Self-interstitials have never been observed in crystalline Si. How about amorphous Si?
- PDF / 116,671 Bytes
- 5 Pages / 612 x 792 pts (letter) Page_size
- 83 Downloads / 195 Views
Self-interstitials have never been observed in crystalline Si. How about amorphous Si? Sjoerd Roorda Département de physique et Groupe de recherche en physique et technologie des couches mines, Université de Montréal CP 6128 succursale Centre-ville Montréal, H3C 3J7 CANADA Abstract In the early days of point defect studies in electron irradiated crystalline silicon, it was surmised that the Si self-interstitial is highly mobile even at 4 K and escapes direct detection. The existence of self-interstitials has of course been confirmed through the diffusion behaviour of a range of impurities and the direct observation of larger interstitial-type clusters. Against this background, the direct observation of self-interstitials in amorphous Si would seem next to impossible. Yet just such an observation may have been made recently, through a comparison of the high-resolution radial distribution function of pure amorphous Si before and after thermal anneal and that of crystalline Si. Introduction The structure of pure amorphous silicon is believed to approach that of a perfect, fully connected, four-fold coordinated, continuous random network. This ideal cannot be obtained, much like that is the case for crystalline materials which are never as ideal as the perfect, defect-free crystal. Real amorphous silicon, therefore, would contain deviations from the ideal random network. These deviations can be defects that are typical for a random network, such as density variations, 5- or -7 member rings or defects much like those encountered in crystalline silicon, including vacancies and interstitials. It has been argued that removal of such defects, which requires atomic mobility just like defect annihilation in crystal silicon does, is in fact the origin of the phenomenon known as structural relaxation [1]. Considerable experimental evidence has been accumulated in support of the notion that vacancytype defects do indeed exist in amorphous silicon. These include the solubility and diffusivity of metal impurities in amorphous silicon [2], the kinetics, energetics, and temperature dependence of structural relaxation [1], and the trapping and annihilation behaviour of positrons in amorphous silicon [3]. However, all of these measurements are in one form or other indirect and, moreover, none of these address the possible existence of interstitial type defects. In this paper, I will discuss recent experimental results [4, 5] and how they may give direct evidence for the existence of vacancies and interstitials in amorphous silicon and their role in structural relaxation. Amorphous silicon by ion implantation: pure, void-free and non-compactable The recent experimental data in question consists of high resolution x-ray diffraction data of pure amorphous silicon, prepared by ion implantation and, in some cases, thermal annealing. These A16.1.1
data have recently been published including a detailed description of the experimental conditions [5] and sample preparation [6] which will not be repeated here. Before discussing these data and
Data Loading...
