Advanced Micro Devices Corporation

Norman L Armour

John K Lowell

Mohammed Anjum

Maung H Kyaw

Valerie A Wenner

Julia Sherry

The present method allows detection of heavy metal impurities introduced into a silicon wafer during an ion implantation procedure. The method is quick, non-destructive, and functions in the presence of extensive lattice damage created during ion implantation. A thermal treatment follows ion implantation to cause any heavy metal impurities to diffuse into near-surface regions adjacent to major surfaces. A major surface of the silicon wafer is subjected to a high-injection SPV frequency sweep procedure before and after the thermal treatment. During each high-injection SPV frequency sweep procedure, the major surface is subjected to a train of light pulses modulated at frequencies within a frequency range of interest spanning from a low frequency cutoff (about 280 Hz) to a high frequency cutoff (about 10 kHz). Surface charge values are derived from surface photovoltages measured at each modulation frequency. Ratios of surface charge values corresponding to endpoints of the frequency range of interest are used to gauge generation lifetimes before and after the thermal treatment. A decrease in generation lifetime following the thermal treatment indicates the introduction and subsequent diffusion of heavy metal contaminants into the near-surface region adjacent to the analyzed surface.

Method of detecting heavy metal impurities introduced into a silicon

A method and device is provided for determining defects within a single crystal substrate. The methodology includes a surface photovoltage (SPV) technique in which the magnitude of non-linearity is quantified and correlated to defects within the crystal lattice. The correlation factor is determined in a rapid and efficient manner using least square correlation methodology without having to determine diffusion length and incur difficulties associated therewith. Obtaining a quantifiable least square correlation factor allows the operator to quickly determine the amount of crystalline damage often encountered by, for example, ion implantation. In addition, the operator can determine the relative depth and position of defective crystalline layers within the substrate based upon demarcations between monotonically and non-monotonically aligned points plotted in a graph of reciprocal photovoltage versus reciprocal absorption coefficient.

Growing community of inventors

Austin, TX, United States of America

Norman L Armour