Oxygen electrodes in solid oxide electrochemical cells such as Solid Oxide Fuel Cells (SOFCs) or electrolysers (SOECs) are typically based on LaBO₃ perovskites, where B is a transition metal such as Co, Mn or Fe. In order to impart oxide ion conductivity and electro-catalytic activity towards oxygen exchange, the La is often partially substituted with a divalent alkaline earth cation such as strontium. Although this Sr substitution is necessary to improve the electrode performance, the Sr tends to segregate to the outer surface of the electrode, where it is implicated in the decrease of surface exchange activity, and hence device performance [1].

            Our group has shown that this segregation occurs rapidly, with significant changes in the surface composition of polished dense ceramics of these materials observed in times as short as 15 minutes at temperatures as low as 400 °C [2,3]. Clearly the driving forces are sufficiently high, and the kinetics (thought to be mediated by grain boundary diffusion [4]) so fast that suppressing this segregation is a challenge.

            One strategy to altogether avoid the problems associated with Sr segregation would be to identify new candidate materials which do not require Sr doping in the first place. In this way, the surface composition, and hence oxygen exchange activity, is expected to be more stable. However, the absence of Sr to introduce oxygen vacancies may severely hamper the oxygen transport and exchange properties. Therefore, it would be necessary to investigate alterative doping strategies to tailor the oxygen transport properties.

            In this work, we study Sr-free perovskite materials, such as LaCo_1-xNi_xO_3-d (LCN), in terms of their surface composition and oxygen exchange properties. The measurements combine Low Energy Ion Scattering (LEIS) spectroscopy studies of the surface composition, combined with ¹⁸O exchange-Secondary Ion Mass Spectrometry (SIMS) measurements of the oxygen transport properties.

            Although preliminary results indicate that the oxygen diffusivity in these materials is not as high as would be desired for a mixed conducting oxide electrode, the transport behavior is superior to that of the Sr-containing perovskite La_0.8Sr_0.2MnO_3+d, which does find commercial application. At low to intermediate temperatures, the oxygen diffusion profiles (figure 1) show a significant contribution from grain boundary diffusion (Harrison type B kinetics [5]), whereas around 700 ºC, the diffusion kinetics transitions towards the Harrison “A” regime, where no grain boundary tailing is evident.

            More interestingly, the obtained surface exchange coefficients (k*) are relatively high. Around 700 ºC, the Sr-free materials show k* values comparable to La_0.6Sr_0.4CoO_3-d. This may be related to the surface composition; LEIS spectroscopy studies of the composition of the very outer atomic surfaces of the samples (fig. 2) show that the transition metal cations persist at the surface, whereas their coverage is very low in Sr-containing materials such as La_0.6Sr_0.4CoO_3-dafter comparable treatments.

            However, although no divalent cation segregation is present, sodium impurities are observed at the outer surfaces after annealing at higher temperatures. It is not clear as of writing whether this is due to the segregation of bulk impurities, or cross contamination in the annealing apparatus. These changes in surface composition may have implications for the longer-term stability of these materials.

REFERENCES:

[1] G.M. Rupp et al., J. Mater. Chem. A 3(45), pp. 22759 – 22769 (2015)

[2] H. Tellez et al., Int. J. Hydrogen Energy 39(35), pp. 20856-20863, (2014)

[3] J. Druce et al., ECS Trans. 66(2), pp 61-68, (2015)

[4] M. Kubicek et al., Phys. Chem. Chem. Phys. 16(6), pp. 2715-2726 (2014)

[5] L.G. Harrison, Trans. Faraday Soc. 57, p. 1191 (1961)