Advanced Synthesis and Scale-up of Earth-Abundant Catalysts for Electrochemical Water Splitting

Abstract

The global energy crisis and pollution from fossil fuel have intensified the demand for sustainable solutions, with green hydrogen production by electrochemical water splitting emerging as a promising alternative. However, current electrolyser technologies face significant challenges particularly due to their reliance on expensive noble metal catalysts. To address these issues, multi-metallic, earth-abundant catalysts such as NiFe- and NiMo-based have emerged as viable alternatives. Despite their potential, these catalysts face barriers to widespread adoption due to complex fabrication and optimization processes. Moreover, scaling them from laboratory setting to industrial applications introduces additional complications, as the performance achieved in lab conditions often fails to translate directly to industrial electrolysers. This thesis aims to investigate scalable methods for synthesising efficient earth-abundant catalysts for water-splitting devices and to bridge the gap between lab-scale innovations and their industrial-scale applications. First, a one-step fabrication process for complex electrodes is investigated, focusing on optimising trimetallic NiFeMoN catalysts for the oxygen evolution reaction (OER). Using magnetron cosputtering, synthesis process is refined through statistical optimisation techniques, design of experiments (DOE) and response surface methodology (RSM). Analysis of variance (ANOVA) is employed to understand the interaction effects of synthesis parameters which are critical for achieving optimal performance. The optimised NiFeMoN exhibits high performance with a low overpotential of 216 mV at 10 mA/cm2 with stability over seven days. Furthermore, when integrated into a decoupled photoelectrode design, the sputtered NiFeMoN not only enhances OER performance but also effectively protects Si photoanodes from photo-corrosion. To develop an efficient transition metal-based multi-metallic carbide for the hydrogen evolution reaction (HER), thesis explores the utilisation of triple-target magnetron co-sputtering process. NiMoC is synthesised in a single step and at room temperature, exhibiting remarkable performance across a broad pH range. It demonstrates a superior overpotential of 26 mV with ten days' stability and 42 mV with 70 hours of stability in alkaline and acidic conditions, respectively. These enhancements are attributed to Mo2C and NiMo as active sites, with disordered and graphite-like carbon improving electrical conductivity and stability. When used as a cathode in an electrolyser, it achieves a cell voltage of 1.78 V at 0.5 A/cm2 and 1.87 V at 1 A/cm2, maintaining stability for nearly 70 hours. To facilitate the transition from lab-scale to industrial-scale catalysts, thesis explores the individual and interaction effects of operating parameters, electrolyte flow rate, KOH concentration, and temperature, along with a synthesis parameter, catalyst deposition time, in a zero-gap electrolyser. An artificial neural network combined with particle swarm optimisation (ANN-PSO) and the Box-Behnken design (BBD) are applied to model the system and uncover the interaction effects of parameters. The findings reveal an extended deposition time shifts the optimal KOH concentration and flow rate by 40.8% and 19%, respectively. Increasing flow rate reduces the optimal deposition time by 23%, while increasing the optimal KOH concentration by 30%. Furthermore, a genetic algorithm is employed to optimise zero-gap electrolyser, resulting in a 31.4% improvement in performance. Analysis using Garson's method shows that KOH concentration has the most substantial impact, while flow rate has the least. Deposition time exhibits an influence comparable to that of temperature. This work provides a systematic approach to catalyst synthesis and scaling, advancing the field of water splitting and paving the way for industrial deployment of cost-effective and efficient green hydrogen production technologies. Show more