Lithographic Mask Inspection via Multi-Focus Intensity Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting defects on lithographic masks are inefficient in distinguishing between lithographically significant and nuisance defects, leading to false positives and increased time and resources in defect disposition.
Innovation Solution
The method involves obtaining pair of intensity images of the reticle with different focus settings, analyzing these images to determine the transmission function, and using a lithography system model to simulate the defect's impact on the photoresist pattern, thereby identifying defects that would cause significant issues on the wafer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional defect detection methods are used to inspect all defects on lithographic masks, then all defects are detected, but lithographically insignificant defects (false positives) are also identified, leading to increased time and resources for defect disposition
Solution Approach 1:
The patent changes the inspection parameters by capturing images at multiple focus settings (different z-heights) rather than a single focus plane. This allows the system to differentiate between significant defects (visible across multiple focus settings) and insignificant defects (visible only at specific focus settings), thereby reducing false positives and improving defect detection accuracy while reducing unnecessary defect disposition time.
2Measurement precision
If multiple focus settings are used to obtain images, then lithographically significant defects are better identified, but the inspection process becomes more complex
Solution Approach 1:
The patent creates multiple copies of the mask image at different focus settings and processes these copies through automated software algorithms. This copying approach allows comprehensive multi-focus inspection without requiring complex manual analysis, as the software automatically compares images across different focus planes to identify significant defects, thereby managing inspection complexity while improving defect identification accuracy.
3Reliability
If stringent defect detection is performed on all mask defects, then wafer quality is improved, but productivity decreases due to increased inspection and correction requirements
Solution Approach 1:
The patent applies partial action by performing stringent inspection only on defects that meet specific criteria (visible across multiple focus settings indicating lithographic significance). Rather than inspecting and correcting all detected defects, the system selectively focuses resources on defects that actually impact wafer quality, thereby maintaining high reliability while improving manufacturing throughput by reducing unnecessary corrections.
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 filters out nuisance defects and identifies only lithographically significant defects, reducing unnecessary corrections and improving the efficiency of the defect detection process.
Implementation Method 1
at least a pair of related intensity images of the reticle in question are obtained using an inspection apparatus
Data Source
AI summary
Disclosed are apparatus and methods for finding lithographically significant defects on a reticle. In general, at least a pair of related intensity images of the reticle in question are obtained using an inspection apparatus. The intensity images are obtained such that each of the images experience different focus settings for the reticle so that there is a constant focus offset between the two focus values of the images. These images are then analyzed to obtain a transmission function of the reticle. This transmission function is then input into a model of the lithography system (e.g., a stepper, scanner, or other related photolithography system) to then produce an aerial image of the reticle pattern. The aerial image produced can then be input to a photoresist model to yield a “resist-modeled image” that corresponds to an image pattern to be printed onto the substrate using the reticle. This resist-modeled image can then be compared with a reference image to obtain defect information. In particular, due to the introduction of the lithography tool and photoresist model, this defect information pertains to lithographically significant defects.