Substrate Shape Simulation for High-Contribution Parameter Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing shape simulations in substrate processing, such as etching, struggle to accurately determine parameters that contribute significantly to achieving an ideal substrate shape, hindering the attainment of optimal processing conditions in actual substrate processing.
Innovation Solution
A computer program and analyzer that utilize two models, the first model reproducing actual substrate processing and the second model targeting an ideal shape, allowing for the identification of parameters with a high contribution degree by comparing their values and adjusting them to bring the substrate shape closer to the ideal shape.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If shape simulation is performed using multiple parameters to predict substrate shape, then the ability to control substrate shape is improved, but it becomes difficult to identify which specific parameter has high contribution degree to achieving ideal shape
Solution Approach 1:
The patent segments the analysis by creating two distinct simulation models: a first model that reproduces actual substrate processing results and a second model that targets ideal substrate shapes. By dividing the parameter analysis into these separate models, the system can identify which parameters contribute most to achieving ideal shapes by comparing parameter values between the two models, thus resolving the information loss about parameter contribution degrees.
2Manufacturing precision
If multiple parameters are adjusted in shape simulation to achieve ideal substrate shape, then the substrate shape accuracy is improved, but the complexity of determining optimal processing conditions increases
Solution Approach 1:
The patent extracts and compares specific parameter values from two different simulation models. By taking out the parameter information from the first model (actual processing) and the second model (ideal shape), and systematically comparing them, the system identifies high-contribution parameters without requiring complex analysis of all parameters simultaneously, thus reducing the complexity of determining optimal processing conditions.
3Measurement precision
If parameter values are adjusted to match actual substrate processing results, then the simulation accuracy is improved, but it becomes harder to distinguish which parameters drive ideal shape formation
Solution Approach 1:
The patent applies local quality by focusing analysis on specific parameter comparisons between the two models rather than treating all parameters uniformly. By identifying parameters where values differ significantly between the actual processing model and the ideal shape model, the system highlights locally important parameters that drive ideal shape formation, preserving information about parameter influence while maintaining simulation accuracy.
Data Source
AI summary
An analyzer that enables specification of a parameter having a high contribution degree for making a substrate shape close to an ideal shape in a shape simulation is used to adjust a parameter in a first model in which substrate processing is simulated using a plurality of parameters, such that a substrate shape after the substrate processing, which is obtained by the simulation, becomes a specific shape obtained by actual substrate processing, adjust a parameter in a second model in which substrate processing is simulated using the plurality of parameters, such that a substrate shape after the substrate processing, which is obtained by the simulation, becomes a predetermined ideal shape, and compare a parameter of the first model and a parameter of the second model to specify a parameter having a high contribution degree for making a substrate shape after substrate processing close to the ideal shape.