Skip to Main Content
Due to the ongoing reduction in device geometries, the statistical properties of a few defects can significantly alter and degrade the electrical behavior of nano-scale devices. These statistical alterations have commonly been studied in the form of random telegraph noise (RTN). Here we show that a switching trap model previously suggested for the recoverable component of the negative bias temperature instability (NBTI) can more accurately describe the bias and temperature dependence of RTN than established models. We demonstrate both theoretically and experimentally, that the recovery following bias temperature stress can be considered the non-equilibrium incarnation of RTN, caused by similar defects. We furthermore demonstrate that the recoverable component is solely constituted by individual and uncorrelated discharging of defects and that no diffusive component exists. Finally it is highlighted that the capture and emission times of these defects are uncorrelated.