Computational Studies of the Effect of Nitrogen Dopants and Oxygen Vacancies on Li+ Conduction in Lithium Lanthanum Titanate Electrolytes

Abstract

Our recent experiments have shown that nitrogen doping can effectively improve the ionic conductivity of lithium lanthanum titanate (LLTO) electrolytes. Herein, computational studies are performed to study the mechanisms of ionic conductivity with density-functional theory and machine learning approaches. To elucidate the relationship between nitrogen concentration and oxygen vacancy, the energy for nitrogen doping into the lattice and the formation energy for oxygen vacancy were computed. Oxygen vacancies were found more likely to be created at higher level of nitrogen dopants. The calculations also discovered the potential formation of nitrogen bonds, generating vacancy-like pathways for Li+ conduction. In addition, the minimum-energy pathways for Li+ migration in pristine and various doped LLTO materials were analyzed using the nudged-elastic band method. To correlate energetics and structural features of the different LLTO materials, several machine learning methods were successfully constructed, where the neighboring sites with oxygen vacancies are identified as the major factor influencing the energies and hopping barriers compared with nitrogen dopants. The new insights can guide the design and synthesis of anion dopped oxide electrolytes in solid-state lithium-ion batteries.