Reduction of MnFe2O4 without and with carbon. Simultaneous measurements of humidity sensor and evolved gas analysis

Article abstract of DOI:10.1023/A:1020605416982

We succeeded in studying the mechanism of hydrogen added carbothermic reduction process of iron-manganese oxide by means of the new technique, simultaneous measurement of evolved gas analysis (EGA) and humidity sensor (HS). Water vapor evolved by the redu

Full text of DOI:10.1023/A:1020605416982

Journal of Thermal Analysis and Calorimetry, Vol. 69 (2002) 1045–1050
REDUCTION OF MnFe O WITHOUT AND WITH
2
4
CARBON
Simultaneous measurements of humidity sensor and evolved
gas analysis
1
*
1
1
2
T. Hashizume , K. Terayama , T. Shimazaki , H. Itoh and Y. Okuno
3
1
Department of Material Systems Engineering and Life Science, Faculty of Engineering,
Toyama University, 3190 Gofuku,Toyama 930-8555, Japan
2
Material and Strength Research Laboratory, Takasago Research and Development Center,
Mitsubishi Heavy Industries, Ltd. 2-1-1 Shinhama Arai-cho, Takasago Hyogo 676-8686, Japan
3
Graduate Student, Department of Material Systems Engineering and Life Science,
Faculty of Engineering, Toyama University, 3190 Gofuku, Toyama 930-8555, Japan
Abstract
We succeeded in studying the mechanism of hydrogen added carbothermic reduction process of
iron-manganese oxide by means of the new technique, simultaneous measurement of evolved gas
analysis (EGA) and humidity sensor (HS). Water vapor evolved by the reduction with hydrogen can
be detected by HS. Other gas was detected by TCD. Without carbon, the hydrogen reduction process
was followed to the formation of the intermediate product between MnO and FeO and finally
reduction to the mixture of MnO and Fe. With carbon, the intermediate products between MnO and
FeO was formed at about 780 K. The methane was formed in higher temperature than 1073 K and
the reduction with carbon proceeded mainly. At higher temperatures, methane decomposed to yield
nascent carbon that tended to result in the acceleration of the reduction rate with carbon.
The study is concerned with the mechanism of the hydrogen reduction of MnFe
2 4
O and the ef-
fect of without and with carbon on this reduction by means of combining EGA and HS.
Keywords: evolved gas analysis, humidity sensor, iron-manganese oxide, reduction process
Introduction
The iron-manganese double oxide, MnFe O , has been studied and summarized by sev-
2
4
eral investigators [1, 2]. However, the characteristics of high temperature carbothermic
reduction process of iron-manganese oxide were not clarified sufficiently. The vapor
pressure of metallic manganese formed in the reduction process is about 890 Pa at
1673 K. Therefore, the application of the gravimetric technique to this reduction process
seemed to be difficult because of the manganese loss through volatilization.
*
Author for correspondence: E-mail: hasizume@eng.toyama-u.ac.jp
1
418–2874/2002/ $ 5.00
Akadémiai Kiadó, Budapest
©
2002 Akadémiai Kiadó, Budapest
Kluwer Academic Publishers, Dordrecht

1
046
HASHIZUME et al.: IRON-MANGANESE OXIDE
It is well known that the hydrogen peak area measured by thermal conductivity
detector (TCD) of gas chromatograph is not in proportional to the hydrogen concen-
tration using helium as a carrier gas. Therefore, it is difficult to study the hydrogen re-
duction process quantitatively by TCD alone. As the evolution of water vapor can be
detected by HS using metal oxidized thin film, the simultaneous measurement of the
two techniques of TCD and HS is carried out to overcome this problem.
The humidity sensor can be used as a detector for various gas analysis inde-
pendently or in combination with gas detector such as TCD. This useful new method
was applied to the reduction process without and with carbon in pyrometallurgy in or-
der to obtain several useful infomations.
Experimental
Specimens
MnFe O was prepared by heating the mixture of MnCO (41–46% purity (as Mn))
and Fe O (99% purity ) powders. First, this mixture powders were formed into the
2
4
3
2
3
pellet and inserted into a ceramic tube placed in an electrical resistance furnace.
Then, the pellet was heated at 1373 K under a helium atmosphere for 43.2 ks. This re-
–
4
action product was crushed into 1.05×10 m particle size. The product was identified
to be single phase of spinel structure by X-ray analysis.
Carbon powder with high purity as a reducing agent was heated at 1473 K under
a helium atmosphere to eliminate any volatile materials.
Experimental apparatus
The isothermal and non-isothermal reduction processes were studied with EGA
method. In the hydrogen reduction process, H O evolved was measured by means of
HS. The experimental apparatus is illustrated in Fig. 1.
2
Fig. 1 Evolved gas analysis system: 1 – H
mixer; 4 – deoxidize; 5 – sample; 6 – thermocouple; 7 – reaction furnace;
– humidity sensor; 9 – ascarite trap; 10 – TCD
2
gas cylinder; 2 – He gas cylinder; 3 – gas
8
–
For the fabrication of HS, tantalum wire (99.95% purity) of 1.0×10 m in diame-
3
–
ter and 1.5×10 m in length was mechanically polished to obtain smooth surface. This
2
J. Therm. Anal. Cal., 69, 2002

HASHIZUME et al.: IRON-MANGANESE OXIDE
1047
wire was cleaned with aceton and oxided at 803 K during 19.8 ks in air. Vacuum
–
evaporation was used for the fabrication of the Au thin film of 1.4×10 m in thickness
7
–
2
and 1.0×10 m in length. Finally, this wire was polished up to remove the Au film for
lower electrode and bonded lead wire. The schematic illustration of HS is shown in
Fig. 2. This HS belongs to a group of surface sensitive type sensors, which detect the
evolution of water vapor by a change in surface electrical conductivity.
Fig. 2 Schematic illustration of humidity sensor
Sample was heated at a constant heating rate in a helium–hydrogen gas stream as a
carrier gas. When interaction such as chemical reaction or adsorption between the carrier
gas and sample occurs, the concentration of the hydrogen in the gas stream decreases,
and the concentration of H O is evolved from sample is measured by HS individually.
After elimination of H O by trap (Ascarite), the decrease of hydrogen concentration in
2
2
passing through the TCD cell was detected by TCD with high sensitivity. In the case of
methane formation, the gas evolved was detected only by gas chromatograph (GC). The
gases evolved were calculated in the same way as GC. The loss of manganese through
volatilization was calculated to be less than 0.12% at 1473 K.
Results and discussion
The typical reduction process of MnFe O in a helium-hydrogen gas stream and the
reduction process of MnFe O with carbon in a helium-hydrogen gas stream are
shown in Figs 3 and 4, respectively. As the evolution of H O was detected clearly by
2
4
2
4
2
HS corresponding to the decrease of hydrogen, it was thought that HS responded only
to the water vapor. In Fig. 3, the hydrogen reductions proceed in the temperature
range of 530–1040 K. The evolution of water vapor in the hydrogen reduction pro-
cess was followed to the formation of the intermediate product between MnO and
FeO and finally reduction to the mixture of MnO and Fe. In Fig. 4, since water vapor
was detected by HS at early stage, the reduction with hydrogen proceeded. At higher
J. Therm. Anal. Cal., 69, 2002

1
048
HASHIZUME et al.: IRON-MANGANESE OXIDE
2 4 2
Fig. 3 Evolved gas analysis curves for the reduction of MnFe O with H and He mix-
ing flow
2 4 2
Fig. 4 Evolved gas analysis curves for the carbothermic reduction of MnFe O with H
and He mixing flow
temperature above about 850 K, water vapor wasn’t evolved, but CO and CO were
2
detected by TCD. Therefore, the reduction with carbon proceeded mainly at higher
temperature of above about 850 K.
When the carbothermic reduction of MnFe O proceeded, non-stoichiometric
2
4
compound was also formed and reduced to metallic Fe and MnO up to 1273 K. At
higher temperatures, the manganese carbides such as Mn C and Mn C were formed.
3
7
3
When carbon was consumed entirely, the reaction between Mn carbide and MnO oc-
curred to yield metallic manganese. If carbon was added excessively to the
carbothermic reduction process, the rate determining reaction seemed to be a chain
reaction between CO and CO gas phases as reported by Maru et al. [3].
For the addition of hydrogen to the carbothermic reduction process of MnFe O ,
this reduction process resulted in the formation of methane gas. This methane forma-
2
2
4
J. Therm. Anal. Cal., 69, 2002

HASHIZUME et al.: IRON-MANGANESE OXIDE
1049
tion reaction between carbon and hydrogen proceeded actually at about 1073 K. It has
been mentioned in the previous paper [4] that at higher temperatures methane decom-
poses to yield the formation of nascent carbon which tended to result in the accelera-
tion of reduction rate with carbon. The rate equation for an interfacial reaction was
applicable to this reduction process.
The effect of hydrogen on the hydrogen added carbothermic reduction process
of MnFe O was examined in the temperature range of 843 to 1473 K. In the lower
temperature range of 843 to 863 K in Fig. 4, the activation energy for measured by HS
2
4
–
1
was estimated to be about 43.1 kJ mol . This value was calculated for interface rate
1
/3
determining process by using the relation of (1–(1–a) ) vs. time, where a is the frac-
tion of the volume that has already reacted. This value is nearly equal to that for the
hydrogen reduction of iron oxide, although the former reduction process differs from
the latter one. Therefore the interactions between MnO and C were often highly com-
plex, and there was a difference in reaction mechanism. At the higher temperature
range of 1373 to 1473 K, the rate equation for an interfacial reaction is applicable to
this system of MnFe O . It has been mentioned in the previous paper [4] that the acti-
2
4
–
1
vation energy for this reduction was estimated to be 188 kJ mol by TCD. This value
is smaller than that for the reduction of MnO with only carbon and nearly equal to
that for Boudouard’s reaction.
In the reduction process of several sorts of iron ores by the reformed natural gas,
it was reported by Kasaoka et al. [5], that the reducing action of natural gas was weak,
but, when the gas was decomposed the reduction became strong. This strong reducing
action would be given by hydrogen and nascent carbon formed by the decomposition
of hydrocarbon. Therefore, the fundamental data obtained in this study seems to be
useful to develop the direct one step reduction process of iron-manganese ores, and
contribute to save the energy of ferroalloy manufacturing industry.
It is concluded that the multi-technique of EGA and HS has made it possible to
clarify the hydrogen addition effect that takes place in the reduction process of man-
ganese oxide with carbon.
Conclusions
The kinetics of hydrogen reduction of iron-manganese oxide and the effect of hydro-
gen addition on the carbothermic reduction of MnFe O were studied by the new
2
4
technique, simultaneous measurement of evolved gas analysis method and the hu-
midity sensor. The results are summarized as follows:
1) The activation energy for the reduction of MnFe O in the hydrogen-helium gas
stream was calculated to be 43.1 kJ mol by HS in the temperature range of 843 to
2 4
–1
–
1
8
63 K. The activation energy for this reduction was estimated to be 188 kJ mol by TCD
at temperatures of 1373 to 1473 K.
) The carbothermic reduction process of MnFe O with hydrogen became clear by
2
2
4
TCD and HS. This reaction occurred in the temperature range of 530–1423 K, and the in-
termediate products between MnO and FeO was formed at about 780 K. The formation
J. Therm. Anal. Cal., 69, 2002

1
050
HASHIZUME et al.: IRON-MANGANESE OXIDE
of methane proceeded at about 1073 K, and the maximum peak was observed at 1323 K.
At higher temperatures, methane decomposed to yield nascent carbon that tended to re-
sult in the acceleration of the reduction rate with carbon.
3
) Simultaneous measurement of EGA and HS is useful and trouble-less tool for
the high temperature reduction processes of active materials such as manganese ox-
ide, since the vapor pressure of manganese is very high.
*
* *
The authors would like to thank Prof. N. Mizutani, Department of Inorganic Materials at Tokyo In-
stitute of Technology for his variable guidance and help rendered during this investigation. The au-
thors wish to thank Mr. M. Yamada who carried out the experiment as a graduate student of Toyama
University.
References
1
2
J. F. Butler, C. L. McCabe and H .W. Paxton, Trans. AIME, 221 (1961) 479.
T. Yagihashi, K. Asada, K. Atarashiya, S. Ichnohe and H. Hanada, Tetsu-to-Hagane,
49 (1963) 971.
3
4
5
6
7
Y. Maru, Y. Kuramasu, Y. Awakura and Y. Kondo, Metal Trans., 4 (1973) 2591.
K. Terayama and T. Ishiguro, Netsu Sokutei, 47 (1975) 1389.
N. Kasaoka, Y. Sakata and S. Takao, Nennryou Kyoukai Kaishi, 58 (1979) 741.
K. Terayama and T. Shimazaki, J. High Temp. Soc., 26 (2000) 36.
T. Hashizume, K. Terayama, T. Shimazaki and H. Itoh, J. High Temp. Soc., 27 (2001) 119.
J. Therm. Anal. Cal., 69, 2002