Reply to comment on 'Thermodynamic and elastic properties of a many-body model for simple oxides'

The comment discussed four issues: (1) The sensitivity of thermoelastic properties from approximate lattice dynamics (ALD) to the adjustable parameter; (2) the applicability of normalized lattice dynamics (NLD) to shear thermoelasticity; (3) the applicability of NLD to low-symmetry lattices; (4) Iab initioR versus empirical-potential models. This reply will reaffirm the following: (1) The mathematical derivation of ALD contains serious flaws that show in the results. The use of an adjustable parameter serves to minimize the errors, but cannot completely eliminate them. (2) Our NLD formulation puts the calculation on a physically justified ground, while eliminating the need in adjustable parameters and the ensuing inaccuracies. That formalism does not depend on the cubic symmetry, and so there is no room for unclarity in the reported computation of shear thermoelastic moduli. (3) NLD is valid to polyatomic lattices of any symmetry. Therefore NLD completely solves the problem of polarization shifts in potential-induced-breathing models. (4) At the present state of the art, Iab initioR potentials cannot compete with empirical potentials on accuracy and reliability.