X-ray diffraction(XRD)is a powerful method for determining a material’s crystal structure in highthroughput experimentation,and is widely being incorporated in artificially intelligent agents for autonomous scientifi...X-ray diffraction(XRD)is a powerful method for determining a material’s crystal structure in highthroughput experimentation,and is widely being incorporated in artificially intelligent agents for autonomous scientific discovery.However,rapid,automated,and reliable analysis of XRD data at rates that match the pace of experimental measurements at a synchrotron source remains a major challenge.To address these issues,we developed CrystalShift for rapid and efficient probabilisticXRD phase labeling employing symmetry-constrained optimization,best-first tree search,and Bayesian model comparison.The algorithm estimates probabilities for phase combinations without requiring additional phase space information or training.We demonstrate that CrystalShift provides robust probability estimates,outperforming existing methods on synthetic and experimental datasets,and can be readily integrated into high-throughput experimental workflows.In addition to efficient phase labeling,CrystalShift offers quantitative insights into materials’structural parameters,which facilitate both expert evaluation and AI-based modeling of the phase space,ultimately accelerating materials identification and discovery.展开更多
基金supported by the Air Force Office of Scientific Research under award number FA9550-18-1-0136This work is based upon research conducted at the Materials Solutions Network at CHESS(MSN-C)which is supported by the Air Force Research Laboratory under award FA8650-19-2-5220+1 种基金Experimental work was performed in part at the Cornell NanoScale Facility,an NNCI member supported by NSF Grant NNCI-2025233Additionally,this research was conducted with support from the Cornell University Center for Advanced Computing.Use of the Stanford Synchrotron Radiation Lightsource,SLAC National Accelerator Laboratory,is supported by the U.S.Department of Energy,Office of Science,Office of Basic Energy Sciences under Contract No.DE-AC02-76SF00515.
文摘X-ray diffraction(XRD)is a powerful method for determining a material’s crystal structure in highthroughput experimentation,and is widely being incorporated in artificially intelligent agents for autonomous scientific discovery.However,rapid,automated,and reliable analysis of XRD data at rates that match the pace of experimental measurements at a synchrotron source remains a major challenge.To address these issues,we developed CrystalShift for rapid and efficient probabilisticXRD phase labeling employing symmetry-constrained optimization,best-first tree search,and Bayesian model comparison.The algorithm estimates probabilities for phase combinations without requiring additional phase space information or training.We demonstrate that CrystalShift provides robust probability estimates,outperforming existing methods on synthetic and experimental datasets,and can be readily integrated into high-throughput experimental workflows.In addition to efficient phase labeling,CrystalShift offers quantitative insights into materials’structural parameters,which facilitate both expert evaluation and AI-based modeling of the phase space,ultimately accelerating materials identification and discovery.
