Octahedral tilting, monoclinic phase and the phase diagram of PZT

Anelastic and dielectric spectroscopy measurements on PbZr1-xTixO3 (PZT) close to the morphotropic (MPB) and antiferroelectric boundaries provide new insight into some controversial aspects of its phase diagram. No evidence is found of a border separating monoclinic (M) from rhombohedral (R) phases, in agreement with recent structural studies supporting a coexistence of the two phases over a broad composition range x < 0.5, with the fraction of M increasing toward the MPB. It is also discussed why the observed maximum of elastic compliance appears to be due to a rotational instability of the polarization linearly coupled to shear strain. Therefore it cannot be explained by extrinsic softening from finely twinned R phase alone, but indicates the presence also of M phase, not necessarily homogeneous. A new diffuse transition is found within the ferroelectric phase near x similar to 0.1, at a temperature T-IT higher than the well established boundary T-T to the phase with tilted octahedra. It is proposed that around T-IT the octahedra start rotating in a disordered manner and finally become ordered below T-T. In this interpretation, the onset temperature for octahedral tilting monotonically increases up to the antiferroelectric transition of PbZrO3, and the depression of T-T(x) below x = 0.18 would be a consequence of the partial relief of the mismatch between the average cation radii with the initial stage of tilting below T-IT.