Dimension Conversion Difference Prediction for Photomask Manufacturing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current dimension conversion difference prediction methods are inadequate for nonlinear features in semiconductor manufacturing, leading to inaccuracies in pattern transfer and potential electrical characteristic degradation and yield loss due to bridges and breaks in patterns.
Innovation Solution
The method involves determining the opening angle or incident amount at a conversion difference prediction point based on design pattern data, correlating these values with actual measurements to predict dimension conversion differences, thereby improving prediction accuracy for both linear and nonlinear features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If stepwise predetermined values are used to predict dimension conversion difference based on line width and space width, then prediction accuracy for linear features is improved, but prediction accuracy for nonlinear features deteriorates
Solution Approach 1:
The patent changes the prediction parameters from simple line width and space width to include opening angle and incident amount parameters that better represent the geometric characteristics of both linear and nonlinear features. This parameter transformation enables accurate prediction across different feature types by capturing the essential geometric properties that influence dimension conversion differences.
Solution Approach 2:
The patent segments the prediction approach by treating linear and nonlinear features differently through distinct parameter sets. Linear features are characterized by line width and space width, while nonlinear features are characterized by opening angle and incident amount. This segmentation allows each feature type to be predicted using its most relevant parameters, resolving the contradiction between accuracy for linear features and applicability to nonlinear features.
2Area of moving object
If design rule shrinkage is pursued to create finer patterns, then device integration density is improved, but pattern transfer accuracy deteriorates
Solution Approach 1:
The patent applies preliminary action by predicting dimension conversion differences before the actual photolithography and etching processes. The predicted values are used to pre-correct the mask pattern dimensions, compensating for expected process variations. This advance correction ensures that even as design rules shrink to finer patterns, the final transferred patterns maintain high accuracy by accounting for process-induced dimensional changes in advance.
Data Source
AI summary
A method for dimension conversion difference prediction includes: determining an opening angle at a conversion difference prediction point on basis of a design pattern data; and predicting a dimension conversion difference on basis of correlation between the opening angle and an actual measurement value of the dimension conversion difference, or a method for dimension conversion difference prediction includes: determining an incident amount of incident objects at a conversion difference prediction point on basis of a design data; and predicting a dimension conversion difference on basis of correlation between the incident amount and an actual measurement value of the dimension conversion difference.