Impact of Sr-Containing Secondary Phases on Oxide Conductivity in Solid-Oxide Electrolyzer Cells

Solid-oxide electrolyzer cells (SOECs) based on yttria-stabilized zirconia (YSZ) oxide electrolytes are devices capable of producing hydrogen with excess thermal energy. However, beginning with initial materials sintering and extending through electrochemical aging, Sr diffusion within the Gd-doped CeO2 (GDC) barrier layer has been observed to lead to the formation of unwanted secondary phases such as SrO and SrZrO3. To establish the impact of these phases on SOEC performance, we perform firstprinciples calculations to determine secondary phase bulk oxide conductivities and compared them to that of the YSZ electrolyte. We find that SrO has a low conductivity arising from poor mobility and a low concentration of oxygen vacancies (V_O^2+), and its presence in SOECs should therefore be avoided as much as possible. SrZrO3 also has a lower oxide conductivity than YSZ; however, this discrepancy is primarily due to lower V_O^2+ concentrations, not V_O^2+ mobility. We find Y-doping to be a viable strategy to increase V_O^2+ concentrations in SrZrO3, with 16% substitution of Y on the Zr site leading to an ionic conductivity on par with that of YSZ. Energy dispersive x-ray spectroscopy obtained using scanning transmission electron microscropy on cross-sections of SOECs indicates that Y is the most common minority element present in SrZrO3 forming near the GDC—YSZ interface. Thus, we expect SrZrO3 to be rich in V_O^2+ and not to hinder long-term device performance. These results from our combined computational–experimental analysis can inform future materials engineering strategies designed to limit the detrimental effects of Sr-induced secondary phase formation on SOEC performance.