264037 Assessing the Ability of Using First Principles to Predict Fe-Ni-O Bulk Phase Stability

Wednesday, October 31, 2012: 12:48 PM
320 (Convention Center )
Zhongnan Xu, Chemical Engineering, Carnegie Mellon University, Pittsburgh, PA and John R. Kitchin, Department of Chemical Engineering, Carnegie Mellon University, Pittsburgh, PA

Mixed iron and nickel oxides surfaces show increased activity for water oxidation, a key technology in the development of hydrogen as an energy carrier, compared to the pure oxide surfaces. Experimental characterization of the mixed metal oxides showed that a NiFe2O4 spinel phase is likely to be the active phase responsible for the enhanced activity. Previous DFT computations have shown the oxygen adsorption energy to be a good descriptor for oxygen evolution activity. In this work, we aim to apply that model to the Ni-Fe oxide surfaces to determine if the improved activity can be attributed to the changes in oxygen adsorption energy. We first evaluated the ability of DFT calculations to model the bulk phase stability of the Fe-Ni-O system with respect to temperature, pressure of oxygen, and composition. We found it is essential to systematically include magnetization in the calculations; even the errors in not including magnetization are not systematic, and the relative trends in energetics are not similar. Though DFT predicts lattice constants of Fe, Ni, FeO, Fe2O3, Fe3O4, and NiO structures that are within 2% of experiments, it fails to predict two key qualities of the experimental phase diagram: the absolute stabilities of oxides with respect to temperature and pressure and the relative stabilities between oxides of different degrees of oxidation. We suggest that the first failure is caused by the overbinding of oxygen predicted by the GGA exchange-correlation functional, and the second failure is caused by different magnitudes of correlation error in oxides of different degrees of oxidation. DFT predicts accurate relative phase stabilities between wustite and bunsenite (FeO and NiO), suggesting that bulk iron and nickel oxide phases can be predicted if they are oxidized to the same degree. We will discuss some approaches to addressing these issues, and their effects on surface reaction energetics.

Extended Abstract: File Not Uploaded
See more of this Session: Computational Catalysis V
See more of this Group/Topical: Catalysis and Reaction Engineering Division