Powder X-Ray Diffraction Assisted Evolutionary Algorithm for Crystal Structure Prediction

Stefano Racioppi,Alberto Otero De la Roza,Samad Hajinazar,Eva Zurek

2024-07-07

Abstract:Experimentally obtained X-ray diffraction (XRD) patterns can be difficult to solve, precluding the full characterization of materials, pharmaceuticals, and geological compounds. Herein, we propose a method based upon a multi-objective evolutionary search that uses both a structure's enthalpy and similarity to a reference XRD pattern (constituted by a list of peak positions and their intensities) to facilitate structure solution of inorganic systems. Because the similarity index is computed for locally optimized cells that are subsequently distorted to find the best match with the reference, this process transcends both computational (e.g. choice of theoretical method, and 0 K approximation) and experimental (e.g. external stimuli, and metastability) limitations. We illustrate how the proposed methodology can be employed to successfully uncover complex crystal structures by applying it to a range of test cases, including inorganic minerals, pure elements ramp-compressed to extreme conditions, and molecular crystals. The results demonstrate that our approach not only improves the accuracy of structure prediction but also significantly reduces the time required to achieve reliable solutions, thus providing a powerful tool for the advancement of materials science and related fields.

Materials Science

What problem does this paper attempt to address?

The paper aims to address the challenging problem of predicting crystal structures from powder X-ray diffraction (PXRD) data, especially when experimental conditions or theoretical limitations lead to a mismatch between the experimental diffraction pattern and the computed most stable structure. To tackle this issue, the authors propose a method that combines a multi-objective evolutionary algorithm, which considers not only the energy (or enthalpy) of the structures but also utilizes a metric for similarity to the reference PXRD pattern.

Specifically, the core of this method lies in the combination of the XtalOpt evolutionary algorithm and the use of the Variable-Cell Gaussian Powder Diffraction Similarity Index (VC-GPWDF). In this way, researchers can optimize structures to achieve the best match with experimental PXRD data while considering the enthalpy of the structures. This method is particularly suitable for handling complex inorganic systems, including pure elements compressed under extreme conditions, inorganic minerals, and molecular crystals.

The paper demonstrates the effectiveness of this method through several examples, such as:

1. **Brookite** (a metastable phase of titanium dioxide) structure prediction: Even though Brookite is a low-symmetry and metastable phase of titanium dioxide, by adjusting the weight parameters in the multi-objective search, the method successfully discovered the structure of Brookite in a relatively short time.

2. **Sodium structure prediction under high pressure**: For sodium samples obtained in dynamic or ramp compression experiments, the experimental data often contain high background noise, and the experimental conditions may deviate from theoretical predictions. By using a limited number of labeled peaks as references, the method successfully identified the structure of sodium under experimental conditions, revealing significant differences between the structure under experimental conditions and the standard 0K density functional theory (DFT) optimized structure.

Through these examples, the paper demonstrates that this method not only improves the accuracy of structure prediction but also significantly reduces the time required to obtain reliable solutions, thereby providing a powerful tool for advancements in materials science and related fields.