We have performed an extensive computational study of gamma-Al2O3, beginning with the geometric analysis of approximately 1.47 billion spinel-based structural candidates, followed by derivative method energy minimization calculations of approximately 122 000 structures. Optimization of the spinel-based structural models demonstrated that structures exhibiting nonspinel site occupancy after simulation were more energetically favorable, as suggested in other computational studies. More importantly, none of the spinel structures exhibited simulated diffraction patterns that were characteristic of gamma-Al2O3. This suggests that cations of gamma-Al2O3 are not exclusively held in spinel positions, that the spinel model of gamma-Al2O3 does not accurately reflect its structure, and that a representative structure cannot be achieved from molecular modeling when the spinel representation is used as the starting structure. The latter two of these three findings are extremely important when trying to accurately model the structure. A second set of starting models were generated with a large number of cations occupying c symmetry positions, based on the findings from recent experiments. Optimization of the new c symmetry-based structural models resulted in simulated diffraction patterns that were characteristic of gamma-Al2O3. The modeling, conducted using supercells, yields a more accurate and complete determination of the defect structure of gamma-Al2O3 than can be achieved with current experimental techniques. The results show that on average over 40% of the cations in the structure occupy nonspinel positions, and approximately two-thirds of these occupy c symmetry positions. The structures exhibit variable occupancy in the site positions that follow local symmetry exclusion rules. This variation was predominantly represented by a migration of cations away from a symmetry positions to other tetrahedral site positions during optimization which were found not to affect the diffraction pattern. This study has provided further insight of the defect structure of gamma-Al2O3 which is necessary for the understanding and optimization of properties. This work also demonstrates the advantages of prior use of geometric analysis and interatomic potentials to assess a large number of structural possibilities, before striving to achieve high accuracy with DFT on promising cases.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据