Note
High Temperature Microscope Observation of Melt Formation at the Interface between Crystalline Olivine and Wüstite
2015 Volume 55 Issue 6 Pages 1210-1212
Details
2015 Volume 55 Issue 6 Pages 1210-1212
In order to elucidate the slag melting behavior of sinters, the in-situ observation of the melt formation behaviors at the interface between synthesized crystalline olivine and wüstite bulks was carried out using a high temperature microscope over a temperature range between 1080 and 1120°C. The melting instantly happens at the interface in the olivine side at the temperatures higher than 1100°C although the thickness of the melting zone does not vary with time during the heating. The thickness of the melting zone is ca. 20 μm irrespective of heating temperature as well as heating duration. The melt formation is considered to be due to the fact that the olivine composition slightely shifts toward 2FeO·SiO2 by substituting Ca2+ with Fe2+ owing to the diffusion of Fe from wüstite into olivine, lowering the liquidus temperature.
In order to reduce CO2 emission in the steel industry, research and development oriented toward low reduction agent ratio (RAR) operation of a blast furnace is significantly required. Currently, Japanese steel industries are developing an innovative technology for low RAR operation through the Japan’s national project COURSE50.1) In the course of the development, the gas permeability in the cohesive zone of a blast furnace is one of the crucial issues to solve. To improve the gas permeability, it is necessary to control the softening behaviors of sinter in the cohesive zone. Softening and melting behaviors of sinter in the cohesive zone, which lead to the shrinkage of the packed bed and the increment in the pressure loss, have been intensively elucidated by many researchers for the last three decades. It has been revealed that the softening and melting is relevant to the melt formation of the slag in the vicinity of wüstite, and that the decrement in the reduction rate of sinter, i.e., the increment in the wüstite volume in the cohesive zone becomes the gas permeability worse.2,3)
Recently, the melt formation of the slag in the vicinity of wüstite has been directly observed using a scanning laser microscope mounted with a heating device by Kawabata et al.4) They have heated mixed powders of premelting slag and wüstite, which simulates the slag in sinter, at a temperature increasing rate of 5°C/min in a flow of Ar gas, and have found that the olivine like melt is initially formed at the interface between two powders at ca. 1075°C. They have also found that the melt formation behavior can be qualitatively explained by the Fact Sage thermodynamic calculation. However, the measurement temperature at which the melt volume abruptly increases has been found to be ca. 40°C higher than the calculated temperature. Their results imply that diffusions and/or chemical reactions do not instantly take place, and that the melt formation of the slag should be considered not only thermodynamically but also kinetically. Such a kinetic contribution might result in the formation of metastable phase as well.
According to the aforementioned reports, it is considered that olivine and wüstite play an important role in the melt formation of the slag. As a consequence, the authors aimed at carrying out the in-situ observation of the melt formation behaviors at the interface between synthesized crystalline olivine and wüstite bulks using a high temperature microscope so as to clarify the lowest temperature at which the melt can be formed and the temperature dependency of the melt formation rate.
Wüstite was synthesized from regent grade Fe and Fe2O3 powders. Powder mixtures of Fe and Fe2O3 (Fe : Fe2O3 = 0.79 : 1) were uniaxially pressed to form a tablet. The tablet was sintered in a flow of CO–CO2 mixture (CO2/CO = 0.5) for 8 hours at 1100°C and then quenched into water. It was confirmed by X-ray diffraction (XRD) analysis that the sample has the single phase of wüstite. A dense wüstite was obtained by melting the sintered tablet in an optical floating zone furnace in a flow of pure argon deoxidized by zirconium sponge heated at 1073 K.
Olivine (CaO·FeO·SiO2) was synthesized from regent grade CaCO3 and SiO2 powders as well as wüstite made in this study. Equimolar mixtures of CaCO3, SiO2 and FeO powders were uniaxially pressed to form a tablet. The tablet was put on a basket made of stainless steel (SUS304) wire, which was suspended by a nichrome wire in the maximum temperature zone of an electric furnace. After suspending the sample in the reaction tube, the sample was heated up to 1300°C at a constant heating rate of 5°C/min in a flow of CO–CO2 mixture, where the CO/CO2 ratio is set to have an equilibrium oxygen partial pressure within the FeO stable region. The temperature of the sample was monitored using a Pt-13 pct Rh/Pt (R type) thermocouple, positioned just above the sample. The sample melted, and dripped from the basket into an iron crucible settled beneath the furnace. It was confirmed by XRD analysis that the sample has the single phase of olivine (see Fig. 1).
XRD profile of the olivine sample.
It is known that olivine has a solid solution with a wide composition range. In order to confirm that the composition of the synthesized olivine is the one having the aimed composition, i.e., CaO·FeO·SiO2, the melting point of olivine (if the olivine is the congruent melting compound CaO·FeO·SiO2) or the solidus temperature of olivine (if the olivine has the solid solution of either 2FeO·SiO2 or 2CaO·SiO2) was determined by differential thermal analysis (DTA). About 40 mg of the pulverized powdery sample with a particle diameter of ca. 80 μm was subjected to the DTA measurement. The measurement was carried out at a constant heating rate of 10°C/min in a flow of CO–CO2 mixture (CO2/CO = 1). The measurement temperature was calibrated using NaF and MgF2 melting points as references.
Wüstite and olivine were both cut into a cuboid of ca. 1.5 mm×ca. 1.5 mm×ca. 1 mm using a precision cutting machine, which were subjected to the high temperature microscope observation. Wüstite and olivine samples were made contact with each other in a crucible made of an alumina plate (0.5 mm in thickness) and an alumina tube (4 mm in inner diameter). A crucible having the samples was set in the high temperature microscope composed of an infrared image furnace and an optical microscope mounted with a CCD camera. The temperature of the sample was measured by a thermocouple fixed at the sample holder. The temperature was calibrated using Au, Ag and Cu melting points as references. In order to carry out the high temperature microscope observation, the sample was heated up to a desired temperature of either 1080, 1090, 1100, 1110 or 1120ºC within a minute in a flow of CO–CO2 mixture (CO2/CO = 0.6) and was kept at the constant temperature for 20 min. Subsequently, the electric power of the furnace was shut off, and the sample was cooled down by a furnace cooling. The cooling rate is estimated to be ca. 50°C/s until 800°C. The cross sectional areas in the vicinity of the interface between two samples were observed by electron probe microanalyzer (EPMA). The back-scattered electron (BSE) images and the line profiles of the constituent elements for the observation areas were obtained, where the diameter of an electron beam was ϕ1 μm. An accelerating voltage of 15 kV and a probe current of 20 nA were used.
It can be concluded from the DTA measurement that the solidus temperature of olivine is 1191°C (see Fig. 2) because the olivine sample is composed of a single olivine phase according to the XRD analysis. The CaO–SiO2–FeO ternary phase diagram demonstrates that the melting point of CaO·FeO·SiO2 is ca. 1220°C,5) and that the olivine with the solidus temperature of 1191°C has the solid solution of 2FeO·SiO2. Namely, the FeO concentration in the olivine is higher than the nominal composition, i.e., CaO·FeO·SiO2.
DTA curve of the olivine sample.
The optical microscope observation of the samples has revealed that the melt cannot be observed at all for duration of 20 min after the onset of heating at constant temperatures of 1080 and 1090°C. On the other hand, the melting takes place along the interface between wüstite and olivine in the olivine side immediately after the temperature reaches a desired temperature higher than 1100°C. Figure 3 shows the optical micrographs of the samples, taken just after the onset of heating, where the temperature is still close to the room temperature (a), and taken in ca. 10 min after the temperature reaches 1100°C (b). The samples in the right and left hand sides are wüstite and olivine, respectively. Although the samples look like as if they have a gap in between as shown in Fig. 3(a), they actually contact with each other below the surface. The melt formation instantly happens, but does not progress afterward. However, the thickness of the melting zone is only about 20 μm, and does not vary during the duration of 20 min at a constant heating temperature.
Optical micrographs of the samples heated at 1100°C, taken just after the onset of heating, where the temperature is still close to the room temperature (a), and at ca. 10 min after the onset of heating (b).
Figure 4 shows the BSE image of the cross section and the line profiles of the elements Ca, O, Fe and Si for the samples heated at 1100°C for 20 min. It can be seen from the BSE image that the meniscus can be observed at the interface in the olivine side, which implies that the olivine sample has melted at the interface. While the Si concentration is almost constant throughout the olivine sample, the Ca and Fe concentrations of the olivine sample at the interface are smaller and larger than those apart from the interface, respectively. It is considered from the line profiles that the olivine composition shifts toward 2FeO·SiO2 by substituting Ca2+ with Fe2+ owing to the diffusion of Fe2+ from wüstite into olivine. As a consequence, the liquidus temperature of olivine at the interface decreases from 1191°C to 1100°C. The concentration gradients observed at the interface in Fig.4 might be relevant to the high viscosity of the melt; the mass transfer due to convection hardly takes place. This might also be the reason that the thickness of the melting zone does not vary with time during the heating.
BSE image of the cross section and the line profiles of the elements Ca, O, Fe and Si for the samples heated at 1100°C for 20 min.
The in-situ observations of the melt formation behaviors at the interface between synthesized crystalline olivine (the solidus temperature of which is 1191°C) and wüstite bulks were carried out using a high temperature microscope. It has been found that the melting takes place along the interface in the olivine side immediately after the temperature reaches a desired temperature higher than 1100°C. However, the thickness of the melting zone is only about 20 μm, and does not vary while the sample is held for the duration of 20 min at a constant heating temperature. The conclusion may imply the possibility that because the mass transfers of Fe2+ and Ca2+ ions in olivne is very slow, the composition of olivine solid solution hardly changes by the reaction between olivine and wüstite. On the other hand, the previous researchers have reported that the olivine-like melt is formed as a primary melt of sinter around 1120°C in a blast furnace.6,7) Consequently, the primary melt of sinter may not be formed by the reaction between olivine and wüstite but formed by the melting of olivines with low liquidus temperatures (FeO-rich olivine) independently. Such olivines are originally contained in sinters.
