In Situ X-ray Absorption Near-Edge Structure Calculation and Machine Learning Analysis of the Structural Evolution in Lithium-Ion Battery Cathode Materials

The development of lithium-ion battery cathode materials has improved the overall performance of the battery, while the unresolved irreversible structural evolution that occurs at high voltage hinders their practical application. In situ X-ray absorption spectroscopy is beneficial to research the electronic and atomic structure information related to the complex mechanism. Massive X-ray absorption near-edge structure spectra calculation based on ab initio code and assisted by machine learning can solve the related problems more effectively. In this work, theoretical calculations of the in situ Mn/Co/Ni K-edge XANES spectra of Li1.2Mn0.54Co0.13Ni0.13O2 during the initial charging process were carried out. We have achieved fine-tuning the ratio of Mn and Ni, XANES simulations were conducted to achieve cluster analysis, which showed the influence of tetrahedral Mn (Mn-tet), octahedral Mn (Mn-oct), and Ni (Ni-2b) in the transition metal layer on the spectra. With several analysis methods, it was determined that Mn-tet does not exist, and only when the Mn-oct/Ni-2b ratio decreased from 0.78/0.4 to 0.48/0.2 did the calculated spectra of Mn/Ni coincide with the experimental spectra. This method provides a qualitative and quantitative analysis model for predicting the local structural evolution of cathode materials during the initial charging process and indirectly evaluates the reliability of the experimental data.