Role of point defects in bipolar fatigue behavior of Bi(Mg_1/2Ti_1/2)O₃ modified (Bi_1/2K_1/2)TiO₃-(Bi_1/2Na_1/2)TiO₃ relaxor ceramics

Lead-free Bi(Mg_1/2Ti_1/2)O₃-(Bi_1/2K_1/2)TiO₃-(Bi_1/2Na_1/2)TiO₃ (BMT-BKT-BNT) ceramics have been shown to exhibit large electromechanical strains under high electric fields along with negligible fatigue under strong electric fields. To investigate the role of point defects on the fatigue characteristics, the composition 5BMT-40BKT-55BNT was doped to incorporate acceptor and donor defects on the A and B sites by adjusting the Bi/Na and Ti/Mg stoichiometries. All samples had pseudo-cubic symmetries based on x-ray diffraction, typical of relaxors. Dielectric measurements showed that the high and low temperature phase transitions were largely unaffected by doping. Acceptor doping resulted in the observation of a typical ferroelectric-like polarization with a remnant polarization and strain hysteresis loops with significant negative strain. Donor-doped compositions exhibited characteristics that were indicative of an ergodic relaxor phase. Fatigue measurements were carried out on all of the compositions. While the A-site acceptor-doped composition showed a small degradation in maximum strain after 10⁶ cycles, the other compositions were essentially fatigue free. Impedance measurements were used to identify the important conduction mechanisms in these compositions. As expected, the presence of defects did not strongly influence the fatigue behavior in donor-doped compositions owing to the nature of their reversible field-induced phase transformation. Even for the acceptor-doped compositions, which had stable domains in the absence of an electric field at room temperature, there was negligible degradation in the maximum strain due to fatigue. This suggests that either the defects introduced through stoichiometric variations do not play a prominent role in fatigue in these systems or it is compensated by factors like decrease in coercive field, an increase in ergodicity, symmetry change, or other factors.