Impedance characterisation of conducting materials
Impedance spectroscopy is a powerful technique for characterising electrical microstructure of materials and has been increasingly useful in material science. The characteristic equivalent circuit elements give useful information on the nature of materials under study. This thesis describes impedance characterisation of conducting materials using non-Debye circuits including constant phase elements (CPE) as a superior data processing method to the conventional Debye approach which is customarily used for impedance characterisation. First, impedance spectroscopy is interpreted by inclusion of CPE components into the Debye circuits. The response was simulated in several impedance formalisms considering many possible changes in parameters. Most responses are totally different from the Debye response and cannot be interpreted in the same manner. Some key features are explained using characteristic parameters in Almond-West theory. Impedance data for yttria-stabilised zirconia with variable Y2O3 content were fitted well with circuits including CPE elements. The non-Debye character of the bulk and grain boundary components was successfully interpreted in terms of the circuits simulated previously. However, some samples showed small, but significant deviations from these approximate circuits. The discrepancy was ascribed to the existence of an additional capacitance, resulting from dipole reorientation. The data were fully fitted using the modified bulk circuit. It is also found that multi-impedance formalism analysis is the best method to obtain the most reliable equivalent circuits and fitted parameters. The conductivity data are analysed with Almond-West theory. The equality of activation energy for conduction and hopping suggests that V"O can contribute to dc conduction using direct hops among Y' defects. This energy can be represented as the sum of the dissociation energy needed to break the Coulomb interaction and the activation energy of local V'o migration. Impedance data of lithium disilicate glass were also examined using the non- Debye approach. The data for the bulk component fit well with the circuit including CPE. Annealed samples showed larger activation energy than that for an as-quenched sample. Using Almond-West theory, the difference is directly attributed to increase in the activation energy for hopping, which is explained by narrowing of the conduction pathway for the conducting cation in the annealed sample; this may be caused by reorganisation of the anion framework with decreasing volume while annealing. Barium titanate ceramics were chosen as the third example. The Debye approach is discussed first with the impedance data above the Curie temperature, Tc. The analysis showed that, in conventional fixed frequency measurements, grain boundary impedances influence the Curie-Weiss plots in two ways: at high temperatures, they increasingly dominate the permittivities; at lower temperatures, closer to Tc, the permittivity contains a contribution from grain boundary effects. Methods for extraction of bulk and grain boundary permittivities using impedance spectroscopy are discussed and the importance of selecting the appropriate equivalent circuit to model the impedance response is stressed. A constriction impedance model for the grain boundary in BaTiO3 ceramics is proposed from the temperature variation of grain boundary capacitance. The grain boundary is ferroelectric, similar to the grains, but its impedance is modified by high impedance electrical inhomogeneity in the region of the necks between grains; the activation energy of the constriction grain boundary impedance differs from that of the bulk, suggesting differences in defect states or impurity levels. The analysis for the same BaTiO3 impedance data is extended to a non-Debye approach. The non-Debye character of the bulk response is successfully analysed using CPE elements with manual fitting while the computer fitting was of limited success for both the bulk and grain boundary responses. The reason is ascribed to data limitation and additional, unknown bulk response appeared at high frequencies; the further development and analysis of BaTiO3 impedance data is a future problem.