Degradation Mechanisms and Accelerated Stress-Test Protocols for Polymer Electrolyte Water Electrolyzers

Abstract

If polymer electrolyte water electrolysis (PEWE) is to be deployed to serve as an energy storage technology in the grid, the potential challenges related to the component durability need to be understood and addressed. The US Department of Energy (DOE) targets a cost of 2 $/kg H2 for the year 2020. Maintaining stack efficiency is key to a competitive end-cost of H2, as the main contribution is projected to come from the cost of electricity (1). The development of tailored PEWE components therefore needs to go hand-in-hand with the development of accelerated stress-tests that trigger specific degradation mechanisms. In this study, different degradation mechanisms were induced by subjecting commercial catalyst-coated membranes (CCMs) to specific operating conditions. PEWE cells were operated for 300 h to estimate the effects of the imposed stressors, and the cell performance was compared to the 300 h non-stressed benchmarks. The cell response was analyzed using electrochemical impedance spectroscopy (EIS). Individual catalyst layer (CL) characteristics were probed in the H2/N2 regime, and the cyclic voltamograms were recorded. Finally, samples were post-test investigated using scanning electron microscopy (SEM). The effect of the cell compression on the CL performance was stessed using cell assemblies with Ti-sintered porous transport layers (PTLs) of different morphological surface properties. Rigid PTL-surface features have induced local variation in CL porosity, and increased CL proton transport resistance (RCLa H+). In the case of PTLs with large voids, CCM creep led to the formation of thin spots and an increase in measured H2 in the O2 outlet. We have found that the increased gas crossover with mechanically stressed and locally thinned CCM results in a higher fluoride release rate over 300 h. The effects of start-stop operation on the CL stability were quantified over 300 h by periodically analyzing the anodic water using ICP-mass spectroscopy for traces of precious metals. Additionally, elemental mapping was done on the post-test CCM cross-sections to deterime if potentially dissolved species had deposited in the PEM (2). As water impurities are an important factor determining the electrolyzer lifetime (3), we have designed Ir2+ doping experiments to simulate the accelerated reverisble contamination of the CCM. A detalied overpotential analysis was done to determine the effect on the ohmic, activation and mass-transport overpotential. A specially designed cell was used to conduct operando neutron imaging of the contaminated CCM, and track the distribution of metal cations in the CCM cross-section under various operating conditions. EIS was done in paralell to correlate the impact of contaminants on the cell performance with their cross-sectional position. We established an comprehensive degradation matrix based on the PEWE degradation-inducing experiments, highlighting the stressors, degradation mechanisms and the affected aspects of PEWE performance.