Ion transport in concentrated electrolyte solutions

The design and optimization of industrial electrochemical reactors can benefit from accurate electrochemical models to predict the current, potential and concentration distributions in the electrolyte solution. An electrochemical model is a set of macroscopic equations for all the relevant physical quantities, like the potential, the concentrations, the temperature and the fluid velocity, that contains all the relevant couplings between fluxes and forces. An important problem in computational electrochemistry is the determination of the coefficients in these equations. The nature of the interactions between the chemical components in the solution determines how the coefficients vary with temperature, pressure and concentrations. Nowadays, the potential model is the most widely used electrochemical model. The demand for more accurate simulations and the ever-increasing power of computers drive electrochemists to look for more advanced - i.e. more general - electrochemical models. Therefore, special attention is given in this work to the macroscopic equations that describe the behaviour of electrolyte solutions. A very general set of equations is presented and the couplings between fluxes and forces are discussed. The simpler electrochemical models are naturally found by treating certain quantities as constant or by neglecting certain couplings between fluxes and forces. The answer to the question of the coefficients in the macroscopic equations lies in statistical mechanics. The basics of statistical mechanics are explained in this work. In the case of electrolyte solutions, the solvent molecules are abundant and may be modeled by a continuum. This simplification permits to treat only the ion-ion interactions explicitly. The interest in this approach lies in the fact that the ion-ion interactions can be modeled as hard sphere Coulomb interactions. In the mean spherical approximation (MSA), this physical model of the ion-ion interactions leads to analytical expressions for the coefficients. This is of course of great interest for computational electrochemistry. Before this work started, only the osmotic coefficient, the activity coefficients and the electrophoretic contribution to the Onsager coefficients were known in the MSA. The relaxation contribution to the Onsager coefficients was missing. An expression for this relaxation contribution is derived here. The charge and the diameter of the ions are microscopic parameters that characterize the ion-ion interactions. A sensitivity study is performed to analyze how the concentration dependence of the coefficients is influenced by these parameters. This also helps to find a procedure to get the diameters from measurements of the coefficients. Several ion diameters are obtained in this way by examining the experimental data on 18 aqueous binary electrolyte solutions. The theoretical plots agree well with the experimental data. The behaviour of mixed electrolyte solutions can be predicted from those same microscopic parameters. The multi-ion transport and reaction model (MITReM) is to be the successor of the potential model. Different MITReM's can be obtained by using different interaction theories for the coefficients. In the pseudo-empirical I-MITReM the ion-ion interactions are treated in a crude fashion that only guarantees a correct conductivity, whereas in the MSA-MITReM the coefficients are given by the MSA expressions. A theoretical comparative study is performed to determine the conditions under which the pseudo-empirical I-MITReM fails and the MSA-MITReM is recommended. A comparison of the pseudo-empirical I-MITReM and the MSA-MITRe