AI Etching Recipe Control for Wafer-to-Wafer Precision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The scaling down of semiconductor devices poses challenges in achieving improved quality, yield, performance, and reliability while reducing complexity, particularly due to issues arising during the etching process.
Innovation Solution
A method for controlling an etching tool using an artificial intelligence module that executes etching recipes, generates and analyzes data from measurement modules, and adjusts recipes on a wafer-to-wafer basis to ensure data is within predetermined ranges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional etching control methods are used, then the manufacturing process is simple to operate, but the manufacturing precision and yield are insufficient due to inability to adapt to scaling challenges
Solution Approach 1:
The patent implements a feedback control system where measurement modules monitor etching process parameters in real-time, and the artificial intelligence module uses this feedback data to dynamically adjust etching recipes. This closed-loop feedback mechanism enables precise control of etching processes while adapting to variations in wafer characteristics and process conditions, thereby improving manufacturing precision without requiring overly complex manual intervention.
Solution Approach 2:
The artificial intelligence module autonomously analyzes measurement data, generates optimized etching recipes, and applies adjustments without human intervention. The system performs self-optimization by learning from process data and automatically adapting recipes for subsequent wafers, reducing the need for complex manual control while maintaining high precision etching results.
2Reliability
If real-time process adjustments are implemented, then the yield and reliability are improved, but the extent of automation and data processing complexity increase
Solution Approach 1:
The artificial intelligence module serves as an intermediary between measurement modules and process control systems. It receives raw measurement data, processes and analyzes it using machine learning algorithms, and translates findings into optimized etching recipes. This intermediary layer enables reliable real-time adjustments by bridging the gap between data collection and process control, managing automation complexity through intelligent data processing rather than direct control.
Solution Approach 2:
The system performs preliminary analysis of measurement data and generates optimized recipes in advance before processing subsequent wafers. By predicting optimal process parameters based on previous wafer results and preparing adjusted recipes beforehand, the system ensures reliable outcomes while managing automation complexity through proactive rather than reactive control.
3Productivity
If wafer-to-wafer optimization is performed, then the productivity and quality are enhanced, but the measurement and data analysis requirements increase
Solution Approach 1:
The artificial intelligence module performs multiple functions including data collection, analysis, recipe optimization, and process control adjustment within a single integrated system. The measurement modules are designed to collect multiple types of process parameters simultaneously. This multi-functionality enables wafer-to-wafer optimization that enhances productivity without requiring separate complex measurement and analysis systems for each function.
Data Source
AI summary
The present application discloses a method for controlling an etching tool. The method includes executing a first etching recipe on a current wafer; generating a first set of data of the current wafer by a first measurement module; analyzing the first set of data by an artificial intelligence module coupled to the first measurement module; generating, by the artificial intelligence module, a second etching recipe and applying the second etching recipe to the etching tool when the first set of data is not within a predetermined range; and executing the second etching recipe on a next wafer.