Multiscale Modeling of the Implantation and Annealing of Silicon Devices

Abstract

Process simulators for silicon devices are benefiting from recent work on several models that cover vastly different length and time scales. In this talk we will focus on molecular dynamics (MD) and Monte Carlo (MC) models. We will describe these models and the applications of the results to commonly used simulators such as SUPREME-IV and PROPHET. Improvements in these continuum simulators are presently being implemented by the introduction of mechanisms and energetics derived from the more fundamental approaches. We will discuss MD methods, which can provide some of the parameters needed for the rates of diffusion, recombination and other phenomena modeled by the simulators. The MC model is less detailed and can simulate entire implantation and diffusion cycles. It is also atomistic in scale, and gives results on the dopant density, clusters and other properties. In addition, the model includes stochastic effects of implantation and diffusion, which are of increasing importance as device components are scaled down in size.