Kinetics of oxychlorination of chromite Part I. Effect of temperature

Article abstract of DOI:10.1016/S0040-6031(01)00414-2

Nonisothermal thermogravimetric (TG) analysis was used to evaluate the reactivity of a chromite mineral [(Fe²⁺, Mg)(Cr, Al, Fe³⁺)₂O₄] towards Cl₂ + CO, Cl₂ + N₂ and Cl₂ + O₂ gaseous mixtures up to 1000°C. Full chlorination and volatilization of reaction products were achieved at about 975°C using Cl₂ + CO, while only about 40% of the sample have reacted at 1000°C using Cl₂ + N₂ and Cl₂ + O₂. The effect of the temperature on the oxychlorination with Cl₂ + O₂ of the chromite mineral was studied between 600 and 1050°C using isothermal TG measurements. The results show that the oxychlorination of chromite occurs in two stages. The initial stage of the oxychlorination was characterized by average values of apparent activation energy of about 151 and 57 kJ/mol for the temperatures lower and higher than 825°C, respectively. While a value of about 262 kJ/mol was found for the second stage of the oxychlorination process between 925 and 1050°C. The effects of temperature on the oxychlorination of the simple chromite constituents (Cr₂O₃, Fe₂O₃ and MgO) were also studied.

Full text of DOI:10.1016/S0040-6031(01)00414-2

Thermochimica Acta 371 (2001) 143±154
Kinetics of oxychlorination of chromite
Part I. Effect of temperature
N. Kanaria, I. Gaballaha,*, E. Allainb
Mineral Processing and Environmental Engineering Team, Laboratoire Environnement et Mineralurgie,
a
Â
Â
Â
Â
Centre National de la Recherche Scienti®que UMR 7569, Ecole Nationale Superieure de Geologie,
Institut National Polytechnique de Lorraine, Boite Postale 40,
Rue du Doyen Marcel Roubault, 54501 Vandúuvre Cedex, France
bSchool of Mines and Metallurgy, Center for Pyrometallurgy, University of Missouri±Rolla,
210 Fulton Hall, 65401 Rolla, MO, USA
Received 10 August 2000; accepted 15 December 2000
Abstract
Nonisothermal thermogravimetric (TG) analysis was used to evaluate the reactivity of a chromite mineral [(Fe2‡, Mg)(Cr,
Al, Fe3‡)2O4] towards Cl2 ‡ CO, Cl2 ‡ N2 and Cl2 ‡ O2 gaseous mixtures up to 10008C. Full chlorination and volatilization
of reaction products were achieved at about 9758C using Cl2 ‡ CO, while only about 40% of the sample have reacted at
10008C using Cl2 ‡ N2 and Cl2 ‡ O2.
The effect of the temperature on the oxychlorination with Cl2 ‡ O2 of the chromite mineral was studied between 600 and
10508C using isothermal TG measurements. The results show that the oxychlorination of chromite occurs in two stages. The
initial stage of the oxychlorination was characterized by average values of apparent activation energy of about 151 and 57 kJ/
mol for the temperatures lower and higher than 8258C, respectively. While a value of about 262 kJ/mol was found for the
second stage of the oxychlorination process between 925 and 10508C. The effects of temperature on the oxychlorination of the
simple chromite constituents (Cr2O3, Fe2O3 and MgO) were also studied. # 2001 Elsevier Science B.V. All rights reserved.
Keywords: TG analysis; Oxychlorination of chromite; Apparent activation energy
1. Introduction
crust as oxides, a reducing agent (C, CO) was
employed in the chlorination process to increase its
ef®ciency.
Chlorine technology is used for the recovery of
several metals such as Ti, Zr, Hf, Nb, Ta, rare-earth
elements, etc., from their bearing materials. Korshu-
nov [1] described, in a recent overview, the theoretical
aspects of chlorination, the methodology and the
apparatus used as well as the industrial application
of the chlorine technology. As the majority of the
above-mentioned elements are present in the earth
However, chlorination in presence of oxygen (oxy-
chlorination) was applied to separate selectively a
group of elements from other ones. In certain cases,
the presence of oxygen in the system modi®es the
thermodynamic feasibility of speci®c oxychlorination
reactions. Recently, Gaballah and Djona [2] extracted
Co, Ni, Mo, and V chlorinated compounds from spent
hydrodesulfurization catalysts using Cl2 ‡ air gas
mixtures without chlorinating the catalyst support
(Al2O3). In previous investigations [3±6], oxychlori-
* Corresponding author. Tel.: ‡33-383-596-336;
fax: ‡33-383-569-585.
0040-6031/01/$ ± see front matter # 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 0 - 6 0 3 1 ( 0 1 ) 0 0 4 1 4 - 2

144
N. Kanari et al. / Thermochimica Acta 371 (2001) 143±154
nation was employed to upgrade the chromite con-
centrates as well as the partial recovery of chromium
as chromium oxychloride (CrO2Cl2).
out in a horizontal experimental setup described also
in [7].
This paper is devoted to the study of temperature
effect on the oxychlorination of chromite mineral
using experimental conditions that minimize the effect
of mass transfer phenomenon. To our knowledge, no
studies concerning the oxychlorination of chromite
were reported in the literature.
3. Results and discussion
3.1. Nonisothermal TGA treatment in various
chlorinating atmospheres
This study was started by nonisothermal TGA
treatment of the chromite mineral in Cl2 ‡ CO
ꢀCl2=CO ˆ 1†, Cl2 ‡ N2 ꢀCl2=N2 ˆ 1†, and Cl2‡
O2 ꢀCl2=O2 ˆ 4† gaseous mixtures to evaluate its
reactivity towards these chlorinating gaseous mixtures
between 25 and 10008C. The sample's heating rate
was about 78C/min. Fig. 1(a) shows the evolution of
the percent weight loss (%WL) of chromite sample as
a function of temperature for the chosen gaseous
mixtures. At 10008C, about 40% of the sample were
chlorinated in Cl2 ‡ N2 and/or Cl2 ‡ O2 gas mixtures
indicating a relatively low reactivity of chromite
towards these chlorinating gaseous mixtures. While
full chlorination of the chromite was possible, at
9758C, using Cl2 ‡ CO gas mixture.
2. Materials and experimental procedure
The sample used in this study is a chromite mineral
obtained from its concentrate described in detail pre-
viously [7]. Microprobe analyses revealed the follow-
ing weight composition of this mineral reported as
simple oxides: 59.4% Cr2O3, 13.9% FeO, 14.2%
MgO, 9.5% Al2O3, and 0.1% SiO2. The ®rst four
oxides are part of the spinel structure of chromite.
Based on the microprobe data, the general formula of
2‡
the chromite sample could be given as: (Fe0:30
,
3‡
Mg0.70)(Cr1.56, Al0.37, Fe0:07 )O4. On the other
hand, it could be represented as spinels having the
following weight composition: 30.9% FeCr2O4,
51.0% MgCr2O4, 13.7% MgAl2O4 and 4.4% Fe3O4.
The thermogravimetric (TG) chlorination tests were
performed using 40 mg of the sample and the experi-
mental setup described earlier [7]. Some oxychlorina-
tion tests, using several grams of sample, were carried
The derivative of weight loss versus temperature is
reported in Fig. 1(b). Two maximums of reaction rate
at about 655 and 8908C are observed during the
carbochlorination of the sample. The chlorination
and the oxychlorination of chromite sample proceed
with the maximal reaction rates at about 725 and
Fig. 1. TG analysis of chromite: (a) in different chlorinating gas mixtures; (b) derivative of weight losses versus temperature.

N. Kanari et al. / Thermochimica Acta 371 (2001) 143±154
145
7858C, respectively. Beyond these temperatures, the
reaction rate decreases sharply with temperature
increase. Examination of Fig. 1(a) and (b) revealed
that, at least, two different constituents of chromite
possess different reactivities towards the three chlor-
inating gas mixtures. As ®rst approximation, it seems
that Fe3O4 and FeCr2O4 were the ®rst to react with
reactive gases. The calculated weight loss for the
formation [10]. Thus, the rate of the chlorides' vola-
tilization will be higher than that of their formation
during the oxychlorination of chromite and conse-
quently the measured weight loss is directly related to
the oxychlorination rate of the sample.
3.2. Isothermal TGA treatment in Cl2 ‡ O2
Ch
chlorination of these compounds (L1 in Fig. 1(a))
3.2.1. Oxychlorination of chromite at 10508C
agrees with the data obtained experimentally at
ꢁ9008C during the reaction of Cl2 ‡ N2 and
Cl2 ‡ O2 gas mixtures with chromite. MgCr2O4 and
MgAl2O4 appear to be less reactive to chlorine in
absence of a reducing agent. This is in good agreement
with thermodynamic predictions [4]. In the case of
carbochlorination, CrCl3, FeCl2 or FeCl3, MgCl2,
AlCl3 are the most probable reaction products of
chromite with Cl2 ‡ CO. Details concerning the
kinetic aspects of the chromite carbochlorination
are given in previous paper [7].
The main products of the chromite oxychlorination
are expected to be CrO2Cl2, FeCl3, and MgCl2. The
evolution of the vapor pressure of chlorides as a
function of temperature is given in Fig. 2 [8,9]. This
®gure suggests that the chromium oxychloride is
volatile at low temperatures and ferric chloride pos-
sesses a vapor pressure of about 1 atm at temperatures
approaching 3008C. On the other hand, using a
Cl2 ‡ O2 with a Cl2/O2 molar ratio equal to 4 led
to a volatilization rate of MgCl2 higher than its rate of
To get an insight about the effect of grain size on the
chromite oxychlorination, isothermal treatment with
Cl2 ‡ O2 gas mixture using ground chromite samples
having a particle size of <10 and 125±180 mm was
conducted at 10508C. The evolution of the sample
%WL versus time is traced in Fig. 3. The time scale is
graduated in minutes up to 30 min and the rest of
chlorination time is given in hours. About 35% of
sample are chlorinated and volatilized for a reaction
time of about 15 min. Then, the reaction rate falls
sharply and about 12 h are required to reach 88%WL
whereas the weight loss is stabilized. In both cases, the
reaction rate is slightly affected by the particles' size.
The oxychlorination residues corresponding to the
extent of reaction `X' (weight ratio of the reacted
fraction to its initial weight) equal to 0.40, 0.60,
0.80 and 0.88 (points a, b, c and d of Fig. 3) were
analyzed by scanning electron microscopy (SEM) and
X-ray diffraction (XRD). The SEM spectrums of these
residues, as well as that of the raw sample, are grouped
in Fig. 4. The residue of point `a' ꢀX ˆ 0:40† is free of
Fig. 2. Evolution of vapor pressure of several chlorides as a function of temperature [8,9].

146
N. Kanari et al. / Thermochimica Acta 371 (2001) 143±154
Fig. 3. Isothermal oxychlorination with Cl2 ‡ O2 of two grain sizes of chromite at 10508C.
Fig. 4. SEM results of oxychlorination residues of chromite at 10508C for different reaction extents.

N. Kanari et al. / Thermochimica Acta 371 (2001) 143±154
147
iron. However, full elimination of iron oxides cannot
justify a reaction extent of 0.40 as the iron oxides'
content of the chromite is about 14%. Only the
chlorination of both FeCr2O4 and Fe3O4 is in good
agreement with an extent of reaction of 0.40 (refer to
3.2.2. Effect of gas ¯ow rate
The effect of the gas ¯ow rate on the reaction rate
was measured to determine the experimental condi-
tions that minimize the effect of mass transfer phe-
nomena. A series of experiments was performed at
10008C using Cl2 ‡ O2 gas mixtures with a Cl2/O2
molar ratio equal to 4. The total gas ¯ow rate was
varied from 15 to 75 l/h. Fig. 5(a) traces the evolution
of the %WL versus temperature. After about 35%WL,
a dramatic decrease of the sample's reactivity was
observed. This decrease is almost independent of the
gas ¯ow rate. The evolution of the reaction rate as a
function of the gas ¯ow rate was determined for
Ch
L1 in Fig. 3). The SEM spectrums of points `b' and
`c' show the decrease of Cr peak intensity and the
increase of Mg and Al ones. Finally, the SEM spec-
trum of points `d' is essentially composed of Al and
Mg without Cr. The evolution of elemental composi-
tion of residues from `b' to `d' indicates that MgCr2O4
was chlorinated leaving MgAl2O4 that is least reactive
compound of chromite towards Cl2 ‡ O2 gas mixture.
Ch
The %WL calculated (refer to L2 in Fig. 3) for the
reaction
extents
0:10 ꢂ X ꢂ 0:25
and
chlorination of ꢀFeCr2O4 ‡ Fe3O4 ‡ MgCr2O4† were
in fair agreement with the observations of Figs. 3 and
4. It is important to note that chlorine was not detected
in the chlorination residues. This con®rmed the
hypothesis that the reaction rate of chromite with
Cl2 ‡ O2 was lower than the rate of volatilization
of generated products.
XRD diagrams of the residues corresponding to
points `a' to `d' were almost identical. This is probably
due to the fact that the different spinels (FeCr2O4,
Fe3O4, MgCr2O4, MgAl2O4) have the same lattice
parameters making their identi®cation almost impos-
sible. Moreover, the simultaneous removal of divalent
(Fe2‡, Mg) and trivalent (Fe3‡, Cr) elements from the
chromite structure seems to compensate the ionic radii
effects on the lattice parameters.
0:40 ꢂ X ꢂ 0:50 as shown in Fig. 5(b). The gas ¯ow
rate has almost no effect on the oxychlorination
reaction rate for a gas velocity (Vg) higher than
40 cm/min. For this reason, determination of the
temperature effect on the chromite oxychlorination
was performed using a chlorinating gas mixture hav-
ing a Vg ꢁ 49:6 cm=min.
3.2.3. Isothermal oxychlorination of chromite at
different temperatures
A series of TGA experiments was carried out in the
temperature range 600±10508C. As the reaction rate
was low for X > 0:35, the experiments carried out up
to 9008C are stopped when the reaction extent of 0.35
is approached. Fig. 6(a) and (b) trace the evolution of
the %WL of the chromite sample versus time for
Fig. 5. Isotherms of the chromite oxychlorination using: (a) different gas ¯ow rate; (b) its effect on the reaction rate.

148
N. Kanari et al. / Thermochimica Acta 371 (2001) 143±154
Fig. 6. Isotherms of the chromite oxychlorination with Cl2 ‡ O2 between: (a) 600 and 8008C ꢀX ꢂ 0:30†; (b) 850 and 10508C ꢀX ꢂ 0:30†; (c)
925 and 10508C ꢀX ꢂ 0:80†.
several isotherms performed for reaction extents lower
than 0.35. It is clear from these ®gures that the time
required to reach a given reaction extent decreases
with the temperature increase. Data of the isothermal
oxychlorination of the chromite at temperatures
higher than 9008C and for reaction extent ꢂ0.80
are displayed in Fig. 6(c). After a rapid weight loss
at the beginning of the reaction, the reaction rate
decreases sharply. At 9508C and for a reaction extent
higher than 0.35, the reaction rate drops by two order
of magnitudes.
The variation of Ea values versus the reaction extent is
shown in Fig. 7 between 925 and 10508C. The begin-
ning of the reaction is characterized by an Ea of about
80 kJ/mol and decreases slightly with the progress of
reaction up to X ꢂ 0:25. An important increase of Ea is
observed for X > 0:35 and the average value of acti-
vation energy is about 265 kJ/mol for 0:40 < X
< 0:70. It should be pointed out that the important
change of the apparent activation energy coincides
roughly with the chlorination of iron chromite :
Ch
FeCr2O4 (refer to L1 in Fig. 6(c)). An example of
The temperature dependency of the reaction rate
was determined according to Arrhenius diagram. It
was interesting to follow the evolution of apparent
activation energy (Ea) as a function of reaction extent.
Arrhenius diagram for the oxychlorination of chromite
sample is shown in Fig. 8. The apparent activation
energy is calculated for 0:00 ꢂ X ꢂ 0:25 between 600
and 10508C and for 0:40 ꢂ X ꢂ 0:50 between 925 and

N. Kanari et al. / Thermochimica Acta 371 (2001) 143±154
149
the evolution of the reaction extent as a function of
time for the reaction extents lower and higher than
0.30. Eq. (1) seems to be the most appropriate to
describe the evolution of the reaction extent
ꢀX ꢂ 0:30† as a function of time for the oxychlorina-
tion of chromite between 600 and 8008C. While,
Eq. (2) matches well the oxychlorination data
ꢀX ꢂ 0:30† for the temperature ranges from 825 to
10508C (Fig. 9(b)). Similarly, Fig. 9(c) gives the data
linearization by using Eq. (2) of the chromite oxy-
chlorination isotherms for X ꢂ 0:30 between 925 and
10508C.
2=3
3ꢀ1 X† ‡ 2ꢀ1 X† ˆ kt
1
(1)
(2)
X ‡ ꢀ1 X† lnꢀ1 X† ˆ kt
Fig. 7. Evolution of the activation energy as a function of the
reaction extent.
where k is the constant and t the chlorination time.
Eqs. (1) and (2) describe pore diffusion control for
spherical and cylindrical pores' shape, respectively, if
either the initial solid or the reaction product are
porous. Microscopic examination of the chromite
samples at 10008C, for different reaction extents,
showed that the particles become porous with the
progress of the reaction. On the other hand, the small
dependency of the reaction rate on the particles' size
of the sample (Fig. 3) is probably due to the pore
formation inside the particles, due to gradual extrac-
tion of iron and chromium from the chromite. This is
also supported by the fact that the particles' size
decreased slightly as the reaction progress for reaction
extents of about 0.80.
10508C. The oxychlorination of chromite for
X ꢂ 0:25 is characterized by an Ea value of about
151 and 57 kJ/mol for temperatures lower and higher
than 8258C, respectively. As shown by Fig. 8, the
apparent activation energy for 0:40 ꢂ X ꢂ 0:50 is
about 262 kJ/mol. This high value of Ea indicates that
the oxychlorination of magnesium chromite is
strongly dependent on temperature probably due to
solid state diffusion.
Mathematical ®tting of experimental data of iso-
thermal oxychlorination of chromite was performed.
Different kinetics equations [11] were used to describe
However, it is dif®cult to clarify the kinetics of the
chromite oxychlorination due to complex composition
and structure of the chromite. Thus, a mixed regime of
pore diffusion and chemical reaction could control the
oxychlorination. Further clari®cation of the kinetics
phenomena of the oxychlorination was performed by
initiating another series of TGA tests of chromite
oxychlorination between 700 and 10408C using a
Cl2 ‡ O2 gas mixture with a Cl2/O2 molar ratio equal
to 0.5. In other words, the partial pressure of chlorine
was decreased, leading probably to a decrease of
reactivity of the chlorinating gaseous mixtures and
limiting the pore diffusion phenomena.
Fig. 10(a) and (b) group the experimental data
obtained for the oxychlorination of chromite with
Cl2 ‡ O2 ꢀCl2=O2 ˆ 0:5† for X ꢂ 0:30. In these con-
ditions, the effect of the temperature on the chromite
Fig. 8. Example of Arrhenius plot of the chromite oxychlorination.

150
N. Kanari et al. / Thermochimica Acta 371 (2001) 143±154
Fig. 9. Mathematical ®tting of the data of the chromite oxychlorination for: (a) T ꢂ 800ꢃC, X ꢂ 0:30 using Eq. (1); (b) T ꢄ 825ꢃC, X ꢂ 0:30
using Eq. (2); (c) T ꢄ 925ꢃC, X ꢄ 0:35 using Eq. (2).
Fig. 10. Isotherms of the chromite oxychlorination with Cl2 ‡ O2 having a Cl2/O2 molar ratio equal to 0.5 for: (a) T ꢂ 800ꢃC; (b) T ꢄ 850ꢃC.

N. Kanari et al. / Thermochimica Acta 371 (2001) 143±154
151
than 8258C, the value of the Ea obtained using
Cl2 ‡ O2 with Cl2=O2 ˆ 0:5 is higher (85 kJ/mol)
than that deduced when Cl2/O2 ratio in the gas mixture
was 4. The increase in the value of Ea could be related
to a mixed regime controlling the reaction mechanism.
The mathematical ®tting of experimental data for
the chromite oxychlorination with Cl2 ‡ O2
ꢀCl2=O2 ˆ 0:5† is represented in Fig. 12. As shown
by Fig. 12(a), experimental data are lineraized using
Eq. (1) for reaction temperatures lower or equal to
8008C. Fig. 12(b) gives the evolution of the reaction
extent versus time related by Eq. (3) for the oxychlor-
ination of chromite at temperatures equal or higher
than 8508C. This result combined to the decrease of Ea
value from 150 to 85 kJ/mol con®rm that the oxy-
chlorination reaction rate of chromite is controlled by
a mixed regime when using Cl2 ‡ O2 ꢀCl2=O2 ˆ 0:5†
at temperatures higher than 8008C.
Fig. 11. Comparison of Arrhenius plots of the chromite oxychlor-
ination with Cl2 ‡ O2 gaseous mixtures having different Cl2/O2
molar ratios.
1=2
1
ꢀ1 X† ˆ kt
where k is the constant and t the chlorination time.
(3)
oxychlorination was determined by the Arrhenius
diagram.
The Arrhenius diagrams of the oxychlorination of
chromite ꢀX ꢂ 0:25† for both oxychlorination gas
mixtures are compared in Fig. 11. As could be
expected, the reaction rate using Cl2 ‡ O2 with
Cl2=O2 ˆ 4 is higher than that obtained using
Cl2 ‡ O2 with Cl2=O2 ˆ 0:5 in the explored tempera-
ture range. The values of the apparent activation
energy at temperatures lower than 8258C are equiva-
lent in both cases. However, at temperatures higher
3.2.4. Effect of temperature on the oxychlorination of
the simple chromite oxides
Although the simple oxides of chromite mineral are
combined in the spinel structure (Fe2‡, Mg)(Cr, Al,
Fe3‡)2O4, it is worthwhile to study the effect of
temperature on the oxychlorination of simple oxides:
Cr2O3, Feꢀ1 O, Fe2O3, MgO and Al2O3. Their com-
x†
position and structures were determined by SEM and
XRD analysis [4]. The tests were carried out in boat
Fig. 12. Mathematical ®tting of the oxychlorination data at: (a) T ꢂ 800ꢃC; (b) T ꢄ 850ꢃC using Eqs. (1) and (3), respectively.

152
N. Kanari et al. / Thermochimica Acta 371 (2001) 143±154
Fig. 13. Oxychlorination of chromite and its simple oxides with Cl2 ‡ air between 600 and 10008C for 2 h using boat experiments.
experiments using the horizontal setup. Several grams
were used for each test. An oxychlorinating gas
mixture composed of Cl2 ‡ air with a Cl2/air molar
ratio equal to 1 was used and the oxychlorination time
was 2 h. The results are given in Fig. 13 as evolution of
%WL versus time. This ®gure contains also the results
of the oxychlorination of chromite in the same experi-
mental conditions as those used in the case of simple
oxides.
Details concerning the behavior of these oxides in
the oxychlorinating gas mixture were given previously
[4]. Iron oxides are the most reactive towards the
oxychlorinating gas mixture and almost full reaction
was obtained at about 9008C. The reaction product
was identi®ed as FeCl3. The reactivity of Cr2O3 is
lower than that of iron oxides especially at high
temperatures and the reaction product was chromium
oxychloride (CrO2Cl2). Magnesium oxide seems to be
inert up to about 8508C and only 20% of sample had
reacted at 10008C generating magnesium chloride. No
reaction of Al2O3 with the oxychlorinating gas mix-
ture was observed even at 10008C. Chromite started to
react with Cl2 ‡ air at about 6008C and less than 30%
of the sample had reacted at 10008C during 2 h. This
series of experiments revealed roughly the reactivity
of these oxides towards oxychlorinating gas mixtures
as well as the main reaction products.
while minimizing the mass transfer phenomena. WuÈs-
tite ꢀFeꢀ1 O† was not tested because it is oxidized
x†
rapidly into hematite (Fe2O3) during its chlorination.
Aluminum oxide was not taken into consideration due
to its refractive nature towards Cl2 in absence of a
reducing atmosphere [4].
The other oxides (Cr2O3, Fe2O3, and MgO) were
oxychlorinated with Cl2 ‡ O2 gas mixture with Cl2/O2
molar ratio equal to 4. Kinetic aspects of the oxy-
chlorination Cr2O3 and MgO were reported in [12] and
[10], respectively. Fig. 14 compares the temperature
effects on the oxychlorination of the chromite and its
simple oxides. The following observations could be
deduced from this ®gure:
1. Concerning the oxychlorination of chromite (®rst
stage), Cr2O3, and Fe2O3:
1.1. chromite oxychlorination is characterized by
an apparent activation energy of about
151 kJ/mol between 600 and 8258C. The
same value of the Ea (ꢁ148 kJ/mol) was
obtained for the hematite oxychlorination.
Chromium (III) oxide oxychlorination pro-
ceeds with an Ea value of about 46 kJ/mol
[12] at temperatures higher than 7008C. The
reaction rate of chromite and that of hematite
with Cl2 ‡ O2 at temperatures lower or equal
to 8008C are of the same order of magnitude.
This suggest that the ®rst stage of the
More consist information was obtained during the
treatment of these oxides using TG measurements

N. Kanari et al. / Thermochimica Acta 371 (2001) 143±154
153
Fig. 14. Comparison of Arrhenius plots of the oxychlorination of chromite, Fe2O3, Cr2O3, and MgO.
chromite (FeCr2O4) reaction with Cl2 ‡ O2
at temperatures ꢂ8008C depend on the
oxychlorination rate of iron oxides;
the reaction products and chromite, the formation of
the lacunaire spinels, etc., affect the chromite oxy-
chlorination.
1.2. at temperatures higher than 8258C, the
oxychlorination of chromite and that of
Cr2O3 were characterized by approximately
the same values of the Ea and the reaction
rates. It seems that the global reaction rate of
chromite, in this temperature range, is
affected by the oxychlorination rate of
Cr2O3.
4. Conclusions
The presence of a reducing atmosphere was neces-
sary to achieve a complete reaction of the chromite
mineral with chlorine. Only iron chromite (FeCr2O4)
could be oxychlorinated at temperatures lower than
10008C using nonisothermal conditions.
The oxychlorination of the chromite occurred in
two stages characterized by widely different reactiv-
ities of chromite towards Cl2 ‡ O2 gaseous mixture.
Similarly, the effect of the temperature changed sig-
ni®cantly with the progress of the reaction. Iron
chromite seems to be more reactive than magnesium
chromite in the oxychlorinating gases.
2. Concerning the oxychlorination of chromite
(MgCr2O4), Cr2O3, and MgO: the value of Ea
deduced from the data of chromite ꢀX ꢄ 0:4†,
Cr2O3, and MgO oxychlorination are about 262,
46 [12] and 214 kJ/mol [10], respectively. The
reaction rate of Cr2O3 is much higher than that of
chromite. Thus, it suggests that the oxychlorina-
tion of chromite for X ꢄ 0:4 (MgCr2O4) depends
more on MgO reaction rate than that of Cr2O3.
The oxychlorination of chromite with Cl2 ‡ O2
ꢀCl2=O2 ˆ 4†, for reaction extents up to 0.30, pro-
ceeded with an apparent activation energy of about
151 and 57 kJ/mol between 600±8258C and 825±
While similarities of the chromite oxychlorination
with the reactions of its simple oxides with Cl2 ‡ O2
exists, one may underline that other phenomena such
as the chemical bonds between the simple oxides in
the chromite spinel structure, the interaction between
10508C, respectively. Oxychlorination with
a
Cl2 ‡ O2 mixture, having a Cl2/O2 ratio of 0.5, at
temperatures lower than 8258C lead to similar tem-

154
N. Kanari et al. / Thermochimica Acta 371 (2001) 143±154
perature effect on the reaction rate. While, a value of
Ea of about 85 kJ/mol was obtained for the oxychlor-
ination at temperatures higher than 8258C. The oxy-
chlorination rate using Cl2 ‡ O2 with Cl2=O2 ˆ 4 is
systematically higher than that obtained using a
Cl2 ‡ O2 with Cl2=O2 ˆ 0:5. The second stage of
the chromite oxychlorination ꢀX > 0:35† is strongly
dependent on the reaction temperature and was char-
acterized by Ea value of ꢁ262 kJ/mol between 925
and 10508C. The effect of temperature on the chromite
oxychlorination may be roughly explained by that of
its simple constituents.
References
[1] B.G. Korshunov, Metall. Rev. MMIJ 8 (2) (1992) 1±33.
[2] I. Gaballah, M. Djona, Metall. Mater. Trans. B 26 (1995) 41±
50.
[3] I. Gaballah, N. Kanari, Fifth Report of European Union
Contract No. BRE2-CT92-0173, Institut National Polythec-
nique de Lorraine, Laboratoire Environnement et Mineralur-
gie, Nancy, France, October, 1995.
[4] N. Kanari, Ph.D. Thesis, Institut National Polythechnique de
Lorraine, Laboratoire Environnement et Mineralurgie, Nancy,
France, November 1995, 209 pp.
[5] N. Kanari, I. Gaballah, in: B. Mishra (Ed.), Paper Presented at
the Proceedings of the TMS Annual Meeting, Orlando, FL,
February 9±13, 1997, EPD Congress 1997, TMS, pp. 57±71.
[6] N. Kanari, I. Gaballah, in: J. Welch (Ed.), Paper Presented at
the Proceedings of the TMS Annual Meeting, San Antonio,
TX, February 15±19, 1998, Light Metals 1998, TMS,
pp. 1333±1341.
Acknowledgements
This work was performed in the frame of contract
No. BRE2-CT92-0173, thanks to the ®nancial support
of the European Union (DG-XII). The authors thank
Dr. H.L. Schmidt for discussion, suggestion and
help.
They also would like to thank Dr. J.C. Mugica, Dr.
R. Solozabal (INASMET, San Sebastian, Spain) and
Dr. M. Coelho (INETI, Lisbon, Portugal) for technical
[7] N. Kanari, E. Allain, I. Gaballah, Metall. Mater. Trans. B 30
(1999) 577±587.
[8] Anonymous, in: R.C. Weast, M.J. Astle, W.H. Beyer (Eds.),
Handbook of Chemistry and Physics, 66th Edition, CRC
Press, Boca Raton, FL, 1986, pp. D193±D194.
Â
[9] P. Pascal, in: Masson et al. (Eds.), Nouveau Traite de Chimie
Â
Minerale, Tome XIV, 1959, p. 135.
[10] N. Kanari, I. Gaballah, Metall. Mater. Trans. B 30 (1999)
383±391.
[11] J. Szekely, J.W. Evans, H.Y. Sohn, in: Gas±Solid Reactions,
Academic Press, New York, 1976, pp. 68±70, 73±88, 109±131
and 232±240.
Â
discussions. Authors are indebted to Drs. A. Bonazebi,
M. Djona, S. Ivanaj, N. Menad, N. Mirghaffari for
discussions and help on different subjects and to Mrs.
C. Richard for technical and administrative support.
[12] I. Gaballah, S. Ivanaj, N. Kanari, Metall. Mater. Trans. A 29
(1998) 1299±1308.