Historically, the ionic conductivity of solids has been insufficient to supplant liquid electrolytes in battery applications. Very recently, however, a small number of solids with ionic conductivities comparable to that of liquids have been identified. The discovery of these fast ion conductors has advanced the prospects for realizing solid-state batteries. Nevertheless, additional study of these materials has, in essentially all cases, unearthed other shortcomings (stability, Li penetration, etc.), suggesting that the discovery of alternative solid electrolytes (SE) remains an important pursuit.

A fundamental question in the identification of new SEs is ‘what chemical/mechanical/structural features promote high ion mobility?’ We hypothesize that lattice disorder represents one such feature. The anti-perovskites (AP) present an ideal venue to explore the connection between disorder and ion mobility: the degree of disorder in the APs can be systematically tuned via atom substitution, yet the relative simplicity of the structure allows for a comprehensive characterization of ion migration pathways.

Here, we systematically probe the connection between ionic mobility and lattice disorder across a series of 24 model APs. Our calculations reveal that ‘disorder tuning’ provides a pathway to higher ionic conductivity. More specifically, a strong correlation is observed between the degree of lattice disorder and the minimum migration barrier for percolating ion migration: compounds with the largest disorder exhibit the lowest barriers for ion migration, independent of the migration mechanism (vacancy or interstitial). At the same time, the energetic spread between the largest and smallest migration barriers for a given compound increases with increasing disorder. Nevertheless, a higher degree of disorder also correlates with diminished thermodynamic stability, signaling a tradeoff with ionic mobility. Thus, realizing higher mobility will benefit from the development of strategies to kinetically stabilize disordered phases.