Automated multiphase identification and refinement in powder diffraction using mismatch-tolerant machine learning
- 1Neutron Scattering Division, Oak Ridge National Laboratory , Oak Ridge, Tennessee 37831,
Abstract
Powder diffraction is a primary structural characterization tool in materials science, yet automated phase identification remains a major bottleneck for autonomous discovery. Existing workflows rely heavily on search–match heuristics and manual Rietveld refinement, and broadly usable end-to-end automation is especially limited for neutron powder diffraction, where comparable tools are largely absent. Here, we introduce Residual-Aware Deep-learning–Assisted Refinement for Powder Diffraction (RADAR-PD), a modality-aware machine learning framework for phase identification and quantification across both x-ray and neutron powder diffraction. RADAR-PD couples a mismatch-tolerant neural network operating on coarse momentum-transfer fingerprints with automated lattice nudging and physics-constrained Rietveld verification, enabling dominant-phase hypotheses to be generated from elemental constraints and secondary phases to be isolated recursively. On an experimental RRUFF powder x-ray diffraction benchmark, RADAR-PD outperforms data-driven automated Rietveld analysis (DARA) in recovering the reference phase. RADAR-PD further provides robust multiphase analysis on complex time-of-flight and constant-wavelength neutron datasets, addressing an important unmet need in automated neutron diffraction analysis. These results establish RADAR-PD as an auditable, instrument-agnostic framework for autonomous structural discovery.
Related
- AutoSolveX: automated batch structure solution, initial refinement and guest screening from diffraction data
- Rongzai agent: A Large Language Model-Based Autonomous Assistant for Rietveld Refinement of Neutron Diffraction Data
- Performance and Design Limits of Membraneless Flow Electrochemical Cells: A Combined Secondary Current and Multiphase Flow Analysis
- Relationships of Multiphase Structure with Flexibility and DC Insulation Performance in Impact-Resistant Polypropylene Copolymers
- Compliance Mismatch Drives Mechanochemical Reactivity in Diels-Alder Reactions
BibTeX
@article{yadav2026automated,
title = {{Automated multiphase identification and refinement in powder diffraction using mismatch-tolerant machine learning}},
author = {Lalit Yadav and Yongqiang Cheng and Mathieu Doucet},
journal = {APL Machine Learning},
year = {2026},
month = sep,
doi = {10.1063/5.0345499},
url = {https://doi.org/10.1063/5.0345499},
note = {CC BY},
}
RIS
TY - JOUR
TI - Automated multiphase identification and refinement in powder diffraction using mismatch-tolerant machine learning
AU - Lalit Yadav
AU - Yongqiang Cheng
AU - Mathieu Doucet
PY - 2026
DA - 2026/09/01
JO - APL Machine Learning
DO - 10.1063/5.0345499
UR - https://doi.org/10.1063/5.0345499
AB - Powder diffraction is a primary structural characterization tool in materials science, yet automated phase identification remains a major bottleneck for autonomous discovery. Existing workflows rely heavily on search–match heuristics and manual Rietveld refinement, and broadly usable end-to-end automation is especially limited for neutron powder diffraction, where comparable tools are largely absent. Here, we introduce Residual-Aware Deep-learning–Assisted Refinement for Powder Diffraction (RADAR-PD), a modality-aware machine learning framework for phase identification and quantification across both x-ray and neutron powder diffraction. RADAR-PD couples a mismatch-tolerant neural network operating on coarse momentum-transfer fingerprints with automated lattice nudging and physics-constrained Rietveld verification, enabling dominant-phase hypotheses to be generated from elemental constraints and secondary phases to be isolated recursively. On an experimental RRUFF powder x-ray diffraction benchmark, RADAR-PD outperforms data-driven automated Rietveld analysis (DARA) in recovering the reference phase. RADAR-PD further provides robust multiphase analysis on complex time-of-flight and constant-wavelength neutron datasets, addressing an important unmet need in automated neutron diffraction analysis. These results establish RADAR-PD as an auditable, instrument-agnostic framework for autonomous structural discovery.
SN - 2770-9019
ER -