Adaptive Weighted Error-Correction Method Based on the Error Distribution Characteristics of Multi-Channel Alignment.

Area of Science:

• Semiconductor Manufacturing
• Integrated Circuit Design
• Metrology

Background:

• Decreasing process nodes (below 10 nm) demand stricter overlay accuracy in integrated circuits.
• Sub-nanometer alignment position accuracy is crucial for wafer alignment sensors.
• Phase Grating Alignment (PGA) offers high precision but is hindered by mark asymmetry errors.

Purpose of the Study:

• To address the challenge of mark asymmetry in achieving sub-nanometer alignment accuracy.
• To develop a novel method for correcting unpredictable alignment errors in PGA technology.

Main Methods:

• Proposed an adaptive weighted method utilizing multi-channel error distribution characteristics.
• Simulated the correlation between multi-alignment error distribution and mark asymmetry.
• Developed a weight construction method based on error distribution, assuming a second-order linear relationship.

Main Results:

• Simulation confirmed a strong correlation between error distribution and mark asymmetry.
• The adaptive weighted method achieved a Root Mean Square (RMS) residual error of 0.01 nm.
• This significantly outperforms the fixed weighted method's RMS of 14.0 nm for typical mark asymmetries.

Conclusions:

• The adaptive weighted method effectively corrects alignment errors caused by unpredictable mark asymmetry.
• This approach offers a stable and highly accurate solution for sub-nanometer alignment in advanced semiconductor manufacturing.
• The proposed method overcomes limitations of fixed-weight approaches in handling complex asymmetry.