Metal contamination detection in nickel induced crystallized silicon by spectroscopic ellipsometry

Abstract

In this paper a new approach is presented for the simulation of spectroscopic ellipsometry (SE) data to estimate the level of nickel (Ni) contamination in silicon crystallized by metal induced crystallization (MIC). The method employs the addition of Ni as reference for a Bruggemann effective medium approximation (BEMA) to simulate the optical response of the crystallized silicon. This new approach is sensitive to changes in the initial average metal thickness used on the crystallization process to thickness values as low as 0.05 nm. This corresponds to a volume fraction of 0.24%, confirmed by Rutherford backscattering spectrometry (RBS) where it was observed that the Ni volume fraction detected by SE varies linearly with the metal amount inside the crystallized films determined by RBS.