Representative Contours for IC Mask Defect Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting defects in integrated circuit (IC) masks, particularly for geometries smaller than 22 nanometers, are limited by the need for visual inspection and are not precise enough due to optical resolution limits and inefficiencies in processing high-resolution images, leading to increased risks of manufacturing defects.
Innovation Solution
Converting images of printed masks into representative contours, which are then analyzed against predetermined contour tolerances to identify and locate defects, allowing for automated generation of instructions to adjust the mask layout and flag compliance with manufacturing requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual inspection methods are used to detect mask defects, then the process is simple to implement, but the detection precision is insufficient due to optical resolution limits
Solution Approach 1:
The patent creates pixelated grayscale images as simplified copies of the original high-resolution mask patterns. These pixelated images serve as representative models that capture the essential defect information while being computationally efficient to process, resolving the contradiction between detection precision and analysis complexity
Solution Approach 2:
The patent segments the mask analysis process into distinct stages: creating pixelated grayscale images from design layouts, comparing these simplified representations against captured images, and identifying defects based on discrepancies. This segmentation allows each stage to be optimized independently, improving overall detection precision without proportionally increasing complexity
2Measurement precision
If high-resolution images are used to capture mask patterns, then the image quality is improved, but the computational processing time and efficiency deteriorate
Solution Approach 1:
The patent generates pixelated grayscale images from the design layout data as simplified computational representations. These pixelated copies retain the essential geometric information needed for defect detection while requiring minimal processing power, thus maintaining measurement precision while dramatically improving processing efficiency
Solution Approach 2:
The patent changes the resolution parameter of the mask representation by creating pixelated versions with coarser granularity. This parameter change reduces the computational complexity of image processing while preserving the critical defect information, thereby improving productivity without sacrificing essential detection accuracy
3Productivity
If pixelated grayscale images are used to represent mask patterns, then the processing efficiency is improved, but the ability to detect subtle defects deteriorates
Solution Approach 1:
The patent applies partial action by focusing the pixelated image analysis specifically on regions where defects are most likely to occur or where contour violations are most significant. Rather than attempting to preserve all fine details uniformly across the entire mask, the method concentrates computational resources on critical areas, maintaining defect detection capability while improving overall processing efficiency
Data Source
AI summary
Embodiments of a method include: converting at least one image of a printed mask to a plurality of representative contours, each corresponding to mask patterns in the printed mask; determining whether the printed mask includes a printing defect based on whether the plurality of representative contours violates a set of contour tolerances for the printed mask; in response to at least one of plurality of representative contours violating at least one of the set of contour tolerances: identifying a location where a representative contour violates the at least one of the set of contour tolerances, and generating an instruction to adjust a layout for the printed mask, based on the violating of the at least one of the set of contour tolerances; and in response to none of the plurality of representative contours violating the set of contour tolerances, flagging a layout for the printed mask as compliant.