Understanding proton-conducting perovskite interfaces using atom probe tomography

by Daniel R Clark

Institution: Colorado School of Mines
Year: 2015
Keywords: Ceramics; Perovskites; Space charge; Grain boundaries; Atom probe tomography; Proton conductors
Posted: 02/05/2017
Record ID: 2120917
Full text PDF: http://hdl.handle.net/11124/20278


Proton-conducting ceramics are under intense scientific investigation for a number of exciting applications, including fuel cells, electrolyzers, hydrogen separation membranes, membrane reactors, and sensors. However, commercial application requires deeper understanding and improvement of proton conductivity in these materials. It is well-known that proton conductivity in these materials is often limited by highly resistive grain boundaries (GBs). While these conductivity-limiting GBs are still not well understood, it is hypothesized that their blocking nature stems from the formation of a positive (proton-repelling) space-charge zone. Furthermore, it has been observed that the strength of the blocking behavior can change dramatically depending on the fabrication process used to make the ceramic. This thesis applies laser-assisted atom probe tomography (LAAPT) to provide new insights into the GB chemistry and resulting space-charge behavior of BaZr0.9Y0.1O3−δ (BZY10), a prototypical proton-conducting ceramic. LAAPT is an exciting characterization technique that allows for three-dimensional nm-scale spatial resolution and very high chemical resolution (up to parts-per-million). While it is challenging to quantitatively apply LAAPT to com- plex, multi-cation oxide materials, this thesis successfully develops a method to accurately quantify the stoichiometry of BZY10 and maintain minimal quantitative cationic deviation at a laser energies of approximately 10–20 pJ. With the analysis technique specifically opti- mized for BZY10, GB chemistry is then examined for BZY10 samples prepared using four differing processing methods: (1) spark plasma sintering (SPS), (2) conventional sintering using powder prepared by solid-state reaction followed by high-temperature annealing (HT), (3) conventional sintering using powder prepared by solid-state reaction with NiO used as a sintering aid (SSR-Ni), and (4) solid-state reactive sintering directly from BaCO3, ZrO2, and Y2O3 precursor powders with 1 wt. % NiO as a reactive sintering aid (SSRS-Ni). It is ob- served that oxygen depletion (oxygen-vacancy accumulation) occurs at all GBs. Segregation of the constituent cations, Ba, Zr, and Y, is found to be variable across all samples although zirconium depletion and yttrium accumulation are most prevalent. Additionally, impurities such as Al, Fe, Mg, Ni, Si, and Sr are shown to generally accumulate at the GBs. Finally, LAAPT-derived GB chemistry data is combined with electrostatic modeling to examine the electronic structure of a BZY10 GB, revealing significant non-uniformity in the space charge region at the GB with an average space-charge potential of approximately 580 mV, extending 5–7 nm in width from the GB core. This result demonstrates how LAAPT can not only be used to further understand the role of GB chemistry within oxide materials, but can also be used to examine the electronic structure, allowing for the possibility of engineering these interfaces to improve their electrochemical performance. Advisors/Committee Members: O'Hayre, Ryan P. (advisor), Sullivan, Neal P. (committee member), Gorman, Brian P. (committee member), Tong, Jianhua (committee member), Almansoori, Ali (committee member).