Low Temperature B Activation in SOI Using Optimised Vacancy Engineering Implants
- PDF / 78,852 Bytes
- 5 Pages / 612 x 792 pts (letter) Page_size
- 93 Downloads / 191 Views
E7.1.1
Low Temperature B Activation in SOI Using Optimised Vacancy Engineering Implants A.J. Smith, B. Colombeau, N. Bennett, R. Gwilliam, N. Cowern and B. Sealy Advanced Technology Institute, University of Surrey, Guildford, Surrey, GU2 7XH
ABSTRACT In this study 300keV and 1MeV Si vacancy engineering implants have been used to optimise the activation of a 2keV 1x1015cm-2 B implant into SOI. Although the two implants generate a similar areal density of excess vacancies in the SOI top layer, Hall Effect measurements show that low temperature activation is possible to a greater level with the 300keV Si co-implant than with the 1MeV implant. Hall and SIMS data are consistent with a high level of activation of the B at 700OC, with no significant diffusion at the metallurgical junction depth.
INTRODUCTION The production of highly active p-type source/drain layers is a challenge for future CMOS technology. Techniques such as pre-amorphisation have proved to increase the electrical activation of a standard boron implant to above solid solubility [1]. However, the resulting end of range damage deactivates the implanted boron when further thermal processing is required after solid phase epitaxial re-growth [2]. As Silicon On Insulator (SOI) increases in importance for future technology nodes due to its ability to overcome leakage and other issues set out in the International Technology Roadmap for Semiconductors (ITRS), it may become difficult to implement techniques such as pre-amorphisation, especially for fully depleted devices. Another technique known as Point Defect Engineering (PDE) involves using a nonamorphising high-energy silicon co-implant in conjunction with the low energy boron implant. The momentum of the impinging high-energy silicon atoms transferred to the host atoms causes a spatial separation of the net concentrations of vacancies and interstitials left after recombination of the silicon implant damage [3]. The result is a net increase in vacancies close to the surface and a corresponding net increase in Si interstitials near the projected range of the high-energy implant. This technique has been shown to reduce transient enhanced diffusion [46], however very few studies have examined the effect on boron activation [7]. Recent studies have used indirect techniques to determine the electrically active fraction, either by making assumptions about the magnitude of the mobility, or by the analysis of SIMS profile shape [8]. Performing PDE implants in SOI substrates has proved to be advantageous due to the buried oxide (BOX) acting as a perfect diffusion barrier [9], restricting the excess silicon interstitials from back diffusing into the silicon overlayer during post implantation annealing. This inherent property of SOI is in contrast to bulk silicon, where previous studies have been limited to ion energies more than 500keV, simply to gain a sufficiently large spatial separation between the two defect regions (vacancies and interstitials) so that these point defects are prevented from interacting and anni
Data Loading...
