Fresnoite, Ba2TiOSi2O7, has been predicted structurally to be a new ferroelectric. Detection of both ac and dc dielectric hysteresis verifies the prediction. The spontaneous polarization Ps estimated from the hysteresis at 1.2 MV m−1 ac is ∼0.2 C m−2 at 295 K, comparable with the minimum Ps observed in one-dimensional ferroelectrics. A reproducible calorimetric anomaly with entropy change 0.19(3) J mol−1 K−1 at 433(2) K in polycrystalline material coincides with a small dielectric and pyroelectric anomaly previously reported in single crystals; an entropy change ∼0.2 J mol−1 K−1 at 810(5) K also accompanies a dielectric anomaly observable in ceramic samples at 805(5) K. Both calorimetric anomalies are ∼60 K wide. Observation of dielectric hysteresis at 875 K shows that neither anomaly corresponds to the Curie temperature; both are likely associated with small changes in atomic position, not with symmetry changes. Melting onset in fresnoite is 1703(1) K with undercooling as deep as 435 K.

Ba6CoNb903o and Ba6FeNbgO3o in space group P4bm are shown to satisfy the structural criteria for ferroelectricity. Ba6CoNb903o undergoes a diffuse phase transition at 660 (11) K, as observed calorimetrically, in addition to a dielectric permittivity anomaly with an onset temperature of 685 (10) K. The demonstration of dielectric hysteresis at room temperature under the application of a varied DC field reaching a maximum of + 300kVm-l, corresponding to a spontaneous polarization of 1.2 (5) x 10-2C m-2, provides unambiguous verification that it is a new ferroelectric. Ba6FeNb9030 also undergoes a diffuse phase transition at 605 (16) K, with a dielectric anomaly at 583(5)K, and exhibits dielectric hysteresis at room temperature under a varied DC field ranging to + 310 kV m-1 corresponding to a spontaneous polarization of 2.2(5) x 10-2Cm-2; it too is demonstrably a new ferroelectric. Although Bas.2Ko.8U2.nNb7.603o has also been reported in space group P4bm, all atomic displacements from the corresponding centrosymmetric positions are less than their refined root-mean-square thermal or static amplitudes. Such an arrangement is likely to be thermodynamically unstable. Either its space group has been incorrectly assigned, and reinvestigation will show the space group is P4/mbm, or the structural refinement is incomplete.

Sr 2 SbMnO 6 is predicted to be ferroelectric with Curie temperature Tc=450(175) K on the basis of a 1990 structure determination. Calorimetry reveals a stepwise heat capacity change at Tc=431(5) K, indicative of a phase transition close to first order. A further small calorimetric anomaly occurs reproducibly at 743(10) K, consistent with an earlier report of a higher-temperature phase transition to cubic symmetry. The dielectric permittivity ε′ increases sharply at Tc=466(10)K in the range 0.1–100 kHz; ε′ rises ∼3 orders of magnitude at 0.1 kHz, ∼1.0 order at 100 kHz, before a stable maximum is reached at T>Tc. The conductivity σ of Sr2SbMnO6 at 295 K is 0.23(5) S m−1, too high for dielectric hysteresis measurement. The linear dependence of ln σ on 1/T between 295 K and Tc in this semiconductor ferroelectric corresponds to a band-gap EG=0.50(2) eV; above Tc, EG increases abruptly to 0.70(2) eV.