Toward Single Particle Phases Mapping in Degraded Layered Oxide Cathodes via 4D-STEM

Lithium-ion batteries (LIBs) have risen to prominence as the primary power solution across a myriad of applications, encompassing consumer electronics, grid-scale energy storage, and a rapidly expanding share of the light-duty automotive sector. Indeed, despite their widespread adoption, the application of LIBs is constrained mainly by capacity limitations and stability as of now. It is widely recognized that even trace amounts of faulted phases and related defects within cathode particles can lead to substantial degradation in battery performance. This underscores the critical importance of precisely characterizing and understanding faulted phases to unlock the full potential of LIB technology. Despite extensive efforts to comprehensively elucidate the phases responsible for failure mechanisms through local nano- and even atomic-scale investigations especially using transmission electron microscopy (TEM) [1-3], the precise distribution and relationship of these faulted phases and defects within a single crystal cathode, at the scale of hundreds nanometer scale remains inadequately characterized. In this presentation, I will take LiNiO2 layered oxide, a promising avenue for advancing (LIB) cathode, as a model system for exploring the intricate phase distribution in single particle scale. Four-Dimensional Scanning Transmission Electron Microscopy (4D-STEM) is utilized as a powerful tool to unravel local structure with sub-nanometer precision across as large scale. In delithiation process at high voltage, LiNiO2 normally degrade from pristine O3 phase to faulted phases such as O1 and subsequent Rock Salt (RS) cubic phase. Simulations aid in elucidating these phase differences, discerning their unique positions of diffraction disks (Figure 1). Through masking their distinctive disks from experimental Convergent Beam Electron Diffraction (CBED) patterns, we quantitatively differentiate between the O3, O1 and rock salt phases at sub-nanoscale (Figure 2). The phase mapping reveals a distinct enrichment of the O1 phase at the edges, suggesting a preference for its formation along the surface region. In contrast, the RS phase exhibits a broad distribution extending from the surface to the bulk of the material. These distributions may be attributed in part to the reversible nature of the O1 phase and the irreversible nature of the RS phase during charging cycling. The surface's susceptibility to lithium-ion exchange facilitates the reversible formation of the O1 phase and subsequent surface enrichment. On the other hand, the loss of oxygen leads to the formation of the RS phase, which extends broadly from the surface to the bulk of the material. Our approach offers unparalleled spatial resolution, allowing for providing a full picture of the degraded phase distribution within the cathode single crystal on a sub-nanometer scale. Moreover, our study sheds light on the detection sensitivity of 4D-STEM on phase identification, showcasing its ability to capture subtle variations in phase distribution with high precision. By integrating 4D-STEM experiments with computational simulations, our research not only enhances the fundamental understanding of LiNiO2 particle degradation at the single crystal level but also provides practical insights for the advancement of high-performance cathodes in next-generation energy storage technologies [4]. Atomic models and simulated CEBDs of LiNiO2 O3 and faulted phases. (A) and (C) Atomic models along a-axis [100] and c-axis [001] direction. The LiNiO2 O3, O1 and Rock Salt (RS) structures are aligned with their phase transition geometrical relationship. (B) and (D) Corresponding simulated CBEDs. The reflection disks exhibit distinctive positions and intensities as highlighted in the red box. The convergence angle is 1.1 mrad for the a-axis and 2.2 mrad for the c-axis case. All simulations are based on a 50 nm thickness model. Note: The intensity of each disk may vary with thickness, but its position remains constant. Phases spatial distribution map of high voltage charged LiNiO2 nanoparticles obtained via 4D-STEM. As shown in the right panels in (A) and (B), CBEDs are taken along two perpendicular directions, [100] and [001], respectively, with convergence angles 1.1 and 2.2 mrad, respectively. Distinctive refection disks are masked by black circles for phase separation. The spatial distributions of phases are revealed in the right panels of (A) and (B). Since the difference of O3 and RS phase along c-axis is subtle, the O3 and RS phase are mapped jointly.