Epitaxial Stabilization
Epitaxial stabilization is the growth of a phase as a thin film on a substrate that favors it, so that a polymorph which is metastable in bulk becomes the preferred phase in the film. In a thin nucleus the free energy difference between two candidate polymorphs includes, besides the bulk volumetric energy difference, terms for strain energy and for the film/substrate and film/vapor interfacial energies. When a metastable phase forms a coherent, low-energy interface with an isostructural or closely matched substrate while the competing stable phase does not, the interfacial term can outweigh a modest bulk penalty. Because the interface term scales with area and the bulk term with volume, the effect is limited to thin films and to polymorphs within some energy window of the ground state. Epitaxy is complementary to high-pressure synthesis: it can reach polymorphs with similar or larger volume than the ground state, which compression cannot. See Oxide polymorphs and Strain and ligand effects.
An energy window for BO2 epitaxial targets (2014)
Mehta, Salvador and Kitchin balanced the bulk molar energy difference against the film/substrate interfacial energy difference for a nucleus of thickness t. Using a typical dioxide volume of about 30 Å3 per formula unit, a 1 nm nucleus and interfacial energies of about 1 J/m2 for an incoherent stable phase versus about 0 for a coherent metastable phase, they estimated that roughly 18 kJ/mol can be overcome, and proposed 10 to 20 kJ/mol of 0 K DFT energy as a working target window. Applied to six polymorphs of TiO2, VO2, RuO2, IrO2 and SnO2, this identified columbite and brookite VO2 and pyrite and columbite RuO2 as prime candidates, with isostructural substrates (for example columbite under-layers for RuO2, or polycrystalline substrates in combinatorial substrate epitaxy) suggested to achieve them. The authors caution that the true window depends on the specific interfaces and on how free energy differences vary with temperature (Mehta et al. 2014). Summary: mehta2014.
Interface energies decide anatase versus rutile TiO2 on perovskites (2017)
Xu, Salvador and Kitchin tested the capillarity model with DFT-computed bulk formation, biaxial strain and film/substrate interface energies for anatase and rutile TiO2 on the (001), (011) and (111) facets of SrTiO3 and BaTiO3, using orientation relationships that align the close-packed planes of film and substrate. Strain penalties for coherent nuclei were large (mostly 8 to 16 kJ/mol), comparable to or larger than the bulk energy difference, and bulk terms alone mispredicted the observed phases. Adding interface energies (about 0.25 J/m2 for anatase (001) on perovskite (001), about 0.1 to 0.3 J/m2 for rutile (100) on (111), and 1 J/m2 assumed for incoherent interfaces) correctly predicted anatase on (001) and rutile on (111) for all three functionals; on (011) BaTiO3 the model predicted anatase where rutile is observed, which the authors attribute to the assumed incoherent rutile interface. Recasting anatase growth on SrTiO3 (011) and BaTiO3 (001) in the idealized form of the Mehta model gave effective penalties of about 20 and 18 kJ/mol, supporting the 20 kJ/mol window as long as it is read as combining bulk, strain and interface terms (Xu et al. 2017). Summary: xu2017.
Scrutinyite versus rutile SnO2 on columbite polycrystals (2017)
Wittkamper and co-workers grew SnO2 by pulsed laser deposition on polycrystalline columbite CoNb2O6 and used EBSD to relate each film grain to its substrate grain. Metastable scrutinyite SnO2 (about 6 kJ/mol above rutile) grew on substrate orientations between the (001) and (100) poles with a unit-cell over unit-cell orientation relationship, while rutile grew near the (010) pole with a {101}<010> rutile parallel to (010)[001] columbite relationship; of 135 grain pairs, 66 were scrutinyite. The observed relationships minimized misfit strain and matched cation packing across the interface. Where strain and interface terms were comparable for both phases, rutile won on bulk stability; elsewhere strain plus cation packing mismatch penalized rutile enough for the metastable phase to form, making cation packing penalties a key criterion for choosing substrates (Wittkamper et al. 2017). Summary: wittkamper2017.