Probing Structural Heterogeneity of Disordered Materials with Machine Learning Interatomic Potentials and X-Ray Absorption Spectroscopy

The precise determination of atomic structural information in functional materials holds transformative potential and broad implications for emerging technologies. Spectroscopic techniques, such as X-ray Absorption Near-Edge Structure (XANES), have been widely used for material characterization. However, extracting chemical information from experimental probes often intractable in disordered or heterogeneous systems. We present an integrated approach that combines atomic simulation, machine learning interatomic potentials, data-driven approaches and XANES calculations to investigate the chemical speciation of amorphous carbon nitride systems as a case study. We discuss the development of machine learning potentials that can efficiently explore the vast configurational space of amorphous carbon nitrides. By employing statistical methods, this structural database enables the elucidation of the most representative local structures, including carbon and nitrogen hybridization, and how they evolve with chemical compositions and density. Our simulations indicate that structure of amorphous carbon nitrides is highly complex, exhibiting non-linear and unexpected behavior at medium density and nitrogen concentration. Density functional theory (DFT)-based XANES simulations were carried out to establish a correlation between local structure and spectroscopic signatures. This serves as the basis for both forward and inverse models; the former predicts XANES spectroscopy when provided with macroscopic properties of density and chemical composition, while the latter enables a high-fidelity interpretation of local chemical motifs from experimental measurements. Although our framework is specifically applied to XANES and carbon nitrides, the approach described herein is readily adaptable to other experimental probes and materials classes.<br/><br/>This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344. Funding was provided by LLNL Laboratory Directed Research and Development (LDRD) Program Tracking No. 22-ERD-014.