Semiconductor Recipe Selection Using AHP and Best Known Methods
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The semiconductor industry faces challenges in efficiently developing recipes for fabricating semiconductor devices, as existing methods rely heavily on time-consuming experimental fine-tuning and limited simulation capabilities, leading to inaccuracies and resource inefficiencies.
Innovation Solution
The proposed system and method utilize experimental results and measurements to develop and optimize recipes for semiconductor wafer fabrication, automating the process through subsystems that generate or find the Best Known Method (BKM) without relying on simulations, and provide recommendations to process engineers using an analytical hierarchy process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If experimental fine-tuning of recipes is performed iteratively, then manufacturing precision is improved, but loss of time and productivity deteriorate
Solution Approach 1:
The system performs preliminary computational analysis and simulations before actual fabrication to predict recipe outcomes and identify optimal parameters in advance, reducing the need for iterative experimental fine-tuning
Solution Approach 2:
The system creates virtual models and digital twins of fabrication processes to simulate and evaluate recipe performance computationally, allowing multiple iterations in silico before implementing changes physically
2Productivity
If simulations are used to predict wafer parameters, then productivity is improved, but measurement precision deteriorates due to simulation limitations
Solution Approach 1:
The system implements feedback loops where actual fabrication measurements are continuously fed back to refine and calibrate simulation models, improving prediction accuracy over time while maintaining rapid simulation-based evaluation
Solution Approach 2:
The system dynamically adjusts simulation parameters and model complexity based on the specific fabrication scenario, using simplified models for rapid screening and more complex models for critical predictions
3Manufacturing precision
If multiple recipes are evaluated and fine-tuned experimentally, then manufacturing precision is improved, but productivity and loss of substance deteriorate
Solution Approach 1:
The system performs preliminary computational screening and evaluation of multiple recipes before experimental testing, identifying the most promising candidates for further development and reducing the number of physical iterations required
Solution Approach 2:
The system uses computational models and virtual experiments to evaluate multiple recipe variations simultaneously, allowing parallel assessment of numerous options without consuming physical materials
Data Source
AI summary
Systems and methods for process development assistance are disclosed. A system includes one or more processors and memory. The system receives a scorecard including a set of criteria for fabricating a semiconductor device. The system obtains, based on the scorecard, a set of recipes stored in a knowledge base. The system obtains a set of feature dimensions associated with the set of recipes. The system obtains, using an analytic hierarchy process on the set of feature dimensions, the set of criteria, and weights of the set of criteria, an objective function value of each of the set of recipes. The system selects a subset of recipes according to the objective function value of each of the set of recipes. The system generates at least one recipe according to the selected subset of recipes and the objective function value. The system displays, via a display device, the at least one recipe.