Recently, reliable Pb (Ti, Zr) O3 (PZT) thin-film capacitors with robust switching endurance were fabricated successfully by incorporating oxidizable metal and oxide electrodes. However, the reasons for the drastic improvement in the switching endurance property was not clear. Degradation of polarizations by switching is called "polarization fatigue. " This paper describes the mechanisms of polarization fatigue, and discusses ways for improving of that property from the standpoints of microstructure, and of interactions between the PZT and the electrode materials of the capacitors. It is clearly identified that the causes of the fatigue are the unexpected formation of a surface transition layer of PZT, which is strongly dependent on the crystallization process, and a decrease in the interfacial capacitances due to the accumulation of oxygen vacancies between the PZT and non-oxidizable metal electrodes with high work functions such as Pt. Oxidizable metal and oxide electrodes suppress by oxidation-reduction reactions the accumulation of oxygen vacancies. Fatigue-free PZT thin-film capacitors can be formed if oxidizable metal or conductive oxide electrodes are incorporated in columnar-grain-structured PZT thin-film. In sharp contrast, fatigue and retention properties of PZT thin-film capacitors with Ir electrodes were degraded by modification of the PZT with 1 atm%-Nb and 1 atm%-La even though its grain structures were columnar.

This paper describes influence of the relaxation current in BaxSr(1-x)TiO3 (BST) thin films on dynamic random access memory (DRAM) operation. The relaxation current is a transient content of dielectric leakage currents. In BST thin films (expected to be a cell capacitor dielectric in 256 Mb DRAM and beyond), the relaxation current often displays the power law behavior I(t)t-1. This leads to the singularity near the time zero. When one attempts to evaluate precisely the influence of this leakage on DRAM operation, the behavior should be estimated on a time-dependent bias. However, such a singular behavior makes analysis based on a linear response difficult. In this analysis, we start by assuming that the behavior of the relaxation current can be modeled as a linear equivalent circuit. We also assume that the relaxation current follows the power law, I(t)t-1 for 1 ns