Adsorption behavior of metal-organic frameworks: From single simulation, high-throughput computational screening to machine learning

Recently, the development and application of porous materials have attracted increasing attention, and metalorganic frameworks (MOFs) have become ?stars? in the emerging material field because of their high porosities and ultra-high specific surface areas. In this review, the calculated works of MOFs in this field in the past ten years were summarized, especially for MOF adsorbents. With the continuous growth of the number of adsorbent materials, simulations have gradually transitioned from the single simulation, high-throughput computational screening (HTCS) to machine-learning (ML)-assisted HTCS. The purpose of this paper is to sort out the research progress and current ideas for the adsorption simulations of MOFs. Finally, we highlight the bottlenecks and challenges for the future commercialization of HTCS based on ML and the ML-assisted HTCS methods that are suitable for solving the research problems in this field. We also speculate about the future development directions of this field, hoping to promote the practical application of porous adsorbents.

Automated summary

This review summarizes the calculated works of MOFs in the field of porous materials in recent years, with a focus on the development of MOF adsorbents and the application of machine learning in high-throughput computational screening. The paper highlights the bottlenecks and challenges for the future commercialization of HTCS based on ML, as well as the potential future development directions of this field.