Quick Buzz Feed
Argonne scientists have created an AI-driven method that dramatically speeds up a powerful X-ray technique. The new approach reduces the number of measurements needed by as much as 80% while maintaining accuracy and reducing human error.
Artificial intelligence is revolutionizing scientific fields, with researchers at the U.S. Department of Energy's (DOE) Argonne National Laboratory at the forefront. Mathew Cherukara, a computational scientist at Argonne's Advanced Photon Source (APS), emphasizes AI's critical role in advancing chemical processes essential to American industry, promising significant breakthroughs.
The Argonne team has innovated an AI-driven method to drastically accelerate X-ray absorption near-edge structure (XANES) spectroscopy. This powerful analytical technique, crucial for understanding the internal chemistry of materials like batteries and catalysts, now performs with greatly reduced risk of human error and minimizes potential sample damage from X-ray beams.
This novel AI approach dramatically cuts the number of measurements previously required for XANES by as much as 80%, all while maintaining exceptional accuracy. The significant reduction in data acquisition time enables scientists to capture rapid chemical changes in materials as they occur, providing real-time insights into dynamic processes.
XANES operates by exposing a material to X-ray beams of progressively increasing energy. When the X-ray energy reaches a threshold sufficient to dislodge a tightly bound electron from an atom, the material exhibits a sharp increase in X-ray absorption, known as the absorption edge. By analyzing absorption changes around this edge, researchers can precisely monitor the chemistry of specific elements within the material, observing phenomena such as metallic catalyst reactions or battery charge state alterations.
Previously, XANES experiments demanded extensive manual decision-making from scientists, who had to determine dozens or hundreds of measurement points across varying X-ray energy levels. This manual process was often inefficient and prone to human error, as optimizing measurement density in chemically rich or sparse regions was challenging. The new AI algorithm eliminates this guesswork by automatically and intelligently selecting the most valuable measurement points. It identifies critical areas where the absorption edge is expected or where significant chemical detail resides, while bypassing regions with little new information. This makes the experimentation process smarter, faster, and more efficient, allowing researchers to focus on broader scientific questions.
Beyond mere speed, the AI-guided system facilitates entirely new capabilities: AI-directed experiments. By continuously comparing an evolving sample's spectrum against established baseline states (e.g., fully charged vs. fully discharged electrodes), the AI can provide real-time updates on the chemical progression. It autonomously determines when sufficient information has been collected, signaling when to advance the experiment. This capability signifies a major step towards increasingly autonomous X-ray beamlines, such as those at the Advanced Photon Source, which can intelligently optimize photon usage and track complex reactions dynamically. Mathew Cherukara highlights Argonne's commitment to developing more AI-driven tools, particularly with the upcoming upgrade to the APS, which will deliver X-ray beams up to 500 times brighter. The research utilized beamlines 25-ID-C, 20-BM, and 10-ID at the APS and was supported by the DOE Office of Science, Office of Basic Energy Sciences, published in npj Computational Materials.
The Advanced Photon Source (APS), a U.S. DOE Office of Science User Facility at Argonne National Laboratory, is a premier X-ray light source globally. It offers high-brightness X-ray beams for over 5,000 researchers annually, leading to significant discoveries in materials science, chemistry, physics, and life sciences. Argonne National Laboratory is a multidisciplinary research center dedicated to solving national science and technology challenges, managed by UChicago Argonne, LLC for the U.S. Department of Energy’s Office of Science. The DOE Office of Science is the leading supporter of physical sciences basic research in the U.S.