Scatterometry APC Model for High-Dimensional Knob Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current advanced process control (APC) methods in semiconductor manufacturing face challenges in achieving high accuracy due to the high dimensional knob space and variability introduced by multiple manufacturing routes and time scales.
Innovation Solution
A machine learning-based APC system is developed, which uses scatterometric training data and process control knob settings to train a machine learning model that recommends changes to process control knob settings to compensate for variations in pre-process scatterometric data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional process control methods are used, then the control process is simple, but the accuracy of manufacturing is insufficient for high dimensional knob space
Solution Approach 1:
The patent replaces traditional mechanical/process-based control methods with machine learning algorithms. The system uses neural networks and other ML techniques to analyze scatterometric data and determine optimal process control knob settings, substituting conventional control mechanisms with intelligent computational systems that can handle high-dimensional parameter spaces and achieve the required manufacturing precision.
2Adaptability or versatility
If multiple manufacturing routes are used to increase flexibility, then adaptability improves, but variability of production results increases
Solution Approach 1:
The patent implements a feedback-based machine learning system that continuously analyzes scatterometric measurements from wafers processed through different manufacturing routes. The ML model learns from actual production data and adjusts process control knob recommendations to compensate for variations introduced by different manufacturing routes, maintaining consistency while preserving flexibility. The system adapts to each route's characteristics and provides route-specific optimization.
3Measurement precision
If scatterometric data analysis is used to improve measurement capability, then pattern parameter detection improves, but the complexity of data processing increases
Solution Approach 1:
The patent replaces complex manual or rule-based data processing methods with machine learning algorithms. The system uses neural networks and other ML techniques to automatically analyze scatterometric data, extract pattern parameters, and determine process control optimizations. This substitution maintains high measurement precision while managing data processing complexity through intelligent computational approaches that can handle the complexity autonomously.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system effectively reduces variations in pattern parameters at wafer sites by optimizing process control knob settings, thereby improving the accuracy and consistency of semiconductor manufacturing processes.
Implementation Method 1
Scatterometric data (also referred to herein as 'spectra data') is typically acquired as reflected light radiation that is indicative of optical properties of patterns at wafer sites.
Data Source
AI summary
A system and methods for Advance Process Control (APC) in semiconductor manufacturing include: for each of a plurality of waiter sites, receiving a pre-process set of scatterometric training data, measured before implementation of a processing step, receiving a corresponding post-process set of scatterometric training data measured after implementation of the process step, and receiving a set of process control knob training data indicative of process control knob settings applied during implementation of the process step; and generating a machine learning model correlating variations in the pre-process sets of scatterometric training data and the corresponding process control knob training data with the corresponding post-process sets of scatterometric training data, to train the machine learning model to recommend changes to process control knob settings to compensate for variations in the pre-process scatterometric data.