Computational study of Li3BO3 and Li3BN2 I: Electrolyte properties of pure and doped crystals

Abstract

Both ${\mathrm{Li}}_{3}{\mathrm{BO}}_{3}$ and ${\mathrm{Li}}_{3}{\mathrm{BN}}_{2}$ materials have promising properties for use in all-solid-state batteries and other technologies dependent on electrolytes with significant ionic conductivity. As the first of a two-part study, this paper reports the analysis of detailed simulations of Li ion diffusion in the monoclinic forms of these materials. Using both NEB and MD methods, it is clear that Li ion migration via vacancy mechanisms provides the most efficient ion transport in each material. While the results suggest that interstitial defects in these materials do not play a direct role in Li ion migration, their relative stability seems to enhance vacancy production via the formation of Frenkel-type defects. This may partially explain why the Li ion conductivities computed from MD simulations of samples initially containing a single Li ion vacancy are in reasonable agreement with measured values of this work for ${\mathrm{Li}}_{3}{\mathrm{BO}}_{3}$ and those reported in the literature for poorly crystalline samples of both materials. The possibility of increasing vacancy concentrations by substitutional doping (F for O in ${\mathrm{Li}}_{3}{\mathrm{BO}}_{3}$ and C for B in ${\mathrm{Li}}_{3}{\mathrm{BN}}_{2}$) is also examined, finding simulated conductivities comparable to those of the ideal vacancy model.