Multi-objective optimization improves crystal structure refinement precision
CrystalMO-TuRBO: Multi-Objective Trust-Region Bayesian Optimization for High-precision Joint Crystal Structure Refinement
Machine Learning
Summary
Crystal structure refinement means adjusting a model of a material's atomic arrangement so it matches experimental data. Traditional methods struggle when the data is complex or noisy, especially when combining different types of measurements like X-ray and neutron diffraction. The authors developed a new method that treats these data sources separately and searches in two phases: a broad exploration of possibilities, then a focused fine-tuning. This new approach achieves more accurate and reliable results on a real crystal sample compared to older techniques.
What this means in practice
- •For materials characterization labs: Refine crystal structures more accurately by jointly optimizing X-ray and neutron diffraction data with improved convergence and precision.
- •For synchrotron facilities: Enhance joint analysis pipelines for multi-modal diffraction experiments by automating the search for optimal structural parameters.
Authors
Joseph Agada, Yishu Wang, Arpan Biswas
Abstract
Crystal structure refinement is a fundamental inverse problem in materials characterization, where structural parameters are optimized to reproduce experimental diffraction data. Conventional approaches, such as least-squares and likelihood-based optimization, rely on local search and often struggle with non-convex, noisy, and highly correlated parameter landscapes, particularly when integrating multiple diffraction modalities. Joint refinement of X-ray and neutron data is especially challenging due to their complementary but competing sensitivities, which are typically combined through scalarized objectives requiring manual weighting and leading to suboptimal solutions. We propose CrystalMO-TuRBO, a multi-objective trust region Bayesian optimization architecture for joint crystal structure refinement. The method models X-ray and neutron discrepancies as separate objectives and transforms the problem into a normalized maximization setting. A two-phase optimization strategy is introduced: Phase 1 performs global exploration using parallel trust-region Bayesian optimization across multiple scalarizations to identify promising regions of the parameter space, while Phase 2 conducts localized refinement within a shrinking region to achieve high-precision solutions. This design explicitly separates global search from fine-grained optimization, addressing the unique accuracy requirements of refinement tasks. We evaluate the proposed method on experimentally collected X-ray and neutron diffraction data from single-crystal Ho2Ti2O7. Results demonstrate improved convergence, robustness, and parameter precision compared to classical refinement methods and Bayesian optimization baselines on refinement of a single-crystal pyrochlore material system.