Ab Initio Atomistic Thermodynamics
Ab initio atomistic thermodynamics combines DFT total energies with gas-phase chemical potentials to find which surface composition and structure has the lowest surface free energy at a given temperature and pressure. Adsorbate chemical potentials are tied to reservoir gases (often treated as ideal gases with tabulated thermochemistry), so phase diagrams plotted against chemical potential can be read in terms of experimental conditions. Bounds on the chemical potentials come from the formation of bulk phases such as oxides or sulfides.
Extension to alloy surfaces with a bulk reservoir (2008)
Kitchin, Reuter and Scheffler extended the formalism to a binary alloy slab in equilibrium with both an O2 gas and the bulk alloy, so the surface free energy depends on the oxygen chemical potential and on the difference of the metal chemical potentials (Kitchin et al. 2008). They bracket the latter between Ag-rich and Pd-rich phase-separation limits (a window equal to the alloy formation energy divided by x(1 - x), 0.28 eV for Ag3Pd), and rewrite the surface free energy as a sum of segregation energy, average adsorbate binding energy and gas-phase terms to separate the driving forces. They present this direct screening of ordered structures as complementary to cluster expansion approaches. Summary: kitchin2008.
Cu surfaces in a water-gas shift environment (2009)
İnoğlu and Kitchin applied the Reuter-Scheffler formalism to Cu(111), (100) and (110) in contact with H2, H2S, H2O, CO and CO2, tying S to H2S/H2, O to H2O/H2 and CO to gas-phase CO (INOGLU 2009). They argued that adsorbate vibrational (2 to 6 meV/Ų) and configurational (at most about 4 meV/Ų) contributions are small enough to neglect, so surface free energies follow from DFT energies alone. Phase diagrams showed S, O and CO phases at 0.25 to 0.75 ML on the different facets, bounded by bulk CuS and Cu oxide formation, and for a representative 50 atm WGS feed with 1 ppm H2S the surfaces are partly CO-covered, with S phases taking over at higher H2S levels. Summary: inoglu2009.
Atomistic thermodynamics as a screening filter for alloys (2011)
İnoğlu and Kitchin used a surface free energy containing the clean host surface energy, an impurity segregation energy, the S adsorption energy and the S chemical potential to decide, for 370 alloy surfaces, at what H2S/H2 chemical potential each becomes S-covered and whether the alloy arrangement is stable against segregation. Adsorption energies came from a cheap d-band width model rather than direct DFT, so the thermodynamic step served as a fast filter; propagated model uncertainty shifted poisoning onsets by about 3%. The authors stress that S binding alone is insufficient, since, for example, a Au monolayer on Cu binds S weakly but risks S-induced Cu segregation (İnoğlu 2011). Summary: inoglu2011.