Quick counting and monitoring method of silicon substrate chip dislocation defects

Abstract

The invention discloses a quick counting and monitoring method of silicon substrate chip dislocation defects. The method comprises the steps of providing a silicon substrate chip sample needing analysis and manufacturing an analysis section. The analysis section is subjected to corrosion processing through crystal defect corrosive liquid, and is subjected to SEM analysis. The method increases the observation range of the dislocation defects on a silicon substrate, and can improve analysis speed of the dislocation defects, improve dislocation monitoring and analysis efficiency and lower cost.

Description

technical field [0001] The invention relates to a semiconductor integrated circuit manufacturing process method, in particular to a fast statistical monitoring method for silicon-based chip dislocation defects. Background technique [0002] There are two commonly used analysis methods for dislocation defects in silicon-based chips: [0003] The first method is to first prepare a sample with a thickness of about 100nm at a fixed point using a focused ion beam electron microscope (Focused Ion Beam, FIB), and then use a transmission electron microscope (Transmission electron microscope, TEM) to observe the structure of the prepared sample. [0004] The second method is to first prepare a thin slice sample with a thickness of about 100nm by using an ion thinner, and then use TEM to observe the structure of the prepared sample. [0005] The FIB sample preparation in the existing first method is: [0006] Use the focused gallium positive ion beam as the incident particle (or pri...

AI Technical Summary

Problems solved by technology

[0008] 1. It takes a long time, and the preparation of the most complex single sample is calculated in hours

[0009] 2. The observable area is small. The length of a single sample is usually about 10 microns, and the depth is 2-4 microns. The thickness corresponds to the width of the sample. Even if it takes a long time to prepare the sample, the observable sample area is only in the dozens of square microns

[0010] 3. The overall efficiency is low

[0016] 3. The overall efficiency is still low

Images