Electrochemical Nitrogen Reduction Under (Near) Ambient Conditions

Abstract

The world population is expected to increase significantly over the next 80 years, reaching potentially 11 billion people before leveling off at the end of this century. All these people will need food, which can be grown more effectively by the application of synthetic fertilizer to the soil compared to traditional manure. Synthetic fertilizer, in the form of ammonia, is currently commercially produced via the large-scale catalytic Haber-Bosch process. This utilizes methane and nitrogen as reagents to produce ammonia over an iron-based catalyst, with carbon dioxide as a side product. Unfortunately, this process releases ~1% of all the CO2 emissions in the world, negatively impacting climate change. Furthermore, due to the large energy barriers required to split the inert dinitrogen molecule, high pressures and  temperatures are needed, leading to huge centralized facilities for the production of ammonia. In this thesis, the process of electrochemical nitrogen reduction is considered, where ammonia can potentially be made directly from nitrogen and hydrogen, using electricity as the energy source. This would theoretically lead to a carbon neutral process, wherein the electricity could be supplied via renewable sources, such as wind or solar power. Furthermore, the process should be able to run continuously at (near) ambient conditions, thereby enabling production at the point of use, cutting out the need for transportation and storage.  Unfortunately, this process is far from optimal, and while the field is relatively new, the activity and selectivity of the electrocatalytic process is horrifically dismal. Ammonia is sadly also ubiquitous in the environment in small, but similar (and often greater), amounts to those made from electrocatalysis, which leads to regrettably many false positives in the literature. Scientists are, for the most part, not nearly careful and stringent enough with background testing, therefore inadvertently misleading the field with wrong results. Therefore a central part of this work was in developing a thorough and rigorous testing protocol, which, if followed properly, can ascertain if the source of the measured ammonia is pervasive contamination or successful synthesization. The crucial and indispensable step of this protocol is the use of gas cleaning combined with quantifiable isotope labeled experiments, which is the definitive proof of a successful electrocatalytic process. Two different proposed processes for gas cleaning were tested successfully (a home-made and a commercial system), along with nuclear magnetic resonance measurements for isotope sensitive quantification of ammonia. Using this protocol, many different catalysts were tested, both in aqueous and non-aqueous conditions, based on published research. None of the tested pure metal candidates in aqueous conditions produced any measurable ammonia above the detection limit of the colorimetric indophenol method. Ruthenium, one of the most promising pure metal catalysts according to theoretical calculations, was tested in alkaline, neutral, and acidic conditions, as well as under different temperatures in both alkaline and acidic media, but to no avail. One of the more promising systems in non-aqueous electrolyte was also tested: the lithium-mediated system first proposed by Tsuneto et al. in 1993. This system finally gave a positive result. Rigorous testing validated the reported result by Tsuneto et al., bringing it up to be included as one of the few trustworthy reports from the electrochemical nitrogen reduction field. Different cathode materials were also tested in the system, along with different salts and proton sources leading to the conclusion that the initial system with a molybdenum working electrode in lithium perchlorate in tetrahydrofuran with ethanol as the proton source was the best in terms of Faradaic efficiency. However, there is a caveat to this ammonia synthesizing process. The lithium-mediated system is not stable over longer measurements, as initially observed by Tsuneto et al., and seen from our data as well. A cyclic stabilization method of the working electrode potential was therefore created. This method utilizes a short lithium deposition pulse, followed by a longer resting pulse below lithium deposition potentials, with the cycle repeated indefinitely. This allows the lithium species to chemically dissolve from the surface during the longer resting pulse, preventing a build-up of passivating species on the surface of the electrode. The electrodes were investigated with various ex-situ characterization techniques to determine the surface composition post-measurement. The continuous lithium deposition process originally proposed Tsuneto et al. was compared to the newly developed cycling method. The new cyclic method was seen to keep the electrode surface clear of accumulated species, which could therefore keep the working electrode potential in a stable regime for days, improving both the Faradaic efficiency and the energy efficiency significantly. This suggested method is an incremental step towards commercialization of the lithium-mediated system, as continuous operation at increased energy efficiency is absolutely necessary for future industrial use.