Accelerated Computational Design of Mixed Protonic and Electronic Conduction for H2 Separation through Tailoring Polaron

Thursday, 5 October 2017: 14:20
National Harbor 7 (Gaylord National Resort and Convention Center)
Q. Bai, Y. Zhu, X. He (University of Maryland, College Park), and Y. Mo (University of Maryland Energy Research Center, University of Maryland, College Park)
The technology of H2 separation is essential for the utilization of hydrogen fuel, which is strongly motivated by the grand challenge of climate change. Electrochemical membrane based on cerate, titanate and zirconate ceramics is a promising technology of H2 separation. In particular, SrCeO3 perovskites with mixed protonic and electronic conductivities are promising materials for H2 separation due to negligible oxygen permeability. However, low electronic conductivity of SrCeO3 often limits the overall H2 flux. The increase of electronic conductivity via B-site doping has been largely based on trial-and-error experiments. This lack of understanding limits further design and development of hydrogen membrane materials. In this presentation, we will share our computation studies of revealing distinctive behaviors of different dopants in enhancing electronic conductivity via novel polaron mechanism, using first principles hybrid functional calculations. Our study shows SrCeO3 can trap excess charge as electron or hole polarons on Ce4+ or certain dopants. The enhancement of different dopants to electronic conductivity can be determined by the ability of polaron formation and migration. Based on the insight of the polaron enhancement mechanism, new promising dopants are predicted to increase electronic conductivity, and the computation predictions are verified by preliminary experiments. Thus, our study demonstrates first principles techniques in accelerating deign of mixed conductors through tailoring polaron, and provides new avenues for the future design of the materials with different polaronic behavior to achieve improved conductivity and other properties.