Semiconductor Process Control via Metrology Data Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for controlling semiconductor manufacturing processes using lithographic apparatuses face instability and erroneous Key Performance Indicator (KPI) determination due to heterogeneous metrology data sets from different processing stages.
Innovation Solution
A method is provided to control semiconductor manufacturing processes by obtaining first metrology data after a initial process step and second metrology data after the initial step and at least one additional process step. The method estimates the contribution of the control action or additional process step to the process using the second metrology data and determines a KPI or correction for the initial process step using the first metrology data and the estimated contribution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If heterogeneous metrology data sets from different processing stages are used for process control, then comprehensive process monitoring is achieved, but process instability and erroneous KPI determination occur
Solution Approach 1:
The patent segments the heterogeneous metrology data into distinct components: data from the first process step, data from additional process steps, and estimated contributions. By separating these data sources and processing them through different computational paths, the system maintains comprehensive monitoring while preventing destabilizing interactions between heterogeneous data sets.
Solution Approach 2:
The patent introduces an intermediary computational layer that estimates the contribution of control actions and additional process steps to the metrology data. This intermediary processing step acts as a mediator that reconciles heterogeneous data from different stages, allowing comprehensive monitoring without directly combining incompatible data sets that would cause instability.
2Loss of information
If heterogeneous metrology data sets from different processing stages are combined, then complete process information is obtained, but erroneous Key Performance Indicator determination occurs
Solution Approach 1:
The patent extracts and separately processes the estimated contribution of additional process steps from the total metrology data. By taking out this contribution and handling it through a separate computational pathway, the system preserves complete process information while preventing the contamination of KPI determination with inappropriate data from subsequent process stages.
Solution Approach 2:
The patent performs preliminary estimation of the contribution of additional process steps before final KPI determination. This preliminary action separates the influence of subsequent process steps from the metrics used to evaluate earlier steps, ensuring that KPIs reflect only the relevant process stage while still incorporating complete process information through the estimated contribution term.
Data Source
AI summary
A method of controlling a semiconductor manufacturing process, the method including: obtaining first metrology data based on measurements performed after a first process step; obtaining second metrology data based on measurements performed after the first process step and at least one additional process step; estimating a contribution to the process of: a) a control action which is at least partially based on the second metrology data and/or b) the at least one additional process step by using at least partially the second metrology data; and determining a Key Performance Indicator (KPI) or a correction for the first process step using the first metrology data and the estimated contribution.