New Simulation Method to Characterize the Recoverable Component of Dynamic Negative-Bias Temperature Instability in p-Channel Metal–Oxide–Semiconductor Field-Effect Transistors

Abstract

Listen

Translate article

We introduce a new simulation technique to investigate the recovery characteristics of dynamic negative-bias temperature instability (NBTI) in conventional silicon dioxide (SiO2) dielectric p-channel metal–oxide–semiconductor field-effect transistors (p-MOSFETs) based on the hydrogen diffusion and hole-trapping mechanisms. In this work, a sequence of train pulses on the gate terminal were applied to simulated p-MOSFETs in single and multiple runs, thus emulating repetitive cycles of stress and recovery. The effects of varying the applied stress voltages, temperatures, and durations were then analyzed. The recoverable component, R, of degradation was found to increase when the magnitudes of the applied stress voltage and temperature were increased. Moreover, the R was reduced when the recovery time was increased for a single run. In contrast, the R increased when the recovery time was increased for multiple runs. The normalized R of the simulated device was found to decrease by 0.7% and to increase by 7% with respect to the shortest recovery duration for a single run and for multiple runs, respectively. In addition, we measured the effects of equivalent oxide thickness (EOT) on the R and found that, in our study, the R for a transistor with a smaller EOT exhibited a substantial increase of 84% compared with that for a transistor with a larger EOT. These characteristics of the R of dynamic NBTI could be explained from the perspectives of the reaction–diffusion (R–D) model and hole-trapping mechanism. The results suggested that the underlying connection between these two mechanisms was the interface trap concentration, which represents the permanent components in both mechanisms.