Plasmonic Responses in Metal Nanoslit Array Fabricated by Interference Lithography

Abstract

Gap tunable gold nanoslit arrays were fabricated by interference lithography and investigated numerically to understand the impact of fabrication errors on plasmonic responses. To fabricate the gap tunable gold nanoslit arrays, photoresist nanoslit arrays on quartz substrate were first formed by laser interference, and then converted to gold nanoslit array on glass substrate by perpendicular gold deposition and photoresist lift-off. Because the photoresist nanoslit has a sinusoidal profile due to the laser light interference lithography, different photoresist development time from 20[Formula: see text]s to 30[Formula: see text]s can tune the photoresist width from 100[Formula: see text]nm to 70[Formula: see text]nm, thus allows the gap-width-tuned metallic nanoslits to be attained accordingly. The optical properties of the fabricated gold nanoslit arrays were investigated experimentally and theoretically by studying the absorption in the transmission spectra. Within the wavelength range of 400[Formula: see text]nm to 860[Formula: see text]nm, the nanoslit in air has two prominent absorption peaks at 500[Formula: see text]nm and 670[Formula: see text]nm. It is found that a simulation model with gold nanoslit fabrication errors such as size variation, chromium adhesive layer and gold residue in nanoslit gaps considered can better match the simulation peaks with the experiments. The simulation of the gold nanoslit array in air indicates that the 500[Formula: see text]nm peak includes the interband transition and surface plasmon polariton (SPP) at air-gold surface, and the other peak at 670[Formula: see text]nm is SPP at glass side. The two SPP peaks are both sensitive to the refractive index of surrounding solution, with sensitivities of the two peaks demonstrated to be 267[Formula: see text]nm/RIU and 111[Formula: see text]nm/RIU in experiments, and 462[Formula: see text]nm/RIU and 180[Formula: see text]nm/RIU by simulation. The lower sensitivity detected by experiments might be due to some air bubbles in the flow cell reducing the effective refractive index around the nanoslit. The shorter wavelength SPP mode is 2.4 (in experiments) or 2.6 times (by simulation) more sensitive than the long wavelength SPP mode because its plasmonic field concentrates on water-gold surface. The plasmonic responses we simulated with fabrication errors explained our experimental investigations, and deepened our understanding on the application of the gold nanoslit array for refractive index-based biosensing.