He ϩ ions were implanted into silicon with a fluence of 5ϫ10 16 cm Ϫ2 at different temperatures ranging from 473 to 1073 K. Samples were analyzed by thermal helium desorption spectroscopy and by transmission electron microscopy. As far as cavity formation is concerned, the behavior can be divided into three stages depending on the implantation temperature. However, it is found that helium release from cavities is governed by a single mechanism regardless of the implantation temperature. © 2002 American Institute of Physics. ͓DOI: 10.1063/1.1525059͔With the miniaturization of microelectronic devices, the purity requirements of semiconductors become extremely severe. In particular, the density of metallic impurities in the active region of the devices must be extremely low. Such an impurity level can only be reached by a gettering treatment.In the last few years, work 1,2 has shown that helium induced cavities in silicon can be used as very efficient gettering sites. Cavities in silicon are usually formed by high dose He ion implantation. During subsequent annealing at temperatures above 700°C, bubbles grow and He is released from them by gas outdiffusion, leading to void formation, i.e., empty cavities. Metallic impurities can be trapped on the cavities at their internal surfaces. 1 Numerous studies have been performed varying implantation parameters 3 to optimize cavity formation. The effects of varying the implantation temperature have, however, not received much attention. In this letter, a combination of thermal helium desorption spectroscopy ͑THDS͒ and transmission electron microscopy ͑TEM͒ was used to study the effect of implantation temperature.All the experiments were performed on commercial n -n ϩ silicon wafers. The n-type layer was epitaxially grown on a ͗111͘ orientated n ϩ substrate of Czochralski silicon. The doping concentration of the 100 m thick n region was 1ϫ10 14 P cm Ϫ3 . These samples were implanted to a constant dose of 5ϫ10 16 ions cm Ϫ2 with 50 keV helium ions (Rpϭ500 nm and ⌬Rpϭ140 nm according to SRIM calculations͒. 4 The beam current was kept at 40 A. The structure of the implantation damage was studied with crosssectional TEM using a JEOL 200 CX operating at 200 kV. In order to study cavities with minimal contrast from the unavoidable accompanying lattice damage, specimens were tilted from their ͗110͘ orientation by few degrees in order to reduce diffraction effects. They were also imaged in underfocus and overfocus conditions to highlight the cavity edges with Fresnel contrast. THDS measurements were performed in an ultrahigh vacuum chamber (10 Ϫ8 Pa͒. Helium desorption was monitored using a quadrupole mass spectrometer Balzers QMG 111B while the sample was annealed with a constant heating rate of 5 K/s.The thermal desorption spectrum of helium from the sample implanted at 473 K is shown in Fig. 1͑a͒. Helium release clearly occurs partially at a low temperature from 700 to 900 K and in a high-temperature regime centered at about 1300 K with all the helium being released by 1400 K. Fo...