Applied Catalysis B: Environmental 72 (2007) 157–165

www.elsevier.com/locate/apcatb

Gold supported on Fe, Ce, and Al pillared bentonites

for CO oxidation reaction
J.G. Carriazo a, L.M. Martı́nez b, J.A. Odriozola b, S. Moreno a,
R. Molina a, M.A. Centeno b,*
a
Estado Sólido y Catálisis Ambiental, Departamento de Quı́mica, Universidad Nacional de Colombia, Carrera 30 No 45-03, Bogotá, Colombia
b
Instituto de Ciencia de Materiales de Sevilla, Centro Mixto Universidad de Sevilla-CSIC, Avda Americo Vespuccio s/n 41092 Sevilla, Spain
Received 21 June 2006; received in revised form 17 October 2006; accepted 21 October 2006
Available online 21 November 2006

Abstract
A series of gold catalysts supported on Fe, Ce, and Al pillared bentonites have been prepared, characterized by means of X-ray diffraction
(XRD), transmission electronic microscopy (TEM), SBET, and X-ray fluorescence spectrometry (XRF) techniques and tested in the CO oxidation
reaction in the absence and in presence of water in the stream. The pillaring process of the natural bentonite results in an increase in the
microporosity and surface area of the solid. In the gold samples, spherical arrangements of gold particles of 2–4 nm around one central particle of
4–5 nm are observed, the catalysts having an average gold particle size of 2.2 nm. All the gold catalysts are very active, those modified with Fe and
Ce being the best. The introduction of iron and cerium by pillaring increases the catalytic performances of the final gold solids. Introduced iron
atoms are accessible and active centers for the gaseous CO oxidation reaction while the cerium ones promote the catalytic activity of Fe and gold
atoms. Catalysts activity increases in the presence of water in the reactant stream at temperatures lower than 120 8C. At room temperature, the
catalytic activity increases continuously with the gaseous water concentration of up to 6700 ppm.
# 2006 Elsevier B.V. All rights reserved.

Keywords: Gold catalysts; Fe, Ce, and Al pillared bentonites; CO oxidation

1. Introduction metal oxides [3,8], oxynitrides [14,15], and zeolites [16–20]. It

In the past few years, the number of studies focusing on gold
catalysts has increased exponentially. The discovery of the high
activity of gold nanoparticles in CO oxidation at low
temperature has been the main factor of such interest, since,
until then, gold had traditionally been considered as an inactive
material [1]. The catalytic activity of these gold catalysts has
been reported to be influenced mainly by gold particle size, and
the nature of the support [2–5]. Support materials have been
classified into active (those able to provide oxygen atoms) and
inactive [6]. Well-studied examples of active supports are ceria
[7–9] and iron oxides [10,11]. For inactive support, the activity
of the system is attributed only to the high dispersion of gold
particles. A typical case is alumina [3,12,13]. A large amount of
materials have been used for supporting gold, among them,
* Corresponding author. Tel: +34 954489543; fax: +34 954460665.
E-mail address: centeno@icmse.csic.es (M.A. Centeno).
seems that small gold particles can be inserted into the cages of
the zeolite [16,18–21] resulting in very active and stable
catalysts for the CO oxidation reaction. To our knowledge, only
three different kinds of zeolites (Y, b, and mordenite) have been
reported as support for preparing gold catalysts [19]. On the
other hand, it has been described that the presence of iron atoms
in these zeolite induces a higher initial activity towards CO
oxidation [18]. The improvement of the CO oxidation activity
of gold catalysts by addition of iron atoms has also been also
reported for Au/CeO2, Au/SnO2, and Au/TiO2 systems [10].
Within this context, the present work explores the utilization
of natural clay as support for preparing active gold catalysts for
CO oxidation. The natural clay, a Colombian bentonite, has an
initial iron content that may contribute to increase the oxidation
capabilities of the final catalysts prepared. In order to increase
the low surface area of the clay, a modification of the bentonite
was performed with Al, Al–Fe or Al–Ce–Fe intercalating
solutions. In this way, pillared clays containing Al, Fe, and Ce
are obtained. Previous studies confirm the success of the

0926-3373/$ – see front matter # 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcatb.2006.10.018
158 J.G. Carriazo et al. / Applied Catalysis B: Environmental 72 (2007) 157–165
pillaring process, the pillars formed essentially by Al–Fe
forming nanoclusters of iron species (oxides or oxyhydroxydes)
[22]. Pillared clays offer interesting catalytic and adsorption
properties. The introduction of inorganic pillars increases
microporosity and offers a greater surface area, improving the
material’s resistance and stability [23]. The addition of small
molar fractions of a second and third cation, iron and cerium in
our case, has the purpose of increasing the thermal, adsorption
and/or catalytic properties of the final solid [24]. The
incorporated iron can act as a new active center while Ce
acts as a catalytic promoter, improving the redox properties and
the metal dispersion of the solid. In fact, in a previous paper it
has been reported that these solids are active for the oxidation of
phenol in liquid phase, the effect of introducing Fe and Ce being
favorable [22]. As stated above, the presence of iron oxides and
ceria in the pillared clays is also expected to result in an
enhancement of the observed CO oxidation performances of the
prepared gold catalysts.
2. Experimental
2.1. Preparation of pillared bentonites
The original clay is a commercial type natural bentonite
issued from Valle del Cauca, Colombia, supplied by Bentocol
(Colombia), previously separated by particle size (<2 mm) and
homoionized with a sodium chloride solution 1 N. This clay has
been described in previous works [25–27]. In order to modify
the clay, different intercalating solutions have been prepared: a
10% Al–Fe solution, expressed as a Fe molar percentage Fe/
(Fe + Al); a 10% Al–Ce solution, expressed as a Ce molar
percentage Ce/(Ce + Al); and a 10% Al–Ce–Fe solutions,
expressed as molar percentages Fe/(Fe + Al + Ce) and keeping
a 10% molar cerium molar percentage. In addition, a
polyhydroxicatonic solution containing only Al is prepared.
All solutions have been prepared from their corresponding
nitrates through the slow addition of NaOH 0.2 M to the
solution of Al, Al–Fe, Al–Ce or Al–Ce–Fe cations, at room
temperature (20 8C), until a ratio OH/metal = 2 is obtained. The
solutions obtained are aged for 2 h at 60 8C.
The clay pillarization process is carried out through the ionic
exchange of the previously aged polymeric solutions, keeping a
ratio of 20 meq. of total metal per gram of clay. The modifying
solution is slowly added to a 2% clay aqueous suspension and
then it is left stirring for 3 h at room temperature. The solids
exchanged are washed by centrifugation until nitrates are
eliminated. They are dried at 60 8C and finally calcined at
400 8C for 2 h. The solids are labelled according to the
following nomenclature: BAl, BAlCe(10), BAlFe(10), BAl-
Ce(10)Fe(10), B referring to bentonite and the values shown in
parentheses indicating the molar percentage for the corre-
sponding element in the intercalating solution.
2.2. Synthesis of gold catalysts
The gold-containing catalysts were prepared by the
deposition–precipitation method. Previous to the process of
the deposition of gold, pillared bentonite was dispersed with
constant stirring in a solution (1 N) of sodium nitrate at a pH of
6 for one night, with the aim of adjusting the surface acidity of
the solid to the pH on which the gold deposition will be carried
out, to achieve an effective deposition of the metallic particles
and obtain stable particles of the smallest possible size [18].
After that, the solid was filtered and dried at 70 8C. The
deposition–precipitation process was done according to the
method described by Tsubota et al. [28]. The adequate amount
of HAuCl43H2O (Alfa, 99.99% pure) was dissolved in 250 mL
of deionized water and the pH of the solution adjusted to 6.0 by
addition of NaOH 0.1 M. The solution obtained was heated at
80 8C and then the support was added, keeping that temperature
for 1 h in constant stirring. After cooling at room temperature,
the solid obtained was separated by filtration and washed with
deionized water up to the end of chlorides and sodium cations.
Finally, the solid was dried in an oven at 100 8C. The catalysts
obtained are labelled Au/X, X being the label of the
corresponding bentonite support.
2.2.1. Characterization techniques
The textural properties were studied by N2 adsorption
measurements at liquid nitrogen temperature. The experiences
were carried out by means of a Micromeritics ASAP 2010
equipment. Before analysis, the samples were degassed for 1 h
at 90 8C and 8 h at 350 8C in vacuum.
X-ray diffraction (XRD) analysis was performed on a
Siemens D-500 diffractometer. Diffraction patterns were
recorded by using the plate-oriented technique with Cu Ka
radiation (40 mA, 40 kV) and a position-sensitive detector
using a step size of 0.058 and a step time of 2 s.
The chemical composition of the samples was determined
by X-ray fluorescence spectrometry (XRF) in a Siemens SRS
3000 sequential spectrophotometer with a rhodium tube as the
source of radiation. XRF measurements were performed onto
pressed pellets (sample included in 10 wt% of wax).
Transmission electronic microscopy (TEM) observations
were carried out with a Philips CM 200 microscope operating at
200 kV. The samples were dispersed in ethanol by sonication
and dropped on a copper grid covered with a carbon film.
2.2.2. Activity measurement
The CO oxidation reaction tests were carried out in a
conventional continuous flow U-shape glass reactor working at
atmospheric pressure. The sample (80 mg) is placed among
glass wools. A thermocouple in contact with the sample assures
the right measure of temperature. The feed mixtures were
prepared using mass flow controllers (Bronkhorst). The
reaction was followed by mass spectrometry, using a Balzers
Thermostar benchtop mass spectrometer controlled by the
software Balzers Quadstar 422 with capabilities for quantitative
analysis.
The light-off curves of CO oxidation (400 8C, 5 8C/min) were
obtained with a mixture of 3.4% CO (Air Liquide, 99.997% pure,
<3 ppm H2O) and 21% O2 (Air Liquide, 99.999% pure, <3 ppm
H2O) balanced by He (Air Liquide, 99.999% pure, <3 ppm H2O)
at a total flow rate of 42 mL/min. Empty reactor shows no activity
 J.G. Carriazo et al. / Applied Catalysis B: Environmental 72 (2007) 157–165 159

under such conditions. The catalysts were pre-activated ‘‘in situ’’

for 1 h at 400 8C with a mixture of 21% O2 in He at a flow of
30 mL/min and then stabilized at room temperature before the
light-off curve started. In order to evaluate the influence of water
in the catalytic performances of the samples, a second test was
carried out in which the reactive stream was saturated with water
at room temperature before admittance into the reactor.

3. Results and discussion

Table 1 shows the chemical composition of the solids

considered, obtained from XRF measurements. The natural
bentonite presents a composition typical of a smectite
containing some amount of metals, mainly iron (8.4%),
titanium (0.7%), and manganese (0.069%). The modified clays
obtained after the pillaring process always display a higher
content in the metals present in the intercalating solutions than
those of the natural bentonite, indicating their effective
introduction in the material. The amount of gold introduced
in the solids after the deposition-precipitation process varies
from 0.56 to 0.9% weight/weight. Nonmodified natural
bentonite (B) and the clay modified with Al solution (BAl)
are the ones with the lower amount of gold deposited. On the
contrary, those clays modified with cerium present the highest
gold content. These results agree with the reported preferential
deposition of gold over ceria in CeO2/Al2O3 systems, arising
from the more basic character of CeO2 compared to alumina
[3,8], and with the observed four times higher deposition of
gold over a ceria coated foam compared to that observed over Fig. 1. XRD patterns of studied solids. (A) Gold catalysts, (B) supports. (a)
an alumina coated foam, prepared in similar conditions [29]. Natural bentonite; (b) BAl; (c) BAlFe(10); (d) BAlCe(10); (e) BAl-
Ce(10)Fe(10).
The basal space of the natural bentonite (B), with or without
gold, calculated from the d001 signal of their XRD pattern
(Fig. 1) is 14.7 Å. In the case of all modified clays, this value in the intensity of the diffraction lines, Fig. 1. These
shifts to 16–18 Å, highlighting the success of the pillarization, observations could indicate that the gold deposition procedure
Table 2. Although the differences in the basal spaces are not could drive to a little alteration of the pillars in the solids. At this
very high, it seems that those solids containing cerium have point we must take into account that this process implies a
bigger basal spaces than the others. This observation has been previous treatment of the bentonite with a 1 N sodium nitrate
reported early and it could indicate that cerium is likely to solution at pH 6, and the subsequent deposition-precipitation
provide favorable conditions for the formation of the alumina procedure with the chloroauric acid solution at 808.
pillar enhancing the thermal resistance of the pillar [24,30,31]. No diffraction peaks due to gold species are detected for the
The introduction of gold does not modify the observed basal catalysts supported on modified solids, pointing out their low
spaces of the bentonites, Table 2, but produces a little decrease crystalline domain size and/or concentration (Fig. 2). However,

Table 1
Chemical composition of the considered solids
Solid Metal content (% w/w)
Al Si Ce Fe Ca Ti Mn Cl K Au
B 11.54 28.65 – 8.39 1.21 0.72 0.069 0.6 1.08 –
Au/B 11.91 28.09 – 8.81 0.65 0.78 0.051 0.37 1.24 0.65
BAl 13.92 27.67 – 7.48 0.71 0.72 0.077 0.16 1.08 –
Au/BAl 14.08 27.72 – 7.13 0.68 0.66 0.060 0.17 0.91 0.56
BAlFe(10) 13.71 25.99 – 10.56 0.64 0.66 0.077 0.2 1.00 –
Au/BAlFe(10) 13.87 25.61 – 10.49 0.61 0.66 0.053 0.16 0.74 0.8
BAlCe(10) 13.87 27.39 0.25 7.97 0.71 0.66 0.055 0.18 1.00 –
Au/BAlCe(10) 13.87 27.67 0.25 6.99 0.67 0.66 0.073 0.18 0.91 0.87
BAlCe(10)Fe(10) 13.34 25.85 0.37 10.91 0.61 0.66 0.076 0.13 1.00 –
Au/BAlCe(10)Fe(10) 13.39 25.90 0.23 10.35 0.67 0.66 0.075 0.16 0.83 0.9
160 J.G. Carriazo et al. / Applied Catalysis B: Environmental 72 (2007) 157–165

Table 2
Textural properties and basal spaces values of the considered solids
Solid Surface area Micropore volume d001
(m2/g) (cm3/g) (Å)
B 44 0.0006 14.7
Au/B 60 0.0070 14.7
BAl 86 0.0253 16.5
Au/BAl 95 0.0176 16.5
BAlFe(10) 118 0.0372 15.8
Au/BAlFe(10) 109 0.0189 15.8
BAlCe(10) 116 0.0301 17.4
Au/BAlCe(10) 84 0.0168 17.4
BAlCe(10)Fe(10) 132 0.0418 17.7
Au/BAlCe(10)Fe(10) 98 0.0199 17.7

diffraction lines characteristic of polycrystalline gold phase

(JCPDS 04-0784) at 2u = 38.2, 44.4 and 64.68 are observed for
Au/B sample. The size of the gold crystallites in this sample is
deduced from the broadening of the reflections peak at 38.28
according to Debye–Scherrer’s equation and assuming the
spherical shape of the gold particles to be 63 nm.
The textural properties of the solids studied are presented in
Table 2. The initial surface area of the natural bentonite
(44 m2 g1) increase more than 100% after all the pillaring
process, showing the success of the pillaring and the subsequent
formation of the microporous structures. In fact, the measured
surface area on the modified solids is determined practically by
the microporous area. Bentonites modified with cerium and/or
iron drive to a higher surface area and micropore volume, the
solid modified with the solution containing Al, Ce, and Fe
cations being the one with the highest values. The addition of
gold results in a decrease in the surface area and micropore
volume of the modified clay. These decreases are higher for the
bentonites modified with cerium and/or iron. This observation
can be explained assuming that gold particles are deposited in
both the meso- and micropores of the solid. It could be
considered that at least some of the gold particles are decorating
the pillars, decreasing in this way the surface area and
Fig. 2. XRD patterns of gold studied solids. (A) Au/B; (B) Au/BAl; (C) Au/
BAlFe(10); (D) Au/BAlCe(10); (E) Au/BAlCe(10)Fe(10); (*) gold.
Fig. 3. Examples of N2 adsorption-desorption isotherms of the solids studied.
(&) Natural bentonite; (*) BAl; (~) BAlCe(10); (5) BAlCe(10)Fe(10).
micropore volume while maintaining the basal space. In fact, as
stated above, cerium and iron atoms are decorating the
aluminium pillars, and gold is preferentially deposited on
cerium and iron atoms. On the contrary, in the case of the
natural bentonite, the introduction of gold induces an increase
in its SBET and micropore volume values, which could be related
with an expansion of the lamellae of the clay due to the gold
deposition process.
The N2 adsorption–desorption isotherms of selected samples
are shown in Fig. 3. All solids present a similar isotherm
characteristic of meso- or macroporous materials [32,33],
although modified solids show a higher adsorption capacity
associated with the formation of microporosity. In all solids, a
type H3 hysteresis, typical of absorbents with slit-shaped pores
[34], is observed.
Typical TEM micrographs of natural and modified
bentonites with gold are shown in Figs. 4 and 5. Unmodified
bentonite presents iron oxides agglomerates of up to 200 nm
(not shown). These iron species are only of about 15–25 nm in
the case of the modified solids (not shown). In these materials,
the longitudinal arrangements at about 4 nm distance related
with grouping of face-to-face structures at long distances can be
seen, a consequence of the successful pillaring of the clay,
Fig. 6 [31,35].
In the case of gold catalysts, important differences in the
gold particle size and distribution are detected between the
modified and unmodified clays. The gold supported on natural
bentonite (Au/B) sample presents a very bad distribution of
metallic particles with high diameters of up to 100 nm. This
agrees with the average size of 63 nm calculated from XRD
results and with the low amount of gold determined from XRF.
However, gold particles are regularly distributed over the
modified bentonites, the particle morphology being similar in
the four solids. It is possible to distinguish some spherical
particles of about 10–20 nm besides spherical arrangements of
gold particles of 2–4 nm around one central particle of 4–5 nm,
Fig. 5. The total dimension of these spherical arrangements of
gold particles is also of 10–20 nm. So, it is possible to claim that
 J.G. Carriazo et al. / Applied Catalysis B: Environmental 72 (2007) 157–165 161

Fig. 4. TEM micrographs of gold catalysts.

the spherical particles formed are produced from condensation
of these spherical arrangements. To our knowledge, this finding
has not yet been described elsewhere. On the other hand,
particles with a very low diameter (below 1 nm) are present on
the whole supports, although the determination of the real
number of such particles is difficult to measure due to the low
mass and diffraction contrast [3]. Probably, most of the gold
particles could be smaller than 1 nm and not detectable.
Keeping in mind these events, the average gold particle size
calculated from the particle size distribution in selected regions
of the TEM micrographs turns out to be 2.2 nm. An example is
displayed in Fig. 7.
Fig. 8 shows the activity towards CO oxidation of the all
solids considered. The activity of the bentonites without gold is
very low and depends on the type and concentration of metallic
atoms introduced in the solid by the pillaring process. Natural
clay is almost inactive. A little increase in activity is observed
for the bentonites pillared with aluminium (BAl) and cerium
(BAlCe(10)), while the iron contained clay (BAlFe(10))
displays considerably activity, pointing out that the iron atoms
162 J.G. Carriazo et al. / Applied Catalysis B: Environmental 72 (2007) 157–165
Fig. 5. TEM micrographs showing the spherical arrangements of gold particles
of 2–4 nm around one central particle of 4–5 nm.
introduced by pillaring are accessible and active centers for the
gaseous CO oxidation reaction. The simultaneous presence of
cerium and iron atoms in the clay results in a higher increase of
its oxidative performances. These results could mean that the
addition of cerium does not generate very active centers, but it
promotes the catalytic activity for Fe, probably as a
consequence of the known redox properties of the CeO2
[8,36]. A more detailed discussion concerning the CO oxidation
activity of the pillared bentonites as a function of the Ce and Fe
content of the clay has been reported elsewhere [31]. In this
paper, the catalytic properties of the iron species are put in
evidence, as well as the promoting effect of the cerium ones.
Also, some considerations concerning the Fe active sites are
reported. In this sense, Mossbauer spectroscopy results indicate
that the active iron oxides species are meanly as Fe+3 in the
hematite phase [31].
Fig. 6. TEM micrograph of modified bentonites.
Fig. 7. Example of gold particle size distribution obtained in a selected region
of the TEM micrographs of a gold-modified bentonite catalysts.
The presence of gold strongly enhances the CO oxidation
activity of the all-pillared solids. In the case of the natural clay,
the introduction of gold does not imply a high increment on
activity, which can be related with the low surface area and
textural properties of the natural clay, which drives to a
noneffective deposition of gold, producing less accessible gold
particles with a high particle size, as deduced from TEM
observations (Fig. 4). On the other hand, the Au/B sample
presents itself a relatively low specific surface area and a low
gold content (0.65%), which contributes to the low catalytic
activity detected. The sequence of activity of the gold-pillared
samples is the same than that observed for the supports:
Au=BAlCeð10ÞFeð10Þ > Au=BAlFeð10Þ
> Au=BAlCeð10Þ > Au=BAl
with T50 values (temperature at which the 50% CO conversion is
achieved) of 47, 49, 52 and, 63 8C, respectively. The intrinsic
activity of the iron and cerium centers introduced in the pillaring
process is thus confirmed, that of iron species being higher
than that of cerium. It is important to note that the difference
in activity between Au/BAlCe(10) and Au/BAlFe(10) is not as
high as the one observed for the corresponding supports. This
agrees with the reported promoting effect of cerium atoms in
the gold catalytic activity for CO oxidation [3,8,29]. This effect,
commonly related with the easiness of CeO2 to undergo fast
 J.G. Carriazo et al. / Applied Catalysis B: Environmental 72 (2007) 157–165
Fig. 8. Conversion of CO on the considered solids, without (support) and with
gold. (A) B; (B) BAl; (C) BAlCe(10); (D) BAlFe(10); (E) BAlCe(10)Fe(10).
reduction/oxidation cycles enhancing the mobility of lattice
oxygen and the capacity of the solid to storage oxygen, which
leads to an increase in the catalytic activity of the metallic sites, is
particularly observed in the case of Au/BAlCe(10)Fe(10), where
the catalytic properties of both iron and gold centers are
improved. This solid presents a 30% CO conversion at room
temperature and the total CO oxidation is achieved at 68 8C.
Besides this, a recent paper has reported that ceria stabilizes a
more oxidized state and a higher dispersion of gold than those
observed for Au/Al2O3, the activity of these Aud+ species being
higher than that of Au0 centers [12,37].
Another important feature is the observation of a wide
shoulder in the CO light-off curve of Au/BAlFe(10), Au/
BAlCe(10), and Au/BAl samples. This pseudo-maximum
detected in the CO conversion appears at around 120 8C
and cannot be ascribed to CO/CO2 adsorption/desorptions
phenomena since the CO and CO2 concentration curves are
symmetrical, and the total carbon balance is close to 100% in
every temperature (not shown). However, this could be related
with the presence of water since it has been recently reported
that at temperatures lower than 120 8C the presence of water in
the feed increases the catalytic activity towards CO oxidation of
a series of Au/CeO2/Al2O3 samples, the effect of water being
negative at higher temperatures [8]. The positive effect of the
presence of water at low temperatures was ascribed to the
improved control of the redox reactions implied in the reaction
mechanism of the catalytic system. In this sense, the role of the
163
presence of water at low temperatures on the surface of the
catalysts can be similar to that of ceria, which means favoring
the mobility of the oxygen species and stabilizing a more
oxidized and active oxidation state of gold. In fact, water has
been described to be able to reoxidize reduced ceria at room
temperature [38]. On the other hand, it has been reported that
water vapor can regenerate deactivated gold catalysts through
the hydrolysis of the Au-carbonate complexes formed on the
surface of Au/Al2O3 catalysts during CO oxidation [13]. The
same authors propose that this may be the reason for the
observed different sensitivity to water and rates of deactivation
for Au on different supports.
In order to verify the existence of a relationship between the
pseudo-maximum activity observed and the presence of water
on the catalysts surface and/or in the reactant feed, some
experiences were carried out.
In the first one, three CO light-off curves were carried out
consecutively over the same catalyst, following simultaneously
the evolution of the intensity of the m/z = 18 signal,
corresponding to water. The results obtained for the most
active solid, Au/BAlCe(10)Fe(10), are shown in Fig. 9. In the
first run, performed over the fresh calcined solid, a complex
evolution of the water signal is observed, showing at least five
peaks of water desorption at 88, 120, 140, 156, and 393 8C.
These desorptions must be related not only with the desorption
of adsorbed water on the external and inner surface and in the
pores of the solid, but also with the different hydroxyls
condensation of the bentonite support. In the second run, the
water signal profile has changed, presenting three maxima at
73, 97, and 354 8C. Not only has the temperature of water
desorption changed but also the intensity of the water signal,
that is, the concentration of water desorbed. In this second run,
water is easily desorbed at low temperatures and the amount of
water released from the catalyst is decreased. Finally, in the
third run, no water desorption is detected at temperatures lower
than 250 8C. Concerning the catalytic activity measured in the
experiment, as said above, in the first run the solid displays total
CO conversion at 68 8C, temperature lower than those of water
desorption. In the second run, the CO conversion decreases
Fig. 9. CO conversion (solid lines) and evolution of water (dash ones) on three
consecutives CO light-off curves obtained for Au/BAlCe(10)Fe(10) catalysts.
164 J.G. Carriazo et al. / Applied Catalysis B: Environmental 72 (2007) 157–165
Fig. 10. CO conversion (solid lines) and evolution of water (dash ones) on Au/
BAlCe(10) catalysts in wet and dry conditions.
significantly, and a small shoulder in the light-off curve is
detected at about 75 8C, the temperature at which the first peak
of water desorption is detected. Finally, the activity of the
catalyst in the third run is similar to that of the second one,
except for little changes in the shape of the conversion shoulder.
From here it is clear that both the activity of the catalysts and
the amount and temperature of water released are altered after
some consecutive light-off experiments. However, since the
observed decrease in activity can also be due to the deactivation
of the catalysts, for instance due to coke deposition or
agglomeration of the gold particles, a second experience is
necessary. Fig. 10 shows the CO light-off curve obtained for
Au/BAlCe(10) catalysts when the reactant feed is saturated
with water at room temperature. At low temperatures, the
presence of water deeply increases the catalytic activity of the
sample, reaching about 85% CO conversion at room
temperature. In dry conditions, the CO conversion of this
sample at room temperature was 20%. This result provides
clear evidence on the positive effect of water in the catalytic
properties of the catalyst at low temperatures. On the other
hand, it is interesting to note that at about 130 8C, a sudden
decrease in the CO conversion is produced when water is in the
feed. After that, in the range of temperature from 137 to 145 8C,
the CO conversion matches that of the system in dry conditions.
Above these temperatures, the conversion achieved in humid
conditions is worse than that of dry ones. These observations
agree with those reported early for Au/CeO2/Al2O3 catalysts
[8]. In that paper, it is reported that water exerts a positive
influence in the catalyst’s activity at temperatures lower than
120 8C but, above this temperature, the water effect becomes
negative. From this observation, it seems that the effect in the
activity of water on the catalysts surface is more important than
the effect of water in the gas phase, as proposed earlier [13,39].
Once the increase in the CO oxidation catalytic activity of
the gold-pillared catalysts at low temperatures due to the
addition of water in the reaction mixture has been confirmed,
another interesting question is to determine the relationship
between the concentration of water in the feed and the CO
conversion achieved. Fig. 11 shows the CO conversion obtained
for Au/BAlCe(10)Fe(10) at 20 8C in the steady state of reaction
Fig. 11. CO conversion obtained for Au/BAlCe(10)Fe(10) at 20 8C in the
steady state of reaction as a function of the water concentration.
as a function of the water concentration. As can be seen, a
sigmoidal relationship is obtained. The CO conversion
continuously increased, sharply at low water concentrations,
until an almost independent value is achieved at concentrations
of water above 6700 ppm.
4. Conclusions
Gold supported on Fe, Ce, and Al pillared bentonites are
active catalysts for the CO oxidation reaction. The pillaring
process of the natural bentonite results in an increase in the
microporosity and surface area of the solid. The introduction of
iron and cerium by pillaring increases the catalytic perfor-
mances of the final gold solids. Iron atoms are accessible and
active centers for the gaseous CO oxidation reaction while
cerium promotes the catalytic activity of Fe and gold atoms.
The activity of the catalysts increases in the presence of water in
the reactant stream at temperatures lower than 120 8C. At room
temperature, the catalytic activity increases continuously with
the gaseous water concentration until 6700 ppm.
Acknowledgements
Financial support by CSIC-Spain/COLCIENCIAS-Colom-
bia (2003CO0002 and 2004CO0014), Junta de Andalucı́a
(TEP106, PAI), and Universidad Nacional De Colombia
(Project DIB No.20201005196 and VRI No 20201007579)
are gratefully acknowledged.
References
[1] M. Haruta, T. Kobayashi, H. Sano, N. Yamada, Chem. Lett. (1987) 405.
[2] N. Weiher, E. Bus, L. Delannoy, C. Louis, D.E. Ramaker, J.T. Miller, J.A.
van Bokhoven, J. Catal. 240 (2006) 100.
[3] M.A. Centeno, M. Paulis, M. Montes, J.A. Odriozola, Appl. Catal. A 234
(2002) 65.
[4] T.V.W. Janssens, A. Carlsson, A. Puig-Molina, B.S. Clausen, J. Catal. 240
(2006) 108.
[5] M. Valden, X. Lai, D.W. Goodman, Science 281 (1998) 1647.
 J.G. Carriazo et al. / Applied Catalysis B: Environmental 72 (2007) 157–165
[6] M. Schubert, S. Hackenberg, A.C. van Veen, M. Muhler, V. Plzak, R.J.
Behm, J. Catal. 197 (2001) 113.
[7] F. Arena, P. Famulari, G. Trunfio, G. Bonura, F. Frusteri, L. Spadaro, Appl.
Catal. B 66 (2006) 81.
[8] M.A. Centeno, C. Portales, I. Carrizosa, J.A. Odriozola, Catal. Lett. 102
(2005) 289.
[9] P. Concepción, S. Carretin, A. Corma, Appl. Catal. A 307 (2006) 42.
[10] F. Moreau, G.C. Bond, Catal. Today 114 (2006) 362.
[11] L. Guczi, D. Horváth, Z. Pászti, G. Petõ, Catal. Today 72 (2002) 101.
[12] Tz. Venkov, Hr. Klimev, M.A. Centeno, J.A. Odriozola, K. Hadjiivanov,
Catal. Commun. 7 (2006) 308.
[13] C.K. Costello, J.H. Yang, H.Y. Law, Y. Wang, J.-N. Lin, L.D. Marks, M.C.
Kung, H.H. Kung, Appl. Catal. A 243 (2003) 15.
[14] M.A. Centeno, I. Carrizosa, J.A. Odriozola, Appl. Catal. A 246 (2003)
365.
[15] M.A. Centeno, M. Paulis, M. Montes, J.A. Odriozola, Appl. Catal. B 61
(2005) 177.
[16] J.-N. Lin, B.-Z. Wan, Appl. Catal. B 41 (2003) 83.
[17] G. Riahi, D. Guillemot, M. Polisset-Thfoin, A.A. Khodadadi, J. Fraissard,
Catal. Today 72 (2002) 115.
[18] J.-N. Lin, J.-H. Chen, C.-Y. Hsiao, Y.-M. Kang, B.-Z. Wan, Appl. Catal. B
36 (2002) 19.
[19] J.-H. Chen, J.-N. Lin, Y.-M. Kang, W.-Y. Yu, C.-N. Kuo, B.-Z. Wan, Appl.
Catal. A 291 (2005) 162.
[20] T.M. Salama, T. Shido, H. Mingawa, M. Ichikawa, J. Catal. 152 (1995)
322.
[21] S. Qui, R. Ohnishi, M. Ichikawa, J. Phys. Chem. 101 (1994) 2719.
[22] J. Carriazo, E. Guélou, J. Barrault, J.M. Tatibouet, R. Molina, S. Moreno,
Catal. Today 107–108 (2005) 126.
[23] M.F. Molina, R. Molina, S. Moreno, Catal. Today 107–108 (2005) 426.
165
[24] A. Gil, L.M. Gandia, M.A. Vicente, Catal. Rev. Sci. Eng. 42 (2000) 145.
[25] J. Carriazo, E. Guélou, J. Barrault, J.M. Tatibouet, S. Moreno, Appl. Clay
Sci. 22 (2003) 303.
[26] J. Carriazo, E. Guélou, J. Barrault, J.M. Tatibouet, R. Molina, S. Moreno,
Water Res. 39 (2005) 3891.
[27] J.A. Pinzón, J. Requena, J. Bol. Soc. Esp. Ceram. Vidrio 35 (1996) 337.
[28] S. Tsubota, D.A. Cunningham, Y. Bando, M. Haruta, in: G. Poncelet, P.
Grange, P.A. Jacobs (Eds.), Preparation of Catalysts VI, Elsevier, Ámster-
dam, 1995, p. p.227.
[29] M.I. Domı́nguez, M. Sánchez, M.A. Centeno, M. Montes, J.A. Odriozola,
Appl. Catal. A 302 (2006) 96.
[30] M.J. Hernando, C. Pesquera, C. Blanco, F. Gonzalez, Langmuir 18 (2002)
5633.
[31] J.G. Carriazo, M.A. Centeno, J.A. Odriozola, S. Moreno, R. Molina, Appl.
Catal. A (2006), in press.
[32] F. Rouquerol, J. Rouquerol, K.-S.-W. Sing, in: F. Schüth, K.-S.-W. Sing,
J. Weitkamp (Eds.), Handbook of porous solids, vol. 1, Wiley-VCH,
Wincheim (Germany), 2002, p. 236.
[33] R. Masel, Principles of adsorption and reaction on solids surfaces, John
Willey and sons, New York (USA), 1996, p. 235.
[34] G. Leofanti, M. Padovan, G. Tzzola, B. Venturelli, Catal. Today 41 (1998)
207.
[35] J. Domı́nguez, M.L. Occelli, in: M.L. Occelli, H.E. Robson (Eds.),
Expanded clays and others microporous solids (Synthesis of microporous
materials), vol. II, Van Nostrand Reinhold, New York, 1992, pp. 81–107.
[36] A. Trovarelli, Catal. Rev. Sci. Eng. 38 (1996) 439.
[37] M.A. Centeno, K. Hadjiivanov, Tz. Venkov, Hr. Klimev, J.A. Odriozola, J.
Molec. Catal. A 252 (2006) 142.
[38] C. Padeste, N. Cant, D. Trimm, Catal. Lett. 18 (1993) 305.
[39] M. Date, M. Haruta, J. Catal. 201 (2001) 221.