Noise-based prognostic design for real-time degradation analysis of nanodevice dielectric breakdown

Prognostics and system health management (PHM), which is an upcoming and potential area of research aimed at analyzing system performance degradation and estimating the time to failure at operating conditions using sensor-based tracking of diagnostic degradation indices in real-time, has been widely applied to various macro electronic and mechanical systems in recent times [1, 2]. However, its application not only spans these macro systems, but can also be effectively applied to nano-systems, as we illustrate in this work. Though initial studies have been carried out to apply PHM to high power field effect transistors [3], an in-depth investigation into its potential for nano-reliability studies is still lacking. This serves as a motivation for us to present a holistic study on the time-dependent dynamic degradation and failure prediction of metal-oxide-semiconductor field effect transistors (MOSFET) using random telegraph noise (RTN) fluctuations that can be measured when the device leakage current is sensed with a nominal voltage during the operation of the transistor / circuit. We will first present the algorithm used to track the degradation of the transistor and highlight the use of RTN signals and their analysis in the time and frequency domains. This will be followed by an experimental case study on a very small area transistor, illustrating how the RTN signals can be used to diagnose the degradation of the device. Finally, we conclude with potential suggestions for further research into this topic in the near future.