Dynamic Sampling Scheme for Wafer Metrology
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The semiconductor industry faces inefficiencies and increased costs due to overutilization of resources from excessive data generation and processing in wafer metrology, particularly in stable processes where unnecessary measurements consume computational and manpower resources, impacting productivity and cycle times.
Innovation Solution
A dynamic sampling scheme is implemented based on process capability index data and measurement history to optimize sampling rates, allowing for reduced or skipped measurements in stable processes without compromising quality control, using a decision engine to determine whether to sample or skip metrological operations at targeted process stages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If wafer metrology measurements are performed on every process run to ensure quality control, then measurement precision and reliability are improved, but productivity decreases and manufacturing costs increase due to excessive data generation and processing
Solution Approach 1:
The patent implements a dynamic sampling scheme where the measurement sampling rate is adjusted based on real-time process stability assessment. When the process is determined to be stable through statistical analysis of process capability indices, the sampling rate is reduced or measurements are skipped entirely. This dynamic adaptation allows the system to maintain quality control during stable periods while maximizing productivity, resolving the contradiction between continuous measurement and manufacturing throughput.
Solution Approach 2:
The system changes the sampling parameter (measurement frequency) based on process conditions. By monitoring process capability indices and measurement history, the system dynamically modifies the sampling rate parameter - increasing it when process variability increases and decreasing it when the process is stable. This parameter adaptation enables the system to maintain adequate measurement coverage for quality control while reducing unnecessary measurements that hinder productivity.
2Manufacturing precision
If comprehensive metrological operations are performed on all wafers, then manufacturing precision is improved, but loss of time increases due to processing delays
Solution Approach 1:
The patent applies partial action by performing metrological operations on only a subset of wafers rather than all wafers. The dynamic sampling scheme determines which wafers require measurement based on process stability assessment, allowing the system to obtain sufficient process control information from selective sampling while minimizing the time loss associated with comprehensive inspection of every wafer.
Solution Approach 2:
The system performs preliminary assessment of process stability using process capability indices before committing to comprehensive measurement. This preliminary evaluation allows the system to anticipate when reduced sampling will be sufficient, thereby avoiding unnecessary measurement time while maintaining manufacturing precision when needed.
3Reliability
If separate metrology stations are established for performing measurements, then measurement reliability is improved, but device complexity and manufacturing costs increase
Solution Approach 1:
The patent makes the metrology system universal by implementing a dynamic sampling framework that can adapt to different process conditions and stability levels. Rather than requiring multiple specialized measurement stations for different scenarios, the single metrology infrastructure is made multi-functional through intelligent sampling decisions, maintaining measurement reliability across varying process states while avoiding the complexity and cost of duplicating metrology equipment.
Solution Approach 2:
The system uses its own measurement history and process capability data to determine when measurements are necessary, making the metrology system self-regulating. This self-service capability allows the system to maintain measurement reliability through intelligent sampling decisions rather than requiring additional external verification infrastructure, thereby reducing device complexity and associated costs.
4Measurement precision
If excessive data is generated and processed from continuous measurements, then quality control is improved, but loss of energy and computational resources increases
Solution Approach 1:
The patent extracts only the essential measurement information needed for quality control by implementing dynamic sampling. Rather than generating and processing continuous data streams from all process runs, the system selectively extracts measurements only when process instability is detected or when statistical sampling requirements are met. This extraction approach maintains quality control while dramatically reducing the energy and computational resources required for data processing.
Solution Approach 2:
The system implements periodic measurement action based on process stability rather than continuous measurement. By using statistical process control and process capability indices to determine when measurements should occur, the system creates a periodic measurement rhythm that maintains quality control while minimizing unnecessary data generation and the associated computational energy consumption during stable process periods.
Data Source
AI summary
A wafer metrology system including a dynamic sampling scheme configured to optimize a sampling rate for measurement of process wafers in an IC fabrication flow based on process capability index data as well as measurement history data. For a stable process, the process wafers may be sampled at a lower rate without negatively affecting quality control.