Abstract
This study investigated the similarities and differences between B 1 monomer and B n cluster ion implantation into silicon. Small polyatomic boron ions (Bn-, n = 1-4) with the same atomic boron kinetic energy (20 keV/atom) and atomic fluence (5 × 10 13 atoms/cm 2 ) were used. In the simulation, the widely-used SRIM computer code was employed to calculate the as-implanted boron and damage depth profiles of B 1 monomer ion implantation in order to make comparisons with experimental results. In the experimental one, the B 1 monomer and B n cluster ions extracted from a tandem accelerator were used to perform ion implantation. Post-annealing methods included one-step (RTA) and two-step (FA + RTA) treatments, where RTA denoted high-temperature rapid thermal annealing at 1050 °C for 10 s and FA represented low-temperature furnace annealing at 550 °C for 1 h. The results revealed that all four as-implanted range parameters (average range, longitudinal range straggling, skewness, kurtosis) increase and tend to saturate as the cluster size increases when compared to those of SRIM-calculated results for the B 1 implant. Furthermore, the peculiar damage structures produced by different B n cluster ions lead to various behaviors in both diffusing and activating boron atoms. © 2004 Elsevier B.V. All rights reserved.