Data-Driven Lithium-Ion Battery Cathode Research with State-of-the-Art Synchrotron X-ray Techniques

Abstract

Metrics & MoreArticle Recommendations CONSPECTUS:The lithium-ion battery (LIB) is a tremendously successful technology for energy storage thanks to its favorable characteristics including high energy density, long lifespan, affordability, and safety. It has been widely adopted in sectors including consumer electronics and electric vehicles, which are featured by an enormous market value. To meet the everincreasing demands for energy density and cycle life, industry and academia are continuously devoting efforts to improve the current LIB technology. This requires an in-depth understanding of the electrochemical reaction processes and degradation/failure mechanisms, to which advanced characterization is pivotal. Combining advanced synchrotron X-ray techniques with machine learning (ML) methods has been demonstrated as a powerful tool for uncovering the fundamental reaction and aging mechanisms in LIB and is emerging as an important research frontier. Our group's research has been focusing on the battery cathode, which is a major limiting factor in today's LIB technology. The degradation and failure of cathode materials in LIB are multiscale. The chemo mechanical processes at these different length scales are intertwined and mutually modulated. Therefore, it is crucial to understand the underlying mechanisms of charge−lattice− morphology−kinetics interactions in battery cathodes as a function of the electrochemical states. Synchrotron X-ray technology has unique advantages. It can detect lattice structure, electronic structure, chemical valence state, and multiscale morphology in different experimental modes, with high resolution and high efficiency. However, the large-scale experimental data bring great challenges in terms of reduction, analysis, and interpretation. Data-driven methods based on ML can greatly assist researchers to understand, control, and predict the electrochemical behavior of the complex battery cathode systems.In this Account, we focus on showcasing the integration of synchrotron and ML techniques for LIB cathode research. We review our recent findings on charge−lattice−morphology−kinetics in LIB cathode materials via this approach. First, the ML-based morphological study of cathode materials is discussed, highlighting a ML-assisted automatic feature recognition, particle identification, and statistical analysis of the prolonged cycling-induced particle damage and detachment from the carbon matrix. Second, we discuss the chemical heterogeneity and lattice deformation in cathode materials revealed by ML-assisted multimodal synchrotron characterizations. The role of ML tools in identifying and understanding chemical outliers and lattice defects in NCM cathodes is highlighted. Third, we provide our perspective on a future "dream" experiment for investigating the spatial distribution of cation−anion redox coupling effects in the battery cathode by means of resonant inelastic X-ray scattering (RIXS) imaging with ML. We anticipate that this new approach will provide new horizons for the development o...

Machine-learning strategies for the accurate and efficient analysis of x-ray spectroscopy

“…Propelled by continuous improvements in hardware, software, and infrastructure, computational spectroscopy has become an indispensable tool for the modern spectroscopist that is capable of providing predictions-and, consequently, interpretations-of experimental spectroscopic observables across the electromagnetic spectrum. The predictive power of computational spectroscopy is perhaps best showcased within x-ray spectroscopy [7][8][9][10] where the transformative effects of next-generation light sources [11,12] are pushing the limits of the technique, facilitating new insights into the structure and dynamics of molecules and materials as well as opening up new possibilities across a wide range of research fields [13][14][15][16][17][18][19][20][21][22]. The remarkable progress in x-ray spectroscopy continues to stimulate concomitant progress in theoretical techniques to ensure that data can be accurately and affordably analysed, setting up one of the most effective experiment-theory feedback loops [23].…”