Phosphated alumina catalysts: Surface properties and reactivity towards 2-PrOH decomposition

Article abstract of DOI:10.1007/s00706-007-0674-6

The influence of the type of precursor, the phosphate content, as well as the source of phosphate ions on the surface texture, acidity, and catalytic activity of phosphated aluminas has been described. Phosphated alumina catalysts were prepared by impregnating two different precursors with two different sources of phosphate in different loading levels w = 3, 6, and 10% PO ₄ ^3-. The characterisation of the catalyst was performed using X-ray powder diffraction (XRD), thermal analysis (TG), and nitrogen adsorption-desorption methods at 77∈K. The surface acidity of the catalysts has been studied by FT-IR spectroscopy of adsorbed pyridine at different temperatures. Moreover, the activity in the catalytic decomposition of isopropanol (2-PrOH) has been investigated at 520∈K. Investigation of the surface properties shows that the addition of phosphate ions does not change the crystal phase (γ-phase) and the samples prepared from gel and phosphoric acid have the highest surface area. An FT-IR study of pyridine adsorption shows both Lewis and Bronsted acid sites on the surface of the samples prepared from gel, while only Lewis acid sites are detected on the samples prepared from crystalline oxide. The catalytic activity by 2-PrOH conversion shows that the conversion of 2-PrOH as well as the selectivity towards propene formation increases from w = 3 to 6% followed by a decline for w = 10%. Moreover, the strongest activity was detected in case of phosphated alumina gel with w = 6% which gives 97.3% propene and 96.1% conversion of 2-PrOH. Springer-Verlag 2007.

Full text of DOI:10.1007/s00706-007-0674-6

Monatshefte fu¨r Chemie 138, 641–648 (2007)
DOI 10.1007/s00706-007-0674-6
Printed in The Netherlands
Phosphated Alumina Catalysts: Surface Properties and Reactivity
towards 2-PrOH Decomposition
ꢀ
Hussein A. Khalaf , Gamal A. Mekhemer, Ahmed K. Nohman, and Seham A. A. Mansour
Chemistry Department, Faculty of Science, Minia University, El-Minia, Egypt
Received July 19, 2006; accepted February 28, 2007; published online May 15, 2007
# Springer-Verlag 2007
Summary. The influence of the type of precursor, the phos-
phate content, as well as the source of phosphate ions on the
surface texture, acidity, and catalytic activity of phosphated
properties and bulk crystallinity of the oxides, re-
aluminas has been described. Phosphated alumina catalysts
amounts of phosphate anions onto metal oxides has
been found to remarkably increase the surface acidic
gardless of the type of phosphate compound used
were prepared by impregnating two different precursors with
viz. (NH4)2HPO4, (NH4)H2PO4, or H3PO4 [1–3].
The use of phosphates as promoting additives for
catalytic aluminas has been reported and their effect
has been suggested to be twofold [4]. Phosphates
play an important role as stabilizer [5]. The so-called
‘‘solid phosphoric acid’’ is a typical acidic cata-
lyst widely used for olefin hydration, alcohol and
cumene conversion, propylene oligomerization, alkyl-
ation, and isomerization reactions [6–8]. Phosphoric
acid impregnated onto alumina and titanium diox-
ide gives rise to ‘‘solid superacids’’ [9]. The super-
acidic character [10] of some solids has also been
addressed at various times and phosphated solid
superacids can be prepared by the introduction of a
calculated amount of phosphate into metal oxides
or hydroxides. The stability and the surface acidity
of phosphated catalysts are indeed directly depen-
dent on the textural properties of the parent of alu-
mina gel and the phosphation process [11]. Most of
the previous studies used similar catalyst prepara-
tion techniques [12, 13], namely alumina gel formed
by precipitation with ammonium hydroxide was
impregnated with a phosphate source of variable
concentration.
two different sources of phosphate in different loading levels
w ¼ 3, 6, and 10% PO43ꢁ. The characterisation of the catalyst
was performed using X-ray powder diffraction (XRD), ther-
mal analysis (TG), and nitrogen adsorption–desorption meth-
ods at 77 K. The surface acidity of the catalysts has been
studied by FT-IR spectroscopy of adsorbed pyridine at differ-
ent temperatures. Moreover, the activity in the catalytic de-
composition of isopropanol (2-PrOH) has been investigated at
520 K. Investigation of the surface properties shows that the
addition of phosphate ions does not change the crystal phase
(ꢀ-phase) and the samples prepared from gel and phosphoric
acid have the highest surface area. An FT-IR study of pyridine
adsorption shows both Lewis and Brønsted acid sites on the
surface of the samples prepared from gel, while only Lewis
acid sites are detected on the samples prepared from crys-
talline oxide. The catalytic activity by 2-PrOH conversion
shows that the conversion of 2-PrOH as well as the selectiv-
ity towards propene formation increases from w ¼ 3 to 6%
followed by a decline for w ¼ 10%. Moreover, the strongest
activity was detected in case of phosphated alumina gel with
w ¼ 6% which gives 97.3% propene and 96.1% conversion of
2-PrOH.
Keywords. Alumina; Phosphate; Acidity; Isopropanol.
Introduction
Many metal oxides are readily prepared in the form
of high-surface area solids. The addition of small
Isopropanol (2-PrOH) conversion is widely inves-
tigated and recently it has been used to characterize
acid–base or redox properties of catalysts [14–18].
ꢀ
Corresponding author. E-mail: husskasar@yahoo.com

642
H. A. Khalaf et al.
The conversion of 2-PrOH is known to occur through
two competitive pathways namely dehydrogenation
and dehydration. Ai [16] has assumed that the dehy-
dration of 2-PrOH probes acid sites, whereas the de-
hydrogenation probes acid and base sites functioning
in a concerted fashion. 2-PrOH usually dehydrates to
propene over acidic catalysts and dehydrogenates to
acetone over basic catalysts. The acid–base pairs re-
sponsible for acetone formation are strongly activated
in the presence of oxygen that indicates that they
are related to redox properties of the material [19].
The aim of this study is studying the effect of
some parameters like: 1) the loading levels of phos-
phate ions w ¼ 3, 6, and 10%; 2) the source of phos-
phate ions from both acids and salts; and 3) the
source of precursor i.e. pure crystalline oxide (ꢀ-
Al2O3) or alumina gel (Al(OH)3) on the characteri-
zation and the surface acidity as well as the catalytic
activity of the alumina and phosphated alumina cat-
alysts in order to have a picture as complete as pos-
sible of the acidic character of these solid catalysts.
with the relevant ASTM standards indicate that all
modified samples assume the ꢀ-structure (ASTM
card No. 29-1486) of alumina and no bulk phase
changes to that structure are observed for all sam-
ples. These results are similar to those previously re-
ported [4, 20, 21]. The introduction of phosphate ions
from ammonium salt into crystalline oxide (yPAlO),
causes a slight increase in the degree of crystallization
of these samples. Furthermore, increasing the loading
of phosphate from w ¼ 3 to 10%, resulted in more
increase in the degree of crystallization (higher peak
intensities). On the other hand, the addition of phos-
phate from phosphoric acid (yxPAlO), Fig. 1, shows
that both w ¼ 3 and 6% exhibit nearly the same
degree of crystallization, but the crystallization in-
creases for the 10xPAlO sample. The same trend was
observed for the addition of phosphate ions (from
salt or acid) into alumina gel but these samples
(yPAlH and yxPAlH) reveal low crystallization in
comparison with that of phosphated aluminum oxide
(yPAlO) as indicated by the low intensity and the
width of the XRD peaks.
Results and Discussion
Thermal Analysis
Characterization of the Catalysts
X-Ray Diffraction (XRD)
The thermal behavior of the pure alumina gel, AlH,
and of the w ¼ 6% phosphated ones has been pre-
viously studied [22], and it was found that the gel
suffers about 30% loss of its original mass on heat-
ing up to 1070 K. This is close to the value (30.6%)
expected for release of 2.5 mol of H2O, and then the
proposed composition of AlH is considered to be
Al2O3 ꢂ 2.5H2O. Thermal analysis results for AlO, in
connection with XRD results revealed that the ob-
XRD diffractograms of phosphated alumina samples
are shown in Fig. 1. These diffractograms indicate
that impregnation with phosphate ions (either from
salt or from acid) followed by calcinations at 870 K
for 1 h does not obviously modify the structure of ꢀ-
Al2O3. Inspecting the diffractograms and matching
Fig. 1. X-Ray powder diffractograms of yxPAlH series of
catalysts
Fig. 2. TG curves of phosphated samples with w ¼ 10%
loading level

Phosphated Alumina Catalysts
643
Table 1. Textural data of alumina and phosphated alumina catalysts
Particle sizea=nm
Sample
CBET
Particle sizea=nm
SBET
m2 gꢁ1
SBET
m2 gꢁ1
Sample
CBET
AlO
187
182
144
130
219
188
167
128
120
129
148
88
33
35
42
48
31
33
38
3PAlO
6PAlO
10PAlO
3PAlH
6PAlH
10PAlH
3xPAlO
6xPAlO
10xPAlO
3xPAlH
6xPAlH
10xPAlH
173
151
125
203
196
170
225
128
138
91
108
112
34
39
50
34
34
36
69
87
a
Particle size derived from Debye equation [24]
tained calcination product at 870 K is structurally ꢀ-
Al2O3. The thermogram exhibited the loss of physi-
sorbed water and partial surface dehydroxylation in
addition to mass-invariable crystallization processes
at ꢃ620 K. The addition of phosphate up to 10% into
crystalline alumina, from either salt or acid, 10PAlO
or 10xPAlO, causes the mass loss percentage to reach
a value of ca. 12% as shown in Fig. 2. Such mass
loss could be due to the partial transformation of
(NH4)2HPO4 to (NH4)H2PO4 via. the loss of NH3 as-
sociated with the thermal events of pure AlO itself,
as in case of the 10PAlO sample. The TG curves of
10PAlH and 10xPAlH, Fig. 2, show that the mass loss
percentage reached to ca. 30% through three steps up
to 870 K which are due to the loss of compositional
water from the gel. Mono- and di-basic phosphates
would exhibit stronger interactions with the surface
of alumina than polyphosphates and hence the fixa-
tion of such species by alumina is expected.
smallest particle size (31nm). Increasing the loading
level of phosphate ions added to the gel resulted in
decrease in the SBET to become 188 m2 gꢁ1 for 6PAlH
and 167m2 gꢁ1 for 10PAlH. It is worth noting that
the hysteresis loops of these samples are of the same
type (H3) and the N2 adsorption isotherms belong to
Type IV. Hence, the noticeable decrease in the SBET
3ꢁ
values with increasing the PO4 loading levels of
these samples could be attributed to the increase in
the particle size rather than any modification in the
pore structure of the pure alumina.
The addition of H3PO4 to pure crystalline oxide
(yxPAlO) and gel (yxPAlH) has a similar effect on the
specific surface area as the addition of (NH4)2HPO4
to yPAlO and yPAlH samples. This is shown in
Table 1. Regarding to yxPAlH, the hysteresis loops
for this series are somewhat mixed H2 and H3, which
indicates mesoporous and microporous nature and,
also, the sorption isotherms are of Type IV, Fig. 3.
Hence, the decrease in the SBET could be attributed
to a sort of modification in the porous nature of the
pure alumina, most probably narrowing of mesopo-
rous portion or even blockage. In comparison between
these data and that from oxide (yxPAlO) it is clear that
Surface Area Measurement
Textural data obtained from the nitrogen sorption iso-
therms and the BET-plots of the phosphated aluminas,
are cited in Table 1. From these data, it is clear that
pure crystalline alumina, AlO, has high specific sur-
face area (187m2 gꢁ1). The addition of (NH4)2HPO4
into crystalline oxide resulted in a decrease in the
surface area. Thus, the increase of the loading of
3ꢁ
PO4 causes decrease in the SBET values. These re-
sult matches well with those obtained from the XRD
[23], which is shown in Table 1 also. The table indi-
cates that the particle size increase with increasing the
loading level of phosphate ions from treating pure
oxide (AlO) by ammonium salt. In case of the addi-
tion of (NH4)2HPO4 into alumina gel, AlH, it is found
that the SBET has been changed from 187 m2 gꢁ1
for AlO to 219 m2 gꢁ1 for 3PAlH which exhibits the
Fig. 3. N2 adsorption–desorption isotherms of yxPAlH series
of catalysts

644
H. A. Khalaf et al.
the samples prepared from gel have higher surface
area. This means that the addition of H3PO4 onto alu-
mina gel has a probable effect on the retardation of
the transformation rate of spinel alumina into the co-
rundum phase and that leads to high surface area [4].
3300 cmꢁ1, Fig. 4a, displays similar bands to those
observed for pure alumina, in addition to new bands
at 3728, 3641, and 3581 cmꢁ1 due to isolated P-OH
groups. This is similar to what has been previously
observed on phosphated zirconia [25]. Inspection of
the IR spectra of phosphated samples shows a de-
crease in the Al-OH band intensities with increasing
phosphate contents, indicating that H3PO4 reacts with
all OH groups, in agreement with the results ob-
tained by Lewis and Kydd [26].
FT-IR Spectroscopy
IR Background Spectra and Surface Groups
of Catalysts
In the phosphate region, Fig. 4b, the comparison of
the IR spectra of phosphated aluminas with that of
pure alumina shows several absorptions due to the
phosphate precursor. As expected, a broad band in the
region around 1200cmꢁ1 due to ꢁ(PO) mode is well
evident for all samples. A weaker band at 1369 cmꢁ1
is assignable to ꢁ(P¼O) of surface phosphoryl groups
present probably in small proportions. In addition, the
broad bands at 1100cmꢁ1 are characteristic of P–O
bands. These results are in a good agreement with the
results obtained by Lavalley et al. [27].
The IR spectrum of AlO (not shown) displays five
bands at 3782, 3736 (shoulder), 3690, 3664, and
3594 cmꢁ1. The first four bands are due to different
types of isolated Al-OH groups, while the latter
band, at 3594 cmꢁ1, is due to associated hydroxyls
(H-bonded) [24]. The IR background spectra of
phosphated alumina samples (yPAlH, yPAlO, yxPAlH,
and yxPAlO) are similar to each other. Thus, we
have chosen the spectra of 6xPAlO and 6xPAlH for
comparison and to illustrate the similarity between
them. The spectra of 6xPAlO and 6xPAlH are shown
in Fig. 4. The spectrum in the range of 3900–
IR Spectra of Adsorbed Pyridine and Surface
Acid Sites
The IR spectrum of Py adsorbed on pure alumina
at room temperature (300 K), Fig. 5a, displays five
bands at 1614, 1595, 1580, 1490, and 1445 cmꢁ1.
These bands were shown by several authors [25, 28–
30] to be assigned to hydrogen bonded and coordi-
natively bonded Py. The band at 1614 cmꢁ1 (8a mode)
is due to Py coordinatively bonded to Lewis acid
sites of moderate strength, for example the authors
in Refs. [4, 29, 31] have assigned them to tetrahedral
aluminum vacancies. While the band at 1595 cmꢁ1 is
due to Py coordinatively bonded to octahedral alu-
minum Lewis acid sites [29, 32]. The other bands at
1580, 1490, and 1445 cmꢁ1 are due to other vibra-
tion modes of Py species coordinatively bonded with
Lewis acid sites. The absence of any absorption
around 1540 cmꢁ1 indicates that there is no Brønsted
acid site on alumina surface.
The spectrum taken from Py adsorption on pure
alumina at 520 K, Fig. 5b, shows common features
of a gradual reduction in intensities accompanied by
shifts to higher frequencies of the bands character-
istic of the adsorbed species of Py at ꢄ 300 K. Addi-
tionally, the IR spectrum of pure alumina displays
new bands at 1622 cmꢁ1. On comparing these bands
with those exhibited by the background spectrum of
Fig. 4. Background spectra of w ¼ 6% phosphated samples
at OH (a) and PO (b) region

Phosphated Alumina Catalysts
645
the spectra of Py adsorbed on yxPAlH samples at
ꢄ 300 K. The figure indicates that these samples,
yxPAlH, display, in addition to the bands displayed
for AlO, two new bands at 1543 and a weak band at
1649 cmꢁ1 due to the interaction of Py with surface
Brønsted acid sites [33]. It is noticeable that increas-
ing the loading levels of phosphate is accompanied
by an increase in the band intensities. On raising the
Py outgassing temperature to 520 K, Fig. 5b, the
spectra of all samples clearly show the disappear-
ance of the two bands at 1543 and 1649 cmꢁ1 which
are characteristic of Brønsted bound Py, and hence
indicating the removal of such sites. These results
indicate that the surface of phosphated aluminas ex-
pose different types of Lewis acid sites with different
strengths, in addition to Brønsted acid sites. It has
been reported that addition of phosphate to amor-
phous alumina increased its Lewis acidity and cre-
ated Brønsted acidity [30], and it was found that the
surface acidity of phosphated alumina was reported
to be increased first to a moderate extent [34–36]
and then to decline [34] depending on the amount
of phosphate added.
The IR spectra of adsorbed Py on the surface of
yxPAlO samples at ꢄ 300 K (not shown) show that
the 3xPAlO sample exhibits the same bands as those
of pure alumina in addition to a new weak (shoulder)
band at 1622 cmꢁ1 which is assignable to a strong
Lewis acid site. Furthermore, the increase in the load-
ing level of phosphate ion causes an increase in its
intensity. No band at 1540 cmꢁ1 was observed, this
indicates that 3xPAlO has no Brønsted acid sites
[20]. IR spectra for higher loading levels of phos-
phate ions (6xPAlO and 10xPAlO) show similar
bands to AlO in addition to two new bands appearing
at 1542 and 1648 cmꢁ1. These two bands are due to
Brønsted acid sites. Hence, the increase in loading
levels of phosphate ions is accompanied by the ap-
Fig. 5. Py adsorption spectra for yxPAlH samples at 300 K
(a) and 520 K (b)
ꢀ-Al2O3, it seems clear that the complete elimination
of hydrogen bonds (HPy) did not restore the back-
ground spectrum. This means that, an amount of Py
species is held more strongly than H-Py species. The
existence of the 8a mode of vibration at 1622 and
1614 cmꢁ1 implies the presence of at least two dif-
ferent types of Lewis acid sites with different acidic
strengths on the surface of ꢀ-Al2O3.
The spectra of Py adsorbed on yxPAlH series
are shown in Fig. 5 and Table 2. Figure 5a shows
Table 2. FT-IR absorption frequencies (ꢁꢀ=cmꢁ1) of Py adsorbed on the surface of yxPAlH samples at 300 and 520 K
Sample
ꢁꢀ=cmꢁ1 at 300 K
ꢁꢀ=cmꢁ1 at 520 K
AlO
1614-1595-1580-1490-1445
Lewis sites, no Brønsted
1622-1617-1603-1576-1493-1451
Lewis sites, no Brønsted
3xPAlH
6xPAlH
10xPAlH
1649-1617-1611-1595-1580-1543-1490-1445
Lewis þ Brønsted sites
1627-1617-1603-1576-1493-1451
Lewis sites, no Brønsted
1632-1617-1603-1576-1493-1451
Lewis sites, no Brønsted
1632-1617-1603-1576-1493-1454
Lewis sites, no Brønsted
1649-1621-1611-1595-1580-1543-1490-1445
Lewis þ Brønsted sites
1649-1621-1611-1595-1580-1543-1490-1445
Lewis þ Brønsted sites

646
H. A. Khalaf et al.
Table 3. 2-PrOH and its decomposition products on alumina
and phosphated alumina catalysts at 520 K
pearance of Brønsted acid sites on the surfaces of
these samples in accordance with previous results
[33]. The spectral patterns of yxPAlO indicating that
the interaction with Py yields different amounts of
three types of Lewis coordinated species whose 8a
mode lies at 1622, 1614, and 1595 cmꢁ1. The two
bands observed near 1614 and 1595 cmꢁ1 are rela-
tive to the most sensitive ꢁ8a mode of two different
adsorbed species. The spectrum of Py adsorption of
phosphated alumina catalysts (yxPAlO) at 520 K (not
shown) is very similar with that of phosphated alu-
mina gel (yxPAlH).
The IR spectra of the two phosphated series
(yxPAlH and yxPAlO), Fig. 5, show that there is a
strong interaction of Py with surface OH groups,
but through a mere perturbation of the H-bonding
type, as monitored by the strong 8a Py mode at
ꢄ 1595 cmꢁ1. And the interaction of Py with phos-
phated alumina yields different types of Lewis acid
sites, at least three families of Lewis coordinated
species, whose 8a modes lie at 1622, 1614, and
1595 cmꢁ1, respectively. The band at 1595 cmꢁ1 is
due to Py interaction with weak Lewis acid sites
identified by Morterra et al. [29, 31] as coordina-
tively unsaturated nearly octahedral Al ions (AlVI).
However, the band at 1614 cmꢁ1 is due to Py species
coordinated on Lewis sites of medium acidity invol-
ving both octahedral and tetrahedral sites (AlVI–
AlIV). The band at 1622 cmꢁ1 is due to the strong
Lewis acid sites (AlIV). Thus, we can conclude that
the surfaces of phosphated aluminas expose different
types of Lewis acid sites of different strengths.
Catalyst Conversion
Selectivity
% ꢅ 0:1
Propene=% ꢅ 0.1 Acetone=% ꢅ 0.1
AlO
3PAlO
6PAlO
10PAlO
3PAlH
6PAlH
10PAlH
3xPAlO
6xPAlO
10xPAlO
3xPAlH
6xPAlH
10xPAlH
79.2
87.7
95.1
93.7
88.6
95.5
94.8
88.1
95.7
94.2
89.2
96.1
95.2
94.2
95.3
96.7
96.2
95.7
97.1
96.7
95.6
97.2
96.9
96.3
97.3
97.0
5.8
4.7
3.3
3.8
4.3
2.9
3.3
4.4
2.8
3.1
3.7
2.7
3.0
be in harmony with those, which are obtained by IR
using Py as probe molecule.
The results obtained from the decomposition of
2-PrOH on yPAlO and yPAlH are cited in Table 3.
From these data, it is clear that the conversion of
2-PrOH has been increased by increasing the load-
ing level of phosphate, and both 6PAlO and 6PAlH
exhibit higher decomposition of 2-PrOH and higher
selectivity towards propene formation than the other
loadings (the lower w ¼ 3% and even the higher
w ¼ 10%). In addition, the dehydration selectivity
(formation of propene) is higher than for pure alu-
mina. Furthermore, the results cited in Table 3, show
that the dehydration of 2-PrOH on the surface of
yPAlH is higher than using yPAlO.
The same trend was noticed on the decomposition
of 2-PrOH on phosphated catalysts prepared from
phosphoric acid (yxPAlO and yxPAlH). From these
results, Table 3, it is also true for this series that the
maximum conversion of 2-PrOH and the maximum
selectivity towards propene formation are obtained
by 6xPAlO and 6xPAlH. At the same time, the
conversion of 2-PrOH and the selectivity towards
propene formation for yxPAlH is higher than for
yxPAlO. Knowing that 2-PrOH dehydrates to pro-
pene over acidic catalysts, this could explain such
behavior of the phosphated catalysts. It is worth re-
calling the results and discussion in the previous part
(Py adsorption) which revealed that the addition of
the phosphate ions to alumina had improved its acid-
ity. Such results explain the higher dehydration se-
lectivity of the phosphated catalysts compared to that
Catalytic Activity by 2-PrOH Conversion
The decomposition of 2-PrOH over the prepared ꢀ-
Al2O3 and phosphated aluminas has been carried out
at 520 K. The obtained results are listed in Table 3 in
terms of conversion, and acetone and propene selec-
tivity. The results obtained from decomposition of
2-PrOH on ꢀ-Al2O3 (Table 3) display mainly con-
version of 2-PrOH to propene, whereas the forma-
tion of acetone (minor product) is found to be very
low (5.8%). Table 3 indicated that the selectivity of
propene is 94.2%, thus, alumina surface has strong
acidic sites, which is evident from the dehydration
process being favored over dehydrogenation. Alu-
mina surface is known to acquire strong Lewis acid
sites that explain the high activity which promotes
the dehydration pathway [37]. Such results seem to

Phosphated Alumina Catalysts
647
metric (TG) curves were recorded up to 1070K at a heating
rate of 10ꢆ=min in static air. Small portions (15–20mg) of test
sample were used in TG measurements.
of pure alumina. Table 3 reveals that the 6% phos-
phated alumina catalysts exhibit higher activity and
selectivity regardless of the phosphate ion precursor.
Herein, it is worth mentioning that the surface acidity
first increases as the added phosphate ions increase,
reaches a maximum, and then decreases as previously
reported [34–36]. Consequently, one could suggest
that the phosphated catalysts under investigation of
the 6% loading level may possibly acquire higher
acidity (amount and strength) than either 3 or 10%
phosphated catalysts. Finally, one could suggest that
the 6% loading level would be used to prepare quite
a good acidic catalyst.
Nitrogen Adsorption Isotherm Measurement
Full nitrogen adsorption=desorption isotherms at 77K were
obtained using a NOVA 2200, version 6.10 high-speed gas
sorption analyzer (Quantachrome Corporation, USA). The cal-
cined samples were first outgassed at 470 K for 1 h. Twenty-
four-point adsorption and desorption isotherms were obtained,
from which BET surface areas were derived using standard
and well-established methods [41–44].
FT-IR Spectroscopy
IR spectra were obtained at a resolution of 4 cmꢁ1, in the
range of 4000–500 cmꢁ1, using a Genesis-II FT-IR spectro-
photometer, Mathson (USA). For pyridine (Py) adsorption
on test samples a wafer of 15–20mg=cm2 was mounted in a
Pyrex vacuum cell fitted with CaF2 windows. The samples
were pretreated at 700 K for 1 h in a stream of O2 followed
by evacuation at 700K for 1 h, and then cooled to room tem-
perature to obtain the background IR spectra. Then, 5 Torr Py
were admitted at various temperatures (300–570 K) for 5 min,
degassed for 5 min at each temperature in order to remove the
fraction of physisorbed Py, and the spectra were then taken at
room temperature.
Experimental
Materials
A pure alumina sample, denoted in the text as AlO, was ob-
tained from alumina gel, AlH, by calcination at 870 K for 3 h.
Aluminum hydroxide (AlH) was prepared according to Lippens
[38] from a 0.3 M aqueous solution of Al(NO3)3 ꢂ 9H2O.
Phosphated alumina catalysts were prepared by impregna-
tion method. Two series of phosphated aluminas were obtained;
the first series was prepared from AlH as a precursor and
the second from AlO (ꢀ-phase) as starting materials. The two
series were obtained by impregnation of an aqueous solution
of the impregnating with (NH4)2HPO4 or H3PO4 for 1 h under
stirring. The solution was adjusted to give w ¼ 3, 6, and 10%
Catalytic Activity (2-Propanol Decomposition)
The catalytic activity experiments for isopropanol (2-PrOH)
decomposition were carried out in a fluidized bed quartz flow
reactor at atmospheric pressure. 0.2 g of the catalyst were ac-
tivated in-situ at 670 K for 1 h in N2. 2-PrOH (Merck, 99.9%)
was introduced at a flow rate of 14.8cm3 minꢁ1 into carrier
gas flow of N2. The reaction products were analyzed by gas
chromatography on a 2 m long 1=800 column packed with 10%
Carbowax and Chromm WHP 80=100 using a model 3400
Varian Gas Chromatograph equipped with a flame ionization
detector (FID).
ꢁ3
PO4 content. The catalysts were obtained by calcination of
the dried samples at 870 K for 1 h. The amount loaded was
calculated to be similar to that needed to cover the entire
support surface in the case of a homogenous dispersion (area
2
ꢁ3
of PO4 ¼ 2.4 nm ) [34, 39] and two loadings were taken
around this figure; one of them is low and the other is high,
w ¼ 3 and 10% PO4ꢁ3, respectively.
The sample names used in this text include the source of
phosphate, the source of precursor, and the nominal phosphate
loading in w%. For example, yPAlO means phosphated alumi-
na with y% of phosphate (y ¼ 3, 6, or 10) and the source of
phosphate is (NH4)2HPO4 salt,ꢁb3ut yxPAlH means phosphated
alumina gel with w ¼ y% PO4 and the source of phosphate
is H3PO4 acid.
Acknowledgements
We acknowledge with appreciation the GC equipment
donation (V- 8151=87039; C=O Prof. M. I. Zaki of Minia
University) by the Alexander-von-Humboldt-Foundation,
Bonn=Germany.
Methods
X-Ray Diffraction
XRD diffractograms were recorded for all samples using a
model JSX-60PA JEOL diffractometer (Japan) using Cu Kꢂ
References
[1] Mekhemer GAH, Ismail HM (2000) Coll Surf A
Physicochem Eng Asp 164: 227
[2] Abbattista F, Delmastro A, Gozzelino G, Mazza D,
Vallino M, Busca G, Lorenzelli V (1990) J Chem Soc
Faraday Trans 86(1): 3653
˚
radiation (l ¼ 1.5418 A). The generator was operated at 35kV
andꢆ20 mA. For identification purposes, diffraction patterns
˚
(I=I ) versus d spacing (A) were matched with the relevant
ASTM standards [40].
[3] Mekhemer GA (1998) Coll Surf A Physicochem Eng
Asp 141: 227
[4] Morterra C, Magnacca G, De Maestri PP (1995) J Catal
152: 384
Thermal Analysis
Thermal analysis was performed using an automatically re-
cording DT-3OH Shimadzu apparatus (Japan). Thermogravi-

648
[5] Gishti K, Iannibello A, Marengo S, Morelli G, Tittarelli
P (1984) Appl Catal 12: 381
H. A. Khalaf et al.: Phosphated Alumina Catalysts
(eds) Proc. 9th Inter. Congr. Catal Calgary. Chem Instit
of Canada, Ottawa, p 1874
[6] Hess A, Kemnitz E (1997) Appl Catal A 149: 373
[7] Friedman P, Pinder KL (1971) Ind Eng Chem Process
Des Dev 10: 548
[28] Yang TS, Chang TH, Yeh CT (1997) J Mol Catal 115: 339
[29] Morterra A, Chiorino A, Ghiotti G, Garrone E (1979)
J Chem Soc Farady Trans I 75: 271
[8] Samantaray SK, Parida K (2001) J Mol Cat A Chemical
176: 151
[30] Morterra C, Coluccia S, Chiorino A, Boccuzzi F (1978)
J Catal 54: 348
[9] Krzywicki A, Marczewski M (1980) J Chem Soc
Faraday Trans 76(1): 1311
[31] Morterra C, Cerrato G, Emanuel C, Bolis V (1993) J
Catal 142: 349
[10] Olah GA, Prakash SK, Sommer J (1985) Superacids.
Wiley, New York
[11] Arata K (1990) Adv Catal 37: 165
[12] Yamaguchi T (1990) Appl Catal 61: 1
[13] Alexander FB, Kenneth JK (1998) J Catal 176: 448
[14] Tanabe K, Misono M, Ono Y, Hattori H (1989) New solid
acids and bases. Kodansha, Tokyo, Elsevier, Amsterdam
[15] Winterbottom JM (1981) Catalysis, Royal Society of
Chemistry. London, 4: 141
[32] Knøzinger H, Ratnasamy P (1978) Catal Rev Sci Eng
17: 31
[33] Corma A, Radellas C, Fornes V (1984) J Catal 88: 374
[34] Busca G, Ramis G, Lorenzelli V, Rossi PF, La Ginestra
A, Patrono P (1989) Langmuir 5: 911
[35] Berteau P, Kellens MA, Delmon B (1991) J Chem Soc
Faraday Trans 87(9): 1425
[36] Ramis G, Rosi PF, Busca G, Lorenzelli V, La Ginestra A,
Patrono P (1989) Langmuir 5: 917
[16] Ai M (1977) J Catal 49: 305
[37] Waqif M, Baachelier J, Saur O, Lavalley JC (1992) J Mol
Catal 72: 127
[38] Lippens BC (1961) PhD Thesis, Delft University, The
Netherlands
[17] Grybowiska-Swierkosz B (1987) Mater Chem 17: 121
[18] Campelo JM, Garcia A, Herencia JF, Luna D, Marinas
JM, Romero AA (1995) J Catal 151: 307
[19] Abdellah O, Coudurier G, Vedrine JC (1993) J Chem
Soc Farad Trans 89(16): 3151
[20] Mekhemer GAH, Nohman AKH, Fouad NE, Khalaf HA
(2000) Coll Surf A Physicochem Eng Asp 161: 439
[21] Mekhemer GAH, Khalaf HA, Mansour SAA, Nohman
AKH (2005) Monatsh Chem 136: 2007
[22] Nohman AKH, Mekhemer GAH, Fouad NE, Khalaf HA
(1999) Adsorption Sci Technol 17: 8
[23] Debye P (1915) Ann Physik 46: 809
[24] Ballinge TH, Yates T (1991) Langmuir 7: 3041
[25] Mekhemer GAH (1995) PhD Thesis, Minia University,
Egypt
[39] Moreno JA, Poncelet G (2001) J Catal 203: 453
[40] Frank W et al (eds) (1981) Powder Diffraction File for
Inorganic Phase. International Center for Diffraction
Data, Philadelphia PA
[41] Gregg SJ, Sing KSW (1982) Adsorption, Surface Area
and Porosity, 2nd edn. Academic Press, London
[42] Lecloux A (1981) Catalysis-Science and Technology. In:
Anderson JR, Boudaart M (eds) Springer-Verlag, Berlin,
Vol. 2
[43] Webb PA, Orr C (1997) Analytical methods in fine par-
ticle technology. Micromere tics, Norcross, GA, U.S.A.,
Chaps. 1–4
[26] Lewis JM, Kydd RA (1991) J Catal 132: 465
[27] Ramis G, Busca G, Lorenzelli V, Rossi PF, Bensitel M,
Saur O, Lavalley JG (1988) In: Phillips MJ, Ternan M
[44] Sing KSW, Everett DH, Haul RAW, Moscou L, Pierotti
RA, Rouquerol J, Siemieniewska T (1985) Pure Apple
Chem 57: 603