Machine learning for the design and discovery of zeolites and porous crystalline materials

Porous crystalline materials, such as zeolites and metal organic frameworks (MOFs), have shown great promises with superior separation, catalysis and upgrading performances in many areas of energy, the environment and health. However, the discovery of new zeolites and MOFs with desired properties is a complex process that often involves trial-and-error experimental/computational approaches. Computational discovery of new materials often involves learning and optimizing more than one objective such as stability, system equilibrium and efficiency. The knowledge gaps between computational prediction and actual synthesis of materials present hurdles to scientific discovery. Advances are underway in machine learning (ML)-in particular, deep reinforcement learning (DRL)-to address these challenges. The goal of this article is to systematically present the key elements and sketch the first steps towards ML-based design and discovery of new zeolites and similar crystalline materials.

Automated summary

Porous crystalline materials like zeolites and metal organic frameworks have shown great potential in energy, environment, and health applications. However, the complex process of discovering new materials with desired properties often involves trial-and-error computational approaches. Machine learning, especially deep reinforcement learning, is making advancements to bridge the knowledge gap between computational prediction and actual synthesis, aiming to revolutionize the design and discovery of new crystalline materials.