Electrolyte Formulation By Design of Experiments: The Hidden Correlations in the Li-Mediated Nitrogen Reduction Reaction for Ammonia Production

Abstract

Ammonia, a fundamental building-block for fertilizers and other many commodities, is responsible, through the Haber-Bosh (HB) process, of around 1.4% of the global greenhouse gas emissions. Indeed, HB operates in severe conditions (200 atm), and is fossil fuel dependent, since the few huge, centralized plants are combined with steam reforming for H2 production. To find a renewable-driven and delocalized electrochemical process for NH3 production, complementary to HB, could be a key solution for our society that is facing climate change crisis and that is demographically growing [1].In view of process electrification, the Li-mediated (Li-m) pathway represents the most promising solution in the N2 reduction reaction (NRR) challenging field. Exploiting the unique reducing power of this alkali metal, this strategy achieves the highest Faradic efficiency (FE) and NH3 production rate [2,3,4].Li+ ions from the aprotic electrolyte are electrodeposited on the cathode, where N2 is reduced and protonated into NH3 thanks to a proton donor or a proton shuttle, as EtOH. On the plated Li, a solid electrolyte interphase (SEI) unavoidably forms, and its nature and morphology are straightly dependent on the electrolyte composition. The different diffusion rate of Li+, N2, and H+ trough the SEI layer determines the selectivity towards NH3 formation. The correct reactant ratio at the electrode surface is essential to hinder both the hydrogen evolution reaction and an excessive Li plating, maximizing the FE [5,6].Design of experiment (DoE) together with response surface methodology (RSM) are a powerful tool that can reveal, with a restricted number of experiments (e.g. 9), the optimal composition of the electrolyte towards the highest FE and the hidden intercorrelations among variables. In particular, we aimed at studying two key variables, i.e. EtOH and Li-salt concentration; to this purpose, the Doehlert design with two factors and three replicates of the central point was chosen and RSM was employed to model the response and determine the optimal conditions.The study was repeated with two different fluorinated Li-salt, LiBF4 and LiFOB. LiBF4 has recently shown an improved stability and FE for the Li-m NRR process, correlated to the LiF presence in the SEI layer. The optimal combination of the electrolyte components obtained from the DoE for LiBF4 is in accordance with previous study, validating the statistical methodology applied. The newly proposed salt showed an increase of the 25% in the FE with respect to LiBF4 in the same conditions [7,8]. Experiments was conducted under 20 bar N2 in an autoclaved glass cell with a proper gas purification. As in previous study with this Li-m strategy in this set-up, even in this case the total NH3 amount obtained overcame the amount of possible impurities coming from the N2 gas, moreover controls tests with Ar instead of N2 were performed.