Synthesis and structural characterization of Pt/amorphous Al 2O3 catalyst

Article abstract of DOI:10.1016/S0021-9517(02)00022-2

The synthesis of Pt supported on nonhydrolytic derived amorphous alumina catalyst was studied. The Pt was introduced both by in situ gelation with the alumina precursor and by impregnation of the amorphous alumina matrix. The in situ precursors were PtBr₄ and (C₆H₅CN) ₂PtCl₂, while the impregnation precursors were H ₂PtCl₆ solution in water and Pt(acac)₂ solution in toluene. For comparison, samples of commercial γ-Al ₂O₃ were impregnated by the same methods. Complete fixation of the Pt precursor from the solution to the amorphous alumina support was achieved for both impregnation methods. After 773 K calcination in air the Pt dispersion, determined by H₂ pulse chemisorption, for the in situ gelation-based catalysts was 32%. The corresponding Pt dispersion of the impregnated catalysts, prepared with H₂PtCl₆ and Pt(acac)₂, was 15 and 49%, respectively. The low dispersion obtained using H₂PtCl₆ solution in water is explained by the partial burial of the Pt by a dissolution reprecipitation of the amorphous alumina and by a condensation reaction which occurs upon contact with water. These processes also cause reduction in the surface area of the amorphous alumina upon subsequent calcination.

Full text of DOI:10.1016/S0021-9517(02)00022-2

Journal of Catalysis 214 (2003) 146–152
www.elsevier.com/locate/jcat
Synthesis and structural characterization of Pt/amorphous Al2O3 catalyst
Alon Dolev, Gennady E. Shter, and Gideon S. Grader ∗
Chemical Engineering Department, Technion, Haifa 32000, Israel
Received 14 June 2002; revised 2 September 2002; accepted 16 September 2002
Abstract
The synthesis of Pt supported on nonhydrolytic derived amorphous alumina catalyst was studied. The Pt was introduced both by
in situ gelation with the alumina precursor and by impregnation of the amorphous alumina matrix. The in situ precursors were PtBr and
4
(C H CN) PtCl , while the impregnation precursors were H PtCl solution in water and Pt(acac) solution in toluene. For comparison,
6
5
2
2
2
6
2
samples of commercial γ -Al O were impregnated by the same methods. Complete fixation of the Pt precursor from the solution to the
2
3
amorphous alumina support was achieved for both impregnation methods. After 773 K calcination in air the Pt dispersion, determined by H
2
pulse chemisorption, for the in situ gelation-based catalysts was 32%. The corresponding Pt dispersion of the impregnated catalysts, prepared
with H PtCl and Pt(acac) , was 15 and 49%, respectively. The low dispersion obtained using H PtCl solution in water is explained by the
2
6
2
2
6
partial burial of the Pt by a dissolution reprecipitation of the amorphous alumina and by a condensation reaction which occurs upon contact
with water. These processes also cause reduction in the surface area of the amorphous alumina upon subsequent calcination.
2003 Elsevier Science (USA). All rights reserved.
Keywords: Amorphous; Alumina; Nonhydrolytic; Pt; Catalyst; Synthesis; Characterization; Dispersion; Impregnation; Gelation
1. Introduction
Pt/Al2O3 catalysts, with additional promoters to tune
the catalytic activity, are common in applications such as
naphtha reforming [14]. These catalysts are also common
in automobile catalytic converters. Utilization of amorphous
alumina for catalytic applications is very scarce and little
has been published regarding the synthesis of amorphous
Al2O3 and its utilization for catalytic applications. Although
the Pt/Al2O3 catalyst system has been studied extensively,
we found only one work dealing with Pt supported on
amorphous Al2O3 [15]. In that work the amorphous alumina
was impregnated by a methanolic solution of H2PtCl6;
the obtained catalyst was used to study the spillover of
hydrogen and oxygen onto amorphous alumina and was not
characterized.
The most common route for the introduction of Pt into
porous alumina is by impregnation with an aqueous solution
of H2PtCl6 precursor. Although the attachment mechanism
of the Pt precursor is not clear [16–21], the Pt precursor is
fixed to the alumina at discrete sites and therefore is initially
dispersed at the atomic level. Contact of the NH-amorphous
alumina with water causes morphology changes in the NH
alumina matrix [22,23]. It was found that the amorphous alu-
mina may undergo crystallization into pseudo-bohemite,
depending on water contact conditions, and upon further
calcination the aluminas may undergo crystallization into
Porous aluminas with large surface areas (SA) are widely
used in catalytic applications. In applications as isomeriza-
tion, hydrocracking, cyclization, and dehydration the alu-
mina serves as a bifunctional carrier or as the catalyst itself
and therefore the acidic sites on the alumina play a major
role in the catalytic activity [1]. The most common alumina
supports are γ and η phases, and the main route to obtaining
large SA aluminas is the hydrolytic sol–gel process [2,3].
In the past decade a new route for the production of
transition metal oxides by a nonhydrolytic (NH) sol–gel
process was developed [4–11]. The NH alumina retains high
SA and is in the amorphous state up to 1073–1123 K.
Additionally, the NH alumina also retains ∼ 2 wt.% Cl
up to the crystallization temperature. For applications such
as naphtha reforming Cl is added in order to enhance the
acidic function of the alumina [12–14]. Therefore the NH
alumina is a promising candidate as a bifunctional support
for applications such as hydrocarbon reforming. Moreover,
the utilization of this type of NH alumina as a catalyst
support has never been examined before.
*
Corresponding author.
E-mail address: grader@tx.technion.ac.il (G.S. Grader).
0021-9517/03/$ – see front matter
doi:10.1016/S0021-9517(02)00022-2
2003 Elsevier Science (USA). All rights reserved.

A. Dolev et al. / Journal of Catalysis 214 (2003) 146–152
147
γ -Al2O3 in the range 673–773 K. The SA of the alumina af-
2.2. Introduction of Pt
ter the second calcination also differs from the initial SA of
the amorphous alumina. Immersion in water at room temper-
ature (RT) for up to several days causes a decrease in final
SA, whereas longer immersion periods result in increased
final SA. An alternative route to obtaining well-dispersed
Pt on the alumina is by impregnating the alumina with a
solution of platinum acetylacetonate [Pt(acac)2] in organic
solvents such as toluene or benzene [24–32]. This chlorine-
free ligand exchange method does not contribute any addi-
tional acidity on the alumina carrier (which does occur when
chloroplatinic acid precursor is used) and does not involve
contact with water. Yet another approach for the introduc-
tion of the noble metal into the support is by introduction of
the metal during the gelation of the support by an adequate
precursor [33–40]. This one-step method, pioneered by the
group of Gomez and Lopez, may result in reduced metal sin-
tering upon thermal treatment [38,39] and in metal–support
interactions which may effect the catalytic activity [33–35].
The main drawback of this approach is that the metal might
be buried within the support and therefore be catalytically
inactive [38,39].
The objectives of this work were the synthesis of Pt/NH-
amorphous Al2O3 catalysts by several methods and their
characterization. The Pt was introduced either by impregna-
tion of the calcined alumina matrix or by in situ gelation with
the alumina precursor. The impregnation solutions were (a)
chloroplatinic acid in water and (b) Pt(acac)2 in toluene. For
comparison purposes commercial γ -aluminas were also im-
pregnated by the same techniques. The in situ gelation meth-
ods included the addition of an organic solution of either
PtBr4 or (C6H5CN)2PtCl2 to the alumina sol before gela-
tion. The catalysts prepared by all the methods above were
characterized in terms of Pt loading, Pt dispersion, and sur-
face area.
2.2.1. In situ introduction of Pt into NH alumina
2.2.1.1. Introduction with cis-bis(benzonitrile)platinum
dichloride, (C6H5CN)2PtCl2 Two hours after the addition
of the alumina precursors, as described above (before gela-
tion), the sol was cooled and a solution of 0.02 g/ml of
(C6H5CN)2PtCl2 in CH2Cl2 was added under argon. The
amount of (C6H5CN)2PtCl2 solution added was equivalent
to a final product of 1.0 wt.% Pt/Al2O3. The reactor was
placed again in an oil bath at 363 K for an additional 22 h for
gelation and aging. After aging the xerogels were dried at RT
under a mild vacuum of 25–40 N/m2. The catalyst was then
calcined at various temperatures for 3–5 h. The 1.0 wt.%
Pt content of the final catalysts was verified by dissolving
the catalyst in concentrated HNO3/HCl and analyzing the
Pt concentration in the solution using a PE-SCIEX ELAN
6000 ICP-MS unit.
2.2.1.2. Introduction with PtBr4 PtBr4 did not dissolve
in either Pri2O nor CH2Cl2, but was found to dissolve in
a Pri2O/CH2Cl2 mixture. However, the time required for
the dissolution was several days. Therefore a stock so-
lution of 0.05 g PtBr4/1 ml Pri2O/2.5 ml CH2Cl2 was
used and the procedure was the same as for introduc-
tion by (C6H5CN)2PtCl2 except that PtBr4 solution in
Pri2O/CH2Cl2 was used instead of (C6H5CN)2PtCl2 solu-
tion in CH2Cl2.
2.2.2. Introduction of Pt by impregnation
For the catalysts prepared by impregnation the above NH
alumina and two commercial γ -Al2O3 of Alfa Aesar, termed
below γ 1 and γ 2, were utilized, γ 1 and γ 2 possess SA of
170 and 227 m2/g, respectively. Prior to the impregnation
the NH and γ 2 aluminas were ground to ∼ 50 µm (γ 1 is in
the form of ∼ 250 µm powder).
2.2.2.1. Impregnation by H2PtCl6 solution in water To
each gram of Al2O3, 50 ml of a 100 or 200 µg Pt/ml,
H2PtCl6 solution in water was added to obtain 0.5 or
1.0 wt.% Pt/Al2Cl3 catalysts, respectively. The mixture was
stirred for 4 h, after which the catalyst was filtered. The
catalyst was then dried at 383 K under vacuum and then
calcined at 773 K for 3 h. The liquid phase was analyzed
for Pt traces by a UNICAM. UV/V is spectrometer operated
at 262 nm or with a VARIAN 400 atomic absorption
instrument.
2. Experimental
2.1. Preparation of nonhydrolytic amorphous alumina
(NH)
AlCl3 (Fluka, 99%), Pri2O (Riedel-deHaen, 99%), and
CH2Cl2 (Carlo Erba, 99.8%) were added into a reactor under
argon in a ratio of 1 g/1.6 ml/4.2 ml, respectively, as in [22].
After dissolution the reactor was immersed in an oil bath at
363 K for 24 h for gelation and aging. The average gelation
time was 6.5 h with a standard deviation of 0.5 h. After aging
the xerogels were dried at RT (293–298 K) under a mild
vacuum of 25–40 N/m2.
The fresh xerogels were calcined in dry air at 923 K for
5 h, to obtain the nonhydrolytic amorphous alumina. The
heating rate was 40 K/h up to 673 K and then 70 K/h up
to 923 K. The average SA of NH alumina was measured as
274 m2/g.
2.2.2.2. Impregnation by Pt(acac)2 solution in toluene The
alumina was initially dehydrated under dry air at 773 K
and cooled under the same atmosphere. The alumina was
then added to 125 or 250 µg Pt/ml, Pt(acac)2 solution in
toluene in a ratio of 40 ml/g alumina to yield 0.5 or 1.0 wt.%
Pt/Al2O3 catalysts, respectively. The mixture was stirred
for 4 h, after which the catalyst was filtered. The catalyst
was then dried at 383 K under vacuum and then calcined at

148
A. Dolev et al. / Journal of Catalysis 214 (2003) 146–152
773 K for 3 h. The liquid phase was analyzed for Pt traces
by a VARIAN 400 atomic absorption instrument. In order
to avoid interference caused by toluene, before the atomic
absorption the toluene was exchanged with methyl isobutyl
ketone.
2.3. Characterization of the catalysts
Thermal behavior of the xerogels was studied by DTA/
TGA measured in a Setaram TG-92 unit, in flowing air,
with a heating rate of 5 K/min, and the composition of the
effluent gas was detected using a mass spectrometer (Balzers
Thermostar GSD300T).
BET specific surface areas were determined by nitrogen
desorption using a Quantachrome Monosorb. XRD patterns
were recorded using a Siemens 5000 instrument employing
Cu-Kα radiation. The catalysts were observed by a JEOL
STEM FX-2000 operated at 200 kV and equipped with a
LINK AN 10000 EDS system. TPR and H2 pulse chemisorp-
tions measurements were held on a Zeton Altamira AMI200
instrument. The TPR was held using 10% H2/Ar gas and
a heating rate of 10 K/min up to 673 K. For the H2 pulse
chemisorptions, samples which were reduced by 10% H2/Ar
gas at 673 K and then saturated by oxygen using a 10%
O2/He stream at 308 K were used. The number of 10%
H2/Ar pulses into Ar which were consumed by the sample
was recorded. The surface reaction [41]
Fig. 1. TGA and DTA curves in the range 300–1250 K for pure NH-alumina
xerogel (a,d), NH alumina with in situ (C H CN) PtCl xerogel (b,e), and
6
5
2
2
NH alumina with in situ PtBr xerogel (c,f ). Heatings were done in air at
4
5 K/min.
without Pt is in the sharp ending of the organic residues
burnout at lower temperatures of 740 and 770 K using
PtBr4 and (C6H5CN)2PtCl2 solutions, respectively. In [42]
the lower burning temperature of organic residues of co-
gelled Pd/SiO2 vs the higher burning temperature of Pd-
free samples was explained by the oxidation power of Pd.
TGAs of PtBr4 and (C6H5CN)2PtCl2 in air (not shown)
revealed that the decomposition to metallic Pt is over at 748
and 763 K, respectively. The mechanism of this platinum
reduction is not clear and the Pt complex fixed to the
fresh xerogel may be different from the original complex.
However, it is logical to assume that at temperatures a little
lower than 773 K, the Pt in the alumina is converted to
the metallic state, and then catalyzes the burnout of the
organic and coke residues at much lower temperatures than
in the xerogels without Pt. This explanation, which is based
on formation of metallic Pt, is supported by the fact that
the XRD of samples in which Pt was introduced in situ,
and then calcined in air at 873 K, revealed the presence of
crystalline Pt (Fig. 2). The intensity of the reduction peaks
observed in the H2 TPR patterns of catalysts which were
previously calcined at 873 K (not shown) correspond to
∼ 25% reduction, which also suggests that part of the Pt was
converted to the metallic state during the calcination in air.
3
2
Pt–O + H2 → Pt–H + H2O
was assumed to calculate the Pt dispersion.
3. Results and discussion
3.1. Catalysts prepared by in situ introduction of Pt into
NH alumina
3.1.1. Effect of in situ Pt introduction into NH alumina on
thermal behavior
The thermal behavior of the fresh xerogel with and
without the Pt upon treatment in air was examined by
TGA/DTA, and is shown in Fig. 1. For the xerogel without
Pt (see a and d), below 470 K the weight loss was due
to evaporation of organic residues from the xerogel. This
region is accompanied by a small endotherm observed in
the DTA (d). For the Pt-free xerogels, the weight loss
between 470 and 870 K is due to oxidation of the organic
residues. This segment is accompanied by two exothermic
peaks observed in the DTA with maximums at 570 and
770 K. The oxidation of the organic residues is completed by
∼ 870 K as seen by the TGA and DTA curves. At a higher
temperature of ∼ 1070 K an exothermic peak is observed
in the DTA. This peak is attributed to crystallization into
γ -Al2O3. The main difference between the TGA/DTA
patterns of fresh xerogels with in situ introduction of Pt
(b, c, e, f) from the TGA/DTA curves of fresh xerogels
3.1.2. SA of calcined catalysts
The average SA of catalysts calcined at 773 and 873 K in
air is summarized in Table 1. As seen in Table 1, there is a
significant decrease in SA for the pure NH alumina samples
from 420 to 320 m2/g when the calcination temperature is
increased from 773 to 873 K. For the platinum-containing
samples the SA remains constant in the same temperature
range (∼ 330 m2/g), similarly to the SA of pure NH alumina
calcined at 873 K. The lower SA of the Pt-containing

A. Dolev et al. / Journal of Catalysis 214 (2003) 146–152
149
Fig. 2. XRD patterns of NH alumina with in situ, introduction of
(C H CN) PtCl and after heat treatments at (a) 773 K air; (b) 773 K air;
6
5
2
2
673 K H ; (c) 873 K air, and of NH alumina with in situ introduction of
2
PtBr and after heat treatments at (d) 773 K air, H 673 K; (e) 873 K air;
(f ) 1023 K air.
4
2
Fig. 3. TEM micrographs of NH alumina with in situ introduction of
(C H CN) PtCl after heat treatment at 873 K under air.
6
5
2
2
samples compared to the NH alumina after calcination at
773 K, may be due to differences caused by the addition of
the organic solvents and/or the Pt precursor during gelation,
which causes the evolution of different microstructures.
Another possible explanation for the reduced SA of the Pt-
containing samples is the catalyzed organic burn-off. The
catalyzed burn-off causes a local increase in the temperature
within the catalyst pellets, which is observed in the DTA
curves (Fig. 1 e, f). This local increased temperature results
in higher actual calcination temperature and therefore causes
lower SA.
of the fact that TEM and the XRD pattern suggest that
the Pt is well dispersed, the Pt might be buried as very
small crystallites within the alumina matrix. These small
crystallites, which will not contribute to the catalytic activity,
due to inaccessibility of the gas phase, are not observed
in the TEM or in the XRD pattern. However, H2 pulse
chemisorption over reduced samples, which is only sensitive
to surface atoms, shows Pt dispersion of ca. 30%, which
means that significant portion of the Pt atoms are exposed
at the surface. We therefore conclude that even if the Pt
precursor is buried within the xerogel during the gelation
process, a major portion of Pt atoms are accessible surface
atoms after thermal treatment of the catalyst.
3.1.3. The effect of thermal treatment on Pt accessibility
and dispersion
Upon higher temperature treatment in air (T = 873 K),
there is extensive sintering of Pt, as can be seen by the
appearance of Pt peaks in the XRD pattern (Fig. 2 c, e).
From the width of the 2θ peak centered at ∼ 40◦, and using
Scherrer’s equation [44] and assuming spherical shape of
Pt crystals, the average Pt size is ∼ 25 nm. This particle
size is slightly larger than the Pt particles observed in the
TEM micrographs (Fig. 3) as dark features of 10–20 nm.
The dark features in the TEM micrographs were proved
to be Pt by EDS in the STEM mode. The larger particle
size obtained by XRLBA is probably due to the fact that
the XRD line-broadening is a volume-average particle size,
which is larger than the sample mean diameter. Further H2
treatment, at 723 K, of samples previously calcined under
air at T ꢀ 873 K did not significantly affect the intensity
of the Pt peaks nor the line broadening of the Pt peaks,
which suggests that the sintering into large Pt crystals occurs
during the high-temperature air treatment and not during
subsequent H2 treatment. Increasing the air calcination
temperature to 1023 K resulted in increased intensity of the
Pt peaks (almost double the intensity obtained at 873 K)
The effect of heat treatments on Pt dispersion (and
particle size) was examined by XRD (Fig. 2), by TEM
(Fig. 3), and by H2 pulse chemisorption (Table 1). In samples
which were calcined under air at temperatures lower then
773 K the Pt is well dispersed, as can be seen from the
absence of Pt peaks from the XRD pattern of the H2 reduced
sample (Fig. 2 b, d), which suggest that Pt crystallites
are smaller than 5 nm [43]. In the TEM micrographs of
samples calcined at 773 K (not shown) Pt particles are not
detectable, which is consistent with the XRD data. In spite
Table 1
Surface area and Pt dispersion—in situ introduction of Pt
2
Pt precursor
Surface area (m /g)
Pt dispersion from pulse
chemisorption (%)
773 K
873 K
773 K
873 K
calcination calcination
calcination calcination
None (pure Al O )
420
320
350
320
330
330
–
32
32
–
13
7
2
3
PtBr
4
(C H CN) PtCl
2
6
5
2

150
A. Dolev et al. / Journal of Catalysis 214 (2003) 146–152
and a further increase of the mean particle size to ∼ 30 nm.
The increase in the Pt peaks’ intensity with increased
calcination temperature suggests that not all of the Pt in
samples calcined at 873 K is in the form of large metallic
Pt crystals, which are detected by XRD. The Pt which was
not detected by XRD might be in the form of small crystals
or of a noncrystalline species. Upon increased temperature
air treatment (1023 K), this noncrystalline/well-dispersed
platinum sinters into large Pt crystals. Support for the
existence of small Pt crystals not detected by XRD after
873 K calcination comes from comparing the dispersions
obtained by XRD to the dispersions obtained by the pulse
chemisorption. The mean particle size of 25 nm obtained
by XRD corresponds to ∼ 4% dispersion. The dispersions
obtained by pulse chemisorption (Table 1) are, however,
higher (7–13%), which suggests the existence of small Pt
crystals which are not detected by XRD but do contribute
most of the Pt dispersion.
alumina, which increases the support acidity, or the higher
SA of the NH alumina. The Pt loading NH > γ 2 > γ 1 is in
the same order as the SA of these aluminas, which is in line
with the second explanation.
3.2.2. SA
After calcination at 773 K all catalysts regained the
surface area of the original alumina supports; i.e., the
aluminas do not undergo morphological changes during
contact with the toluenic solution of Pt(acac)2. As will
be discussed later, the SA of the NH alumina changes
significantly during impregnation with aqueous solution.
3.2.3. Pt dispersion
The Pt dispersion was evaluated after calcination at
773 K. In the TEM micrographs no Pt particles were ob-
served, which is consistent with the good Pt dispersions mea-
sured by the H2 pulse chemisorptions (Table 2). Increasing
the Pt loading did not affect the Pt dispersion for the γ 1, γ 2
catalysts. However, for the NH alumina, there is a decrease
in the Pt dispersion with increase of Pt uptake. This may
suggest that the enhanced Pt adsorption on the NH alumina
is on weak adsorption sites, which release the Pt rather read-
ily upon thermal treatment, resulting in sintering and lower
Pt dispersion.
3.2. Catalysts impregnated by Pt(acac)2 solution in toluene
3.2.1. Pt uptake during impregnation
The Pt uptake by the catalyst was calculated by a mass
balance on the Pt content in the liquid phase (detected by
atomic absorption) before and after contact with the alumina
and is shown in Table 2. For catalysts impregnated with
a solution containing Pt in amounts adequate to 0.5 wt.%
Pt/Al2O3, almost all of the Pt is fixed on the alumina
surface. According to [29,30] this Pt is fixed by the following
surface reaction:
3.3. Catalysts impregnated by H2PtCl6 solution in water
3.3.1. Pt uptake during impregnation
As in the previous section, the Pt uptake by the catalyst
was calculated by a mass balance on the Pt content in the liq-
uid phase (detected by atomic absorption or by spectropho-
tometry) before and after contact with the alumina and is
shown in Table 3. For all three aluminas and for the nomi-
nal loading of 0.5 and 1.0 wt.%, more than 98% uptake of
the Pt from the water was obtained. The nature of Pt uptake
from aqueous H2PtCl6 solution, although studied for many
decades, is not fully understood [16–21], however, the full
Pt uptake for all aluminas may suggest that the fixation of Pt
is of the same nature for the NH amorphous alumina and the
γ -aluminas.
Surface–OH + Pt(acac)2
⇐⇒ surface–O–Pt(acac)+ Acacadsorbed
.
However, for catalysts impregnated with a solution contain-
ing Pt in amounts adequate to 1.0 wt.% Pt/Al2O3, there is
a significant amount of Pt residue in the solution after con-
tact with γ 1, γ 2 aluminas, while the NH alumina practi-
cally fixes all of the Pt. This means that the NH alumina
possesses more sites that can adsorb the Pt(acac)2 complex.
The Pt(acac)2 fixation reaction with the support is an acidic
reaction. The origin of the higher number of fixation sites on
the NH alumina may be the high chlorine content in the NH
3.3.2. SA
After calcination in air at 773 K the catalysts obtained
from impregnation of the commercial γ 1, γ 2 aluminas
regained the original surface area (Table 3). However, the
catalysts obtained by impregnation of the NH amorphous
alumina followed by calcination at 773 K had a significantly
lower surface area (175 m2/g) than the original NH support
(274 m2/g). In a previous study [22], it was been found that
upon contact of the NH amorphous alumina with water, at
RT, for periods of up to several days and subsequent thermal
treatment the SA of the alumina decreases. The suggested
explanation for the decrease in SA was a condensation
reaction, which may block some of the pores, resulting in
lower SA.
Table 2
SA, Pt uptake and dispersion—impregnation by Pt(acac) /toluene
2
Alumina Amount of Pt in Actual Pt Surface
Pt dispersion
from pulse
a
initial solution
(wt.%)
loading
(wt.% )
area
2
b
(m /g) chemisorption (%)
NH
γ 1
γ 2
0.5
1.0
0.46
0.96
267
252
49
28
0.5
1.0
0.41
0.57
180
176
37
37
0.5
1.0
0.45
0.71
219
213
43
39
a
After calcination at 773 K.
Based on actual Pt loading.
b

A. Dolev et al. / Journal of Catalysis 214 (2003) 146–152
151
Table 3
PtBr4 or (C6H5CN)2PtCl2, are exposed at the surface
after calcination at 773 K. Increasing the calcination
temperature to 873 K results in sintering of part of the
Pt to crystals of an average size of ∼ 20 nm.
SA, Pt uptake and dispersion—impregnation by H PtCl /water
2
6
Alumina
Amount of Pt in Actual Pt Surface
a
Pt dispersion
from pulse
initial solution
(wt.%)
loading
(wt.% )
area
2
(m /g)
chemisorption (%)
(2) The NH alumina fixes Pt from Pt(acac)2 solution in
toluene more readily than the γ -aluminas. Upon calci-
nation at 773 K of 0.5 wt.% Pt/alumina catalysts the
obtained Pt dispersions are ∼ 50% for the NH alumina
and ∼ 40% for the γ -aluminas. Increasing the Pt load-
ing does not affect the dispersion obtained with both γ -
aluminas, whereas the sintering of the Pt leading to a
lower dispersion is more pronounced for the NH alu-
mina.
NH
0.5
1.0
0.5
0.99
177
173
15
21
γ 1
0.5
1.0
0.5
0.99
189
178
45
39
γ 2
0.5
1.0
0.49
0.99
227
213
41
39
a
After calcination at 773 K.
3.3.3. Pt dispersion
(3) The NH and the γ aluminas fixe Pt from aqueous
H2PtCl6 solution. Complete Pt uptake from the water
was obtained for the 1.0 wt.% catalysts. However, the
dispersion obtained after calcination at 773 K of the
Pt/NH-alumina (∼ 20%) is inferior to the dispersion ob-
tained with γ -aluminas (∼ 40%). The NH alumina also
suffers a decrease in SA upon contact with water. These
findings are explained by a condensation reaction, which
may block some of the pores, and by the dissolution–
reprecipitation process of the NH alumina which buries
the Pt.
The Pt dispersion was evaluated after calcination at
773 K. In the TEM micrographs no Pt particles were ob-
served and in the XRD pattern of these catalysts after re-
duction in H2 at 673 K no Pt peaks are observed, which
is consistent with the good Pt dispersions measured by the
H2 pulse chemisorptions for the γ 1, γ 2 aluminas (Table 3).
For the NH alumina, however, the ∼ 20% dispersion cor-
responds to crystallites of average size 5–6 nm, which are
expected to be observed in TEM and to cause the appear-
ance of wide Pt peaks in the XRD. However, no Pt crystals
were observed in the TEM, and no traces of Pt peaks were
observed in the XRD pattern of the Pt/NH-alumina samples.
We therefore conclude that the Pt is spread as small species
(not detectable by TEM or XRD), which are partially buried
within the alumina matrix. The mechanisms by which the Pt
may be buried within the alumina are:
The effect of the synthesis method and the type of alu-
mina used on catalytic activity is currently under investiga-
tion.
Acknowledgments
(a) The condensation reaction, described above, which may
block some of the pores.
The authors thank the Israel Science Foundation, the
Technion Grand Water Institute, the Technion Catalysis
Project, and the Center for Adsorption in Science of the
Ministry of Immigrant Absorption (under the framework
of the KAMEA program) for their support. In addition,
the authors thank Dr. Michael Reisner for help in X-ray
measurements.
(b) Upon contact with water the NH amorphous alumina
undergoes a dissolution–reprecipitation process [22].
At RT this process takes up to several weeks. The
adsorption of Pt from the water is over at a much shorter
time. We therefore suggest that after the Pt complex is
fixed to the alumina from the water it may be buried by
the dissolution–reprecipitation of the NH alumina. As a
result of this process the Pt is spread as small species
partially buried within the alumina matrix.
References
[1] R.K. Oberlander, in: B.E. Leach (Ed.), Applied Industrial Catalysis,
Vol. 3, Academic Press, London, 1984, p. 63.
4. Summary
[2] B.E. Yoldas, J. Appl. Chem. Biotechnol. 23 (1973) 803.
[3] C.J. Brinker, G.W. Scherer, Sol–Gel Science: The Physics and
Chemistry of Sol–Gel Processing, Academic Press, New York, 1990.
[4] Y. De Hazan, G.E. Shter, Y. Cohen, C. Rottman, D. Avnir, G.S. Grader,
J. Sol–Gel Sci. Technol. 14 (3) (1999) 233.
[5] G.S. Grader, Y. De Hazan, D.B. Zhivotovskii, G.E. Shter, J. Sol–Gel
Sci. Technol. 10 (1997) 127.
[6] S. Acosta, R. Corriu, D. Leclercq, P.H. Mutin, A. Vioux, J. Sol–Gel
Sci. Technol. 2 (1994) 25.
[7] S. Acosta, P. Arnal, R.J.P. Corriu, D. Leclercq, P.H. Mutin, A. Vioux,
in: Better Ceramics through Chemistry VI, in: Mater. Res. Soc. Symp.
Proc., Vol. 346, 1994, p. 43.
[8] S. Acosta, R.J.P. Corriu, D. Leclercq, P. Lefevre, P.H. Mutin, A. Vioux,
J. Non-Cryst. Solids 170 (1994) 234.
The synthesis of Pt/NH–amorphous-alumina catalysts
by (a) in situ introduction of PtBr4 or (C6H5CN)2PtCl2,
during gelation of the NH alumina, (b) by impregnation with
Pt(acac)2 solution in toluene, or (c) by impregnation with
H2PtCl6 solution in water was investigated. Pt/γ -alumina
catalysts prepared by the same impregnation methods were
used as a reference for comparison. The main findings are:
(1) About 30% of the Pt atoms of Pt/NH–amorphous-
alumina catalysts prepared by in situ introduction of

152
A. Dolev et al. / Journal of Catalysis 214 (2003) 146–152
[9] M. Andrianainarivelo, R.J.P. Corriu, D. Leclercq, P.H. Mutin, A.
[26] L.C. de Menorval, A. Chaqroune, B. Coq, F. Figueras, J. Chem. Soc.
Faraday Trans. 93 (20) (1997) 3715.
Vioux, J. Mater. Chem. 7 (2) (1997) 279.
[27] Z. Gandao, B. Coq, L.C. de Menorval, D. Tichit, Appl. Catal. A 147 (2)
(1996) 395.
[10] M. Andrianainarivelo, R.J.P. Corriu, D. Leclercq, P.H. Mutin, A.
Vioux, Chem. Mater. 9 (5) (1997) 1098.
[28] P. Reyes, G. Pecchi, M. Morales, J.L.G. Fierro, Appl. Catal. A 163
(1997) 145.
[11] M. Andrianainarivelo, R. Corriu, D. Leclercq, P.H. Mutin, A. Vioux,
J. Mater. Chem. 6 (10) (1996) 1665.
[29] J. Berdala, E. Freund, J.P. Lynch, J. Phys. Colloq. C 8 (1) (1986)
C8/265.
[30] J.P. Boitiaux, J. Cosyns, S. Vasudevan, in: G. Poncelet, P. Grange, P.A.
Jacobs (Eds.), Preparation of Catalysts III, Elsevier, Amsterdam, 1983,
p. 123.
[31] B. Coq, A. Tijani, F. Figueras, J. Mol. Catal. 71 (3) (1992) 317.
[32] B. Coq, A. Chaqroune, F. Figueras, B. Nciri, Appl. Catal. A 82 (2)
(1992) 231.
[33] T. Lopez, M. Asomoza, P. Bosch, E.G. Figueroa, R. Gomez,
J. Catal. 138 (1992) 463.
[34] T. Lopez, P. Bosch, M. Moran, R. Gomez, J. Phys. Chem. 97 (1993)
1671.
[35] T. Lopez, R. Gomez, J. Navarrete, R.D. Gonzalez, J. Mater. Synth.
Process. 2 (1994) 305.
[36] O. Bokhimi, A. Aceves, O. Novaro, T. Lopez, R. Gomez, J. Phys.
Chem. 99 (1995) 14403.
[37] E. Sanchez, T. Lopez, R. Gomez, O. Bokhimi, A. Morales, O. Novaro,
J. Solid State Chem. 122 (1996) 309.
[38] S. Castillo, M.M. Pineda, V. Molina, R. Gomez, T. Lopez, Appl. Catal.
B 15 (1998) 203.
[39] I.H. Cho, S.B. Park, S.J. Cho, R. Ryoo, J. Catal. 173 (1998) 295.
[40] C. Gorsmann, U. Schubert, J. Leyrer, E. Lox, Mater. Res. Soc. Symp.
Proc. 435 (1996) 625.
[41] J.E. Benson, M. Boudart, J. Catal 4 (1965) 704.
[42] B. Heinrichs, F. Noville, J.P. Pirard, J. Catal. 170 (1997) 366.
[43] T.A. Dorling, R.L. Moss, J. Catal. 7 (1967) 378.
[12] G.E. Costa, N.S. Figoli, M.R. Sad, G.T.O. Zwiener, L.M. Krasnogor,
J.M. Parera, React. Kinet. Catal. Lett. 20 (12) (1982) 171.
[13] J. Hancsok, G. Gardos, J. Perger, Hung. J. Ind. Chem. 15 (3) (1987)
309.
[14] J.P. Boitaux, J.M. Deves, B. Didillon, C.R. Marcilly, in: G.J. Antos,
A.M. Aitani, J.M. Parera (Eds.), Catalytic Naphtha Reforming Science
and Technology, Dekker, New York, 1995, p. 79.
[15] D. Maret, G.M. Pajonk, S.J. Teichner, in: G.M. Pajonk, S.J. Teichner,
G.E. Germain (Eds.), Spillover of Adsorbed Species, Elsevier, Ams-
terdam, 1983, p. 215.
[16] R.D. Gonzalez, H. Miura, Catal. Rev. Sci. Eng. 36 (1) (1994) 145.
[17] J.P. Brunelle, Pure Appl. Chem. 50 (1978) 1211.
[18] J.C. Summers, S.A. Ausen, J. Catal. 52 (1978) 445.
[19] J.R. Regalbuto, A. Navada, S. Shadid, M.L. Bricker, Q. Chen,
J. Catal. 184 (1999) 335.
[20] B.N. Shelimov, J.F. Lambert, M. Che, B. Didillon, J. Mol. Catal. 158
(2000) 91.
[21] B. Shelimov, J.F. Lambert, M. Che, B. Didillon, J. Am. Chem.
Soc. 121 (3) (1999) 545.
[22] G.S. Grader, G.E. Shter, D. Avnir, H. Frenkel, D. Sclar, A. Dolev,
J. Sol–Gel Sci. Technol. 21 (3) (2001) 157.
[23] F. Abbattista, S. Delmastro, G. Gozzelino, D. Mazza, M. Vallino, G.
Busca, V. Lorenzelli, G. Ramis, J. Catal. 117 (1989) 42.
[24] B. Coq, E. Crabb, M. Warawdekar, G.C. Bond, J.C. Slaa, S. Galvagno,
L. Mercadante, J.G. Ruiz, M.C.S. Sierra, J. Mol. Catal. 92 (1) (1994)
107.
[44] J.L. Lemaitre, P.G. Menon, F. Delannay, in: F. Dellanay (Ed.),
Characterization of Heterogeneous Catalysts, Dekker, New York,
1984, p. 299.
[25] C.G. Walter, B. Coq, F. Figueras, M. Boulet, Appl. Catal. A 133 (1)
(1995) 95.