Machine Learning Helps Organize MOF Databases
(EPFL, September 11, 2020)
Due to their extreme versatility and considerable range of potential uses, material scientists have been rapidly developing, synthesizing, studying, and cataloguing metal-organic frameworks (MOFs) – a class of materials that contain record-breaking internal surface areas, thanks to their nano-sized pores. However, although exciting, the sheer number of MOFs – currently over 90,000 have been published – is making it difficult to tell if a newly synthesized MOF is “truly a new structure and not some minor variation of a structure that has already been synthesized,” as explained by EPFL Professor Berend Smit. To address the issue, Smit teamed up with MIT Professor Heather Kulik, and used machine learning to develop a “language” for comparing two materials and quantifying the differences between them, thereby allowing the researchers to explore chemical diversity in MOF databases.