Gradient-based source and mask optimization in optical lithography.

Area of Science:

• Semiconductor Manufacturing
• Optical Lithography
• Computational Lithography

Background:

• Conventional 193 nm lithography faces limitations in extending its lifespan.
• Source and Mask Optimization (SMO) is a computationally intensive but effective solution.
• Existing SMO methods require further efficiency improvements.

Purpose of the Study:

• To develop a highly effective and efficient Source and Mask Optimization (SMO) framework.
• To improve upon existing mask optimization techniques.
• To enhance image fidelity and increase the depth of focus (DOF) in lithography.

Main Methods:

• An integrated SMO framework combining novel source and improved mask optimization.
• Utilizing pixel-based source and mask representations with objective function gradients for optimization.
• Incorporating additional penalties in the objective function to boost DOF and regularize patterns.
• Implementing an alternative optimization sequence for source and mask patterns.

Main Results:

• Achieved convergence in just two to three iteration cycles.
• Demonstrated significant improvements in image quality and DOF for 45 nm critical dimension patterns (contact holes and cross gate).
• Verified the robustness of the proposed SMO method across different initial conditions.

Conclusions:

• The proposed SMO framework offers a computationally efficient and effective solution for extending 193 nm lithography.
• The method significantly enhances lithographic performance, particularly in image quality and DOF.
• The algorithm's robustness and efficiency make it suitable for advanced semiconductor manufacturing processes.