Phase Stability of SrTi1-XFexO3- δ Under Solid Oxide Cell Fuel-Electrode Conditions: Implications for Related Exsolution Electrode Materials

Abstract

Perovskite oxides with catalytically active metal nanoparticle exsolution are receiving considerable attention as alternative Solid Oxide Cell fuel-electrode materials. For exsolution, a small amount of a cation is substituted on the B-site of the base oxide and then is reduced out of the host lattice to form nanoparticles that promote electrochemical processes. During exsolution, the host lattice generally becomes more B-site deficient. In some cases, the oxide is made initially A-site deficient to compensate for the loss of B-site cations. Thus, the stability of these oxides under a range of stoichiometries and fuel electrode conditions is of interest. Here we focus on one perovskite host material of interest, SrTi 1-x Fe x O 3- δ (STF), which has shown good fuel electrode characteristics that can be enhanced by the substitution of Ru or Ni, resulting in exsolution to form Ru-Fe or Ni-Fe alloy nanoparticles, respectively. (1, 2) Although the amounts of Ru or Ni substituted are small, typically ~ 7% of the B-site cations, the co-exsolution of a comparable amount of Fe is usually observed, resulting in substantial B-site deficiency, potentially de-stabilizing the host perovskite phase. The stability of Fe in STF is also of interest because it helps to determine the Fe content of the alloy nanoparticles. (3) This study seeks to determine the stability of STF in various reducing fuel environments. While STF has been mostly studied as SrTi 0.3 Fe 0.7 O 3-δ (STF-7), it is not known if this composition provides the best combination of stability and electrochemical performance. Thus, a few additional compositions, SrTi 0.5 Fe 0.5 O 3- δ (STF-5), SrTi 0.4 Fe 0.6 O 3- δ (STF-6), and SrTi 0.2 Fe 0.8 O 3- δ (STF-8), have been studied. In general, the Fe-rich compositions are expected to possess higher electronic and ionic conductivity, leading to good electrochemical performance, whereas the more Ti-rich compositions are expected to provide better stability. Exposure to 97% H 2 – 3% H 2 O at 850°C for 4 h resulted in STF decomposition into BCC α-Fe and a Ruddlesden-Popper (RP) phase, but the decomposition became more limited for the more Ti-rich compositions (Figure 1a). The main Fe peak overlaps with one of the RP peaks, but the presence of α-Fe particles is clearly visible in STEM energy dispersive x-ray spectroscopy chemical mapping, shown for STF-7 in Figure 1b. Figure 1c shows that there is a critical p(O 2 ), 1.3 x 10 -20 atm, below which decomposition of perovskite STF-7 occurs. In addition to ex situ XRD as shown in Figure 1, in situ XRD results will be presented and used to confirm and quantify phase changes in combination with thermogravimetric analysis (TGA). Additionally, the conductivity and electrochemical performance of the various STF compositions will be reported and discussed relative to the phase change from perovskite to Ruddlesden Popper and the Fe exsolution. The other main materials variable to be explored is the A-to-B site stoichiometry, which will be studied over the range expected during exsolution. The implications of these results for various exsolution electrode compositions will be discussed. Figure 1: a) Reductions of STF-5 through STF-8 all yield significant decomposition as the initially pristine perovskite acquires several additional peaks. b) Using energy dispersive x-ray spectroscopy, it can be shown that significant Fe deposits appear in STF-7 after reduction at 850°C (effective pO 2 1.2e-21 atm for 4 hours). c By increasing pO 2 by a factor of 10, XRD patterns for STF-7 show no clear Ruddlesden-Popper or α-Fe peaks, indicating that decomposition requires significantly reducing conditions and that STF may remain stable under more oxidizing conditions. References R. Glaser, T. Zhu, H. Troiani, A. Caneiro, L. Mogni and S. Barnett, Journal of Materials Chemistry A , 6 , 5193 (2018). T. Zhu, H. Troiani, L. V. Mogni, M. Santaya, M. Han and S. A. Barnett, Journal of Power Sources , 439 (2019). T. Zhu, H. E. Troiani, L. V. Mogni, M. Han and S. A. Barnett, Joule , 2 , 478 (2018). Figure 1