Solid electrolytes offer significant safety advantages and potential for higher energy density compared to liquid electrolytes in lithium-ion batteries. However, achieving comparable ionic conductivity remains a challenge. A comprehensive understanding of lithium-ion transport requires investigation beyond the ideal periodic lattice structure to include non-periodic features such as point defects and grain boundaries, which are known to significantly impact ionic conductivity. While some studies have explored these known features, a complete picture necessitates the identification and characterization of additional non-periodic structural elements that might influence lithium-ion transport. This study aims to uncover such previously unconsidered features and elucidate their effect on ionic conductivity in solid-state electrolytes.

Existing research has highlighted the crucial role of non-periodic features like point defects and grain boundaries in influencing ionic conductivity. Studies have demonstrated that grain boundaries can drastically reduce conductivity in various solid electrolytes, including Li2OX anti-perovskites, Li7La3Zr2O12 garnets, and LLTO-based perovskites. Conversely, the introduction of interstitial Li+ point defects can enhance conductivity through cooperative mechanisms. Other non-periodic features, such as H+ substitutional defects from hydration or lithium-halide Schottky defect pairs, have also been shown to impact Li-ion migration. Despite these advancements, a significant gap remains in our understanding of various non-periodic features and their atomistic mechanisms influencing Li-ion migration. Most mechanistic studies are still limited to point defects and grain boundaries, leaving many other potential non-periodic features, such as dislocations, stacking faults, and twin boundaries, largely unexplored.

This study employed a multi-faceted approach combining advanced electron microscopy techniques and ab initio calculations. Li0.33La0.56TiO3 (LLTO), a prototype solid electrolyte known for its high bulk conductivity, was selected for investigation. The LLTO ceramics were synthesized using a sol-gel approach, and their structural and electrochemical properties were characterized using X-ray diffraction, inductively coupled plasma spectroscopy, and electrochemical impedance spectroscopy. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), energy-dispersive X-ray spectroscopy (EDX), and electron energy loss spectroscopy (EELS) were used to characterize the atomic structure and composition of the discovered defects. Ab initio calculations, including density functional theory (DFT) calculations and ab initio molecular dynamics (AIMD) simulations, were used to investigate the Li-ion transport properties in the presence of the defects and in the bulk material. The calculations employed the projector augmented-wave (PAW) approach with Perdew-Burke-Ernzerhof (PBE) generalized-gradient approximation (GGA) implemented in the Vienna ab initio simulation package (VASP). The geometrical phase analysis (GPA) software was used for the calculation of the strain map from the STEM image and ImageJ software was used for volume fraction calculation.

High-resolution STEM imaging revealed the ubiquitous presence of closed loops formed by single-atom-layer two-dimensional defects within the LLTO grains. These loops were termed "single-atom-layer traps" (SALTs). EDX mapping indicated a negligible La content within the defects, while EELS line scans revealed a significant enrichment of Li in the defect regions compared to the surrounding LLTO. Further EELS analysis indicated that Ti remained in the 4+ oxidation state within the defects, suggesting the presence of TiO6 octahedra similar to those in bulk LLTO, albeit with slightly different distortions. Atomic-resolution STEM observations, combined with the chemical analysis, strongly suggested that the two-dimensional defects are isostructural with the {001} plane of the rock-salt-structured γ-Li2TiO3. However, charge balance considerations and Li concentration analyses from EELS and DFT simulation imply a composition of approximately [Li0.37Ti0.33O]0.31, reflecting cation vacancies likely originating from Li deficiency. AIMD simulations indicated that Li-ion migration is effectively blocked across the single-atom-layer defects, and that Li-ion diffusion within the defect layer is significantly slower than in the bulk LLTO, attributed to shorter Li-O distances and higher migration barriers. The presence of SALTs isolates significant volumes of the material (estimated at 15.7 vol%), effectively rendering these regions non-conductive, thus causing a substantial reduction in overall ionic conductivity. This is comparable to the conductivity decrease expected from increasing porosity from 3.7% to 18.8%.

The discovery of SALTs as a significant non-periodic feature impacting ionic conductivity significantly advances our understanding of Li-ion transport in solid-state electrolytes. The finding demonstrates that the impact of non-periodic features on conductivity extends beyond previously considered factors such as grain boundaries and point defects. While further research is needed to compare the relative influence of SALTs and grain boundaries on ionic conductivity, the widespread presence and significant volume fraction of SALTs suggest a substantial contribution to the overall conductivity degradation observed in LLTO. The inability of Li-ions to traverse the SALT layers indicates a critical need to develop strategies to mitigate SALT formation during the synthesis of solid electrolytes to enhance their performance.

This study uncovered single-atom-layer traps (SALTs) as a new type of non-periodic feature that severely limits lithium-ion conductivity in solid-state electrolytes. The findings emphasize the need for a comprehensive investigation of various non-periodic structural elements to fully understand and improve the ionic transport properties of these materials. Future research should explore strategies for minimizing SALT formation during synthesis and further investigate the relative impact of SALTs compared to other non-periodic features on conductivity.

While the study provides strong evidence for the detrimental effect of SALTs on ionic conductivity, the quantitative comparison between the impact of SALTs and grain boundaries requires further investigation. The simulations were conducted at 2000 K, and further studies at lower temperatures might provide additional insights. The volume fraction of SALTs was estimated based on a limited number of randomly selected grains; a larger-scale statistical analysis might yield a more precise estimation.