High-throughput computational screening of MOFs for carbon dioxide capture and hydrogen purification

Metal organic frameworks (MOFs) constitute a novel class of porous materials that are formed by the combination of inorganic nodes and organic linkers. The number of newly synthesized MOFs increases exponentially each year, and it is not feasible to experimentally test the gas separation performances of each MOF adsorbent and membrane. Therefore, a computational approach that can identify the promising MOFs out of thousands of materials is essential. In this thesis, by performing high-throughput computational screening, CO2/H2 separation performances of MOF adsorbents and membranes were investigated in detail, and the top promising MOFs were identified. In the first part of this thesis, molecular simulations were performed to identify adsorption- and membrane-based CO2/H2 separation performances of 3857 unique MOFs. Results showed that all 3857 MOFs were CO2 selective when considered as adsorbents, whereas 899 MOFs overcame the performances of polymeric membranes as H2 selective membranes. H2/CO2 selectivities and H2 permeabilities of MOF membranes varied between 2.1×10−5-6.3 and 2.30-1.7×106 Barrer, respectively. Structure-performance relationships revealed that MOFs with pore size <7.5 Å performed well as CO2 selective adsorbents whereas MOFs with pore size >15 Å were more suitable to be used as H2 selective membranes. In the second part of this thesis, Grand Canonical Monte Carlo (GCMC) and Equilibrium Molecular Dynamics (EMD) simulations were performed on the updated MOF database to identify adsorption and membrane-based CO2/H2 separation performances of 10221 unique MOFs. The applicability of Ideal Adsorbed Solution Theory to MOFs for CO2/H2 separation at temperature and pressure swing adsorption conditions, and the effects of inaccessible local pores, catenation in the frameworks, and presence of impurities (CO, CH4 and H2O) in gas mixture on the selectivity, working capacity, and regenerability of MOFs were examined. MOFs that are recently synthesized and added to the updated MOF database were shown to have higher CO2/H2 selectivities and working capacities than the previously reported MOFs. In the third part of this thesis, new MOFs were designed by using in-silico metal exchange techniques and CO2/H2 and CO2/CH4 separation performances of these MOFs were studied. Results showed that the type of the metal site affects the CO2/H2 selectivities of MOFs. As a result, CO2/H2 selectivity of a commonly studied MOF in the literature, HKUST-1, was significantly enhanced by 11% and 38% when Cu metal was exchanged with Cr and Cd metals, respectively. The exchange of Zn with V increased the selectivity of HIFTOG02 from 119 to 355. Overall, results showed that high-throughput computational screening techniques introduced in this thesis can be used to (i) shortlist potentially promising MOFs among thousands of MOFs for pre-combustion CO2 capture, (ii) identify the structural properties of the promising MOFs with high CO2/H2 separation performances, (iii) design new MOFs with exceptional CO2 capture and H2 purification properties. The results presented in this thesis will serve as a catalyst for future computational and experimental studies on MOFs which will efficiently capture and sequester CO2.