Sub-pixel Alignment of Inspection and Design Data for Wafer Defect Binning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current semiconductor manufacturing processes face challenges in accurately aligning inspection data with design data, particularly at sub-pixel levels, leading to difficulties in identifying and correcting defects, especially as design rules shrink, resulting in increased nuisance defects and reduced inspection system sensitivity.
Innovation Solution
A computer-implemented method and system that aligns inspection data with design data by determining the position of alignment sites in design data space, using predetermined alignment sites and attributes, to achieve sub-pixel accuracy, allowing for context-based functions such as defect classification, filtering, and sensitivity adjustment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional alignment methods are used to align inspection data with design data, then the alignment process is simple and fast, but the alignment accuracy deteriorates at sub-pixel levels leading to difficulty in identifying defects
Solution Approach 1:
The alignment process is segmented into multiple hierarchical levels: first aligning at the wafer level using alignment marks, then at the die level using extracted features, and finally at the sub-pixel level using correlation techniques. This segmentation allows achieving high sub-pixel alignment accuracy while keeping each individual alignment step relatively simple and manageable.
Solution Approach 2:
The system performs preliminary actions by pre-processing inspection data to extract alignment features and pre-calculating transformation parameters before actual defect detection. Alignment sites are identified and marked in advance, and transformation matrices are computed beforehand, which simplifies the subsequent defect identification process while maintaining high alignment precision.
2Productivity
If design rules are shrunk to increase device density, then productivity increases, but measurement precision deteriorates due to increased nuisance defects and reduced inspection sensitivity
Solution Approach 1:
The system applies local quality by adjusting inspection parameters and sensitivity thresholds based on local design characteristics. Different regions of the wafer with different design rule densities receive customized inspection settings, allowing high sensitivity in critical areas while maintaining acceptable performance in less critical regions, thus preserving defect detection accuracy despite overall design rule shrinkage.
Solution Approach 2:
The system transitions from two-dimensional spatial inspection to three-dimensional analysis by incorporating depth information and multi-layer correlation. By analyzing defects in the context of multiple design layers and using 3D transformation matrices, the system can distinguish real defects from nuisance signals, maintaining measurement precision even as design rules shrink and device density increases.
3Measurement precision
If alignment accuracy is improved to sub-pixel levels, then defect classification accuracy improves, but the complexity of determining positions in design data space increases
Solution Approach 1:
The system introduces intermediary elements such as alignment marks and extracted feature points that serve as mediators between inspection data and design data. These intermediaries provide stable reference points for correlation-based alignment, enabling sub-pixel accuracy without requiring direct complex matching of all design features, thus reducing the overall complexity of position determination.
Solution Approach 2:
The system creates simplified copies or representations of design data at the inspection level, such as extracted alignment features and transformed design patterns. These copies retain the essential geometric relationships needed for accurate alignment while being computationally simpler to handle, enabling sub-pixel position determination without processing the full complexity of the original design data.
4Measurement precision
If inspection sensitivity is increased to detect smaller defects, then measurement precision improves, but the number of nuisance defects increases reducing overall effectiveness
Solution Approach 1:
The system implements feedback mechanisms where alignment accuracy and defect patterns are continuously monitored and used to adjust inspection parameters. By feeding back alignment transformation quality and defect distribution information, the system can dynamically optimize sensitivity thresholds to distinguish real defects from nuisance signals, maintaining high detection sensitivity while minimizing false positives.
Solution Approach 2:
The system changes inspection parameters dynamically based on local design characteristics and alignment quality. By adjusting sensitivity thresholds, inspection modes, and analysis parameters according to the specific region and design complexity, the system can maintain high defect detection sensitivity in critical areas while reducing nuisance defect generation in less sensitive regions, effectively managing the trade-off between sensitivity and false positives.
Data Source
AI summary
Various methods and systems for utilizing design data in combination with inspection data are provided. One computer-implemented method for binning defects detected on a wafer includes comparing portions of design data proximate positions of the defects in design data space. The method also includes determining if the design data in the portions is at least similar based on results of the comparing step. In addition, the method includes binning the defects in groups such that the portions of the design data proximate the positions of the defects in each of the groups are at least similar. The method further includes storing results of the binning step in a storage medium.