From Defect Analysis to Gate-Level Fault Modeling of Controllable-Polarity Silicon Nanowires

Abstract

Controllable-polarity silicon nanowire transistors (CP-SiNWFETs) are among the promising candidates to complement or even replace the current CMOS technology in the near future. Polarity control is a desirable property that allows the online configuration of the device polarity. CP-SiNWFETs result in smaller and faster logic gates unachievable with conventional CMOS implementations. From a circuit testing point of view, it is unclear if the current CMOS and FinFET fault models are comprehensive enough to model all the defects of CP-SiNWFETs. In this paper, we explore the possible manufacturing defects of this technology through analyzing the fabrication steps and the layout structure of logic gates. Using the obtained defects, we then evaluate their impacts on the performance and the functionality of CP-SiNWFET logic gates. Out of the results, we extend the current fault model to a new a hybrid model, including stuck at p-type and stuck-at n-type, which can be efficiently used to test the logic circuits in this technology. The newly introduced fault model can be utilized to adequately capture the malfunction behavior of CP logic gates in the presence of nanowire break, bridge, and float defects. Moreover, the simulations revealed that the current CMOS test methods are insufficient to cover all faults, i.e., stuck-Open. We proposed an appropriate test method to capture such faults as well.