Transport In Dispersed Ionic Conductors: Effect Of Insulating Particle Size

Dispersed ionic conductors are random mixtures of a solid salt, e.g. Agl, LiI, with fine particles of an insulating second phase, like A1203 or Si02. These composites can show a dramatic increase in ionic conductivity compared to the pure homogeneous system. Generally, this observation is attributed to an increased conductivity along the internal interface between the conducting salt and the insulating material. In this work a three-component random resistor network (RRN) model for dispersed ionic conductors is reviewed. In the model, the ionic conductor is represented by normally conducting bonds, the insulating material by non-conducting bonds and the interface between the two phases by highly conducting bonds. A special feature of the model is the existence of two critical concentrations of the insulating phase, p'c and p'c for interface percolation and bulk conduction, respectively, where critical transport properties corresponding to conductor/superconductor and conductor/insulator networks are predicted. The model describes satisfactorily the dependence on composition of the conductivity and activation energy of dispersed ionic conductors. Furthermore, the observed effect on the conductivity of the size of dispersed particles can be described qualitatively well by a generalized version of the RRN model, which in addition predicts a sensitive dependence of the critical thresholds on particle size. Non-universality features in the critical exponents for the conductivity are also discussed within a continuum percolation analog of the model. A three-component percolation model for DIC has been studied. In the model, DIC are considered as a matrix of bonds which can be normally conducting (representing the ionic conductor), non-conducting (representing the inert particles dispersed in the conductor) or highly conducting (representing the interface between the two phases). The model describes correctly the macroscopic transport properties of these composites, such as the dependence of ionic conductivity and activation energy on concentration p and size of the dispersed particles. In addition, the resistor model predicts critical behaviour of the conductivity close to the critical concentrations, p“cand p”, for interface percolation and conductor-insulator transition, respectively. Experimentally, the critical behaviour near pC'(which is very close to one) does not seem accessible since the composite is not stable at such large concentrations. There. © 1990, Taylor & Francis Group, LLC. All rights reserved.