr50
Al50Zr50
Al25Zr75
Al0Zr100
50
55
55
28
65
54
28.5
11
33
70
24
21.5
34
12
2
11
13
1.00
1.22
1.22
0.39
1.86
1.17
0.75
3.09
0.36
0.03
0.45
0.35
33
a
a
Calcined at 600 ◦C.
Calcined at 1000 ◦C.
Calcined at 600 ◦C.
Calcined at 1000 ◦C.
b
b
Al2O3–ZrO2 composites was evaluated for the dehydration of a
mixture of 2-octanol (98 wt.%) and DPP (2 wt.%) using a continu-
ous flow, fixed bed, micro reactor at 280 ◦C. 2-Octanol (secondary
alcohol) and DPP (tertiary alcohol) are very useful models for
investigating the catalyst behavior (reactivity vs selectivity). The
low concentration of DPP (with higher boiling temperature) pre-
vents saturation over catalyst surface. Results are summarized in
Tables 1–3 and compared with those of pure oxides. The conver-
sion of 2-octanol and DPP over pure ␥-alumina (Al100Zr0) was 44%
and 98% and over pure zirconia (Al0Zr100) was 13% and 98%, respec-
tively. The high conversion of DPP over both ␥-alumina and zirconia
must be a result of stronger adsorption of DPP over the surface in
comparison to 2-octanol which is due to the higher electron den-
sity of phenyl groups in comparison to the alkyl chain. This feature
was recently investigated theoretically over ␥-alumina by Dabbagh
Cis-2-octene is the major products of 2-octanol dehydration
over all catalysts (Table 2). The maxima amount of this prod-
uct (49%) was obtained over the Al75Zr25 composite. The minima
amount of cis-2-octene (35%) was obtained over Al50Zr50 com-
posite. The second major product of 2-octanol conversion was
1-octene. The maxima amount of this product (40%) was observed
over Al25Zr75 composite and minima amount (26%) was obtained
over alumina. The 1-ene/2-ene selectivity for the reaction of 2-
octanol over pure ZrO2 prepared under acidic condition was 0.47
(Table 2). This value for zirconia prepared under basic condition
was 4.24 (1-octene 80.5%, trans-2-octene 11.5%, and cis-2-octene
7.5%) which is recently reported by Davis and Chokkaram [37].
Apparently, the basic sites of zirconia are protonated under acidic
condition which promote the formation of 2-ene. Stronger basic
condition or basic sites are required for the formation of 1-ene.
The difference between this work and Davis and Chokkaram results
are reasonable and complement to the Ferino et al. [35,38] report
which examine the effect of preparation conditions of ZrO2 cata-
lysts for the dehydration reaction of 4-methylpentane-2-ol. They
reported that 2-ene is a major product over ZrO2 prepared using
zirconyl nitrate. The best selectivity for 1-ene was obtained when
the precursors were immersed in NaOH. Apparently, the basic sites
of zirconia are protonated under acidic condition (this work) which
promote the E2 elimination. Stronger basic condition or basic sites
are required for E1cb mechanism [35–38].
For DPP dehydration (Scheme 1, Table 3), the major product in all
catalysts (>50%) is 1-alkene; and the best selectivity was observed
for Al25Zr75 composite. For Al75Zr25 composite, the second major
product was Z-2-alkene. These results predict high Z/E selectivity
for Al75Zr25 composite.
Conversion percentage of 2-octanol and DPP over amorphous
(calcined at 600 ◦C) and crystalline (calcined at 1000 ◦C) Al50Zr50
composite as a function of reaction time (60 min) is presented
in Figs. 7 and 8. The amorphous catalyst converted 25% of
2-octanol to a mixture of cis-2-octene (35%), 1-octene (34%),
trans-2-octene (24%), dioctyl ethers (1.3%), isomerization product
(3-octenes, 1.3%), dehydrogenation product (2-octanone, 2.7%) and
1.7% unknown side products. The crystalline catalyst converted
1.20% of 2-octanol to a mixture of cis-2-octene (28%), 1-octene
(26%), trans-2-octene (17%), 3-octenes (0%), dioctyl ether (0%) and
unknown (29%) side products (Tables 1 and 2). These results indi-
et al. [25]. The reactivity and selectivity of amorphous Al75Zr25
,
Al50Zr50, and Al25Zr75 composites do not resemble the pure alu-
mina or zirconia. This aspect was previously investigated by XRD
and SEM analysis (Section 3.1). Conversion of 73%, 25%, 18% (for 2-
octanol) and 58%, 88%, 74% (for DPP) were obtained over Al75Zr25
,
Al50Zr50, and Al25Zr75 composites, respectively. These results indi-
cated that the Al75Zr25 composite is more selective than other
composites for the conversion of secondary alcohol (2-octanol).
Interestingly, the selectivity of DPP for 1-alkene formation was
increased over Al25Zr75 catalyst (Scheme 1, Table 3). These find-
ing predicts that 25% content (alumina or zirconia) has pronounce
effect on reactivity and selectivity. Reactivity is decreased with
increase in % Zr (Table 1). For example, mixed Al75Zr25 converted
73% of 2-octanol and 58% of DPP but mixed Al25Zr75 converted
18% of 2-octanol and 74% of DPP. Cis/trans ratio is increased by 3-
fold depending upon % zirconia. 1-alkene/E + Z-2-alkenes ratio was
increased for 75% zirconia. What causes these changes is the subject
of future investigation.
Generally, major products of dehydration of alcohols over alu-
mina and zirconia are alkenes, but isomerization, dehydrogenation,
and ether formation as by products are increased with increasing
branch of the alcohol hydrocarbon chain or at higher conversion.
Therefore, in this study, 2-octanol was used as a solvent and an iso-
merization preventer. Moreover, the product selectivity (%) of the
above pathways depends on the type and preparation method of
the catalyst as well as the reaction conditions.
Table 2
Products distribution of dehydration reaction of 2-octanol over pure and mixed Al2O3–ZrO2 composites at 280 ◦C.
Catalyst
1-Octene
Trans-2-octene
Cis-2-octene
3-Octene
2-Octanone
Dioctylether
Others
1-Ene/2-ene
Cis/trans
a
a
a
b
a
Al100Zr0
Al75Zr25
Al50Zr50
Al50Zr50
Al25Zr75
Al0Zr100
26
36
34
26
40
28
23
14
24
17
11
21
46
49
35
28
46
38
4
0
1.3
0
0
0
0
2.7
0
0
0.4
0.5
1.3
0
1.1
1.4
0.6
0.5
1.7
29
2.1
2.3
0.38
0.57
0.58
0.58
0.70
0.47
2.00
3.50
1.46
1.65
4.18
1.61
a
5.6
3.7
a
Calcined at 600 ◦C.
Calcined at 1000 ◦C.
b

H.A. Dabbagh, M. Zamani / Applied Catalysis A: General 404 (2011) 141–148
147
cated moderate activity for 2-octanol conversion over amorphous
mixed oxide catalyst with interesting product selectivity (equal
distribution of cis-2-octene and 1-octene) and 7% side products.
Conversion of 2-octanol over crystalline phase was very low pro-
ducing similar distribution of octenes as amorphous catalyst and
29% side products under the same conditions. The conversion of
2-octanol and DPP over pure ␥-alumina which was prepared at
600 ◦C was 44% and 98% and over pure zirconia (mixture of t- and m-
phases) was 13% and 98%, respectively. The product selectivity for
the reaction of 2-octanol over pure ␥-alumina was 1-alkene (26%),
trans-2-octene (23%), cis-2-octene (46%), 3-octenes (4%), dioctyl
ethers (0.4%), 2-octanone (0%) and side products (0.6%) and for the
reaction of 2-octanol over ZrO2 was 1-octene (28%), trans-2-octene
(21%), cis-2-octene (38%), 3-octenes (5.6%), dioctyl ethers (1.4%),
2-octanone (3.7%) and side products (2.3%).
Ph
Ph
Hb
Ha
HO
Al2O3-ZrO2 = M
CH3
Ph
Ph
Ph
Hb
Ph
Ha
Ha
H
H
Hb
Ph
CH2Ph
MO
M
MO
O
CH3
CH3
H
DPP was converted (88%) over amorphous Al50Zr50 catalyst to a
mixture of 2-alkene (33% E-alkene and 12% Z-alkene) and 1-alkene
(55%). The crystalline catalyst converted 87% of DPP to 1-alkene
(28%), E-2-alkene (70%) and Z-2-alkene (2%). The conversion of DPP
over both catalysts was very high producing mostly 1-alkene over
amorphous catalyst and E-alkene (which increased at the longer
time on stream) over crystalline form (Figs. 7 and 8). The Z-2-
alkene (%) approached zero at the final times on stream. The product
distribution (%) of the reaction of DPP was 1-alkene (54%), E-2-
alkene (33%) and Z-2-alkene (13%) over pure ZrO2. This selectivity
for the DPP conversion over pure ␥-alumina was 1-alkene (50%),
E-2-alkene (28.5%) and Z-2-alkene (21.5%), Table 3.
H
H
CH2Ph
Ph
Ph
H
CH3
Ph
E-Alkene
H
CH3
Ph
Z-Alkene
Ph
1-Alkene
Scheme 1. Schematic representation of DPP dehydration reaction.
Pure ZrO2 (mixture of m and t-ZrO2 phases) converted 13% 2-
octanol and 98% DPP, while crystalline 50:50 mixed oxide (mixture
of t-ZrO2 and ␣-alumina) converted 1.2% 2-octanol and 87% DPP.
␣-Al2O3 is an inactive catalytic toward the dehydration of alco-
hols. This indicates that zirconia (of the mixed oxide prepared
at 1000 ◦C) is active catalyst for the dehydration of alcohols. The
low reactivity of 2-octanol over the crystalline Al50Zr50 catalyst
is consistent with low surface area (there are limited active sites,
pores and crevices to undergo concerted elimination to kinetic
controlled products). The small amount of kinetically controlled
octenes was produced over this surface. Adsorption and isomer-
ization of these alkenes over the surface also was ruled out. At
very low conversion of 2-octanol and relatively low tempera-
ture, the rate of 2-octanol adsorption over the surface is much
faster than the adsorption of alkenes. Generally, isomerization
of alkene over the catalyst surface produces the thermodynami-
cally controlled products which were not observed in this work.
In the case of 2-octanol conversion over the crystalline catalyst,
both E1 and E2 mechanisms are competing. The E1 mechanism
is responsible for the formation of the large amount of secondary
products (29%).
The high conversion and unusual product selectivity of DPP
over both amorphous and crystalline Al50Zr50 catalyst were unpre-
dictable. The high reactivity of DPP over both catalysts must be a
result of stronger adsorption of DPP than 2-octanol over the sur-
face which is due to the higher electron density of phenyl groups
in comparison to the alkyl chain (minimum surface is required for
high conversion of DPP). The unusual selectivity must stem from the
structural feature of transition state for each isomer on the surface
which favors the formation of 1-alkene (55%) over the amorphous
catalyst and E-2-alkene (70%) over the crystalline form. These two
pathways must have the most feasible conformation and geome-
try in the transition state with the lowest activation energy. The
thermodynamically controlled mixture of alkenes was expected to
form over the low surface area catalyst. The ZrO2 tetragonal phase
favors the formation of 1-alkene; and ␣-Al2O3 is an inactive cat-
alytic toward the dehydration of alcohols. This indicates that in the
crystalline mixed Al2O3/ZrO2, alumina is apparently poisoned by
the catalytic activity of tetragonal zirconia. In other words, doping
Fig. 7. % Conversion of 2-octanol ( crystalline;
amorphous) and DPP ( crys-
talline; amorphous) over Al50Zr50 catalysts at 280 ◦C (reaction temperature).
80
1-alkene
E-2-alkene
Z-2-alkene
70
60
50
40
30
20
10
0
10
20
30
40
50
60
Time (min)
Fig. 8. Products distribution (% E-2-alkene ( ), Z-2-alkene ( ) and 1-alkene (
)) for the conversion of DPP over amorphous (solid line, —) and crystalline (dotted
line, . . .) Al50Zr50 catalysts.

148
H.A. Dabbagh, M. Zamani / Applied Catalysis A: General 404 (2011) 141–148
of t-ZrO2 with aluminum atoms generates a catalyst with specific
activity and selectivity.
References
[1] J.S. Valente, X. Bokhimi, F. Hernandez, Langmuir 19 (2003) 3583–3588.
[2] D. Susnik, J. Holc, M. Hrovat, S. Zupancic, J. Mater. Sci. Lett. 16 (1997) 1118–1120.
[3] Q.H. Zhang, Y.Q. Feng, S.L. Da, Anal. Sci. 15 (1999) 767–772.
[4] M. Sugiura, Catal. Surv. Asia 7 (2003) 77–87.
4. Conclusions
[5] M.Y. He, J.G. Ekerdt, J. Catal. 87 (1984) 238–254.
[6] M.D. Rhodes, K.A. Pokrovski, A.T. Bell, J. Catal. 233 (2005) 210–220.
[7] M. Yoshimura, Ceram. Bull. 67 (1988) 1950–1955.
[8] E.V. Stefanovich, A.L. Shluger, C.R.A. Catlow, Phys. Rev.
11560–11571.
[9] G. Stapper, M. Bernasconi, N. Nicoloso, M. Parrinello, Phys. Rev. B 59 (1999)
797–810.
[10] J.H. Bitter, K. Seshan, J.A. Lercher, J. Catal. 183 (1999) 336–343.
[11] H.G. Scott, J. Mater. Sci. 10 (1975) 1527–1535.
[12] R.T. Yang, Adsorbents: Fundamentals and Applications, John Wiley & Sons Inc.,
2003, pp. 146–150.
[13] D. Sarkar, D. Mohapatra, S. Ray, S. Bhattacharyya, S. Adak, N. Mitra, Ceram. Int.
33 (2007) 1275–1282.
[14] M. Zawadzki, D. Hreniak, J. Wrzyszcz, W. Mista, H. Grabowska, O.L. Malta, W.
Strezka, Chem. Phys. 291 (2003) 275–285.
[15] M. Kagawa, M. Kikuchi, Y. Syono, T. Nagae, J. Am. Ceram. Soc. 66 (1983) 751–754.
[16] G. Kalonji, J. McKittrick, L.W. Hobbs, in: N. Claussen, M. Ruhle, A.H. Heuer
(Eds.), Advances in Ceramics, vol. 12, American Ceramic Society, USA, 1984,
pp. 816–825.
[17] V. Jayaram, T. Whitney, C.G. Levi, R. Mehrabian, Mater. Sci. Eng. A 124 (1990)
65–81.
[18] J. McKittrick, G. Kalonji, T. Ando, J. Non-Cryst. Sol. 94 (1987) 163–171.
[19] S. Soisuwan, J. Panpranot, D.L. Trimm, P. Praserthdam, Appl. Catal. A: Gen. 303
(2006) 268–272.
[20] J.K. Han, F. Saito, B.T. Lee, Mater. Lett. 58 (2004) 2181–2185.
[21] B.T. Lee, J.K. Han, F. Saito, Mater. Lett. 59 (2005) 355–360.
[22] A. Beitollahi, H. Hosseini-Bay, H. Sarpoolaki, J. Mater. Sci. Mater. Electron. 21
(2010) 130–136.
[23] H.A. Dabbagh, J. Mohammad Salehi, J. Org. Chem. 63 (1998) 7619–7627.
[24] H.A. Dabbagh, M.S. Yalfani, B.H. Davis, J. Mol. Catal. A: Chem. 238 (2005) 72–77.
[25] H.A. Dabbagh, K. Taban, M. Zamani, J. Mol. Catal. A: Chem. 326 (2010) 55–68.
[26] P. Souza Santos, H. Souza Santos, S.P. Toledo, Mater. Res. 3 (2000) 104–114.
[27] K. Jiratova, L. Beranek, Appl. Catal. 2 (1982) 125–138.
[28] H.J.M. Bosman, E.C. Kruissink, J. von der Spoel, F. von den Brink, J. Catal. 148
(1994) 660–672.
[29] M.M. Amini, M. Mirzaee, J. Sol-Gel. Sci. Technol. 36 (2005) 19–23.
[30] L. Le Bihan, F. Dumeignil, E. Payen, J. Grimblot, J. Sol-Gel. Sci. Technol. 24 (2002)
113–120.
[31] D. Sarkar, D. Mohapatra, S. Ray, S. Bhattacharyya, S. Adak, N. Mitra, J. Mater. Sci.
42 (2007) 1847–1855.
[32] S. Roy, J. Sol-Gel. Sci. Technol. 44 (2007) 227–233.
[33] M. Araki, K. Takahashi, T. Hibi, Sumitomo Chemical Co., Eur. Pat. Appl. (1985)
0150832.
Mixed Al2O3–ZrO2 composite with different compositions were
prepared by sol–gel method and characterized by XRD, SEM, TGA
and FT-IR. Chemically mixed oxide composites calcined at 600 ◦C
are amorphous with no distinct XRD pattern. At 1000 ◦C the amor-
phous phase of mixed oxide was transformed to a crystalline
phase which consists of t-ZrO2 and trace of ␣-Al2O3. This feature
was observed by TGA and XRD pattern of Al50Zr50. Mechanically
mixed Al50Zr50 composite have XRD pattern similar to those of
pure ␥-alumina and t, m-zirconia. The specific BET surface area for
the Al50Zr50 after calcination at 600 and 1000 ◦C was 104.2 and
0.3 m2/g, respectively. The NH3-TPD data complements the BET
finding.
Evaluation of catalytic activity and selectivity for dehydration
of mixture of 2-octanol and DPP was investigated. Reactivity of
2-octanol was decreased by increasing the zirconium content.
This trend was not observed for DPP dehydration. Selectivity for
the formation of 1-octene and cis-2-octene was increased for
Al25Zr75 composite. Octenes isomerization and dehydrogenation
was observed only for the Al50Zr50 composite and pure zirconia.
Minimum ether was formed for the reaction of 2-octanol over
Al75Zr25 composite. The formation of 1-alkene was increased at
the expense of Z-2-alkene for the dehydration of DPP over Al25Zr75
composite.
B 49 (1994)
The amorphous and crystalline Al50Zr50 catalysts showed mod-
erate and very low reactivity for the conversion of 2-octanol
respectively, and high conversion for DPP. The crystalline form
selectively dehydrated the tertiary alcohol (DPP) in the pres-
ence of secondary alcohol (2-octanol) producing E-alkene. The
amorphous mixed oxide produced the least stable kinetically
controlled 1-alkene. This feature of new crystalline Al2O3-ZrO2
(50 wt.%) composite has not been reported elsewhere up to this
date.
[34] M. Araki, T. Hibi, Sumitomo Chemical Co., Eur. Pat. Appl. (1986) 0222356.
[35] I. Ferino, M.F. Casula, A. Corrias, M.G. Cutrufello, R. Monaci, G. Paschina, Phys.
Chem. Chem. Phys. 2 (2000) 1847–1854.
Acknowledgements
[36] K. Tanabe, T. Yamaguchi, Catal. Today 20 (1994) 185–198.
[37] S. Chokkaram, B.H. Davis, J. Mol. Catal. A: Chem. 118 (1997) 89–99.
[38] I. Ferino, A. Auroux, P. Artizzu, J. Chem. Soc. Faraday Trans. 91 (1995)
3263–3267.
[39] B. Shi, H.A. Dabbagh, B.H. Davis, J. Mol. Catal. A: Chem. 141 (1999) 257–262.
[40] H.A. Dabbagh, C.G. Hughes, B.H. Davis, J. Catal. 133 (1992) 445–460.
[41] H. Knozinger, H. Buhl, K. Kochloefl, J. Catal. 24 (1972) 57–68.
[42] V. Macho, M. Kralik, E. Jurecekova, J. Hudec, L. Jurecek, Appl. Catal. A: Gen. 214
(2001) 251–257.
We would like to thank Isfahan University of Technol-
ogy (IUT) research council for the financial support (Grant
# 87/500/9143). We also thank Dr. K. Ghani for his helpful
assistant.
Appendix A. Supplementary data
[43] H.A. Dabbagh, M. Zamani, B.H. Davis, J. Mol. Catal. A: Chem. 333 (2010) 54–68.
[44] M.L.G. Franco, S.R. Andrade, R.G. Alamilla, G.S. Robles, J.M.D. Esquivel, Catal.
Today 65 (2001) 137–141.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.apcata.2011.07.024.