INVESTIGATING DENSITY FUNCTIONAL THEORY'S EFFECTIVENESS IN STUDYING METAL-ORGANIC FRAMEWORKS STRUCTURES

In combatting human induced climate change, carbon capture provides the potential to more slowly ease away from the dependence on hydrocarbon fuel sources, while mitigating the amount of CO2 released into the atmosphere. One promising material to use is metal-organic frameworks (MOF's). MOF's offer an immense variety in potential exceptionally porous structures, a property important in separation. As a result of practical experimental measurements being expensive and time consuming, interest in accomplishing the same goal through modeling has also increased. Using density functional theory to optimize the approximate experimentally measured atomic geometries has been shown to have sufficient accuracy. A previous study by Nazarian et al. was performed to optimize structures on the CoRE MOF Database using a supercomputer. The purpose of this study was to attempt to replicate their work done with a single MOF using computational resources more commonly available. Furthermore, as time tends to be the limiting factor in conducting these studies, the use of a smearing function was adjusted for two optimizations to see if any considerable improvement on the efficiency of the optimizations could be made. Our results show both optimizations improved the bond length accuracy relative to the raw data compared with the optimization from Nazarian, et al. The optimization with a more present smearing effect was able to converge the electron field in roughly half the time, while still showing nearly the same results, except for slightly more variability in the bond lengths involving transition metals. Unfortunately, the improvement in bond length, did not correspond in consistent improvement of the larger cell defining metrics. This shows that either a different energy minimum was found or the relationship between the larger cell parameters, with the more local parameters such as bond length is too complex for the method to effectively solve.