Investigating the Effects of Anode Catalyst Conductivity and Loading on Performance for Anion Exchange Membrane Water Electrolysis

Abstract

Water electrolysis (WE) is a promising route for carbon-free production of green hydrogen, which is increasingly considered to be an important energy carrier. Among the low-temperature WE technologies, anion exchange membrane water electrolysis (AEMWE) has the potential advantages of high operating current densities and low cost due to the use of zero-gap assembly and platinum group metal (PGM)-free components, respectively [1]. However, significant challenges remain for this technology, including low efficiency, insufficient durability, and poor performance in pure water operation. Recent efforts have yielded improvements in the individual components, particularly the oxygen evolution reaction (OER) catalysts and anion exchange polymers; however, to further improve the performance of the AEMWE, it is also necessary to understand and improve the integration of these materials into the membrane electrode assembly (MEA). Within the MEA, the anode and cathode catalyst layers (CLs) incorporate catalyst and ion-conducting and/or binding polymers, which are deposited onto either the membrane or the porous transport layer. The material properties of each component, as well as the loadings of the catalyst and ionomer and methods of dispersion and deposition, have significant impact on the catalyst layer morphology and uniformity [2]. These CL properties affect the transport of electrons, hydroxide ions, and evolved gases, ultimately controlling the accessibility and utilization of catalyst active sites, with significant consequences for both the performance and durability of AEMWEs.Most of the effort in catalyst development for AEMWE has focused on the OER, aiming to improve sluggish kinetics and improve stability at the highly oxidizing operating potentials. A number of PGM-free catalysts have shown promising activity in rotating disk electrode (RDE) tests, particularly oxides with varying ratios of Ni, Fe, and Co. The activity improvements with tuning the composition are often attributed to the increase in surface area and shift in redox potentials for the in-situ conversion to higher valence sites with higher intrinsic OER activity. However, these activity trends do not always translate into device-level performance because of differences in the operating environment, e.g. hydroxide concentration, reaction rates, etc. In the AEMWE, where typically an order of magnitude higher loadings and current densities are required, the challenges of conductivity and mass transport to allow for full site access are exacerbated. Electrochemical impedance spectroscopy (EIS) is a valuable tool for understanding the origin of voltage losses in the AEMWE, typically broken down into ohmic, kinetic, and mass transport losses. Recently, a transmission line model has been applied to non-Faradaic EIS to isolate another contribution to voltage loss: the catalyst layer resistance (CLR), which corresponds to in-plane and through-plane electronic and ionic resistances in the catalyst layer [3]. Using this approach, the roles of intrinsic catalyst activity and catalyst layer properties can be decoupled, helping to guide materials integration efforts.In this study, we investigate two types of OER catalyst materials, NiFe- and Co-based oxides. Within each composition, the catalysts have similar active sites but vary in electronic conductivity, crystal structure, and surface area. By comparing the AEMWE performance at different catalyst loadings, we are able to determine the role of catalyst properties in determining active site utilization and electronic resistance within the catalyst layer. Figure 1 shows the trends in performance and ohmic and catalyst layer resistance as a function of loading for high-conductivity (Co@Co3O4) and low-conductivity (Co3O4) catalysts. Despite nominally having the same surface chemistry and similar particle morphology, they display both different performance and different trends with loading. Using in-situ EIS diagnostics and ex-situ characterization of the catalyst layer morphology, the relationship between catalyst utilization, CLR, and the physical properties of the catalyst and catalyst layer is elucidated. This work provides insight into the catalyst properties that most strongly affect AEMWE performance, as well as how materials integration strategies must be tailored based on these properties.