Determination of temperature effect of oxychlorination of Cr2O3 and MgO using non-isothermal conditions

Article abstract of DOI:10.1016/S0040-6031(00)00430-5

The effect of temperature on the oxychlorination of Cr₂O₃ and MgO using Cl₂+O₂ was evaluated using non-isothermal conditions. A mathematical treatment of the experimental data was performed. Results were compared with those obtained using isothermal conditions. The oxychlorination of Cr₂O₃ up to 650°C was characterized by an apparent activation energy 'E(a)' of ca. 83 kJ/mol. Beyond 650°C, the reaction proceeded with an E(a) of ca. 51 kJ/mol. Between 850 and 1025°C, the reaction of MgO with Cl₂+O₂ was strongly dependent on temperature and its E(a) was equal to ca. 215 kJ/mol. The values of the apparent activation energies, found for the oxychlorination of both oxides using non-isothermal conditions, were comparable to those obtained during isothermal treatment. (C) 2000 Elsevier Science B.V.

Full text of DOI:10.1016/S0040-6031(00)00430-5

Thermochimica Acta 351 (2000) 131±137
Determination of temperature effect of oxychlorination of
Cr2O3 and MgO using non-isothermal conditions
N. Kanaria,1, I. Gaballaha,*, E. Allainb
aMineral Processing and Environmental Engineering team, LEM, CNRS, UMR 7569, ENSG, INPL, BP 40, 54501 Vandoeuvre, France
bUniversity of Missouri-Rolla, School of Mines and Metallurgy, Center for Pyrometallurgy, 210 Fulton Hall, 65401 Rolla, MO, USA
Received 1 August 1999; accepted 1 February 2000
Abstract
The effect of temperature on the oxychlorination of Cr2O3 and MgO using Cl2‡O2 was evaluated using non-isothermal
conditions. A mathematical treatment of the experimental data was performed. Results were compared with those obtained
using isothermal conditions.
The oxychlorination of Cr2O3 up to 6508C was characterized by an apparent activation energy `Ea' of ca. 83 kJ/mol. Beyond
6508C, the reaction proceeded with an Ea of ca. 51 kJ/mol. Between 850 and 10258C, the reaction of MgO with Cl2‡O2 was
strongly dependent on temperature and its Ea was equal to ca. 215 kJ/mol.
The values of the apparent activation energies, found for the oxychlorination of both oxides using non-isothermal
conditions, were comparable to those obtained during isothermal treatment. # 2000 Elsevier Science B.V. All rights reserved.
Keywords: Thermogravimetric analysis; Oxychlorination; Chromium(III) oxide; Magnesium oxide; Non-isothermal conditions; Apparent
activation energy
1. Introduction
trial application of this speci®c reaction. Generally,
the effects of temperature on the gas±solid reactions
are determined using isothermal conditions. Thus,
solids are ®rst heated at a predetermined temperature
under an inert atmosphere and when a steady state is
reached the reactive gases are introduced to react with
the solid isothermally. The dependency of the reac-
tions on the temperature is evaluated by Arrhenius
plot. This method was recently used by Djona et al.
[1], Gaballah et al. [2], Allain et al. [3], and Kanari
et al. [4] for the determination of the kinetics para-
meters of the chlorination and carbochlorination of
some refractory metals oxides.
The activation energy of a speci®c reaction is an
important engineering data. It will condition the reac-
tor size and the optimum temperature for any indus-
* Corresponding author. Tel.: ‡33-383-596-336;
fax: ‡33-383-569-585.
E-mail address: gaballah@ensg.u-nancy-fr (I. Gaballah)
1
Â
LEM, Laboratoire Environnement et Mineralurgie, rue du
Doyen M. Roubault, BP 40, 54501 Vandoeuvre Cedex, France;
CNRS, Centre National de la Recherche Scienti®que, 3 rue Michel-
Ange, 75794 Paris Cedex, France; ENSG, Ecole Nationale
 Â
Superieure de Geologie, rue du Doyen M. Roubault, BP 40,
The literature survey revealed that some research
works were devoted earlier to the determination of the
temperature effect on the gas±solid reactions using
54501 Vandoeuvre Cedex, France; INPL, Institut National Poly-
Ã
technique de Lorraine, 2 rue de la Foret de Haye, 54501
Vandoeuvre Cedex, France.
0040-6031/00/$ ± see front matter # 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 0 - 6 0 3 1 ( 0 0 ) 0 0 4 3 0 - 5

132
N. Kanari et al. / Thermochimica Acta 351 (2000) 131±137
one single non-isothermal experiment. Chaubal and
Sohn [5] developed a theoretical approach to de®ne
the rate-controlling step of the chalcopyrite oxidation
up to 8778C using non-isothermal conditions. Simi-
larly, Wang et al. [6] investigated the kinetics of the
oxidation of cobaltite between 300 and 9008C using
the non-isothermal technique. Bustens et al. [7]
obtained consistent kinetics results testing isothermal
and non-isothermal reduction of molybdenum and
tungsten oxides by hydrogen.
pA the partial pressure of the reactive gas(es) and d the
mean particle size.
The effects of partial pressure of the reactive gases,
reaction extent and particle size on the reaction rate
are described by f1(pA), f2(X) and f3(d), respectively.
Arrhenius law gives the temperature effect on the
reaction rate:
ꢀ
ꢁ
Ea
RT
k ˆ k0 exp
(2)
The advantage of using one single non-isothermal
test, to determine the temperature effect on gas±solid
reactions, is time and material saving. However, if the
gas reactivity towards the solid is high and the heating
rate of furnace is low, the temperature effect will be
de®ned only in a small temperature range. On the
other hand, higher furnace heating rates will lead to a
large difference of temperature between the sample
and the bulk of reactive gas mixture. To avoid such
disadvantage in determining the value of the apparent
activation energy, at least two non-isothermal tests
using different furnace heating rates are required. The
determination of the temperature effect by the non-
isothermal technique will be especially appreciated
when the solids undergo various thermal phenomena
such as: decomposition; sublimation; volatilization;
dehydration, etc.
Integration of Eq. (1) in isothermal conditions gives:
Z
The conversion function g(X) generally depends on
the geometric change appearing in the solid as the
reaction proceeds. The different types of g(X) were
summarized earlier by Szekely et al. [10].
Z
X
t
dX
ˆ gꢀX† ˆ
kf1ꢀpA†f3ꢀd†dt
0
(3)
f2ꢀX†
0
If the temperature is raised linearly with respect to
time the temperature evolution may be given by Eq. (4):
T ˆ ft
(4)
or
dT ˆ f dt (5)
Substituting Equations (2) and (4) in Eq. (1) and
rearranging the resulting equation, it was obtained:
This paper is focused on the determination of
temperature effect of the oxychlorination rate of
Cr2O3 and MgO using non-isothermal conditions.
The results were compared with those obtained
for the isothermal oxychlorination of the same
oxides.
ꢂꢀ
ꢁ
ꢃ
ꢂ
ꢃ
dx
dt
k0f1ꢀpA†f3ꢀd†
Ea
ln
g0ꢀX† ˆ ln
f
RT
(6)
In the above equation, g0(X) is the derivative of g(X)
with respect to X. The activation energy can be
determined from the slope of data ®tted by Eq. (6)
when the appropriate type of g(X) is determined.
2. Theoretical approach
The following paragraphs describe the theoretical
approach [5] for the determination of the activation
energy of a gas±solid reaction under non-isothermal
conditions.
The reaction rate of a process can be expressed as
follow:
3. Materials and experimental procedure
The samples of chromium (III) oxide and magne-
sium oxide used for this study were of pure analytical
grade. Results of physico-chemical analysis of these
two samples were given earlier [8,9]. They con®rmed
the purity of both those oxides.
The thermogravimetric analysis (TGA) of oxy-
chlorination was investigated using 40 mg of sample.
The experimental set described in Ref. [9] was
dx
ˆ kf1ꢀpA†f2ꢀX†f3ꢀd†
(1)
dt
where X is the reaction extent (de®ned to be the weight
of the reacted fraction to the initial solid weight ratio),

N. Kanari et al. / Thermochimica Acta 351 (2000) 131±137
133
employed in this study. Its main unit was a CAHN
1000 micro-balance having a sensitivity of 10 mg. The
oxychlorination gas mixture composed of Cl2‡O2 was
puri®ed using H2SO4 and CaCl2 columns before its
introduction in the reactor. The sample was homo-
geneously distributed in a quartz crucible having a
section of ca. 0.5 cm2. This crucible was hooked to the
balance using quartz rods. For TGA isothermal runs,
the sample was preheated in nitrogen atmosphere up to
the desired temperature before the introduction of the
chlorinating gas mixture. In the case of non-isothermal
TGA runs, using a linear heating rate, the sample was
directly heated in the chlorinating atmosphere.
4. Results
Fig. 1. TGA of Cr2O3 oxychlorination using linear temperature
raise.
4.1. Oxychlorination of Cr2O3
An oxychlorinating gas mixture of Cl2‡O2 having a
linear velocity of ca. 33 cm/min and a molar ratio Cl2/
O2ˆ4 was used for the non-isothermal oxychlorina-
tion test of Cr2O3. This gas composition was chosen to
allow the formation of chromium oxychloride accord-
ing to Eq. (7). Taking into account the reactivity of
Cr2O3 towards the oxychlorinating gas [8], the heating
rate of the sample (fˆdT/dt) was kept constant at ca.
28C/min.
Chromium trioxide starts to react with the oxychlor-
inating gas mixture at temperatures higher than 5008C
and ca. 80% of the sample were oxychlorinated at
8508C. A higher heating rate of furnace would be
necessary to get a lower reaction rate at lower tem-
peratures. However, as mentioned in the introduction,
higher furnace heating-rates would have led to a
large temperature gradient between the sample and
the gas bulk.The data obtained during the isothermal
oxychlorination of Cr2O3 with Cl2‡O2 (Cl2/O2ˆ4)
are available in Ref. [8] and grouped in Fig. 2a
and b. It must be reported that the evolution of the
reaction extent as function of time could be described
by Eqs. (8) and (9) at temperatures lower and higher
12 Cr2O3 ‡ Cl2 ‡ 14 O2 ! CrO2Cl2
(7)
Fig. 1 shows the evolution of the percent weight
loss (pct WL) of the Cr2O3 sample as function of
the temperature using the conditions given above.
Fig. 2. Isotherms of oxychlorination of Cr2O3 for (a) Tꢁ7008C and (b) T ꢂ7508C.

134
N. Kanari et al. / Thermochimica Acta 351 (2000) 131±137
Table 1
Table 2
Kinetics data obtained during the isothermal oxychlorination of
Cr2O3 between 525 and 6508C
Kinetics data obtained during the isothermal oxychlorination of
Cr2O3 between 680 and 10008C
a
a
Temperature
Value of Cc of the data
fitted by: 1 (1 X)1/3ˆkt
Value of the rate
Temperature
Value of Cc of the data
fitted by: 1 (1 X)1/2ˆkt
Value of the rate
(8C)
constant, `k'
(8C)
constant, `k'
4
3
525
550
575
600
625
650
0.998
0.997
0.995
1.000
0.999
0.998
2.61Â10
680
700
750
800
850
900
950
1000
1.000
1.000
1.000
1.000
1.000
1.000
0.999
0.999
3.11Â10
4
3
3.44Â10
3.56Â10
4
3
5.93Â10
4.59Â10
4
3
8.49Â10
5.96Â10
3
3
1.26Â10
7.71Â10
3
3
1.61Â10
9.15Â10
2
1.15Â10
a Correlation coef®cient.
2
1.45Â10
a Correlation coef®cient.
than 6508C, respectively.
1=3
ꢀ1 X† ˆ kt
and trying different possible formulations given by
Szekely et al. [10] concerning the conversion function
g(X) at temperatures up to 6508C. The expression
given by Eq. (8) was the best ®tting for the linear-
ilization of ln [(dX/dT)g0(X)] as function of 1/T. The
result is given in Fig. 3b. Avalue of Ea of ca. 83 kJ/mol
was obtained for the non-isothermal oxychlorination
of Cr2O3 up to 6508C. This value is in fairly good
agreement with the value obtained from isothermal
treatment tests.
The kinetics data concerning the isothermal oxy-
chlorination of Cr2O3 between 680 and 10008C for
Xˆ0.8 are given in Table 2. While Fig. 4a traces the
evolution of the rate constant as function of tempera-
ture in Arrhenius ordinates. The value of Ea deduced
from Fig. 4a is ca. 48 kJ/mol. The decrease of the
value of the apparent activation energy at temperatures
1
1
(8)
(9)
1=2
ꢀ1 X† ˆ kt
where X is reaction extent, k a constant and t a
chlorination time.
Table 1 gives the ®tting curves obtained using Eq.
(8) (for Xˆ0.8) and concerning the isothermal oxy-
chlorination of Cr2O3 between 525 and 6508C. Fig. 3a
shows the Arrhenius diagram of the Cr2O3 isothermal
oxychlorination, plotted as the evolution of the rate
constant as function of temperature. Avalue of apparent
activation energy `Ea' of ca. 94 kJ/mol was obtained
fortheisothermaloxychlorinationofCr2O3 upto6508C.
A similar value of Ea was reported earlier [8] for the
initial reaction rate of Cr2O3 with Cl2‡O2.
Data of Fig. 1, concerning the non-isothermal oxy-
chlorination of Cr2O3, were processed using Eq. (6)
Fig. 3. Effect of temperature on the oxychlorination of Cr2O3 up to 6508C using isothermal (a) and non-isothermal (b) conditions.

N. Kanari et al. / Thermochimica Acta 351 (2000) 131±137
135
Fig. 4. Effect of temperature on the oxychlorination of Cr2O3 at temperatures higher than 6508C using isothermal (a) and non-isothermal (b)
conditions.
higher than 6508C (Fig. 3a and Fig. 4a) is due to the
decrease of the sample speci®c surface area with the
raise of temperature [8].
4.2. Oxychlorination of MgO
Isothermal oxychlorination of MgO with Cl2‡O2
between 850 and 10258C was described in details in
Ref. [11] and summarized by Fig. 5a and b. The
chlorination of MgO by chlorine can be represented
by Eq. (10). The generated magnesium chloride is
characterized by its melting and boiling points of 714
and 14128C, respectively [12]. To insure that the rate
of MgCl2 volatilization was higher than that of its
formation, it was suggested to increase the partial
pressure of oxygen in the system to decrease the rate
of the chlorination of MgO by chlorine. Thus, the
chlorination of MgO was achieved using an oxychlor-
inating gas mixture Cl2‡O2 having a molar ratio
Cl2/O2ˆ4.
Non-isothermal data of Cr2O3 oxychlorination
represented in Fig. 1a for the temperatures higher
than 6508C were also tried to be ®tted by Eq. (6)
for different types of the function g(X). The appro-
priateness of the ®tting was determined by calculating
the coef®cient correlation. As for the isothermal test
case, the best correlation was found using Eq. (9). Fig.
4b represents the plot of Eq. (6) for non-isothermal
treatment of Cr2O3 with Cl2‡O2. A value of Ea of ca.
51 kJ/mol was calculated from the slope of this plot.
The comparison of the results exhibited in Fig. 4a and
b shows that the values of the apparent activation
energy within the experimental error.
Fig. 5. Isotherms of oxychlorination of MgO for (a) Tꢁ9258C and (b) T ꢂ9508C.

136
N. Kanari et al. / Thermochimica Acta 351 (2000) 131±137
Table 3
Kinetics data obtained during the isothermal oxychlorination of
MgO between 850 and 10258C
a
Value of Cc of the data
Temperature
Value of the rate
(8C)
fitted by: 1 (1 X)1/2ˆkt
constant, `k'
4
850
875
0.999
0.999
0.997
0.997
0.999
0.999
0.999
0.998
3.65Â10
4
6.18Â10
4
900
925
9.72Â10
3
1.50Â10
3
950
2.33Â10
3
975
3.69Â10
3
1000
1025
5.32Â10
3
7.98Â10
a Correlation coef®cient.
to react with the oxychlorinating gas mixture at tem-
peratures higher than 7508C and ca. 70% of the sample
had reacted at 10258C.
Fig. 6. TGA of MgO oxychlorination using linear temperature
raise.
MgO ‡ Cl2 ! MgCl2 ‡ 12 O2
(10)
To determine the effect of temperature on the
oxychlorination rate of MgO one single non-isother-
mal test under Cl2‡O2 (Cl2/O2ˆ4) was carried out
from ca. 575 to 10258C. As the reactivity of MgO
towards oxychlorinating gas mixture was relatively
low [11], the heating rate of the sample was chosen to
be ca. 18C/min. This heating rate generates accurate
data as the temperature gradient between the sample
and the gas bulk is small. The evolution of the
temperature and pct WL of the sample as a function
of time are shown in Fig. 6. Magnesium oxide started
The main kinetics data obtained during the isother-
mal treatment of MgO in Cl2‡O2 (Cl2/O2ˆ4) are
given in Table 3. The evolution of the reaction extent
(Xˆ0.7) as function of time for all isotherms was
appropriately described by Eq. (9). The Arrhenius plot
of Fig. 7a reveals the temperature effect of the MgO
isothermal oxychlorination. An apparent activation
energy of ca. 213 kJ/mol was deduced from this
Arrhenius diagram.
Using Eq. (6), it is clear that the activation energy
could be determined only after ®nding the appropriate
Fig. 7. Effect of temperature on the oxychlorination of MgO between 850 and 10258C using isothermal (a) and non-isothermal (b) conditions.

N. Kanari et al. / Thermochimica Acta 351 (2000) 131±137
137
form of g(X). As for the isothermal tests, different
models of the conversion function g(X) were tried and
the best data ®tting was obtained using Eq. (9). The
data were plotted in Fig. 7b using Eq. (6). Avalue of Ea
of ca. 215 kJ/mol was deduced for the non-isothermal
oxychlorination of MgO, between 850 and 10258C,
for a reaction extent up to 0.7. Fig. 7 indicates that the
activation energy values calculated using the two sets
of data, for MgO oxychlorination, were within the
experimental error (213 and 215 kJ/mol). This result
con®rms once again the validity of the non-isothermal
measurements to determine the activation energy of a
reaction.
They also would like to thank Drs. J.C. Mugica,
R. Solozabal (INASMET, San Sebastian, Spain) and
M. Coelho (INETI, Lisbon, Portugal) for technical
Â
discussions. Authors are indebted to Drs. A. Bonazebi,
M. Djona, S. Ivanaj, N. Menad, N. Mirghaffari
for discussions and help on different topics and to
Mrs. C. Richard for technical and administrative
support.
References
[1] M. Djona, E. Allain, I. Gaballah, Metall. Mater. Trans. B 26B
(1995) 703±710.
[2] I. Gaballah, M. Djona, E. Allain, Metall. Mater. Trans. B 26B
(1995) 711±718.
5. Conclusions
[3] E. Allain, M. Djona, I. Gaballah, Metall. Mater. Trans. B 28B
(1997) 223±233.
The non-isothermal technique was found to be
suitable for studying the effect of temperature on
the oxychlorination of chromium and magnesium
oxide. Results were in good agreement with those
obtained during isothermal treatment.
Oxychlorination of chromium (III) oxide was char-
acterized by an apparent activation energy values of
ca. 83 and 51 kJ/mol for temperatures lower and
higher than 6508C, respectively.
[4] N. Kanari, B.R. Reddy, I. Gaballah, Metall. Mater. Trans. B
29B (1998) 729±737.
[5] P.C. Chaubal, H.Y. Sohn, Metall. Trans. B 17B (1986) 51±60.
[6] G.X. Wang, D. Chandra, M.C. Fuerstenau, Metall. Mater.
Trans. B. 26B (1995) 523±527.
[7] J.A. Bustnes, Du Sichen, S. Seetharaman: Paper presented at
Proc. of the TMS Annual Meeting, San Francisco, California,
27 February±3 March 1993, EPD Congress 1994, G. Warren
(Ed.), Published by TMS, pp. 581±600.
[8] I. Gaballah, S. Ivanaj, N. Kanari, Metall. Mater. Trans. A 29A
(1998) 1299±1308.
The oxychlorination of magnesium oxide between
850 to 10258C was strongly depended on the tem-
perature with an apparent activation energy of ca.
215 kJ/mol.
[9] N. Kanari, I. Gaballah: Paper presented at Proc. of the TMS
Annual Meeting, San Antonio, Texas, 15±19 February 1998,
EPD Congress 1998, B. Mishra (Ed.), Published by TMS,
pp. 595±609.
[10] J. Szekely, J.W. Evans, H.Y. Sohn, in Gas±Solid Reactions,
Academic Press, New York, NY, 1976, 68±70, 73±88, 109±
131 and 232±235.
Acknowledgements
[11] N. Kanari, E. Allain, I. Gaballah: Paper presented at Proc. of
the TMS Annual Meeting, San Diego, California, 28
February±4 March 1999, EPD Congress 1999, B. Mishra
(Ed.), Published by TMS, pp. 29±37.
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.
[12] D.R. Lide, Editor-in-Chief, Handbook of Chemistry and
Physics, 74th Edition (CRC Press, Florida, USA, 1993±1994),
pp. 4±71 and 6±69.