Open-Circuit Voltage Anomalies in Yttria-Stabilized Zirconia and Samaria-Doped Ceria Bilayered Electrolytes

Abstract

The open-circuit voltage (OCV) in samaria-doped ceria (SDC) electrolytes has been explained by Wagner’s equation. According to this theory, even if there is no external current, two types of internal currents are created by electrons and oxygen ions. Due to the ohmic loss caused by the ionic internal short-circuit current, the OCV is 0.80 V, which is lower than the theoretical voltage of 1.15 V. Coating a thin electron-blocking layer of YSZ (yttria-stabilized zirconia) onto the SDC electrolyte is effective at stopping the electronic leakage current and improving the OCV. For cells with YSZ on the anode side, the OCV is 1.05 V-1.15 V. For cells with YSZ on the cathode side, the OCV is 0.81-0.87 V. Therefore, the electronic leakage currents have been thought to be not entirely stopped. YSZ films deposited on the anode side are very effective, but YSZ films deposited on the cathode side are clearly less effective. Thus, we proposed experiments to confirm whether electronic leakage currents can be blocked by the YSZ films deposited on the cathode side. A polished YSZ film (500 mm thickness) on the cathode side and a polished SDC film (970 mm) on the anode side were physically contacted. The OCV was measured to be 819 mV at 1073 K, 825 mV at 973 K and 874 mV at 873 K. Then, we noticed that the large voltage loss was not due to electronic leakage currents. However, there were objections in these experimental results. The surrounding air could enter the contact space between the YSZ film and the SDC film, which indicated that due to air entering the contact, an electronic leakage current could arise in SDC electrolytes. By measuring the transient process, we showed that the electronic leakage currents were stopped entirely. After changing the anode gas from N2 gas to H2 gas including 3% H2O, the OCV was changed from 85 mV to 755 mV with a short time constant (12 seconds) and from 755 mV to 819 mV with a long time constant (100 seconds) at 1073 K. The equilibrium time based on the electron diffusion coefficient equation should be longer than 49 minutes. Thus, the voltage loss (0.35 V) in SDC could not be explained by the electronic leakage current. Furthermore, a higher OCV (819 mV) was not due to stop the electronic leakage current. Figure 1