A Viscosity Estimation Model for Molten Slags in Al2O3‐CaO‐MgO‐SiO2 System

A model for viscosity estimation of molten slags in the Al₂O₂‐CaO‐MgO‐SiO₂ system is presented in this work. The model is an extension to the viscosity estimation model of molten slags in the CaO‐FeO‐MgO‐MnO‐SiO₂ system developed before by the present author. The present model has explicitly taken charge compensation into consideration. It is postulated that Al exists in a structural unit MAl₂O₄ when MO/ Al₂O₃ >1 for the Al₂O₃‐MO‐SiO₂ system (MO=CaO, MgO). MAl₂O₄ has a similar behaviour as SiO₂, i.e. it can form an Al‐O‐Al network and be depolymerised by network modifying oxides (CaO, MgO). The present model is applied in viscosity estimation of some slags within the Al₂O₃‐CaO‐MgO‐SiO₂ system. A mean deviation of less than 25% is achieved for the present model.     

Viscosities of the Quaternary Al2O3‐CaO‐MgO‐SiO2 Slags

Viscosities of some quaternary slags in the Al₂O₃‐CaO‐MgO‐SiO₂ system were measured using the rotating cylinder method. Eight different slag compositions were selected. These slag compositions ranging in the high basicity region were directly related to the secondary steel making operations. The measurements were carried out in the temperature range of 1720 to 1910 K. Viscosities in this system and its sub‐systems were expressed as a function of temperature and composition based on the viscosity model developed earlier at KTH. The iso‐viscosity contours in the Al₂O₃‐CaO‐MgO‐SiO₂ system relevant to ladle slags were calculated at 1823 K and 1873 K for 5 mass% MgO and 10 mass% MgO sections. The predicted results showed good agreement with experimental values and the literature data.     

Activity of MnO in MnO‐CaO‐MgO‐SiO2‐Al2O3 Slags at 1500 °C

A thermodynamic study was made on the MnO‐CaO‐MgO‐SiO₂‐Al₂O₃ slags that are typical of the production of ferromanganese in submerged arc furnaces. The Al₂O₃ content of the slags was kept constant at 5 per cent by mass. The activity‐composition relationship in Pt‐Mn binary alloys were re‐determined for calibration purposes at 1300, 1400 and 1500°C and po₂ values between 5.40×10^?6 and 4.54×10^?13 atm. A linear regression equation was derived to predict the activity coefficients of manganese, in Pt‐Mn alloys at 1500°C. The effect of concentration, basicity ratio and CaO‐to‐MgO ratio on MnO activities in above mentioned complex slags was investigated at 1500 °C and at two different po₂ values of 4.76×10^?7 and 5.80×10^?8 atm. It was found that a_Mno values increase with increasing MnO, and tend to increase with an increasing CaO‐to‐MgO ratio. The a_MnO values also increase with increasing basicity ratio. The activity coefficient of MnO increases with an increase in its mole fraction in the slag. Quadratic multivariable regression model equations which represent the activity data successfully and which can be used to predict the MnO activities in the compositional range of this study were developed. The MnO activity data was interpreted in terms of a slag model which describes the thermodynamic properties of the slag successfully.     

The Effect of Carbon in Slag on Steel Reoxidation by CaO‐SiO2‐Al2O3‐MgO‐MnO‐FetO Slags

An equilibrium study was carried out at 1873K to ascertain the effect of carbon in CaO‐SiO₂‐Al₂O₃‐MgO‐MnO‐Fe_tO slag systems on their Fe_tO and MnO activity coefficients, representing the slags’ thermodynamic potential for steel reoxidation. Both γf e_to and γm no showed not only a sharp increment but also a simultaneous slow decrement by increasing carbon content in slag, suggesting opposite roles of the carbon according to its stable forms. XPS (X‐ray photoelectron spectroscopy) was introduced to clarify the stable forms of carbon in slag. XPS results prove that carbon dissolves in slag as carbonate, and carbide ions under oxidizing and reducing atmospheres, respectively. It was concluded that carbonate ions increase γf e_to and γm no , but that carbide decreases them. This paper suggests an application method of the present results to actual ladle refining processes, in order to enhance steel cleanliness with maintaining (Fe_tO + MnO) in slag to some allowable amount.     

Thermodynamic Behaviour of Chlorine in CaO‐SiO2‐MgO‐Al2O3(‐CaF2) Slags

The solubility of chlorine in CaO‐SiO₂‐Al₂O₃‐MgO(‐CaF₂) slag was measured at 1673 ‐1823 K. By estimating the chloride capacity of slags, thermodynamic behaviour of chlorine in the molten slag was investigated. Chloride capacity increased with increasing CaO / SiO₂ ratio (C/S). An increase in MgO content decreased chloride capacity at C/S≥1.0 because it lowered the activity of Ca²⁺ which seemed to have strong affinity with Cl^? in molten slag. Also, the chloride capacity decreased with increasing Al₂O₃ content. The affinity between the Ca²⁺ and Cl^? ions was confirmed by measuring the infrared spectra of slags. The dissolution reaction of chlorine into slag was exothermic and its molar enthalpy was evaluated from the experimental results at 1673 ‐ 1823 K. Based on the result obtained in the present study, the quantitative prediction of chlorine distribution during the blast furnace process was performed. It was estimated that almost all chlorine in the blast furnace would be absorbed into molten slag even if the PCI ratio was increased or low quality coal with chlorine content less than 1.0 mass% was injected.     

Sulphide Capacities of CaO‐SiO2‐Al2O3‐MgO Slags

In Japanese steelworks, hot metal is now being produced by a scrap melting process. With this process, removals of sulphur is very much handicapped because of very high sulphur levels (0.04‐ to 0.09‐ pct by weight) and relatively low tapping temperatures (1623 to 1723 K). In order to overcome such handicaps, the authors explored on the respective phase diagrams. These explorations revealed that {CaO‐SiO₂‐Al₂O₃‐MgO} slags with Al₂O₃ contents of 30‐ to 35‐pct by weight would be good candidates as reagents for sulphur removal from high sulphur hot metal at relatively low temperatures. For better understanding of the thermodynamic properties of the candidate slags, in this study, sulphide capacities were determined through gas/slag equilibrium technique. The experimental results suggest that there would be, at least, a “window” to remove sulphur from high sulphur hot metal as relatively low temperatures.     

The Dissolution of Alumina in CaO‐SiO2‐Al2O3 Slags

The dissolution of alumina inclusions in CaO‐SiO₂‐Al₂O₃ based slags have been measured using laser scanning confocal microscopy (LSCM). Experiments were carried out over a temperature range of 1477 to 1577°C. It was found that diffusion coefficients calculated from the experimental results showed a dependence on the slag viscosity. This is considered strong evidence that the dissolution process is at least in part controlled by mass transfer in the slag phase. The diffusion coefficients for the alumina particles are estimated to be in the range 10^?11 to 10^?10 m²/s.     

Thermodynamics of manganese distribution between liquid iron and the CaO–Al2O3–SiO2–FeO–MnO system

The thermodynamics of distribution of constituents between liquid iron and the CaO–Al₂O₃–SiO₂–FeO–MnO system at 1600°C was studied using electrochemical indication of the equilibrium partial pressure of oxygen in both phases. The results show that oxidation potential of the Fe(l)–CaO–Al₂O₃–SiO₂–FeO–MnO system, expressed in terms of log p(O₂), is directly proportional to log (x(MnO) · x(FeO)/w| Mn |). Manganese distribution coefficient, L'_mn, in intersection CaO/Al₂O₃ = 1 decreases with increasing slag basicity expressed in terms of activity a(CaO) or 1/γ(MnO). Experimentally determined equilibrium constant K_Mn/Fe is equal to 2.7 for 1600°C. The number of exchanged electrons between Fe-O-Mn-Si electrode and the slag approaches the theoretical value.     

The Activities of FexO in {CaO‐SiO2‐Al2O3‐MgO‐FexO} Slags at 1723 K

In Japanese steelworks, hot metal is now produced by scrap melting process. With this process removal of sulphur is very much handicapped because of very high sulphur levels (0.04 to 0.09 pct by weight) and relatively low tapping temperatures (1623 to 1723 K). In order to overcome such disadvantages, the authors explored on the phase diagrams of {CaO‐SiO₂‐Al₂O₃‐MgO} slags, and this research revealed that those slags at 35 wt%‐Al₂O₃ would be good candidates as reagents for the removal of sulphur from high sulphur hot metal at relatively low temperatures. For better understanding of the thermodynamic properties of the candidate slags, in this study, activities of Fe_xO were determined by using solid‐state electrochemical cells incorporating MgO‐stabilized zirconia and Mo + MoO₂ reference electrode.     

On the Formation of Vanadium Ferrites in CaO–SiO2–FeO–V2O5 Slags

Selective extraction of valuable elements (such as V, Cr, Mn, etc.) and their compounds from metallurgical slag is in a focus of many researchers. Although vanadium may be present in slag as oxides and/or complex spinels with Fe, Mn, etc. during an alloyed steel production, the majority of vanadium in metallurgical slags typically exists as V₂O₅, which comprises up to 3–5 wt% of the slag in some cases. Due to the vanadium toxicity, these slags are forbidden in many civil engineering applications. As a result, hundreds of thousand tonnes of V₂O₅‐bearing slags are landfilled every year. In the present work, the formation of vanadium ferrites (FeV₂O₄ and Fe₂VO₄) in synthesized CaO–SiO₂–FeO–V₂O₅ slags containing 5 wt% V₂O₅ was examined under different partial pressures of oxygen. For the current slag chemistry range, an XRD analysis confirmed the presence of vanadium ferrite in slag samples treated at 1773 K in an argon atmosphere (PO₂ = 10^?1–10^?2 Pa) while no solid was noted in samples treated in air. Results were discussed based on thermodynamic consistency.     

Nitride capacities in CaO–SiO2 and CaO–SiO2–Al2O3 melts

Nitride capacities, , defined by (mass-% N) · in the CaO–SiO₂ and CaO–SiO₂–AI₂O₃ melts were measured in the temperature range of 1 723 to 1 923 K by a gas-slag equilibration technique using CaO, Al₂O₃, and Mo crucibles. Nitrogen content in slag, (mass-% N), was proportional to oxygen partial pressure, , to the power of ?3/4 at constant nitrogen partial pressure, . The values for increased linearly with increasing temperature and increased with the content of nitride formers, SiO₂ and Al₂O₃, but the effect of SiO₂ on value was found to be greater than that of Al₂O₃. The activity coefficients of Si_3/4N in the CaO–SiO₂ melts tended to increase with increasing the content of SiO₂.     

Thermodynamic assessment of CaO‐SiO2‐Al2O3‐MgO‐Cr2O3‐MnO‐FetO slags for refining chromium‐containing steels

The slag system of CaO‐SiO₂‐Al₂O₃‐MgO‐Cr₂O₃‐MnO‐Fe_tO relevant to refining chromium‐containing steels such as bearing steel is thermodynamically assessed at 1873 K. The activity coefficient of Fe_tO shows an initially rapid increment followed by a gradual reduction according to Cr₂O₃ content at a constant basicity, and decreases with increasing slag basicity. γ_MnO is decreased abruptly by increasing Cr₂O₃ content and thereafter, maintains a nearly constant level. From the standpoint of inclusion control, the Cr₂O₃ presence in ladle refining slags is thermodynamically harmful in that it minimizes the inclusion level by inducing the increment of γ_FetO even though Cr₂O₃ exists in extremely small amounts. However, it is beneficial in that it diminishes AI reoxidation by decreasing γ_MnO. The presence of carbon in slag decreases γ_FetO and γ_MnO, which turns out to be favourable for the reduction of Al reoxidation. The thermodynamic equilibria of chromium and manganese are quantified in terms of Fe_tO and Cr₂O₃ content as well as slag basicity by using multiple regression analysis. L_Cr and L_Mn are increased by the presence of Cr₂O₃, indicating a low recovery efficiency of Cr and Mn in the treatment of ferroalloy addition. In determining L_S values, Cr₂O₃ is not so important as the basicity of slags.     

Activity of VO1.5 in CaO‐SiO2‐MgO‐Al2O3 Slags at Low Vanadium Contents and Low Oxygen Pressures

In the present work, the gas‐slag equilibration technique was employed for the measurement of the thermodynamic activity of vanadium oxide. The vanadium‐containing slag kept in a platinum crucible was equilibrated with a gas mixture of CO, CO₂ and Ar, with well‐defined oxygen partial pressure at a pre‐determined temperature. The slag sample was quenched and the composition of the final slag was determined by chemical analysis. From the value of the oxygen partial pressure, the thermodynamic activity of VO_1.5 could be calculated using the value for the activity of vanadium in V‐Pt alloy. The measurements were carried out in the temperature range 1823～1923K and the oxygen partial pressures employed were 10^‐3, 10^‐4, 10^‐5 Pa. The present results show that the activity of vanadium(III) oxide in slag exhibits a negative deviation from ideality in the present composition range. With increasing basicity of the slag, the final content of vanadium oxide in the slag was found to show an initial increase followed by a constant content. The activities of vanadium(III) oxide did not exhibit any significant change with increasing temperature. The activity coefficient of vanadium(III) oxide decreased sharply with slag basicity approximately up to a basicity of 1, beyond which it showed a near–constant value. Increase in basicity was found to cause a change in the distribution of vanadium between the slag and the alloy phases even though this effect was less pronounced. From the present results, a mathematical relationship for estimating the vanadium content in slag for a given activity of vanadium in the molten metal phase was developed.     

Dissolution of Lime in Synthetic ‘FeO’‐SiO2 and CaO‐‘FeO’‐SiO2 Slags

Dissolution of different CaO samples into molten synthetic ‘FeO’‐SiO₂ and ‘FeO’‐SiO₂‐CaO slags was carried out in a closed tube furnace at 1873 K. The slag was kept stagnant. It was found that the dissolution rate was very fast when CaO rod was dipped into ‘FeO’‐SiO₂ slag. In the case of ‘FeO’‐SiO₂‐CaO slag, the dissolution of CaO rod in the stagnant slag was retarded after the initial period (2 minutes). Only less than 16 percent CaO reacted with the slag, irrespective of the type of lime. Three phase‐regions were identified in the reacted part of the lime rod by SEM‐EDS analysis. The formation of these regions was explained thermodynamically. A dense layer of 2CaO · SiO₂ was found to be responsible for the total stop of the dissolution. It could be concluded that constant removal of the 2CaO · SiO₂ layer would be of essence to obtain a high dissolution rate of lime. In this connection, it was found necessary to study the dissolution of lime in moving slag to reach a reliable conclusion regarding the relevance of the reactivity obtained by water ATSM test to the real reactivity of lime in high temperature slag.     

Water Capacity Model of Al2O3‐CaO‐MgO‐SiO2 Quaternary Slag System

The focus of the present work was to develop a water capacity model for the quaternary slag system Al₂O₃‐CaO‐MgO‐SiO₂. In the model, a silicate melt was considered to consist of two ion groupings, viz. cation grouping and oxygen ion. The water capacity of a melt is supposed to depend on the interactions between the cations in the presence of oxygen ions. These interactions were determined on the basis of the experimentally measured water solubility data. Only binary interactions were employed in the model. For the system CaO‐SiO₂, disagreement in the literature data was found. Since the interaction between Ca²⁺ and Si⁴⁺ would play an important role, experiments were carried out to determine the water capacities of some CaO‐SiO₂ slags. For this purpose a thermogravimetric method was employed. Iso‐lines of water capacities at constant MgO contents were predicted by the model and compared with the experimental data from literature. The model calculations agreed well with the experimental results.