Conference Dates

September 4-8, 2016


Atomic scale simulations are used to predict how excess oxygen is accommodated across three distinct groups of oxides: group II monoxides [1], fluorite di-oxides [2] and zirconate perovsites [3]. In addition to the crystallographic position and orientation of the peroxide molecule, transport of the species is also considered. For each group, three different cations are considered in order to determine how stability and structure change as a function of cation size. Peroxide molecular vibrational frequencies are also predicted to facilitate experimental investigation of the various structural models. For all simulations, the density functional code VASP was employed.

Starting with the monoxides, in all cases, the preference is to form a peroxide ion centered at an oxygen site, rather than a single oxygen species, although the peroxide molecular orientation changes from <100> to <110> to <111> with increasing cation radius. The enthalpy for accommodation of excess oxygen in BaO is strongly negative, whereas in SrO it is only slightly negative and in CaO and MgO the energy is positive. Interestingly, the increase in material volume due to the accommodation of oxygen (the defect volume) does not vary greatly as a function of cation radius. Calculations of the BO2 dioxide structures have also been carried out. For these materials the oxygen vacancy formation energy is always positive (1.0–1.5 eV per oxygen removed) indicating that they exhibit a surprisingly small oxygen deficiency.

For the di-oxides, accommodation of hyperstoichiometry is considered in CeO2, ThO2 and UO2. Calculations indicate a preference for the peroxide species over an isolated interstitial in CeO2 and ThO2 but not in UO2. Frenkel pair defects are investigated to understand if the interstitial component could assume a peroxide like configuration in the vicinity of the vacancy. While it is expected that this would not be the case for UO2 since peroxide was not stable, it is also not found to be the case for CeO2 and ThO2 with the peroxide disassociating into a lattice species and a separate interstitial ion.

For the zirconate perovskites, again group II cations are the variable: BaZrO3, SrZrO3 and CaZrO3. While facilitated by peroxide, in contrast to the monoxide, the solution energy of O2 is predicted to be positive (though close to zero) for BaZrO3. Similar to the monoxides, the peroxide molecule is less favourably accommodated in SrZrO3 or CaZrO3. This trend is tested experimentally by exposing SrZrO3 and BaZrO3 to hydrogen peroxide solution and carrying out Raman spectroscopy measurements to look for a peak indicative of peroxide ions. A peak was observed at 1000 cm-1 in both compositions, suggesting the theoretically predicted peroxide ion is present. The transport of the excess oxygen through the perovskite lattice was predicted to proceed with activation energies of less than 1 eV in each of the systems.

1. Middleburgh S. C., Lagerlof K. P. D. & Grimes R. W. “Accommodation of Excess Oxygen in Group II Monoxides” J. Am. Ceram. Soc. 96, 308 (2013).

2. Middleburg S. C., Lumpkin G. R. & Grimes R. W. “Accommodation of excess oxygen in fluorite dioxides” Solid State Ionics, 253, 119 (2013).

3. Middleburgh S. C., Karatchevtseva I., Kennedy B. J., Burr P. A., Zhang Z., Reynolds E., Grimes R. W. & Lumpkin G. R. “Peroxide defect formation in zirconate perovskites” J. Mater. Chem. A, 2, 15883 (2014).