Misregistration Metrology Parameter Selection Using Measurement Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for measuring misregistration in semiconductor device manufacturing are inefficient in optimizing measurement parameter configurations, leading to inaccuracies and suboptimal results.
Innovation Solution
A data-driven method that simulates multiple measurement scenarios for semiconductor devices using various parameter configurations, identifies the best configurations based on quality metrics like inaccuracy, Qmerit, focus sensitivity, and contrast precision, and applies these to misregistration metrology tools for precise measurement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple measurement simulations are performed with various parameter configurations to identify optimal settings, then measurement precision and accuracy improve, but measurement time and processing complexity increase
Solution Approach 1:
The system performs measurement simulations and identifies optimal parameter configurations before actual misregistration measurements are taken. By pre-determining the best measurement parameters through simulation data analysis, the system eliminates the need to test multiple configurations during actual measurements, thus improving accuracy while reducing measurement time.
Solution Approach 2:
The system creates simulation models that replicate actual measurement conditions and device characteristics. These simulated measurements provide a copy of the measurement process that can be analyzed to determine optimal parameters without consuming actual measurement time or resources.
2Measurement precision
If multiple parameter configurations are tested to find the best measurement settings, then measurement accuracy improves, but device complexity and system requirements increase
Solution Approach 1:
The system automatically analyzes simulation data and identifies optimal measurement parameter configurations without requiring manual intervention or complex external systems. The automated analysis process handles the complexity internally, providing accurate measurement parameters while keeping the overall system relatively simple.
3Reliability
If simulation data is compared with actual measurement data to identify root causes of mismatches, then measurement reliability improves, but processing time and computational requirements increase
Solution Approach 1:
The system compares simulation data with actual measurement data and uses the discrepancies as feedback to identify root causes of mismatches and refine measurement parameters. This feedback loop improves measurement reliability by continuously validating and adjusting parameters based on actual performance, while the automated nature of the comparison minimizes additional processing time.
Data Source
AI summary
A data-driven misregistration parameter configuration and measurement system and method including simulating a plurality of measurement simulations of at least one multilayered semiconductor device, selected from a batch of multilayered semiconductor devices intended to be identical, using sets of measurement parameter configurations, generating simulation data for the device, identifying recommended measurement parameter configurations selected from sets of measurement parameter configurations, providing a multilayered semiconductor device selected from the batch, providing the at least one recommended set of measurement parameter configurations to a misregistration metrology tool having multiple possible sets of measurement parameter configurations, measuring at least one multilayered semiconductor device, selected from the batch, using the recommended set, thereby generating measurement data for the device, thereafter identifying a final recommended set of measurement parameter configurations and measuring misregistration of at least one multilayered semiconductor device, selected from the batch, using the final recommended set.