Electrochemical and Mechanistic Characterization of Thiolate Ligand Modifications of [2fe-2s]-Based Electrocatalysts for Hydrogen Evolution

Abstract

Green hydrogen generation is currently a leading topic of research in the field of renewables and sustainable energy. The development of synthetic hydrogenase mimetics as catalysts for hydrogen evolution has emerged as a promising approach due to their relatively low cost and sustainability. Mimetics of the [2Fe-2S] active site, which are originally based on the active site of the [FeFe]-hydrogenase enzyme, have been shown to efficiently reduce protons to molecular hydrogen and help highlight interesting synthetic and spectroscopic insights into the improvement of catalytic capabilities of this class of complexes. These synthetic mimetics are typically defined by the composition and connectivity of the thiolato ligands to both each other and the [2Fe-2S] catalytic core. This dissertation focuses primarily on the effects that differences in the architecture of the thiolate attachments have on catalytic performance and mechanism for both molecular and metallopolymer type [2Fe-2S] catalysts.The complexes discussed in this dissertation underwent new electrochemical and spectroscopic experimentation that defined differences in their catalytic behavior. New spectroelectrochemical equipment and methodology were developed to make the first in situ spectroelectrochemical observations of the initially reduced and protonated catalytic intermediates for aryl dithiolate [2Fe-2S] catalysts in the presence of weak acids. These structural assignments were supported by improved DFT computations and led to the first refinement of the mechanistic paradigm for aryl dithiolate [2Fe-2S] catalysts in the presence of weak acids. Comparisons of new and existing aryl dithiolate [2Fe-2S] active sites in both the organic molecular and aqueous metallopolymer phases revealed the influence of metallopolymer system factors on active site catalytic behavior. These studies also highlighted limitations in the predictive power of [2Fe-2S] molecular system studies to predict catalytic properties of [2Fe-2S] metallopolymers. Active site 21 centered differences in electron and proton transfer rates were shown to be dwarfed by metallopolymer characteristics such as polymer length and identity. The synthesis and characterization of the first ATRP generated bis-monothiolate [2Fe-2S] metallopolymers that have two unlinked thiolate ligands is presented. These new metallopolymers exhibited severely reduced catalytic current outputs, but also provided the first example in this family of metallopolymer electrocatalysts of a slightly lowered overpotential for electrocatalytic hydrogen production in neutral aqueous conditions. Additional DFT computations show that the reduced catalytic activity of these systems corresponds with the new ability of the polymer arms to engage in intramolecular hydrogen bonding interactions with each other. These interactions are possible due to the increased degree of conformational freedom afforded by the lack of a dithiolate linker. This study highlights the critical role of a metallopolymer’s conformational form in determining how the active site orients itself relative to the electrode surface which in turn significantly aids or limits the ability of the catalyst to be reduced and undergo catalytic production of H2. Finally, the methodologies for the characterization and prediction of infrared (IR) spectra using DFT computations in previous studies were applied to the problem of prediction of IR transparency properties in materials. A quantitative measure of the sample thickness dependent transparency of a material in a given frequency range of the IR region was defined and referred to as the window transparency (wT). A methodology for the computational and experimental characterization of the long wave IR wT of molecules was developed and applied to a series of straight chain alkanes. A comparison of DFT predicted wT values with those determined experimentally for the series of n-alkanes showed an average difference of ~3% which is promising in its initial accuracy without the consideration of other complicating factors.