Computational Discovery of Metal-Organic Frameworks for Separations of Organic Molecules
Gee, Jason Alan
MetadataShow full item record
The separation of para-xylene from a stream of mixed xylenes and ethylbenzene is critical for the large-scale production of plastics in the petrochemical industry. Several groups have identified metal-organic frameworks (MOFs) as having desirable characteristics for this separation. In this thesis, we demonstrate that molecular simulations can be used to efficiently screen large databases of MOFs to identify promising materials for this separation. We validated our approach in conjunction with our experimental collaborators and discovered that two of the top-performing materials from our screening procedure have similar performance to the zeolites used in industrial practice for xylene separations. We also developed a classical force field parameterization approach for refining the interactions between C8 alkyl aromatic hydrocarbons and MOFs using Density Functional Theory (DFT) calculations. We demonstrate that our DFT-based force field gives better predictions of some adsorption properties than generic force fields. A major technological hurdle to using small alcohols as biofuels is in their separation from aqueous fermentation broths. To address this issue, we developed classical models to identify hydrophobic MOFs capable of efficiently performing this separation. We were then able to use our models in a different context to understand the factors governing the thermodynamic stability and structural flexibility of MOFs. The methods developed in this thesis provide unique insight into chemical separations and material properties that would be challenging to obtain from experiments and promote the development of MOFs for industrial applications.