Plasmonic Bi-Cavity Nanostructure for Efficient Light Collection and Localization

TL;DR

High-NA optics are often required for on-axis radially polarized excitation in TERS, constraining sample geometry. The PTTP is a bi-cavity plasmonic probe with plateaus and $L$ that supports a hybrid antenna-cavity mode when co-tuned, funneling energy to the apex under radially polarized excitation. Finite-element simulations and graphene experiments show that plateau length $W$ forms an in-plane SPP Fabry-Pérot-like cavity, yielding a locus of apex $|E|^2$ maxima as $L$ and $W$ are varied. Relaxing the NA to $0.75$ preserves robust apex enhancement and NF signals, broadening practical deployment of nano-Raman instrumentation.

Abstract

Tip-enhanced Raman spectroscopy (TERS) typically relies on high-NA excitation to generate a strong axial field at the tip apex, which shortens the working distance and constrains sample geometries. We show that a plasmonic bi-cavity tip, the plasmon-tunable tip pyramid (PTTP), co-tuned in nanopyramid length L and plateau length W, supports a hybrid antenna-cavity mode that funnels energy to the apex under radially polarized, on-axis excitation, even with a dry objective of NA = 0.75. Finite-element simulations identify W as a design-critical parameter that sets an in-plane surface-plasmon-polariton (SPP) Fabry-Pérot-like resonance; co-tuning (L,W) yields a periodic series of maximal apex |E|^2. Experiments on monolayer graphene confirm near-field enhancement and reproduce the characteristic annular TERS point spread function (PSF) with NA = 0.75. Relaxing the NA requirement increases working distance and compatibility with constrained environments, pointing to practical, deployment-ready nano-Raman instrumentation.

Figures (5)

• Figure 1: Computational model of the nanostructure highlighting the plateau and the optimization parameters. Being L the nanopyramid length and W the plateau length.
• Figure 2: $|$E$|^2$ sampled along the optical axis (see double arrow in inset) for NA = 0.75 (black-dashed) and NA = 1.4 (green-solid) configurations. Data normalized by the maximum amplitude of $|$E$|^2$ for NA = 1.40.
• Figure 3: Tip up (FF, in blue-dashed) and tip down (NF, in red-solid) Raman spectra measured from a monolayer graphene with the two illumination configurations: NA = 1.4 in a) and NA = 0.75 in b). All spectra were normalized by the amplitude of the 2D peak in the tip-up configuration, and the background was removed.
• Figure 4: Tip Enhanced Raman Hyperspectral maps of graphene measured with the 0.75 NA dry objective lens. a,c) Full width at half maximum ($\Gamma$) and b,d) position ($\omega$) of the G and 2D bands, respectively. e) Atomic force microscopy (AFM) topography map. f) Topography line profile along segment 1 shown in e) of the feature responsible for the localized strain.
• Figure 5: i) Optimization heatmap revealing the field intensity($|$E$|$$^2$, color code) 5 nm below the tip apex for different tip configurations, varying the nanopiramid length (L) and plateau length (W). a-e locate specific configurations displayed in (ii). ii) from a) to e) Tangential $|$E$|$$^2$ is displayed for each point of interest along the surface of the plateau and nanopyramid as shown in figure \ref{['pttp']} b) (i). The region between the concentric dashed red circles represents the half-maximum contour of the radial excitation of the nanostructure.