Etching Recipe Prediction Using Position-Optical Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current etching operations in semiconductor manufacturing require multiple actual tests to determine etching endpoint and maintenance time, leading to increased operating steps and testing costs, especially when etching recipes change or new materials are introduced.
Innovation Solution
A method and system that collect historical etching recipes and position-optical measurement values to build a supervised learning model, allowing for the prediction of etching recipes for new products based on input specifications, thereby reducing the need for actual etching tests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If actual etching tests are conducted to determine etching endpoint and maintenance time, then etching operation reliability is improved, but operating steps and testing costs increase
Solution Approach 1:
The system performs preliminary actions by collecting historical etching data and building a prediction model in advance. When a new etching recipe is needed, the system can predict the endpoint and maintenance time directly from the model without conducting actual tests, thus maintaining reliability while avoiding time-consuming testing steps
Solution Approach 2:
The system creates a virtual copy of the etching process through data collection and model building. Instead of physically testing each new recipe, the system uses the prediction model to simulate and determine optimal parameters, replacing physical tests with virtual predictions that maintain accuracy while reducing operational overhead
2Manufacturing precision
If multiple actual etching tests are performed to obtain etching endpoint curve, then etching parameter accuracy is improved, but productivity decreases
Solution Approach 1:
The system performs preliminary data collection and model training using historical etching data from multiple sources. Once the prediction model is established, it can rapidly predict endpoint and maintenance time for new recipes with high accuracy without requiring new physical tests, thus maintaining precision while improving productivity
Solution Approach 2:
The prediction model serves itself by learning from historical data and automatically generating predictions for new etching recipes. The system uses accumulated knowledge to independently determine optimal parameters without requiring external testing interventions, maintaining accuracy while accelerating the recipe development process
Data Source
AI summary
A method and a system for predicting etching recipe are provided, wherein the method includes steps as follows: Firstly, a plurality of etching recipes of existing etched products and a plurality sets of position-optical measurement values corresponding to the plurality of etching recipes are collected. Then, a supervised learning training is performed according to a plurality of optical measurement values in each set of the position-optical measurement values to build a predicting model. A specification data of a product to be etched including a position-optical parameter is input into this predicting model to obtain a prediction result. Subsequently, according to the prediction result, one of the plurality of etching recipes of the existing etched products is selected as a suggested etching recipe for the product to be etched.