The conductivity and heat capacity in the superionic phase were predicted by computer simulation, using an ab initio interaction model which included polarization effects. The results were compared with those obtained by Gillan, where an empirical effective pair potential had been used. In spite of large differences between the potentials at the pair-wise interaction level, the properties predicted by the 2 models were essentially identical. The simulation predictions were clarified with regard to the magnitude of the heat capacity anomaly at the superionic transition, and the number of defects present in the superionic state. In this regard, the Gillan results were considered to disagree with the body of experimental evidence on other fluorite-structured materials.

Simulation of CaF2 in the Superionic State - Comparison of an Empirical and Realistic Potential. A.Dent, P.A.Madden, M.Wilson: Solid State Ionics, 2004, 167[1-2], 73-81