Mobile silicon di-interstitial: Surface, self-interstitial clustering, and transient enhanced diffusion phenomena

In the light of recent theoretical calculations predicting the mobile silicon di-interstitial (I₂) we have carried out a thorough analysis of its implications within a self-consistent, comprehensive modeling framework using atomistic Monte Carlo. In the study presented here, we investigate its role in surface recombination and self-interstitial clustering processes, including boron transient enhanced diffusion under conditions where no boron clustering occurs. We conclude that, under such conditions, the mobile I₂ does not play a critical role and the single self-interstitial can be made to account for the missing mobile I₂. The main requirement for introducing the mobile I₂ is a bottleneck of two consecutive, relatively unstable, self-interstitial cluster sizes. We demonstrate that the different surface recombination lengths measured under oxidation/implantation conditions cannot be attributed to different recombination lengths for single interstitial and di-interstitial.