Machine Learning-Aided Multimodal Spectroscopy for Advanced Local Structure Characterization

Atomistic structures of materials offer valuable insights into their functionality. Determining these structures remains a fundamental challenge in materials science, especially for systems with defects. While both experimental and computational methods exist, each has limitations in resolving nanoscale structures. Core-level spectroscopies such as x-ray absorption (XAS) or electron energy-loss spectroscopies (EELS), in particular, have been used to determine the local bonding environment and structure of materials. Recently, machine learning (ML) methods have been applied to extract structural and bonding information from XAS/EELS, but most of these frameworks rely on a single data stream, which is often insufficient. In this work, we address this challenge by integrating multimodal ab initio simulations, experimental data acquisition, and ML techniques for structure characterization. Our goal is to determine local structures and properties using EELS and XAS data from multiple elements and edges. To showcase our approach, we used various lithium nickel manganese cobalt (NMC) oxide compounds which are used for lithium ion batteries, including those with oxygen vacancies and antisite defects, as the sample material system. We successfully inferred local element content, ranging from lithium to transition metals, with quantitative agreement with experimental data. Beyond improving prediction accuracy, we find that multimodal data, compared to a single data source, exhibits higher sensitivity to local structural changes and is more robust to noise, enabling the characterization of complex structures that might be challenging or even impossible to capture using only a single data source. Furthermore, our framework is able to provide physical interpretability, bridging spectroscopy with the local atomic and electronic structures.