Lithography Control Parameter Optimization for Semiconductor Yield
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current lithographic processes face challenges in optimizing yield due to suboptimal control strategies that do not effectively utilize performance data from all dies on a substrate, leading to inefficiencies in process corrections and yield optimization.
Innovation Solution
A method is introduced to determine control parameters for lithographic processes by analyzing performance data and die specification data, allowing for the identification of manufacturable dies and adjusting control parameters to prioritize those that are likely to meet specifications, thereby optimizing process settings and improving yield.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If control parameters are determined using traditional optimization techniques that adjust imaging parameters to optimize root mean square error across the substrate, then process control is improved, but yield optimization is not achieved because the approach does not account for die-level specifications and manufacturability
Solution Approach 1:
The patent applies local quality by transitioning from global RMS error optimization to die-level control parameter optimization. Each die's control parameters are determined based on its specific performance data and die specification data, allowing different regions of the substrate to be optimized according to their individual manufacturability characteristics rather than applying a uniform optimization approach across the entire substrate.
Solution Approach 2:
The patent implements parameter changes by modifying the optimization objective from minimizing RMS error to maximizing expected yield. The control parameter determination process incorporates die specification data and performance data to calculate expected yield values, and adjusts control parameters to optimize this yield metric rather than focusing solely on error minimization.
2Stability of the object's composition
If control efforts are applied uniformly across all dies on a substrate, then process consistency is maintained, but resources are wasted on dies that cannot meet specifications regardless of process adjustments
Solution Approach 1:
The patent applies local quality by differentiating control strategies based on die-level characteristics. Dies are evaluated individually using their performance data and die specification data to determine expected yield, allowing the system to apply appropriate control efforts only to dies that can benefit from process adjustments, rather than uniformly treating all dies the same.
Solution Approach 2:
The patent implements partial action by selectively applying control parameter optimization only to dies with reasonable expected yield potential. Dies that are determined to be non-manufacturable based on their performance data and specifications are excluded from control efforts, avoiding wasted resources on processes that cannot achieve specification compliance regardless of optimization.
3Area of stationary object
If all dies on a substrate are included in control parameter calculations, then comprehensive process coverage is achieved, but the accuracy of yield optimization is reduced due to inclusion of non-manufacturable dies
Solution Approach 1:
The patent applies the taking out principle by extracting and excluding non-manufacturable dies from control parameter calculations. Based on die specification data and performance data comparison, dies that cannot meet specifications are identified and removed from the optimization process, ensuring that control parameters are calculated only from dies with reasonable expected yield potential, thereby improving optimization accuracy.
Solution Approach 2:
The patent implements parameter changes by modifying the input dataset for control parameter calculations. Instead of using all dies uniformly, the system dynamically adjusts which dies are included based on their expected yield values derived from performance data and die specifications, changing the composition of the calculation dataset to improve optimization accuracy.
Data Source
AI summary
A method for determining a control parameter for an apparatus used in a semiconductor manufacturing process, the method including: obtaining performance data associated with a substrate subject to the semiconductor manufacturing process; obtaining die specification data including values of an expected yield of one or more dies on the substrate based on the performance data and/or a specification for the performance data; and determining the control parameter in dependence on the performance data and the die specification data. Advantageously, the efficiency and/or accuracy of processes is improved by determining how to perform the processes in dependence on dies within specification.