Towards atomistic understanding of Iron phosphate glass: a first-principles based DFT modeling and study of its physical properties

Abstract

Iron phosphate glasses (IPG) have been proposed as futuristic glass material for nuclear waste immobilization, anode material for lithium batteries and also as bioactive glass. In the last decade, there have been attempts to propose atomistic models of IPG to explain their properties from atomistic viewpoint and to predict their behavior in radioactive environment. In this paper, we seek to produce small scale models of IPG that can be handled within the scheme of Density Functional Theory (DFT) to study the electronic structure of this material. The starting models generated using Monte Carlo (MC) method [S. Singh and S. Chandra, Comp. Mat. Sci., 202, 110943, (2022)] were subsequently annealed (at 1000 K) using ab-initio molecular dynamics (AIMD). This removes coordination defects present in the MC models. The equilibrated structure at this temperature was then force-relaxed using conjugate-gradient (CG) optimization. This hybrid approach (MC + AIMD + 0K DFT-CG optimization) produced good atomistic models of IPG which can reproduce experimentally observed electronic band-gap, vibrational density of states (VDOS), magnetic moment of Fe, the elastic constants as well as optical and dielectric properties. Computationally expensive melt-quench simulation can be avoided using present approach allowing the use of DFT for accurate calculations of properties of complex glass like IPG.