DONUT: Physics-aware Machine Learning for Real-time X-ray Nanodiffraction Analysis
This addresses a critical problem for materials scientists by enabling faster, automated analysis of nanoscale structural data, though it is incremental as it builds on existing physics-aware ML approaches.
The paper tackles the bottleneck of real-time analysis in scanning X-ray nanodiffraction microscopy, which is hindered by artifacts and computational demands, by introducing DONUT, a physics-aware neural network that predicts crystal lattice strain and orientation over 200 times more efficiently than conventional methods.
Coherent X-ray scattering techniques are critical for investigating the fundamental structural properties of materials at the nanoscale. While advancements have made these experiments more accessible, real-time analysis remains a significant bottleneck, often hindered by artifacts and computational demands. In scanning X-ray nanodiffraction microscopy, which is widely used to spatially resolve structural heterogeneities, this challenge is compounded by the convolution of the divergent beam with the sample's local structure. To address this, we introduce DONUT (Diffraction with Optics for Nanobeam by Unsupervised Training), a physics-aware neural network designed for the rapid and automated analysis of nanobeam diffraction data. By incorporating a differentiable geometric diffraction model directly into its architecture, DONUT learns to predict crystal lattice strain and orientation in real-time. Crucially, this is achieved without reliance on labeled datasets or pre-training, overcoming a fundamental limitation for supervised machine learning in X-ray science. We demonstrate experimentally that DONUT accurately extracts all features within the data over 200 times more efficiently than conventional fitting methods.