Process Proximity Effect Correction Using Machine Learning Sensitivity Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
As semiconductor processes become more complex and miniaturized, the increasing number of patterns in semiconductor layouts leads to process errors during etching, which existing methods struggle to efficiently correct, especially with the rapid increase in computation required to modify layouts.
Innovation Solution
A process proximity effect correction method using a machine learning module to train a sensitivity model with layout images and critical dimensions, estimate after cleaning inspection critical dimension sensitivity prediction values, and determine correction rates for layout critical dimensions, enabling efficient pattern dispersion improvement in semiconductor processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the number of patterns in semiconductor layout is increased to achieve higher integration levels, then the integration level is improved, but process errors during etching increase due to pattern dispersion
Solution Approach 1:
The patent applies preliminary action by performing process proximity effect correction before the etching process. A sensitivity model is trained using layout images and critical dimension data to predict and correct pattern dispersion issues in advance. The correction rate is calculated and applied to modify the layout CD, so that when etching occurs, the patterns are already optimized to compensate for expected dispersion, thus maintaining etching accuracy despite high integration levels.
2Manufacturing precision
If traditional methods are used to modify layout patterns to compensate for process errors, then etching accuracy is improved, but computation time increases rapidly
Solution Approach 1:
The patent replaces traditional mechanical/computational layout modification methods with a machine learning-based sensitivity model. Instead of using complex computational algorithms to calculate and adjust each pattern, the system trains a neural network model that can quickly predict pattern dispersion and determine correction rates. This substitution of mechanical computation with intelligent modeling dramatically reduces computation time while maintaining or improving etching accuracy.
3Manufacturing precision
If layout critical dimension is corrected to improve pattern distribution, then manufacturing precision is improved, but the complexity of the correction process increases
Solution Approach 1:
The patent applies parameter changes by focusing correction efforts on the critical dimension parameter of the layout. Rather than attempting to optimize all layout parameters simultaneously, the sensitivity model specifically analyzes and adjusts the CD parameter based on trained relationships between layout images and measured critical dimensions. This targeted parameter adjustment simplifies the correction process while effectively improving pattern distribution and manufacturing precision.
Data Source
AI summary
Provided is a process proximity effect correction method capable of efficiently improving the dispersion of patterns. There is a process proximity effect correction method according to some embodiments, the process proximity effect correction method of a process proximity effect correction device for performing process proximity effect correction (PPC) of a plurality of patterns using a machine learning module executed by a processor, comprising: training a sensitivity model by inputting a layout image of the plurality of patterns and a layout critical dimension (CD) of the plurality of patterns into the machine learning module; estimating an after cleaning inspection critical dimension (ACI-CD) sensitivity prediction value of the plurality of patterns by inferring an ACI-CD prediction value of the plurality of patterns; and determining a correction rate of the layout CD of the plurality of patterns using the estimated sensitivity prediction value.