Time-Dependent 3-D Statistical KMC Simulation of Reliability in Nanoscale MOSFETs

Abstract

This paper presents a thorough numerical investigation of statistical effects associated with charge trapping dynamics and their impact on the reliability projection in decananometer MOSFETs. By means of a novel 3-D kinetic Monte Carlo TCAD reliability simulation technology, we track the time-dependent variability associated with granular charge injection and trapping into the oxide traps. We consider the interactions between the statistical variability of the virgin transistors, introduced by the discreteness of charge and granularity of matter, and the stochastic nature of the trap distribution and the trapping process itself. As a result, the path to device failure (PtDF), defined as the stochastic succession of trapping events that bring the device parameters to a predefined failure criteria, is analyzed in detail. In particular, we show that the two stochastic components determining the PtDF, namely the traps' capture time constants and threshold voltage shifts, are uncorrelated. The charge injection variability is shown to play a dominant role in determining the statistical dispersion of the reliability behavior. Furthermore, we show that the short- and long-term reliability behaviors are uncorrelated. Finally, 3-D fringing and percolative effects are shown to play an important role in determining the statistical degradation of nanoscale MOSFETs.