Alignment Mark Detection Using Waveform Correlativity Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In semiconductor manufacturing, the alignment of shot areas on a wafer is challenging due to poor overlay accuracy between circuit layers, leading to defective chips and reduced yield, primarily because existing methods struggle to accurately recognize alignment marks in noisy waveform data and require time-consuming adjustments of processing parameters.
Innovation Solution
An evaluation system and method that photoelectrically detects potential positions of alignment marks within waveform data, specifies the most likely positions based on specifying information, and evaluates the search operation to improve mark recognition accuracy by simulating mark recognition operations at designated positions, allowing for parameter adjustments to prevent mark detection errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If mark recognition operation is performed using waveform data with noise components, then mark detection may fail, but increasing processing parameters adjustment time is required to resolve detection errors
Solution Approach 1:
The patent performs mark recognition operations at multiple potential positions before final detection, extracting candidate positions and evaluating correlativity in advance. This preliminary action at multiple positions prevents detection errors that would require time-consuming parameter adjustments later.
Solution Approach 2:
The patent evaluates correlativity between template waveforms and waveform data at each potential position, using this feedback to determine whether marks are recognized. This feedback mechanism allows automatic identification of detection failures without requiring manual parameter adjustment.
2Measurement precision
If multiple processing parameters are adjusted to improve mark recognition, then detection accuracy may improve, but the complexity of parameter setting increases
Solution Approach 1:
The patent performs self-evaluation of mark recognition operations by automatically calculating correlativity between template waveforms and detected waveforms. This self-service mechanism eliminates the need for complex manual parameter setting and evaluation, as the system automatically identifies detection quality.
Solution Approach 2:
The patent changes the approach from adjusting multiple processing parameters to evaluating correlativity as a detection metric. By using correlativity values to assess mark recognition quality, the system avoids the complexity of tuning multiple parameters while maintaining high detection accuracy.
3Reliability
If mark search operation is performed at all potential positions, then detection coverage is improved, but processing time increases
Solution Approach 1:
The patent performs mark recognition operations at multiple potential positions (excessive action) to ensure reliable detection, then uses correlativity evaluation to identify the most likely position. This approach maintains high reliability while avoiding the need to process all possible positions equally.
Solution Approach 2:
The patent extracts candidate potential positions from waveform data based on characteristics of alignment marks, then focuses recognition operations on these extracted candidates. This extraction approach maintains comprehensive coverage while reducing the total number of positions requiring full processing.
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 mark recognition, reduces mark detection errors, and streamlines the process of adjusting processing parameters, thereby improving the yield and quality of semiconductor devices by facilitating precise alignment and overlay exposure.
Implementation Method 1
a waveform data that represents photoelectric conversion signals corresponding to areas on the wafer including the alignment marks is obtained
Data Source
AI summary
First, an operator inputs various parameters required for a mark recognition operation including a designated position coordinate designated via a mouse and a keyboard (step 201). Then, an edge potential position closest to the designated position is selected (step 205), or an edge potential position within a predetermined range having the designated position as a datum is selected (step 207), and a mark recognition operation is performed at the selected edge potential position (step 209). The recognition results are shown on a display (step 211).