Charged Particle Image Alignment for Multi-Defect Wafer Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current charged particle beam inspection systems face challenges in accurately identifying and locating defects in integrated circuit components due to limitations in resolution and misalignment of inspection images with reference images, especially when multiple defects are present, leading to inefficient defect detection and image analysis.
Innovation Solution
The implementation of a method that involves obtaining and analyzing images of samples using charged particle beam systems, determining defect characteristics, generating updated images by mapping or restoring them to defect-free images, and aligning these images with reference images using machine learning models to enhance defect detection and location identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If optical microscope inspection systems are used, then the inspection system is simple and easy to operate, but the resolution is limited to a few hundred nanometers due to light wavelength
Solution Approach 1:
The patent replaces optical inspection systems with charged particle beam systems (electron or ion beams). This substitution enables resolution below the diffraction limit of light (achieving sub-100nm and even sub-10nm resolution) by using particle beams instead of photons, directly resolving the resolution limitation while accepting increased system complexity as a necessary trade-off for advanced inspection capabilities
Solution Approach 2:
The patent changes the fundamental inspection parameter from optical wavelength to charged particle beam properties. By using electrons or ions with de Broglie wavelengths much shorter than visible light, the system achieves higher resolution. The patent also utilizes multiple beam parameters (beam energy, beam current, detection angles) to extract comprehensive defect information, resolving the resolution contradiction
2Measurement precision
If traditional defect detection methods are used, then the detection process is simple, but misalignment between inspection images and reference images occurs, especially with multiple defects, leading to inaccurate defect location identification
Solution Approach 1:
The patent performs preliminary actions by: (1) detecting defects in the inspection image before alignment, (2) generating a defect map indicating defect locations and characteristics, and (3) using this defect information to guide the alignment process. This preliminary defect detection and mapping enables accurate defect location identification even when multiple defects are present, resolving the measurement precision contradiction
Solution Approach 2:
The patent implements feedback mechanisms where defect detection results from the inspection image are fed back into the alignment process. The detected defect characteristics and locations serve as reference points to optimize image alignment, creating an iterative improvement loop that enhances defect location identification accuracy while managing analysis complexity
3Productivity
If multiple defects are present in the sample, then the inspection comprehensiveness increases, but the misalignment problem worsens, leading to inefficient defect detection and image analysis
Solution Approach 1:
The patent segments the defect analysis process by: (1) dividing the inspection image into regions containing defects, (2) separately analyzing each defect's characteristics, and (3) processing each defect location independently. This segmentation approach prevents defects from interfering with each other during alignment and analysis, maintaining both high productivity and measurement precision when multiple defects are present
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 improves the accuracy of defect detection and location identification by mitigating misalignment issues and enabling the use of characteristics from inspection images, even when multiple defects are present, thereby increasing the yield of functional integrated circuits.
Implementation Method 1
The primary electrons interact with the wafer and may be backscattered or may cause the wafer to emit secondary electrons
Implementation Method 2
The primary electrons interact with the wafer and may be backscattered or may cause the wafer to emit secondary electrons
Data Source
AI summary
Apparatuses, systems, and methods for providing beams for defect detection and defect location identification associated with a sample of charged particle beam systems. In some embodiments, a method may include obtaining an image of a sample; determining defect characteristics from the image; generating an updated image based on the determined defect characteristics and the image; and aligning the updated image with a reference image.