Bond susceptibilities and ionicities in complex crystal structures

Abstract

The concept of crystal ionicity has proved to be a useful unifying concept for understanding chemical trends in diverse problems in chemistry and solid state physics. In particular, the dielectric description of ionicity developed by Phillips and Van Vechten (PV) has been successfully employed in a wide variety of areas. However, the applicability of this PV dielectric analysis has been limited to only the simple AN B8-N compounds which contain only one type of bond. We have therefore generalized this approach to include such diverse structures as CsCl, PbS, GeTe, MgF2, SnO2, CaF2, Li2O, Al2O3, Ga2Se3, Cd3As2, Si3N4, HgS, Se, Te, LiGaO2, ZnGeP2, NaClO3, and AIPO4. Our approach is to decompose the bulk crystal properties into parameters which can be associated with individual bonds such as the polarizability, the ionicity, and the average number of electrons per bond. In order to determine the functional dependence of these bond properties on the fundamental variables such as the bond length, we use the basic dielectric description of PV for each type of bond, attempting to make the smallest number of additional assumptions. For example, we use the same relationship between the homopolar potential (Eh) and bond length (d) that PV do (i.e, Eh ∝ d−2.48). In addition to obtaining the individual bond properties we also present some striking empirical relationships between them. For example, we find that the Thomas-Fermi prescreening factor, which plays a fundamental role in the PV theory, is simply related to the average coordination number. Hopefully, this systemization of the dielectric properties will prove useful for understanding a variety of problems in physics and chemistry.