Targeted Grain Boundary Modifications to Overcome Conductivity and Stability Challenges in Lithium Solid-State Electrolytes

With the growing demand for energy-dense batteries, inorganic solid-state electrolytes have sparked academic and industrial interests owing to their inherently high energy densities and better safety. Among these, Li-garnet LLZO (Li₇La₃Zr₂O₁₂) and Li-perovskite LLTO (Li_0.33La_0.56TiO₃) have emerged as frontrunners based on their promising bulk properties. Not only do these oxides improve existing battery systems, but they also enable the applications of energy-dense anodes and high-voltage cathodes. Despite the promise, grain boundaries in both materials hinder ionic conduction and promote electronic conduction, reducing the local ionic-to-electronic transference number. This impacts key battery performance metrics such as power density and overall lifespan. A mechanistic understanding is a key prerequisite for developing strategies to improve detrimental grain boundary interfaces. In this study, we show that low transference numbers at Li oxide grain boundaries arise from space charge formation. Charged point defects tend to segregate and accumulate at the grain boundaries leading to excess charges, which create Schottky-type potential barriers (measuring 0.15 V for LLZO and 0.32 V for LLTO at 20 °C). These barriers cause a depletion of ionic carriers and an accumulation of electronic carriers in the adjacent space charge layers. Our space charge model provides a clear explanation for the thermodynamic origins of low ionic conductivity in LLTO and high electronic conductivity in LLZO at the local boundaries. In light of these insights, we propose new electrolyte design principles for tailoring the local defect chemistry, carrier density, and transport dynamics of Li ceramic grain boundaries: First, by selectively choosing the sintering atmosphere (oxygen partial pressure), we influence intrinsic defect equilibrium and lower space charge accumulation at the grain boundaries. Second, by using aliovalent doping (donor dopants), we introduce net-opposite extrinsic charges that counterbalance the intrinsic space charge at these grain boundaries. Collectively, the fundamentals guide engineering strategies leading to control of local potentials at grain boundaries, enabling four times faster ion conduction for Li-perovskite LLTO and a threefold improvement in short-circuit current limit for Li-garnet LLZO.