Synthesis, characterisation and catalytic performance of nanocrystalline Co3O4 for gas-phase chlorinated VOC abatement

Article abstract of DOI:10.1016/j.jcat.2011.04.005

Several nanocrystalline Co₃O₄ catalysts were investigated for their activity and selectivity during the oxidation of 1,2-dichloroethane, which was selected as a model chlorinated volatile organic compound. A wide number of synthesis routes starting from cobalt(II) nitrate were examined, namely calcination of the precursor salt, solid-state reaction, precipitation and sol-gel. The catalysts prepared by precipitation decomposed the chlorinated feed at the lowest temperatures. Activity was observed to be chiefly governed by a small crystallite size which may give rise to more easily accessible active sites (oxygen -O^- or O^2-- species), which were not present on the more highly crystalline Co₃O ₄ catalysts. Additionally, surface Lewis acidity played a relevant catalytic role. Interestingly, the behaviour of some of the nanocrystalline oxides was superior to that of supported noble metal catalysts and other bulk oxide catalysts. Conversion to deep oxidation products was complete (CO ₂, HCl and Cl₂), and no appreciable deactivation with time on stream was noticed.

Full text of DOI:10.1016/j.jcat.2011.04.005

Journal of Catalysis 281 (2011) 88–97
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Synthesis, characterisation and catalytic performance of nanocrystalline Co3O4
for gas-phase chlorinated VOC abatement
⇑
Beatriz de Rivas, Rubén López-Fonseca, Cristina Jiménez-González, José I. Gutiérrez-Ortiz
Department of Chemical Engineering, Faculty of Science and Technology, Universidad del País Vasco/EHU, P.O. Box 644, E-48080 Bilbao, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Several nanocrystalline Co3O4 catalysts were investigated for their activity and selectivity during the oxi-
dation of 1,2-dichloroethane, which was selected as a model chlorinated volatile organic compound. A
wide number of synthesis routes starting from cobalt(II) nitrate were examined, namely calcination of
the precursor salt, solid-state reaction, precipitation and sol–gel. The catalysts prepared by precipitation
decomposed the chlorinated feed at the lowest temperatures. Activity was observed to be chiefly
governed by a small crystallite size which may give rise to more easily accessible active sites (oxygen
–Oꢀ or O2ꢀ– species), which were not present on the more highly crystalline Co3O4 catalysts. Additionally,
surface Lewis acidity played a relevant catalytic role. Interestingly, the behaviour of some of the
nanocrystalline oxides was superior to that of supported noble metal catalysts and other bulk oxide
catalysts. Conversion to deep oxidation products was complete (CO2, HCl and Cl2), and no appreciable
deactivation with time on stream was noticed.
Received 10 February 2011
Revised 4 April 2011
Accepted 5 April 2011
Available online 10 May 2011
Keywords:
VOC oxidation
1,2-Dichlorethane
Ethylene dichloride
Co3O4
Synthesis
Crystallite size
Ó 2011 Elsevier Inc. All rights reserved.
1. Introduction
An alternative to commonly applied but expensive noble metals
is transition metal oxides. Despite their high activity for the abate-
Over the past several years, environmental legislation has im-
posed increasingly stringent limits on permitted atmospheric
emission levels. In particular, the release of volatile organic com-
pounds (VOCs) has received much attention. VOCs are a widely
ranging class of chemicals commonly occurring in many commer-
cial waste streams and containing over 300 compounds. Their re-
lease has widespread environmental implications and has been
linked to the increase in photochemical smog, the depletion in
atmospheric ozone and the production of ground-level ozone.
Additionally, many VOCs have highly noxious effects on human
health. The emission sources associated with industrial activity
(manufacture of organic chemicals and polymers, vent air from
operations in which VOCs are used for cleaning and degreasing
purposes in metal processing, machining and finishing, and
exhaust air from ground water and/or soil remediation processes)
reveal a wide variety of compounds with different chemical
compositions, though the ease or difficulty involved in removing
them varies considerably. In this sense, chlorinated VOCs, which
have enjoyed widespread acceptance as reagents and solvents
due to their compatibility with most substrate materials and
non-flammability, are, on account of their high stability, amongst
the most difficult to abate compounds by catalytic combustion.
ment of chlorinated VOCs, the disadvantages associated with noble
metals (high cost, low thermal stability and susceptibility to
poisoning [1]) cannot be overcome. Thus, substantial efforts are
currently being made to develop transition metal oxides catalysts
with a comparable activity. Cobalt oxide catalysts are effective
low-temperature combustion catalysts of non-chlorinated hydro-
carbons [2–5], carbon monoxide [6,7] and diesel soot [8]. However,
much less consideration has been given to investigating the appli-
cability of this class of catalysts for the gas-phase oxidation of chlo-
rinated VOCs [9,10].
The active behaviour of Co-based catalysts in the above applica-
tions, typically as Co3O4, is most likely related to high bulk oxygen
mobility and the facile formation of highly active oxygen (Oꢀ or
O2ꢀ) species. These properties are mainly related to crystallite size.
Hence, the controlled preparation of Co3O4 nanoparticles (1–50 nm)
has attracted considerable interest owing to its impact on
governing their ultimate performance and applications. In the
present work, the suitability of cobalt oxide catalysts for the abate-
ment of chlorinated VOCs was investigated. Particularly, the cata-
lytic behaviour of nanocrystalline Co3O4 systems on the total
oxidation of 1,2-dichloroethane, which was chosen as a model
chlorinated VOC, was examined in a fixed-bed flow reactor. 1,2-
Dichloroethane (C2H4Cl2, DCE), also known in the industry as
ethylene dichloride (EDC), is probably one of the most important
chlorinated VOCs emitted in gaseous industrial waste streams
since it is used as an intermediate for the production of polyvinyl
⇑
Corresponding author. Fax: +34 94 6015963.
E-mail address: joseignacio.gutierrez@ehu.es (J.I. Gutiérrez-Ortiz).
0021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.04.005

B. de Rivas et al. / Journal of Catalysis 281 (2011) 88–97
89
chloride, the most produced plastic in the world after polyethylene.
Less important uses are as a solvent in textile cleaning and metal
degreasing and paint remover, a starting material for paint, varnish,
and finish removers, a cleaner for upholstery and carpets, a fumigant,
a lead scavenger in antiknock gasoline and as a dispersant for plas-
tics and elastomers such as synthetic rubber.
eritics ASAP 2010 apparatus. The specific areas of the samples were
determined in line with standard BET procedure, using nitrogen
adsorption taken in the relative equilibrium pressure interval of
0.03–0.3. Mean pore size was calculated using the BJH method.
The samples were previously degassed overnight under high vac-
uum conditions. X-ray diffraction (XRD) studies were conducted
on a X’PERT-MPD X-ray diffractometer with Cu K
a radiation
2. Experimental
(k = 1.5406 Å) and Ni filter. The X-ray tube was operated at 30 kV
and 20 mA. Samples were scanned between 20° (2h) and 70°
(2h), and the X-ray diffraction line positions were determined with
a step size of 0.02° and a counting time of 2.5 s per step. Phase
identification was conducted by comparison with JCPDS (Joint
Committee on Powder Diffraction Standards) database cards. The
thermo-oxidative degradation of the cobalt precursors was investi-
gated by means of dynamic thermogravimetry using Setaram
Setsys Evolution apparatus under atmospheric pressure. The mass
loss and the sample temperature were continuously recorded by a
computerised data acquisition system. The studies were carried
2.1. Catalyst preparation
A number of routes including thermal decomposition, solid-
state reaction, wet-chemical and sol–gel methods were employed
to synthesise bulk cobalt oxide catalysts. Firstly, Co3O4 was pre-
pared by simple calcination in air of the selected cobalt precursor
(hexahydrated cobalt(II) nitrate, Sigma–Aldrich). This catalyst
was denoted as DC. Secondly, the cobalt oxide was prepared by so-
lid-state reaction following two different procedures. Hence, on
one hand, the GB catalyst was prepared by grinding a mixture of
cobalt(II) nitrate (9 g) with ammonium hydrogen carbonate (Fluka)
(6.1 g) with a molar ratio of 2:5 in an agate mortar for 30 min [11].
The solids were washed thoroughly with distilled water and col-
lected by filtration. The GC catalyst was obtained by grinding a
mixture of citric acid (5.3 g) (Sigma–Aldrich) with the Co(II) basic
carbonate precursor (8.7 g) (Sigma–Aldrich) [12]. The mixture
was first premixed by hand grinding for 5 min. The grinding was
then carried out in a planetary ball mill (Retsch) at a speed of
600 rpm for 6 h. The as-ground citrate precursors were collected
and washed thoroughly.
out from 25 to 550 °C at a constant heating rate of 10 °C minꢀ1
.
The oxidant stream was dry air (50 cm3 minꢀ1) flowing down-
wards onto the cylindrical sample holder.
Raman spectra, acquired using a Leica 50ꢂ N Plan (0.75 aper-
ture) lens, were recorded with a Renishaw InVia Raman spectrom-
eter coupled to a Leica DMLM microscope. The spectrometer was
equipped with a 514-nm laser (ion-argon laser, Modu-Laser) with
a nominal power at the source of 50 mW, with maximum power at
the sample of 20 mW. Ten seconds were employed for each spec-
trum, and 20 scans were accumulated with 10% of the maximum
power in the spectral window from 150 to 1150 cmꢀ1
Temperature-programmed desorption (TPD) of ammonia was
performed on Micromeritics AutoChem 2920 instrument
.
Co3O4 catalysts were also prepared by wet-chemical methods.
The first route (OW) consisted of a precipitation–oxidation reac-
tion in an aqueous solution. The precipitation process was con-
a
equipped with a quartz U-tube coupled to a thermal conductivity
detector. Prior to adsorption experiments, the samples were first
pre-treated in a 5%O2/He stream at 500 °C and then cooled to
100 °C in a He flow (20 cm3 minꢀ1). Later, the NH3 adsorption step
was performed by admitting a flow of 10%NH3/He at 100 °C up to
saturation. Subsequently, the samples were exposed to a flow of
helium (50 cm3 minꢀ1) for 1 h at 100 °C to remove reversibly and
physically bound ammonia from the surface. Finally, desorption
was carried out from 100 to 500 °C at a heating rate of 10 °C minꢀ1
in an He stream (50 cm3 minꢀ1). This temperature was maintained
for 1 h until the adsorbate was completely desorbed. The amount
of gases desorbed was determined by time integration of the TPD
curves. Diffuse reflectance (DRIFT) spectra of pyridine adsorbed
on the oxide samples were obtained with a Nicolet Protegé 460
ESP spectrometer, equipped with a Spectra-Tech high-temperature
chamber and a nitrogen-cooled MCT detector. All spectra were re-
corded in the range 1700–1300 cmꢀ1 averaging 400 scans with a
1 cmꢀ1 resolution and analysed using OMNIC software. After the
sample was evacuated at 550 °C under high vacuum conditions
for 1 h, pyridine was admitted at 200 °C at the equilibrium pres-
sure of 3 mbar. After removing physisorbed pyridine, the spectra
were then recorded. Difference spectra were obtained by subtract-
ing the spectrum of the clean sample from the spectra obtained
after pyridine adsorption.
Redox behaviour was examined by temperature-programmed
reduction (TPR), and the experiments were also conducted on a
Micromeritics AutoChem 2920 instrument as well. Firstly, all the
samples were pre-treated in an oxygen stream (5%O2/He) at
500 °C for 1 h and then cooled to room temperature. The reducing
gas used in all experiments was 5%H2/Ar, with a flow rate of
50 cm3 minꢀ1. The temperature range explored was from room
temperature to 500 °C, with a heating rate of 10 °C minꢀ1. This
temperature was maintained for 0.5 h. The water produced by
reduction was trapped in a cold trap, and consumption of H2 was
quantitatively measured by time integration of the TPR profiles.
ducted by the drop-by-drop addition of 100 ml of
a 3.2 M
solution of NaOH (Sigma–Aldrich) into an aqueous solution (50
cm3) of 0.6 M Co(NO3)2ꢁ6H2O at 50 °C. Next, 100 ml of H2O2
(Sigma–Aldrich, 50 wt.%) was also introduced drop-by-drop under
constant stirring [13]. The precipitate was then filtered and
washed with deionised water. The alternative procedure (CC) in-
volved an aqueous hydroxycarbonate precipitation [12]. Thus, a
1.2 M aqueous (200 cm3) solution of Na2CO3 (Fluka) was added
into a 0.5 M aqueous (100 cm3) solution of cobalt(II) nitrate under
vigorous stirring. The temperature was kept at 80 °C during the
precipitation (0.5 h), the pH was fixed at 8.5 and the precipitates
were collected and washed thoroughly. Finally, a sol–gel method
(SG) was employed. The oxide was prepared by an aqueous sol–
gel citrate procedure involving complexation of cobalt(II) nitrate
(0.1 M, 250 cm3) and citric acid. An excess of citric acid (3.5 g)
was used to ensure complete complexation. Water was removed
on a rotary evaporator at 40 °C until the formation of a gel or vis-
cous material. The temperature was then increased up to 70 °C and
maintained overnight. A spongy, highly hygroscopic, amorphous
citrate was obtained [14].
All the catalyst precursors were dried at 110 °C overnight and
then calcined at 300 or 500 °C in static air for 4 h at a heating rate
of 1 °C minꢀ1. For the samples calcined at 500 °C, an intermediate
calcination at 300 °C for 0.5 h was performed. Next, catalyst pellets
with a 0.3–0.5 mm diameter were prepared by a process of
compressing the oxide powders into flakes in a hydraulic press
(Specac), crushing and sieving. All samples obtained prior to
catalytic activity and selectivity experiments were characterised
using several analytical techniques.
2.2. Characterisation techniques
Textural properties were evaluated from the nitrogen adsorp-
tion–desorption isotherms, determined at ꢀ196 °C with a Microm-

90
B. de Rivas et al. / Journal of Catalysis 281 (2011) 88–97
During this test, the reduction of Co3O4 to Co was assumed to be
complete since no hydrogen consumption was registered when a
second TPR run was performed. Oxygen storage complete capacity
(OSCC) measurements were taken using Micromeritics ASAP 2010
equipment. The samples were first calcined at 500 °C for 1 h in a
5%O2/He stream to clean the surface and then subjected to a reduc-
tion treatment with a pure hydrogen stream at constant tempera-
ture (150–350 °C) for 1 h. The sample was then purged with
helium, and the measurement of oxygen consumption at 500 °C
was taken.
selectivity was calculated based on either chlorine or carbon atoms
present in that product divided by the total chlorine or carbon
atoms present in the product stream (expressed as %). Feed and
effluent streams were analysed using an online 7980A Agilent
Technologies gas chromatograph equipped with thermal conduc-
tivity (CO and CO2) and an electron capture detector (chlorinated
hydrocarbons). Analysis of HCl and Cl2 was conducted using ion
selective electrode and titration, respectively. Further details on
analytical procedures are described elsewhere [15].
The quantification of chlorine content present in the used cata-
lysts was evaluated by energy disperse X-ray (EDX also known as
EDS, or EDAX) analysis using a JEOL JSM-6400 scanning electron
microscope coupled with analysis software INCA Energy 350 from
Oxford Instruments.
3. Results
3.1. Physicochemical characterisation of the Co3O4 samples
Thermogravimetric analysis was conducted in order to deter-
mine the temperature for the oxidative decomposition of the var-
2.3. Catalyst activity determination
ious cobalt catalytic precursors. As a result, the activation
procedure could be defined so as to obtain the desired cobalt oxide
(Co3O4). Several weight losses were recorded from 110 to 300 °C
(not shown here), though no further change in weight was ob-
served at higher temperatures. Thus, calcination at a temperature
of at least 300 °C was required to obtain cobalt oxide phases as
Co3O4. Two calcination temperatures were selected, namely 300
and 500 °C in air for 4 h, to analyse the thermal stability of the oxi-
des. Hence, the samples were denoted as ꢀ300, for example DC-
300, and as ꢀ500, for example DC-500.
Catalytic tests were performed in a bench-scale fixed-bed
reactor (Microactivity modular laboratory system provided by PID
Eng&Tech S.L.) operated at atmospheric pressure and fully monitored
by computer. The reactor was made of quartz with an internal
diameter of 10 mm and a height of 300 mm, in which the temper-
ature was controlled with a thermocouple place in the catalyst bed.
Typically 0.85 g of catalyst in powdered form (0.3–0.5 mm) was
loaded. The catalyst bed was diluted with inert quartz (0.85 g, 1–
1.25 mm). The reaction feed consisted of 1000 ppm of DCE in dry
air with a total gas flow of 500 cm3 minꢀ1. The amount and granu-
lometric fraction of catalyst and the total gaseous stream were
chosen selected to be in a true kinetic regime, i.e. beyond the inter-
nal and external diffusion limits, and to reach a gas hourly space
velocity of 15,000 hꢀ1, which corresponds to the conditions usually
met in the exhaust gases from industrial units. Catalytic activity
was measured over the range 150–500 °C, and conversion data
were calculated by the difference between inlet and outlet concen-
trations. Conversion measurements and product profiles were ta-
ken at steady state, typically after 30 min on stream. Product
The structural properties were characterised by XRD using the
JCPDS files as a reference. Fig. 1 shows the XRD patterns of the co-
balt samples calcined at 300 and 500 °C. As for the samples acti-
vated at the lower temperature, it was checked that all the
reflection peaks located at 2h 31.3°, 36.8°, 44.8°, 59.3° and 65.2°
could be perfectly indexed to a pure cubic phase of Co3O4 spinel
(JCPDS 78-1970), irrespectively of the synthesis route. No diffrac-
tion peaks related to a CoO phase were detected [12,16]. The aver-
age crystallite size, estimated from the full width half maximum of
the characteristic diffraction peaks by applying the Scherrer
equation (Table 1), was about 5–6 nm for all the samples except
Co3O4
DC-500
GB-500
GC-500
GB-300
GC-300
DC-300
OW-500
OW-300
SG-500
CC-500
SG-300
CC-300
20
30
40
50
60
70
30
40
50
30
40
20
50
70
60
60
Angle, 2θ
Fig. 1. XRD diffraction patterns of the Co3O4 samples.

B. de Rivas et al. / Journal of Catalysis 281 (2011) 88–97
91
for DC-300 (18 nm). After calcination at 500 °C, all the samples
maintained their structure, i.e. Co3O4 cobalt oxide in the absence
of other cobalt species. However, significantly more intense well-
resolved reflections were noticeable in larger particle sizes (higher
crystallinity) [13,17]. The influence of calcination on the crystallite
size growth varied depending on the preparation procedure.
Hence, the samples obtained by precipitation (CC-500 and OW-
500 samples) exhibited a less pronounced growth of crystallite size
(Table 1), increasing by a factor of 1.8–2 in comparison with the
GC-500 sample whose crystallite size was six times greater. On
the other hand, the DC-500 sample showed the largest size with
a value of 38 nm.
(Table 1), thereby showing an overall loss of the total number of
acid sites (15–65%). This feature was aligned with surface sinter-
ing. The sample CC-500 was the oxide that preserved a larger
amount of acid sites (0.34 mmol NH3 gꢀ1) due to a less noticeable
decrease in its surface area. The DC-500 oxide presented the most
dramatic decrease in the amount of acid sites in line with the ob-
served severe sintering. As far as acid strength was concerned, it
could be concluded that the majority of synthesised oxides exhib-
ited equal amounts of weak and strong acid sites. In contrast, the
samples obtained by precipitation (CC and OW) exhibited an
appreciably larger amount of strong acid sites, and their distribu-
tion was not appreciably modified by the activation temperature.
It is important to highlight that to the best of our knowledge, this
is the first time that results on surface acidity of bulk Co3O4 cata-
lysts have been reported in the technical literature. The nature
(Brønsted/Lewis) of the acid sites was investigated by DRIFT spec-
troscopy of adsorbed pyridine at 200 °C. Infrared spectra were
characterised by an intense absorption band at about 1445–
1460 cmꢀ1 associated with the presence of Lewis sites.
The redox properties of cobalt catalysts were investigated by
temperature-programmed reduction with hydrogen. Fig. 3 com-
pares the reduction behaviour of the oxides calcined at 500 °C as
a function of the preparation method. Hydrogen consumption pro-
files were quite similar in most cases, corresponding to a two-step
reduction process initiated at approximately 200 °C. The first peak
was centred at about 250 °C and was associated with the reduction
of Co3+ ions to Co2+ with the concomitantly structural change to
CoO (Eq. (1)), whilst the second peak was located at around
350 °C and attributed to subsequent reduction of CoO to metallic
cobalt (Eq. (2)) [18]. Only for the sample DC-500 was the distinc-
tion between the two peaks not as evident as in the case of the
other cobalt oxides.
The evolution of the textural properties as a function of the cal-
cination temperature in terms of specific surface area, pore volume
and average pore size is shown in Table 1. As for the samples cal-
cined at 300 °C, the surface area varied between 21 and 80 m2 gꢀ1
.
The GC-300 and CC-300 samples had the highest surface area,
whereas the DC-300 sample showed the lowest value. Thermal
treatment at 500 °C in air provoked notable modifications, involv-
ing a progressive drop in the surface area (by 45–85%) and pore
volume along with a significant increment in the pore size in some
cases. In line with XRD analysis, these results suggested that the
samples underwent noticeable sintering. The DC-550 sample was
shown to be the most affected catalyst, with a surface area as
low as 3 m2 gꢀ1. In contrast, both CC-500 and OW-500 displayed
better behaviour in terms of thermal stability. All the adsorption/
desorption isotherms were typical for mesoporous materials, with
significantly reduced hysteresis loops for the oxides activated at
500 °C.
Acid properties were evaluated by means of NH3-TPD analysis,
with the corresponding profiles shown in Fig. 2. The total amount
of desorbed ammonia, expressed as mmol of adsorbed NH3 per
gram of catalyst and estimated by integrating the area of the TPD
curve, was taken as a measurement of the total acid sites concen-
tration. The NH3-TPD profiles revealed the presence of acid sites
with varying strength on each sample. Characteristically, two
desorption peaks were observed at 190–200 and 300–305 °C. A
shoulder at 250 °C was also evident. However, two additional
peaks at 150 and 500 °C were noticed for the OW-500 and CC-
500 samples. Hence, the sum of the areas under the first two peaks
(150 and 190 °C) was assumed to be the number of weak acid sites,
and the sum of the areas under the other two peaks (300 and
500 °C) was associated with the number of strong acid sites. The
results are listed in Table 1.
Co3O4 þ H2 ! 3CoO þ H2O
3CoO þ 3H2 ! 3Co þ 3H2O
ð1Þ
ð2Þ
A quantitative evaluation of the amounts of hydrogen consumed
during reduction (about 170
mol H2 gꢀ1) revealed that in all cases,
l
Co3O4 was reduced almost completely into metallic cobalt at about
400 °C (425 °C for DC-500 sample) As predicted by Eqs. (1) and (2),
the ratio between the areas of the two peaks was about 1:3. To gain
a better understanding of the reducibility of the samples, an at-
tempt was made to quantify the amount of hydrogen consumed
in each step. The area under each curve made it possible to estimate
the relative fraction of reduced cobalt species with temperature. It
was observed that the H2 consumed at low temperatures (200–300 °C)
by CC-500 and OW-500 samples was 25% and 23% of the total
amount (peak H2 uptake at 260–270 °C), respectively, compared
with 10–20% for the other samples (peak H2 uptake at 280–340 °C).
It was observed that the amount of acid sites varied as a func-
tion of the preparation method. Hence, the GB-300 sample exhib-
ited the highest total acidity whilst the DC-300 oxide showed a
considerably reduced acidity. The area under the TPD curves
dropped significantly for all the oxides after calcination at 500 °C
Table 1
Textural, structural and acid properties of the synthesised Co3O4 samples.
Catalyst
BET surface
area (m2 gꢀ1
Pore volume
Average pore
size (Å)
Crystallite
size (nm)
Total acidity
Relative amount of
strong sites (%)
)
(cm3 gꢀ1
)
(mmol NH3 gꢀ1
)
DC-300
DC-500
GB-300
GB-500
GC-300
GC-500
OW-300
OW-500
CC-300
CC-500
SG-300
SG-500
21
3
0.06
0.004
0.30
0.07
0.24
0.05
0.13
0.05
0.25
0.13
0.16
0.04
175
106
160
144
96
314
131
113
159
197
93
18
38
6
16
5
30
5
10
5
0.10
0.03
0.53
0.28
0.27
0.19
0.23
0.16
0.39
0.34
0.28
0.18
46
39
51
49
47
43
80
79
86
85
47
52
80
35
78
18
53
26
78
42
65
20
9
5
13
113

92
B. de Rivas et al. / Journal of Catalysis 281 (2011) 88–97
500
400
300
200
100
GB-500
DC-500
GC-500
GC-300
DC-300
GB-300
400
300
200
100
SG-500
SG-300
OW-500
CC-500
OW-300
CC-300
0
25
50
75 100
0
25
50
75 100 0
25
50
75
100
Time, min
Fig. 2. NH3-TPD profiles of the Co3O4 samples.
SG-500
DC-500
DC-300
GB-500
GB-300
SG-300
CC-500
GC-500
OW-500
OW-300
100
GC-300
100
CC-300
100
0
300
0
300
0
300
500
Temperature, °C
Fig. 3. H2-TPR profiles of the Co3O4 samples.
Furthermore, the reduction of Co3O4 in these two samples started at
noticeably lower temperatures, with a new small peak at about
100 °C, which was not detected for the other cobalt oxides. This
peak was attributed to very active species on the surface formed
during the preoxidation treatment [19]. Therefore, it could be con-
cluded that the synthesis of cobalt oxides having a small crystallite
size favoured the formation of easily reducible sites, which are less
abundant on more highly crystalline Co3O4 catalysts.
For the sake of comparison, TPR profiles of the samples calcined
at 300 °C are also included in Fig. 3. All of them showed a two-step
reduction profile. As expected, it was observed that reducibility
was promoted by about 25 °C due to the lower crystallite size of
this series of oxides. Interestingly, no differences in reduction tem-
peratures were observed between CC-300 and CC-500 samples,
and OW-300 and OW-500 samples. These results were seen to be
aligned to less marked growth of the crystallite size, thus revealing

B. de Rivas et al. / Journal of Catalysis 281 (2011) 88–97
93
Table 2
that both precipitation routes led to relatively stable bulk oxides
T50 and T90 values of DCE oxidation over Co3O4 samples.
without a noticeable inhibition of reduction after calcination at
500 °C.
Catalyst
T50 (°C)
T90 (°C)
DC-500
GB-500
GC-500
OW-500
CC-500
380
305
320
275
250
305
>500
435
370
375
320
300
350
ꢄ500
3.2. Catalytic performance of Co3O4 samples
The oxidation of 1,2-dichloroethane was chosen as a measure to
assess the performance of the prepared Co3O4 catalysts. 1,2-
Dichloroethane was taken as a model for the category of double
carbon chlorinated VOCs with a H/Cl ratio of >1. This compound
is typically ranked intermediate-high on its ease of destruction
when compared, for instance, with double carbon olefins such as
dichloroethylene and trichloroethylene [20]. Before testing the
activity and selectivity of the cobalt oxide catalysts, the transport
effects were evaluated to ensure that experimental results were
not significantly influenced by interphase transportation. Calcula-
tion of the theoretical external transfer rate of reactants to the cat-
SG-500
No catalyst
to attain 50 and 90% conversion, respectively). The light-off curves
are plotted in Fig. 4, whilst the T50 and T90 values are listed in
Table 2. The samples calcined at 500 °C were first examined, since
it was a priori expected that the oxidation process under the tested
experimental conditions would require relatively high reaction
temperatures due to the known stability of chlorinated VOCs.
The results from the homogeneous reaction (the catalyst was re-
placed by inert quartz particles) are also included for comparative
purposes. DCE conversion in the absence of any catalyst did not
start prior to 350 °C and reached only 20% conversion at 500 °C.
In contrast, the use of Co3O4 oxide as catalyst significantly acceler-
ated the desired reaction, and in all cases the chlorinated com-
pound was completely abated between 325 and 500 °C, which
compared very favourably with temperatures in excess of 500 °C,
which would be required for thermal combustion.
Nevertheless, substantial differences were clearly evident
depending on the used synthesis route, thus revealing that the
preparation method to be a key factor for obtaining a highly active
Co3O4 catalyst. Hence, the range at which the 50% conversion was
achieved varied between 250 and 380 °C, whereas the temperature
for complete removal (90% conversion) was in the 325–500 °C
range. The best behaviour was exhibited by the CC-500 sample,
which was obtained via aqueous hydroxycarbonate precipitation,
with a T50 value of 250 °C, whilst the poorest performance was that
of the DC-500 oxide showed (T50 value of 380 °C). The catalytic
activity followed this trend: CC-500 > OW-500 > SG-500 ꢃ GB-500 >
GC-500 > DC-500. In addition to the CC-500 oxide, the OW-500
sample also emerged as an oxide with an attractive behaviour for
DCE oxidation with a T50 value of 275 °C. The precipitation routes
(via sodium carbonate or sodium hydroxide in the presence of
oxygenated water) therefore appeared as adequate methods for
preparing active nanocrystalline cobalt oxides for chlorinated VOC
abatement. The notable activity of these cobalt oxides indicated
that excessively high temperatures were not required.
alytic particles (at
a typical temperature 300 °C) based on
estimated mass-transfer coefficients gave a value three orders of
magnitude larger than the measured reaction rates, indicating pro-
cess conditions were far from external diffusional limitations. The
effects of external mass-transfer resistances were experimentally
evaluated by repeating a set of process conditions whilst employ-
ing a different linear velocity. Results of these experiments indi-
cated that conversion was not affected for linear velocity higher
than 7 cm sꢀ1, to within the experimental error. Likewise, esti-
mates of interphase temperature gradients showed fluid–solid dif-
ferences of less than 1 °C, thereby limiting heat transfer limitation
concerns. The possibility of internal pore diffusion was examined
by measuring conversions at fixed conditions for catalyst particles
of different size. Results showed that pore diffusional resistance
was absent for particles less than 1 mm in diameter. Intraparticle
mass-transfer resistances were theoretically evaluated by comput-
ing effectiveness factors, which were calculated to be greater than
0.98, indicating that intraparticle mass-transfer resistances were
not significant. Finally, internal thermal gradients also proved to
be negligible over the range of conditions evaluated in this study.
The activity of the bulk catalysts was examined by means of the
corresponding ignition or light-off curves between 150 and 500 °C,
characterised by the T50 and T90 parameters (temperature needed
100
GB
OW
DC
80
SG
400
380
360
340
320
300
280
260
240
DC-500
GC-500
60
40
GB-500
CC
GC
Homog.
SG-500
20
OW-500
CC-500
0
150
250
350
450
0
10
20
30
40
50
Temperature,
°
C
Crystallite size, nm
Fig. 4. Light-off curves of DCE oxidation over Co3O4 samples activated at 500 °C in
air.
Fig. 5. Relationship between the crystallite size and T50 value for the Co3O4 samples
activated at 500 °C in air.

94
B. de Rivas et al. / Journal of Catalysis 281 (2011) 88–97
Although all the cobalt oxide catalysts were comprised of the
200
150
100
50
same phase, Co3O4, Fig. 5 suggests that a certain dependence of
activity on crystallite size exists, which in turn is very closely re-
lated to the redox properties of the samples. As a general behav-
iour, a lower crystallite size decreased the temperature for DCE
50% conversion. On the basis of the assumption that the reaction
occurs through a Mars–Van Krevelen mechanism involving lattice
oxygen via a redox cycle [18,21,22], it is believed that the nano-
crystalline CC-500 and OW-500 catalysts presented more facile
(Oꢀ or O2ꢀ) oxygen species that were very reactive for chlorocar-
bon oxidation. As revealed by H2-TPR analysis, the reducibility at
low temperatures was promoted, and a low-temperature feature
on these oxides at 100 °C was also identified, which was not pres-
ent on other Co3O4 samples. The reduced crystallite size of these
samples would give rise to a greater number of low coordination
defect lattice oxygen sites when compared with more highly crys-
talline catalysts, and these may significantly contribute to the high
activity level [11]. The presence of this type of sites is usually re-
lated to the presence of Co2+ in the structure of the spinel. For com-
parative purposes amongst the various synthesised Co3O4 samples,
the relative abundance of Co2+ can be estimated by Raman spec-
troscopy, since a shift of the A1g vibration towards lower frequen-
cies can be taken as a sensitive indication of a highly defective
structure [12]. Typically, five Raman bands at 198, 484, 522, 622
and 694 cmꢀ1 are visible in the range 100–800 cmꢀ1, which corre-
spond respectively to the F12g, Eg, F22g; F33g and A1g modes of crystal-
line Co3O4. Results included in Fig. 6 suggest that both highly active
CC-500 and OW-500 samples (with a more significant shift of the
band, from 694 cmꢀ1 for DC-500 to 692–679 cmꢀ1) could present
a higher concentration of cobalt in low state, which in turn could
indicate oxygen defects.
Taking into account that the second step of the redox model of
Mars–Van Krevelen consists of the re-oxidation of the reduced me-
tal oxide sites by the oxygen present in the gas phase, it is impor-
tant to establish the ability of the catalyst to recover its original
oxidised phase (Co3O4). The oxygen storage complete capacity
(OSCC) is proposed as a meaningful analysis, since it represents
the total amount of O2 consumed after reduction at constant tem-
perature. Fig. 7 includes the OSCC values at 500 °C for some se-
lected samples, namely CC-500, OW-500, SG-500 and DC-500.
These values were taken after isothermal reduction between 150
and 350 °C. It was found that CC-500 sample had a superior oxida-
tion behaviour followed by OW-500. However, the other two
examined oxides a significantly lower O2 uptake after reduction.
In short, analysis from both H2-TPR and OSCC showed a better re-
dox behaviour for CC-500 and OW-500 oxides, resulting in an
excellent catalytic activity.
CC-500
OW-500
SG-500
DC-500
0
150
200
250
300
350
Temperature, °C
Fig. 7. OSCC of selected Co3O4 samples (activated at 500 °C) at varying reduction
temperatures.
the overall behaviour of a given bulk cobalt catalyst. This observa-
tion stems from the following evidence. On one hand, it was found
that both CC-500 and OW-500 samples showed a significantly dif-
ferent activity in spite of having a virtually identical crystallite size
(9–10 nm). This behaviour could be related to the notably higher
acidity of the CC-500 sample (0.34 mmol NH3 gꢀ1) in comparison
with the OW-500 oxide (0.16 mmol NH3 gꢀ1). The relevant cata-
lytic role of acidity was also assessed when analysing the activity
of SG-500 and GB-500 samples. These two oxides showed a similar
T50 value, although the crystallite size of SG-500 was appreciably
smaller (13 nm versus 16 nm for GB-500). It is believed that the
higher overall acidity of GB-500 oxide (0.28 mmol NH3 gꢀ1 in com-
parison with 0.16 mmol NH3 gꢀ1 for SG-500) was responsible for
its notable activity. Acid sites can act as effective chemisorption
sites, thus promoting the conversion of the chlorinated feed.
Numerous evidence is available to demonstrate the importance
of acidity in the overall chlorinated VOC oxidation reaction. For
example, both pure Al2O3 and ZrO2 presented a superior activity
than non-acidic SiO2 [23], but proved to be less active when com-
pared with H-zeolites [24,25]. For the latter class of catalysts, it
was also found that acid properties could be tuned by selective
chemical dealumination for enhancing catalytic performance
[26]. Note that the contribution of active oxygen species from the
surface of these metal-free samples was negligible. Moreover,
when using Ce/Zr or Mn/Zr mixed oxides surface acidity helped
in promoting the activity, which was chiefly controlled by the re-
dox properties (accessible lattice oxygen species) at low tempera-
tures [27,28].
Results included in Fig. 5 also suggest that, apart from crystal-
lite size, other catalytic properties may play a role in determining
3.3. Comparison with other catalysts
Considering the significantly high activity shown by some of the
prepared Co3O4 catalysts, it is important to compare their behav-
iour with other previously used examples for chlorinated VOC
destruction under identical condition in terms of DCE concentra-
tion (1000 ppm in air), total gas flow rate (500 ml minꢀ1) and
DC-500
GC-500
GB-500
ꢀ
ꢁ
ꢀ1
SG-500
OW-500
WHSV 635 g h mol
, namely alumina supported noble metal
DCE
catalysts [29], protonic zeolites [24,25], Ce/Zr mixed oxides
[27,30,31] and Mn/Zr mixed oxides [28]. These results were ob-
tained within the framework of wider-ranging work conducted in
our laboratories on chlorinated VOC oxidation using supported
and bulk catalysts (Table 3). It was observed that cobalt oxide sam-
ples resulted appreciably more active (T50 were lowered by about
15–115 °C with respect to the other catalysts). It is also worth
CC-500
750
700
650
600
550
500
450
400
350
Raman shift, cm-1
Fig. 6. Raman spectra of the Co3O4 samples activated at 500 °C.

B. de Rivas et al. / Journal of Catalysis 281 (2011) 88–97
95
Table 3
100
Comparison of DCE oxidation over Co3O4 samples with catalysts reported in the
literature.
OW
GC
CC
Catalyst
T50 (°C)
Reference
80
60
40
20
H-Y(Si/Al = 2.6)
H-ZSM-5(Si/Al = 27.4)
H-MOR(Si/Al = 5.2)
Dealuminated H-Y(Si/Al = 3.3)
Dealuminated H-Y(Si/Al = 4.3)
Dealuminated H-Y(Si/Al = 6.2)
Dealuminated H-Y(Si/Al = 8.9)
CeO2
Ce0.8Zr0.2O2
Ce0.68Zr0.32O2
Ce0.5Zr0.5O2
Ce0.15Zr0.85O2
ZrO2
Mn0.1Zr0.9O2
Mn0.2Zr0.8O2
Mn0.25Zr0.85O2
Mn0.4Zr0.6O2
Mn0.5Zr0.5O2
Mn2O3
Al2O3
325
275
290
300
280
265
350
345
335
330
310
315
365
375
355
335
305
320
345
380
320
300
275
250
275
[24]
[24]
[25]
[26]
[26]
[26]
[26]
[27]
[27]
[27]
[27]
[27]
[27]
[28]
[28]
[28]
[28]
[28]
[28]
[29]
[29]
[29]
[30]
This work
This work
0
150
200
250
300
Temperature, °C
Pt(0.44wt.%)/Al2O3
Pd(0.42wt.%)/Al2O3
Fig. 8. Light-off curves of DCE oxidation over selected Co3O4 samples activated at
300 °C in air.
j-CeZrO4 (pyrochlore)
Co3O4 nanoparticles (CC catalyst)
Co3O4 nanoparticles (OW catalyst)
3.5. Catalyst stability and characterisation of used samples
pointing out that although H-zeolites showed considerable activity
in DCE abatement with relatively low T50 values (265–325 °C), this
was somewhat deceptive since the chlorinated feed was mainly
converted into chlorinated intermediates, principally vinyl chlo-
ride. In fact, 50–70% of the feed was decomposed to this chlori-
nated species, the complete decomposition of which required
temperatures as high as 450 °C. Interestingly, our work shows that
selected synthesised Co3O4 catalysts can provide a viable alterna-
tive to precious metal-based catalysts, the typical reference cata-
lysts for this application.
The catalyst stability of the samples was investigated by analys-
ing the evolution of conversion in a series of consecutive tempera-
ture cycles from 150 to 500 °C, and the evolution of conversion
with time on stream at 250 °C for 140 h. In the first case, the
behaviour of CC-500, OW-500 and SG-500 samples was evaluated.
The cycle was repeated four times using the same catalyst. Conver-
sion data for each cycle were super-imposable with no change in
the light-off temperature. These observations indicated that the
catalyst exhibited stable activity. The used catalysts were charac-
terised by BET measurements, XRD and EDX (Table 4). The BET sur-
face area was seen to decrease by 15–20%, whilst Co3O4 crystallite
size enlarged significantly, though no evidence of the formation of
CoO, with an expected lower catalytic activity, was noted. More-
over, EDX analysis indicated the presence of a significant amount
of Cl on the samples between 1.5% and 3.5%. All these changes
did not seem to strongly influence the performance of the catalysts.
Fig. 9 shows the evolution of conversion over CC-500 catalyst as a
function of time on stream (140 h) at 250 °C. This temperature
(T50) was selected as it provoked a conversion of less than 100%,
thus providing a more sensitive indication of changes of the cata-
lyst performance with time online. These new experiments
showed that conversion remained stable (50%) with no appreciable
deactivation. By comparing the characterisation results of this
sample with those of the sample submitted to four temperature cy-
cles, it was observed that operation over a prolonged time did not
lead to a larger decrease in surface area or an increase in crystallite
size. Nevertheless, a slightly higher chlorine content was detected
(4.2%). An attempt was made to identify the nature of the chlorine
present in the sample. Basically, two main possibilities exist in
principle, since chlorine may remain adsorbed on the support or
interact with the support forming chlorinated species. Results from
Raman spectroscopy suggested the formation of CoCl2 character-
3.4. Influence of calcination temperature on catalytic performance
In view of the key role played by the Co3O4 crystallite size in the
combustion of chlorinated hydrocarbons, the behaviour of selected
samples calcined at a lower temperature (300 °C for 4 h) was ana-
lysed as well. As previously mentioned, this temperature was the
minimum temperature required to obtain the desired Co3O4 cobalt
oxide. In particular, the study was focused on the following active
catalysts, namely CC-300, OW-300 and GC-300 oxides. Fig. 8 illus-
trates the conversion of DCE with reaction temperature for this set
of bulk cobalt oxides. These catalysts proved to be highly active,
but at 300 °C they only attained 90% conversion. Similar to the con-
version trend seen in the series of samples calcined at 500 °C, the
CC-300 sample again exhibited the highest activity. On the basis
of T50 values, the following conversion pattern was found CC-
300(200 °C) > OW300(225 °C) > GC-300(250 °C). All these catalysts
oxidised DCE at significantly lower temperatures when compared
with their counterparts calcined at 500 °C. These samples calcined
at 300 °C preserved the same and unique structure (cubic Co3O4)
but with a substantial higher surface area and smaller crystallite
size. Sintering was therefore responsible for this detrimental effect
on conversion since it led to an important loss of the amount of
accessible catalytic sites. Whilst T50 increased from 250 to 320 °C
for the GC oxide, a shift of about 50 °C was found for both CC
and OW samples. This behaviour was evidence of better thermal
stability for these two samples, as the loss of both acid and redox
sites was less noticeable compared with that of the GC oxide
(Table 1).
ised by the Raman frequencies at 984 and 1014 cmꢀ1
.
3.6. Product distribution
High conversion is not the only criterion for determining good
chlorinated VOC destruction catalysts. The desired reaction was

96
B. de Rivas et al. / Journal of Catalysis 281 (2011) 88–97
Table 4
Textural, structural and chlorine content of the used Co3O4 samples.
Catalyst
BET surface area (m2 gꢀ1
)
Pore volume (cm3 gꢀ1
)
Average pore size (Å)
Crystallite size (nm)
Chlorine content (%)
DC-500a
GB-500a
GC-500a
OW-500a
CC-500a
SG-500a
CC-500b
2
0.003
0.05
0.04
0.03
0.19
0.04
0.19
194
174
325
114
197
158
198
52
27
41
17
13
22
13
1.4
2.2
2.9
1.5
3.5
3.1
4.1
17
15
17
34
16
34
a
After four consecutive temperature cycles from 150 to 500 °C.
After 140 h time on stream at 250 °C.
b
higher temperatures, significant amounts of molecular chlorine
were also formed due to the occurrence of the Deacon reaction
(2HCl + O2 M Cl2 + H2O). As a consequence, selectivity to Cl2 was
about 60–65% at 450–500 °C. It should be pointed out that all of
chlorine atoms present in the feed were transformed into HCl or
Cl2, and no other chlorinated by-products were obtained at 350–
400 °C.
Product distribution of the oxides activated at 300 °C was also
monitored. It was observed that although at 300 °C the chlorinated
feed was almost completely decomposed, it was also true that sig-
nificant amounts of chlorinated by-products were formed. The oxi-
dation of these compounds needed higher temperatures which, in
all probability, will modify the structure of the catalyst activated at
this low temperature. In contrast, it was seen that for the samples
calcined at 500 °C, destruction of the chlorinated by-products oc-
curred below this temperature (500 °C), which points to the exis-
tence of an optimum temperature for activation between 300
and 500 °C at which the conversion of the parent molecule to deep
oxidation products could be achieved more effectively.
100
80
60
40
20
0
0
20
40
60
80
100
120
140
Time on stream, h
Fig. 9. DCE oxidation over CC-500 sample at 250 °C as a function of time on stream.
Cl-VOC decomposition to deep oxidation products, i.e. CO2 and HCl
or Cl2. It is indeed useful to provide some basic information about
the nature and the amount of reaction products, as the combustion
of chlorinated VOCs may be accompanied by the concomitant pro-
duction of carbon monoxide and highly chlorinated by-products,
sometimes more toxic and recalcitrant than the starting material.
Next, the product distribution corresponding to the CC-500 sample
will be analysed. DCE oxidation gave rise to CO2, HCl and Cl2 as ma-
jor products with a minimal yield of CO. Nevertheless, decomposi-
tion was accompanied by the generation vinyl chloride, methyl
chloride and carbon tetrachloride at mild temperatures (between
175 and 375 °C) according to Eqs. (3)–(7). Vinyl chloride was
formed by dehydrochlorination of the feed molecule, and this
by-product could be further partially oxidised to methyl chloride.
Carbon tetrachloride was formed as a result of the chlorination
of methyl chloride. The peak concentrations were 10–25 ppm
(200–225 °C) for vinyl chloride, 250–275 ppm (175 °C) for methyl
chloride and 80–100 ppm (350–375 °C) for carbon tetrachloride.
4. Conclusions
The catalytic performance of a series of Co3O4 catalysts pre-
pared from a variety of routes was examined for the gas-phase oxi-
dation of 1,2-dichloroethane. This compound was selected as a
model chlorinated VOC due to its widespread use in commercial
applications and distinct toxicity. A wide number of synthesis
routes starting from cobalt(II) nitrate were examined, namely cal-
cination of the precursor salt, solid-state reaction with ammonium
hydrogen carbonate or citric acid, precipitation with sodium
hydroxide and oxygenated water or sodium carbonate and sol–
gel via complexation with citric acid. The resulting oxides were
characterised by thermogravimetry, X-ray diffraction, BET mea-
surements, NH3 temperature-programmed desorption, adsorption
of pyridine followed by infrared spectroscopy, Raman spectros-
copy, temperature-programmed reduction with hydrogen, oxygen
storage complete capacity and energy disperse X-ray analysis.
It was found that although all the routes led to the same phase,
substantial differences existed in terms of surface area, crystallite
size, and redox and acid properties. In particular, a reduced crystal-
lite size, related in turn to improved redox properties at low tem-
peratures, was seen to be the key factor in determining the activity
of the bulk cobalt catalysts. In this sense, of the synthesis proce-
dures investigated, those based on precipitation (via sodium car-
bonate or sodium hydroxide in the presence of oxygenated
water) gave rise to the most active samples with a crystallite size,
after thermal stabilisation at 500 °C, about 10 nm. The considerable
simplicity of this preparation strategy is noteworthy when taking
into account the eventual scale-up of the process. A high total (Le-
wis) acidity was also observed to be relevant for promoting the
conversion of the chlorinated feed. Another important catalytic
feature of Co3O4 catalysts was that they showed an excellent
C2H4Cl2 ! C2H3Cl þ HCl
ð3Þ
ð4Þ
ð5Þ
ð6Þ
ð7Þ
C2H3Cl þ 5=2O2 ! 2CO2 þ HCl þ H2O
C2H3Cl þ O2 ! CH3Cl þ CO2
CH3Cl þ 3=2O2 ! CO2 þ HCl þ H2O
CH3Cl þ 3HCl ! CCl4 þ 3H2
With respect to the conversion to carbon oxides, selectivity towards
CO2 was 100% at 400–500 °C. The active performance of Co3O4 co-
balt oxide for CO oxidation has been previously reported by several
authors [6,7]. These results are in stark contrast with those obtained
with other bulk catalysts such as protonic zeolites or Ce/Zr mixed
oxides. Hence, CO selectivity over the zeolite catalysts was as high
as 50–60% [24,26], whilst CO2 selectivity was about 70–80% over
Ce/Zr oxides [27]. On the other hand, conversion of chlorine atoms
to HCl was favoured at temperatures lower than 350–400 °C. At

B. de Rivas et al. / Journal of Catalysis 281 (2011) 88–97
97
[9] S. Chatterjee, H.L. Greene, Appl. Catal. A 98 (1993) 139–158.
[10] M. Kang, M.W. Song, C.H. Lee, React. Kinet. Catal. Lett. 80 (2003) 123–129.
[11] B. Solsona, I. Vázquez, T. Garcia, T.E. Davies, S.H. Taylor, Catal. Lett. 116 (2007)
116–121.
[12] Q. Liu, L. Wang, M. Chen, Y. Cao, H. He, K. Fan, J. Catal. 263 (2009) 104–113.
[13] C.-B. Wang, C.-C. Lee, J.-L. Bi, J.-Y. Siang, J.-Y. Liu, C.-T. Yeh, Catal. Today 146
(2009) 76–81.
[14] J. Kirchenerova, M. Alifanti, B. Delmon, Appl. Catal. A 231 (2002) 65–80.
[15] R. López-Fonseca, S. Cibrián, J.I. Gutiérrez-Ortiz, M.A. Gutiérrez-Ortiz, J.R.
González-Velasco, AIChE J. 49 (2003) 496–504.
[16] X. Wang, X. Chen, L. Gao, H. Zheng, Z. Zhang, Y. Qian, J. Phys. Chem. B 108
(2004) 16401–16404.
[17] H. Yang, Y. Hu, X. Zhang, G. Qiu, J. Phys. Chem. B 107 (2003) 12643–12649.
[18] B. Solsona, T.E. Davies, T. Garcia, I. Vázquez, A. Dejoz, S.H. Taylor, Appl. Catal. B
84 (2008) 176–184.
100% selectivity towards CO2. On the other hand, HCl and Cl2 were
the major chlorinated oxidation products with reduced amounts of
by-products, which were removed at relatively low temperatures
(350–400 °C). Interestingly, these samples (those obtained via pre-
cipitation) proved to be considerably more active than supported
noble (Pt, Pd) metals, protonic zeolites, and Ce/Zr and Mn/Zr mixed
oxides, and were found to be highly stable despite the retention of
chlorine on the catalyst surface. This behaviour was evidence of the
high potential of this class of catalysts for its widespread applica-
tion in chlorinated VOC abatement systems. Consequently, it is
proposed that future research should be focused on the develop-
ment of new synthesis routes leading to thermally stable samples
with an even lower crystallite size (<10 nm).
[19] X. Xie, W. Shen, Nanoscale 1 (2009) 50–60.
[20] R. López-Fonseca, J.I. Gutiérrez-Ortiz, M.A. Gutiérrez-Ortiz, J.R. González-
Velasco, J. Chem. Technol. Biotechnol. 78 (2003) 15–22.
[21] G. Busca, M. Daturi, E. Finocchio, V. Lorenzelli, G. Ramis, R.J. Willey, Catal.
Today 33 (1997) 239–249.
Acknowledgment
[22] N. Bahlawane, Appl. Catal. B 67 (2006) 168–176.
[23] J.I. Gutiérrez-Ortiz, R. López-Fonseca, U. Aurrekoetxea, J.R. González-Velasco, J.
Catal. 218 (2003) 148–154.
[24] J.R. González-Velasco, R. López-Fonseca, A. Aranzabal, J.I. Gutiérrez-Ortiz, P.
Steltenpohl, Appl. Catal. B 24 (2000) 233–242.
The authors wish to thank the UPV/EHU-Gobierno Vasco (SAIO-
TEK S-PE09UN23) for the financial support.
[25] R. López-Fonseca, A. Aranzabal, P. Steltenpohl, J.I. Gutiérrez-Ortiz, J.R.
González-Velasco, Catal. Today 62 (2000) 367–377.
References
[26] R. López-Fonseca, B. de Rivas, J.I. Gutiérrez-Ortiz, A. Aranzabal, J.R. González-
Velasco, Appl. Catal. B 41 (2003) 31–42.
[27] J.I. Gutiérrez-Ortiz, B. de Rivas, R. López-Fonseca, J.R. González-Velasco, Appl.
Catal. A 269 (2004) 147–155.
[28] J.I. Gutiérrez-Ortiz, B. de Rivas, R. López-Fonseca, S. Martín, J.R. González-
Velasco, Chemosphere 68 (2007) 1004–1012.
[29] J.R. González-Velasco, A. Aranzabal, J.I. Gutiérrez-Ortiz, R. López-Fonseca, M.A.
Gutiérrez-Ortiz, Appl. Catal. B 19 (1998) 189–197.
[30] B. de Rivas, R. López-Fonseca, M.A. Gutiérrez-Ortiz, J.I. Gutiérrez-Ortiz, Appl.
Catal. B 101 (2011) 317–325.
[31] B. de Rivas, R. López-Fonseca, C. Sampedro, J.I. Gutiérrez-Ortiz, Appl. Catal. B 90
(2009) 545–555.
[1] J.J. Spivey, in: G. Ertl (Ed.), Handbook of Heterogeneous Catalysis, vol. 5, Wiley-
VCH Verlag, Weinheim, 2008, pp. 2394–2411.
[2] A. Szegedi, M. Popova, A. Dimitrova, Z. Cherkezova-Zheleva, I. Mitov, Micropor.
Mesopor. Mater. 136 (2010) 106–114.
[3] P. Konova, M. Stoyanova, A. Naydenov, St. Christoskova, D. Mehandjiev, Appl.
Catal. 298 (2006) 109–114.
[4] L.F. Liotta, G. Di Carlo, G. Pantaleo, G. Deganello, Appl. Catal. B 70 (2007) 314–
322.
[5] Y. Xia, H. Dai, H. Jiang, L. Zhang, Catal. Commun. 11 (2010) 1171–1175.
[6] S. Sun, Q. Gao, H. Wang, J. Zhu, H. Guo, Appl. Catal. B 97 (2010) 284–291.
[7] X. Xie, Y. Li, Z-Q. Liu, M. Haruta, W. Shen, Nature 458 (2009) 746–749.
[8] J. Liu, Z. Zhao, J. Wang, C. Xu, A. Duan, G. Jiang, Q. Yang, Appl. Catal. B 84 (2008)
185–195.