Semiconductor Pattern Defect Detection via Brightness Index
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for inspecting defects in semiconductor integrated circuits using voltage contrast are unreliable due to variations in brightness of non-defective products and inability to detect defects based on pattern brightness in die-to-database comparisons.
Innovation Solution
A pattern defect detection method that generates images of semiconductor specimens using a scanning electron microscope, extracts and compares target patterns with reference patterns from design data, calculates brightness index values, and determines standard ranges to accurately detect defects by classifying patterns into groups based on shape and connectivity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If design data is used to generate reference images for defect inspection, then inspection can be performed without non-defective products, but brightness variations in non-defective products cause incorrect defect detection
Solution Approach 1:
The patent changes the parameter used for defect detection from brightness comparison to edge position comparison. By extracting edge information from both the inspection target image and reference image, and comparing corresponding edge positions, the method eliminates the influence of brightness variations while maintaining the ability to detect defects accurately.
Solution Approach 2:
The patent substitutes the optical/brightness-based detection mechanism with a geometric/edge-based detection mechanism. Instead of comparing brightness intensities which are affected by variations, the method compares the positions of extracted edges, providing a more reliable defect detection that is insensitive to brightness fluctuations.
2Measurement precision
If edge positions are compared in die-to-database comparison, then structural defects can be detected, but defects based on pattern brightness cannot be detected
Solution Approach 1:
The patent merges two detection approaches by combining edge position comparison with brightness index value comparison. The method extracts both edge information and brightness information, and uses both for comprehensive defect detection, thereby detecting both structural defects and brightness-based defects simultaneously.
Solution Approach 2:
The patent creates a universal defect detection system that can handle multiple defect types through a single integrated method. By incorporating both edge position comparison and brightness index comparison, the system achieves multi-functionality in detecting various defect types including open defects, short defects, and brightness variations.
3Reliability
If brightness of contact holes is used for defect detection, then open and short defects can be detected, but variations in brightness of non-defective products reduce inspection reliability
Solution Approach 1:
The patent introduces a new parameter - edge position - to supplement the brightness parameter. By extracting edge positions from images and comparing them with reference edge positions, the method provides an additional detection dimension that is not affected by brightness variations, thereby improving measurement precision for defect detection.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This method enhances the accuracy of defect inspection in semiconductor integrated circuits by establishing standard brightness ranges for each group, allowing for precise detection of defects based on calculated brightness values.
Implementation Method 1
a scanning electron microscope generates an image of a specimen
Implementation Method 2
an inspection method using a voltage contrast using a scanning electron microscope. This technique utilizes the fact that a brightness of a pattern on an SEM image decreases when there is an open defect or a short defect in an interconnect underlying the pattern
Data Source
AI summary
A pattern defect detection method capable of detecting a pattern defect of a semiconductor integrated circuit with higher accuracy is disclosed. The pattern defect detection method includes: extracting an image of an inspection target pattern from an image of a specimen; identifying a reference pattern from design data, the reference pattern having the same shape and the same position as those of the inspection target pattern; calculating a brightness index value indicating a brightness of an entirety of the inspection target pattern; repeating said extracting an inspection target pattern, said identifying a reference pattern, and said calculating a brightness index value, thereby building mass data containing brightness index values of inspection target patterns and corresponding reference patterns; determining a standard range of brightness index value based on the brightness index values contained in the mass data; and detecting a defect of the inspection target pattern based on whether or not the calculated brightness index value is within the standard range.