Modeling and control of current inrush in PTCR barium-lead titanate

The research described here explores the relationships between inrush current and processing conditions in positive temperature coefficient of resistance (PTCR) barium-lead titanate (Ba:Pb = 0.7:0.3) doped with 0.2 mole % CeO;A thermopower measurement indicated that the cerium-doped barium lead titanate is an n-type semiconductor. This result, along with ion size considerations, led to the conclusion that at least some of the cerium exists in the trivalent state on barium sites, as do other rare earths such as La in barium titanate. However, the ratio of apparent carrier concentration to Ce concentration at room temperature is less than 1%, suggesting the coexistence of Ce atoms substituting on the Ti sublattice.;Scanning Auger microprobe line scans showed that titanium and cerium segregated to the grain boundaries of the sintered material. Further point analysis across the grain boundaries to obtain a relative concentration ratio of Ba, Ti, and Ce revealed that the grain boundaries are barium deficient.;An equivalent circuit model, incorporating the temperature-dependence of the resistance and capacitance of the grain and grain boundary, was used to predict current versus time (I(t)) characteristics. The input values for the circuit elements were based on data from capacitance and AC impedance measurements. Comparison of calculated and measured I(t) curves gave excellent agreement, supporting the use of the model as a basis for device design with a goal of reduced current inrush.;The model predicts that high overall resistance and low capacitance will reduce the inrush current. It was also concluded that the slope of the NTCR region significantly affects the average rate of inrush. The experimental results indicate that high sintering temperature and slow cooling rate raise the resistivity and lower the capacitance. The I(t) characteristics of the samples processed in this way showed that these processing changes led to reduced current inrush, as predicted by the model.