You are on page 1of 11

Available online at www.sciencedirect.

com

Microporous and Mesoporous Materials 111 (2008) 323–333


www.elsevier.com/locate/micromeso

Synthesis, characterization and catalytic applications of mesoporous


c-alumina from boehmite sol
a,b
Qian Liu , Aiqin Wang a, Xuehai Wang a,b
, Peng Gao a,b
, Xiaodong Wang a, Tao Zhang a,*

a
State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, P.O. Box 110, Dalian 116023, China
b
Graduate School of the Chinese Academy of Sciences, Beijing, China

Received 24 July 2006; received in revised form 22 November 2006; accepted 10 August 2007
Available online 17 August 2007

Abstract

In this study, boehmite sols were used as aluminum precursors for preparing mesoporous alumina (MA) having crystalline framework
walls in the presence of non-ionic surfactants as structure directing agents. Nitrogen physisorption showed that aluminas prepared in this
way displayed very rich porosities with large mesopores, and both the pore volumes and the pore sizes increased with the surfactant con-
centration. The improved textural parameters in the samples should be attributed to the three-dimensional interconnected scaffold-like
channels, which were formed by randomly ordered stacking and condensing of rigid boehmite nanoparticles with the aid of the surfac-
tant. TEM observations revealed that the precursor morphology had an important effect on the textural properties of the mesoporous
alumina. The sample with a corrugated platelet-like morphology exhibited a large surface area of 463 m2/g, which was reduced to 81 m2/g
after calcination at 1200 °C, indicating a strong resistance to sintering. This material, with its improved textural properties, crystalline
framework walls and high thermal stability, not only could increase the dispersion of the active catalytic species, but also could enhance
the diffusion efficiency and mass transfer of reactant molecules when employed as catalyst supports. As examples, our MA samples dem-
onstrated a remarkable enhancement in the catalytic performances for both reactions of SO2 catalytic reduction by CO and catalytic
combustion of methane.
Ó 2007 Elsevier Inc. All rights reserved.

Keywords: Mesoporous c-alumina; Boehmite sol; Catalyst support; SO2 reduction by CO; Methane combustion

1. Introduction surface acidic–basic properties can often result in favorable


enhancements in the catalytic performances [2,3]. There-
Aluminas are extensively used as catalyst supports due fore, synthesis of mesoporous aluminas (MA) with high
to their favorable textural properties and intrinsic acid– surface areas and uniform mesopores has attracted much
base characteristics. In particular, c-alumina, which has a attention [3–8]. Many synthesis routes have been developed
large surface area and a crystalline structure, is an impor- for the preparing of MA. Among them, organic–inorganic
tant catalyst support in automotive and petroleum indus- assemblies involving complicated sol–gel processes by
tries [1]. The catalytic performances of alumina-supported using surfactants as structure-directing agents are regarded
catalysts are largely dependent on the textural properties as one of the most promising approaches [3–6]. Various
of the alumina supports. Alumina supports with large sur- neutral and ionic surfactants have been used as templates
face areas, large pore volumes, narrow pore size distribu- for the preparation of MA.
tions within the mesoporous range, as well as suitable Differing from ordered mesoporous silicas [9], many MA
materials prepared via surfactant templates are lacking in
*
Corresponding author. Tel.: +86 411 8437 9015; fax: +86 411 8469
long-range order, and the inorganic walls are often amor-
1570. phous [3,5]. Comparing with amorphous aluminas, crystal-
E-mail address: taozhang@dicp.ac.cn (T. Zhang). line aluminas can afford special surface acid–base

1387-1811/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.micromeso.2007.08.007
324 Q. Liu et al. / Microporous and Mesoporous Materials 111 (2008) 323–333

characteristics and proper aluminum coordinations which dea Chemie Co.) was dispersed in 200 ml of deionized
are very important for their applications in catalysis [1,4]. water at 80 °C. Then 1 M nitric acid was added into the
Recently, Pinnavaia and co-workers have reported a three- mixture with the molar ratio of [H+]/[Al3+] = 0.3–3.0,
step assembly pathway to obtain MA composed of and the mixture was kept under stirring at 80 °C for at least
crystalline and lath-like c-Al2O3 nanoparticles [4]. Since 6 h to obtain a boehmite sol of 1 mol/l, which is denoted as
then, several similar routes have been reported for preparing the BP sol.
mesoporous c-Al2O3 [10,11]. In their synthesis procedures, In another procedure, the boehmite sol was prepared by
expensive and toxic aluminum alkoxides were usually used peptizing aluminum hydroxide precipitate with nitric acid,
as the aluminum precursors, and an additional hydrother- which is denoted as the AN sol. In detail, 75.03 g of
mal treatment was necessary for transforming the as-synthe- Al(NO3)3 Æ 9H2O was dissolved in 100 ml of deionized
sized amorphous walls to a surfactant/boehmite mesophase. water. Then 2.5% ammonia was dropwise added into the
Motivated by the fact that heating the boehmite phase solution until pH > 3.8. After that, 350 ml of ammonia
to above 450 °C can result in the formation of c-Al2O3, was poured into the solution and a white precipitate was
we have developed a facile and economic route for prepar- formed immediately. The precipitate was separated by cen-
ing mesoporous c-Al2O3 in our previous work [8] by using trifugation, washed several times with deionized water, and
boehmite sol as an inorganic precursor and hydro-carbox- then dispersed in 200 ml of deionized water and peptized
ylic acid as the structure directing agent. In that synthesis, with 1 M nitric acid at 80 °C under vigorous stirring for
the pore structures could be tuned by adjusting the coordi- 4 h to obtain a stable boehmite sol of 1 mol/l.
nation interaction between the hydro-carboxylic acid and
the boehmite particulates.
2.1.2. Synthesis of the MA materials
In the present work, we took a new step to synthesize mes-
A desired amount of surfactant EO20PO70EO20 (P123,
oporous c-Al2O3, in which the boehmite sol was used as the
Aldrich) was dissolved in the boehmite sol at room temper-
precursor and a non-ionic surfactant as the template. Then,
ature and then aged for 3 h. The resulting mixture was
by simple drying and calcining the mixture of the boehmite
dried at 110 °C overnight to form the as-synthesized sam-
sol and the non-ionic surfactant, mesoporous c-Al2O3 with
ple. For comparison, samples drying at 30 °C and 150 °C
large surface areas and rich porosities could be obtained.
were also investigated. The mesoporous products were
Preliminary catalytic tests showed that the c-Al2O3 materi-
obtained by calcining the as-synthesized samples at
als prepared in such a way exhibited superior performance
500 °C for 4 h in air, with b = 1 °C/min. The detailed prep-
in SO2 catalytic reduction by CO when they were used as
aration conditions and the corresponding sample I.D. are
the supports for Fe catalysts. It was also found that the sam-
listed in Table 1.
ple with a peculiar platelet-like morphology presented high
To examine the thermal stabilities of the MA materials,
thermal stability and therefore acted as an excellent support
the samples that had been subjected to calcination at
for Pd catalysts in methane catalytic combustion.
500 °C were further calcined at a higher temperature rang-
ing from 700 to 1200 °C for 6 h, with b = 20 °C/min in air.
All of the MA samples were pelletized and sieved to 40–60
2. Experimental
mesh for use as catalyst supports.
2.1. Preparation methods
2.1.3. Preparation of alumina supported catalysts
2.1.1. Preparation of aluminum precursor The calcined MA supports were impregnated with an
The boehmite sols were prepared by two different proce- aqueous solution of Fe(NO3)3 Æ 9H2O or PdCl2, then the
dures. In one procedure, 13.67 g of boehmite powder (Con- samples were dried at 110 °C overnight and calcined at
Table 1
The synthesis conditions and textural properties of the MA samples
Sample I.D. Sol precursor P123/Al (mol ratio) Tdrya (°C) SBETb (m2/g) SMesopc (m2/g) Vpd (cm3/g) d100 (nm) Dpe (nm)
1 BP 0 110 265 355 0.30 – 3.4
2 BP 0.01 110 306 368 1.00 25.8 10.4
3 BP 0.02 110 339 422 1.19 27.2 11.3
4 BP 0.02 30 323 419 1.19 27.2 11.4
5 BP 0.02 150 319 418 1.23 27.6 11.7
6 BP 0.05 110 319 411 1.29 29.5 14.0
7 AN 0 110 269 359 0.31 – 4.2
8 AN 0.02 110 463 604 2.60 – 17.8
a
The drying temperature of the bohmite sol–P123 mixture.
b
The BET surface areas were calculated using the BET equation.
c
The mesoporous surface areas were calculated using the BJH equation based on the desorption branches of the isotherms.
d
The total pore volumes were determined at P/P0 value of 0.995.
e
The mean pore size distributions were determined by BJH model applied to the desorption branches of the isotherms.
Q. Liu et al. / Microporous and Mesoporous Materials 111 (2008) 323–333 325

550 °C for 5 h to yield Fe/Al2O3 catalysts with an Fe load- ½SO2 in  ½SO2 out
X ¼  100%
ing of 20 wt.%, and Pd/Al2O3 catalysts with Pd loading of ½SO2 in
1 wt.%, respectively.
For comparison, an Fe/Al2O3 catalyst was also ½SO2 in  ½SO2 out  ½COS
S sulfur ¼  100%
prepared by a one-pot synthesis procedure, where ½SO2 in  ½SO2 out
Fe(NO3)3 Æ 9H2O was dissolved in the boehmite sol before
where [SO2]in is the inlet concentration of SO2, [SO2]out and
the addition of the surfactant P123. The subsequent treat-
[COS]out are the outlet concentrations of SO2 and COS,
ment procedures were the same as described in Section
respectively.
2.1.2.

2.3.2. Pd/Al2O3 for methane catalytic combustion


2.2. Characterization techniques The catalytic activity of Pd/Al2O3 for methane combus-
tion was also tested in a fixed-bed flow reactor system.
Transmission electron microscopy (TEM) images were 30 mg of the catalyst was mixed with 0.5 g of quartz beads
obtained on a JEOL 2000 EX electron microscope operat- and then placed in a quartz reactor (5 mm i.d.). Prior to the
ing at an accelerating voltage of 200 kV. The samples were test, the catalyst was first reduced in situ with 30 ml/min H2
first dispersed ultrasonically in ethanol and then dropped at 300 °C for 1 h, and then purged with He (30 ml/min) at
onto the carbon-coated copper grids prior to the 300 °C for 30 min. After that, the catalyst was cooled to
observations. 200 °C, and a reaction mixture composed of 1 vol.%
Nitrogen adsorption–desorption experiments were per- CH4, 20.2 vol.% O2 and 78.8 vol.% He was fed into the
formed at 196 °C on a Micromeritics ASAP 2010 appara- reactor at a space velocity of 48,000 h1. The effluent gas
tus. The samples were evacuated firstly at 110 °C for 3 h was analyzed by an on-line gas chromatography equipped
and then at 350 °C for 5 h prior to their analysis. The spe- with TCD. Partial oxidation products such as CO were not
cific surface areas of the samples were calculated by the found in any of the experiments.
BET equation.
Powder X-ray diffraction (XRD) patterns were collected
3. Results and discussion
with a D/Max-bb diffractometer using a Cu Ka radiation
source (k = 0.15432 nm). Low-angle diffractions (2h =
3.1. Synthesis and characterization of the MA materials
0.1–10°) and wide-angle diffractions (2h = 10–80°) were
recorded at a scanning speed of 2°/min and 5°/min,
3.1.1. Morphologies
respectively.
Fig. 1 shows typical TEM images of the two precursors
Solid state 27Al MAS NMR spectra were recorded on a
as well as the resulting alumina samples. Obviously, the
Varian InfinityPlus spectrometer at a frequency of
two precursor sols contained different shapes of boehmite
104.17 MHz using a 4 mm zirconia rotor. The samples
crystallites. The BP sol was mainly built up of rod-like
were spinning at the magic angle with a spinning rate of
nanoparticles (Fig. 1a), while the AN sol was composed
10 kHz. 1.5 ls pulse width was used with a pulse delay of
of corrugated nano-platelets (Fig. 1b) [12]. After calcina-
2 s. The signal of an aqueous solution of 1%Al(NO3)3
tion at 500 °C, the boehmite precursors were transformed
was used as the reference for chemical shifts.
to c-Al2O3 through dehydration between the adjacent
boehmite particulates, but their morphologies were
2.3. Catalytic tests retained (Fig. 1c and D). This is in agreement with the
observations of Dı́az et al. [12] and Zhu et al. [13].
2.3.1. Fe/Al2O3 for catalytic reduction of SO2 by CO TEM images of the samples prepared with the aid of
The catalytic reaction between SO2 and CO was carried surfactant are shown in Fig. 2. The morphologies of the
out in a fixed-bed flow reactor system at atmospheric pres- precursors only changed very little after introducing the
sure. 0.2 g of the catalyst was placed in a quartz reactor non-ionic surfactant into the boehmite sol (Fig. 2a and
(6 mm i.d.). After being purged with a He flow (36 ml/min) b). It is known that the boehmite layers, formed by Al3+
at 550 °C for 2 h, the catalyst was presulfided in situ with surrounded octahedrally by O2 and OH, are linked
a reaction gas mixture containing 10,000 ppm CO and through hydrogen bonds and packed to give the particular
5000 ppm SO2 at 550 °C for 2 h. After the presulfidation, morphology [14]. Therefore, the fact that the morphologies
the catalyst was cooled to 130 °C for conducting activity of the precursors did not change obviously by the addition
tests. The effluent gas passed through an ice-water trap, of the surfactant suggested that the interaction between the
where elemental sulfur was condensed. SO2 and COS in boehmite and the surfactant was not strong enough to
the effluent gas were separated by Gaspro Capillary and destroy the hydrogen bonding interaction between
detected by FPD, while CO and CO2 were separated by the boehmite layers. According to the literature [15], the
Porapack Q and detected by TCD. The percent conversion hydrophilic poly ethylene oxide (PEO) headgroups of
(X) of SO2 and the selectivity (Ssulfur) to elemental sulfur the surfactant can interact with the OH groups on the sur-
are defined as follows: face of the boehmite layers through hydrogen bonding
326 Q. Liu et al. / Microporous and Mesoporous Materials 111 (2008) 323–333

Fig. 1. TEM images of (a, b) the as-synthesized and (c, d) the calcined blank samples. a and c: Sample 1 derived from the BP sol; b and d: Sample 7 derived
from the AN sol.

interactions. On the other hand, it has been reported that All the samples display classical IV type isotherms with
alkylene oxide segments can form crown-ether-type com- hysteresis loops, which is typical for mesoporous material
plex with many kinds of inorganic ions through weak coor- [17]. Comparing with the blank Sample 1 and Sample 7,
dination bonding [16]. In our case, we infer that analogous Sample 3 and Sample 8 prepared by using P123 as the tem-
hydrogen bonding and coordination interactions might plate show larger mesopores with broader pore size distri-
exist between the surfactant P123 and the boehmite partic- butions, which is accordant with their broad hysteresis
ulates. Though such interactions were not strong enough to loops centering at higher relative pressures.
change the intrinsic morphologies of the precursors, yet Table 1 lists the textural parameters of the samples pre-
they could induce effectively a stacking of the boehmite col- pared with different precursors and various surfactant
loids. As can be seen in Fig. 2c and d, the resulting Sample concentrations. The blank Sample 1 and Sample 7 give
3 and Sample 8 displayed respectively a lath-like and a specific surface areas of about 260 m2/g and pore volumes
scaffold-like configuration through loose stacking of the of ca. 0.30 cm3/g, which are quite similar to the conven-
boehmite ‘‘building blocks’’. tional c-alumina. This can also be seen from the relatively
dense morphologies caused by the close stacking of the
3.1.2. Textural properties boehmite particulates, as shown in Fig. 1c and d. On
Fig. 3 illustrates the nitrogen adsorption–desorption iso- the contrary, the MA samples prepared with P123 showed
therms and pore size distributions of the calcined samples. much larger pore volumes (1.0–2.6 cm3/g) and higher
Q. Liu et al. / Microporous and Mesoporous Materials 111 (2008) 323–333 327

Fig. 2. TEM images of (a, b) the as-synthesized and (c, d) the calcined samples prepared with P123/Al = 0.02. a and C: Sample 3 derived from the BP sol;
b and d: Sample 8 derived from the AN sol.

specific surface areas (300–470 m2/g) than the two blank ture of Sample 8. Hence, the large porosity of Sample 8
samples. should be attributed to the abundant nanospace between
Two trends can be found from Table 1. On the one the interpenetrated particles. In contrast, the BP sol was
hand, with an increase of the surfactant concentration, mainly comprised of rod-like nanoparticles, which yielded
both the pore volumes and the mean pore sizes increased. a lath-like mesostructure of Sample 3 with a smaller
On the other hand, different precursors caused great differ- porosity.
ences in the textural parameters. For example, Sample 8 In our previous work, we have found that the drying
derived from the AN sol precursor exhibited an extremely procedure had an important effect on the pore structures
large pore volume of 2.6 cm3/g and a high specific surface of the MA samples yielded from the boehmite/hydro-car-
area of 463 m2/g, whereas Sample 3 deriving from the BP boxylic acid reaction system [8]. With an increase in the
sol precursor showed a pore volume of 1.2 cm3/g and a sur- drying temperature, the MA samples changed from bimo-
face area of 339 m2/g. Such differences in the textural prop- dal mesoporous into mesoporous, and then into meso-
erties caused by the precursor can be explained by their microporous structures, which was attributed to the
different morphologies. From Section 3.1.1, we know that improved coordination interaction due to the increase of
the AN sol consisted of platelet-like colloids, which led to the drying temperature. However, in our present case of
a three-dimensional interconnected scaffold-like mesostruc- the boehmite/P123 reaction system, the trend was rather
328 Q. Liu et al. / Microporous and Mesoporous Materials 111 (2008) 323–333

1600 Table 2
8 0.3 The synthesis conditions and textural properties of the MA samples
Volune adsorbed (cm /g STP)

8
prepared using the AN sol as aluminum precursor in the presence of
1200 3 different structure-directing agents
0.2

dV/dD
3

1 Templates Template/Al SBET (m2/g) Vp (cm3/g) Dp (nm)


7
(mol ratio)
0.1
800
3 P123 0.02 463 2.60 17.8
0.0
Brij30 0.20 385 2.32 15.4
CTAB 0.125 413 1.91 12.4
0 5 10 15 20 25 30 35
400 7 SDS 0.125 368 1.30 11.7
Pore diameter (nm)
AOT 0.10 382 1.49 13.3
1
0
0.0 0.2 0.4 0.6 0.8 1.0 nanocrystallites, the surfactant amount and the surfactant
Relative pressure type.
Fig. 3. N2 adsorption–desorption isotherms and pore size distribution
patterns for the blank Sample 1 and Sample 7; Sample 3 and Sample 8 3.1.3. Crystalline framework walls
prepared with P123/Al = 0.02. Sample 1 and Sample 3 derived from the Fig. 4 presents the powder XRD patterns of the as-syn-
BP sol; Sample 7 and Sample 8 derived from the AN sol.
thesized and calcined samples. All of the as-synthesized
samples display diffraction lines of the boehmite phase
(JCPDS Card 21-1307), with the exception of a broad
different from the case of the boehmite/hydro-carboxylic peak centering at 22° that corresponds to the configura-
acid reaction system. In the boehmite/P123 system, varying tion of oxide groups in the surfactant molecules [12]. After
the drying temperature from 30 °C to 150 °C gave nearly calcination at 500 °C, the resulting MA samples showed
the same textural properties for the MA samples (see Sam- only one broad peak in the low-angle range of the XRD
ples 3–5 in Table 1). This observation clearly indicates that patterns (not given here), implying the existence of disor-
the drying temperature has a negligible influence on the dered mesostructures [18]. On the other hand, the wide-
properties of the MA materials. angle XRD patterns of the MA samples revealed that after
On the other hand, the effect of surfactant types has also the removal of surfactant molecules, the boehmite crystal-
been investigated. As shown in Table 2, comparing with lites were transformed into well-crystallized c-alumina par-
ionic surfactants (including cationic CTAB and anionic ticles (JCPDS Card 10-0425). The surfactant amount had
SDS and AOT), the employment of non-ionic surfactants no marked impact on this structural transformation.
(such as P123 and Brij30) will lead to larger textural Selected area electron diffraction (SAED) micrograph of
parameters. Obviously, the textural properties of the MA the typical MA Sample 3 (Fig. 2c, inset) exhibited concen-
materials obtained from the boehmite/surfactant system tric diffraction ring patterns corresponding to the strongest
depended largely on the morphologies of the boehmite (4 4 0) and (4 0 0) diffractions of c-alumina, suggesting once

as-synthesized calcined

sample 8
sample 8

sample 6
Intensity

sample 6

sample 3 sample 3

sample 2 sample 2

10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80
2θ (degree) 2θ (degree)

Fig. 4. Powder XRD patterns for as-synthesized and calcined samples prepared in the presence of P123.
Q. Liu et al. / Microporous and Mesoporous Materials 111 (2008) 323–333 329

rather than the widely accepted supermolecular self-assem-


bly mechanism.
To get an insight into the pore-forming mechanism, we
have performed further characterizations on our samples.
Fig. 6 shows typical TG-DTA curves of the as-synthesized
Sample 2. The slight weight loss at below 120 °C was
caused by the removal of the adsorbed water. Two exother-
mic peaks centering at 150 °C and 250 °C, corresponding
to a main weight loss of about 42.3% in the TG curve,
should be attributed to the oxidation and elimination of
embedded P123 in the mixture. Besides, the gradual weight
100 50 0 -50 -100 loss of ca. 12.5% at 325–500 °C could be ascribed to the
δ (ppm) dehydroxylation of the boehmite to the c-Al2O3 phase.
Fig. 5. 27Al MAS NMR of the calcined Sample 3 derived from the BP sol FTIR spectra of the as-synthesized Sample 1 and Sam-
with P123/Al = 0.02. ple 3 are given in Fig. 7. For the blank Sample 1, the
absorption bands at 473 and 620 cm1 were due to the
vibration of Al–O–Al, while the peaks at 742 and
more the presence of c-alumina nanocrystallites in the 1070 cm1 can be assigned respectively to the bending
framework walls. In addition, the assignment of crystal vibrations of Al–O and OH in the boehmite. In addition,
phase can be confirmed further by the 27Al MAS NMR the bending mode peaks of the intercalated molecular
spectra shown in Fig. 5. Only tetrahedral (63 ppm) and water in the boehmite phase were positioned at 1640 and
octahedral (6 ppm) aluminum sites were observed, with 3080 cm1 [21]. By comparison, the Sample 3, prepared
an intensity ratio of 25:75, which was characteristic of c- under the presence of P123, shows several new absorption
alumina [19]. It should be mentioned that calcining the bands besides the characteristic bands of the boehmite
boehmite precursor at temperatures higher than 450 °C phase. The peaks at 650, 1107, 1250, 1462 and 1730 cm1
will lead directly to the formation of c-Al2O3 [20]. Obvi- were associated with the EO and PO chains in P123, and
ously, the addition of P123 into the boehmite precursor the bands at 2800–2900 cm1 were caused by the vibration
did not affect the crystalline phase of the resulting alu- of –CH2– groups in P123 [13].
mina, manifesting again the weak interaction between The above characterization results revealed that the
the surfactant and the boehmite. interaction between the boehmite precursor and the P123
molecules was comparatively weak, and had only negligible
3.1.4. Pore-forming procedure impact on the intrinsic morphology and crystalline struc-
The synthesis of MA materials in the presence of non- ture of the boehmite precursor. However, such a weak
ionic surfactants has previously been reported by several interaction could induce a change in the stacking mode
groups using organic aluminum alkoxide [15,18] and inor- of the boehmite layers as building blocks [8]. This mecha-
ganic Al(NO3)3 or AlCl3 [4] as precursors. By comparing nism was more or less analogous to the formation of the
our MA samples with those reported in the literature in boehmite–surfactant ‘‘sandwich’’ structure proposed by
terms of textural parameters, it can be noted that there González-Peña and his co-workers [15]. In detail, the
are two important advantages for our MA samples over non-ionic surfactant might intercalate into the boehmite
those reported by other investigators. The first advantage
of our samples is the large pore volume. As shown in
30
Tables 1 and 2, all the MA samples prepared with the aid
0
of surfactants had pore volumes greater than 1.0 cm3/g.
In particular, Sample 8 even had a pore volume of -10 20
2.6 cm3/g. On the contrary, most MA samples reported
Weight loss (%)

-20
in the literatures possessed pore volumes only ranging from 10
0.4 to 0.7 cm3/g [15,18]. The second advantage of our MA
Exo.

-30
samples is that all of the three textural parameters, i.e., sur-
-40 0
face area, pore volume and average pore size, exhibit com-
paratively high values. In other words, the large surface -50
areas (300–470 m2/g) were accompanied with the high pore -10
volumes (1.0–2.6 cm3/g) and large pore sizes (10.0– -60

17.8 nm). However, for the MA samples reported in the lit-


erature, large pore sizes were usually accompanied by lower 100 200 300 400 500 600 700 800

surface areas. Such exceptionally large porosities in our Temperature (oC)


samples suggested that a different pore-forming mechanism Fig. 6. TG-DTA curves for the as-synthesized Sample 2 prepared with
might be at work in our present boehmite/P123 system, P123/Al = 0.01 derived from the BP sol.
330 Q. Liu et al. / Microporous and Mesoporous Materials 111 (2008) 323–333

100

sample 3 80

SO2 conversion (%)


Absorbance

60

sample 1 40

20

500 1000 1500 2000 2500 3000 3500 4000 100 150 200 250 300 350 400 450
Wavenumber (cm-1) o
Temperature ( C)
Fig. 7. FTIR spectra of the as synthesized Sample 1 and Sample 3 derived
from BP sol.
100

layers through their interactions, leading thus to a random

Sulfur selectivity (%)


80
packing of a boehmite–surfactant sandwich-like structure.
Upon calciation, partial collapse and interpenetration of 60
the layers could give rise to large-sized porosity and a
broad pore size distribution. In this mechanism, the 40
boehmite morphology, the surfactant amount and the sur-
factant type would impose influences on the textural 20

parameters of the resulting MA materials with the altering


0
of the boehmite interlayer space.
200 250 300 350 400 450
Temperature (oC)
3.2. Application of catalysts for SO2 reduction by CO
Fig. 8. SO2 conversion and sulfur selectivity over 20 wt.% Fe supported
It has been known that the textural properties of meso- on (j) Sample 1, (d) Sample 3 and (m) Sample 8 (Feed compositions:
porous c-aluminas have great influence on the catalytic 5000 ppm SO2 and 10,000 ppm CO, WHSV = 18,000 ml/g h).
performance when they are used as a catalyst support.
Especially for hydrodesulfurization reactions, a high sur- Table 3
face area and a large pore volume can often lead to an Catalytic activities of 20 wt.% Fe supported on the alumina samples for
enhancement in catalytic activity [22]. In the present work, SO2 reduction by CO
we have chosen the catalytic reduction of SO2 with CO as a Support I.D. 1 2 3 6 7 8
model reaction for evaluating the influence of the textural T10%a(°C) 250 190 180 184 245 135
properties of c-Al2O3 on the catalytic performance. This T100%b(°C) 380 340 330 330 380 300
reaction was of environmental importance, as it can elimi- a
The temperature that SO2 be 10% reduced by Fe/Al2O3 catalysts.
nate the pollutions caused by both SO2 and CO, while at b
The temperature that SO2 be totally catalytic reduced to elemental S.
the same time yielding sulfur as a commercial product
[23–25]. Thus, with reference to our previous work [25],
we have prepared catalysts of 20 wt.% Fe supporting on increased from 265 m2/g (Sample 1) to 463 m2/g (Sample
the MA materials obtained in this study. 8). To the best of our knowledge, such an activity level is
Fig. 8 illustrates the SO2 conversion and sulfur selectiv- the highest reported so far [24,26].
ity as a function of the reaction temperature. Comparing Fig. 9 demonstrates the XRD patterns of the catalysts
with Sample 1 prepared without surfactant, both Sample after being presulfidated. A single peak was observed in
3 and Sample 8 showed remarkable improvements in their the low-angle region, implying that the mesostructures of
catalytic performance. The temperatures at which SO2 was the MA materials were still retained after supporting a
completely converted were reduced drastically; while the large amount of Fe. Besides, several peaks appeared in
selectivity to elementary sulfur was increased. As summa- the wide-angle region, which should be assigned to the
rized in Table 3, the temperature at which SO2 began to active FeS2 phase [27]. When the MA materials were
be reduced by CO (T10%) decreased from 250 °C to employed as the support, the characteristic peaks of FeS2
135 °C; while the temperature at which SO2 was completely phase were weakened and broadened, suggesting a higher
converted (T100%) was reduced from 380 °C down to dispersion of the active phase. Hence, an improvement in
300 °C when the surface areas of the supports were textural properties of the supports could effectively increase
Q. Liu et al. / Microporous and Mesoporous Materials 111 (2008) 323–333 331

Fig. 9. Powder XRD patterns in wide-angle regions (left) and low-angle regions (right) for 20 wt.% Fe supported on (A) Sample 1, (B) Sample 3, (C)
Sample 8 and (D) the one-pot synthesized sample after sulfidation at 550 °C for 2 h (h: FeS2 phase).

the activities of the catalysts by enhancing the dispersion of Table 4, after calcining at 1000 °C for 6 h, Sample 8 had
the active phase. a BET specific surface area of 174 m2/g and a pore volume
It should be mentioned that the catalyst prepared by a of 0.99 cm3/g, and its surface area was still as high as
one-pot synthesis procedure did not show any activity even 81 m2/g even after being calcined at 1200 °C for 6 h. On
above 400 °C. The single peak in the low-angle range of the the other hand, Sample 3 derived from the BP sol showed
XRD patterns (see Fig. 9D) suggested that the mesostruc- poor thermal stability. Its surface area dropped to 10 m2/g
ture remained well even in the presence of a large amount after calcination at 1200 °C.
of Fe in the precursor. But after presulfidation at 550 °C, The high thermal stability of Sample 8 should be
no FeS2 phase could be observed in the XRD pattern. attributed to its peculiar scaffold-like morphology. The
We inferred that the Fe atoms might be embedded or low contact areas between the platelet-like ‘‘building
located in the framework of the alumina during the one- blocks’’ could hinder their sintering [12]. This could be
pot synthesis process and could not be accessible by the further confirmed by their morphological changes after
reactants during the presulfidation or in the reduction reac- thermal treatments. As shown in Fig. 11, after calcining
tion. As a result, the Fe/MA catalyst from one-pot synthe- at 1200 °C, the nanoparticles of Sample 8 became spher-
sis was not active for the catalytic reduction of SO2. ical within the range of 20–50 nm due to sintering,
accompanied by a phase transformation from c-Al2O3
3.3. Thermal stability of MA and application for methane to a-Al2O3. Contrarily, Sample 3 showed a textural struc-
catalytic combustion ture of random stacking of large particles with sizes well
above 100 nm after calcined at 1200 °C. Clearly, the
One of the limitations of the mesoporous materials is smaller particles of Sample 8 formed by calcination at
their low thermal stabilities. For the above MA samples 1200 °C could give rise to larger surface area and pore
prepared in our present work, the single XRD peak in volume. Therefore, the difference in morphology between
the low-angle range disappeared completely after calcina- Sample 8 and Sample 3 could lead to a great difference in
tion at 700 °C, indicating the collapse of the mesostruc- their textural properties after high temperature
tures. However, the crystalline phase of the c-Al2O3 did treatments.
not change until calcining at 800 °C, as indicated by the It is expected that alumina materials with high thermal
wide-angle XRD patterns (Fig. 10). When the calcination stabilities will find important applications in reactions
temperature was increased to 1200 °C, only the a-phase, involving high temperature processes. As an example, we
the most thermodynamically stable alumina phase, could illustrate here the advantages of aluminas with high ther-
be observed. mal stability as the support for a Pd catalyst in methane
It was worth noting that Sample 8 derived from the AN catalytic combustion. As we all know, methane combustion
sol exhibited high resistance to sintering. As shown in is a strongly exothermic reaction and therefore requires a
332 Q. Liu et al. / Microporous and Mesoporous Materials 111 (2008) 323–333

F
E

Intensity D

10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80
2θ (degree) 2θ (degree)

Fig. 10. XRD patterns of Sample 8 derived from the AN sol with P123/Al = 0.02 after calcined at (A) 500 °C, (B) 600 °C, (C) 700 °C, (D) 800 °C, (E)
1000 °C and (F) 1200 °C for 6 h.

Table 4 with the Pd/Al2O3 catalyst is activity loss at high tempera-


The textural properties of the samples calcined at different temperatures tures due to the sintering of the alumina support, the for-
Sample I. D. Sample 3 Sample 8 mation of spinels or the agglomeration of the active
Tcal (°C) 700 800 1000 1200 700 800 1000 1200 phase [31].
SBET (m2/g) 305 177 128 10 380 227 174 81 By considering the high thermal stability of Sample 8,
Vp (cm3/g) 1.14 0.83 0.33 0.02 2.14 1.57 0.99 0.31 we supported Pd onto it and investigated its catalytic per-
Dp (nm) 16.5 17.5 – – 20.4 22.4 – – formance in CH4 combustion. As references, Sample 1 and
Sample 3 were also employed as supports. Table 5 lists
thermally stable and active catalyst. Pd/Al2O3 is one of the their activity data in terms of T10, T50 and T90. It can be
most investigated catalysts for this reaction due to its high seen that the three characteristic temperatures were the
activity [28–30]. However, the major problem associated lowest for Sample 8, but the highest for Sample 1, indicat-

Fig. 11. TEM images of the samples prepared with P123/Al = 0.02, calcined at 500 °C for 4 h and then at 1200 °C for 6 h. (a) Sample 8 derived from the
AN sol; (b) Sample 3 derived from the BP sol.
Q. Liu et al. / Microporous and Mesoporous Materials 111 (2008) 323–333 333

Table 5 Acknowledgments
Catalytic activities for methane combustion of 1 wt.% Pd supported on the
alumina samples
Supports of National Science Foundation of China
Support I.D. Tcala(°C) Catalytic activityb(°C) (NSFC) for Distinguished Young Scholars (No.
T10 T50 T90 20325620) and NSFC Grant (No. 20303017) are gratefully
1 500 393 475 646 acknowledged. We also thank Prof. D.B. Liang for his kind
3 500 351 445 507 help in English improvement.
8 500 325 430 482
1 1000 353 425 516
3 1000 329 412 486 References
8 1000 298 337 422
a
The calcination temperature of the alumina supports. [1] C. Misra, Industrial Alumina Chemicals, ACS Monograph 184,
b
Temperatures at which methane conversions reach 10%, 50% and 90%, American Chemical Society, Washington, DC, 1986.
respectively. [2] F. Schüth, K. Unger, in: G. Ertl, H. Knözinger, J. Weitkamp (Eds.),
Preparation of Solid Catalysts, Wiley-VCH, Weinheim, 1999, pp. 77–
80.
ing that when Pd was supported on Sample 8, it showed the [3] J. Čejka, Appl. Catal. A 254 (2003) 327, and references therein.
[4] Z. Zhang, T.J. Pinnavaia, J. Am. Chem. Soc. 124 (2002) 12294.
highest activity towards methane combustion. Moreover, if
[5] B. Tian, X. Liu, B. Tu, C. Yu, J. Fan, L. Wang, S. Xie, G.D. Stucky,
the alumina samples were precalcined in air at 1000 °C for D. Zhao, Nature Mater. 2 (2003) 159.
6 h and then supported with the Pd active phase, the cata- [6] K. Niesz, P. Yang, G.A. Somorjai, Chem. Commun. (2005) 1982.
lysts could yield even higher activities, as compared with [7] Q. Liu, A. Wang, X. Wang, T. Zhang, Chem. Mater. 18 (2006)
those supports calcined at 500 °C. Especially for Sample 5153.
[8] Q. Liu, A. Wang, X. Wang, T. Zhang, Microporous Mesoporous
8, the T10 was as low as 298 °C and the T90 was 422 °C.
Mater. 92 (2006) 10.
The possible reason for such a promotion in catalytic per- [9] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck,
formance is supposed to be associated with the larger pore Nature 359 (1992) 710.
sizes generated by the high-temperature treatment, which [10] H.C. Lee, H.J. Kim, C.H. Rhee, K.H. Lee, J.S. Lee, S.H. Chung,
favored molecular diffusion and mass transfer during the Microporous Mesoporous Mater. 79 (2005) 61.
[11] J. Aguado, J.M. Escola, M.C. Castro, B. Paredes, Microporous
reaction. However, we can not exclude other factors, such
Mesoporous Mater. 83 (2005) 181.
as the modification in morphology of the palladium parti- [12] I. Dı́az, V. González-Peña, C. Márquez-Alvarez, J. Pérez-Pariente
cles by the different alumina crystalline phases [28], the Collect. Czech. Chem. Commun. 68 (2003) 1937.
alteration in metal-support interaction during high-temper- [13] H.Y. Zhu, J.D. Riches, J.C. Barry, Chem. Mater. 14 (2002) 2080.
ature treatments, and so on [32]. This will be further eluci- [14] H.Y. Zhu, X.P. Gao, D.Y. Song, Y.Q. Bai, S.P. Ringer, Z. Gao, Y.X.
Xi, W. Martens, J.D. Riches, R.L. Frost, J. Phys. Chem. B 108 (2004)
dated in our future work. Nevertheless, the above results
4245.
have illustrated the obvious advantages of mesoporous [15] V. González-Peña, I. Dı́az, C. Márquez-Alvarez, E. Sastre, J. Pérez-
aluminas with high thermal stability for applications in Pariente, Microporous Mesoporous Mater. 44–45 (2001) 203.
methane combustion. [16] P. Yang, D. Zhao, D.I. Margolese, B.F. Chmelka, G.D. Stucky,
Nature 396 (1998) 152.
[17] S.J. Gregg, K.S.W. Sing, Adsorption, Surface Area and Porosity,
4. Conclusions Academic Press, London, 1982.
[18] S.A. Bagshaw, T.J. Pinnavaia, Angew. Chem. Int. Ed. Engl. 35 (1996)
Mesoporous c-aluminas were prepared using boehmite 1102.
sols as the precursors and non-ionic surfactants as the [19] C.S. John, N.C.M. Alma, G.R. Hays, Appl. Catal. 6 (1983) 341.
[20] M. Digne, P. Sautet, P. Raybaud, H. Toulhoat, E. Artacho, J. Phys.
structure-directing agents. With the aid of the surfactant,
Chem. B 106 (2002) 5155.
the pre-formed boehmite nanocrystallites could act as [21] L. Ji, J. Lin, K.L. Tan, H.C. Zeng, Chem. Mater. 12 (2000) 931.
‘‘building blocks’’ to form the mesostructure with crystal- [22] L. Kaluža, M. Zdražil, N. Žilková, J. Čejka, Catal. Commun. 3
line framework walls. The interactions between the boehm- (2002) 151–157.
ite particulates and the surfactant were not strong enough [23] S.E. Khalafalla, L.A. Haas, J. Catal. 24 (1972) 121.
[24] S.X. Zhuang, H. Magara, M. Yamazaki, Y. Takahashi, M. Yamada,
to destroy the intrinsic crystalline structure of the boehm-
Appl. Catal. B 24 (2000) 89.
ite, but could effectively direct the loose stacking of the [25] X. Wang, A. Wang, N. Li, X. Wang, Z. Liu, T. Zhang, Ind. Eng.
boehmite particulates, leading to mesostructured aluminas Chem. Res. 45 (2006) 4582.
with extremely large pore volumes and pore sizes. The mes- [26] C.L. Chen, H.S. Wang, Appl. Catal. B 55 (2005) 115.
oporous c-alumina with the platelet-like morphology had a [27] L.A. Hass, S.E. Khalafalla, J. Catal. 29 (1973) 264.
[28] R.F. Hicks, H. Qi, M.L. Young, R. G Lee, J. Catal. 122 (1990) 295.
very large surface area and a high resistance to sintering,
[29] T.R. Baldwin, R. Burch, Appl. Catal. 66 (1990) 337.
and therefore demonstrated a great advantage as a catalyst [30] P. Briot, M. Primet, Appl. Catal. 68 (1991) 301.
support for SO2 reduction by CO and methane catalytic [31] P. Euzen, J.H. Le Gal, B. Rebours, G. Martin, Catal. Today 47 (1999) 19.
combustion. [32] K. Eguchi, H. Arai, Appl. Catal. A 222 (2001) 359.

You might also like