Skip to content

A first-principles study of stability of surface confined mixed metal oxides with corundum structure (Fe\(_2\)O\(_3\), Cr\(_2\)O\(_3\), V\(_2\)O\(_3\))

Summary

Surface-confined mixed oxides with corundum structure appear in oxidation catalysis and corrosion contexts, yet composition segregation and termination stability are difficult to probe experimentally at atomic resolution. This collaboration combines density functional theory with surface-science experiments to evaluate (0001) terminations of Fe₂O₃, Cr₂O₃, and V₂O₃ and Fe substitution in V₂O₃(0001). Plane-wave PAW calculations with Hubbard corrections on localized d states capture transition-metal oxide electronic structure. Thermodynamic analyses using oxygen chemical potentials compare surface-confined configurations to bulk and subsurface exchange scenarios, predicting when Fe enriches at the surface versus subsurface under aggressive temperature and oxygen-potential conditions.

Methods

DFT uses the PBE functional with Hubbard U parameters appropriate to Fe, Cr, and V oxides. Basis set / potentials: plane-wave expansions with PAW pseudopotentials (VASP-style setup per PCCP Computational Details). k-point sampling: Monkhorst–Pack k-point meshes specified separately for bulk references and (0001) slab supercells. Slab models represent (0001) surfaces with dipole corrections as needed. Surface energies and segregation energies enter thermodynamic constructions versus oxygen chemical potential. Experimental collaborators provide spectroscopic grounding discussed in the PCCP article. Adri van Duin’s role connects reactive modeling expertise to the broader oxide stability question even though the core results here are DFT-forward.

Surface-confined mixed oxides are represented as thin corundum-structured films where cation segregation can lower surface energy relative to homogeneous compositions, motivating the oxygen-potential diagrams used to predict stable terminations.

Dispersion / pathways. N/A — Grimme DFT-D style corrections not highlighted in this summary—confirm in pdf_path if needed for weakly bound adsorbates. NEB / reaction coordinates: N/A as the primary deliverable; the focus is surface energies, segregation energies, and oxygen chemical potential diagrams.

Findings

Outcomes. Corundum (0001) terminations of Fe\(_2\)O\(_3\), Cr\(_2\)O\(_3\), and V\(_2\)O\(_3\), plus Fe substitution in V\(_2\)O\(_3\)(0001), are ranked using DFT energies combined with μ\(_O\) diagrams to predict reduction-friendly configurations.

Comparisons. Surface-science spectroscopy/microscopy in the PCCP article benchmark predicted terminations and reduction behavior against experiment.

Sensitivity. Fe surface enrichment appears only in aggressive temperature/low μ\(_O\) windows; Fe doping increases reducible character relative to endmembers in the model space shown in the figures.

Limitations and PDF grounding. 0 K DFT omits entropy/kinetic trapping; Hubbard U shifts absolute energies. Quantitative boundaries belong to the PCCP PDF figures (pdf_path), not extrapolated here.

Limitations

0 K DFT with mean-field thermodynamics omits entropic contributions from adsorbates and kinetic trapping; U values influence absolute energies.

Corpus notes

When cross-walking to ReaxFF simulations of the same oxides, document which surface terminations were fixed in DFT versus allowed to react in MD, because the thermodynamic diagrams here assume static slabs without explicit solvent.

Surface oxygen chemical potentials in the figures should be translated to experimental \(p_{\mathrm{O_2}}, T\) windows using the same thermodynamic conventions the authors document, rather than ad hoc scalings from unrelated catalysis datasets.

Relevance to group

van Duin-group participation in oxide catalysis stability alongside experimental surface-science partners.

Citations and evidence anchors