Metal-Organic Frameworks for Water Harvesting: Machine Learning-Based Prediction and Rapid Screening

Top-performing metal-organic frameworksfor atmosphericwater harvesting are predicted by a machine learning method. Atmosphericwater harvesting based on metal-organic frameworks(MOFs) is an emerging technology to potentially mitigate water scarcity.Because of the tremendously large number of existing MOFs, it is challengingto find suitable candidates. In this context, a data-driven approachto identify top-performing MOFs represents an important direction.Herein, we develop a machine learning (ML) method to predict wateradsorption in MOFs and screen out top-performing MOFs for water harvesting.First, experimental water adsorption isotherms in MOFs are collectedand water adsorption properties are extracted. Quantitative structure-propertyrelationships are analyzed in terms of pore structure and frameworkchemistry, providing task-specific design principles. Then, ML modelsare trained and interpreted to predict water adsorption propertiesby using structural and chemical features, as well as operating conditionsas descriptors. The transferability of the ML models is validatedby out-of-sample predictions in seven newly reported MOFs. Finally,the ML models are applied to screen similar to 8000 "Computation-Ready,Experimental" (CoRE) MOFs. Top-performing candidates are identifiedincluding 149 MOFs with the maximum adsorption capacity >= 35mmol/g, 39 MOFs with working capacity >= 10 mmol/g in a relativepressure window 0.1-0.3, and 139 MOFs with working capacity >= 8.7 mmol/g in a relative pressure window 0.6-0.9. Thedeveloped ML-based method would advance task-oriented design and rapiddiscovery of reticular materials for energy and environmental applications.