Application of molecular dynamics for low-energy ion implantation in crystalline silicon

Abstract

Molecular dynamics (MD) is set to replace Monte Carlo (MC) methods utilizing the binary collision approximation (BCA) in modeling dopant distributions after ion implantation in the low energy regime. Simultaneous nonbinary interactions come into play as the ion slows down; unlike BCA, MD automatically accounts for multiple collisions between ion and its neighboring atoms. In this work, the energy limit below which BCA fails is estimated from density functional theory (DFT) calculations for a wide range of dopants. Impurity profiles are generated using the MD code, MDRANGE. A database consisting of secondary ion mass spectrometry (SIMS) profiles covering a wide range of dopants (B, C, F, N, P, As, Ge, In, and Sb) over the energy regime of 0.5-10 keV at critical channeling directions have been set up. The MD simulated profiles show good agreement with SIMS data, which have been obtained either with a quadrupole-or magnetic-sector-based mass spectrometer. © 2006 American Vacuum Society.