Towards Light Driven Water Splitting

Abstract

Energy storage technologies to overcome the intermittent nature of renewable power sources are essential for economic and environmental security. The conversion of energy from solar to fuel follows the blueprint of evolution, which has demonstrated sustainability over a three billion year timescale and inspired researchers for over one hundred years. There are major challenges to efficient solar fuel synthesis leading to it being described as a Holy Grail of science. The synthesis of a fuel occurs via the endothermic reduction of a suitable reactant (e.g., H+, CO2, N2), thus an an electron source is required with H2O the ideal molecule to be oxidised due to its abundance and innocuous nature. However, the oxidation of water (the oxygen evolution reaction, OER) is kinetically complex and active catalysts are required for it to proceed in an energy efficient manner. An objective of global scale artificial photosynthesis also necessitates that the catalysts and all other components are composed of Earth abundant elements. First row transition metal oxides (i.e. Mn, Co, Ni) have demonstrated high catalytic activity towards the OER and have been studied in detail as electrocatalysts. The focus herein is on metal oxides prepared through oxidative electrodeposition, specifically non-stoichiometric Co-oxide (CoOx). CoOx electrosynthesised from neutral and near-neutral, pH buffered electrolyte solutions has been demonstrated as an active water oxidation catalyst and received tremendous attention within the literature, including integration into devices. Herein, ultra-thin and transparent CoOx films are described, with cobalt complexes used as precursors to allow conformal catalyst coatings. This facilitates integration of CoOx films onto photoanodes via photo-electrodeposition and addresses previously identified limitations from opaque CoOx layers. Systematically varying the stability of the precursor allows control of the deposition rate at po- tentials more positive than the OER. The precursor can thus be tailored to the photopotential of an n-type light harvester. Furthermore, the films with lower Co loadings demonstrated higher activity per metal amount at high positive poten-tials, which is due to proton transport becoming the limiting step in thick catalyst films. The mechanism of electrocatalysis and quantification of parameters was then herein examined using large amplitude Fourier transformed alternating current (ac) voltammetry. In contrast to typical direct current (dc) voltammetry, ac voltam- metry filtered out water oxidation dc current and allowed the resolution of redox transformations of CoOx, MnOx and NiOx coupled to catalytic water oxidation. A general water oxidation model was then developed for these surface confined redox transformations coupled to a catalytic reaction with a substrate in solution. The model emphasises the role of the Bronsted base in proton abstraction within a proton coupled electron transfer mechanism. Comprehensive comparisons be- tween experimental and theoretical data revealed comparable effective reversible potentials of 1.9-2.1 V vs the reversible hydrogen electrode for each type of catalyst, i.e. for the formation of the species that undergoes catalytic turnover accompanied by the evolution of O2. The pseudo-first order forward rate constants for these species were also determined to be between 2 ·103 to5 ·104 s–1 for all three metal oxides, which is higher than any previously reported OER catalyst. Retaining a focus on CoOx, the relationship between structural disorder and catalytic activity was examined with samples synthesised in systematic steps through a bulk chemical oxidation. Electrochemical testing revealed the more disordered cobalt oxides were less active for OER catalysis but stronger oxidants, viz. they were more readily involved in non-catalytic chemical reactions. This supports more disordered CoOx being less thermodynamically stable, but challenges previously proposed correlations about disorder in metal oxide OER catalysts being beneficial to catalytic activity. Finally, effective pairing of Ni electrodes to a III/V type photovoltaic (PV) allowed overall water splitting at 22% solar to fuel power conversion efficiency (SFE) using commercially available components, while avoiding the use of Pt group metals and eclipsing the 12% SFE demonstrated in the previous year. The approach was defined with PV-electrolyser pairing parameters that demonstrate the point of power loss from photon to current through to SFE and identify the directions for improvement. Additionally, the parameters allow the maximum SFE of a system to be determined from characterisation of the PV.