Application of Potentiometric Ion-Breakthrough to Assess Individual Diffusion Coefficients of Ions in Ion-Selective Membranes

Abstract

Ion-selective membranes (ISMs) are at the core of ion-selective electrode development. Fundamentally, two groups of parameters determine the response of ISMs: selectivity coefficients and diffusion coefficients of mobile species in the membrane. It is possible to assess both by performing a single potentiometric ion-breakthrough experiment. Basically, the ISM is placed between two contacting electrolyte solutions that do not contain the ion that the ISM is selective for (primary ion). After primary ion is added the potential trace carries valuable information about the thermodynamics and the kinetics of the membrane. So far, extracting parameters from the experimental results was possible only after unrealistic simplifications (e.g. assuming all of the diffusion are the same). The state-of-the-art simulation technique the Nernst-Planck-Poisson finite element method is utilized to give insight on how the different physico-chemical processes generate the measured potential. Numerical simulations are used to train a feedforward neural network, in order to learn the connection between the physico-chemical parameters (e.g., thickness, diffusion coefficients, selectivity coefficients, coextraction etc.) and the shape of ion-breakthrough potential trace. By using the trained neural network it was possible to quickly obtain for the first time the diffusion coefficient of all of the mobile species in the ISM.