Metal-organic frameworks (MOFs) are porous hybrid organic-inorganic materials, which are usually formed by connecting metal nodes and organic ligands to create an extended periodic framework. Their modular building block formation gives rise to a large chemical space of possible combinations and topologies with a wide variety of properties. Recent approaches have demonstrated the capability of redox-active MOFs for energy storage devices such as batteries and supercapacitors. Their intrinsically low electronic conductivity, however, remains a challenge to overcome to improve their viability for device implementation.
This thesis investigates the simulation of redox-active MOFs, focusing on the prediction of conductive frameworks and calculation of relevant parameters that can be measured in experiment. It introduces the theoretical framework for the simulation of these materials: Density Functional Theory (DFT) as the central part of the performed calculations, the calculations necessary for the prediction of band and hopping transport parameters, as well as a short outline of machine learning basics. Firstly, it demonstrates the possibility of predicting the effects of building block substitution on the conductivity of MOFs. Then, it explores how band transport can reproduce experimental conductivity value trends, as long as structural features are accurately taken into account, thus emphasizing the importance of solvent counterions in the framework pore. The ensuing results further explore methods for the calculation of hopping transfer rates that take the full periodic structure into account. The last chapter investigates the prediction of MOF electrode voltages based on pristine structure calculations, employing machine learning to predict the most likely ion adsorption site in a framework and the corresponding energy of intercalation. Taking these results into account, a framework for the exploration of MOF chemical space for energy storage purposes is suggested.