Development of a thermodynamic database for nickel-containing oxide systems for simulation of nickel extraction from laterite ores

An extensive collaborative research program was focused on the development of a self-consistent thermodynamic database for simulation of nickel extraction from laterite ores. Chemical systems relevant to laterite ore processing were experimentally investigated and thermodynamically optimized.;In general, laterite ore deposits consist of heterogeneous mixtures of hydrated iron oxides and hydrous magnesium silicates. The basis of the laterite ore is olivine (Fe,Mg)2SiO4, which often contains small amounts of nickel due to the proximity of ionic radii of Fe2+, Mg2+ and Ni2+. Existing technologies that currently produce nickel worldwide utilize only about half of the nickel laterite deposits; other sources are not utilized due to complex mineralogy. For creation of cost-effective, environmentally-friendly and energy-efficient processes of Ni recovery from laterite ores, it is important to be able to perform reliable calculations of phase equilibria in the Al--Ca--Cr--Fe--Mg--Ni--O--Si system, where Fe, Al2O3, MgO, NiO and SiO2 are the major components of slags and oxide phases in nickel pyrometallurgy.;The current research program met this challenge. It was aimed to develop a thermodynamic database for NiO-containing oxide systems in the Al2O3--CaO--FeO--Fe2O3--MgO--NiO--SiO2 (Al--Ca--Fe--Mg--Ni--O--Si) multi-component system at high temperature of interest to nickel pyrometallurgical processes.;The project was accomplished by the joint efforts of two research groups. The database development was carried out at the Centre for Research in Computational Thermochemistry, Montreal, Canada by means of thermodynamic modeling, which was closely related to experimental study of phase equilibria performed by our colleagues from the Pyrometallurgy Research Centre (PyroSearch), Brisbane, Australia. This collaborative approach greatly increased the effectiveness of the overall program by reducing the amount of the required experimental work and providing specific experimental data for thermodynamic modeling. Thermodynamic assessments were applied to identify priorities for experiments and experimental measurements were planned to provide specific data for thermodynamic modeling.;A literature review and critical assessment of the previously published thermodynamic and phase equilibrium data on the NiO-containing systems were performed using thermodynamic modeling. The parameters of the models were optimized to fit a large variety of the literature data collected from the literature, including phase equilibrium data, thermodynamic properties (heat capacity, entropy, enthalpy, Gibbs energy) and cation distribution data. However, for several low-order subsystems, data were missing in the literature because they are of no direct importance for practical applications. In case there was not enough data to constrain the model parameters, or significant discrepancies in the available data were revealed, an experimental program was suggested to our colleagues from the PyroSearch Centre. A limited number of experimental measurements were planned for temperatures and compositions which were found to be most useful for thermodynamic modeling. In this way, the amount of the work required to obtain an accurate thermodynamic description of a multicomponent system was significantly reduced.;The experimental procedure involved the high-temperature equilibration in controlled gas atmospheres and ultra rapid quenching followed by electron probe X-ray microanalysis (EPMA) of quenched samples. Since the analysis took place after equilibration, the changes in composition during equilibration did not affect the accuracy of the results. Tie-lines between equilibrated liquid and solid phases were measured directly, providing essential data for subsequent thermodynamic modeling.;The whole set of experimental data, including the new experimental results and previously published data, was taken into consideration in thermodynamic modeling of oxide phases in the Al2O3--CaO--FeO--Fe2O3--MgO--NiO--SiO2 (Al--Ca--Fe--Mg--Ni--O--Si) multi-component system at a total pressure of 1 atm and a wide range of temperatures and oxygen partial pressures. The thermodynamic modeling part was undertaken using the FactSage thermochemical software and its databases. All solid and liquid phases of 4 binary: CaO-NiO, MgO-NiO, NiO-SiO2, Al2O3-NiO, 7 ternary: CaO-MgO-NiO, CaO-NiO-SiO2, MgO-NiO-SiO2, Al2O3-NiO-SiO2, Al2O3-MgONiO, Al2O3-FeO-Fe2O3, FeO-Fe2O3-NiO, 5 quaternary: CaO-MgO-NiO-SiO2, Al2O3-FeO-Fe2O3-NiO, FeO-Fe2O3-MgO-NiO, FeO-Fe2O3- NiO-SiO2, CaO-FeO-Fe2O3-NiO, and 2 quinary systems, Fe-Mg-Ni-O-Si and Ca-Fe-Ni-O-Si, of the multi-component chemical system were optimized in the present study. The optimizations are self-consistent and consistent with existing FToxide and FSstel metallic databases of the FactSage software. The applied models are based on the structure of the corresponding solution. The Modified Quasichemical.;Model, which takes into consideration second-nearest-neighbor short-range cation ordering, was used for the slag (molten oxide) phase. The models based on the Compound Energy Formalism have been developed for the olivine, spinel and pyroxene solid solutions. A simple random mixing model with a polynomial expansion of the excess Gibbs energy was used for the monoxide and corundum solid solutions. A set of self-consistent Gibbs energy functions was obtained that provides the best possible description of thermodynamic properties and phase equilibria in the chemical system. The literature data are reproduced within experimental error limits. Using optimized model parameters, valuable predictions of phase equilibria in the multicomponent systems have been made.;The properties of the spinel, monoxide, olivine and corundum solutions in the Al--Fe--Ni--O, Al-- Mg--Ni--O, Fe--Mg--Ni--O, Ca--Fe--Ni--O and Fe--Mg--Ni--O--Si chemical systems and of slag in the Ca--Fe--Ni--O--Si system have been predicted solely from the optimized model parameters for the corresponding binary (ternary) solutions. The subsequent comparison with available thermodynamic and phase equilibrium data has shown that applied physically meaningful models have excellent ability to predict phase relations in multicomponent systems. In this way, the effectiveness of the applied coupled experimental/modeling approach has been demonstrated.;The current database has been incorporated into the existing FactSage databases. By this means, the range of applications of FactSage databases has been expanded, and the existing databases have been updated to describe the most recent and accurate experimental data of interest to industrial operations.;The obtained database along with software for Gibbs energy minimization allows the prediction of liquidus and solidus phase equilibria as well as thermodynamic properties under the conditions, such as range of compositions, temperatures and oxygen partial pressures, which are most useful for metallurgical operations and engineering practice. The database is relevant to various kinds of pyrometallurgical processes for laterite ores, such as reduction roasting as well as electric furnace smelting.