Data-Driven Discovery of New Zeolite Imidazolate Frameworks

Abstract

The search for new, advanced materials is becoming an important research theme in energy science. Synthetic strategy based on self-assembly has been highly successful but often take years to develop new materials. Sorting synthetically feasible materials among numerous hypothetical ones requires scientific intuition and careful analysis. Zeolitic imidazolate frameworks (ZIFs)1 are chosen for applying our new data-driven-discovery strategy. Interestingly, the number of topologies in ZIFs (~40) are much smaller than that (~235) found in zeolites.2 In this presentation, we suggest synthetically accessible (but notyet-synthesized) ~200 ZIFs based on hypothetical zeolite database using three chemical descriptors. 3 Together with possible topologies, we also predict CH4, CO2, and Xe/Kr adsorption properties from computational screening. Some ZIF candidates have exceptional adsorption properties.