AI-Driven Structure Refinement of X-ray Diffraction

Bin Cao; Qian Zhang; Zhenjie Feng; Taolue Zhang; Jiaqiang Huang; Lu-Tao Weng; Tong-Yi Zhang

cond-mat.mtrl-sciFeb 18, 2026

AI-Driven Structure Refinement of X-ray Diffraction

Authors: Bin Cao, Qian Zhang, Zhenjie Feng, Taolue Zhang, Jiaqiang Huang, Lu-Tao Weng, Tong-Yi Zhang

Organizations: Guangzhou Municipal Key Laboratory of Materials Informatics, Advanced Materials Thrust, Sustainable Energy and Environment Thrust, The Hong Kong University of Science and Technology (Guangzhou), Guangzhou, 511453, China · Materials Genome Institute, Shanghai University, Shanghai 200444, China · Hefei National Research Center for Physical Sciences at the Microscale, and Hefei National Laboratory, University of Science and Technology of China, Hefei, Anhui 230026, China · Material Characterization and Preparation Facility, Hong Kong University of Science and Technology (Guangzhou), Guangzhou 511400, China

Abstract

Artificial intelligence can rapidly propose candidate phases and structures from X-ray diffraction (XRD), but these hypotheses often fail in downstream refinement because peak intensities cannot be stably assigned under severe overlap and diffraction consistency is enforced only weakly. Here we introduce the whole-pattern expectation--maximization (WPEM) algorithm, a physics-constrained whole-pattern decomposition and refinement workflow that turns Bragg's law into an explicit constraint within a batch expectation--maximization framework. WPEM models the full profile as a probabilistic mixture density and iteratively infers component-resolved intensities while keeping peak centres Bragg-consistent, producing a continuous, physically admissible intensity representation that remains stable in heavily overlapped regions and in the presence of mixed radiation or multiple phases. We benchmark WPEM on standard reference patterns (PbSO $_4$ and Tb $_2$ BaCoO $_5$ ), where it yields lower $R_p/R_{wp}$ than widely used packages (FullProf and TOPAS) under matched refinement conditions. We further demonstrate generality across realistic experimental scenarios, including phase-resolved decomposition in multiphase materials, quantitative recovery of mixture compositions, separation of crystalline peaks from amorphous backgrounds in semicrystalline systems, high-throughput operando lattice tracking, automated refinement of compositionally disordered solid solutions, and quantitative phase-resolved analysis of complex archaeological samples from synchrotron powder XRD. By providing Bragg-consistent, uncertainty-aware intensity partitioning as a refinement-ready interface, WPEM closes the gap between AI-generated hypotheses and diffraction-admissible structure refinement on challenging XRD data.

Figures & tables

Figure 1: From AI-generated hypotheses to physically admissible refined structures. Given a batch of experimental diffraction patterns, the AI phase/structure identification model Xqueryer [ 8 ] proposes an initial structural hypothesis from the raw PXRD data. The subsequent EM–Bragg refinement performs Bragg-law-constrained whole-pattern decomposition to enforce diffraction consistency and yield a physically grounded refined structure. Finally, the property-prediction model PRDNet [ 5 ] estimates key materials properties (e.g., formation energy, band gap and elastic moduli) to support downstream screening and optimization.

Figure 2: Benchmarking and quantitative decomposition. a,b Whole-pattern decomposition of reference data (

\text{PbSO}{\vphantom{\text{X}}}_{\smash[t]{\text{4}}}

and

\text{Tb}{\vphantom{\text{X}}}_{\smash[t]{\text{2}}}\text{BaCoO}{\vphantom{\text{X}}}_{\smash[t]{\text{5}}}

) and comparison of agreement factors against FullProf and TOPAS. c Summary of profile factors for WPEM, FullProf and TOPAS (FPA). d Decomposition of a multiphase Ti–15Nb thin film into

\beta

\alpha

and

\alpha^{\prime}

contributions and the corresponding phase fractions estimated from integrated intensities. e Quantitative composition analysis of NaCl –

\text{Li}{\vphantom{\text{X}}}_{\smash[t]{\text{2}}}\text{CO}{\vphantom{\text{X}}}_{\smash[t]{\text{3}}}

powder mixtures showing agreement between inferred and nominal mass ratios.

Figure 3: High-throughput PXRD workflows enabled by WPEM. a In situ tracking of lattice-parameter evolution in $\mathrm{Li(Ni_{0.8}Co_{0.1}Mn_{0.1})O_{2}}$ during electrochemical cycling. b Automated solid-solution refinement of a Ru-substituted Mn oxide using a $3\times 3$ supercell model (CCDC 2530452).

Figure 4: Deciphering an ancient Egyptian make-up sample. a, Synchrotron PXRD profile (points) and WPEM fit (line) collected with

\lambda=0.96270\,\text{\AA}

. b–f, Phase-resolved component patterns recovered by WPEM for

\text{CaSO}{\vphantom{\text{X}}}_{\smash[t]{\text{4}}}\,{\cdot}\,\text{2}\,\text{H}{\vphantom{\text{X}}}_{\smash[t]{\text{2}}}\text{O}

(gypsum),

\text{Pb}{\vphantom{\text{X}}}_{\smash[t]{\text{2}}}\text{Cl}{\vphantom{\text{X}}}_{\smash[t]{\text{2}}}\text{CO}{\vphantom{\text{X}}}_{\smash[t]{\text{3}}}

(phosgenite),

\text{PbCO}{\vphantom{\text{X}}}_{\smash[t]{\text{3}}}

(cerussite), PbS (galena) and PbOHCl (laurionite), respectively. The decomposition quantitatively separates strongly overlapped reflections in this complex multiphase mixture, enabling phase-level interpretation directly from the full profile. The Egyptian portrait was generated by ChatGPT-5o and does not infringe any third-party copyrights.

Figure 5: Quantitative analysis of semicrystalline diffraction. a Decomposition of the PE form I pattern into four crystalline peaks and three amorphous halos. b Decomposition of the PE form II pattern into three crystalline peaks and three amorphous halos. c Decomposition of the PB-I diffraction profile into form I, form II and amorphous-halo contributions.

Materials Science

Materials science studies how composition, structure, processing, and defects determine material properties and how to design useful materials.

See all 301 papers

Diagnosing and narrowing the simulation-to-real gap in powder X-ray diffraction with a wet-dry agentic loop
Aug 23, 2026Shaoguang Wang, Weiyu Guo, Ben Fei +3Synthetic-to-Real Domain Adaptation Materials Science
The Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China. · The Chinese University of Hong Kong, Hong Kong SAR, China. · Suzhou National Laboratory, Suzhou, China. +2
PXRD-Bench: How Close Are We to Determining Crystal Structures from Powder Diffraction?
Oct 12, 2026Dongyeop Woo, Nofit Segal, Nayoung Kim +3Crystal Structure Prediction Materials Science
Korea Advanced Institute of Science & Technology (kaist.edu) · Massachusetts Institute of Technology (mit.edu)
XRDiff: Crystal Structure Prediction from Powder X-Ray Diffraction Data Using Diffusion Models
Jun 12, 2026Nofit Segal, Mingda Li, Benjamin Kurt Miller +1Crystal Structure Prediction Diffusion Models
Department of Materials Science and Engineering, MIT, Cambridge, MA, USA. · Department of Nuclear Science and Engineering, MIT, Cambridge, MA, USA. · FAIR, Meta, San Francisco, CA, USA.

Aug 23, 2026J/K move · Enter open · S save

Diagnosing and narrowing the simulation-to-real gap in powder X-ray diffraction with a wet-dry agentic loop

Shaoguang Wang, Weiyu Guo, Ben Fei, Xiaohong Shao+2

The Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China. · The Chinese University of Hong Kong, Hong Kong SAR, China. · Suzhou National Laboratory, Suzhou, China.+2

Synthetic-to-Real Domain Adaptation Materials Science Iterative Refinement

Open paper AskPDF arXiv

Powder X-ray diffraction (PXRD) is the routine probe of crystalline matter, yet its analysis is the rate-limiting step as laboratories automate acquisition. Deep-learning analyzers excel on simulated patterns and degrade on measured ones. This simulation-to-real gap is structural, not additive: synthetic denoising gives no measurable lift on real spectra, whereas correcting a small peak-position drift more than doubles median retrieval correlation. Real-spectrum fine-tuning, peak-aligned reranking, and recalibration narrow what remains and restore the coverage synthetic anchors lose. Xtalyst integrates these in an agent-orchestrated system spanning phase identification, refinement, and calibrated property prediction. On a frozen held-out partition (n=534) each module measured on both splits reproduces its development finding -- including the synthetic-anchor under-coverage, whose magnitude differs between the two pools -- while held-out refinement converges and preserves symmetry without reaching profile-quality fits, and on a diffractometer its wet-dry recommend-rescan-reanalyze loop flips a blinded silicon standard to a gated PASS and changes which minor phase is resolved on a multi-metal alloy.

Abstract

Materials Science

Explore similar work

Diagnosing and narrowing the simulation-to-real gap in powder X-ray diffraction with a wet-dry agentic loop

PXRD-Bench: How Close Are We to Determining Crystal Structures from Powder Diffraction?

XRDiff: Crystal Structure Prediction from Powder X-Ray Diffraction Data Using Diffusion Models