Functional catalysts by design for renewable fuels and chemicals production

Abstract

In the course of mitigating our dependence on fossil energy, it has become an urgent issue to develop unconventional and innovative technologies based on renewable energy utilization for fuels and chemicals production. Due to the lack of fundamental understanding of catalytic behaviors of the novel chemical compounds involved, the task to design and engineer effective catalytic systems is extremely challenging and time-consuming. One central challenge is that an intricate balance among catalytic reactivity, selectivity, durability, and affordability must be achieved pertinent to any successful design. In this dissertation, density functional theory (DFT), coupled with modeling techniques derived from DFT, is employed to gain insights into molecular interactions between elusive intermediates and targeted functional catalytic materials for novel electrochemical and heterogeneous catalytic processes. Two case studies, i.e., electroreduction of furfural and step-catalysis for cyclic ammonia production, will be discussed to demonstrate the capability and utility of DFT-based theoretical modeling toolkits and strategies. Transition metal cathodes such as silver, lead, and nickel were evaluated for furfuryl alcohol and 2-methylfuran production through detailed DFT modeling. Investigation of the molecular mechanisms revealed that two intermediates, mh6 and mh7 from mono-hydrogenation of furfural, are the key intermediates that will determine the product formation activities and selectivities. Nickel breaks the trends from other metals as DFT calculations suggested the 2-methylfuran formation pathway is most likely different from other cathodes. In this work, the Brønsted–Evans–Polanyi relationship, derived from DFT energy barrier calculations, has been found to be particularly reliable and computationally efficient for C-O bond activation trend predictions. To obtain the solvation effect on the adsorptions of biomass-derived compounds (e.g., furfural and glycerol), influence of explicit solvent was probed using periodic DFT calculations. The adsorptions of glycerol and its dehydrogenation intermediates at the water-platinum surface were understood via various water–adsorbate, water–water, and water–metal interactions. Interestingly, the bond-order-based scaling relationship established in solvent-free environment is found to remain valid based on our explicit solvent models. In the second case study, step-catalysis that relies on manganese’s ability to dissociate molecular nitrogen and as a nitrogen carrier emerges as an alternative route for ammonia production to the conventional Haber-Bosch process. In this collaborative project, DFT was used as the primary tool to produce the mechanistic understanding of NH3 formation via hydrogen reduction on various manganese nitride systems (e.g., Mn4N and Mn2N). Both nickel and iron dopants have the potential to facilitate NH3 formation. A broader consideration of a wide range of nitride configurations revealed a rather complex pattern. Materials screening strategies, supported by linear scaling relationships, suggested the linear correlations between NHx (x=0, 1, 2) species must be broken in the development of optimal step catalysis materials. These fundamental findings are expected to significantly guide and accelerate the experimental material design. Overall, molecular modeling based on DFT has clearly demonstrated its remarkable value beyond just a validation tool. More importantly, its unique predictive power should be prized as an avenue for scientific advance through the fundamental knowledge in novel catalysts design.