Lithographic Pattern Defect Detection via Feature Counting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting defects in lithographic patterns used in semiconductor production are inefficient due to limited resolution of optical inspection tools and the time-consuming, error-prone nature of manual SEM analysis, leading to challenges in pinpointing exact defect locations and low throughput.
Innovation Solution
A method involving the acquisition and comparison of images from die areas to count the number of pattern features, using a reference number derived from statistical analysis or design layouts, to automatically detect defects by flagging deviations, with features touching the image edge removed to enhance accuracy and reduce processing resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If optical inspection tools are used for defect detection, then the inspection coverage is improved, but the measurement precision deteriorates due to limited resolution
Solution Approach 1:
The inspection process is segmented into two stages: optical inspection for broad coverage and SEM inspection for precise defect characterization. This segmentation allows each method to operate in its optimal performance range, with optical providing wide area coverage and SEM providing high-resolution defect analysis.
Solution Approach 2:
A coordinate mapping system acts as an intermediary between optical and SEM inspection systems. The system captures optical images with coordinate information, maps defect locations to SEM coordinates, and enables precise defect localization without requiring manual intervention, thus resolving the precision limitation of optical tools.
2Measurement precision
If manual SEM analysis is used for defect detection, then the measurement precision is improved, but the productivity deteriorates due to time-consuming manual inspection
Solution Approach 1:
Optical inspection is performed as a preliminary action to identify potential defect locations before SEM analysis. This preliminary screening reduces the number of regions requiring detailed SEM inspection, maintaining high detection accuracy while significantly improving throughput by avoiding exhaustive manual SEM analysis of entire wafers.
Solution Approach 2:
The system implements automated feedback loops where optical inspection results feed into coordinate mapping, which then guides targeted SEM analysis. The results from SEM analysis are fed back to update inspection parameters and priorities, creating an automated workflow that maintains precision while eliminating manual inspection bottlenecks.
3Measurement precision
If feature outline or area analysis is used for defect detection, then the measurement precision is improved, but the device complexity and processing resources worsen
Solution Approach 1:
The invention extracts the critical defect indicator (feature count) from complex image analysis. Instead of analyzing entire feature outlines, areas, or shapes, the system simply counts the number of features in each die, which is computationally much simpler while remaining highly effective for detecting common defects like bridges and cuts.
Solution Approach 2:
The inspection approach changes from analyzing complex geometric parameters (outline shape, area, perimeter) to counting a simple discrete parameter (number of features). This parameter transformation dramatically reduces computational complexity while maintaining the ability to detect defects, as any change in feature count immediately indicates a potential defect.
Data Source
AI summary
Example embodiments relate to methods for detecting defects of a lithographic pattern. One example embodiment includes a method for detecting defects of a lithographic pattern on a semiconductor wafer that includes a plurality of die areas. Each of the die areas has a region of interest (ROI) that includes a plurality of features forming the lithographic pattern. The method includes acquiring an image of at least one of the ROIs. The method also includes removing features touching an edge of the image. Further, the method includes counting a number of remaining features in the image.