UC Irvine

The oxygen evolution reaction (OER) is anodic backbone of the majority of aqueous electrochemical processes, e.g., electrolytic H2 production, CO2 reduction, and electrowinning. As devices exclusively employ iridium oxide to catalyze the OER in acidic and near neutral electrolyte, the extreme scarcity of iridium oxide can become a bottleneck for the scale up of aqueous electrochemical processes, most notably proton exchange membrane water electrolysis (PEMWE). Despite the importance of developing alternatives to iridium oxide for the OER in acidic media, such platinum-group-metal-free (PGM-free) alternatives have not been commercially deployed. While there is room for improvement in the catalytic activity of such PGM-free catalysts, the main limitation is operational stability; at low pH (< 2) and oxidizing potentials (> 1.5 V vs. RHE), the majority of transition metals dissolve. This dissertation explores how cobalt-based oxides can be modified to improve their activity and stability in acidic media. The first chapter of explores the bulk modification of cobalt spinels with manganese and copper: transition metals of different polarizability (and thus, in principle, different levels of covalency when bonding with oxygen) that also exhibit similar coordination behavior as cobalt. Therein, we find that the presence of manganese is associated with enhanced stability under dynamic polarization conditions, as well as a correlation between the onset of Co2+/3+ oxidation (from cyclic voltammetry) and overall durability (operational lifetime while cycling). Despite the improvements with Mn incorporation, we nonetheless concluded that such materials were still insufficiently stable to justify further bulk elemental modification studies with cobalt spinels. Thus, in the second chapter, the opposite experimental approach was employed. Instead of incorporating different transition metals into the same cobalt lattice, elemental composition was fixed and the lattice instead varied between the cobalt spinel (Co3O4) and cobalt delafossite (HCoO2) phases. Aside from dissolution at open circuit conditions (ca. 1 V vs. RHE), the delafossite demonstrated substantially higher stability across both galvanostatic and potential cycling operating regimes. Such liquid cell results (0.1 M HClO4) were corroborated by testing in a PEMWE. With these results, we conclude that cobalt delafossites, despite exhibiting a slightly lower activity at low overpotentials, are a more suitable platform for further catalyst development. Chapter 3 thus represents our efforts exploring three parameters: Mn concentration, Mn locality (in the bulk or on the surface), and post-synthetic calcination (up to 180 °C). Increasing bulk Mn content was associated with heightened catalyst dissolution at 100 mA cm–2, whereas surface Mn content was associated with lower dissolution. Except for the bulk Mn-modified catalysts, calcination created spinel-like surfaces that both exhibited the highest initial activity and highest dissolution rates—as expected based on Chapter 2’s findings. Interestingly, though, the surface Mn-modified delafossite calcined at 180 °C defied this, exhibiting both high activity and a total metal dissolution level lower than that of unmodified HCoO2. Thus, the most promising catalyst of this study was not the result of bulk modification with Mn, but rather that with an active, spinel-like skin phase stabilized by Mn, while employing the cobalt delafossite as a stable support.