Abstract
In the soft mode picture, the ferroelectric phase transitions in most of ABO3 perovskites are triggered by the long wavelength oscillations of the B ions against their oxygen cage. There is now a lot of experimental evidence that this picture is incomplete. An extra contribution has to be included in the dynamical response. This contribution is usually thought to arise from local lattice distorsions; this is what will be called local disorder. A complete understanding of this local disorder requires the experimental determination of the microscopic origin, of the dynamics and spatial extension of the lattice distorsions.
A useful way to summarize these requirements is to use the concept of clustering. Although this is a rather old concept, it is shown in this paper that the continuous improvement of both the crystals quality and experimental devices led to new developments in the last ten years. The key examples are pure and Fe doped BaTiO3. pure and Nb, Na, Li doped KTaO3.
The microscopic origin of the clusters in pure BaTiO3 and KTaO3 is still an open question.
In the case of doped KTaO3, the random interaction between the impurity induced clusters may result in a polar glass phase. Even when the clusters do not percolate, they may lead to some unusual properties.
Key Words: