Surface polarization, rumpling, and domain ordering of strained ultrathinBaTiO3(001) films with in-plane and out-of-plane polarization

Abstract

BaTiO 3 ultrathin films (thickness ≈ 1.6 nm) with in-and out-of-plane polarization are studied by first-principles calculations. Out-of-plane polarization is simulated using the method proposed by Shimada et al. [Phys. Rev. B 81, 144116 (2010)], which consists in building a supercell containing small domains with alternating up and down polarization. This allows one to investigate the properties of defect free BaTiO 3 ultrathin films with polarization perpendicular to the surface, as a function of in-plane lattice constant, i.e., epitaxial strain. The configurations with polarization perpendicular to the surface (c phase) are found stable under compressive strain, while under tensile strain, the polarization tends to lie in-plane (aa phase), along [110]. In the c phase, the most stable domain width is predicted to be 1 to 2 lattice constants, and the magnitude of the surface rumpling varies according to the direction of the polarization (upwards versus downwards), though its sign is unchanged, the oxygen anions pointing in all cases outwards. Finally, all the surfaces studied are found to be insulating. Analysis of the atom-projected electronic density of states gives insight into the surface contributions to the electronic structure. An important reduction of the Kohn-Sham band gap is predicted at TiO 2 terminations in the c phase (≈ 1 eV with respect to the aa phase). The Madelung potential at the surface plays the dominant role in modifications of the surface electronic structure.

Influence of hole depletion and depolarizing field on theBaTiO3/La0.6Sr0.4MnO3interface electronic structure revealed by photoelectron spectros

“…This geometry allows us to obtain in a single calculation the relevant electronic quantities of both electrode/P + and electrode/P − interfaces for the out-of-plane direction of the ferroelectric polarization. It also avoids the well-known issue of how to realistically screen free ferroelectric surfaces in finite supercells [34]. Of course, this approach does not offer information on the relaxation mechanisms, rumpling, or other surface effects, but within this paper we are mostly concerned with the interface properties, and our treatment gives insight into the LSMO/BTO interface region [6,7,30].…”

Influence of ferroelectric order on the surface electronic structure ofBaTiO3films studied by photoemission spectroscopy

“…The BTO films grown on a STO substrate have a compressive strain of 2% at the interface. As a result, the film is expected to stay in a single tetragonal phase below the ferroelectric phase transition [37,38]. Tetragonal BTO can host a ferroelectric polarization along the 001 out-of-plane as well as the 100 and 010 in-plane directions.…”

“…In films, the transition temperature is raised for compressive as well as tensile strain [35]. For tensile strain, it has been demonstrated that solely an orthorhombic phase exists below the ferroelectric transition temperature [35][36][37]. For compressive strain, the tetragonal phase [ Fig.…”

“…For compressive strain, the tetragonal phase [ Fig. 1(a)] is the only ferroelectric phase present below the transition temperature, with a preferred polarization along the outof-plane axis in films with a thickness of around of 5 unit cells (uc) [37,38]. With increasing film thickness, strain and growthdefect relaxation will be responsible for a mixture of in-plane and out-of-plane domains of the tetragonal phase.…”

Comparative ab initio calculations of SrTiO3/BaTiO3 and SrZrO3/PbZrO3 (0 0 1) heterostructures

“…The BaTiO 3 /SrTiO 3 and PbZrO 3 /SrZrO 3 (0 0 1) interfaces present an excellent possibility for developing novel materials with outstanding properties, as well as they are extremely important in studying the fundamental physics of ferroelectric materials. Despite the huge technological importance of SrTiO 3 (STO), BaTiO 3 (BTO), SrZrO 3 (SZO) and PbZrO 3 (PZO) perovskites, and numerous ab initio studies of their (0 0 1) surfaces, [1][2][3][4][5][6][7][8][9][10][11][12][13], it is hard to understand, why only a small amount of ab initio and experimental studies exist dealing with BTO/STO (0 0 1) interface [14][15][16][17][18][19]. There are no experimental studies available, and only single ab initio calculation [20] exist dealing with PZO/SZO (0 0 1) interface.…”