“…The idea of a similar phase transition on a surface has a long history [27,28], but it was robustly confirmed, apparently for the first time, only in Ref. [29] using the ellipsometric measurements on the STO surface. The surface, a very powerful defect, imposes its own symmetry and strongly influences, in particular, the orientations of the oxygen octahedra in the first layers, resulting in rotations and surface reconstruction (and therefore surface properties) very different from the bulk behavior.…”