Junsuke Tomioka

Formation process of grown-in defects in Czochralski grown silicon crystals

Abstract We have determined the set of the diffusion coefficients ( D v : vacancies, D i : self-interstitials) and equilibrium concentrations ( C eq v : vacancies, C eq i : self-interstitials) of point defects which has been satisfied with the dependence of two-dimensional defect patterns on V and G (V: growth rate, G: axial temperature gradient) and with the reported product values of D v C eq v and D i C eq i . Recently, it has been reported by direct TEM (Transmission Electron Microscopy) observation that the grown-in defects in the vacancy dominant region are the voids of octahedral shape. The idea that the grown-in defects are the voids formed by the vacancy aggregation has been examined by the simulation model. It is shown that this model can well describe the behaviors of grown-in defect formation during the crystal growth and that it is possible to form the void as the grown-in defect in CZ silicon crystals.

Simulation of point defect distributions in silicon crystals during melt-growth

Simulations of the point defect diffusion during the crystal growth process are reported for the investigation of the relationship between the distribution of the oxygen precipitation and point defect concentration in CZ silicon crystals. Distribution of the oxygen precipitation is strongly affected by the point defect concentrations during the growth process. There have been many proposed models for the point defect diffusion during the crystal growth process, and the correspondence between point defect concentrations and grown-in defects has been extensively investigated. In this paper, we have compared the distribution of the oxygen precipitation and the ratio of super-saturation between vacancies and self-interstitials. The experimental results agreed with the results obtained by calculation.

Effect of Oxygen Precipitates on the Surface-Precipitation of Nickel on Cz-Silicon Wafers

This paper presents a model for the analysis of the surface nucleation and growth of Ni silicide on silicon wafers contaminated by Ni. The model can additionally be used to characterize the gettering reaction of Ni induced by oxygen precipitates. We also discuss the relation between the surface precipitation of Ni silicide and the gettering ability of oxygen precipitate. The surface precipitation of Ni silicide depends on the total surface area of oxide precipitates. When the total surface area of the oxide precipitates exceeds the critical value, the surface precipitation is rapidly suppressed. Our model can explain the phenomenon of the gettering threshold in the following manner. 1) The gettering of Ni by oxygen precipitates is a reaction-limited process at the interface between oxygen precipitate and silicon, as Sueoka proposed. 2) The residual Ni concentration in this reaction-limited gettering process continuously decreases as the total surface area of the oxide precipitates increases. 3) The surface precipitation of Ni silicide is rapidly suppressed when the residual Ni concentration falls below the critical concentration. Our calculation results correspond well with the experimental results.

Analysis of Internal Gettering of Iron Based on the Nucleation Model of Iron Precipitation

We propose a new simulation model to describe the behavior of the internal gettering of Fe by oxygen precipitates. The internal gettering of Fe has typically been described by the diffusion-limited model proposed by Gilles et al. While the diffusion-limited model is effective in describing the gettering behavior at low temperature (< 400°C), it fails to do so at practical gettering temperature of 600°C and above. The model proposed here accurately describes the internal gettering behavior of Fe at both low and practical temperatures by considering the nucleation of iron silicide on the oxygen precipitates. Our model also explains how the gettering behavior is influenced by the product of NRo at low temperature and by the product of NRo 3 at high temperature, where N and Ro are the density and radius of the oxygen precipitates, respectively.