Wafer Defect Detection Using Light Point Density Ratios
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting specific defects on wafer surfaces are unreliable due to the mixing of light points from specific and non-specific defects, making it difficult to determine the presence of specific defects, especially when non-specific defects like COPs are present, leading to incorrect determinations based on light point density thresholds.
Innovation Solution
A method that calculates the light point density ratio between a determination region and a reference region in a light point map to reliably detect specific defects, using an overlap light point map of multiple wafers to enhance accuracy, allowing for the differentiation of specific defects from non-specific ones.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If light point density threshold is used to detect defects, then detection coverage is improved, but detection precision deteriorates due to mixing of specific and non-specific defects
Solution Approach 1:
The wafer surface is divided into multiple regions: a determination region where specific defects are suspected and a reference region where general defect density is measured. This segmentation allows the system to distinguish between specific defects (high density in determination region) and non-specific defects (distributed throughout), resolving the contradiction by enabling precise detection without requiring complex analysis of all light points.
Solution Approach 2:
The invention applies different detection criteria to different regions: the determination region uses a threshold based on local light point density to identify specific defects, while the reference region provides contextual information about overall defect density. This local quality approach allows precise detection of specific defects without being misled by the presence of non-specific defects elsewhere on the wafer.
2Reliability
If light point density threshold is set to detect specific defects, then detection reliability is improved, but false positive rate increases due to non-specific defects like COPs
Solution Approach 1:
The system performs preliminary measurement of light point density in the reference region before making a determination about specific defects in the determination region. This preliminary action provides contextual information that allows the system to adjust its detection threshold or make more informed decisions, reducing false positives while maintaining high reliability for actual defect detection.
Solution Approach 2:
The invention uses feedback from the reference region measurement to inform the detection process in the determination region. By comparing the light point density in the determination region against the context provided by the reference region, the system can feedback-adjust its detection criteria, ensuring that only genuine specific defects are identified while filtering out false positives from non-specific defects.
3Ease of operation
If visual observation method is used to identify specific defects, then operational simplicity is improved, but detection accuracy deteriorates due to skill dependency and difficulty in high-density defect environments
Solution Approach 1:
The system performs self-service by automatically calculating light point densities and making defect determinations without requiring skilled human observers. The automated algorithm independently analyzes the light point maps, computes density ratios between determination and reference regions, and makes objective decisions, thereby maintaining operational simplicity while significantly improving detection accuracy and eliminating skill dependency.
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 approach effectively perceives the contrast between specific defects and their surroundings, enabling more reliable detection of specific defects regardless of the number of non-specific defects, thereby improving the accuracy of defect identification.
Implementation Method 1
a surface inspection system using the principle of laser scattering. This laser scattering type surface inspection system irradiates a wafer surface with laser. When a defect is formed at the irradiated position, light is scattered from the wafer surface.
Implementation Method 2
The scattered light is detected by a photodetector such as a light receiving element or a photomultiplier tube, and is converted into an electrical signal by a photoelectric conversion element.
Data Source
AI summary
The present invention provides a detection method which allows specific defects that would occur on a wafer surface to be detected more reliably. A method of detecting a specific defect of the present invention includes the steps of: acquiring a light point map which is in-plane position information of a light point detected in a position corresponding to a defect on a surface of a wafer by irradiating the surface of the wafer with light (S101); specifying a determination region where a specific defect is expected to be formed and a reference region which is a given region other than the determination region in the light point map, and calculating a ratio of a light point density of the determination region to a light point density of the reference region (S102); and determining whether or not the specific defect is formed based on the calculated ratio (S103).