Carbon in Crystalline Silicon

  • PDF / 968,947 Bytes
  • 15 Pages / 420.48 x 639 pts Page_size
  • 37 Downloads / 169 Views



CARBON IN CRYSTALLINE SILICON RONALD C. NEWMAN J. J. Thomson Physical Laboratory, University WhIteknights, Reading. RG6 2AF, UK

of Reading,


Box 220.

ABSTRACT the Early measurements are reviewed relating to the solubility. the infrared precipitation, Internal diffusion. coefficient, segregation absorption and the effect on the lattice parameter of silicon due to substitutional carbon. Interactions with oxygen and group III Impurities are mentioned. A survey is given of the effects of high energy irradiation and heat treatments on both FZ and CZ silicon containing carbon. A rich variety but in general atomic structures have not been of defects can form, determined. The process of site switching from substitutional to interstitial locations which occurs by the selective trapping of self-Interstitials is proving to be invaluable in elucidating the mechanisms of various defect reactions, particularly those Involving oxygen Impurities. INTRODUCTION AND HISTORICAL SURVEY At about the time that silicon was first analysed chemically to determine its oxygen content, related measurements were carried out which impurity might also be present in a high suggested that carbon 3 (see for example Papazian and concentration of up to 1019 atom cmWolsky [1]). It is almost certain that such high estimates exceeded the true concentrations, probably because of contamination Introduced during the analytical procedures adopted, but nevertheless they stimulated other work which Is continuing to the present day. Somewhat later Newman and Wakefield [2] showed that heat treatments of single crystal silicon in vacuum produced surface coatings of cubic $-silicon carbide particles which were detected by reflection electron diffraction and electron microscopy using extraction replica techniques. These observations could of course be explained away by invoking a chemical reaction between the silicon and organic contamination in the ambient. However it was pointed out that It was necessary to consider the possiblity that carbon dissolved in the silicon could have diffused to the external surface to contribute to the coating. Quantitative estimates of this process were not then possible because neither the carbon content of Work by Hall samples nor the diffusion coefficient of carbon were known. (3], Dash (4] and Scace and Slack (5], indicated that silicon crystals might 3 The dissolution of SIC Into contain up to 1018 atom cm- of the Impurity. an oxygen-free silicon melt led to an estimate of the solubility of carbon in 18 the liquid of about 3 x 10 cm-3, and It was assumed that the segregation It was Imperative to gain further coefficient might be around 0.3. As a next step. It was shown by Newman (6] using electron Information. diffraction techniques applied to specially etched and doped crystals that precipitated particles of 2-SiC could also grow along the cores of grown-in 0 If the carbon had diffused dislocations during heat treatments at 1150 C. from the surrounding matrix it was again implied that the concentration 3 would have had to