A Generation Framework with Strict Constraints for Crystal Materials Design

TL;DR

This paper introduces a new constrained generation framework for crystal materials that uses large language models to generate intermediate constraints, resulting in more targeted and reliable crystal structure generation with strict adherence to chemical properties.

Contribution

The work presents a novel framework combining LLM-based constraint generation with structure synthesis, significantly improving control and property adherence in crystal design.

Findings

Generation probability of target properties more than doubles.

Nearly 100% of generated crystals meet chemical constraints.

Framework reduces post-processing steps for stable candidate identification.

Abstract

The design of crystal materials plays a critical role in areas such as new energy development, biomedical engineering, and semiconductors. Recent advances in data-driven methods have enabled the generation of diverse crystal structures. However, most existing approaches still rely on random sampling without strict constraints, requiring multiple post-processing steps to identify stable candidates with the desired physical and chemical properties. In this work, we present a new constrained generation framework that takes multiple constraints as input and enables the generation of crystal structures with specific chemical and properties. In this framework, intermediate constraints, such as symmetry information and composition ratio, are generated by a constraint generator based on large language models (LLMs), which considers the target properties. These constraints are then used by a subsequent crystal structure generator to ensure that the structure generation process is under control. Our method generates crystal structures with a probability of meeting the target properties that is more than twice that of existing approaches. Furthermore, nearly 100% of the generated crystals strictly adhere to predefined chemical composition, eliminating the risks of supply chain during production.