Thermodynamic modelling of complex solutions

摘要

Keith Brimacombe's renowned work on the modeling of metallurgical processes has served as an inspiration to all of us involved in this field. Our own work has focussed on thermodynamic modeling. With current software and databases, the modeling of the complex equilibria involved in real metallurgical processes is now a reality. General software for Gibbs energy minimization is capable of calculating equilibria involving many phases and hundreds of components. This software accesses large databases of evaluated properties of compounds and solutions. The databases for non-ideal solutions (slag, mattes, metals, etc.) are developed by evaluation and optimization of available experimental data, mainly for 2-component and 3-component sub-systems. These optimizations yield parameters of solution models which best describe all the available data. The parameters are stored in the databases, and the models are then used to predict the properties of multicomponent solutions. The predictions are only as good as the models. The proper choice of a model which corrctly reflects the structure, ordering, and speciation of a given solution is essential. In the present article we discuss the models which we have found to be most successful for liquid solutions of oxides (slags), slfides (mattes), metals, and molten salts, and we review the present state of our database development for these solutions. An example of application to Cu smelting and converting is also briefly presented.