Comparison of bulk basic properties with different existing Ni-Fe-O empirical potentials for Fe3O4 and NiFe2O4 spinel ferrites

Accurate empirical potentials for the simulation of magnetite Fe3O4 and nickel-ferrite NiFe2O4 spinel systems are of fundamental importance for understanding their structural stability. To better understand how existing empirical potentials for Ni-Fe-O systems describe the spinel physics, we perform comparisons of some of the most important bulk properties. Elastic constants, lattice parameters, energies and Debye temperatures are computed and compared with previously published data of density functional theory (DFT) and experiments found in the literature. We find that all the potentials predict the spinel geometry well whereas there are discrepancies in bulk properties. The MEAM becomes unstable at high temperature for NiFe2O4 , although it gives the best prediction of static properties at zero temperature whereas under induced pressure or high temperature, Buckingham types offer more stability. In general, for static properties and if computational speed is required -and in the case of Fe3O4 no distinction between normal or inverse is demanded- MEAM should be preferable. However, if dynamics at some temperature and specific ordering are important, Buckingham types, although more computa-tionally expensive, should be used.

Automated summary

Accurate empirical potentials for magnetite and nickel-ferrite spinel systems are crucial for understanding their structural stability. Comparisons of different potentials show that MEAM predicts static properties well at zero temperature, while Buckingham types offer more stability under high temperature or pressure.