Superionic conduction in low-dimensional-networked anti-perovskites

Fast cationic conduction is necessary for a number of solid-state technologies and is particularly critical for solid-state batteries as solid electrolytes are typically poor conductors. This article illustrates the concept of low-dimensional networked (low-DN) anti-perovskite via first-principles computations and shows that superionicity (i.e. the conductivity is above 1 ​mS ​cm⁻¹ at room temperature) can be achieved by lowering the dimensionality of the connected octahedra of the anti-perovskite. In particular, we use the Li–O–Cl model system and study the diffusion of Li in 3DN-Li₃OCl, 2DN-Li₄OCl₂, 1DN-Li₅OCl₃, and 0DN-Li₆OCl₄. We find that the lower the dimensionality, the lower the Li migration barriers and the higher the diffusion coefficients are. We attribute this improved ionic conduction to the decreased size of the bottlenecks and the softening of the rotation modes of the octahedra in the structure. To further explore the concept of low-DN anti-perovskites, we screen 256 model materials in the I-VI-VII group chemical space by computing the phase stability and bandgap of 3DN-X₃BA, 2DN-X₄BA₂, 1DN-X₅BA₃, and 0DN-X₆BA₄ (X ​= ​Li, Na, K, Rb; B ​= ​O, S, Se, Te; and A ​= ​F, Cl, Br, I). The calculations suggest that 20% of the structures might be synthesized and a number of them possess reasonable cationic migration barriers (<400 ​meV). This study puts forward a new principle for designing solid superionic conductors by lowering the dimensionality of the primitive structural units.