Vibrating Powders: Eqcm Measurement of Potential Dependent IrO₂ Mass Changes for Voltage Reversal Tolerant PEMFC Anodes

Oxygen evolution reaction (OER) catalysts are used to protect carbon components of polymer electrolyte membrane fuel cell (PEMFC) anodes during transient fuel starvation conditions such as start-up/shut-down [1]. OER catalysts such as IrO2 or RuO2 promote the oxidation of water instead of the corrosion of carbon components [1, 2]. However, the conditions in a PEMFC during normal operation (ca. 0 V and low pH) can induce instability in these metal oxides. IrO2 in particular can form soluble Ir3+ species [3] that migrate across the membrane electrode assembly and deposit on the cathode causing poisoning of the oxygen reduction reaction (ORR) catalyst (typically Pt/C). Currently, OER catalysts for reversal tolerant anodes (RTAs) are investigated using prototype MEA single unit cell testing which is time consuming and expensive to perform [4]. Previously, our group investigated the dissolution of different IrO2 based powder catalysts using a potential stepping stress test. The dissolution of Ir3+ from this test was compared with reversal tolerance of a membrane electrode assembly fabricated using the same IrO2 catalyst [5]. To further investigate the dissolution of IrO2 powders, we have devised a series of experiments (with potential cycling between 0.05 to 1.2, 1.4 and 1.6 V, at 100 mV s-1 30 times each and 298 K) using the electrochemical quartz crystal microbalance (EQCM) technique. The EQCM uses a vibrating quartz crystal and measures its frequency (inversely proportional to crystal mass) as a function of potential. Our method was validated using a commercially available IrO2 powder (Alfa Aesar) heat treated in air at 350, 450 and 550°C which improves durability while negatively effecting OER activity [6]. The powders were deposited (24 μ cm-2) onto a gold EQCM crystal to investigate the dissolution behavior as a function of particle size, crystallinity and electrode potential. The untreated IrO2 exhibited a large irreversible increase in vibration frequency indicating a loss of mass during the initial polarization from 1.2 to 0.05 V. The increase was on the order of 400 to 700 Hz or -7.1 to -12.4 μg cm-2 assuming a perfectly uniform and rigid layer (figure a). This frequency increase could equally be due to loss of adhesion from the surface, a change in the rigidity of the layer or dissolution of Ir3+. To confirm that frequency changes were not due to adhesion or rigidity of the film, samples of the electrolyte were taken from near the electrode surface and Ir3+ concentration was determined by inductively coupled plasma mass spectrometry (ICP-MS). We observed that the frequency of the untreated IrO2 increased by ca. 90 Hz (-1.6 μ cm-2) during a single polarization from 0.05 V to 1.2 V (figure a). Birss et al. reported the formation of an iridium oxo-hydroxide which causes a mass increase when polarizing to 0 V [6]. Conversely, the frequency of heat treated layers first decreased by ca. 200-500 Hz (1.4 to 3.5 μg) then increased again when polarizing to 1.6 V. Heat treatment induced IrO2 crystallinity changes could account for this difference in frequency response. From the first cycle to the last, there was a frequency gain for untreated IrO2 and a frequency loss for all of the heat treated IrO2 catalysts; suggesting Ir3+ dissolution and increased IrO2 layer rigidity or coverage respectively. The capacitive current of the untreated catalyst decreased slightly over the course of the test and was more than five times higher than the 550°C heat treated catalyst in its cyclic voltammogram (figure c and d). The main challenge to using EQCN for powder catalyst investigation is producing layers with consistent and repeatable frequency behavior. In this work, the IrO2 deposition conditions were tightly controlled and each experiment was repeated to ensure the consistency of the results. If these obstacles can be overcome, EQCN presents a novel, rapid and powerful tool to learn about the potential dependent reactions taking place using a myriad of catalyst materials. Ralph TR, Hudson S, Wilkinson DP (2006) ECS Trans 1:67–84. Patterson TW, Darling RM (2006) Electrochem Solid-State Lett 9:A183–A185. Danilovic N, Subbaraman R, Chang K-C, et al (2014) J Phys Chem Lett 5:2474–2478. Makharia R, Kocha S, Yu P, et al (2006) ECS Trans 1:3–18. Moore CE, Eastcott J, Cimenti M, Kremliakova N, Gyenge EL (2019) J Power Sources in review da Silva GC, Perini N, Ticianelli EA (2017) App Cat B: Env 218 287–297. Birss VI, Elzanowska H, Gottesfeld S (1991) J Electroanal Chem and Interfac Electrochem 318:327–333. Figure 1