Translational dependence of the geometry of metallic mono- and bilayers optimized on semi-ionic supports: the cases of Pd on -Al₂O₃(110), monoclinic ZrO₂(001), and rutile TiO₂(001)

A simplified computational scheme for "deposition" of a metallic (Pd) monolayer on semi-ionic supports is considered. An arbitrary selection of the initial position of the deposited layer relative to the surface of the support cannot provide the best solution for an optimal atom-atom binding. The energy and/or geometry changes were checked relative to a slide of the mono- and bilayers in parallel to the support layer and re-optimization. Two types of possible consequences are shown at the slide length of 1-3 angstrom for gamma-Al2O3(110), monoclinic ZrO2(001) (or m-ZrO2), and rutile TiO2(001) (or TiO2 below). First consequence option: either the Pd monolayer completely collapses after the shift/optimization, or it yields a new geometry (at gamma-Al2O3(110) and m-ZrO2) at different steps. Second consequence option: the shift leads to a loss of the Pd stabilization while the ordered monolayer's geometry is retained (at TiO2 and m-ZrO2). The extent of such destabilization varies depending on the strength of the interaction with the support. It is thus recommended to verify Pd stabilization energy by sliding/optimizing the monolayer along one or two monolayer directions in order to avoid its energy underestimation. The activity of such monolayers for modelling catalysis is demonstrated using H2O dissociation in the Pd monolayer/rutile system.