SPECIFIC FEATURES OF PHASE TRANSFORMATIONS
IN THE Mn-O SYSTEM
G.A. Kozhina1, N.V. Korchemkina2, V.B. Fetisov1, A.V. Fetisov2,
E.A. Pastukhov2
1Ural State Economic University, Ekaterinburg, Russia
2Institute of Metallurgy, Ural Department of Russian Academy of Sciences, Ekaterinburg, Russia
It is generally agreed [1, 2] that the reaction
6-Mn2O3→ 4-Mn3O4 + О2(1)
is virtually irreversible in the Mn-O system. A reverse reaction was observed only in the presence of -Mn2O3 or MnO2 as a catalyst. On the other hand, some data indicate that -Mn3O4 may oxidize to -Mn2O3 in a pure form at a considerable rate already at 400 C [3]. The experiment was performed using finely ground samples comprising the CuxMn3-xO4 system (0 < x < 1). We showed earlier [4] that the reverse of the reaction (1) was observed in the absence of nucleation centers of -Mn2O3 or MnO2. This observation refutes once more the proposition about the catalytic effect of these oxides on oxidation of hausmannite. The present study deals with verification of two hypotheses relating to factors that facilitate the -Mn3O4-Mn2O3 transformation. Is this transformation a manifestation of the mechanochemical effect or the catalytic effect of impurities contaminating the test material as samples are ground in a mill?
The objects of study were powder samples of kurnakite -Mn2O3 and hausmannite -Mn3O4. The samples were ground by three methods: by hand in an agate mortar (1) and in planetary micromills made of a WC + 6 mass % Co composite alloy (3) or a low-carbon steel (2). In the first case, particles were 5 m in size on the average. In the second and third cases the average size of particles was 2.3 m after the 30-min grinding run. X-ray diffraction and chemical analyses showed that samples 2 contained up to 15% WC, samples 3 included about 1% Fe, and samples 1 were free of impurities.
The manganese oxides were reduced and oxidized in a Q-1500D derivatograph in air at temperatures from 20 to 1000 C. The experiments were performed in platinum crucibles. Al2O3, which was calcined at 1300 C, served as the reference sample. All measurements were corrected for idle running of the installation.
The X-ray diffraction analysis was performed over the temperature interval from room temperature to 1000 C in situ in air using a DRON-UM1 automated diffractometer (CuK radiation, reflected-beam monochromator of pyrolytic graphite) with a high-temperature attachment type UVD-2000. Silicon with a = 5.43106(2) Å served as the external standard.
Figure 1 presents thermogravimetric curves for samples of hausmannite Mn3O4, which were prepared by manual grinding in an agate mortar (1), grinding in the Fe mill for 30 minutes (2), and grinding in the W mill for 30 minutes (3). The curves were recorded as the samples were heated to 980 C at a rate of 3.75 C/min and then cooled at the same rate. It is seen that the mass of the sample 1 remained almost unchanged both during heating and cooling. This fact confirms stability of the -Mn2O3 phase [1, 2].
The weight of the sample 2 increased slightly starting from 200C. Its weight stabilized at 600 C and did not change upon further heating and cooling. Probably, iron, which penetrated to the material during mill grinding, was oxidized first,
2Fe + 3/2O2 Fe2O3.(2)
Then hausmannite was oxidized. The X-ray diffraction analysis (XDA) at room temperature of samples 2, which were quenched from 700 C after 30-min isothermal annealing, revealed 10% of the -Mn2O3 phase and 90% of the -Mn3O4 phase. XDA was insensitive to a small amount of iron (about 1%) in the initial sample and, hence, to a small content of Fe2O3 in the oxidized sample. The sample 2 differed from the sample 1 by the grinding fineness. Probably, the mechanochemical effect did take place in this case, but only at low temperatures. At temperatures of 900-1000C the powder was sintered to a ceramic. Therefore, the mechanochemical effect was impossible in these experimental conditions. Hausmannite was not oxidized during cooling.
Fig.1. Thermogravimetric curves for samples of hausmannite Mn3O4,
prepared by different ways of grinding.
The sample 3 did not differ from the sample 2 by the grinding fineness, but the difference was the impurity content. Therefore, it was reasonable to compare these samples from the viewpoint of oxidation-reduction processes. The thermogravimetric curve of the sample 3 shows that the mass started growing at 400C and was a maximum at 900C, which was probably due to the oxidation process. The maximum gain in the weight was considerable as compared to the sample 2. As the sample was heated further, the weight decreased. This decrease was probably connected with reduction of the material. Remarkably, the mass of the sample increased again upon subsequent cooling, indicating reversibility of the process. The sample 3 obviously presented a special interest and detailed studies were performed using this sample.
Figure 2 presents thermogravimetric curve, which was recorded in the "heatingcooling" cycle at a rate of 7 C/min for an -Mn2O3 sample prepared by grinding in the W mill. It is seen from Fig. 2 that a gain in the weight took place at temperatures from 500 to 650 C. The maximum growth rate corresponded to 600C and was accompanied by the isothermal effect. The weight of the sample dropped starting from 940C. This was probably due to reduction of kurnakite. The process stopped at a temperature of 990C and was accompanied by the endothermal effect. A gain in the weight was observed during cooling at temperatures from 800 to 650C. The occurrence of the exothermal effect at 750C (which is attested to by the maximum rate of the process) pointed to the oxidation process.
Fig.2. Thermogravimetric curves for an -Mn2O3 sample
prepared by grinding in the W mill.
To see changes in the phase composition of the sample during the experiment described above, an additional high-temperature XDA was carried out. The following facts were established. The initial sample contained 85% -Mn2O3 and 15% WC. The tungsten carbide was oxidized at temperatures from 500 to 650C:
4WC + 2Mn2O3 + 9O2 4MnWO4 + 4CO2.(3)
This process ended before 700C. At this temperature the X-ray diffraction pattern contained lines corresponding to two phases: kurnakite -Mn2O3 and oxide MnWO4 having a monoclinic structure. This phase composition was preserved up to 940C. At 1000 C the X-ray diffraction pattern suggested the presence of MnWO4 and hausmannite -Mn3O4. In other words, kurnakite was reduced completely over the temperature interval from 940 to 1000C. As the sample was cooled from 1000C, its phase composition changed. The oxidation reaction
4-Mn3O4 + О2 → 6-Mn2O3(4)
took place, which was in agreement with the thermogravimetric data (Fig. 2). The composition of the sample, which was cooled to room temperature, included MnWO4, -Mn2O3, and just traces of the -Mn3O4 phase. Notice that the initial powder sample was sintered at temperatures from 900 to 1000C. Nevertheless, this sample was oxidized (almost completely) upon subsequent cooling. This observation invalidates the opinion about the mechanochemical effect of oxidation. Fineness of samples may have its effect only at temperatures below the sintering point. Most likely, the catalytic effect of W was decisive in this case.
In addition to high-temperature XDA, we also determined phase compositions of samples, which were quenched after high-temperature annealing, at room temperature. The obtained data agreed in general with results of high-temperature XDA. The only exception was that reduction of kurnakite started at a temperature slightly below 900C and ended at 980C.
Thus, the following conclusions can be drawn:
1) The mechanochemical effect does not explain comprehensively the obtained experimental results.
2) One may reasonably think that tungsten has a catalytic effect on oxidation of -Mn3O4 to -Mn2O3.
Acknowledgement
The authors are grateful to Dr. S.A. Petrova and Prof. A.Ya.Fishman for useful discussions.
The work was supported by the Program “Leading Scientific Schools”, grants 2022.2003.3 and 468.2003.3.
References
- Tretiakov Yu.D. Thermodynamics of Ferrites. Leningrad, Khimiya, 1967. 304 p.
- Balakirev V.F., Barkhatov V.P., Golikov Yu.V., Maizel S.G. Manganates: Equilibrium and Unstable States. Ekaterinburg, Ural Branch RAS, 2000. 397 p.
- Gillot B., Kharroubi M., Metz R., Legros R., Rousset A. // J. Solid State Chem. 1991. V.91. P.375-384.
- Fetisov V.B., Fetisov A.V., Korchemkina N.V., Ovchinnikova L.A., Pastukhov E.A., Fishman A.Ya. // Dokl. Aakad. Nauk. 202. V. 387. No. 3. P. 361-363.
3 - 1