Exploring covalent organic frameworks for H2S+CO2 separation from natural gas using efficient computational approaches

Abstract

Covalent organic frameworks (COFs) are emerged as strong adsorbent candidates for industrial gas separation applications due to their highly porous structures. In this work, we explored H2S+CO2 capture potentials of synthesized and computer-generated COFs from a natural gas mixture using an efficient, multi-level computational screening approach. We computed the adsorption data of a six-component natural gas mixture, CH4/C2H6/CO2/C3H8/H2S/H2O, for 580 synthesized COFs by performing Grand Canonical Monte Carlo (GCMC) simulations under industrially relevant conditions. H2S+CO2 selectivities and working capacities of COFs were computed to be 0.4-12.4 (0.2-8.5) and 0.01-5.36 (0.04-2.5) mol/kg at pressure-swing adsorption (vacuum-swing adsorption) condition. NPN-3 was identified as the best performing COF due to the competitive adsorption of H2S+CO2 over C2H6 and C3H8 as revealed by density functional theory (DFT) calculations. Structural (pore sizes, porosities, and topologies) and chemical properties (linker units and heats of gas adsorption) of the best-performing synthesized COFs were used to efficiently screen the very large number of hypothetical COFs (hypoCOFs). Results showed that isosteric heats of adsorption can be used to discover high performing hypoCOFs for H2S+CO2 separation from natural gas. Finally, we compared COFs, hypoCOFs, zeolites, carbon nanotubes, metal organic frameworks (MOFs) and concluded that several synthesized and computer-generated COFs can outperform traditional adsorbents in terms of H2S+CO2 selectivities. Our results provide molecular-level insights about the potential of COFs for natural gas purification and direct the design and development of new COF materials with high H2S+CO2 selectivities.