The purpose of this work is twofold: (i) to study the damage induced by ion implantation, with special attention to low implanted doses; (ii) to study the efficiency of annealing techniques — particularly incoherent light annealing — in order to relate the electrical activity of implanted atoms to damage annealing. We have used three methods to study the damage induced by ion implantation: (1) RBS (or nuclear reactions) in random or in channeling geometry (2) RX double crystal diffractometry and (3) electrical measurements (free carrier profiling). Damage induced by silicon implantation at doses >10 14at/cm 2 can be monitored by all three techniques. However, the sensitivity of RBS is poor and hence this technique is not useful for low implantation doses. As device technology requires dopant levels in the range of 5 × 10 12 atoms/cm 2, we are particularly interested to the development of analytical techniques able to detect the damage at this implantation level. The sensitivity of such techniques was checked by studying homogeneously doped (5 × 10 16 e −/cm 3) and semi-insulating GaAs samples implanted with 3 × 10 12 silicon atoms/cm 2 at 150 keV. The substrate temperature during implantation was 200°C. The damage produced in these samples and its subsequent annealing are evidenced by strong changes in X-ray double crystal diffraction spectra. This method hence appears as a good monitoring technique. Annealing of the implanted layers has been performed using incoherent light sources (xenon lamps) either in flash or continuous conditions. Reference samples have also been thermally annealed (850°C, 20 min in capless conditions). The results are compared, and the electrical carrier profiles obtained after continuous incoherent light irradiation indicate that the implanted silicon atoms are almost dully activated. The advantages and disadvantages of incoherent light irradiation are discussed (surface oxidation, surface damage) in comparison with standard thermal treatment.
Read full abstract