(572i) High-Throughput Screening of Hypothetical Functionalized-Irmofs for Separation of Alkanes

Metal organic frameworks (MOFs) have gained wide popularity for their applications, such as gas storage and gas separation, due to their high porosity, large surface area, and tunability of pore size. The development of MOFs with enhanced selectivity and permeability for n-alkanes could result in significant energy and cost savings associated with alkane separation compared to distillation in the petrochemical industry. In this study, we performed high-throughput screening using dual force zone non-equilibrium molecular dynamics (DFZ-NEMD) simulations of hypothetical and known MOFs from existing databases for the separation of linear hydrocarbons, e.g., hexane and nonane. Data generated from this screening was used to develop deep machine learning models. Then these ML models were integrated with our in-house structure generation code that systematically added functional groups on selected MOFs to predict the separation performance of functionalized MOFs. The best candidates were then validated by performing DFZ-NEMD simulations. Our design scheme and the molecular-level insights from the results of this research might help in the design of functionalized MOFs at the molecular level that could be used to separate different hydrocarbons from mixtures.