Spectra-to-Structure and Structure-to-Spectra Inference Across the Periodic Table

2025-06-13

AI Summary

X-ray absorption spectroscopy (XAS) analysis has long suffered from heavy reliance on domain experts, costly first-principles simulations, and poor generalizability of existing machine learning models—typically restricted to single elements or isolated absorption edges. To address these limitations, we propose XAStruct, the first framework enabling high-accuracy bidirectional inference between XAS spectra and local crystal structures. Methodologically, it integrates: (1) a spectrum-to-neighbor-atom-type classifier covering >70 periodic-table elements without element-specific hyperparameter tuning; (2) a unified bond-length regression model eliminating elemental bias; and (3) a dual-path training strategy combining a deep neural network (spectrum → structure), a lightweight regressor (structure → spectrum), a large-scale multi-element XAS–structure paired dataset, and physics-informed feature encoding. Results demonstrate significantly improved generalizability and accuracy across diverse elements and chemical environments, drastically reducing expert intervention and simulation overhead—thereby advancing XAS analysis toward a scalable, data-driven paradigm.

Technology Category

• Machine Learning: Structured Learning
• Humans and AI: Explainable AI (XAI) for Human Understanding
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Web Mining and Content Analysis: Bridging structured and unstructured data
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

X-ray Absorption Spectroscopy (XAS) is a powerful technique for probing local atomic environments, yet its interpretation remains limited by the need for expert-driven analysis, computationally expensive simulations, and element-specific heuristics. Recent advances in machine learning have shown promise for accelerating XAS interpretation, but many existing models are narrowly focused on specific elements, edge types, or spectral regimes. In this work, we present XAStruct, a learning framework capable of both predicting XAS spectra from crystal structures and inferring local structural descriptors from XAS input. XAStruct is trained on a large-scale dataset spanning over 70 elements across the periodic table, enabling generalization to a wide variety of chemistries and bonding environments. The model includes the first machine learning approach for predicting neighbor atom types directly from XAS spectra, as well as a unified regression model for mean nearest-neighbor distance that requires no element-specific tuning. While we explored integrating the two pipelines into a single end-to-end model, empirical results showed performance degradation. As a result, the two tasks were trained independently to ensure optimal accuracy and task-specific performance. By combining deep neural networks for complex structure-property mappings with efficient baseline models for simpler tasks, XAStruct offers a scalable and extensible solution for data-driven XAS analysis and local structure inference. The source code will be released upon paper acceptance.

Problem

• Interpreting XAS spectra without expert-driven analysis
• Overcoming element-specific limitations in XAS models
• Predicting local atomic structures from XAS spectra

Innovation

• Machine learning framework for XAS spectra prediction
• Predicts neighbor atom types from XAS spectra
• Unified regression model for neighbor distance