O_O^X → v_O^•• +1/2 O₂ (g) + 2e' (1)

v_O^•• + O_O^X + H₂O → 2(OH)_O^• (2)

The occurrence of chemical expansion in solid state ionic materials and devices can cause large chemical stresses, leading to mechanical failure. To prevent the formation of cracks and delamination, this chemical stress should be minimized. One way to achieve this goal involves minimization of stoichiometry gradients within materials, which is dependent on processing and operating conditions, their relationship to the material’s point defect thermodynamics, and the chemical diffusivity of the material. Another approach to minimize chemical stress, which we take, is to focus on an intrinsic material property, the coefficient of chemical expansion. For oxides that change their oxygen stoichiometry, the CCE can be defined as:

CCE = ε_C / Δδ (3)

while for oxides that change their hydration or proton content, the CCE can be defined as:

CCE = ε_C / Δ[(OH)_O^•]. (4)

We seek to identify and understand the factors that affect the CCE, so that its magnitude may be rationally tailored.

Measurements have focused on a wide array of bulk perovskite-structured titanate, stannate, zirconate, cerate, and gallate compositions as well as on nanoparticles of fluorite-structured (Ce,Pr)O_2-δ. Chemical strains have been measured macroscopically by dilatometry and at the crystal-structure level by in situ X-ray and neutron diffraction as a function of gas atmosphere. For nanoparticle measurements, a specially-designed cell was employed for simultaneous dilatometric and impedance measurements, leading to possibly the first assessment of nanopowder chemical expansion by this approach. Stoichiometry changes in corresponding conditions have been assessed by thermogravimetric analysis.

In initial studies, factors including the size of the oxygen vacancy radius, charge localization, and temperature were identified as variables that could be tailored to modify the CCE. In more recent work, we have been focusing on the effect of nanostructuring, the impact of the choice of multivalent cation, and the role of crystal symmetry in perovskites. Regarding the latter variable, the results are consistent with a general pattern where deviation of the tolerance factor away from 1 (the ideal A- to B-site cation ratio giving cubic symmetry) correlates with lower CCE values. For nanopowders, despite well-documented differences in their point defect chemistry vs. bulk materials, no measurable deviation from bulk CCE behavior was observed. Implications for durable processing and use of these mixed conductors will be addressed.