Reticle Defect Detection via Multi-Parameter Aerial Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current reticle inspection techniques fail to effectively detect pattern anomalies and design errors in reticle design patterns, particularly those resulting from errors in the design database, and are inadequate in locating defects in single die or multi-die reticles.
Innovation Solution
A computer-implemented method and system that acquire images of the reticle design pattern using a sensor positioned near the image plane of an exposure system, detecting defects by comparing images at different values of parameters such as focus, exposure, and partial coherence, and utilizing a processor to identify anomalies based on these comparisons.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional reticle inspection techniques are used, then the inspection process is simple, but the detection accuracy of pattern anomalies and design errors is insufficient
Solution Approach 1:
The inspection process is segmented into multiple distinct stages: aerial image acquisition at different process parameters, image registration/alignment, defect detection through comparison, and classification. This segmentation allows each stage to be optimized independently, improving overall detection accuracy while managing system complexity through modular architecture
Solution Approach 2:
The inspection method transitions from traditional single-parameter inspection to multi-dimensional inspection by varying multiple process parameters (focus, exposure, partial coherence) simultaneously. This dimensional expansion in parameter space enables detection of defects that are invisible under conventional single-condition inspection, significantly improving measurement precision
2Reliability
If focus exposure matrix techniques are used for image qualification, then coverage of process conditions is improved, but the ability to detect defects farther than one exposure field is limited
Solution Approach 1:
The method creates multiple copies of the aerial image under different process conditions (focus, exposure, partial coherence) and compares them. This copying approach allows defects to be detected by their inconsistent appearance across copies, extending detection capability beyond the limitations of single-field comparison while maintaining comprehensive process condition coverage
Solution Approach 2:
The aerial image acquisition system is designed to be universal, capable of capturing images under multiple process parameters simultaneously. This multi-functional capability allows the same system to cover various process conditions and detect defects across the entire reticle area, rather than requiring separate specialized systems for different inspection tasks
3Ease of operation
If design simulation technique is used for single die reticle inspection, then signal comparison is simplified, but errors in the design database cannot be detected
Solution Approach 1:
Instead of comparing the actual reticle image against a simulated design reference (traditional approach), the invention inverts the approach by acquiring actual aerial images under multiple process conditions and comparing them against each other. This inversion eliminates dependence on the design database, allowing detection of design database errors while maintaining operational simplicity through direct image comparison
Data Source
AI summary
Computer-implemented methods and systems for detecting defects in a reticle design pattern are provided. One computer-implemented method includes acquiring images of the reticle design pattern using a sensor disposed on a substrate arranged proximate to an image plane of an exposure system configured to perform a wafer printing process using the reticle design pattern. The images illustrate how the reticle design pattern will be projected on a wafer by the exposure system at different values of one or more parameters of the wafer printing process. The method also includes detecting defects in the reticle design pattern based on a comparison of two or more of the images corresponding to two or more of the different values.