AI accelerates MOF synthesis

Researchers have trained ChatGPT to create a chemistry lab assistant that summarizes with high accuracy information about synthesis from papers ( J. Am. Chem. Soc. 2023, DOI: 10.1021/jacs.3c05819 ). The program extracts over 26,000 parameters from peer-reviewed articles and supporting information about metal-organic frameworks (MOFs) . Once trained, it is able to answer questions about the preparation of MOFs quickly and accurately. The ChatGPT models mined the supporting information of hundreds of MOF papers in which data on synthesis are unstructured and sparse, often extended over hundreds of pages. The researchers developed a filtering strategy that increases the efficiency of that process, says Omar Yaghi of the University of California, Berkeley, the lead author of the study. Large language models like ChatGPT can be prone to responses that seem correct but aren’t. The team minimized these misleading affirmations with careful prompt engineering. “It’s a means of training ChatGPT,” Yaghi says.