Cobalt and cobalt-iron spinel oxides as bulk and silica supported catalysts in the ethanol combustion reaction

Article abstract of DOI:10.1016/j.molcata.2016.11.005

Cobalt and mixed iron-cobalt spinel oxides Co_xFe_(3-x)O₄ supported on SiO₂ ‘IMP-Co_x-500' and their bulk analogues ‘CP-Co_x-500' were prepared via impregnation and co-precipitation methods respectively and calcined at 500 °C. Thermogravimetric analysis ‘ATG/DTG’, X-ray Diffraction method ‘XRD’, Texture measurements ‘BET/BJH’ and Temperature Programmed Reduction technique ‘TPR’ were used for their characterization. Their catalytic behavior in ethanol combustion was investigated. Results showed that nanostructured oxide materials with spinel structure were obtained with a good dispersion of the active phase at the surface of silica in the case of IMP-Co_x-500 compounds. The reduction behavior of the prepared samples is highly dependent on the oxide composition and interaction with the support; which is directly related to the catalytic behavior. The use of supported catalysts improves the catalytic performance and leads the reaction towards a total oxidation by decreasing greatly the formation of partial oxidation products. The obtained values of activation energies are relatively low in comparison with the literature data. A compensation effect was found by carrying out ethanol combustion through the spinel systems considered in this work.

Full text of DOI:10.1016/j.molcata.2016.11.005

Accepted Manuscript
Title: Cobalt and Cobalt-Iron spinel oxides as bulk and silica
supported catalysts in the ethanol combustion reaction
Author: Yasmina Hammiche-Bellal Nabila Zouaoui-Mahzoul
Ibtissem Lounas Amel Benadda Rafik Benrabaa Aline Auroux
Laaldja Boukhobza Amar Djadoun
PII:
S1381-1169(16)30461-7
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.molcata.2016.11.005
MOLCAA 10102
To appear in:
Journal of Molecular Catalysis A: Chemical
Received date:
Revised date:
Accepted date:
12-7-2016
29-9-2016
3-11-2016
Please cite this article as: Yasmina Hammiche-Bellal, Nabila Zouaoui-Mahzoul,
Ibtissem Lounas, Amel Benadda, Rafik Benrabaa, Aline Auroux, Laaldja Boukhobza,
Amar Djadoun, Cobalt and Cobalt-Iron spinel oxides as bulk and silica supported
catalysts in the ethanol combustion reaction, Journal of Molecular Catalysis A:
Chemical http://dx.doi.org/10.1016/j.molcata.2016.11.005
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Cobalt and Cobalt-Iron spinel oxides as bulk and silica supported
catalysts in the ethanol combustion reaction
Yasmina HAMMICHE-BELLAL1,*, Nabila ZOUAOUI-MAHZOUL2, Ibtissem
LOUNAS3, Amel BENADDA1, Rafik BENRABAA1, Aline AUROUX4, Laaldja
MEDDOUR- BOUKHOBZA1, Amar DJADOUN5.
1Laboratoire des Matériaux Catalytiques et Catalyse en Chimie Organique, Faculté de
Chimie, USTHB. BP32 El Alia, Bab Ezzouar, 16111, Alger, Algérie.
2Laboratoire Gestion des Risques et Environnement (LGRE, EA2334), Institut de
Recherche Jean-Baptiste Donnet, Université de Haute-Alsace, 3bis rue Alfred Werner,
68093 Mulhouse Cedex, France.
3Laboratoire de Chimie du Gaz Naturel, Faculté de Chimie, USTHB. BP32 El Alia, Bab
Ezzouar, 16111, Alger, Algérie.
4Université Lyon 1, CNRS, UMR 5256, IRCELYON, Institut de Recherches sur la
Catalyse et l’Environnement de Lyon, 2 Avenue Albert Einstein, F-69626 Villeurbanne,
France.
5Laboratoire de Géophysique, FSTGAT, USTHB. BP32 El Alia Bab Ezzouar 16111,
Alger, Algérie.

Graphical abstract

HIGHLIGHTS
 Pure Co3O4 and CoFe2O4 phases are obtained by co-precipitation method
at calcination temperature 500°C.
 FeCo2O4 material undergoes spinodal decomposition at calcination
temperatures 500-800°C.
 The silica supported solids CoxFe(3-x)O4 (x=1, 2, 3) show an important
increase in the specific surface area compared to the bulk catalysts.
 Supported catalysts display a better catalytic activity than their bulk
analogs in the total ethanol oxidation and Co2+ ions are identified as the
active species.
 The values of the activation energies obtained are relatively low compared
with those of literature.

ABSTRACT
Cobalt and mixed iron-cobalt spinel oxides CoxFe(3-x)O4 supported on SiO2
‘IMP-Cox-500’ and their bulk analogues ‘CP-Cox-500’ were prepared via
impregnation and co-precipitation methods respectively and calcined at 500°C.
Thermogravimetric analysis ‘ATG/DTG’, X-ray Diffraction method ‘XRD’, Texture
measurements ‘BET/BJH’ and Temperature Programmed Reduction technique ‘TPR’
were used for their characterization. Their catalytic behavior in ethanol combustion
was investigated. Results showed that nanostructured oxide materials with spinel
structure were obtained with a good dispersion of the active phase at the surface of
silica in the case of IMP-Cox-500 compounds. The reduction behavior of the prepared
samples is highly dependent on the oxide composition and interaction with the
support; which is directly related to the catalytic behavior. The use of supported
catalysts improves the catalytic performance and leads the reaction towards a total
oxidation by decreasing greatly the formation of partial oxidation products. The
obtained values of activation energies are relatively low in comparison with the
literature data. A compensation effect was found by carrying out ethanol combustion
through the spinel systems considered in this work.
Keywords: Spinel oxides - precipitation - impregnation - ethanol combustion
* Corresponding Author: Y. Hammiche-Bellal Tel: +213 560131918
Fax: +213 21310436
Email address: yasminahammiche@gmail.com

1. Introduction
Volatile organic compounds (VOCs) are a large group of carbon-based
chemicals that easily evaporate at room temperature as alcohols, aldehydes, ketones,
halogenated hydrocarbons, aromatics, alkanes, etc [1]. VOCs are classified as
pollutants; they are toxic and contribute significantly to the formation of the
photochemical smog, which has remarkable impact to the air quality [2]. Regulations
on air quality have been implemented to control VOCs in many countries such as the
European Union, the United States of America and China [3-4]. VOCs originate from
many places such as biogenic sources, solvents, industrial products and automobile
emissions.
The main sources of VOCs are vehicles; they are accounting for 32% of the total
VOCs mass in USA and 35% in Europe [5]. Formation of vehicle emissions comes
from the combustion process and evaporation of the fuel. Ethanol is a renewable fuel
that can be used as an additive to gasoline or its substitute, with the advantage of
octane enhancement and reduced carbon monoxide exhaust emissions [6-7].
However, the exhaust from vehicles fed with such fuel contains residue of unburned
alcohol and some of its partially oxidized derivatives such as aldehydes and ketones,
so it is important to evaluate the catalytic performance in ethanol combustion.
Catalytic combustion is one of the most important and attractive technologies for
reducing volatile organic compounds (VOCs) emissions due to its many advantages,
such as high efficiency, low operating temperatures and extremely low emissions of
NOx and unreacted hydrocarbons [8]. Transition metal oxides are an inexpensive
alternative to conventional catalysts based on noble metals, which are highly active
but rather expensive. Furthermore, they are susceptible to deactivation by poisoning,
especially in the presence of water, one of the combustion products [9-10].

As reported in literature, mixed oxides of transition metals present more
favorable properties compared to single-metal oxides which make them more active
and more resistant to deactivation [11-16]. Ferrite oxides have been widely studied,
and it is well established that ferrites properties depend strongly on the nature and
amount of the substituting metal and also its position in the crystalline structure, [17-
18]. Recently, catalytic properties of MeFe2O4 spinel ferrites (Me = Cu, Cd, Ni, Zn
and Co) in the VOCs combustion were evidenced, opening a new field of application
for these materials [19-20]. Particularly, the cobalt spinel ferrite CoFe2O4 showed the
best reactivity among the studied ferrites in the total oxidation of ethanol.
In this work, we propose a comparative study between bulk and supported
CoxFe(3-x)O4 catalysts (x=1,2,3). the aim is to (i) examine the effect of the structural
phase composition of iron-cobalt spinel and the effect of the use of silica support on
the catalytic properties of the developed solids (ii) correlate the physico-chemical
characteristics to the catalytic activity in the total oxidation of ethanol.
From the phase diagram of the CoFe2O4-Co3O4 system, the thermal stability of
the spinel phase CoxFe(3-x)O4 is closely related to its composition and cobalt oxide
Co3O4 and cobalt ferrite CoFe2O4 present stable single spinel phases up to 900 and
1300°C respectively [21]. Contrarily, the iron cobaltite FeCo2O4 exists as a single-
phase spinel only in the temperature range 850-950°C. Below 850°C, the iron
cobaltite undergoes spinodal decomposition in two average compositional spinel
phases approaching the CoFe2O4 and Co3O4 spinel oxides. The synthesis of these
materials is widely reported in the literature and has been used in different
applications such as photodegradation of dyes, catalytic and lithium storage properties
[22-25]. In this work, we adopted the co-precipitation method for the synthesis of
bulk powders and the incipient wetness impregnation method for the preparation of

supported catalysts. The so-prepared materials were characterized by X-ray
diffraction, thermogravimetric analysis, surface area measurements and temperature-
programmed reduction. Their catalytic activities were evaluated in the reaction of
ethanol combustion.
2. Experimental
a. Samples preparation
Iron nitrate Fe(NO3)3.9H2O (Fluka, 99% ) and cobalt nitrate Co(NO3)2.6H2O
(Merck, 99 %) were used as precursor salts, sodium hydroxide as precipitating agent
and silica SiO2 (Sigma-Aldrich, 99.5%, 211 m2 /g) as support.
For the preparation of bulk catalysts, precursor salts were weighted in the
adequate molar ratios Co/Fe and dissolved in double-distilled water. NaOH aqueous
solution (2M) was added till pH = 12 under constant stirring at room temperature.
After filtering and repeated washing with double-distilled water, the precipitates were
dried at 80°C for 12 h and then calcined at 500°C for 5h with a heating rate of
10°C/min. The as-prepared precursors are named CP-Cox and the calcined samples
CP-Cox-500 (x = 1, 2, 3).
The wet impregnation method was employed to synthesize the CoxFe(3-
x)O4/SiO2 catalysts with 10 wt.% CoxFe(3-x)O4 phase. They were prepared by
impregnating the silica powder with aqueous solutions containing the required
amounts of precursor salts. The solutions were stirred constantly at 80°C up to a
maximum water evaporation. The obtained samples were dried at 80°C for 12 h and
calcined at 500°C for 5h with a heating rate of 10°C/min. The resulting powders will
be named IMP-Cox-500 (x = 1, 2, 3).

b. Characterization techniques
Thermogravimetric analysis (ATG/DTG) of the solids was performed in a
SETARAM TG-DTA12 system; samples were heated from 25 to 900 °C, at a heating
rate of 10 °C/min, in flowing Argon.
The X-ray Diffraction (XRD) pattern analysis was carried out with Cu Kα
radiation, using a Philips X-ray diffractometer equipped with a graphite
monochromator. The XRD patterns were recorded in the 2 range of 10°–90° with a
scanning step of 0.02° and a counting time of 1s. The crystalline phases of the
prepared solids were identified by comparison with the JCPDS files. The unit cell
parameter was obtained using Rietveld analysis of the powder XRD data (Xpowder
2004.04.59 Pro software). The mean crystallite sizes of the phases detected by XRD
were determined from the broadening of the line of attributed peaks using the
Williamson-Hall method.
The nitrogen adsorption-desorption isotherms were recorded on
a
Micromeritics ASAP 2010. Prior analysis, the samples were out gassed overnight
under vacuum at 623 K. The Specific surface areas were determined from the
adsorption-desorption data by BET procedure and the pore size distributions by BJH
method.
Temperature-programmed reduction (TPR) analyses were carried out in a
TPD/R/O-1100 instrument (Thermo-Fisher), using H2/Ar (5% v/v) as reducing gas
mixture, flowing at 20 mL min-1. The reduction behavior of the samples was studied in
the range 40-1000°C at a heating rate of 5°C.min-1, and sample weights were chosen
according to literature recommendations [26-27] to improve the resolution of the peaks.

c. Catalytic activity and kinetic factors determination
The catalytic activity of the solids was evaluated in the total oxidation of
ethanol using a glass fixed bed reactor (30 cm long, inner = 1 cm). Prior to catalytic
tests, solids were activated by heating in flowing air up to 300°C for 15 min, to
remove any eventual physisorbed water molecules. Air was used as carrier gas for
ethanol vapor and as oxidant reactant. The catalyst (0.2 g, catalytic bed height about 2
mm) was tested under a gas hourly space velocity (GHSV) of 30 m3h−1kg−1 in the
temperature range 125-350°C. The reactant mixture was analyzed using a Clarus500
Perkin-Elmer chromatograph equipped with a FID detector and a 20%
carbowax/Chromosorb packed column. Each temperature has been maintained for 90
min during which several analysis of the reaction mixture have been performed.
The conversion of ethanol, X%, is defined as the percentage of ethanol feed
that has reacted, the reported values were calculated after stationary regime was
reached, by the following relation:
X (%) = [(Ai - Ar) / Ai] * 100
’Ai’ and ‘Ar’ are the ethanol chromatographic peaks areas in the reactional mixture
before and after passing through the reactor respectively.
The calculation of selectivity (Si%) and yield (Yi%) was performed for the
partial oxidation products of ethanol, which may be formed during the reaction of
ethanol with oxygen, according to the relation :
Si (%) = (ni / reacted nT) * 100

‘ni’ is the moles number of the product i, ‘reacted nT’ is the total moles number of all
reaction products.
Y% = [X (%) * Si (%)] / 100
The specific reaction rate (V), expressed per gram of catalyst per unit time (s)
is calculated as follows [28] :
V = (C * F) / m for low rate conversions C <0.2
V = (F / m) * (1-C) * Ln (1 / (1-C) for conversion of C> 0.2
C = Conversion rate
F = ethanol flow = 1.085 µmol / s
m: mass of the active phase in the catalyst (0.2 g for a bulk catalyst, 0.02 g for a
supported catalyst)
To access the values of the apparent activation energies (Eapp) from the values
of specific reaction rate, the following classical equation was used:
LnV = (-Eapp / RT) + Ln Aapp
Aapp: apparent Arrhenius constant.
3. Results
a. Characterization of the bulk catalysts
The thermal decomposition of the precursor powders was examined by
TGA/DTA (Figures 1a, 1b and 1c). The curves obtained show basically three main
mass losses associated respectively with:
 An endothermic step in the temperature interval 100-200°C attributed to
removal of physisorbed and chemisorbed water

 An exothermic peak situated between 300 and 500°C, it corresponds to the
decomposition of metal precursors giving rise to the corresponding oxides
CoO and Fe2O3 [7].
 An endothermic decomposition at about 850°C in the case of CP-Co2 sample.
This phenomenon reflects the transformation of the two-phase system in a
single-phase system in accordance with XRD results exposed below.
The X-ray diffractograms of the calcined samples are represented in Figure 2.
The solids CP-Co1-500 and CP-Co3-500 exhibit characteristic peaks of the spinel
phases CoFe2O4 and Co3O4 respectively (F3dm space group). On the other hand,
XRD profiles of the CP-Co2-500 powder display also the existence of diffraction
peaks positioned at 2θ angles consistent with a spinel phase but they are splitted
when calcination temperature ranges between 600-800°C, implying the presence
of two distinct spinel phases; one rich in iron and the second rich in cobalt, in
agreement with the phase diagram, shown in Figure 3, established by S. Hirano
and coll. [21]. Indeed, the composition of the oxides obtained in the temperature
range 600-800°C, are in the miscibility gap in which the material exists in the
form of two spinel phases. Furthermore, the monophasic structure shown by the
diffractogram of the sample calcined at 900°C is in agreement with the phase
diagram; this solid is well located in the spinel single-phase domain. Note that the
spinel species CoFe2O4 and Co3O4 are stable till 1300 and 900°C respectively.
The values of crystallite size of the calcined solids are listed in Table 1, they are
varying between 8.5 and 9.5 nm.
The N2 adsorption–desorption isotherms and pore size distributions are shown
in Figures 4a, 4b, 4c , the textural properties are listed in Table 1. The isotherms

are of type IV with a clear H1-type hysteretic loop, which are characteristic of
mesoporous materials. The mixed oxides CP-Co1-500 and CP-Co2-500 have a
pore volume of 0.21 and 0.19 cm3g-1 respectively and a similar surface area of
about 60 m2g-1 which are greater than those of the single oxide having a surface
area of 42.7 m2g-1 and a pore volume of 0.22 cm3g-1. The pore size distribution
curves show that the pure spinel oxides CP-Co1-500 and CP-Co3-500 have a very
narrow pore size distribution ranging between 10 and 30 nm while the solid CP-
Co2-500 has a wider distribution pore size ranging from 10 to 90 nm.
In order to investigate the reduction behavior of the prepared samples, H2-TPR
analysis was performed. The TPR profiles of the calcined samples are shown in
Figure 5. As it can be seen, the shape of the profiles is similar for all of them but
reduction take place at different temperatures. It can be observed that the peaks
maxima of H2 consumption for the mixed oxides were shifted to higher
temperatures compared to the cobalt oxide reduction peaks, this result is similar
to those reported in the literature [29].
The pure cobalt oxide sample (Co/Fe = 1/0) has been shown to be constituted
by mainly Co3O4 phase, with a TPR profile displaying two main peaks. The first
peak at 315°C corresponds to the reduction of Co3O4 to CoO and the second peak
at 370°C is related to the reduction of CoO to Co0 [30-31].
The CP-Co1-500 sample revealed four reduction peaks that were identified in
our previous study [32] according to literature Data. The reduction peaks at lower
temperatures (328 and 396°C) correspond to the more easily reducible cations (in
this case Co3+ and Co2+) [33], present in the form of cobalt oxide phases [34]. The

last two peaks at 463 and 518°C are assigned to the successive reduction of Fe3+
and Fe2+ oxides species, corresponding to the reduction sequences :
Fe2O3
Fe3O4
Fe [35].
The TPR plot of the solid CP-Co2-500 showed three reduction steps at 398,
444 and 523°C respectively. According to the results previously obtained for the
cobalt ferrite, the peaks at 398 and 523°C are attributed to the reduction of cobalt
species, the last peak at higher temperature result from the reduction of ferric
ions.
One can observe that the reduction temperatures of CP-Co2-500 are relatively
higher than those of CP-Co1-500 meaning that FeCo2O4 phase was more difficult
to be reduced than CoFe2O4 phase. Generally, the shift of the reduction curves to
higher temperature can be ascribed to an increase in the crystallite size [33, 36-
37]. However, the samples CP-Co2-500 and CP-Co1-500 revealed similar
crystallite size and surface area values as shown in Table 1. So, their reducibility
could be assigned to different metal Fe–Co interactions in catalysts structures
which may reshape the TPR profile [37].
b. Characterization of the supported catalysts
X-ray diffraction patterns obtained with the support and the supported
catalysts are shown in Figure 2. The silicon oxide material manifests an amorphous
structure throughout the considered diffraction field. The position of the observed
peaks for the supported catalysts confirmed that spinel phases are present in all the
samples and the support remained unchanged after metal impregnation and
calcination step. As listed in Table 1, the crystallite size of IMP-Co1-500 cannot be

determined by the Williamson-Hall method because of the poorly crystalline phase
obtained.
Recently, Bragança et al. reported the synthesis and characterization of Co-Fe
nanoparticles supported on mesoporous silica [38-39]. They have found that
diffraction peak positions were observed through characteristic lines of Co3O4
crystalline phase for monometallic cobalt based catalyst, while diffraction peaks for
the Co-Fe/SiO2 samples indicated the presence of two phases; Co3O4 and CoFe2O4
spinel oxides. In this work a similar result was obtained for IMP-Co3-500 sample
which showed a well crystallized Co3O4 phase whose diffraction angles are perfectly
consistent with those reported in JCPDS file 421467. For bimetallic iron-cobalt based
catalysts, only lines corresponding to spinel phases have been clearly indentified.
Nevertheless, the presence of secondary phases, namely cobalt spinel oxide Co3O4
and hematite Fe2O3, cannot be excluded but their identification is not obvious due to
small proportions.
…
The N2 adsorption–desorption isotherms and pore size distributions of the
supported catalysts are revealed in Figures 4a, 4b and 4c. Textural data, reported in
Table 1, show an important increase in the specific surface area compared to bulk
catalysts; the values obtained are between 152 and 172 m2g-1. The shape of the
adsorption-desorption isotherms approaches the type IV isotherm with a H3 hysteresis
loop indicating the formation of slit-like mesopores.
The supported catalysts were subjected to a temperature programmed
reduction under the same experimental conditions as bulk catalysts. This analysis

allowed us to examine the influence of the support on the redox properties of the
materials investigated. TPR profiles obtained are shown in Figure 5.
A clear shift of the reduction curves is observed for supported catalysts as
compared to the bulk ones. In addition to the reduction peaks corresponding to cobalt
and iron oxide species previously identified in the study of the bulk solids, the TPR
profiles of the supported catalysts show the presence of peaks, at higher temperatures,
related to the reduction of the spinel phase interacting with the support. This
interaction is of electrostatic type between silanol groups present on the silica surface
and the metal cations [40-41]. The reduction peaks occur at the temperatures maxima
849, 822 and 688°C for IMP-Co1-SiO2, IMP-Co2-SiO2 and IMP-Co3-SiO2
respectively. The peaks of the supported materials are much broader than the bulk
ones, which confirm the existence of a wide particle size distribution as shown by the
BJH analysis [36]. The reduction sequences of IMP-Co2-500 take place at lower
temperatures than those observed for IMP-Co1-500 sample; this may be due to the
higher proportion of cobalt in the former solid. Indeed, Xiangdong Ma et al [42]
observed that the TPR profiles of Fe-Co/SiO2 were shifted toward the lower
temperatures when the cobalt content increased; this was interpreted in terms of
reducibility enhancement by cobalt species in the Fe-Co/SiO2 catalyst.
c. Catalytic activities and reaction kinetics
Figure 6 displays catalytic activity results for the ethanol combustion reaction
over the bulk and supported catalysts.
Depending on the solid composition, ethanol combustion via the bulk catalysts
started at about 150-200°C and full conversion of ethanol was achieved between 275
and 350°C. As expected, the activity of the mixed oxides CoFe2O4 and FeCo2O4 is

generally better than that of the single oxide Co3O4. For a conversion below 15%, the
three oxides have the same reactivity, in contrast to higher conversion levels, where
the reactivity closely depends on the solid composition. Cobalt ferrite exhibits
significantly the best activity; a conversion jump from 30 to about 100% is obtained
via the CP-Co1-500 sample in the temperature range 250-275°C. The activity of iron
cobaltite is intermediary between cobalt ferrite CoFe2O4 and cobalt oxide Co3O4; 90%
of ethanol is converted at 300°C over CP-Co2-500 catalyst against only 50% on CP-
Co3-500 sample, this latter solid achieved a total conversion at about 350°C.
The catalytic performances of the supported solids are greatly different from
that of bulk catalysts. Reactivity is substantially similar, for all the solids, till 30% of
conversion. Further than, the conversion reached 96, 77 and 60% at 275°C, for the
samples IMP-Co2-500, IMP-Co3-500 and IMP-Co1-500 respectively.
Yields of ethanol partial oxidation products were calculated and illustrated in
Figure 7 for the solids CP-Co1-500 and IMP-Co1-500. For all the tested bulk catalysts,
the secondary product detected corresponds to acetaldehyde, which is quite significant
at low temperature and rather small at high temperature. The second partial oxidation
product is acetic acid, present in lower amounts in the whole considered temperature
range. On the supported catalysts, the partial oxidation products are present at very
low levels in the total studied temperature range, indicating that the supported
catalysts orient preferentially the reaction towards a total oxidation even at low
temperatures.
Because of the small proportion of spinel phase in the supported catalysts (a
supported catalyst contains only 10% of spinel phase), comparing reactivity by means
of conversion is not significant, the determination of specific reaction rates is
essential. Their values, expressed per gram of spinel phase per unit time, permit a

meaningful comparison of bulk and supported catalysts activities. The determination
of the specific reaction rates was performed by applying the relations previously
mentioned in section 1.c. The calculations were made at ethanol conversions below
30%. The values obtained are summarized in Table 3.
The determination of apparent activation energies and Arrhenius prefactors
was performed by plotting the curves lnV = f (1/T) as shown in Figure 8. The slope of
the line is ‘-Ea/R’ and the intercept represents the apparent values of ‘In A’; the
values obtained are summarized in Table 4. The apparent activation energies vary
from 26.7 kJ mol-1 (over CP-Co3-500) to 76.9 kJ.mol-1 (over IMP-Co2-500). Note that
the specific reaction rates seen in these plots correspond to the overall ethanol
conversion; their values apply to the determination of the activation energy of the
reaction path of combustion because of the low quantities detected of secondary
products, namely acetaldehyde and acetic acid.
The Constable plot of apparent activation energies of ethanol combustion
reaction over all the studied catalysts is presented in Figure 8. A straight-line
relationship between ln A and Ea was obtained, fitting the Cremer-Constable relation
(ln Aapp = a Eapp + b), i.e. a compensation effect is evidenced, by carrying out the
combustion of ethanol through the spinel systems considered in this work. It should
be mentioned that a compensation effect was also found during oxidation of ethanol
by several authors [43-44].
According to the analysis of Lynggaard et al. [45], a good agreement has been
found between the average slope of the Constable plot ‘a’ and the value of the term
(RT)-1. The temperature range in which measurements were carried out was 200–
250°C assuming an average temperature of 225°C. The (RT)-1 term is 0.241 mol kJ-1,

very close to the slope a of the Constable plot which was determined to be 0.251 mol
kJ-1.
4. Discussion
The less reactive catalyst CP-Co3-500 exhibited also the lowest surface area in
comparison with the mixed oxides; this result is in agreement with previous studies
[2-3]. According to the TPR results, the reduction of CP-Co2-500 sample occurs at
higher temperatures than CP-Co1-500. An easy reducibility reflects directly the
presence of surface defects and therefore oxygen mobility [46-47]. Consequently, an
oxidation process (ethanol combustion in our case) was more difficult on the iron
cobaltite catalyst than on the cobalt ferrite. A good correlation can therefore be
established between the catalytic activity and the redox properties for the bulk
catalysts. This result is in good agreement with those reported in the literature; indeed,
several authors have related the catalysts activity to their reducibility in numerous
studies as Mn-based catalysts for total oxidation of VOCs [48-51] and Co3O4/CeO2,
Co3O4/CeO2–ZrO2 composite oxides in methane combustion [52]. Moreover, the
reactivity of the spinel oxides has also been connected to the cationic distribution.
Cobalt ferrite CoFe2O4 adopts an inverse spinel structure with a cationic distribution
such that 80% of Co ions occupy octahedral sites [53-55], while FeCo2O4 is considered
to be a normal spinel structure in which the octahedral sites are shared between cobalt
and iron ions [56]. According to a number of papers, the octahedral sites are almost
exclusively exposed on spinel crystallites surface and thus, the catalytic activity is
mainly due to the octahedrally coordinated cations [57-59], implying that the catalytic
activity can be improved on systems having active species located in octahedral sites.
In a recent paper, we have attributed the catalytic activity of CoFe2O4 spinel oxide in

the ethanol oxidation reaction to cobalt ions [32]. These species have the lower
reduction temperatures in the Co-Fe-O systems and are therefore responsible for
oxygen mobility on the crystallites surface. So, the weaker activity of FeCo2O4
compared with CoFe2O4 is certainly due to a less proportion of cobalt ions in
octahedral sites and therefore the diminishing of the active species.
A reversal reactivity order is observed between the supported iron cobaltite
and the cobalt ferrite, compared with bulk catalysts but as the reducibility is also
inverted as shown by TPR analysis, the correlation reducibility-catalytic activity is
maintained.
Moreover and unlike the bulk catalysts, the reactivity of IMP-Co3-500 is very
close to that of IMP-Co1-500 as shown by the values of T90 summarized in Table 2.
The difference between the T90 values is reduced from 62°C to 9°C by going from
the bulk catalyst to the supported ones, although the IMP-Co3-500 BET surface area is
lower than that of IMP-Co1-500 (Table 1). The present result implies that the
reactivity of a single oxide can be comparable to that of a mixed oxide when they are
both dispersed on a support; this improvement is certainly related to a similar
dispersion of the active species on the support.
It can be observed that the reactivity of a supported catalyst is clearly better
than that of similar bulk. The active sites are certainly more accessible in the
supported catalyst, due to a better dispersion of the active species on the surface of the
support. This factor is usually in favor of the reactivity of a catalyst.
Data from the literature relating to kinetic parameters of ethanol combustion
reaction are generally rare and almost non-existent for the catalyst systems of spinel
oxides. However, some authors report activation energy values that we took as a
reference for comparison with our results.

Delimaris and Ioannides [43-44] studied the combustion of ethanol over
catalysts based on manganese, copper and cerium, with different molar ratios of
Mn/Ce and Cu/Ce. The values of activation energy, as reported by these authors, are
identified in the field 85-122 kJ / mol for the Mn/Ce catalyst and 101-135 kJ / mol for
the Cu/Ce systems. Oyama et al. [60] report a value of 89 kJ / mol, by carrying out the
combustion of ethanol on manganese-based catalysts. By comparing the results
obtained in this work with those found in the literature, it appears that spinel oxides
are good candidates for catalyzing the ethanol combustion reaction; the values of the
activation energies obtained are relatively low.
5. Conclusion
CoxFe(3-x)O4 materials were synthesized as bulk and silica supported catalysts
using co-precipitation and impregnation methods respectively.
The results show that the operating conditions have a great effect on the
structure and catalytic performance of the synthesized materials.
For the bulk catalysts, cobalt oxide Co3O4 and cobalt ferrite CoFe2O4 powders
display single spinel phases while iron cobaltite FeCo2O4 undergoes spinodal
decomposition into two average compositional spinel phases between the cobalt
ferrite CoFe2O4 and Co3O4 spinel oxide. The catalytic performance in the ethanol
combustion of the mixed oxides CoFe2O4 and FeCo2O4 is generally better than that of
the single oxide Co3O4 and cobalt ferrite exhibits the best activity. The catalytic
activity was attributed to the Co2+ ions; the more easily they are reduce the more
active is the solid in the ethanol combustion. The comparison of specific reaction rates
indicated that the reactivity of a supported catalyst is clearly better than the
corresponding bulk oxide. Moreover, the partial oxidation products namely

acetaldehyde and acetic acid, are formed in lower amounts on the supported catalysts,
indicating that supported catalysts orient further the reaction towards the total
oxidation reaction.
A kinetic study revealed that the values of the activation energies obtained are
relatively low compared to those of literature. Besides compensation effect was
evidenced during oxidation of ethanol with both the bulk and supported spinel oxides
catalysts.
References
[1] F.I. Khan, A.K. Ghosal, J. Loss Prev. Process Indust. 13 (2000) 527–545.
[2] P. Lakshmanan, L. Delannoy, V. Richard, C. Methivier, C. Potvin, C. Louis, Appl.
Catal., B 96 (2010) 117-125.
[3] K. Brodzik, J. Faber, D. Lomankiewicz, G.K Anna, J. Environ. Sci. 26 (2014)
1052-1061.
[4] K.W You, Y.S Ge, B. Hu, Z.W Ning, S.T Zhao, Y.N Zhang, P. Xie, J. Environ.
Sci. 19 (2007) 1208-1213.
[5] H.E. Kebin, Z. Quiang, H.U.O Hong, Point sources of pollution, Vol.1- Vehicular
Emissions.
[6] F.N. Aguero, A. Scian, B. P. Barbero, L. E. Cadus, Catal. Today 133–135 (2008)
493–501.
[7] R. Magnusson, C. Nilsson, B. Andersson, Environ. Sci. Technol. 36 (2002) 1656.
[8] T. Takeguchi, S. Aoyama, J. Ueda, R. Kikuchi, K. Eguchi, Top. Catal. 23 (2003)
159-162.
[9] W.B. Li, J.X. Wang, H. Gong, Catal. Today 148 (2009) 81.
[10] M. S. Kamal, S. A. Razzak, M. M. Hossain, Atmos. Environ. 140 (2016) 117-
134.
[11] P. Doggali, Y. Teraoka, P. Mungse, I. K. Shah, S. Rayalu, N. Labhsetwar, J. Mol.
Catal. A: Chem. 358 (2012) 23– 30.
[12] C. Liu, L. Luo, X. Lu, Kinet. Catal. 49 (2008) 676–681.

[13] F. G. Duran, B. P. Barbero, L. E. Cadus, C. Rojas, M. A. Centeno, J. A.
Odriozola, Appl. Catal., B 92 (2009) 194–201.
[14] W. Tang, X. Wu, S. Li, W. Li, Y. Chen, Catal. Commun. 56 (2014) 134–138.
[15] W. Tang, X. Wu, S. Li, X. Shan, G. Liu, Y. Chen, Appl. Catal., B 162 (2015)
110–121.
[16] A. Białas, M. Mazur, P. Natkanski, B. Dudek, M. Kozak, A. Wach, P.
Kustrowski, Appl. Surf. Sci. 362 (2016) 297–303.
[17] K. E. Sickafus, J. M. Wills, J. Am. Ceram. Soc., 82 (1999) 3279–92.
[18] V. Rusanov, Gushterov, S. Nikolov, A.X. Trautwein, Hyperfine Interact 191
(2009) 67–74.
[19] N. Rezlescu, E. Rezlescu, P.D. Popa, E. Popovici, C. Doroftei, M. Ignat, Mater.
Chem. Phys. 137 (2013) 922-927.
[20] Y. Hammiche-Bellal, A. Benadda, L. Meddour-Boukhobza, S. Barama, A.
Djadoun, A. Barama, Catal. Commun. 42 (2013) 62–67.
[21] S. Hirano, T. Yogo, K. Kikuta, E. Asai, K. Sugiyama, H. Yamamoto, J. Am.
Ceram. Soc. 76 (1993) 1788.
[22] M.M. El-Okr, M.A.Salem, M.S. Salim, R.M. El-Okr, M. Ashoush, H.M. Talaat,
J. Magn. Magn. Mater. 323 (2011) 920–926.
[23] N. N. Binitha, P. V. Suraja, Z. Yaakob, M. R. Resmi, P. P. Silija, J Sol-Gel Sci.
Technol. (2010) 53:466–469.
[24] A. A Mirzaei, A. B. babaei, M. Galavy, A. Youssefi, Fuel Process. Technol. 91
(2010) 335–347.
[25] H. Zhu, Yanfang Sun, Xiao Zhang, Lin Tang, Jinxue Guo, Mater. Lett. 166
(2016) 1–4.
[26] D. A. M. Monti, A. Baiker, J. Catal. 83(1983) 323.
[27] P. Malet, A. Caballero, J. Chem. Soc. Faraday Trans. 84 (1988) 2369.
[28] D. Cornet, Techniques de l’ingénieur, traité de génie des procédés chimiques, J.
12501 (1994) 1.
[29] L. Wang, Y. Hisada, M. Koike, D. Li, H. Watanabe, Y. Nakagawa, K.
Tomishige, Appl. Catal., B 121– 122 (2012) 95– 104.
[30] D. Potoczna-Petru, L. Kepinski, Catal. Lett. 73 (2001) 41-46.
[31] Y. Lv, Y. Li, W. Shen, Catal. Commun. 42 (2013) 116–120.

[32] [50] Y. Hammiche-Bellal, A. Djadoun, L. Meddour-Boukhobza, A. Benadda,
A. Auroux, M.H. Berger, F. Mernache, Materials Mater. Chem. Phys. 177 (2016)
384-397.
[33] E. Manova, D. Paneva, B. Kunev, Cl. Estournes, E. Riviere, K. Tenchev, A.
Leaustic, I. Mitov, J. Alloys Compd. 485 (2009) 356–361.
[34] A. Biabani-Ravandi, M. Rezaei, Z. Fattah, Process Safety and Environmental
Protection 91 (2013) 489–494.
[35] J. Zielinski, I. Zglinicka, L. Znak, Z. Kaszkur, Appl. Catal., A 381 (2010) 191–
196.
[36] E. Manova, T. Tsoncheva, Cl. Estournès, D. Paneva, K. Tenchev, I. Mitov, L.
Petrov, Appl. Catal., A 300 (2006) 170–180.
[37] Q. Yang, H. Choi, S. R. Al-Abed, D. D. Dionysiou, Appl. Catal., B 88 (2009)
462-469.
[38] L.F.F.P.G. Bragança, R.R. Avillez, C.R. Moreira, M.I. Pais da Silva, Mater.
Chem. Phys. 138 (2013) 17-28.
[39] L.F.F.P.G. Braganc, M. Ojeda, J.L.G. Fierro, M.I. Pais da Silva, Appl. Catal., A
423– 424 (2012) 146– 153.
[40] L.T. Zhuravlev, Colloids Surf., A 173 (2000) 1–38.
[41] J. Taghavimoghaddam, G. P. Knowles, A. L. Chaffee, J. Mol. Catal. A: Chem.
358 (2012) 79– 88.
[42] X. Ma, Q. Sun, W. Ying, D. Fang, Journal of Natural Gas Chemistry 18 (2009)
232–236.
[43] D. Delimaris, T. Ioannides, Appl. Catal., B 84 (2008) 303.
[44] D. Delimaris, T. Ioannides, Appl. Catal., B 89 (2009) 295.
[45] H. Lynggaard, A. Andreasen, C. Stegelmann, P. Stoltze, Prog. Surf. Sci. 77
(2004) 71.
[46] N. Ballarini, F. Cavani, S. Passeri, L. Pesaresi, A. F. Lee, K. Wilson, Appl.
Catal., A 366 (2009) 184–192.
[47] R. Dumitru, F. Papa, I. Balint, D. C. Culita, C. Munteanu, N. Stanica, A.
Ianculescu, L. Diamandescu, O. Carp, Appl. Catal., A 467 (2013) 178– 186.
[48] Q. Tang, C. Wu, R. Qiao, Y. Chen, Y. Yang, Appl. Catal., A 403 (2011) 136
141.

[49] S.A. Hosseini, A. Niaei, D. Salari, S.R. Nabavi, Ceram. Int. 38 (2012) 1655–
1661.
[50] R. Dula, R. Janik, T. Machej, J. Stoch, R. Grabowski, E.M. Serwicka, Catal.
Today, 119 (2007) 327-331.
[51] A. de la Peña O'Shea, M.C. Alvarez-Galvan, J.L.G. Fierro, P.L. Arias, Appl.
Catal., B, 57 (2005) 191-199.
[52] L.F. Liotta, G. Di Carlo, G. Pantaleo, G. Deganello, Catal. Commun. 6 (2005)
329-336.
[53] V. A. M. Brabers, Handbook of Magnetic Materials, vol. 8 (1995).
[54] L. Braicovich, A. Tagliaferri, G. van der Laan, G. Ghiringhelli, N. B. Brookes,
Phys. Rev. Lett. 90 (2003) 117401.
[55] N.S. Gajbhiye, S. Bhattacharyya, G. Balaji, R.S. Ningthoujam, R.K. Das, S.
Basak, J. Weissmüller, Hyperfine Interact. 165 (2005) 153-159.
[56] P.J. Murray, J.W. Linnett, J. Phys. Chem. Solids 37 (1976) 619.
[57] K. Lazar, T. Mathew, Z. Koppany, J. Megyeri, V. Samuel, S.P. Mirajkar, B.S.
Rao, L. Guczi, Phys. Chem. Chem. Phys. 4 (2002) 3530–3536.
[58] T. Mathew, S. Shylesh, S.N. Reddy, C.P. Sebastian, S.K. Date, B.S. Rao, S.D.
Kulkarni, Catal. Lett. 93 (2004) 155–163.
[59] G. Maniak, P. Stelmachowski, J.J. Stanek, A. Kotarba, Z. Sojka, Catal. Commun.
15 (2011) 127–131.
[60] S.T. Oyama, W. Li, W. Zhang, Stud. Surf. Sci. Catal. 121 (1999) 105.

Figure Captions

Figure 1 : TGA/DTA curves of as prepared precursors for (1a) CoFe2O4, (1b)
FeCo2O4, (1c) Co3O4.

Figure 2: XRD diffractograms of bulk and supported calcined solids.

Figure 3: phase diagram of the system CoFe2O4 / Co3O4 (S. Hirano and coll. [21]).