Exploring the role of hyperbolicity in surface enhanced Raman sensing

TL;DR

This study questions the practical benefit of pristine Type-I hyperbolic metamaterials for visible-range SERS by comparing 140 nm-periodicity HMM AuNR arrays with 400 nm-periodicity non-HMM AuNR arrays, both fabricated via electron-beam lithography and nano-electroplating. Using 532 nm excitation and Rhodamine 6G as the probe, it combines COMSOL electromagnetic simulations with Raman mapping to assess field enhancements and spectral intensities, incorporating anisotropic permittivity concepts such as $\epsilon_{z}^{eff}$, $\epsilon_{x,y}^{eff}$, ENZ regions, and $\lambda_{res}$. The key finding is that hyperbolicity provides no extraordinary SERS enhancement over non-HMM configurations; both substrates show comparable SERS gains driven largely by localized surface plasmon resonances at nanoscale gaps (hotspots). This points to LSPR-dominated mechanisms as the primary driver of SERS performance in the visible range and raises questions about the practicality of pristine HMMs for SERS applications, while offering a low-cost, reproducible AuNR-based SERS substrate for detecting low-concentration analytes.

Abstract

A plasmonic nanostructure-based substrate, serving as a surface-enhanced Raman scattering (SERS) substrate, enhances the Raman scattering of molecules. By employing an electron beam lithography followed by our recently developed nano-electroplating protocol, a gold nanorod array SERS substrate can be fabricated to detect lower molecular analyte concentrations, such as Rhodamine 6G (R6G) solution. As the critical dimensions of the nanorod array decrease, they exhibit hyperbolic metamaterial (HMM) characteristics with anisotropic permittivity behavior. In our study, we fabricated two sets of nanorod arrays: one in the HMM regime (140 nm periodicity) and the other in the non-HMM regime (400 nm periodicity), aiming to evaluate the performance of each set based on R6G detection. The obtained results are compared and analyzed using COMSOL simulation and Raman mapping and the role of hyperbolicity is discussed.

Figures (3)

• Figure 1: (a) Schematic of AuNR array fabricated on quartz substrate via electron beam lithography (EBL) followed by Au electroplating protocol. The electroplating requires a seed layer (Au film), and a Ti-adhesive layer holds the Au film on the cleaned quartz substrate. (b) Simulated HMM- and non-HMM-AuNR array for an incident wavelength of 532 nm. (c) Comparison of simulated electric field enhancement for various FFs of AuNR array.
• Figure 2: (a) SEM images of HMM and non-HMM AuNR array fabricated for various FFs. (b) Effective permittivity ($\epsilon_{x, y}^{eff}$, $\epsilon_z^{eff}$) are calculated analytically using effective medium theory (EMT) in R6G medium (n = 1.33) for all the FFs of HMM structure only (as the EMT is not valid for non-HMM structures). The epsilon near zero (ENZ) region is the zero-crossing of Re ($\epsilon_z^{eff}$) curve, after which HMM region proceeds and the resonance wavelength ($\lambda_{res}$) occurs at the peak of Im ($\epsilon_{x, y}^{eff}$) curve. For FF = 0.25, ENZ and $\lambda_{res}$ occur at 542 nm and 517 nm wavelength; and for FF = 0.40, ENZ and $\lambda_{res}$ occur at 507 nm and 532 nm wavelength, respectively. Solid- and dash-line indicate 0.25 and 0.40 FF of HMM AuNR array.
• Figure 3: (a) Raman spectra of R6G molecules (one $\mu$M) detected from Au thin film, HMM-, and Non-HMM-AuNR array. Raman mapping of R6G molecules (one $\mu$M) was carried, and the highest intensity R6G peak (i.e., 613 cm-1) was plotted for (b) Au thin film, and (c) various FFs of HMM-, and Non-HMM-AuNR array. (d) Raman intensity of R6G at 613 cm-1, 776 cm-1, 1186 cm-1, 1364 cm-1, 1510 cm-1, and 1651 cm-1 peaks detected from Au thin film, and various FFs of HMM-, and Non-HMM-AuNR array.