CrystalGRW: Generative Modeling of Crystal Structures with Targeted Properties via Geodesic Random Walks

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Tangsongcharoen, Krit, Pakornchote, Teerachote, Atthapak, Chayanon, Choomphon-anomakhun, Natthaphon, Ektarawong, Annop, Alling, Björn, Sutton, Christopher, Bovornratanaraks, Thiti, Chotibut, Thiparat
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908478186979328
author Tangsongcharoen, Krit
Pakornchote, Teerachote
Atthapak, Chayanon
Choomphon-anomakhun, Natthaphon
Ektarawong, Annop
Alling, Björn
Sutton, Christopher
Bovornratanaraks, Thiti
Chotibut, Thiparat
author_facet Tangsongcharoen, Krit
Pakornchote, Teerachote
Atthapak, Chayanon
Choomphon-anomakhun, Natthaphon
Ektarawong, Annop
Alling, Björn
Sutton, Christopher
Bovornratanaraks, Thiti
Chotibut, Thiparat
contents Determining whether a candidate crystalline material is thermodynamically stable depends on identifying its true ground-state structure, a central challenge in computational materials science. We introduce CrystalGRW, a diffusion-based generative model on Riemannian manifolds that proposes novel crystal configurations and can predict stable phases validated by density functional theory. The crystal properties, such as fractional coordinates, atomic types, and lattice matrices, are represented on suitable Riemannian manifolds, ensuring that new predictions generated through the diffusion process preserve the periodicity of crystal structures. We incorporate an equivariant graph neural network to also account for rotational and translational symmetries during the generation process. CrystalGRW demonstrates the ability to generate realistic crystal structures that are close to their ground states with accuracy comparable to existing models, while also enabling conditional control, such as specifying a desired crystallographic point group. These features help accelerate materials discovery and inverse design by offering stable, symmetry-consistent crystal candidates for experimental validation.
format Preprint
id arxiv_https___arxiv_org_abs_2501_08998
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CrystalGRW: Generative Modeling of Crystal Structures with Targeted Properties via Geodesic Random Walks
Tangsongcharoen, Krit
Pakornchote, Teerachote
Atthapak, Chayanon
Choomphon-anomakhun, Natthaphon
Ektarawong, Annop
Alling, Björn
Sutton, Christopher
Bovornratanaraks, Thiti
Chotibut, Thiparat
Materials Science
Statistical Mechanics
Machine Learning
Computational Physics
68T07
Determining whether a candidate crystalline material is thermodynamically stable depends on identifying its true ground-state structure, a central challenge in computational materials science. We introduce CrystalGRW, a diffusion-based generative model on Riemannian manifolds that proposes novel crystal configurations and can predict stable phases validated by density functional theory. The crystal properties, such as fractional coordinates, atomic types, and lattice matrices, are represented on suitable Riemannian manifolds, ensuring that new predictions generated through the diffusion process preserve the periodicity of crystal structures. We incorporate an equivariant graph neural network to also account for rotational and translational symmetries during the generation process. CrystalGRW demonstrates the ability to generate realistic crystal structures that are close to their ground states with accuracy comparable to existing models, while also enabling conditional control, such as specifying a desired crystallographic point group. These features help accelerate materials discovery and inverse design by offering stable, symmetry-consistent crystal candidates for experimental validation.
title CrystalGRW: Generative Modeling of Crystal Structures with Targeted Properties via Geodesic Random Walks
topic Materials Science
Statistical Mechanics
Machine Learning
Computational Physics
68T07
url https://arxiv.org/abs/2501.08998