DiffSyn: A Generative Diffusion Approach to Materials Synthesis Planning

TL;DR

DiffSyn is a novel generative diffusion model that predicts feasible synthesis routes for crystalline materials like zeolites, leveraging extensive literature data to improve synthesis planning and successfully synthesize a new high Si/Al UFI material.

Contribution

The paper introduces DiffSyn, a diffusion-based generative model for materials synthesis planning, capturing complex structure-synthesis relationships from large literature data.

Findings

DiffSyn achieves state-of-the-art synthesis route prediction.

Successfully synthesizes a high Si/Al UFI material.

Generates synthesis routes validated by DFT calculations.

Abstract

The synthesis of crystalline materials, such as zeolites, remains a significant challenge due to a high-dimensional synthesis space, intricate structure-synthesis relationships and time-consuming experiments. Considering the one-to-many relationship between structure and synthesis, we propose DiffSyn, a generative diffusion model trained on over 23,000 synthesis recipes spanning 50 years of literature. DiffSyn generates probable synthesis routes conditioned on a desired zeolite structure and an organic template. DiffSyn achieves state-of-the-art performance by capturing the multi-modal nature of structure-synthesis relationships. We apply DiffSyn to differentiate among competing phases and generate optimal synthesis routes. As a proof of concept, we synthesize a UFI material using DiffSyn-generated synthesis routes. These routes, rationalized by density functional theory binding energies, resulted in the successful synthesis of a UFI material with a high Si/Al$_{\text{ICP}}$ of 19.0, which is expected to improve thermal stability and is higher than that of any previously recorded.