Plasmonic Polarization Rotation in SERS Spectroscopy

Abstract

Surface-enhanced Raman optical activity (SEROA) has been extensively investigated due to its ability to directly probe stereochemistry and molecular structure. However, most works have focused on the Raman optical activity (ROA) effect arising from the chirality of the molecules on isotropic surfaces. Here, we propose a strategy for achieving a similar effect: i.e., a surface-enhanced Raman polarization rotation effect arising from the coupling of optically inactive molecules with the chiral plasmonic response of metasurfaces. This effect is due to the optically active response of metallic nanostructures and their interaction with molecules, which could extend the ROA potential to inactive molecules and be used to enhance the sensibility performances of surface-enhanced Raman spectroscopy. More importantly, this technique does not suffer from the heating issue present in traditional plasmonic-enhanced ROA techniques, as it does not rely on the chirality of the molecules.

Keywords: Raman scattering; SEROA; SERS; metallic nanostructures; metasurfaces; optical activity; plasmons.

Figure 1

Schematic of the nanoclock array. (a) Au nanoclock structures are placed on the silica substrate (n = 1.45). H represents the thickness of the Au nanoclock structure. (b) Schematic of a single Au nanoclock particle. Each unit cell consists of a disc of diameter D, a long arm of length L₁ and width W, and a short arm of length L₂ and width W. The angles between the two arms are 90 and 180°, as shown in (c) and (d). The incoming beam is incident from the substrate side, normal to the surface. Light can impinge with polarization perpendicular (black arrow) and parallel (red arrow) to the long arm of the nanostructure, as shown in (b). The period in both x and y directions is denoted by Λ and has a value of either Λ = 500 nm or Λ = 1500 nm. The samples with the configuration in (c) are labeled as CL 90° 500 and CL 90° 1500, while those with the configuration in (d) are labeled as CL 180° 500 and CL 180° 1500, depending on the period. (e) Scanning electron microscopy image of the Au nanoclock array, when Λ = 1500 nm and θ = 180°.

Figure 2

Experimental extinction spectra of the nanoclock arrays (a) CL 180° 500 and (b) CL 90° 500 recorded under two incident polarization states as labeled by arrows. Solid and dashed lines are experimental and simulated data, respectively. The vertical one-way arrows in the panels correspond to the incident excitation wavelengths. Insets: schematics of the structures.

Figure 3

Theoretical results of the extinction cross section (CS) for a single Au nanoclock particle in four configurations. The corresponding distributions of z components of the electric field (E_z) at the resonances, which correspond to charge distribution, are also provided. The fields are monitored in a plane located 1 nm above the structures and perpendicular to the incident wave vector. The corresponding wavelengths used in each plot are as follows: (a) 559, 784, and 1200 nm, (b) 564, 640, and 1188 nm, (c) 534 and 689 nm, and (d) 603 and 847 nm. The shape of the Au nanoclock particles is the same for θ = 180° in (a) and (c), while the shape is the same for θ = 90° in (b) and (d). The arrows indicate the polarization of the incidence. Horizontal polarization is used in (a) and (b), while vertical polarization is used in (c) and (d).

Figure 4

Far-field polarization calculation (transmission mode) for nanoclock particle arrays (θ = 90°) with a period of Λ = 500 nm. The polarization of the normal incidence is represented by arrows in (a) and (b). The light beam is impinging from the substrates (see Figure 1). The inset in (a) is a schematic of the elliptical polarization. α denotes the ellipse major axis angle (blue curve, left axes), and E_ξ/E_η denotes the major/minor axis ratio (brown curve, right axes). The elliptical polarization is shown for representative wavelengths, including 780, 840, and 1062 nm in (a) and 811, 837, and 855 nm in (b).

Figure 5

SERS depolarization ratio for nanoclock particle arrays with (a–d) period Λ = 500 nm (particle array) and (e–h) period Λ = 1500 nm (single particle). Orange stars denote the experimental depolarization ratio, while the sample prediction (red squares) and the simulation prediction (green dots and curves) are also plotted. The excitation wavelength is 638 nm. The insets shows the structures and the incident polarizations for each case.