Con-CDVAE: A method for the conditional generation of crystal structures

Cai-Yuan Ye,Hong-Ming Weng,Quan-Sheng Wu

2024-03-19

Abstract:In recent years, progress has been made in generating new crystalline materials using generative machine learning models, though gaps remain in efficiently generating crystals based on target properties. This paper proposes the Con-CDVAE model, an extension of the Crystal Diffusion Variational Autoencoder (CDVAE), for conditional crystal generation. We introduce innovative components, design a two-step training method, and develop three unique generation strategies to enhance model performance. The effectiveness of Con-CDVAE is demonstrated through extensive testing under various conditions, including both single and combined property targets. Ablation studies further underscore the critical role of the new components in achieving our model's performance. Additionally, we validate the physical credibility of the generated crystals through Density Functional Theory (DFT) calculations, confirming Con-CDVAE's potential in material science research.

Materials Science

What problem does this paper attempt to address?

The paper focuses on how to use machine learning models to generate crystal structures with specific properties, in order to overcome the limitations of existing methods in efficiently generating crystal structures based on target properties. The study proposes a model called Con-CDVAE, which is an extension of Crystal Diffusion Variational Autoencoder (CDVAE). The paper introduces innovative components, a two-step training approach, and three unique generation strategies to enhance the model's performance. The effectiveness of Con-CDVAE under different conditions, including single and combined attribute targets, is demonstrated through extensive testing and ablation studies. In addition, the physical credibility of the generated crystals is verified through Density Functional Theory (DFT) calculations, confirming the potential of the model in materials science research. The core contribution of the paper lies in the development of a method that can generate crystal structures based on desired property conditions, which helps overcome the limitations of traditional methods in searching complex chemical and structural spaces. Compared to existing crystal generation models such as FTCP representation based on 2D grids, atom-by-atom generation, and CDVAE combined with graph neural network diffusion models, Con-CDVAE demonstrates its advantages in conditional generation. The paper also discusses the two stages of model training and the effects of different generation strategies, and validates the rationality of the generated crystals through DFT calculations. Finally, the authors point out potential future directions for improving the model, including enhancing the recognition of crystal symmetry and reducing structural changes in generated crystals before and after relaxation.

Con-CDVAE: A method for the conditional generation of crystal structures

Crystal Diffusion Variational Autoencoder for Periodic Material Generation

Data-driven discovery of novel 2D materials by deep generative models

Data-driven discovery of 2D materials by deep generative models

Deep learning generative model for crystal structure prediction

An effective method for generating crystal structures based on the variational autoencoder and the diffusion model

Diffusion probabilistic models enhance variational autoencoder for crystal structure generative modeling

VQCrystal: Leveraging Vector Quantization for Discovery of Stable Crystal Structures

Generative Design of Crystal Structures by Point Cloud Representations and Diffusion Model

Generative Inverse Design of Crystal Structures via Diffusion Models with Transformers

Self-Supervised Generative Models for Crystal Structures

Unleashing the power of novel conditional generative approaches for new materials discovery

Data-Driven Score-Based Models for Generating Stable Structures with Adaptive Crystal Cells

3-D Inorganic Crystal Structure Generation and Property Prediction via Representation Learning

Physics guided deep learning for generative design of crystal materials with symmetry constraints

PCVAE: A Physics-informed Neural Network for Determining the Symmetry and Geometry of Crystals

Crystal Structure Generation Based On Material Properties

Scalable Diffusion for Materials Generation

Vector Field Oriented Diffusion Model for Crystal Material Generation

Compositional Search of Stable Crystalline Structures in Multi-Component Alloys Using Generative Diffusion Models