Research Of Silicon Nanowire Transistor Threshold Voltage Extraction Technology

In the past forty years, the integrated circuit (IC) technology has achieved great success. As the device’s size on IC scales down continuously, the silicon plane technique will trend to its limit. A lot of research work has been developed, on which the substitution technology of silicon plane technique has become the final goal. In the thesis, our research work focus on the performance of silicon nanowire transistor. Combined with Savitzky-Golay (SG) filtering technique, the threshold voltage extraction process has been analyzed and discussed.The on-state currents of nanometer metal-oxide semiconductor field-effect transistor (MOSFET) become very small, which results in the obvious fluctuations of measured data from the noise of the IC device or the measurement system. These fluctuations will make the extraction of parameter difficult. After introducing SG filtering method to the threshold voltage extraction, the data can be smoothed to a certain extent, and the extraction process will be stable and easy to operate.In the smoothing process of the Id-Vg or gm-Vg curve, a criterion of the minimum mismatch rate was used to determine the best smoothing window and to extract the threshold voltage accordingly. The SG filtering algorithm with mismatch rate criterion is very suitable for the extraction of threshold voltage of nano-MOSFET.During the application of the second-derivative method, we analyze the relationship between selected values of threshold voltage and filter window length. Using the mismatch rate criterion, the best fitting curve could be found and the stable threshold voltage parameter could be extracted.This thesis also studied the relations between the threshold voltage and channel length, channel widths, temperatures and substrate biases for the silicon nanowire transistor. This technique will be in favor of analyzing the characteristics of nano-MOSFET and designing the integrated circuits.