Constrained Crystal Structure Prediction

The aim of first-principles crystal structure prediction (CSP) is computational identification of a compound’s experimentally isolable polymorph(s). The structure generation stage of CSP typically leads to hundreds of thousands of candidate structures across multiple space groups. However, in many use cases, experimental powder X-ray diffraction (PXRD) data is available from high-throughput polymorph screening studies, but the crystal structure is not yet solved. For such cases, we propose using indexed lattice parameters to constrain the CSP search space. Searching only specific unit-cell dimensions allows a large reduction in the number of structures generated, whilst adding confidence that the experimental structure is retained in the search pool when advancing to density-functional reranking of the most likely candidates. Examples are shown for the cases of several small organic molecules, specifically uracil, paracetamol, progesterone, and four forms of the prolific polymorph-former, ROY. The accuracy of the constrained CSP (C-CSP) workflow is found to be highly dependent on the quality of the force field used for initial structure generation. Provided the force field is sufficiently reliable, C-CSP can provide more than a tenfold reduction in computational cost relative to conventional CSP.

Authors

Institutions

Publication Details

Journal
Crystal Growth & Design
Published
2026-11-18
DOI
https://doi.org/10.1021/acs.cgd.6c01435
Primary Topic
Machine Learning in Materials Science
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Constrained Crystal Structure Prediction

Erin R. Johnson, Grace M. Sparrow, R. Alex Mayo
1 links
Crystal Growth & Design
Machine Learning in Materials Science
article

Constrained Crystal Structure Prediction

Erin R. Johnson, Grace M. Sparrow, R. Alex Mayo
article en

Abstract

The aim of first-principles crystal structure prediction (CSP) is computational identification of a compound’s experimentally isolable polymorph(s). The structure generation stage of CSP typically leads to hundreds of thousands of candidate structures across multiple space groups. However, in many use cases, experimental powder X-ray diffraction (PXRD) data is available from high-throughput polymorph screening studies, but the crystal structure is not yet solved. For such cases, we propose using indexed lattice parameters to constrain the CSP search space. Searching only specific unit-cell dimensions allows a large reduction in the number of structures generated, whilst adding confidence that the experimental structure is retained in the search pool when advancing to density-functional reranking of the most likely candidates. Examples are shown for the cases of several small organic molecules, specifically uracil, paracetamol, progesterone, and four forms of the prolific polymorph-former, ROY. The accuracy of the constrained CSP (C-CSP) workflow is found to be highly dependent on the quality of the force field used for initial structure generation. Provided the force field is sufficiently reliable, C-CSP can provide more than a tenfold reduction in computational cost relative to conventional CSP.

Crystal Growth & Design
Dalhousie University (CA), University of Cambridge (GB)
Royal Society, Natural Sciences and Engineering Research Council of Canada
Machine Learning in Materials Science
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon