Curvature Relaxation Measurements of the Oxygen Surface Exchange Coefficient of Thin Film Gadolinium-Doped Cerium Oxide

Introduction

Previous investigations of Gadolinium-doped Cerium Oxide (GDC) have demonstrated its applicability for anode as well as electrolyte applications in Solid Oxide Fuel Cells (SOFCs) [1-3]. In particular, GDC has shown to be a desirable catalyst for CH₄ fuel oxidation due to its ability to suppress carbon precipitation on SOFC anodes [4]. When used as an anode material, GDC must have a high oxygen surface exchange coefficient, k_chem, to facilitate oxygen exchange.

 Previous efforts to measure GDC k_chem values have utilized techniques such as thermo-gravimetric analysis (TGA) [2, 5] and electrical conductivity relaxation (ECR) [6]. At identical testing conditions the results of these measurements often differ by an order of magnitude, or more. The present work aims to help resolve these discrepancies by determining GDC k_chem values through a recently-developed curvature relaxation (κR) technique [7].

Experimental Methods

Following the procedure described in Ref. [7], the κR response of a mechano-chemically active film [8] (in this case, a 570 nm thick sputtered GDC thin film) on an inert substrate (in this case a 200 micron thick (100)-oriented (Y₂O₃)_0.13(ZrO₂)_0.87 single crystal) equilibrating to an abrupt step in oxygen particle pressure, p(O­₂), was used to measure k_chem.  Here, p(O₂) was altered by switching between 2∙10^-27/bar and 2∙10^-28/bar at 600^oC, and between 3.2∙10^-29/bar and 3.2∙10^-30/bar at 550^oC. The p(O₂) was controlled by switching between a 68 sccm dry H₂-32 sccm 25^oC humidifed H₂ mixture and a 100 sccm 25^oC humidified H₂gas stream.

 

Results and Discussion

Fig. 1 shows the changes in bilayer curvature over multiple p(O₂) cycles at 550^oC and 600^oC. The reproducibility in the long-term (i.e. equilibrium) curvatures of the oxidized and reduced stages suggest an absence of factors leading to irreproducibility; such as sample damage through cracking or delamination.

The k_chem values obtained by the κR technique are compared with literature in Fig. 2. The activation energy, E_a, of the κR-measured k_chem is 0.4±0.1 eV, compared to 0.7 eV for literature ECR-measured k_chem values [6].  The k_chem values reported here are noticeably lower than those found in the literature for similar temperatures. While it is possible that this mismatch is due to the difference in the applied p(O₂), the relationship between p(O₂) and k_chem in Fig. 2 does not immediately follow the reported trends. Identification of the source of these discrepancies is on-going, and will be discussed in this work.

Acknowledgements

This work was funded by the Army Research Office Award #W911NF-13-1-0404. The authors would also like to thank Christopher Traverse for assistance with film sputtering.  

References

[1] M. Mogensen, N.M. Sammes, G.A. Tompsett, Solid State Ionics, 129 (2000) 63-94.

[2] K. Yashiro, S. Onuma, A. Kaimai, Y. Nigara, T. Kawada, J. Mizusaki, K. Kawamura, T. Horita, H. Yokokawa, Solid State Ionics, 152–153 (2002) 469-476.

[3] M. Sahibzada, B.C.H. Steele, K. Zheng, R.A. Rudkin, I.S. Metcalfe, Catalysis Today, 38 (1997) 459-466.

[4] V.D. Belyaev, T.I. Politova, O.A. Mar'ina, V.A. Sobyanin, Applied Catalysis A, 133 (1995) 47-57.

[5] S.R. Bishop, K.L. Duncan, E.D. Wachsman, Electrochimica Acta, 54 (2009) 1436-1443.

[6] A. Karthikeyan, S. Ramanathan, Applied Physics Letters, 92 (2008) 243109-1 to 243109-3.

[7] Q. Yang, T.E. Burye, R.R. Lunt, J.D. Nicholas, Solid State Ionics, 249-250 (2013) 123-128.

[8] S.R. Bishop, K.L. Duncan, E.D. Wachsman, Acta Materialia, 57 (2009) 3596-3605.