Chemical expansion is a strain induced by a change in stoichiometry, such as oxygen loss, and it can have a significant impact on device performance and lifetime. While large coefficients of chemical expansion (CCE) are needed for high displacements in an actuator, the same, large CCE can be deleterious to device longevity in a fuel cell where large chemical potential gradients exist across very small thicknesses. The breadth of CCE values needed in various devices calls for the development of design rules to tailor CCE for optimal material response, so our work targets the establishment of such structure-property insights.Oxygen-loss-induced, stoichiometric chemical expansion in oxides involves the formation of an oxygen vacancy; when oxygen leaves the lattice, charge compensating electrons are left behind and can localize on nearby multivalent cations. As cations are reduced, their atomic radii and the surrounding lattice expand. An empirical formula describing the pseudo-cubic lattice constant of perovskite materials has been developed [1] which relates the lattice parameter to the ionic radii of cation and anion components. This equation predicts that changes in the B-site cation size will have a larger effect on the lattice parameter than an equal change at the A-site. If the multivalent cation is the only one changing size during redox processes, this equation suggests that its placement on the A or B site will have a significant effect on the magnitude of the overall lattice strain during oxygen loss or gain.In an effort to develop and understand design rules to tailor CCE, two compositions, PrGa0.9Mg0.1O3 (PGM) and BaPr0.9Y0.1O3 (BPY), have been synthesized. These compositions allowed for a comparison between A and B-site multivalent Pr (nominally 3+/4+); however, we found that the empirical model did not adequately predict the differences in CCEs on this basis. Other factors including crystal symmetry, charge localization, and location of charge (anion or cation) were instead found to be significantly impactful for both compositions [2]. Values of CCE have been determined by characterizing isothermal changes in stoichiometry with thermogravimetric analysis (TGA) and corresponding changes in strain with dilatometry and in situ, high temperature XRD (HTXRD) as a function of oxygen partial pressure (pO2). The degree of charge localization has been interpreted from impedance measurements of the temperature dependence of conductivity, and the experimental results have been compared to density functional theory (DFT+U) calculations. Over the pO2 and temperature range studied, PGM and BPY have low CCEs, therefore making them of potential interest for fuel/electrolysis cell electrodes. The effects of the abovementioned design rules are discussed to provide insights into rational material design for tailored CCE.[1] Marrocchelli, D., Perry, N. H., & Bishop, S. R. (2015). Understanding chemical expansion in perovskite-structured oxides. Physical Chemistry Chemical Physics, 17(15), 10028-10039.[2] Ricote, S., Hudish, G., O’Brien, J. R. & Perry, N. H. Non stoichiometry and lattice expansion of BaZr0.9Dy0.1O3-δ in oxidizing atmospheres. Solid State Ionics (2019) doi:10.1016/j.ssi.2018.12.006.
Read full abstract