I-line photolithographic metalenses enabled by distributed optical proximity correction with a deep-learning model

Area of Science:

• Optics and Photonics
• Materials Science
• Nanotechnology

Background:

• High pattern fidelity is crucial for metalens and metasurface performance.
• Economic photolithography faces challenges with subwavelength structure resolution and process windows.
• Optical proximity correction (OPC) is essential but requires accurate lithographic models for metasurfaces.

Purpose of the Study:

• To develop an accurate lithographic model for metasurface patterning.
• To implement a distributed optical proximity correction (OPC) flow for near-infrared (NIR) metalens fabrication.
• To improve pattern fidelity and focusing efficiency in NIR metalenses using i-line photolithography.

Main Methods:

• Utilized Hopkin's image formulation and fully convolutional networks (FCN) for lithographic modeling.
• Employed a distributed OPC flow to control critical dimension (CD) patterning.
• Fabricated a 1.5 mm diameter NIR metalens with a numerical aperture (NA) of 0.45 using i-line photolithography.

Main Results:

• Achieved an average critical dimension (CD) error of 1.69% due to process variations.
• Successfully patterned a 260 nm CD in the metalens layout, corresponding to a lithographic constant k1 of 0.46.
• Fabricated NIR metalens demonstrated a measured focusing efficiency of 64%, closely matching the calculated 69%.

Conclusions:

• The developed lithographic model and OPC flow enable precise patterning for NIR metalenses.
• The study demonstrates high focusing efficiency for metalenses fabricated using i-line photolithography.
• This approach overcomes limitations of conventional photolithography for advanced metasurface applications.