Kinetics of reduction of magnesia with carbon

Article abstract of DOI:10.1016/S0040-6031(02)00128-4

Reduction kinetics of magnesia with carbon powder was studied using the thermal gravimetric technic in the temperature range of 1743-1883 K. The reduction ratio was determined as a function of time. The effects of compact-forming pressure, composition, partial pressure of CO, sample height, and temperature on reduction ratio were investigated. Experimental results revealed that the gas diffusion including Mg vapor, CO and CO₂ through the porous medium was the rate-determining step of the overall reduction process. Activation energy of the gas diffusion was estimated to be about 30.59 kJ mol^-1.

Full text of DOI:10.1016/S0040-6031(02)00128-4

Thermochimica Acta 390 (2002) 145±151
Kinetics of reduction of magnesia with carbon
Li Rongtia, Pan Weia,*, Masamichi Sanob, Jianqiang Lia
aState Key Laboratory of New Ceramics and Fine Processing, Department of Materials Science and Engineering,
Tsinghua University, Beijing 100084, PR China
bDepartment of Materials Process Engineering, Graduate School of Engineering, Nagoya University,
Furo-cho, Chikusa-ku, Nagoya, Aichi-ken 464-8603, Japan
Received 26 November 2001; received in revised form 14 February 2002; accepted 20 February 2002
Abstract
Reduction kinetics of magnesia with carbon powder was studied using the thermal gravimetric technic in the temperature
range of 1743±1883 K. The reduction ratio was determined as a function of time. The effects of compact-forming pressure,
composition, partial pressure of CO, sample height, and temperature on reduction ratio were investigated. Experimental results
revealed that the gas diffusion including Mg vapor, CO and CO2 through the porous medium was the rate-determining step of the
overall reduction process. Activation energy of the gas diffusion was estimated to be about 30.59 kJ molÀ1. # 2002 Elsevier
Science B.V. All rights reserved.
Keywords: Reduction kinetics; Magnesia and carbon; Thermal gravimetric technic; Diffusion
1. Introduction
metallic elements, such as oxygen, sulfur, phosphor-
ous, and nitrogen, which can be used in steel-making
industry to obtain high quality steels [3±7]. In our
research group, the investigation on deoxidation [8]
and desulfurization [9] of molten iron was carried out
by using a porous immersion tube, the inside of which
was charged with MgO±C pellets. Magnesia in the
pellets was reduced by carbon to produce magnesium
vapor, which was injected directly into the melt. It
was found that the deoxidation and desulfurization
rates and ef®ciency depend on the reduction rate of
magnesia with carbon. Therefore, the kinetics of the
reduction of magnesia with carbon is one key point in
controlling the process of desulfurization and ®nding
more ef®cient ways to improve the desulfurization rate
and ef®ciency. Although, numerous studies have been
conducted on carbothermic reduction of metal oxide
[10±15], the mechanism of reduction of magnesia with
carbon is not fully understood. The complexity of the
MgO±C refractories bricks are widely used in steel-
making industry as basic oxygen furnace (BOF) lin-
ings because of their good resistance both to slag
corrosion and to thermal stresses [1]. However, a
serious problem is that the reaction between magnesia
and carbon takes place, and the strength and service
life decrease when it is used under the steel-making
condition at about 1600 8C for a long time [2]. Until
now, little research work dealing with the reduction
kinetics of magnesia with carbon has been reported.
Recently, the magnesium produced by the reduction
of MgO with carbon attracts many researchers' inter-
est owing to its strong chemical af®nity to those non-
* Corresponding author. Tel.: ‡86-10-6277-2858;
fax: ‡86-10-6277-1160.
E-mail address: panw@mail.tsinghua.edu.cn (P. Wei).
0040-6031/02/$ ± see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 0 - 6 0 3 1 ( 0 2 ) 0 0 1 2 8 - 4

146
L. Rongti et al. / Thermochimica Acta 390 (2002) 145±151
reduction process makes it dif®cult to gain a clear
insight of the various steps involved in the overall
reduction reaction.
The purpose of the present study was to gain more
understanding of the reduction kinetics of magnesia
with carbon. Effects of compact-forming pressure,
sample composition, CO partial pressure, sample
height, and temperature on the reduction ratio would
also be discussed.
the sample. Then the temperature was raised to the
experimental temperature at the heating rate of
50 K minÀ1 and this temperature was kept for 1 h. To
correct the effect of the graphite crucible on the mass
loss of the sample, blank experiments were carried out
using the blank graphite crucible. The calibration runs
were carried out under the same conditions as those
for the samples.
3. Results and discussion
2. Experimental
Based on our previous study, the interaction
between magnesia and carbon can be represented
by the following reactions:
2.1. Specimen preparation
The magnesia powder (0.4 mm in average diameter,
purity higher than 99.99%), graphite (®xed carbon
99.12%, ash 0.15%, water 0.4%, average diameter
3.9 mm, purity higher than 99.5%) and charcoal pow-
der (44 mm in average diameter, purity higher than
99.5%) were used in this study as starting materials.
These powders were supplied by the High Purity
Chemical Institute of Japan.
Direct reduction:
MgOꢀs† ‡ Cꢀs† ! Mgꢀg† ‡ COꢀg†
Indirect reduction:
(1)
MgOꢀg† ‡ COꢀg† ! Mgꢀg† ‡ CO2ꢀg†
CO2ꢀg† ‡ Cꢀs† ! 2COꢀg†
(2)
(3)
Magnesia powder and graphite were mixed at molar
ratios of 1:1 and 1:2, and magnesia powder and
charcoal were mixed at a molar ratio of 1:1. The
mixture was shaped into a 5 mm diameter pellet under
sample-forming pressures of 100 and 150 MPa by
using a cold isostatic press for 1 h. The size of the
graphite crucible used in this study was 6.4 mm o.d.,
5.8 mm i.d., and the height of the graphite crucible
ranged from 4.2 to 7.3 mm.
The kinetics of a solid±solid reaction is in¯uenced
by many factors, such as sample composition, particle
size and distribution, reaction products, compact-
forming pressure, and heating rate. During the
thermo-gravimetric reduction experiment, the mass
loss of the sample is measured as a function of time.
The reduction ratio a at a given instance is de®ned as:
Dm
m0
a ˆ
(4)
where m0 and Dm represents the initial mass of
magnesium in the sample and the magnesium mass
change at a given instance, respectively.
2.2. Apparatus and experimental procedure
The reduction was carried out using a thermo-
balance (NS95, Sinku-riko Incorporation of Japan),
which was connected to a data acquisition and analysis
system. The sample was put into the graphite crucible,
which was placed on the top of the thermo-couple. A
layer of graphite (about 3 mg) was put on the top of
the sample in order to minimize the reoxidation of
magnesium. The inert atmosphere was maintainedÀb1y
3.1. Effect of compact-forming pressure
The effect of compact-forming pressure on the
reduction ratio at the temperature 1848 K holding
for 1 h is given in Fig. 1. It shows that the rate of
reduction decreases signi®cantly with the increment
of compact-forming pressure. It is known that the
contacting condition between magnesia and carbon
particles would be improved with an increase of the
sample compact-forming pressure. In our previous
study on the reduction of magnesia with carbon under
blowing argon at a ¯ow rate of 1:67 Â 10À6 m3 s
.
The air in the furnace was substituted with argon for
30 min, and then the sample was heated from room
temperature to 773 K and kept at this temperature for
10 min to remove the absorbed gases and water from

L. Rongti et al. / Thermochimica Acta 390 (2002) 145±151
147
Fig. 1. Effect of sample-forming pressure on the reduction ratio at the temperature 1884 K holding for 1 h.
non-isothermal condition, the reaction rates are sen-
sitive to the compact-forming pressure and the reduc-
tion rates increase with the increment of the contacting
condition between particles. Therefore, the rate-deter-
mining step is changed at the present experimental
condition. In this process, the possible rate-controlling
step must be related to reactions between gas and solid
or the gas diffusion including Mg vapor, CO and CO2
through the porous media, because the path of gas
diffusion becomes longer and more tortuous with the
improvement of the contacting condition between
magnesia and carbon particles, resulting in the
decrease of the gas diffusion rate. The gas±solid sur-
face reaction can usually be broken down into the
following elementary steps:
sample composition has no signi®cant effect on the
reduction rate at the temperature 1848 K holding for
1 h. It is worth noticing that the reduction rate is
much affected by the sample composition before the
temperature reaches 1848 K, and the rate of the
reaction between magnesia and graphite is slower
than that of the reaction between magnesia and
charcoal. In our previous study, it was found that
at the initial stage of the reduction under the non-
isothermal condition, the solid boundary reaction is
the rate-determining step and its reaction rate is
sensitive to the sample composition. The reaction
activity of charcoal is higher than that of graphite, so
the reaction between magnesia and charcoal is faster
than that between magnesia and graphite before the
temperature reaches 1848 K. But at the temperature
holding step, their reaction rates are becoming very
close. The only explanation for this phenomenon is
that the reaction rate is independent of the sample
composition and the possible rate-determining step is
the gas diffusion including Mg vapor, CO and CO2
through the porous media.
(1) diffusion of the reactants to the surface;
(2) adsorption of reactants at the surface;
(3) chemical reactions on the surface;
(4) desorption of products from the surface;
(5) diffusion of products away from the surface.
Although, neither the diffusion rate of reactants to
the solid surface nor the diffusion rate of products
away from the solid surface seems to be directly rate-
determining, the decrease of the gas diffusion rate
leads to the decrease of the overall reduction rate.
3.3. Effect of CO partial pressure
The effect of CO partial pressure on the reduction
ratio is shown in Fig. 3. It indicates that the CO partial
pressure has no evident effect on the reduction rate at
1848 K holding for 1 h. However, before the tempera-
ture reaches 1848 K, the reduction rate increases with
the increase of the CO partial pressure. This result is in
3.2. Effect of sample composition
Fig. 2 shows the effect of the sample composition
on the reduction ratio. It seems that changing of the

148
L. Rongti et al. / Thermochimica Acta 390 (2002) 145±151
Fig. 2. Effect of composition on the reduction ratio at temperature 1884 K holding for 1 h.
a good agreement with our previous work. As we have
stated, the possible rate-determining step is the gas
diffusion including Mg vapor, CO and CO2 through
the porous media, therefore the effect of every possible
factor on the gas diffusion coef®cient must be taken
into consideration. However, the effect of the CO
partial pressure on the diffusion coef®cient is so little
that it can be neglected.
sensitive to the height of the sample. In order to
prove this hypothesis, an experiment was designed
by carrying out the reduction reaction at different
sample heights. The schematic diagram of the
experiment is shown in Fig. 4. Both the sample
weight and sample height (d) were changed (d ran-
ging from 0.5 to 3.5 mm) while h was kept constant
(3.6 mm). Fig. 5 shows that the sample height has a
signi®cant effect on the reduction rate. When d is in a
range of 0.5 to 2.5 mm, the increase of the sample
height reduces the reduction rate greatly. However,
when d is above 2.5 mm, the reduction rate is only
changed slightly. The relation between the sample
3.4. Effect of sample height
If the gas diffusion through the porous media is the
rate-determining step, the reaction rate should be
Fig. 3. Effect of pressure ratio of PAr/PCO on the reduction ratio at 1848 K holding for 1 h.

L. Rongti et al. / Thermochimica Acta 390 (2002) 145±151
149
3.5. Effect of temperature
In order to investigate the in¯uence of the reaction
temperature on the reduction ratio, experiments were
performed at 1743, 1773, 1813, 1848, and 1883 K,
respectively. The results are shown in Fig. 6. It
indicates that the reaction rate increases only slightly
with the increase of temperature.In this experiment,
the reduction rate is a constant. It can be expressed
as:
da
dt
ˆ k
(6)
Fig. 4. The schematic diagram of the experiment.
where da/dt is the reaction rate and k the rate constant.
As the reaction rate is determined by the gas diffu-
sion, the relationship between the reaction rate and the
diffusion coef®cient can be given as following:
height and the reduction rate can be expressed as
following:
da
dt
ˆ ꢀ2:7  10À5
†
da
ˆ k ˆ BD
(7)
‡ ꢀ1:5  10À4† expꢀ0:85 À 0:715d†
(5)
dt
where B is a constant and D the diffusion coef®cient,
whose relation with the temperature can be given by
the Arrhenius equation:
where d (mm) is the sample height and da/dt the
reduction rate.
With the increment of the sample height, the total
diffusion path of Mg, CO and CO2 gases through the
porous media is prolonged, so the molar ¯ux of gases
is decreased, resulting in the decrease of the reduction
rate. This result is in a good agreement with the
hypothesis that the gas diffusion through the porous
media is the rate-determining step.
ꢀ
ꢁ
E
D ˆ D0 exp À
(8)
RT
where D0 is the frequency factor and E the activation
energy. Both D0 and E are essentially constant over a
wide temperature range.
Fig. 5. Relationship between sample height and reduction rate at 1848 K holding for 1 h.

150
L. Rongti et al. / Thermochimica Acta 390 (2002) 145±151
Fig. 6. Effect of the holding temperature on the reduction ratio.
Fig. 7. Relation between reduction rate and temperature.
By substituting Eq. (8) into Eq. (7) and taking
logarithms, the following equation can be obtained:
4. Conclusion
In the present work, reduction kinetics of mixtures
of magnesia and carbon powder was investigated
using the isothermal gravimetric technic in the
temperature range of 1743±1883 K. It is found that
the sample composition and CO partial pressure
have no signi®cant effect on the reduction rate, the
reduction rate decreases with the increase of the
da
E
ln ˆ ln BD0 À
dt
(9)
RT
Eq. (9) indicates a linear relationship between
ln(da/dt) and 1/T. By plotting ln(da/dt) versus 1/T
(Fig. 7), the activation energy E is estimated to be
30.59 kJ molÀ1
.

L. Rongti et al. / Thermochimica Acta 390 (2002) 145±151
151
[5] W.H. Duquette, N.R. Grif®ng, T.W. Miller, Open Hearth
Proc. 56 (1973) 79.
compact-forming pressure, the reaction rate increases
only slightly with the temperature, and the sample
height has a strong effect on the reduction ratio.
The gas diffusion (including Mg vapor, CO and
CO2) through the porous media is believed to be
the rate-determining step in the overall reduction
process.
[6] W.G. Wilson, A. McLean, Desulfurization of Iron and Steel
and Sul®de Shape Control, The Iron and Steel Society of
AIME, 1980.
[7] P.J. Koros, R.G. Peirushka, R.G. Kerlin, Iron Steelmaker 4
(1977) 34.
[8] J. Yang, K. Okumura, M. Sano, in: Proceedings of the
International Conference on Asia Steel, Vol. C, 2000, p. 80.
[9] J.Y. Shan, K. Okumura, M. Sano, in: Proceedings of the
International Conference on Asia Steel, Vol. C, 2000, p. 94.
[10] D.J. Min, J.W. Han, W.S. Chung, Metall. Metal Trans. B 30
(1999) 215.
References
[11] M.A. Naiyang, N.A. Warner, Can. Metall. Quart. 38 (1999)
165.
[1] I. Peretz, B. Breany, Am. Ceram. Soc. Bull. 7 (1992) 1383.
[2] G.D. Picking, J.D. Batchelor, Am. Ceram. Soc. Bull. 50
(1971) 611.
[12] N.S. Sundar Murti, V. Seshadri, Trans. ISIJ 22 (1982) 926.
[13] Y.K. Rao, Chem. Eng. Sci. 29 (1974) 1435.
[14] Y.K. Rao, Metall. Trans. 2 (1971) 1439.
[15] L. Hong, K. Okumura, M. Sano, Metall. Metal Trans. B 30
(1999) 1003.
[3] M. Ozawa, Y. Outsuka, R. Hori, R. Mori, The Report for
Japan Society for the Promotion of Science, June, 1975.
[4] S. Yamaguchi, T. Uemura, H. Nashiwa, H. Sugita, Ironmak-
ing Steelmaking 5 (1977) 276.