Semiconductor Care Area Generation Using Modulation Transfer Function Rules
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methodologies for generating care areas in semiconductor device inspection are limited, as they only address a low percentage of the chip area and require pre-selection of devices and patterns, leading to suboptimal defect detection sensitivity due to noise issues.
Innovation Solution
A method using modulation transfer function-based hierarchical spacing rules to generate care areas that cover the entire chip, allowing for the detection of defects in various pattern spacings without pre-selecting devices, and adjusting these rules for optimal sensitivity and noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If pre-selected devices and patterns are used for inspection, then the inspection process is simplified, but the coverage area is limited to a low percentage of the chip
Solution Approach 1:
The care area generation methodology is designed to be universally applicable across the entire chip area, transforming the inspection system from one that only handles pre-selected devices to one that can automatically generate and inspect care areas for any device pattern throughout the full chip area, thereby achieving both simplicity and comprehensive coverage
Solution Approach 2:
The chip area is divided into multiple care areas that can be systematically generated and inspected, allowing the inspection process to cover the entire chip by segmenting it into manageable inspection units rather than requiring pre-selection of individual devices
2Ease of manufacture
If conventional care area generation is used, then the process is straightforward, but defect detection sensitivity is reduced due to noise issues
Solution Approach 1:
The methodology optimizes care area generation locally for each region of the chip, taking into account the specific pattern spacing and device characteristics of each area, which improves defect detection sensitivity by reducing background noise while maintaining process straightforwardness through systematic local optimization
Solution Approach 2:
The methodology changes key parameters such as care area size, shape, and positioning based on local pattern characteristics, thereby improving defect detection sensitivity by adapting to different regions while maintaining a straightforward generation process through parameter-based control
3Device complexity
If pre-selected patterns are inspected, then the inspection setup is simple, but the signal-to-noise ratio is suboptimal
Solution Approach 1:
The care area generation methodology dynamically adapts care area parameters based on the actual device patterns present in each region, transforming the static pre-selected pattern approach into a dynamic system that automatically optimizes signal-to-noise ratio while keeping inspection setup simple through automated adaptation
Data Source
AI summary
A method comprises: defining a set of rules for an inspection and detection of a defect in two or more electronic devices on a semiconductor chip, the set of rules being based on a modulation transfer function providing a response as contrast versus spatial frequency of the pattern spacings of the two or more electronic devices on the semiconductor chip; generating two or more care areas for two or more pattern spacings of the electronic devices on the semiconductor chip using a hierarchical set of spacing rules; and inspecting the two or more pattern spacings of the electronic devices on the semiconductor chip for defects.