Spectroscopic study of gap-surface plasmons in a metallic convex groove array and their applications in nanofocusing and plasmonic sensing

Plasmonic tapered grooves have been proven to be good candidates for the excitation of gap surface plasmons (GSPs), surface plasmons trapped vertically inside a metallic tapered groove or slit. GSPs have attracted tremendous interest due to their unique properties of concentrating light in nanosized gaps with significant field enhancement, thus offering potential applications such as ultracompact nanocircuits, broadband light absorbers, and plasmonic sensors. In this paper, we focus on GSPs supported by periodic arrays of narrow convex grooves and study in detail their properties by using visible-near-infrared (VIS-NIR) spectroscopy. We identify strong second- and third-order GSP modes excited in ultrasharp convex grooves. The dependence of GSP resonances on the groove profile is analyzed with the help of detailed full-wave simulations, revealing the fact that an ultrasharp, but finite, gap exists at the groove bottom and plays a crucial role in determining both the GSP resonance positions and the nanofocusing capability with much improved field enhancement inside the grooves. Spectral shifts of the observed GSP resonances relative to the simulation results are found in a shorter wavelength range and are qualitatively explained as nonlocal effects originating from the nonclassical microscopic behavior of local currents and charges at imperfect interfaces. Utilizing such strong and distinguishable GSP resonance line shapes in an otherwise flat reflectivity spectral baseline, we experimentally demonstrate the capability of convex groove arrays to perform dual-band refractive index sensing in the VIS-NIR range.

Figure 1

(a) Three-dimensional schematic illustration of the metallic convex groove array. Four types of groove profiles are given in the inset: linear (I) and convex-shaped (II) taper with zero-width bottom (i) and round apex (ii). (b) Scanning electron micrograph of a 1D array of metallic convex grooves (scale bar is $500\mathrm{nm}$) with a period of $p$. (c) Zoom-in image of one groove with groove depth $h$, trench width $w$, and bottom width $\alpha$. The half ellipse (red curve) with semimajor axis $h$ and semiminor axis $b$ marks the groove profile.

Figure 2

(a) Top: measured (black) and simulated (red) reflectance spectra at normal incidence for asymmetrical metallic groove array of sample 1 in air. Bottom: simulated reflectance spectra for symmetrical (solid line) and tapered (dashed line) metallic groove array in air. The inset shows the geometrical details of the groove tip. Magnetic field in color scale (b) at second-order GSP resonance ${\lambda }_{\mathrm{GSP}}=840\mathrm{nm}$ and (c) at third-order GSP resonance ${\lambda }_{\mathrm{GSP}}=660\mathrm{nm}$. The white lines mark the gold-dielectric interface.

Figure 3

(a) and (b) Simulated reflectance spectra for sample 1 at normal incidence with parameters $p=600\mathrm{nm}$, $w=58\mathrm{nm}$, $h=176\mathrm{nm}$, and $b=27.9\mathrm{nm}$. (c) The measured reflectance spectra of sample 1 (black) and sample 2 (red) and (d) the corresponding simulated reflectance spectra of sample 1 (black) and sample 2 (red) with $\alpha =2.4\mathrm{nm}$. The inset gives the resonance wavelengths of the two samples obtained from the experiments (circles) and simulations (triangles).

Figure 4

(a) and (b) Field enhancement (black curve) calculated at the groove bottom as a function of wavelength for sample 1 and sample 2 with $\alpha =2.2\mathrm{nm}$ (black dash-dotted line), respectively. The corresponding measured reflectance spectra are given by red curves. (c) and (d) Calculated normalized field magnitude distributions for sample 1 at ${\lambda }_{\mathrm{GSP}}=840\mathrm{nm}$ and sample 2 with ${\lambda }_{\mathrm{GSP}}=940\mathrm{nm}$.

Figure 5

(a) and (d) Measured and (b) and (e) simulated reflectance spectra of (a) and (b) sample 1 and (d) and (e) sample 2 with different refractive indices. (c) and (f) The GSP resonance positions as a function of refractive indices for (c) sample 1 and (f) sample 2 obtained from the experiments (circles) and simulations (triangles).