Co-optimization method to reduce the pattern distortion caused by polarization aberration in anamorphic EUV...

Area of Science:

• Semiconductor Manufacturing
• Optical Engineering
• Nanotechnology

Background:

• Extreme ultraviolet lithography (EUVL) is crucial for fabricating semiconductor devices at the 7 nm node and below.
• Polarization aberrations (PAs) arising from multilayer coatings in high-numerical-aperture anamorphic lenses cause significant pattern distortions.
• Existing methods struggle to fully compensate for these PAs, limiting lithographic resolution and process margins.

Purpose of the Study:

• To develop a co-optimization method for source, mask, and process parameters in EUVL.
• To compensate for pattern distortions induced by polarization aberrations (PAs).
• To enhance the process window (PW) and reduce pattern placement error (PPE) in high-resolution lithography.

Main Methods:

• An asymmetric light source, represented by Zernike polynomial superposition, was designed to mitigate pattern placement error.
• A weighted cost function was developed, integrating the impact of polarization aberrations.
• A gradient-based statistical optimization approach was employed to iteratively refine lithography system parameters.

Main Results:

• Simulations for a 1D mask pattern at the 7 nm node demonstrated substantial improvements.
• Critical dimension error was reduced by 75.0%, and pattern placement error (PPE) decreased by 82.4% compared to previous work.
• The process window (PW) was significantly increased by 97.4% for systems affected by marginal field PAs.

Conclusions:

• The proposed source, mask, and process parameter co-optimization method effectively compensates for polarization aberrations in EUVL.
• This approach significantly enhances lithographic accuracy and expands the process window for sub-7 nm node manufacturing.
• The developed technique offers a viable solution for achieving high-fidelity patterning in next-generation semiconductor fabrication.