Semiconductor Inspection Alignment Using Built-In Targets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing inspection systems for semiconductor manufacturing face challenges in accurately determining the position of defects in design data space, leading to reduced alignment accuracy and propagation of bad data, especially as design rules shrink and defect sizes become smaller.
Innovation Solution
A system and method that utilize an inspection subsystem with an energy source and detector, along with computer subsystems to select alignment targets from design data, align images, determine offsets, and position output generated by the inspection subsystem in design data space, using built-in targets for improved accuracy and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If post-processing alignment is performed on inspection image patch and design clip, then alignment accuracy may be improved, but the method is dependent on having sufficient information in the inspection patch and design clip which is often not met
Solution Approach 1:
The patent applies preliminary action by embedding alignment targets into the design data before inspection occurs. These targets are pre-positioned at known locations and with known characteristics, so when inspection is performed, the alignment can be immediately established without requiring post-processing alignment operations that depend on sufficient information in the inspection patch.
Solution Approach 2:
The patent introduces alignment targets as intermediary elements that facilitate the alignment between inspection data and design data. These targets serve as a common reference that both the inspection system can detect and the design system can locate, enabling accurate coordinate transformation without direct dependency on patch or clip information sufficiency.
2Measurement precision
If inspection systems focus on detecting all defects as design rules shrink, then defect detection sensitivity increases, but the population of nuisance defects increases dramatically making correction difficult and expensive
Solution Approach 1:
The patent applies local quality by using alignment targets with specific local characteristics that differ from actual defects. The targets have known positions, sizes, and patterns that allow the inspection system to distinguish them from real defects. This enables the system to focus on relevant defects while filtering out nuisance defects that don't match the characteristics of actual process variations.
Solution Approach 2:
The patent converts the challenge of increased defect detection into a benefit by using the same high-sensitivity inspection data to both detect actual defects and identify alignment targets. The nuisance defects that would normally reduce accuracy are distinguished from real issues through their spatial relationship to the known alignment targets, turning the high defect population into an opportunity for more comprehensive monitoring.
3Measurement precision
If coordinate registration accuracy between inspection system and design is improved, then alignment precision increases, but the method does not necessarily provide the alignment accuracy that is needed
Solution Approach 1:
The patent replaces the mechanical coordinate registration system with an optical/image-based alignment system. Instead of relying on physical coordinate transformations between inspection and design systems, the method uses image processing to locate alignment targets in the inspection data and mathematically transform coordinates based on the known positions of these targets, achieving higher precision through computational rather than mechanical means.
Solution Approach 2:
The patent changes the parameter of alignment from global coordinate registration to local target-based positioning. By using multiple alignment targets distributed across the inspection area and calculating offsets based on their known positions, the system achieves higher overall alignment accuracy through localized measurements rather than relying on a single global coordinate transformation.
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 enhances the accuracy of pattern-to-design alignment, allowing for higher reliability and precision in defect detection and analysis, even at the sub-pixel level, thereby improving the sensitivity and relevance of inspection results to the design.
Implementation Method 1
The detector is configured to detect energy from the specimen and to generate output responsive to the detected energy
Data Source
AI summary
Methods and systems for determining a position of output generated by an inspection subsystem in design data space are provided. One method includes selecting one or more alignment targets from a design for a specimen. At least a portion of the one or more alignment targets include built in targets included in the design for a purpose other than alignment of inspection results to design data space. At least the portion of the one or more alignment targets does not include one or more individual device features. One or more images for the alignment target(s) and output generated by the inspection subsystem at the position(s) of the alignment target(s) may then be used to determine design data space positions of other output generated by the inspection subsystem in a variety of ways described herein.