Abstract
Ferroelectric materials have established themselves as indispensable in key applications such as piezoelectric transducers and energy storage devices. While the use of ferroelectrics in these fields dates back more than 50 years, little progress has been made to extend applications of ferroelectrics into new fields. To a large extend the observed slow progress is not caused by a lack of potential applications, but to by the inherent complexity associated with a structural phase transition, combined with strong coupling of polarisation, strain and temperature, and the strong modification of the phenomena by material defects. This thesis takes a look at prospective applications in energy storage for pulse power applications, solid state cooling and non-volatile random access memory and identifies key issues that need to be resolved. The thesis delivers time-domain based approaches to determine ferroelectric switching behaviour of bulk materials and thin films down to sub-ns time scales. The approach permitted study of how information written to a ferroelectric memory decays as a result of multiple non-destructive read operations. Furthermore simultaneous direct measurements of temperature and ferroelectric switching established a direct link between the retarded switching phenomenon observed in ferroelectrics and temperature changes brought by the electrocaloric effect. By comparison with analytical models and numerical simulation a large localised temperature change on the scale of individual domains is postulated. It implies a much larger coupling between switching and local temperature than has been previously considered. In extension of the model the frequency dependence of polarisation fatigue under bipolar conditions is explained by the occurrence of large temperature gradients in the material.