1 Center for Atomic-scale Materials Design, Center, Technical University of Denmark2 Department of Energy Conversion and Storage, Technical University of Denmark3 Atomic scale modelling and materials, Department of Energy Conversion and Storage, Technical University of Denmark4 Department of Physics, Technical University of Denmark5 SLAC National Accelerator Laboratory
This thesis is dedicated to the investigation and design of new catalyst materials for electrochemical ammonia production and especially the properties of the under-coordinated reaction sites on nanoparticles has been studied in great detail. Additionally, a universal transition state relation governing a large number of (de)hydrogenation reactions has been developed. The approach throughout the work is to use Density Functional Theory (DFT) to understand chemical reactions occurring at surfaces. I present an analysis of the transition state energies for 249 hydrogenation/ dehydrogenation reactions of atoms and simple molecules over close-packed and stepped surfaces as well as nanoparticles of transition metals. Linear energy scaling relations are observed for the transition state structures leading to transition state scaling relations for all the investigated reactions. With a suitable choice of reference systems the transition state scaling relations form a universality class that can be approximated with one single linear relation describing the entire range of reactions over all types of surfaces and nanoclusters. Theoretical studies of producing ammonia electrochemically at ambient temperature and pressure without direct N2 dissociation are presented. The computational hydrogen electrode was used to calculate the free energy profile for the reduction of N2 admolecules and N adatoms with an applied potential on transition metal nanoclusters in contact with an acidic electrolyte. In a study, the extreme under-coordinated M12 nanocluster was used as a model system for very under-coordinated reaction sites. The work resulted in establishing linear scaling relations for reaction intermediates for the dissociative and the associative reaction mechanism, and for the key adsorbates hydrogen and nitrogen. These scaling relations and free energy corrections are used to establish volcanoes describing the onset potential for electrochemical ammonia production and hence describe the potential determining steps for the electrochemical ammonia production. The competing hydrogen evolution reaction has also been analyzed for comparison. The most promising candidates from this study were found to be Mo and Fe which would require potentials of -0.5 V to electochemically form ammonia. These metals still have a tight competition with hydrogen evolution reaction. On the basis of the M12 study, a molybdenum nanocluster of 13 atoms in the cuboctahedral structure was analyzed. Pathways for electrochemical am-monia production via direct protonation of N adatoms and N2 admolecules was analyzed and the results was found to match the predictions from the M12 nanocluster study. Calculations presented here, show that N2 dissociation at either nitrogen vacancies or a clean molybdenum particle, is unlikely to occur. The calculations suggest that nitrogen will be favored at the surface compared to hydrogen even at potentials used to produce ammonia and the Faradaic losses due to hydrogen evolution reaction should be low. For electro-catalysts the presence of water is very difficult to avert. Water will give rise to oxygen adsorption on most surfaces and the oxygen atoms will occupy important surfaces sites, which results in a decrease or a total hindrance of other chemical reactions taking place at that site. We therefore present theoretical investigations of the influence of oxygen adsorption and reduction on a molybdenum nanocluster with the purpose of understanding the issues with oxygen poisoning the catalyst. The calculations show that the molybdenum nanocluster will preferentially bind oxygen over nitrogen and hydrogen. The potentials required to reduce oxygen off the surface are -0.72 V or lower for all oxygen coverages studied. Exposure of the molybdenum nanoclusters to air or water, should not be an issue, since it is possible to reduce oxygen blocking the active sites off the surface. At lower oxygen coverage, nitrogen molecules can adsorb to the surface and electrochemical ammonia production at potentials between -0.45 V and -0.7 V are observed. My conclusion is that the molybdenum nanocluster should be a good electrocatalyst for electrochemical ammonia production at ambient conditons. This should open for another pathway for the production of ammonia.