Oxygen transport membranes (OTM) are dense ceramics oxides that can move oxygen along a chemical potential gradient. The membrane must be oxygen-ion conductive to transport oxygen, but should also be electronically conductive in order to maintain internal charge neutrality. In the simplest design, the ceramic membrane separates two atmospheres with different oxygen partial pressures, creating the gradient which drives oxygen flux. At the oxygen-rich side, oxygen is adsorbed on membrane surface, then diffuses through the bulk, and is finally released at the oxygen-lean side.

OTMs have been used to precisely control the amount of oxygen in the mixture supplied for the partial oxidation of methane. In oxy-fuel combustion, using oxygen supplied by an OTM instead of air increases combustion temperature and decreases NOx emissions. These, and most other OTM applications, operate under conditions of temperature, gas pressure and oxygen partial pressures, which can lead to thermo-chemical degradation and mechanical failure of the OTM. Therefore, to be an economically viable option, OTMs must excel in two requirements which fundamentally drive cost: performance and stability. In perovskite materials, the oxygen diffusion is proportional to the concentration of mobile vacancies in the structure. However, vacancies are also responsible for creep, which often limits the lifetime of the ceramic. Consequently, the improvement of performance comes at the expense of stability, and vice-versa.

CaTi_0.9Fe_0.1O_3−δ (CTF) exhibits an interesting tradeoff between performance and stability, where its lower oxygen flux can be compensated by its superior thermal, mechanical and chemical stability [3]. To optimize flux performance of CTF it is important to establish the rate limiting step during the transport of the oxygen, which may be limited by the rate of the oxygen exchange at either side, or by bulk diffusion. In this study, we will characterize the oxygen semi-permeability, and establish the rate limiting step of CTF membranes as a function of temperature.

CaTi_0.9Fe_0.1O_3−δ powder was prepared by electrofusion and checked for chemical purity and composition by X ray fluorescence. The resulting powders have an orthorhombic crystal structure, increasing symmetry with temperature increase, becoming cubic at 1300 °C. Sintered pellets with > 96 %TD retain the same crystalline structure, and were used for conductivity and oxygen flux measurements.

Conductivity was measured between 150 and 700 °C and between 1 and 10^-27atm. CTF is a p-type conductor, limited mostly by ionic conductivity at lower temperature, transitioning to electronic conductivity above 700 °C, also verified during flux measurements. Regardless, the conduction behavior is never really dominated by either type of charge carrier, and can be better characterized as mixed-limited under the experimental conditions.

Semi-permeability measurements were performed under synthetic air (0.21 bar) and argon (10^-5 bar), between 550 and 950 °C. Oxygen semi permeation at 900 °C is 3.2 × 10^-3 ml.min^-1.cm^-1, consistent with the literature [4]. Additionally, electrochemical surface probes were used to measure the oxygen activity of both surfaces of the membrane. The surface polarization of the oxygen-rich side is negligible compared to the oxygen-lean side, indicating that the former is not limiting.

The limiting step on the lean side can be inferred from the ratio between chemical potential drop of oxygen near the surface and across the bulk. Figure 1 show that below 750 °C CTF is mixed limited by the surface exchange as well as bulk diffusion, and that at higher temperature it is increasingly limited by bulk diffusion.

Assuming that conductivity is limited by ionic charge carriers, we calculated oxygen diffusion, D*, and surface exchange, k, coefficients and compared them with isotopic oxygen exchange between 700 and 900 °C. The characteristic thickness, L_c = D*/k, is superior to 0.6 mm above 700 °C, which confirms that the oxygen semi-permeability is limited by bulk diffusion at OTM operation temperatures.

References

[1] J. Fouletier, P. Fabry, and M. Kleitz, J. Electrochem. Soc. 123  (1976) 204.

[2] P.M. Geffroy, M. Reichmann, T. Chartier, J.M. Bassat, and J.C. Grenier, J. Membrane Sci. 451 (2014) 234.

[3] F. M. Figueiredo, J. Waerenborgh, V. V. Kharton, H. Näfe, and J. R. Frade, Solid State Ionics, vol. 156, no. 3–4, pp. 371–381, 2003.

[4] J. M. Polfus, W. Xing, M. F. Sunding, S. M. Hanetho, P. I. Dahl, Y. Larring, M.-L. Fontaine, and R. Bredesen, “Doping strategies for increased oxygen permeability of CaTiO3 based membranes,” J. Memb. Sci., vol. 482, pp. 137–143, May 2015.