Plasmon-enhanced Raman scattering is a prominent measurement tool for sensitively obtaining the chemical information of chemical and biological materials (e.g., single molecule [1,2], DNA [3], and proteins [4]) by analyzing the frequency shifts owing to molecular vibrations. Plasmonic structures effectively excite localized surface plasmon resonance (LSPR), facilitated by the optical field of light-oscillating free electrons at the surface of the plasmonic structure [5]. LSPR is the collective oscillation of free electrons that enhances the optical field near a plasmonic structure. LSPR properties vary depending on the size, shape, and material of the plasmonic structure. Because Raman scattering is highly dependent on the intensity of the optical field, an LSPR corresponding to the Raman frequency is required to measure the Raman scattering signal sensitively. Therefore, various plasmonic structures (e.g., nanoparticles on mirrors [6], bowties [7,8], nanoparticles [9], and tip-enhanced plasmonic structures [10–12]), each with unique benefits for engineering LSPR, have been investigated extensively. Among these plasmonic structures, a one-dimensional (1D) nanogap [13] is a highly excited LSPR that strongly confines light, which effectively increased the Raman scattering signal [14]. Moreover, a 1D nanogap enables large-area sensing because the gap plasmonic structure is extended one-dimensionally to the excitation beam spot, whereas the sensing area of the other gap plasmonic structure is extremely small for large-area sensing.

In this study, we demonstrate an effective nanogap-enhanced Raman scattering platform to enhance the Raman scattering signal using a 1D Au nanogap. We performed finite-difference time-domain (FDTD) simulations of the 1D Au nanogap (gap width = 20 nm) to investigate the enhanced field distribution and scattering spectra of the 1D Au nanogap. We measured the Raman spectra of a WSe₂ monolayer and brilliant cresyl blue (BCB) molecules via a 1D Au nanogap.

The 1D Au nanogaps were loaded onto a three-axis positioning stage (XYZ Linear Stage, M-562-XYZ, Newport) for XY scanning. To obtain a spatially coherent excitation beam, a He-Ne laser (594.5 nm, < 1.0 mW) was coupled and passed through a single-mode fiber (core diameter of ∼3.5 µm) and collimated again using an aspheric lens. Finally, the beam was focused on the sample using a microscope (NA = 0.8; LMPLFLN100X). The Raman responses were collected using the same microscope objective (backscattering geometry) and passed through an edge filter (FEL0550) to eliminate the fundamental laser line. The Raman signals were dispersed onto a spectrometer (f = 328 mm, Kymera 328i) and imaged using a thermoelectrically cooled chargecoupled device (CCD, iDus 420) to acquire the Raman spectra. Prior to the experiment, the spectrometer was calibrated using an argon–mercury lamp. A 300 g/mm grating blazed to 500 nm (spectral resolution of 0.31 nm) was used for the Raman measurements. For the plasmon-enhanced Raman scattering experiment, BCB molecules were prepared by spin coating (3,000 rpm for 20 s) an ethanol solution (10 mM) on the 1D Au nanogap, and the WSe₂ monolayer was transferred onto the 1D Au nanogap.

A polyethylene terephthalate (PET) substrate was cut to a thickness of 250 µm with a size of 20 mm × 20 mm. A 110-nanometer-thick gold thin film was deposited on the PET substrate using an e-beam evaporator at a deposition rate of 0.3 Å/s. The Au film on the PET substrate was patterned using photolithography with a photoresist (PR, AZ5241e). Furthermore, a 150-nanometer-thick vanadium thin film was deposited onto the patterned photoresist, and a vanadium pattern was developed using acetone via a lift-off process. The gold film was patterned via ion milling, and a second gold thin film with a thickness of 100 nm was deposited on the sample. Plugging the patterned exposed area with gold created a 1D nanogap between the metals. The sacrificial mask was chemically etched to flatten the ground and eliminate any remaining parts of the sample.

We performed FDTD simulations (Lumerical Solutions, Inc.) to quantify the enhancement of the optical field and scattering resonance at the 1D Au nanogap. The nanogap was modeled as 20 nm wide and 110 nm deep. Furthermore, we used a broadband source covering the visible and near-infrared frequency regions to estimate the scattering resonance.

3.1. Pre-characterizations of the 1D Au nanogap

Figure 1(a) illustrates the plasmon-enhanced Raman scattering of the WSe₂ monolayer and BCB molecules through the 1D Au nanogap. The Raman scattering signal in the Au nanogap was significantly enhanced via incident light confinement and a substantially excited LSPR. Figure 1(b) shows the scanning electron microscopy images of the 1D Au nanogap (scale bar is 1 µm).

Figure 1. Enhanced Raman scattering via a 1D Au nanogap. (a) Illustration of the enhanced Raman scattering for a WSe2 monolayer and BCB molecules on the 1D Au nanogap. (b) Scanning electron microscopy images of the 1D Au nanogap.

3.2. Electromagnetic simulations of the 1D Au nano-gap

We performed FDTD simulations to investigate the optical characteristics of the 1D Au nanogap, which exhibited a sufficiently excited LSPR. Electromagnetic simulation was performed using perpendicularly polarized incident light at the nanogap (gap width = 20 nm) to investigate the optical field distribution, which significantly enhanced the Raman signal, as shown in Fig. 2(a). Because the incident light was polarized perpendicular to the nanogap axis, all the electrons in the nanogap oscillated owing to the optical field of the incident light. Therefore, we used light polarized perpendicular to the nanogap axis for the simulation. The optical field distribution between the nanogaps showed strong Raman enhancement with an excitation laser wavelength of 594 nm. We simulated the LSPR spectrum that emerged at a 20 nm gap distance in the 1D nanogap. The LSPR varies with the gap distance of the 1D Au nanogap because the dipole–dipole interaction at the nanogap weakened as the gap distance decreased. Figure 2(b) shows the resonance of the LSPR spectrum when 594 nm was incident on the nanogap, which overlapped with the Raman frequency of the BCB molecules at an excitation wavelength of 594 nm. Therefore, the 1D Au nanogap was sufficient for sensitively analyzing the WSe₂ monolayers and BCB molecules.

Figure 2. Simulation of the optical properties of the 1D Au nanogap. (a) Finite-difference time-domain simulation of the optical field distribution at the 1D nanogap (gap distance = 20 nm). (b) Scattering spectrum of the 1D Au nanogap.

3.3. Plasmon-enhanced Raman scattering of a WSe₂ monolayer

We conducted an experiment to measure the enhanced Raman scattering of a WSe₂ monolayer on a 1D Au nanogap to demonstrate that the 1D Au nanogap was sufficient for the sensitive measurement of Raman scattering. Figure 3(a) shows the spectral Raman image of the WSe₂ monolayer in the 1D Au nanogap. The Raman peak (v ≈ 251 cm⁻¹) intensity of the WSe₂ monolayer at the nanogap exhibits enhanced intensity owing to the localization of the incident light, as indicated in Fig. 2. Because the beam spot area of the incident light was larger than the nanogap width, we measured the enhanced Raman signal throughout an area larger than the nanogap (gap width= 20 nm). Using 2D materials for Raman scattering measurements, the field enhancement emerged throughout the 1D Au nanogap. Thus, a 1D Au nanogap efficiently enhances a Raman scattering signal by effectively exciting the LSPR, which is crucial for detecting weak Raman signals, as shown in Fig. 3(b). Because the WSe₂ monolayer was placed on the Au nanogap, the enhanced optical field strongly enhanced the Raman scattering signal of the WSe₂ monolayer, as shown in Fig. 3(c). This enabled us to measure an enhanced Raman spectrum with a laser excitation power of 400 µW and an acquisition time of 10 s, as shown in Figs. 3(b) and 3(c).

Figure 3. Spectral Raman imaging of the WSe₂ monolayer. (a) Raman peak (ν ≈ 251 cm⁻¹) intensity image of the WSe₂. (b) Raman spectrum of the WSe₂ at the Au film. (c) Raman spectrum of the WSe₂ at the nanogap.

3.4. Plasmon-enhanced Raman scattering of BCB molecules

To investigate the Raman enhancement effect in 1D nanogaps for various materials, chemical molecules such as BCB molecules were measured. The BCB molecules were spin-coated onto the 1D Au nanogaps. Because the Raman scattering cross section was insignificant, the intensity of the Raman signal outside the nanogap was weak, as shown in Fig. 4(a). A 1D Au nanogap was used to efficiently excite the LSPR and confine the incident light, which enhanced the Raman scattering signal. Figure 4(b) shows the Raman spectrum of BCB molecules in a 1D Au nanogap. The enhanced optical field induced by the LSPR significantly enhanced the Raman scattering signals of the BCB molecules. Using a 200 µW laser excitation power and 10 s acquisition time, the Raman spectrum of the BCB molecules enhanced through the enhanced optical field could be measured.

Figure 4. Enhanced Raman spectra of the BCB molecules outside the nanogap (a) and at the nanogap (b).

In this study, we discuss an efficient plasmonic platform that allows the sensitive analysis of various materials owing to an excited LSPR and confinement of incident light. We investigated the optical properties of a 1D Au nanogap, such as the intensity of the optical field, field distribution, and scattering spectra, using electromagnetic simulations. Additionally, the sensitivities of the Raman spectra of a WSe2monolayer and the BCB molecules in the 1D Au nanogap were determined. A 1D Au nanogap enables a sensitive analysis of various materials. Moreover, 1D Au nanogaps combined with a flexible substrate can selectively enhance the different Raman modes of molecules on the same sample using the gap-distance dependency of LSPR.

This study was supported by National Research Foundation of Korea (NRF) grants (No. 2020R1C1C1011301 and 2021R1A6A1A10042- 944).

The authors declare no conflicts of interest.