Impact of the carrier distribution function on hot-carrier degradation modeling

We employ a physics-based model for hot-carrier degradation (HCD), which includes three main sub-tasks: the carrier transport module, a module describing interface state generation and a module for the simulation of the degraded devices. We examine different realizations of the model: with the transport module represented by Monte-Carlo, energy transport and drift-diffusion schemes. The main version, based on the Monte-Carlo approach, is able to represent HCD observed in different MOSFETs using the same set of the model parameters. These parameters have reliable and physically reasonable values. Therefore, we check whether two other versions are capable of the same representation (with the same parameters) or not. It appears that the simplified treatments fail to describe the degradation in devices of the same architecture but with different channel lengths employing a unique set of parameters. This circumstance suggests that a comprehensive HCD model has to be based on a rigorous solution of the Boltzmann transport equation (e.g. by means of a Monte-Carlo method).