Lithography Metrology with Dynamic Measurement Location Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current lithographic processes face challenges in accurately correcting overlay errors due to varying process fingerprints across substrates, leading to suboptimal measurement location selection and increased noise from anomalous samples, which affects throughput and accuracy.
Innovation Solution
A method for dynamically selecting measurement locations based on recognized fingerprints and statistical analysis to optimize measurement relevance without increasing the total number of measurements, incorporating dynamic weighting to reduce outlier influence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a higher spatial density of position measurements is used to correct non-linear distortions, then manufacturing precision improves, but measurement time increases and productivity decreases
Solution Approach 1:
The patent segments the measurement process into two distinct phases: a preliminary measurement phase using a first subset of measurement locations to capture process fingerprints, and a final measurement phase using a second subset of locations. This segmentation allows the system to obtain necessary correction information without requiring measurements at all possible locations, thereby maintaining precision while improving throughput.
Solution Approach 2:
The patent performs preliminary measurements at a first subset of measurement locations before conducting final measurements. These preliminary measurements capture process-specific fingerprints that inform the selection of subsequent measurement locations. This preliminary action enables the system to adapt its measurement strategy to each specific process run, improving accuracy without increasing overall measurement time.
2Productivity
If a fixed set of measurement locations is used across all substrates, then measurement speed increases, but measurement precision decreases due to varying process fingerprints
Solution Approach 1:
The patent implements a dynamic measurement location selection system where the set of measurement locations is not fixed but adapts based on process fingerprints captured during preliminary measurements. The system dynamically determines which locations to measure next based on the specific characteristics of each substrate and process run, thereby maintaining high precision without sacrificing measurement speed.
Solution Approach 2:
The patent changes the parameters of the measurement system by selecting different measurement locations based on process conditions. Instead of using a static measurement recipe, the system modifies its measurement parameters (locations) in response to observed process fingerprints, allowing it to optimize for each specific substrate while maintaining efficient throughput.
3Measurement precision
If measurements are taken at all possible locations to capture process fingerprints, then measurement precision improves, but loss of time increases
Solution Approach 1:
The patent extracts only the essential information needed for accurate overlay correction by measuring at a selective subset of locations rather than all possible locations. The preliminary measurements at the first subset of locations extract the critical process fingerprint information, allowing the system to proceed with final measurements at a second subset without requiring exhaustive measurement of every possible location.
Solution Approach 2:
The patent applies partial action by performing measurements at only the necessary subset of locations rather than all possible locations. The two-subset approach ensures that sufficient measurements are taken to capture process fingerprints and achieve accurate correction, while avoiding the time cost of measuring at every possible location on the substrate.
Data Source
AI summary
Measurements are obtained from locations across a substrate before or after performing a lithographic process step. Examples of such measurements include alignment measurements made prior to applying a pattern to the substrate, and measurements of a performance parameter such as overlay, after a pattern has been applied. A set of measurement locations is selected from among all possible measurement locations. At least a subset of the selected measurement locations are selected dynamically, in response to measurements obtained using a preliminary selection of measurement locations. Preliminary measurements of height can be used to select measurement locations for alignment. In another aspect, outlier measurements are detected based on supplementary data such as height measurements or historic data.


