Exposure Dose Prediction Model for Lithography Critical Dimension Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional photolithographic processes struggle to accurately control critical dimensions due to differences in mask dimensions, leading to deviations in the final exposure pattern, which increases production costs and reduces yield.
Innovation Solution
A prediction model and method for exposure dose compensation using the mean-to-target (MTT) value of mask dimensions to adjust the exposure dose, calculated as E=E0+EC, where EC=[(MTTdiff/X)/(CDmask/X)]×(ES/A′)×(Wlast+Wavg), to optimize the exposure process and align the final pattern with target values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If conventional exposure dose control is used based on last measured critical dimension, then the process can operate with standard feedback control, but the critical dimension of the final exposure pattern deviates from target value due to mask dimension differences
Solution Approach 1:
The patent applies preliminary action by calculating the exposure dose compensation value before the exposure process based on mask dimensional data and historical process data. The compensation value is pre-computed using the formula EC=[(MTTdiff/X)/(CDmask/X)]×(ES/A′)×(Wlast+Wavg), allowing the exposure dose to be adjusted in advance to account for mask dimension differences, thereby preventing critical dimension deviation rather than correcting it after measurement
Solution Approach 2:
The patent implements feedback by using the measured critical dimension from the previous lot and mask dimensional data to calculate the exposure dose compensation value for the current lot. The system continuously feeds back process data (ES, Wlast, Wavg) and mask data (MTTdiff, CDmask) to dynamically adjust the exposure dose, creating a closed-loop control system that adapts to variations in mask fabrication
2Manufacturing precision
If reworking step of exposure process is executed to correct critical dimension deviation, then the critical dimension can be adjusted to meet specifications, but production time and cost increase
Solution Approach 1:
The patent prevents the need for reworking by calculating and applying the exposure dose compensation value before the exposure process. The compensation is determined in advance using mask dimensional data and historical process parameters, ensuring the critical dimension is correct from the first exposure attempt, thereby eliminating rework steps and maintaining high production efficiency
Solution Approach 2:
The patent converts the potentially harmful effect of mask dimension variations into a beneficial outcome by using the mask dimensional data (MTTdiff, CDmask) as input to calculate the compensation value. Instead of treating mask variations as problems requiring rework, the system uses them as information to pre-adjust the exposure dose, turning a source of deviation into a basis for precise control
3Manufacturing precision
If exposure dose is increased to compensate for mask dimension differences, then the critical dimension accuracy improves, but the process parameters become less stable
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the exposure dose parameter based on calculated compensation values. The system changes the exposure dose from the preset value E0 to the optimized value E=E0+EC, where EC is derived from mask dimensional parameters (MTTdiff, CDmask) and process history (ES, Wlast, Wavg). This controlled parameter change achieves higher precision while maintaining stability through the systematic, formula-based adjustment approach
Solution Approach 2:
The patent maintains process stability through feedback by using historical process data (ES, Wlast, Wavg) and mask data to calculate the compensation value. The feedback mechanism ensures that adjustments are based on actual measured deviations rather than arbitrary changes, allowing the system to adapt to variations while maintaining overall process control and stability
Data Source
AI summary
A prediction model for exposure dose is indicated by the following formula, E=E0+EC, wherein E represents an optimized exposure dose, E0 represents a preset exposure dose of a process control system, and EC represents an exposure dose compensation value, andEC=[(MTTdiff/X)/(CDmask/X)]×(ES/A′)×(Wlast+Wavg),wherein MTTdiff represents the differences between the MTT value of a previous lot and the MTT value of a next lot, CDmask represents the actual critical dimension of the mask, X represents the magnification of the mask, ES represents the actual exposure dose of a previous lot, A′ represents an experimental value obtained from the results of different lots, Wlast represents the last batch of weights and Wavg represents an average weight, and CDmask, ES, A′, Wlast and Wavg are set parameters built into the process control system.