Broadband dielectric spectroscopy from Hz up to the infrared (IR) range and temperature interval 10-300 K was carried out for xBaZrO3-(1-x)BaTiO3 (BZT-x, x = 0.6, 0.7, 0.8) solid solution ceramics and compared with similar studies for x = 0, 0.2, 0.4
, 1 ceramics published recently (Phys. Rev. B 86, 014106 (2012)). Rather complex IR spectra without appreciable mode softening are ascribed to Last-Slater transverse optic (TO) phonon eigenvector mixing and possible two-mode mixed crystal behavior. Fitting of the complete spectral range requires a relaxation in the 100 GHz range for all the samples. Below 1 GHz another relaxation appears, which is thermally activated and obeys the same Arrhenius behavior for all the relaxor BZT samples. The frequently reported Vogel-Fulcher behavior in BZT relaxors is shown to be an artifact of the evaluation from the permittivity or loss vs. temperature dependences instead of its evaluation from loss vs. frequency maxima. The relaxation is assigned to local hopping of the off-centered Ti4+ ions in the frozen BTO clusters, whose size is rather small and cannot grow on cooling. Therefore BZT is to be considered as a dipolar glass rather than relaxor ferroelectric.
We performed factor-group analysis of all phonons in possible monoclinic C2/c and C2 structures of BiMnO3 and compared it with our experimental infrared and Raman spectra. We conclude that the crystal structure is centrosymmetric C2/c in the whole in
vestigated temperature range from 10 to 550 K, therefore BiMnO3 cannot be ferroelectric. We revealed a dielectric relaxation in THz spectra above the structural phase transition taking place at T_C1=475 K giving evidence in strong lattice anharmonicity and a large dynamical disorder of Bi cations above T_C1. Step-like dielectric anomaly observed at T_C1 in THz permittivity reminds antiferroelectric phase transition. Nevertheless, the low-temperature dielectric studies did not reveal any antiferroelectric or ferroelectric hysteresis loop. Our experimental results support theoretical paper of P. Baettig et al. (J. Am. Chem. Soc. 129, 9854 (2007)) claiming that BiMnO3 is not multiferroic, but only antipolar ferromagnet.
