Accelerating inverse materials design using generative diffusion models with reinforcement learning

TL;DR

This paper introduces MatInvent, a reinforcement learning workflow that enhances diffusion models for goal-directed crystal generation, significantly reducing data needs and improving multi-property optimization in materials design.

Contribution

The paper presents MatInvent, a novel reinforcement learning approach that efficiently optimizes diffusion models for targeted crystal structure generation with minimal labeled data.

Findings

Rapid convergence to target properties within 60 iterations

Successful multi-objective optimization for complex material properties

Up to 378-fold reduction in property computation demands

Abstract

Diffusion models promise to accelerate material design by directly generating novel structures with desired properties, but existing approaches typically require expensive and substantial labeled data ($>$10,000) and lack adaptability. Here we present MatInvent, a general and efficient reinforcement learning workflow that optimizes diffusion models for goal-directed crystal generation. For single-objective designs, MatInvent rapidly converges to target values within 60 iterations ($\sim$ 1,000 property evaluations) across electronic, magnetic, mechanical, thermal, and physicochemical properties. Furthermore, MatInvent achieves robust optimization in design tasks with multiple conflicting properties, successfully proposing low-supply-chain-risk magnets and high-$\kappa$ dielectrics. Compared to state-of-the-art methods, MatInvent exhibits superior generation performance under specified property constraints while dramatically reducing the demand for property computation by up to 378-fold. Compatible with diverse diffusion model architectures and property constraints, MatInvent could offer broad applicability in materials discovery.