Sparse Modeling analysis of Extended X-ray Absorption Fine Structure data using two-body expansion

TL;DR

This paper introduces a sparse modeling approach for analyzing EXAFS data that utilizes a two-body expansion without prior structural assumptions, successfully extracting key structural parameters from experimental signals.

Contribution

It presents a novel sparse modeling method for EXAFS analysis based on two-body expansion, enabling structure determination without prior assumptions.

Findings

Successfully extracted radial distribution function peak positions.

Determined Debye-Waller factors for first neighbors.

Applied to metals and oxides experimental data.

Abstract

Analysis of extended X-ray absorption fine structure (EXAFS) data by the use of sparse modeling is presented. We consider the two-body term in the n-body expansion of the EXAFS signal to implement the method, together with calculations of amplitudes and phase shifts to distinguish between different back-scattering elements. Within this approach no a priori assumption about the structure is used, other than the elements present inside the material. We apply the method to the experimental EXAFS signal of metals and oxides, for which we were able to extract the radial distribution function peak positions, and the Debye-Waller factor for first neighbors.