Halide Substitution Effects on Lithium-Ion Diffusion in Protonated Antiperovskites

Abstract

Solid-state electrolytes (SSEs) for all-solid-state lithium-ion batteries are generating intense interest because these batteries can improve the safety and performance compared with devices fabricated with conventional, flammable liquid electrolytes. In these SSEs, it has been suggested that Li+ ion diffusion in the grain boundaries is hindered and is a critical determinant of the overall ionic conductivity (σ). However, Li+ ion diffusivities in the grain (DG) and the grain boundary (DGB) are difficult to determine experimentally, with few techniques capable of distinguishing the individual contributions. Here, we distinguished the DG and DGB for the protonated lithium antiperovskites (pLiAPs) SSEs: Li2OHCl, Li2OHBr, Li2OHF0.1Cl0.9, Li2OHF0.1Br0.9, and Li2OHCl0.37Br0.63. The measurements were obtained directly from 7Li pulsed-field gradient nuclear magnetic resonance (PFG-NMR) at 353 K. The 7Li PFG-NMR echo profiles were composed of two primary components with additional secondary oscillatory components – the so-called NMR diffraction phenomenon. The length scale separating the two main components corresponds to a diffusion length of ∼1.7 μm, which is thought to be the average grain size (by diameter). The short-range (≤1.7 μm) diffusion component associated with DG (≈10–11 m2/s) varied minimally with halide substitution, while the long-range (≥1.7 μm) component DGB (≈10–12 to 10–15 m2/s) was highly sensitive to the substitution of halides and closely correlated with σ. In addition, from the comparison of the ratio DGB/DG to Dτ (the Li+ ion diffusion coefficient estimated from the rotational correlation time, τc), it was determined that the fractional contribution of DG to σ is negligible; 0.01–0.04 in the pLiAPs studied here. These insights provide a fundamental understanding of the halide substitution effects on Li+ ion grain vs grain boundary diffusion and suggest that careful engineering of the grain boundaries at the microscopic level is necessary to achieve high-performance pLiAP SSEs.