| The first part of this thesis presents a theory of dielectric relaxation current, which is a fundamental nonideality in the response of glass capacitors to step voltages. The theory is a semiclassical treatment of the double well model traditionally used to describe the properties of glasses at low temperature that are anomalous compared to those of perfect insulating crystals, the classic example being a specific heat capacity varying linearly with temperature.; Dielectric relaxation is of technological interest, as the relaxation of the gate dielectric in a metal-insulator-semiconductor transistor can cause a time-dependent shift in the threshold voltage of the transistor. The next part of the thesis presents room-temperature measurements of dielectric relaxation current for the high-k gate dielectrics hafnia (HfO 2) and zirconia (ZrO2), and it is shown that both materials produce dielectric relaxation currents more than an order of magnitude larger than that of silicon dioxide (SiO2). The threshold voltage shifts calculated from the dielectric relaxation current are on the order of 10 to 30 mV per decade in time, compared to less than 1 mV per decade for SiO 2, making HfO2 and ZrO2 seem unsuitable for use as gate dielectrics.; A later chapter of the thesis shows that the theory described above also gives a straightforward account of other phenomena observed in glass capacitors, including (i) an AC capacitance with a power-law dependence on frequency; (ii) a time-dependent change in AC capacitance following the application of a DC electric field or mechanical strain, and (iii) a dielectric memory effect. In addition, phenomenon (ii) is shown experimentally to occur at room temperature for HfO2, whereas it has previously been observed only at very low temperature.; The final chapters of the thesis concern two topics related to high- k dielectrics but unrelated to the theory of glasses. The first topic is how dielectric susceptibility in thin films and interfaces differs from that in bulk materials, and the second topic is the electronic structure of hafnium and zirconium silicates. Both topics are addressed theoretically using tight-binding theory. |
