Super-resolution imaging using surface plasmon resonance cavity lithography

Surface plasmonic lithography (SPL) utilizes the lateral propagation of light on the surface of metals, which generates transverse excitation, resulting in evanescent waves that participate in imaging to break the diffraction limit. However, a challenge is the difficulty in achieving stable evanescent wave imaging. This paper proposes a surface plasmon resonance cavity lithography (SPRCL) technique, which achieves super-resolution imaging and explores the effects of various factors on this technique. The general dispersion relation for surface plasmon polaritons in a double-resonator cavity structure is derived using traditional electromagnetic field theory. Through optical transfer function (OTF) analysis, the essence of the super-resolution imaging phenomenon is revealed-the OTF of the resonator-based super-resolution imaging system exhibits ultra-high transfer efficiency in the high-frequency evanescent wave region. In the numerical simulation, illumination light with a wavelength of 436 nm and a mask with a period of 740 nm were used to successfully obtain a feature size of 26.4 nm (∼1/17 light wavelength) and period of 52.8 nm. This result is smaller than the resolution of a conventional 193 nm immersion lithography machine. Additionally, at wavelengths of 532 nm and 633 nm, this technique achieved a stripe resolution of less than 1/11 light wavelength, with a lithography contrast greater than 0.9 and a normalized image log-slope (NILS) greater than 1.6. Analysis indicates that parameters including the mask duty cycle, incident angle, SiO₂ thickness of the photoresist (PR) upper layer, and PR layer thickness significantly influence super-resolution imaging performance. Notably, the PR layer thickness enables optical resolution tuning without altering metal grating dimensions. Compared with traditional plasmonic lithography, this approach demonstrates enhanced pattern uniformity and contrast while exhibiting improved process robustness against parameter variations.