Predicting monolayer oxide stability over low-index surfaces of TiO\(_2\) polymorphs using *ab initio* thermodynamics
Summary¶
Oxide-on-oxide coatings are pervasive in catalysis and corrosion protection, yet predicting whether a transition-metal dioxide will wet a titania support as a coherent monolayer—and under which polymorph and facet combination—requires balancing surface energies, epitaxial strain, and electronic structure. This Langmuir article uses plane-wave DFT together with ab initio thermodynamics to compare MO\(_2\) (M = Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn) monolayer configurations on low-index surfaces of anatase, brookite, and rutile TiO\(_2\). The study is explicitly framed as a screening exercise: it identifies which oxide coatings are thermodynamically favored as monolayers versus bulk oxide precipitation, and it discusses when polymorph matching between film and support and relative substrate surface energy emerge as useful qualitative trends. The authors also emphasize that no single structural or electronic descriptor explains all computed stabilities across the broad composition space, pointing toward multivariate structure–property relationships for future descriptor and machine-learning work.
Methods¶
Electronic-structure calculations use VASP with the PBE GGA, PAW pseudopotentials, and plane-wave expansions with Monkhorst–Pack k-point grids (per bulk and slab construction in the article). Calculations are spin-polarized where appropriate, and dipole corrections are applied for asymmetric slabs. DFT+U is applied to Ti 3d states with literature-informed U values tuned against band gaps and reduction energetics, using Vosko–Wilk–Nusair interpolation for the exchange-correlation treatment as stated in the methods. Surface models span low-index facets of the three TiO\(_2\) polymorphs; monolayer free energies are evaluated relative to bulk oxide and particle references using the thermodynamic constructions described in the paper (including corrections noted there). Force and energy convergence criteria are set to stringent thresholds (for example 0.05 eV/Å on forces, with additional criteria in-text). Together, these pieces define a reproducible static-lattice, 0 K** thermodynamic picture of monolayer stability for the MO\(_2\) family on TiO\(_2\).
Dispersion / other DFT knobs. N/A — semi-empirical DFT-D3 (or similar) dispersion add-on not called out in this summary; verify in pdf_path if vdW contributions matter for the MO\(_2\) monolayers. Reaction pathways: N/A — NEB not central; the work is ab initio thermodynamics over relaxed slab geometries.
Findings¶
Outcomes. Monolayer stability on anatase, brookite, and rutile facets is reported as a function of M in MO\(_2\), with interface energies referenced to bulk oxide reservoirs in the ab initio thermodynamics construction.
Comparisons / heuristics. Polymorph alignment between film and support and relative substrate surface energy are useful screening rules, yet the authors stress multifactor electronic/strain effects—no single descriptor fits the entire dataset.
Sensitivity. Facet choice and polymorph alter predicted wetting/adhesion rankings at 0 K; finite-temperature and solvent effects are deferred (see Limitations).
Limitations and PDF grounding. PBE+U band-gap and reduction energetics carry known uncertainty; static 0 K diagrams omit configurational entropy and explicit water for operando wetting. Tabulated stabilities and phase boundaries should be read from the Langmuir PDF (pdf_path).
Limitations¶
PBE+U accuracy limits quantitative oxidation/reduction energetics; 0 K static models omit configurational entropy, explicit solvation, and finite-temperature vibrational free energy that can matter for wetting at operating conditions.
Relevance to group¶
van Duin-group coauthored oxide interface thermodynamics that connects cleanly to later ML descriptor work (2018jonayat-acs-discovery-descriptors) and related first-principles oxide studies.
Citations and evidence anchors¶
- DOI:
10.1021/acs.langmuir.8b02426.