In Search of Lost Iridium: Quantification of Anode Catalyst Layer Dissolution in Proton Exchange Membrane Water Electrolyzers

Abstract

Understanding catalyst dissolution pathways in proton exchange membrane water electrolyzers is paramount for developing mitigation strategies aiming toward higher durability and lower catalyst loadings. To this end, Ir dissolution has been extensively studied using aqueous model systems but not in real devices. Aiming to bridge this knowledge gap, we use a metal-free water electrolysis setup to determine the mass balance of the dissolved Ir in an electrolyzer when applying a protocol mimicking intermittent operation. We find that the main Ir sinks are the cathode catalyst layer and the membrane, while Ir dissolution into the water lines is significantly lower. Although reproducible estimation of Ir present in the membrane is challenging, quantification of Ir within the cathode is reliable and efficient. This new finding implies that tracking Ir present in the cathode can be used to estimate anode catalyst stability, thus accelerating catalyst development and operational parameters optimization.

Porosity as a Morphology Marker to Probe the Degradation of IrO2 Anode Catalyst Layers in Proton Exchange Membrane Water Electrolyzers

“…While the increase in iR -free voltage for the porous-IrO 2 -390 sample during the first 3000 cycles could be correlated with the Ir dissolution in water collected at the anode outlet, no obvious correlation is found between the Ir dissolution trend and the evolution in performance for other MEAs. The very low values of iridium detected in anode outlets, comparable with those recently reported for commercial MEA samples submitted to an accelerated stress test by Milosevic et al do not seem to have an impact of MEA performance in these conditions. H 2 and O 2 crossover were evaluated by measuring H 2 and O 2 content in anode and cathode gas outlets using a micro-GC for the sample prepared with porous IrO 2 -450.…”

“…Whatever the operating conditions, iridium dissolution is one of the agreed mechanisms of iridium oxide nanoparticle degradation, in particular for hydrous oxides and metallic iridium, when submitted to OER potential. , This dissolution has been evidenced by coupling electrochemical measurements with inductively coupled plasma mass spectrometry (ICP-MS) . Indirect dissolution evidence has also been reported in MEAs, and upon dissolution, iridium-oxidized species can be eluted in the anode water outlet or can migrate into the solid electrolyte and further precipitate into the membrane or even into the cathode . Iridium deposition into the membrane or the cathode can be evidenced by scanning or transmission electron microscopy coupled with energy dispersive X-ray spectroscopy (EDS), but this remains only a qualitative analysis.…”

Benchmarking Electrocatalyst Stability for Acidic Oxygen Evolution Reaction: The Crucial Role of Dissolved Ion Concentration

“…EC-MS thus resembles MEA (Figure c, see Alfa Aesar RuO 2 ) as both have a small working volume, and thus a high Ru concentration, resulting in a higher estimated stability number compared to H-cell and in situ ICP flow cell. This observation is supported by the following: the stability number of IrO x measured by MEA (5 × 10) is ∼3 orders of magnitude higher than that measured by the in situ ICP flow cell (6 × 10) and H-cell (1 × 10, electrolyte volume 28 mL); , the stability number of RuO 2 (Alfa Aesar) in MEA is ∼2 orders of magnitude higher than in situ ICP flow cell (replotted in Figure c) and H-cell; and the stability number of RuO 2 in Figure a demonstrates that EC-MS gives a higher stability number than H-cell.…”

Standardizing OER Electrocatalyst Benchmarking in Aqueous Electrolytes: Comprehensive Guidelines for Accelerated Stress Tests and Backing Electrodes

“…Given the potential oxidation of backing electrodes at these values, various LPLs lower than 1.4 V RHE were also tested, specifically 1.23, 1.1, and 0.9 V RHE . Furthermore, 0 V RHE was employed to simulate the effect of H 2 crossover in an MEA. , Each LPL value was associated with a distinct protocol evaluated in both the SFC-ICP-MS and RDE setups. In the SFC-ICP-MS setup, the AST was limited to 200 pulses due to setup limitations.…”