A physics-based modeling approach for the simulation of anomalous boron diffusion and clustering behaviors

Boron doped structures are difficult to model due to transient enhanced diffusion (TED) and defect driven clustering. The purpose of this work is to develop a new model which predicts both the defect enhanced diffusion and the defect enhanced clustering. This will enable better prediction of both threshold adjust implants and shallow boron profiles for P+ source/drain structures. This paper presents a novel approach to the development of predictive boron diffusion and clustering models based upon fundamental physical calculations performed at Lawrence Livermore National Laboratories'. A continuum model has been developed, based entirely upon these energetic calculations, which is capable of accurately simulating the diffusion, clustering and subsequent reactivation of boron for a wide variety of implant and anneal conditions. As a direct consequence of this work, a mechanism has been discovered which has been shown capable of accurately modeling the surface dose loss of boron in silicon following the annealing of damage induced by shallow boron implants.