Effects of vacuum and calcination temperature on the structure, texture, reactivity, and selectivity of alumina: Experimental and DFT studies

Article abstract of DOI:10.1016/j.molcata.2010.04.007

The effects of vacuum and/or calcination temperature on the structure and texture of alumina were investigated using XRD, SEM, BET, DTG and FT-IR. Elucidation of the observed reactivity and selectivitythe catalyst was carried out via elimination reaction of secondary (2-octanol) and tertiary (1,2-diphenyl-2-propanol, DPP) alcohols. The mechanism of (R)-, (S)-2-octanol and (R)-, (S)-DPP dehydration(1 0 0) and (1 1 0) surfaces of alumina was modeled using density functional theory (DFT).

Full text of DOI:10.1016/j.molcata.2010.04.007

Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Effects of vacuum and calcination temperature on the structure, texture,
reactivity, and selectivity of alumina: Experimental and DFT studies
Hossein A. Dabbagh∗, Keivan Taban, Mehdi Zamani
Catalysis Research Laboratory, Department of Chemistry, Isfahan University of Technology, Isfahan 841548311, Iran
a r t i c l e i n f o
a b s t r a c t
Article history:
The effects of vacuum and/or calcination temperature on the structure and texture of alumina were
investigated using XRD, SEM, BET, DTG and FT-IR. Elucidation of the observed reactivity and selectivity
over the catalyst was carried out via elimination reaction of secondary (2-octanol) and tertiary (1,2-
diphenyl-2-propanol, DPP) alcohols. The mechanism of (R)-, (S)-2-octanol and (R)-, (S)-DPP dehydration
over (1 0 0) and (1 1 0) surfaces of alumina was modeled using density functional theory (DFT).
© 2010 Elsevier B.V. All rights reserved.
Received 8 December 2009
Received in revised form 12 April 2010
Accepted 13 April 2010
Available online 20 April 2010
Keywords:
Alumina
Morphology
Vacuum
Syn/anti-elimination
SEM
DFT
1. Introduction
The bohemite was obtained in the absence of surfactant. At 600,
1000 and 1200 ◦C, the predominant oxides are ␥-, ␪- and ␣-Al2O3,
Alumina compounds are currently used as industrial catalysts,
catalyst supports, adsorbents, and ion exchangers because of their
thermal, chemical, and mechanical stability [1]. They are gener-
ally produced industrially by precipitation, drying, and calcination
of aluminum oxy-hydroxides. The catalytic properties of alumina
largely depend on their crystalline structure and texture. Therefore,
great efforts have been devoted to master these physicochemi-
cal properties. Among the synthesis methods that can be used
for preparing alumina, the sol–gel method provides an attrac-
tive and convenient route, one of its advantages being accurate
control over such structural and textural properties as high spe-
cific surface area, homogeneous pore size distribution, and high
purity. Despite these advantages, application of the method to
prepare materials of catalytic interest is not yet as mature as it
is in preparing glasses and ceramics [2]. Amini and Mirzaee [3]
used various aluminum alkoxide precursors for the preparation
of boehmite by hydrothermal assisted sol–gel processing. They
reported crystallized boehmite prepared from aluminum isobu-
toxide or aluminum isopropoxide by hydrothermal hydrolysis at
150 ◦C. Park et al. [4] reported the preparation of aluminum oxide
nanoparticles by the hydrolysis of aluminum isopropoxide fol-
lowed by calcinations, in the presence of surface stabilizing agents.
respectively. Le Bihan et al. [5] examined the introduction of chelat-
ing agents such as butan-1,3-diol or acetylacetone in the aluminum
isopropoxide solution. A decrease in the crystallinity of alumina
was observed when the added amount of complexing agent was
increased. Kim [6] prepared bohemite by using the hydrolysis of
aluminum isopropoxide in a dilute solution of isopropyl alcohol
followed by the addition of ammonium hydroxide. According to
his report the amorphous alumina was transformed to a crys-
talline when calcination temperature increased to 500 ◦C. Between
500 and 900 ◦C, crystalline alumina was converted to ␦-Al2O3. The
phase conversions and morphological changes of alumina aero-
gels during the heat treatment were investigated by Keysar et
al. [7]. The phase conversions in the range 20–1300 ◦C were as
following: boehmite/pseudoboehmite → ␩-Al2O3 → ␪-Al2O3 → ␣-
Al2O3. Furthermore, several authors have investigated the texture
of alumina materials [8–18]. Unfortunately, most of their works
were restricted to the synthesis and morphology of aluminas and a
detailed understanding of their catalytic properties is still lacking.
Surprisingly, there was no reported method for the preparation of
alumina under vacuum.
Dehydration of alcohols to alkenes over alumina dates back a
nearly century ago [19]. During this period large numbers of pub-
lications have addressed the mechanism of these reactions and
physicochemical properties of alumina [20–27]. The surface and
interior structure of metal oxides are complex and remains to be
defined. Understanding the structure of aluminum oxide and the
∗
Corresponding author. Tel.: +98 311 3913257; fax: +98 311 3912350.
E-mail address: dabbagh@cc.iut.ac.ir (H.A. Dabbagh).
1381-1169/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2010.04.007

56
H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
way atoms arrange in space in these oxides is the cornerstone of
investigations for the past century. The widely accepted model, pro-
posed by Peri [23] assumes a random configuration of hydroxyl
groups after dehydration leaving adjoining residual oxide ions,
oxide vacancies and exposed aluminum ions. Recently, we depicted
␥-Al2O3 formation from aluminum isopropoxide as a polymeriza-
tion process and initiated a detail analysis of elimination reaction
of secondary and tertiary alcohols over pure and modified ␥-Al2O3
[26,27]. The major difference in aluminum oxide polymer is that
the bonds have more ionic character. Loss of hydroxyl groups from
polymer chain would increase the cross-links producing larger par-
ticle size. Further, it was concluded that the dehydration of alcohols
over metal oxides depends on not only the steric interaction of the
intermediate and/or transition state but is also strongly dependent
on the preparation conditions of the catalyst and reaction con-
ditions. Dabbagh and Yalfani investigated the effect morphology
of alumina and mixed alumina–thorium oxide on reactivity and
selectivity of alcohol dehydration [27].
A defect spinel-like structure is commonly accepted for ␥-Al2O3
[28]. Spinel has 24 cations and 32 anions in the cubic unit cell.
In the ideal spinel structure the aluminum atoms occupy 10% of
the tetrahedral sites and 46.7% of the octahedral sites formed by
the cubic close-packed array of oxygen atoms. However, to satisfy
the Al2O3 stoichiometry, some Al vacancies must be introduced.
This is the origin of the historical formula for the conventional
cubic cell Al21+1/3ꢀ2+2/3O32, where ꢀ denotes a vacancy at the spinel
sites.
Recently, theoretical calculations have developed new models
of ␥-Al2O3 [29–45]. Ionescu et al. [29] and Maresca et al. [30] pro-
vide an exhaustive study about structural and electronic properties
of ␥-Al2O3 ideal spinel structure. The distribution of vacancies into
this structure has been the subject of great controversy [31]. Some
studies indicate that the vacancies are at tetrahedral sites, others
at octahedral, and others show different proportions of tetrahedral
and octahedral vacancies. The correlation between the positions
of the cation vacancies and the energies of the possible structures
was examined by Gutierrez et al. [31] and Taniike et al. [32]. They
concluded that an octahedral vacancy was more stable than a tetra-
hedral vacancy.
Sohlberg et al. [33–36] supported the presence of various
amounts of hydrogen within the bulk structure of spinel ␥-Al2O3.
While Wolverton and Hass [37] indicated that hydrogen spinel is
thermodynamically unstable with respect to decomposition into
an anhydrous defect spinel plus boehmite. Digne et al. [38–41]
reported a complete non-spinel structure based on molecular
dynamic simulations and DFT calculations of the dehydration of
boehmite. Nelson et al. performed DFT and simulated XRD calcu-
lations for identification of three different spinel-related ␥-Al2O3
structures (fully and partly hydrogenated structures and defect
spinel) versus non-spinel models [42,43]. They have concluded that
the spinel-related structure model is better than the non-spinel
model for describing the structure of the bulk ␥-Al2O3. This idea
immediately was criticized by Paglia et al. [44]. They showed that a
non-spinel structure matches data from neutron diffraction exper-
iments [45]. Recently, a single crystal X-ray diffraction study of
␥-Al2O3 was reported by Smrcok et al. [46]. Refined occupancy
parameters indicate that, in addition to the ideal spinel positions,
approximately 6% of Al ions also occupy non-spinel positions. Such
a cation distribution is in accord with the results of Paglia et al.
[45].
was investigated over the (1 0 0) and (1 1 0) surface of defect spinel
␥-Al2O3 using DFT/BLYP (Becke–Lee–Yang–Parr) [47,48] level of
calculation. These results and experimental finding were utilized
for detailed understanding of the mechanism of elimination of
these alcohols over ␥-Al2O3.
2-Octanol and DPP were selected for the following reasons. First,
the dehydration of secondary and tertiary alcohols was investigated
in one run under identical conditions. Second, high concentration
of 2-octanol was used as solvent to prevent isomerization. Third,
DPP has the potential for producing carbocation intermediate (sta-
bilizes by resonance and hyperconjugation). This helps predict the
mechanism of dehydration.
2. Experimental
2.1. Catalyst preparation
Aluminum isopropoxide was hydrolyzed in deionized water
and stirred for 30 min at room temperature until homogenous
slurry was obtained. The resulting suspension was dried at
120 ◦C for 2 days. The powder was then milled and the parti-
cle size was determined (mesh size = 100–200). Boehmite was
exposed to calcination temperatures and pressures. Samples were
calcined at 120 ◦C (BCT120NV), 250 ◦C (BCT250NV), and 350 ◦C
(BCT350NV) under atmospheric pressure (NV, no vacuum) for 24 h;
at 250 ◦C (BCT250LV) and 350 ◦C (BCT350LV) under 0.01 bar for
24 h; and at 600 ◦C (BCT600NV) under the atmospheric pressure
(BCT600LV) under 0.01 bar (LV, low vacuum), and (BCT600HV) under
1 × 10−6 bar (HV, high vacuum) for 6 h.
2.2. Catalyst characterization
X-ray diffraction (XRD) patterns were collected on powder
diffractometer model Philips Xpert MPD using a Cu K␣ radiation
source (ꢀ = 1.51418 Å). Scanning electron microscopy (SEM) exper-
iments were performed using a Philips XLC electron microscope.
Samples were vapor-deposited with gold before analysis. Infrared
(IR) spectra were recorded on a Jasco-FT-IR-680 plus apparatus
using KBr transparent discs. TGA and DTG curves were analyzed on
a LINAEIS-L70/2171 apparatus. The single point BET analysis was
performed by nitrogen physisorption at 298 K after outgassing at
300 ◦C using an AFAP 2000.
2.3. Catalytic activity and selectivity measurements
1,2-Diphenyl-2-propanol (DPP) was prepared according to the
Grignard method reported earlier [26]. Elimination reactions of 2-
octanol and DPP were carried out over the catalysts (BCT250NV,
BCT250LV, BCT350NV, BCT350LV, BCT600NV, BCT600LV and BCT600HV)
at the reaction temperature 280 ◦C (the reaction temperature over
BCT250 catalysts was 250 ◦C). The catalytic reactions were per-
formed in a vertical plug flow reactor made of Pyrex glass and
fitted with a thermal well extended to the center of the catalyst bed.
About 20–30 cm3 of the reactor volume above the catalyst bed con-
tained Pyrex glass beads to serve as preheater. A 4% solution of DPP
in 2-octanol was exposed to the catalyst at a flow rate of 18 cm3/h.
Liquid products were collected at room temperature at increasing
time intervals. Alcohol conversion and product distribution were
monitored using the Agilent 6890N gas chromatograph equipped
with a capillary HP-5 column. The column properties were: 30 m
long, 0.32 mm in inner diameter, and 0.25 ␮m film thick. The col-
umn stationary phase was (5% phenyl)-methyl polysiloxane. The
products were identified using the Fisons 8060 GC-MASS instru-
ment equipped with a 30 m HP-5 capillary column.
Unlike many reported studies that concentrate on hydrothermal
process adjustments, the present work focuses on the calcination
process and the effects of temperature and vacuum on subse-
quent properties (structure, texture, reactivity and selectivity) of
the synthesized alumina. Finally, the molecular adsorption of alco-
hols (2-octanol and DPP) stereo isomers with the key conformers

H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
57
Fig. 3. XRD patterns for the boehmite after calcination at 350 ◦C. BCT350NV (top)
and BCT350LV (bottom).
Fig. 1. XRD pattern of boehmite dried at 120 ◦C (BCT120NV) under one atmosphere.
title molecules over the dehydroxylated surface. Most experimen-
tal results of alcohol dehydration are reported over the catalysts
calcined at 600 ◦C. This catalyst shows small number of hydroxyl
over the surface. This model surface can play an important role at
higher temperatures.
2.4. Computational details
Potentialenergy scan(PES)ofalcohols in ordertosearch the con-
formers with minimum energy was performed at HF/6-31G** level
of calculation using the GAUSSIAN 03 program [49]. The final DFT
calculations were performed on the more stable conformers using
the DMOL3 program [50,51]. The double numerical plus polariza-
tion function (DNP) [50,51] and BLYP [47,48] generalized gradient
approximation were used in all calculations. It should be men-
tioned that the DNP basis set includes a double-zeta quality basis set
that a p-type polarization function added to hydrogen and d-type
polarization functions added to heavier atoms, and is equivalent to
6-31G** Gaussian basis sets [52]. Each basic function was restricted
to a cutoff radius of 4.5 Å. Effective core potentials (ECP) were used
The adsorption energy (ꢁEads) of alcohol was calculated as:
ꢁEads = E(adsorbed alcohol on surface) − E(alcohol) − E(surface)
where E(adsorbed alcohol on surface) refers to the energy of the system
which is formed by an adsorbed alcohol and the surface. E(alcohol)
and E(surface) refer to the energy of an isolated alcohol and the bare
surface.
To determine the activation energy for a specific reaction path-
way, a transition state was identified by the complete linear
synchronous transit (LST) and the quadratic synchronous transit
(QST) methods.
to treat the core electrons and a k-point set separation of 0.07 Å−1
.
The tolerance of the energy change was set to medium (2.0 e−5 Ha)
for all calculations.
The (1 0 0) and (1 1 0) surface orientations from defect spinel
␥-Al2O3, based on the earlier Ionescu et al. model [29,53], were
cleaved. The effects of the surface defects previously reported in
Ref. [53]. We imposed a vacuum of 15 Å between slabs in the direc-
tion of the crystal lattice, perpendicular to the surface plane, and
periodically repeated the unit cell through space. Also, we selected
four layers of atoms (Al64O96 for (1 0 0) and Al48O64 for (1 1 0) sur-
faces) while, the two bottom layers of the slab were constrained.
The main specific feature of these surfaces is the asymmetric prop-
erty. This feature stems from the holes and vacancies of the defect
␥-Al2O3 bulk structure.
3. Results and discussion
3.1. XRD measurement
We assessed the XRD spectra in order to realize which kind
of transition alumina had formed. In addition, according to peak
heights and peak widths, we presented some qualitative expla-
nation of crystallinity and particle size. The XRD pattern of the
dried powder (BCT120NV) prepared under the atmospheric pres-
sure at 120 ◦C (Fig. 1) showed reasonable compatibility with the
boehmite spectrum reported in the literature [55–58]. Evaluation
of XRD diagrams of catalysts prepared at 250 ◦C (Fig. 2) predicted
that ␥-Al2O3 phase would not be completed, but that the existing
phase (of the catalyst prepared under vacuum, BCT250LV) would
There are several literature reports describing the molecular
adsorption of alcohols over the hydroxylated ␥-Al2O3 surfaces
[34,54]. This study focuses on the molecular adsorption of the
Fig. 2. XRD patterns of boehmite calcined at 250 ◦C for 24 h: (left, BCT250NV) under atmosphere (right, BCT250LV) under 0.01 bar.

58
H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
Fig. 5. FT-IR spectra of boehmite calcined at temperatures 120, 250, 350, and 600 ◦
under atmospheric pressure.
C
Fig. 6. FT-IR spectra of boehmite calcined at different temperatures under 0.01 bar.
Al–O–Al bonds developed resulted from catalyst dehydroxyla-
tion.
Fig. 4. XRD spectrums of boehmite calcined at 600 ◦C under different pressures: one
atmosphere (top); low vacuum (middle); high vacuum (bottom).
The FT-IR spectrum of boehmite (Fig. 5) clearly shows a broad
OH stretching mode from 2500 to 3840 cm−1, which stands for
surface hydroxyls and adsorbed water. The peaks at 1072 cm−1
and a shoulder at 1166 cm−1 follow from the Al–O–Al asymmetric
and symmetric bending modes. The observed decrease in the OH
peak region is a good indication of the temperature effect on the
dehydroxylation process. As we will conclude from thermal anal-
ysis experiments below, dehydroxylation and ␥-Al2O3 formation
are two events taking place as a result of boehmite thermal treat-
ment. The decrease in the OH peak area and the peaks developing
between 500 and 1000 cm−1 an indication of ␯-AlO4 and ␯-AlO6
are convincing evidences supporting TGA and DTG (predicts signifi-
cant water loss at 250–491 ◦C). Several publications have addressed
the hydroxyl surface of alumina using IR spectroscopy [56,59,60].
The IR spectra of boehmite calcined at 600 ◦C under reduced pres-
sure (0.01 and 1 × 10−6 bar) show large reduction of hydroxyl group
(major loss of water at 250–350 ◦C) and appearance of a small peak
at 1023 cm−1 (Al–O–Al bending vibrations) (Figs. 6 and 7). This peak
is absent in the spectrum for alumina prepared under atmospheric
pressure. This is an indication of morphologic alteration. Forma-
tion of Al–O–Al bond might be a consequence of the condensation
of two OH groups lying on a particle’s pore surface (inner-structure
change) or two hydroxyls located on the outer surface of two dif-
ferent alumina particles (larger ␥-Al2O3 particles). Reactivity and
selectivity data (Section 3.6) predicts the latter case to be the dom-
inating factor under vacuum.
be closer to ␥-Al2O3 model than the boehmite (BCT120NV) sam-
ple.
The XRD spectra of boehmite calcined at 350 ◦C (BCT350NV and
BCT350LV) are not quite similar to the pattern observed for a defined
␥-Al2O3 structure; however, the main peaks, that are characteris-
tic of ␥-Al2O3, did appear (Fig. 3). The only difference observed
was that the two small peaks at 2ꢂ ≈ 39◦ and 2ꢂ ≈ 62◦ were not
fairly apparent as they lied in the ␥-Al2O3 spectrum. The XRD pat-
terns of boehmite calcined at 600 ◦C (BCT600NV) (Fig. 4) show the
entire characteristic features of ␥-Al2O3 as reported in the literature
[4,55–58].
A qualitative measure of crystallinity and particle size was
obtained based on peak heights and peak widths of 2ꢂ ≈ 67◦. A
decrease in peak widths was observed with an increase in tem-
perature. Similarly, a decrease in peak height was observed under
reduced pressure at a constant calcination temperature, which is a
good indication of low crystallinity (Fig. 4).
3.2. Infrared spectra
FT-IR technique was used to help obtain a thorough under-
standing of the structural features of the prepared catalysts. For
instance, identification of adsorbed water, determination of the
amount of OH groups present on the surface, and evaluation of

H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
59
ature). The surface area for ␥-Al2O3 was increased under reduced
pressure.
3.5. Thermal analysis
Thermal analysis was performed both to evaluate the thermal
behavior of boehmite and to collect more information on the dehy-
dration process during catalyst preparation. Alphonse and Courty
pointed out that four steps taken into consideration in thermally
stimulated transformation of nanocrystalline boehmite into alu-
mina [57]. Bokhimi et al. recorded DTA curves of various boehmites
and observed two endothermic and one exothermic peak. The
two endothermic peaks corresponded to water desorption and
boehmite transition to alumina, respectively. The exothermic peak
that appeared at above 1100 ◦C was assigned to ␣-Al2O3 formation
[61].
Fig. 7. FT-IR spectra of boehmites calcined at 600 ◦C under atmosphere (top),
0.01 bar (middle) and 1 × 10−6 bar (bottom).
In this research, the TG and DTG curves were recorded in the
temperature range between 22 and 1200 ◦C. A significant weight
loss was observed at temperatures between 200 and 500 ◦C (IR
predicts major loss of water at 250–350 ◦C). This temperature
range encompasses important thermal processes such as transi-
tion alumina creation which starts below 300 ◦C (Figs. 9 and 10).
Examination of the DTG diagram justifies the conclusion that
the imperative changes during boehmite heat treatment include
physisorbed water desorption, chemisorbed water desorption, and
formation of ␥-Al2O3. Depending on the starting aluminum oxy-
hydroxide, treatment at higher temperatures could result in more
dehydroxylation, appearance of other transition aluminas, and,
finally at above 1100 ◦C, formation of a ␣-Al2O3. These thermal
findings complement the FT-IR data presented above which also
indicated the phase transformation at higher temperature or under
reduced pressure.
3.3. SEM studies
In order to find proof for the possible morphologic changes, as
already considered in the FT-IR study, SEM images of BCT600HV and
BCT600NV were recorded at a magnification of 33,872× (Fig. 8), in
which the effect of reduced pressure on particle size is apparent.
While particle size change confirms our idea about external surface
hydroxyl condensation, condensation of inner surface hydroxyls
is still possible. One can notice that lower water vapor pressure
under vacuum simply facilitates dehydroxylation, including also
the external OH groups. Another explanation which was inferred
from the SEM images is that when particles gain energy during
calcinations, the air present between particles is an effective fac-
tor in reducing particle energy. When air is sucked out by the
vacuum pump, there is no fluid available for particles to collide
with. Consequently, the only way for particles to transfer their
energy is to have collisions with each other. The more the colli-
sions, the greater is the possibility for particle size growth. The
role of collision is to bring two external OH groups closer and to
increase the chances for water loss and creation of an Al–O–Al
bond.
Interpretation of the DTG pattern indicates that boehmite trans-
formation to ␥-Al2O3 begins at around 240 ◦C and comes to an end
at 491 ◦C. It is apparent that the highest rate of dehydroxylation
occurs at 402 ◦C (Fig. 10).
3.6. Effect of calcination temperature and vacuum on catalyst
activity and selectivity
Reaction of 1,2-diphenyl-2-propanol (DPP) in 96% 2-octanol was
examined over calcined catalysts (BCT250 NV, BCT250 LV, BCT350NV,
BCT350LV, BCT600NV, BCT600LV, and BCT600HV) at a flow rate of
18 cm3/h at 280 ◦C on a Pyrex vertical flow reactor. 2-Octanol was
used as the solvent and, more importantly, as an isomerization
preventer. Results are summarized in Tables 1–3. Furthermore,
reaction of 2-octanol helped us to gain important information on
the dehydration of secondary alcohols. Elimination reactions of
alcohols were carried out under identical conditions including tem-
3.4. BET analysis
The Brunauer–Emmett–Teller (BET) specific surface area for
the boehmite (BCT120NV) was measured 352.0 m2/g. This value
for catalysts BCT250LV, BCT350NV, BCT350LV, BCT600NV, BCT600LV,
and BCT600HV was measured 242.7, 246.2, 273.2, 150.0, 164.5 and
179.4 m2/g, respectively. These results indicate that the surface area
of boehmite drastically was reduced at 250 ◦C (calcinations temper-
Fig. 8. SEM images of BCT600NV (left) and BCT600HV (right).

60
H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
Fig. 9. TG curve recorded for boehmite (mass = 15 mg, heating rate = 10 ◦C/min).
perature (280 ◦C with the exception of the catalysts which were
calcined at 250 ◦C for which the reaction temperature was kept at
250 ◦C), mesh size, flow rate, amount of catalyst, concentration, and
reactor type.
The questions raised here include whether the reaction of alu-
minum oxide (prepared under vacuum and/or different calcination
temperatures) with secondary alcohol (2-octanol) and tertiary
alcohols (DPP) affects the dehydration reaction. If so, what would
then be the mechanism of elimination for these reactions? What
would the effect of vacuum be on the structure and morphol-
ogy of the catalysts? To answer these questions, we studied the
effects of changes in vacuum pressure and calcination tempera-
Fig. 10. DTG curve recorded for boehmite (mass = 15 mg, heating rate = 10 ◦C/min).

H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
61
Table 1
Effect of calcinations temperature on alcohol conversions under one atmosphere or
vacuum at 280 ◦C.
Catalyst
2-Octanol%a
DPP%
DPP/2-octanol
BCT250NVb
BCT250LVb
BCT350NV
BCT350LV
BCT600NV
BCT600LV
BCT600HV
27
41
50
44
47
64
46
89
100
95
96
97
3.3
2.4
1.9
2.2
2.1
1.5
2.1
99
97
a
Percent calculated by GC.
b
Reactions were carried out at 250 ◦C.
Table 2
Effect of calcinations temperature on product distribution of dehydration of DPP
under one atmosphere or vacuum at 280 ◦C.
Catalyst
1-Alkene
E-2-Alkene
Z-2-Alkene
1-Alkene/
2-alkenes
E/Z
BCT250NVa
BCT250LVa
BCT350NV
BCT350LV
BCT600NV
BCT600LV
BCT600HV
60.80
57.37
46.14
50.29
46.18
49.12
48.42
22.46
17.10
33.73
31.40
36.14
34.28
32.91
16.73
25.53
20.13
18.30
17.68
16.78
18.67
1.55
1.34
0.86
1.01
0.86
0.96
0.94
1.34
0.67
1.68
1.72
2.04
2.04
1.76
Reactions were carried out at 250 ◦C.
a
Fig. 11. Total electron density of alumina (1 0 0) surface after adsorption of the most
stable conformer of R-DPP (top) and S-2-octanol (bottom).
respectively (Table 1). Generally, tertiary alcohols reacted faster
(formation of a stable 3◦ cation) than secondary alcohols under E1
elimination. We concluded that the elimination reactions of tertiary
and secondary alcohols over ␥-Al2O3 proceeds through E2 mech-
anism [25,26,62]. Then, what other factor(s) may be involved in
controlling the very high conversion rate of DPP over 2-octanol?
One may assume the stronger adsorption of DPP plays a major part.
The interaction of the two phenyl groups with the surface favors the
conversion of DPP. The repulsion of long chain octyl moiety lowers
the conversion.
Reactivity and selectivity (E/Z, trans/cis and 1-alkene/2-alkenes)
were slightly improved under reduced pressure for both alcohol
dehydration with the exception of catalysts calcined at 250 ◦C (27%
conversion of 2-octanol under no vacuum; 41% under vacuum)
and for catalyst BCT600LV higher conversion (64%) of 2-octanol
was observed under low vacuum (47% under no vacuum). The
cis/trans ratio of 4.2, 3.0, 3.1 and 2 was measured for 2-octanol over
BCT250LV BCT350LV, BCT600HV and BCT600NV, respectively. Isomer-
ization (formation of 3-octenes) was decreased for the catalysts
Scheme 1. Schematic representation of DPP dehydration reaction.
ture on reactivity (conversion of alcohols), stereoselectivity (e.g.,
the ratio of syn/anti-elimination), regioselectivity (e.g., the ratio of
Saytzeff/Hofmann elimination), and the structure of the catalysts.
Conversions of 27, 50, 47% (for 2-octanol) and 89, 95, 97% (for
DPP) were obtained under one atmosphere at 250, 350, and 600 ◦C,
Table 3
Effect of calcinations temperature on product distribution of dehydration of 2-octanol under one atmosphere or vacuum at 280 ◦C.
Catalysts
1-Octene
t-2-Octene
c-2-Octene
3-Octene
1-Ene/2-ene
Cis/trans
Dioctyl ether A
Dioctylether B
BCT250NVa
BCT250LVa
BCT350NV
BCT350LV
BCT600NV
BCT600LV
BCT600HV
37
38.
30
31
26
30
31
13
11
19
17
23
19
16
46
45
48
50
46
48
50
0.23
0.11
2.7
2.0
4.0
0.6
0.7
0.4
0.5
0.4
0.5
0.5
3.7
4.2
2.5
3.0
2.0
2.5
3.1
2.2
3.0
0.2
0.4
0.2
0.6
0.3
2.15
3.18
0.20
0.32
0.17
0.55
0.31
1.2
2.3
a
Reactions temperature at 250 ◦C.

62
H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
Table 4
Relative energy, ꢁErel (kcal/mol), of alcohol conformers over the (1 0 0) and (1 1 0) surfaces of alumina before and after adsorption, adsorption energy, and the Alalumina–Oalcohol
(Al–OHR) bond distance (Å); values in parentheses are related to (1 1 0) surface.
Conformers
ꢁErel (kcal/mol) before adsorption
ꢁErel (kcal/mol) after adsorption
Adsorption energy (kcal/mol)
Alalumina–Oalcohol distance (Å)
(S)-2-Octanol
1a
1b
1c
1d
0.871
0.871
0.342
0.000
0.000 (0.757)
1.032 (2.339)
12.693 (24.124)
5.247 (0.000)
−38.227 (−62.890)
−37.195 (−61.308)
−25.005 (−38.995)
−32.109 (−62.776)
2.051 (1.985)
2.043 (1.901)
2.052 (1.914)
2.082 (1.928)
(R)-2-Octanol
2a
2b
2c
0.000
0.342
0.871
0.907 (6.332)
3.292 (25.721)
0.000 (0.000)
−32.021 (−60.939)
−29.977 (−41.891)
−33.800 (−68.143)
2.072 (1.916)
2.091 (1.931)
2.072 (2.051)
(S)-DPP
3a
3b
0.000
0.014
1.419
8.447 (4.252)
1.875 (0.000)
0.000 (23.818)
−37.947 (−61.448)
−43.365 (−64.547)
−46.945 (−42.433)
2.178 (1.981)
2.128 (1.904)
2.130 (2.198)
3c
(R)-DPP
4a
4b
0.000
0.014
1.419
0.000 (1.924)
1.165 (0.000)
7.467 (19.703)
−49.382 (−72.447)
−44.320 (−74.385)
−39.354 (−56.087)
2.082 (1.964)
2.064 (1.980)
2.293 (2.024)
4c
prepared at 250 ◦C under vacuum and was increased for catalyst
BCT600NV. Ether formation was increased for catalysts calcined at
lower temperatures under vacuum (Scheme 1 and Table 2).
There are several reactions that may proceed under vacuum at
elevated temperature. Obviously, more hydroxyls group are lost
(loss of water) under vacuum which could enhance an increase
in the number of cross-links. These links could form new bonds
within the alumina particle (intraparticle cross-links, formation
of smaller pores and crevices) and/or between the adjacent par-
ticles (interparticles cross-links) which make larger particle size
(see SEM under high vacuum). Initially we assumed that the for-
mer should make the catalyst more selective. Catalyst with smaller
pores or crevices should be more selective toward reacting with
larger molecules. However, much smaller change in activity and
selectivity was observed than it was anticipated. Apparently, the
small number of pores and crevices formed under vacuum can-
Table 5
The distance between eliminable hydrogens and basic sites of the (1 0 0) and (1 1 0) alumina surfaces.
Conformers
Eliminable hydrogens
Distance (Å)
(1 0 0)
Product
(1 1 0)
(S)-2-Octanol
H␣
H␤
2.616
4.220
3.044
–
2.740, 2.953
3.556, 3.816
3.814
3.443
2.988
4.201, 4.298
–
2.693, 3.483
3.239, 3.709
2-Octanone
cis-2-Octene
cis-2-Octene
2-Octanone
1-Octene
2-Octanone
cis-2-Octene
trans-2-Octene
1-Octene
1a
2.656, 3.014, 3.067
4.107, 4.306
3.102
2.545, 3.859
3.845
2.633, 2.993
2.586, 2.931, 3.349
4.021, 4.338
3.809
3.522, 3.873
–
ꢀ
H␤
H␣
1b
1c
ꢀ ꢀ
H␤
H␣
H␤
ꢀ
H␤
ꢀ ꢀ
H␤
H␣
2-Octanone
trans-2-Octene
1-Octene
1d
ꢀ
H␤
ꢀ ꢀ
H␤
(R)-2-Octanol
H␤
3.422, 3.455, 4.080
3.765, 3.858
2.788, 3.151, 3.498
2.532, 3.044
2.799, 3.169, 3.752
4.419, 4.633
2.623, 3.239, 3.400
2.689, 3.474
3.896, 4.005
2.993, 3.415
3.606
trans-2-Octene
1-Octene
trans-2-Octene
cis-2-Octene
1-Octene
2a
ꢀ ꢀ
H␤
H␤
2b
2c
ꢀ
H␤
ꢀ ꢀ
H␤
–
ꢀ
H␤
3.872, 4.412
3.311, 3.493, 3.851
cis-2-Octene
1-Octene
ꢀ ꢀ
H␤
(S)-DPP
H␤
3.617, 3.687
3.810
3.641, 4.247
3.763, 4.192
2.773, 3.962
2.523, 2.837
4.286, 4.348
4.487
E-2-Alkene
1-Alkene
E-2-Alkene
Z-2-Alkene
1-Alkene
3a
ꢀ ꢀ
H␤
H␤
2.589, 3.673, 3.958
4.117, 4.456
2.945, 3.064, 3.381, 3.094
2.620, 3.937, 3.837
2.396, 3.553, 3.351
3b
3c
ꢀ
H␤
ꢀ ꢀ
H␤
ꢀ
H␤
Z-2-Alkene
1-Alkene
ꢀ ꢀ
H␤
4.058, 4.253
(R)-DPP
ꢀ ꢀ
4a
H␤
2.653, 2.769, 3.327, 3.415
4.315, 4.442
4.423
3.261, 3.667, 3.113
3.433, 4.003, 4.072
2.609, 2.735, 3.480, 3.536
4.571
4.529
3.150, 3.082, 3.536
3.679, 3.846
1-Alkene
Z-2-Alkene
E-2-Alkene
1-Alkene
1-Alkene
H␤
4b
4c
ꢀ
H␤
ꢀ ꢀ
H␤
ꢀ ꢀ
H␤

H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
63
Fig. 12. Portion of S-2-octanol conformers over (1 0 0) and (1 1 0) alumina surfaces.
not compete with medium and large size pores and crevices which
control the overall reactivity and selectivity of the bulk alumina.
showed that all four alcohols chemisorb over the alumina surface
when they come sufficiently close to the surface with proper orien-
tation. Feng et al. [54] have presented a detailed theoretical study
on isopropanol adsorption on both clean and hydrated ␥-Al2O3
(1 0 0) and (1 1 0) surfaces. All possible adsorption configurations
were considered in their calculation.
In this study computation of the optimized geometry of the
defect spinel alumina predicted an unsymmetrical (1 0 0) and (1 1 0)
surfaces. This promoted an investigation of the effect of the stere-
ochemistry (R or S configuration) of the adsorbed chiral alcohol
over the unsymmetrical surfaces (Figs. 11–15). This raises the ques-
3.7. Quantum chemical calculation
The adsorption of alcohols over ␥-Al2O3 was investigated by
several research groups [34,54,63]. The adsorption of methanol
was studied by De Vito et al. [63]. They concluded that methanol
adsorbs on a tetrahedral aluminum ion forming a covalent bond. Cai
and Sohlberg [34] computed the adsorption of methanol, ethanol,
propanol, and isopropanol over the ␥-Al2O3 (1 1 0) surface. They

64
H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
Fig. 13. Portion of R-2-octanol conformers over (1 0 0) and (1 1 0) alumina surfaces.
tion, whether; the surface with the adsorbed chiral alcohol could
produce diasteroselective property.
Lewis acid site and the face of a simple ␲ system with an electro-
static origin. These interactions stabilize the alumina surface and
they are more important than the weak C–H· · ·Al or C–H· · ·O inter-
actions of octyl moiety. Fig. 11 shows the total electron density
of ␥-Al2O3 (1 0 0) surface after adsorption of the most stable con-
former of R-DPP and S-2-octanol. In the former alcohol two Al· · ·␲,
two O· · ·␲, seven C–H· · ·O and a C–H· · ·Al distances are almost 3.8,
4.6, 3.4, 4.1, 2.6 2.8, 3.0, 4.3, 3.4, 3.5, 2.7 and 3.2 Å, respectively. For
the latter alcohol two C–H· · ·Al and six C–H· · ·O distances are 3.3,
4.4, 2.6, 3.2, 2.9, 3.3, 2.9 and 3.6 Å, respectively.
Two additional questions were raised earlier regarding the fac-
tors controlling the conversion (reactivity) and product distribution
(selectivity) of 2-octanol and DPP over ␥-Al2O3. These questions
can now be satisfactorily answered along the following lines. The
low conversion of the former alcohol is related to the repulsion of
the long chain octyl group from the surface. The high conversion of
the latter alcohol is presumed to be a result of strong adsorption of
DPP over the surface. The pi-electron clouds of phenyl groups inter-
act with the surface stabilizing the acidic sites (Al· · ·␲ interactions
[64]). The Al· · ·␲ interaction is defined as the attraction between a
A third question raised involved the reason for this state of
affairs, which might be answered now along the following lines: the

H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
65
Fig. 14. Adsorption of S-DPP conformers over (1 0 0) and (1 1 0) alumina surfaces.
major product of 2-octanol is cis-2-alkene but for DPP it is 1-alkene.
And the last question awaiting an answer is ‘what is the effect of
the configuration of a chiral alcohol on the adsorbed species and on
product distribution of the elimination reaction?’
In this section, these questions are confronted using density
functional theory. The calculated adsorption energy, bond length,
relative energy of the title alcohol conformers before and after
adsorption (ꢁEads) and the probable products of elimination are
shown in Tables 4 and 5 and in Figs. 12–15.
The method predicted a lower adsorption energy for the
adsorbed 2-octanol (in the range of −25 to −38 kcal/mol,
Figs. 12 and 13) than for the adsorbed DPP (in the range of −38
to −49 kcal/mol) (Figs. 14 and 15) over the (1 0 0) surface. In
other words, the more negative adsorption energy value indicates
stronger adsorption of alcohol over the surface. The phenyl groups
of DPP with a high electron density adsorb more strongly over the
surface than the long chain of 2-octanol. The adsorption energy of
2-octanol conformers over the (1 1 0) surface are in the range of
−39 to −68 kcal/mol. These values are higher than those obtain
over the (1 0 0) surface. This difference is related to the nature
of alumina surface. The upper layer of (1 0 0) surface is formed
by penta-coordinated aluminum atoms; while (1 1 0) surface has
both tri- and tetra-coordinate geometry. The tri-coordinated alu-
minum atoms of the (1 1 0) surface are stronger Lewis acids than
penta-coordinated aluminum atoms [39]. These results comple-
ment the prediction of Feng et al. [54] about the adsorption of
isopropanol over the (1 0 0) and (1 1 0) surfaces. They concluded
that isopropanol was adsorbed better over the (1 1 0) surface
(ꢁEads = −45.588 kcal/mol versus −20.100 kcal/mol). In the case
of DPP conformers, the adsorption energy over the (1 1 0) sur-

66
H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
Fig. 15. Adsorption of R-DPP conformers over (1 0 0) and (1 1 0) alumina surfaces.
face (tri-coordinated aluminum atoms) is in the range of −42 to
−74 kcal/mol. These results indicated that the most stable conform-
ers of DPP have higher adsorption energy over the (1 1 0) surface
than the (1 0 0) with the exception of conformer 3b (Fig. 14b).
Bond distances (Al–OHCR) of the adsorbed (S)-2-octanol con-
formers over the (1 0 0) surface were calculated 2.051, 2.043, 2.052,
and 2.082 (Fig. 12a–d), respectively, and 2.072, 2.091, and 2.072
(Fig. 13a–c) for (R)-2-octanol, respectively. The corresponding val-
ues over (1 1 0) surface are 1.985, 1.901, 1.914 and 1.928 for
(S)-2-octanol, and 1.916, 1.931 and 2.051 for (R)-2-octanol con-
formers, respectively.
The relative energy of 2-octanal conformers (listed in Table 4)
predicts that the most stable conformation of the adsorbed alcohol
over (1 0 0) or (1 1 0) surface is different from that of the most stable
conformation of a free alcohol. For instance, the most stable con-
former of free and adsorbed alcohol over (1 0 0) and (1 1 0) surface
is 1d (CH3 and C5H11 groups are anti), 1a (CH3 and C5H11 groups
are gauche) and 1d for (S)-stereo isomer and is 2a (CH3 and C5H11
groups are anti), 2c (CH3 and C5H11 groups are gauche) and 2c for
(R)-2-octanol (Figs. 12 and 13).
Bond distances (Al–OHCR) between the oxygen atom of the
adsorbed alcohol and alumina (1 0 0) surface for the three (S)-DPP
conformers 3a, 3b, and 3c were calculated to be 2.178, 2.128, and
2.130 Å with adsorption energy values of −37.947, −43.365, and
−46.945 kcal/mol, respectively (Fig. 14a–c). These bond distances
for (1 1 0) surface are 1.981, 1.904 and 2.198 Å with adsorption
energy values of −61.448, −64547, and −42.433 kcal/mol, respec-
tively. The most stable conformer of free and adsorbed alcohol over
(1 0 0) and (1 1 0) surface is 3a (two phenyl groups are anti), 3c
(two phenyl groups are gauche) and 3b (two phenyl groups are
gauche) for (S)-stereo isomer (Figs. 14 and 15). The longer bond
distance and the lower adsorption energy were obtained for the
conformer 3a (which the two phenyl groups are anti) over (1 0 0)
surface, while the reverse of this observation was calculated for (R)-

H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
67
␤-hydrogen and basic site. Therefore, it was concluded that the
adsorption of alcohol conformers over the surface and the distance
between the basic sites and the eliminable hydrogens is a key step
to understand the mechanism of alcohol dehydration. A compre-
hensive investigation is now under way.
The eliminable hydrogens are assigned ␣, ␤, ␤ꢀ, ␤ꢀꢀ and ␥. The
distances between each of the eliminable hydrogens from the basic
site of the surface (Å) are listed in Table 5. For (S)-2-octanol, elim-
ꢀ
ination of H␤ and H␤ from conformer 1a and H␤ from conformer
1c (Fig. 12a and c) produced cis-2-octene (experimentally was pro-
ꢀ
duced as the major product). Elimination of H␤ from conformer 1c
of S-enantiomer produced trans-2-octene. For (R)-2-octanol, elim-
ination of two H␤ from the conformers 2a and 2b (Fig. 13a and b)
ꢀꢀ
favored the formation of trans-2-octene. Elimination of H␤ from
the conformers 1b and 1d of S-enantiomer (Fig. 14b and d) and the
conformers 2a, 2b and 2c of R-enantiomer (Fig. 13a–c) produced 1-
octene (Table 5). 2-Octanone is produced in very small amounts via
elimination of H␣. ␥-Elimination is less likely but possible in very
small amounts via the four conformers 1a, 1d and 2a, 2c shown in
Figs. 12a, d and a, c, respectively.
All six conformers of (R)- and (S)-DPP depicted in Figs. 14 and 15
produced 1-alkene (experimentally was reported as the major
product). The conformer 3c of (S)-DPP produced Z-alkene (exper-
imentally was produced as the minor product). Figs. 14b and 15b
both predict the formation of E- and Z-alkenes from the conform-
ers 3b and 4b of DPP. However, elimination of H␤ is favored to give
E-isomer. The distance between H␤ and the basic site of alumina
ꢀ
is about 1 Å shorter than that between H␤ (which gives Z-alkene)
and the basic site.
Fig. 16. Correlation between activation energy [for three path of products forma-
The minimum distances between eliminable hydrogens
(Table 5) and the oxygen of alumina (1 0 0) surface were found to
be 2.396 Å (conformer 3c) and 2.653 Å (conformer 4a) leading to
the 1-alkene from the most stable conformers. This distance over
the (1 1 0) surface is 2.609 Å (conformer 4a). The distance for the
formation of Z-alkene and E-alkene over (1 0 0) surface are 2.620
(conformer 3c), 2.589 (conformer 3b) Å, respectively and over
(1 1 0) surface are 2.523 (conformer 3b), 2.773 (conformer 3b) Å,
respectively.
All of the quantum chemical calculations predict that (1 0 0) and
(1 1 0) surfaces of defect spinel ␥-Al2O3 with the adsorbed chiral
alcohol have diasteroselective property. In other word, alcohols
with (R)- or (S)- configuration behave differently over the surface.
The experimental proof of this phenomenon is a challenge and
remains to be investigated hopefully by us or other research group.
tion: 1-butene (
Oalumina–H␤ 2-butanol
), cis-2-butene (
) and trans-2-butene (
)] and distance of
.
DPP that predicted the 4a conformer (which two phenyl groups are
anti) has maximum of adsorption energy. Bond distances between
the oxygen atom of the adsorbed alcohol and alumina (RCHO–Al)
for the three (R)-DPP conformers 4a, 4b, and 4c over (1 0 0) surface
were calculated to be 2.082, 2.063, and 2.292 Å with adsorption
energy (ꢁEads) values of −49.382, −44.320, and −39.354 kcal/mol,
respectively (Fig. 15a–c). These bond distances for (1 1 0) surface are
1.964, 1.980 and 2.024 Å with adsorption energy values of −72.447,
−74.385 and −56.087 kcal/mol, respectively.
Investigating the transition state (TS) for large molecules (DPP
or 2-octanol) requires super computer and state of art soft-
ware and time. Therefore, 2-butanol was used in this work as a
model molecule for describing the transition states and correla-
tion between the activation energy and the distances of basic sites
of alumina and eliminable hydrogens (␤, ␤ꢀ and ␤ꢀꢀ) of alcohol.
These results help shed some light on the elimination reaction
pathway (starting material (2-butanols/␥-Al2O3) → transition state
[TS]‡ → products) over each basic site of the alumina surface. It
should be mentioned that calculation of transition stats for 2-
butanol conformers is simple, because one need very small surface
area of alumina (containing 30 atoms, Al16O24) to gain com-
plete interaction of alcohol with the surface. A very large surfaces
(Al64O96 for (1 0 0) and Al48O64 for (1 1 0)) were required for calcu-
lating the activation energy for DPP or 2-octanol dehydration.
The correlation between activation energy (Ea) and the distance
between basic sites and eliminable hydrogens are shown in Fig. 16.
4. Conclusions
In summary, our results suggest that altering the boehmite’s
calcination conditions affects certain properties of alumina. XRD
spectra show lower crystallinity for the catalysts prepared under
reduced calcination pressure. The observed differences between
the BCT600NV and BCT600HV XRD peak widths at 2ꢂ ≈ 67◦ were
related to particle size change. This was verified by SEM. The BET
surface area of the catalysts was increased when calcined under
vacuum. A decrease in number of surface hydroxyls was obvious
(by IR) as pressure declined at constant calcination temperature
(600 ◦C).
Reactivity, stereoselectivity (E/Z or trans/cis) and regioselec-
tivity (1-alkene/2-alkenes) were slightly increased under reduced
pressure or at higher temperature for both alcohol dehydration
with the exception of catalysts calcined at 250 ◦C and for catalyst
BCT600LV. The cis/trans ratio of 4.2, 3.0, 3.1 and 2 was measured
for 2-octanol over BCT250LV BCT350LV, BCT600HV and BCT600NV,
respectively. Isomerization was decreased for the catalysts pre-
pared at 250 ◦C under vacuum (0.11%) and was increased for
ꢀꢀ
E2 elimination of H␤ of (R)-2-butanol via a basic site of alumina
produces 1-butene (TS1). E2 elimination of ␤ and ␤ꢀ hydrogens of
(S)-2-butanol leads to the formation of cis (TS2) and trans (TS3)
products, respectively. The activation energy of 1-butene formation
is smaller than that of the corresponding value for 2-butenes.
A general increase was observed for the activation energy for
elimination of a ␤-hydrogens with increase in the distance between

68
H.A. Dabbagh et al. / Journal of Molecular Catalysis A: Chemical 326 (2010) 55–68
catalyst BCT600NV (4.0%). Ether formation was increased for cat-
alysts calcined at lower temperatures under vacuum (6.18 for
BCT250LV and 0.61for BCT600HV).
Theoretical values complement the observed reactivity and
selectivity. This method predicts that (1 0 0) and (1 1 0) surfaces
with the adsorbed chiral alcohol show diasteroselective property.
The adsorption of alcohol conformers over the surface and the
distance between the basic sites-eliminable hydrogens is a key step
to understand the mechanism of alcohol dehydration.
[30] O. Maresca, A. Allouche, J.P. Aycard, M. Rajzmann, S. Clemendot, F. Hutschka, J.
Mol. Struct. Theochem. 505 (2000) 81–94.
[31] G. Gutierrez, A. Taga, B. Johansson, Phys. Rev. B 65 (2002) 0121011–0121014.
[32] T. Taniike, M. Tada, Y. Morikawa, T. Sasaki, Y. Iwasawa, J. Phys. Chem. B 110
(2006) 4929–4936.
[33] K. Sohlberg, S.J. Pennycook, S.T. Pantelides, J. Am. Chem. Soc. 121 (1999)
7493–7499.
[34] S. Cai, K. Sohlberg, J. Mol. Catal. A: Chem. 193 (2003) 157–164.
[35] S. Cai, K. Sohlberg, J. Mol. Catal. A: Chem. 248 (2006) 76–83.
[36] S. Cai, V. Chihaia, K. Sohlberg, J. Mol. Catal. A: Chem. 275 (2007) 63–71.
[37] C. Wolverton, K.C. Hass, Phys. Rev. B 63 (2000) 241021–2410216.
[38] M. Digne, P. Sautet, P. Rayboud, P. Euzen, H. Toulhoat, J. Catal. 211 (2002) 1–5.
[39] M. Digne, P. Sautet, P. Rayboud, P. Euzen, H. Toulhoat, J. Catal. 226 (2004) 54–68.
[40] M.C. Valero, P. Rayboud, P. Sautet, J. Phys. Chem. B 110 (2006) 1759–1767.
[41] M. Digne, P. Rayboud, P. Sautet, B. Rebours, H. Toulhoat, J. Phys. Chem. B 110
(2006) 20719–20720.
Acknowledgements
We would like to thank Isfahan University of Technology (IUT)
Research Council for their financial support (Grant # 86/500/9143).
The authors are grateful to Dr. Abbas Teimuri and Professor
Mohammad Ali Golozar for valuable discussions.
[42] M. Sun, A.E. Nelson, J. Adjaye, J. Phys. Chem. B 110 (2006) 2310–2317.
[43] A.E. Nelson, M. Sun, J. Adjaye, J. Phys. Chem. B 110 (2006) 20724–20726.
[44] G. Paglia, C.E. Buckley, A.L. Rohl, B.A. Hunter, R.D. Hart, J.V. Hanna, L.T. Byrne,
Phys. Rev. B 68 (2003) 1441101–14411011.
[45] G. Paglia, C.E. Buckley, A.L. Rohl, J. Phys. Chem. B 110 (2006) 20721–20723.
[46] L. Smrcok, V. Langer, J. Krestan, Acta Cryst. C62 (2006) i83–i84.
[47] A.D. Becke, J. Chem. Phys. 104 (1996) 1040–1046.
References
[48] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785–789.
[49] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
J.A. Montgomery, T. Vreven Jr., K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson,
H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian,
J.B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J.
Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A.
Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels,
M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman,
J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham,
C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen,
M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian 03, Revision B.02, Gaussian, Inc.,
Pittsburgh, PA, 2003.
[1] Y. Kim, C. Kim, P. Kim, J. Yi, J. Non-Cryst. Solids 351 (2005) 550–556.
[2] J. Sanchez-Valante, X. Bokhimi, F. Hernandez, Langmuir 19 (2003) 3583–3588.
[3] M.M. Amini, M. Mirzaee, J. Sol–Gel Sci. Technol. 36 (2005) 19–23.
[4] Y.K. Park, E.H. Tadd, M. Zubris, R. Tannenbaum, Mater. Res. Bull. 40 (2005)
1506–1512.
[5] L. Le Bihan, F. Dumeignil, E. Payen, J. Grimblot, J. Sol–Gel Sci. Technol. 24 (2002)
113–120.
[6] Y. Kim, Sep. Sci. Technol. 35 (2000) 2327–2341.
[7] S. Keysar, G.E. Shter, Y. de Hazan, Y. Cohen, G.S. Grader, Chem. Mater. 9 (1997)
2464–2467.
[8] D. Mishra, S. Anand, R.K. Panda, R.P. Das, Mater. Lett. 42 (2000) 38–45.
[9] J.J. Fitzgerald, G. Piedra, S.F. Dec, M. Seger, G.E. Maciel, J. Am. Chem. Soc. 119
(1997) 7832–7842.
[10] L. Ji, J. Lin, K.L. Tan, H.C. Zeng, Chem. Mater. 12 (2000) 931–939.
[11] D.J. Suh, T.J. Park, Chem. Mater. 9 (1997) 1903–1905.
[12] S. Rezguit, B.C. Gates, Chem. Mater. 6 (1994) 2386–2389.
[13] G. Paglia, C.E. Buckley, A.L. Rohl, R.D. Hart, K. Winter, A.J. Studer, B.A. Hunter,
J.V. Hanna, Chem. Mater. 16 (2004) 220–236.
[14] G. Paglia, C.E. Buckley, T.J. Udovic, A.L. Rohl, F. Jones, C.F. Maitland, J. Connolly,
Chem. Mater. 16 (2004) 1914–1923.
[15] T.F. Baumann, A.E. Gash, S.C. Chinn, A.M. Sawvel, R.S. Maxwell, J.H. Satcher,
Chem. Mater. 17 (2005) 395–401.
[16] G. Lefevre, M. Duc, P. Lepeut, R. Caplain, M. Fedoroff, Langmuir 18 (2002)
7530–7537.
[17] M.L. Guzman-Castillo, X. Bokhimi, A. Toledo-Antonio, J. Salmones-Blasquez, F.
Hernandez-Beltran, J. Phys. Chem. B 105 (2001) 2099–2106.
[18] C.J. Harlan, M.R. Mason, A.R. Barron, Organometallics 13 (1994) 2957–2969.
[19] H. Adkins, P. Perkins, J. Am. Chem. Soc. 47 (1925) 1163–1167.
[20] H. Knözinger, Angew. Chem. Int. Ed. 7 (1968) 791–805.
[21] H. Knözinger, R. Köhne, J. Catal. 5 (1966) 264–270.
[22] H. Knözinger, H. Buhl, K. Kochloefl, J. Catal. 24 (1972) 57–68.
[23] J.B. Peri, J. Phys. Chem. 69 (1965) 220–226.
[24] H.A. Dabbagh, C.G. Hughes, B.H. Davis, J. Catal. 133 (1992) 445–460.
[25] B. Shi, H.A. Dabbagh, B.H. Davis, J. Mol. Catal. A: Chem. 141 (1999) 257–262.
[26] H.A. Dabbagh, J. Mohammad Salehi, J. Org. Chem. 63 (1998) 7619–7627.
[27] H.A. Dabbagh, M.S. Yalfani, B.H. Davis, J. Mol. Catal. A: Chem. 238 (2005) 72–77.
[28] E.J.W. Verway, Z. Kristallogr. 91 (1935) 65–69.
[50] B. Delley, J. Chem. Phys. 92 (1990) 508–517.
[51] B. Delley, J. Chem. Phys. 113 (2000) 7756–7764.
[52] E. Kim, P.F. Weck, S. Berber, D. Tománek, Phys. Rev.
1134041–1134044.
B 78 (2008)
[53] A. Ionescu, A. Allouche, J.P. Aycard, M. Rajzmann, R.L. Gall, J. Phys. Chem. B 107
(2003) 8490–8497.
[54] G. Feng, C. Huo, C. Deng, L. Huang, Y. Li, J. Wang, H. Jiao, J. Mol. Catal. A: Chem.
304 (2009) 58–64.
[55] H.S. Potdar, K.W. Jun, J.W. Bae, S.M. Kim, Y.J. Lee, Appl. Catal. A: Gen. 321 (2007)
109–116.
[56] M.L. Guzman-Castillo, F. Hernandez-Beltran, J.J. Fripiat, A. Rodriquez-
Hernandez, R. Garcia de Leon, J. Navarrete-Bolanos, A. Tobon-Cervantes, X.
Bokhimi, Catal. Today 107 (2005) 874–878.
[57] P. Alphonse, M. Courty, Thermochim. Acta 425 (2005) 75–89.
[58] H.C. Lee, H.J. Kim, S.H. Chung, K.H. Lee, H.C. Lee, J.S. Lee, J. Am. Chem. Soc. 125
(2003) 2882–2883.
[59] P.H. Colomban, J. Mater. Sci. Lett. 7 (1988) 1324–1326.
[60] G. Urretavizcaya, A.L. Cavalieri, J.M. Porto Lopez, I. Sobrados, J. Sanz, J. Mater.
Syn. Process. 6 (1998) 1–7.
[61] X. Bokhimi, J.A. Toledo-Antonio, M.L. Guzman-Castillo, B. Mar-Mar, F.
Hernandez-Beltran, J. Navarrete, J. Solid State Chem. 161 (2001) 319–326.
[62] H.A. Dabbagh, K. Faghihi, Tetrahedron 56 (2000) 3611–3617.
[63] D.A. De Vito, F. Gilardoni, L. Kiwi-Minsker, P.Y. Morgantini, S. Porchet, A. Renken,
J. Weber, J. Mol. Struct. Theochem. 469 (1999) 7–14.
[29] A. Ionescu, A. Allouche, J.P. Aycard, M. Rajzmann, F. Hutschka, J. Phys. Chem. B
106 (2002) 9359–9366.
[64] H. Aghabozorg, F. Manteghi, S. Sheshmani, J. Iran. Chem. Soc. 5 (2008) 184–227.