Silicon damage studies due to ultra-low-energy ion implantation with heavy species and rapid thermal annealing

Abstract

There have been many studies of electron volt implants of B + into silicon. This focus on boron was due to enhanced diffusion phenomena and the particular difficulty in constructing high conductivity, very shallow layers with B + light ion. We have also analysed some features of high-dose implantation of As + to form n + layers with shallow junction depths (30–40 nm) with carrier concentration of >IE20 cm −3. Some of our heavy ion work is presented here. We characterised the surface damage region (SDR) and identified several non-linear phenomena. High-conductivity layers of 150–300 Ω/square can be made with rapid thermal annealing. The critical limiting factors are range shortening, sputtering and out-diffusion. The range shortening is evident in the saturation behaviour and the out-diffusion is seen as a build up of non-substitutional arsenic in the oxide or the oxide–silicon interface after annealing. We have used Rutherford backscattering (RBS), medium-energy ion scattering (MEIS) and high-resolution transmission electron microscopy (HRTEM) to study crystal micro-structure and damage, as well as secondary ion mass spectrometry (SIMS), spreading resistance profiling (SRP) and sheet resistance methods to study both the diffusion and activation of the dopant. We have observed new features in the diffusion profile with various implant temperatures and offer some explanations for this behaviour. We have also studied Sb + and In + implants because they are becoming increasingly important, at moderate energies, for hyper-abrupt channel and channel engineering with controlled lateral diffusion. Models have been developed to describe the non-linear behaviour of heavy ion doping at low energies and our results demonstrate that implants in the energy range 300 eV–2.5 keV can provide solutions when combined with short rapid thermal anneals for the manufacture of very shallow junctions with high activation of dopant. Interestingly, implants at room temperature do not produce the best results.