Chrome Poisoning of Non-Manganiferous Cathode Materials in Solid Oxide Fuel Cells (SOFCs)

Abstract

Chrome poisoning is considered to be one of the main degradation issues of SOFC cathode materials and thus a major cause for decreased SOFC component performances in combination with a reduced durability [1, 2]. For this reason, deeper understanding of the related chrome degradation mechanisms is of fundamental interest. Since (La,Sr)(Co,Fe)O3 based cathodes show both, a good electronic and sufficient ionic conductivity, these materials are considered to be less susceptible towards chrome poisoning than pure electronic conductors such as (La,Sr)MnO3. Nevertheless, recent performed long-term stack tests showed similar poisoning effects for LSCF materials as seen for LSM cathodes, urging us to perform another in-depth analysis of the mechanisms of chrome poisoning in non-manganiferous cathode materials [3]. LSCF materials with different stoichiometry's (variable cobalt and iron contents) have been synthesized and analyzed regarding their phase purity and microstructure. Anode supported SOFCs were manufactured with these LSCF compositions and characterized by electrochemical measuring techniques (DC/AC measurements). Here, we mainly focus on impedance spectroscopy in a broad range of different operating conditions (e.g. temperature, current density, and humidity on the air side) in presence and absence of a suitable chromium source. Electrochemical impedance measurements performed at 900 °C and a cell voltage of 0.7 V in dry and humid air conditions (Figure 1) reveal that besides an initial formation step (I in Fig. 1) the ohmic and polarization resistances only show small deviations in absence of a chrome source (Rohm: dry ~ 0.125 Ω*cm², humid 0.127 - 0.130 Ω*cm²; Rpol: dry ~ 0.045 Ω*cm², humid 0.047 Ω*cm²; see Fig. 1). During a step-wise increase of current density, i.e. cathode polarization, over 48 hour periods both Rohm and Rpol remained stable (see Fig. 1). Besides the above mentioned 4-point measurements on full cells, 3-point measurements with a reference electrode were performed on symmetrical electrolyte supported half cells. By this means contributions from the Ni/YSZ anode to the overall cell impedance can be excluded and the individual cathode contributions investigated in more detail. In combination, both measuring setups should enable detailed information about the individual resistance contributions of the synthesized cathode materials with respect to the operating conditions.