2.1 Pyramidal pitted silicon preparation

First, the <100> polished silicon coated with 300?nm SiO₂ was cleaned on the surface, dried with nitrogen, and baked in the oven at 90?°C to remove the moisture in the silicon. The appropriate amount of positive photoresist drops is added to the center area of the silicon wafer. Then, a three-stage spin coating process was applied successively for 5?s at 500?rmp, 30?s at 4,000?rmp, and another 5?s at 500?rmp. The silicon wafer coated with the photoresist was then baked again at a constant temperature to ensure that all solvent in the photoresist fully volatilized.

After natural cooling to room temperature, ultraviolet lithography exposure machine with mask plate exposure was performed. Subsequently, development took place in a solution containing 3?‰ sodium hydroxide followed by thorough cleaning with deionized water to complete fixing. The silicon wafer was then baked at a constant temperature for 20–30?min. Afterward, BOE solution (buffer oxide etching solution, 40?% NH₄F aqueous solution to 49?% HF volume ratio 7:1) was used to etch the exposed SiO₂ on the silicon wafer. Once etching was complete, acetone was used to remove any remaining photoresist from the surface before immersing it into heated KOH solution (12.5?%) in a water bath set between temperatures of 55–60?°C for anisotropic wet etching until a pyramidal pit structure had formed on the silicon wafer (reaction time of approximately 40?min?for every 10?μm). Finally, the silicon wafer underwent immersion in BOE solution once more to remove any remaining SiO₂ mask from its surface, resulting in the production of PPSi substrates.

2.2 Synthesis process of Au@Ag NSs

The seed growth method is applicable for the preparation of various types of nanomaterials, allowing for precise control over the size and morphology of nanoparticles [34]. The fabrication of Au-core Ag-shell nanostructures involves a stepwise process of seed growth and oxidative etching to achieve the desired core–shell structure and shape tuning.

Initially, 0.6?mL of strong reducing agent NaBH₄ (10?mM) was used in a CTAB system (10?mL, 100?mM) to reduce the precursor HAuCl₄ (0.25?mM), resulting in the aggregation of 0-valent Au atoms into nanoclusters (3?nm). Subsequently, 1.5?mL of weak reducing agent AA (100?mM) was employed to further stabilize the reduction of chloroauric acid (2?mL, 0.5?mM) on the clusters, leading to the growth of 10?nm Au seeds. The solution containing these 10?nm Au seeds was then collected through centrifugation and dispersion [35].

Then, in the CTAC (20?mL, 100?mM) system, the reducing agent AA (1.3?mL, 10?mM) continued to slowly reduce HAuCl₄ (20?mL, 0.5?mM) on the Au seed surface, resulting in uniform growth of Au. Surfactant CTAC can be used as a soft template to guide Au along specific crystal faces, ultimately forming a regular cubic morphology [36]. To reduce the introduction of other ions into the system, HAuCl₄ (250?μL, 10?mM) was used as the oxidant to preferentially etch the sharp edges and corners of the nanocubic, converting the Au nano cubes into 80?nm diameter highly spherical nanoballs.

In the final step, the reducing agent AA (2.5?mL, 100?mM) was utilized to gradually reduce the precursor AgNO₃ (0.5?mM, 50?mL), resulting in the formation of Ag atoms that deposited onto the uniform nano Au nucleus. Similarly, within the CTAC (50?mL, 20?mM) system, chloride ions effectively shielded the <100> facet, allowing for Ag growth along the Si <111> facet and leading to nanostructures with a regular cubic shape. By employing hydrogen chloride at a concentration of 0.2?% as an etching agent, the Au@Ag nano cubes underwent transformation into Au@Ag NSs. The degree of etching on the Ag shell depended solely on the concentration of the etching agent and not on time duration; this method provides effective control over the etching process [37].

2.3 Self-assembly of Au@Ag NSs monolayer film

The prepared Au@Ag NSs solution was centrifuged and then dispersed into a 1?% mass ratio PVP ethanol solution to convert CTAC-coated Au@Ag NSs into PVP-coated ones. After another round of centrifugation and dispersion in ethanol, the nanoparticles were placed in a CH₂Cl₂ oil system in a centrifuge tube, with an appropriate amount of deionized water added. Following shaking and oscillating, a loosely packed metal film formed at the interface. Subsequently, hexane was added to create an oil–water–oil three-phase system. By tilting and rotating the centrifuge tube on a small scale, the nanoparticles spontaneously move to the hexane/water interface due to interfacial tension gradients, resulting in a dense monolayer film with bright metallic luster. This film was then transferred onto PPSi substrates for assembly into composite SERS substrates.

2.4 Substrate characterization

The surface morphology and size distribution of the pyramidal pitted silicon substrate and the Au@Ag nanoparticle structure were characterized using a scanning electron microscope (SEM, Zeiss AURIGA) at an accelerating voltage of 5?kV. The composition and elemental content of the substrate microregion were analyzed using an X-ray energy dispersive spectrometer (EDS, Oxford X-Max) at an accelerating voltage of 20?kV. The prepared nanoparticles were respectively drop-cast on a TEM grid to perform high-resolution morphology characterization and EDX element mapping using a transmission electron microscope (TEM, FEI Talos F200X G2) operating at 200?kV. Raman mapping was carried out by a confocal microprobe Raman system (Horiba Odyssey, 785?nm). The Raman scattering spectrum was measured using a portable Raman spectrometer (Optosky ATR3000, excitation wavelength: 785?nm, adjustable excitation power: 0–500?mW).

2.5 Bacterial sample culture and SERS detection

The preserved strain was inoculated onto Luria-Bertani (LB) slope culture medium for preculture to obtain the first-generation strain for further experiments. Subsequently, the cultured bacteria were inoculated into LB liquid culture medium and shaken at 220?rpm on a shaker at 37?°C for 24?h. Following this, the cultured liquid was centrifuged to remove the supernatant, and 5?mL of physiological saline was added for a second centrifugation to wash the bacteria. Afterward, 2?mL of bacterial suspension was prepared using physiological saline, and 300?μL was added to a spectrophotometer cuvette and diluted tenfold. The OD value of the bacterial suspension at 600?nm was measured to be 0.287 using a spectrophotometer, indicating that the initial undiluted concentration of S. aureus was 2.87 × 10⁹?CFU/mL. The bacterial suspension at concentrations ranging from 10⁴ to 10⁸?CFU/mL was then stepwise diluted. Following this, drops containing 20?μL of the 10⁴ − 10⁸?CFU/mL S. aureus suspension were placed on to the prepared composite SERS substrate and allowed to dry naturally at room temperature. SERS spectra were collected using a portable Raman spectrometer (wavelength 785?nm, intensity 100?mW, integration time 2000?ms) after the samples were dried. Five points were randomly selected from each sample for detection to obtain the average value as the final result. At the same time, the blank sample prepared by adding 20?μL of physiological saline to the composite SERS substrate was collected under the same conditions.

2.6 Detection of Raman reporters by SERS substrate

The SERS substrates were evaluated for enhancement and uniformity using the Raman reporter molecule R6G. A 10⁻¹?M R6G solution was prepared and stepwise diluted to obtain a low concentration solution of 10⁻⁴ − 10⁻⁹?M. The bare silicon wafer was treated with a 20?μL 10⁻¹?M R6G solution, dried at room temperature, and then analyzed using a portable Raman spectrometer with an excitation wavelength of 785?nm, excitation intensity of 100?mW, and integration time of 5,000?ms as the control group. Similarly, a 20?μL solution of 10⁻⁴ − 10⁻⁹?M R6G was applied to the prepared SERS substrate, dried naturally at room temperature, and then analyzed using a portable Raman spectrometer under the same conditions. To ensure reliable results, spectra were collected from five random points each time and averaged for the final result.

3.1 Optimization and characterization of PPSi substrates

The crystal structure of silicon is a cubic crystal system, where the most common crystal directions are <100> and <111>. Due to the difference of crystal structure, the reaction rate of <100> and <111> crystalline orientation silicon to hydroxide ion is different, which leads to the anisotropic etching phenomena. Generally, the <100> crystal direction reacts more readily with hydroxide ions compared to the <111> crystal direction, resulting in a faster etching rate. As shown in Figure 2, the angle between the <111> plane and the <100> plane of the <100> silicon wafer is 54.74° [38]. When a mask with regularly arranged square holes is applied on the surface of silicon using ultraviolet lithography, a PPSi structure with regular arrangement and uniform shape is formed on the silicon wafer through anisotropic etching process.

Figure 2:

Schematic diagram showing the angle of <100> silicon wafer direction.

Figure 3(a) and (c) show the SEM images of the PPSi structures after etching for different periods. It can be observed that the silicon wafer is etched along the <111> crystal orientation, and the four <111> crystal facets extend downward and converge to form regular pyramid-shaped pits. The PPSi structure array was evenly distributed, with intact pit surfaces and complete structures, where the edges, vertices, and bottom edges were clearly defined, indicating a satisfactory etching effect. The Nano Measure 1.2 software was used to measure the bottom side lengths of 50 randomly prepared PPSi structures and calculate the size distribution. As depicted in Figure 3(b) and (d), the size distribution of the PPSi array is remarkably uniform, with a very small relative standard deviation (RSD) of the side lengths (all within 2?%), which indicates that the uniformity of the prepared PPSi array is excellent.

Figure 3:

SEM picture of the PPSi substrates. (a) PPSi prepared by a 10?μm square hole with a spacing of 10?μm mask. (b) The size distribution histogram of PPSi prepared by a 10?μm square hole with a spacing of 10?μm mask. (c) PPSi prepared by a 20?μm square hole with a spacing of 10?μm mask. (d) The size distribution histogram of PPSi prepared by a 20?μm square hole with a spacing of 10?μm mask.

However, there was a reduction in boundary width after etching compared to the mask size, leading to a corresponding increase in the length of the pyramid base while maintaining an unchanged overall period. For instance, a square mask with a side length of 10?μm results in a 13?μm base for the pyramid and reduces the boundary from 10?μm to 7?μm. Similarly, a PPSi etched from a square mask with a side length of 20?μm had a base edge of 24?μm and reduced the boundary from 10?μm to 6?μm. This phenomenon can be attributed to potassium hydroxide solution corroding silica simultaneously as it etches silicon but at a slower rate.

If the boundary mask is reduced so that the etching amount of the silica mask reaches a critical value, the width of the boundary platform will be significantly reduced, and the slope area of the PPSi array will be further increased as it becomes approximately boundary free. The size of the mask plate was optimized based on actual experimental data. The spacing between 2?μm and 4?μm square holes was adjusted to 2?μm (due to limitations in the mask plate preparation process), while the spacing between 10?μm square holes was set to 3?μm, and the spacing between 20?μm square holes was set to 4?μm. Figure 4 depicts the SEM image of a base prepared using this optimized mask plate, showing that the base length (boundary width) of pyramids measures at 3.5?μm (0.5?μm), 5.6?μm (0.4?μm), 11.9?μm (1.1?μm), and 22.6?μm (1.4?μm). Additionally, corresponding pyramids depths were measured at 2.46?μm, 4?μm, 8.4?μm, and 16?μm, respectively. At this point, the boundary width has been significantly reduced, yet the etching effect on the boundaries remains satisfactory, with sharp and clear edges without any collapse. The size and morphology of the PPSi array exhibit good uniformity.

Figure 4:

SEM picture of the PPSi substrates. (a) PPSi prepared by a 2?μm square hole with a spacing of 2?μm mask. (b) PPSi prepared by a 4?μm square hole with a spacing of 2?μm mask. (c) PPSi prepared by a 10?μm square hole with a spacing of 3?μm mask. (d) PPSi prepared by a 20?μm square hole with a spacing of 4?μm mask (the illustrations show the histogram of the size distribution corresponding to the PPSi of different sizes).

3.2 PPSi substrates with deposition of Au@Ag nanoparticles characterization

Au@Ag NSs were assembled into a highly homogeneous monolayer film using the liquid–liquid interface self-assembly technique. This film was then constructed on a PPSi substrate, utilizing the Marangoni effect. The result was a 3D SERS substrate (Au@Ag NSs/PPSi) with tunable shape and position.

Before interfacial self-assembly, the prepared nanoparticles need to undergo ligand exchange. Due to the amphoteric nature of CTAC and its interaction with both aqueous and organic phases, it is challenging for CTAC-coated Au@Ag NSs to complete self-assembly directly at the interface. Therefore, PVP-coated Au@Ag NSs need to be prepared through a ligand exchange process.

In a water–oil system where the oil phase density is higher than that of the water phase, dichloromethane and water were found to be immiscible, causing the nanoparticles to remain in the organic oil phase. Consequently, they were adsorbed at the interface of the two phases, forming a random, nontightly packed single-layer metal film. However, upon adding another oil phase such as n-hexane with a lower density than water, driven by the interfacial tension gradient of the oil–water–oil three-phase system, the nanoparticles can spontaneously climb from the lower interface to the upper interface. This allows for their transfer and compression, resulting in a single layer of metal nanoparticles with close packing state distribution and bright metallic luster. The membrane assembled by Au@Ag NSs was transferred and constructed on PPSi to obtain 3D SERS substrate. The SEM images in Figure 5 showed composite substrates assembled with PPSi substrates of different sizes. It can be observed from these images that Au@Ag NSs are effectively attached to the PPSi substrate with regular morphology and roughly uniform distribution. Overall, this indicates that composite substrate morphology was excellent for further study purposes.

Figure 5:

SEM picture of the PPSi substrates with deposition of Au@Ag nanoparticles. (a) PPSi prepared by a 2?μm square hole with a spacing of 2?μm mask. (b) PPSi prepared by a 4?μm square hole with a spacing of 2?μm mask. (c) PPSi prepared by a 10?μm square hole with a spacing of 3?μm mask.

In Figure 6(a) and (b), the prepared Au@Ag nanocubes and Au@Ag NSs were depicted. The Nano Measurer 1.2 software was utilized to measure both random diameter and size distribution for 100 prepared nanospheres as shown in Figure 6(c). Statistical calculation determined that these nanospheres had an average diameter of 85?nm when assembled on a composite substrate with an RSD?of 5.88?%. This uniform size distribution indicated consistency among prepared nanospheres. To verify the Au@Ag core–shell structure, high-resolution morphology characterization and elemental mapping using a FEI Talos F200X G2 TEM system operating at 200?kV were performed. The core–shell structure of nanoparticles can be clearly characterized, and the element mapping visually demonstrates the nanostructure of gold core silver shells in Figure 6(d). Figure 6(e) and (f) further reveal the EDS spectral line to Au@Ag NCs and Au@Ag NSs, a comparison of the ratio of Ag to Au content between nanocubes revealed that after etching there was a decrease in weight ratio due to etching away of sharp edges from the Ag shell.

Figure 6:

Characterization of Au@Ag nanoparticles. (a) The SEM image of Au@Ag nano cubes. (b) The SEM image of Au@Ag NSs. (c) The size distribution histogram of Au@Ag NSs. (d) The TEM image and elemental mapping of Au@Ag NSs. (e) The EDS spectral line of Au@Ag nano cubes. (f) The EDS spectral line of Au@Ag NSs.

3.3 Enhance performance and uniformity of the SERS substrate

In Figure 7(a), the distinct Raman responses of Si wafer, PPSi, and Au@Ag NSs/PPSi were clearly observed. The peak at around 523?cm⁻¹ is attributed to the silicon lattice vibration. In comparison to the untreated Si wafer, the Raman spectrum of PPSi exhibits a significant enhancement in the intensity of the 523?cm⁻¹ peak, which was twice as strong as that of the silicon wafer. For the 3D composite SERS substrate, characteristic peaks at 241, 1,006, 1,238, and 1,397?cm⁻¹, attributed to the metal nanoparticles (Au@Ag NSs), were detected in the spectrum. Simultaneously, due to reduced exposed silicon area after attachment of Au@Ag NSs, there is a decrease in intensity of the 523?cm⁻¹ peak. We applied CV solutions at concentrations of 1,000?ppm onto silicon wafer and 0.1?ppb onto a 3D composite SERS substrate, respectively. The major characteristic peaks of CV molecules at 1,589 cm⁻¹ and 1,623?cm⁻¹, which correspond to the in-plane stretching vibrations of the C–C bond, are used as subjects of study typically. As shown in Figure 7(b), the characteristic peaks of 0.1?ppb CV on the 3D composite SERS substrate can be easily captured, whereas the peak intensity of 1,000?ppm on the silicon wafer remained particularly weak.

Figure 7:

SERS spectra of SERS substrate with different probe molecules. (a) Raman spectra of Si, PPSi, and Au@Ag NSs/PPSi. (b) Raman spectra of 1,000?ppm CV and SERS spectra of 0.1?ppb CV. (c) Raman spectra of 10⁻¹?M R6G solution. (d) SERS spectra of 10⁻⁴?M R6G on Au@Ag NSs/flat Si and Au@Ag NSs/PPSi.

Subsequently, we selected Rhodamine 6G (R6G) as a probe molecule to thoroughly investigate the enhancement performance of the Au@Ag NSs/PPSi composite substrate. And 1?mL 10^−1?M Rhodamine 6G (R6G) solution was used as the standard, and the characteristic Raman peak of R6G in the region of 300–2000?cm⁻¹ was observed. In Figure 7(c), the peak at 623?cm⁻¹ corresponds to the C–C–C in-plane vibration of the R6G molecule, while the peaks at 790 and 1,190?cm⁻¹ can be attributed to the out-of-plane vibration and in-plane vibration of the C–H bond, respectively. Additionally, peaks observed at 1362, 1514, and 1,654?cm⁻¹ can be classified as aromatic C–C tensile vibration modes [39]. As shown in Figure 7(d), the SERS signal of Au@Ag NSs/PPSi was more than ten times that of Au@Ag NSs/flat Si. The comparison confirmed the signal amplification capability of PPSi array structure and its effective effect on SERS performance.

As shown in Figure 8(a) and (c), when the concentration of R6G test solution decreased from 10⁻⁴?M to 10⁻⁹?M, the SERS intensity of R6G molecules gradually decreased with it. However, even at a concentration of 10⁻⁹?M, the characteristic peak of R6G could still be detected on the composite SERS substrate with the height of 4?μm (Figure 8(a)) and 8.4?μm (Figure 8(c)). As depicted in Figure 8(b), for the 4?μm 3D composite SERS substrate, a good linear SERS dependence was observed between SERS intensity at 1,514?cm⁻¹ and R6G concentration ranging from 10⁻⁴?M to 10⁻⁹?M showing a high linear regression fitting (R² = 92.95?%). Similarly, as shown in Figure 8(d)–(f), for the 8.4?μm SERS substrate, the peak intensities at 1,362?cm⁻¹, 1,514?cm⁻¹, and 1,654?cm⁻¹ showed linear regression R² values of 94.89?%, 93.2?%, and 95.9?%, respectively, in relation to the R6G concentration. These results demonstrated that the 3D composite SERS substrates can effectively predict analyte concentrations and enable quantitative analysis of molecular quantities.

Figure 8:

SERS spectra of R6G and  calibration curves  of Raman intensity versus R6G concentration. (a) SERS spectra with R6G concentration of 10⁻⁴ − 10⁻⁹?M on the 4?μm composite SERS substrate. (b) The calibration curves of Raman intensity versus R6G concentration (10⁻⁴ − 10⁻⁹?M) at 1,514?cm⁻¹ on the 4?μm composite SERS substrate. (c) SERS spectra with R6G concentration of 10⁻⁴ − 10⁻⁹?M on the 8.4?μm composite SERS substrate. (d) The calibration curves of Raman intensity versus R6G concentration (10⁻⁴ − 10⁻⁹?M) at 1,362?cm⁻¹ on the 8.4?μm composite SERS substrate. (e) The calibration curves of Raman intensity versus R6G concentration (10⁻⁴ − 10⁻⁹?M) at 1,514?cm⁻¹ on the 8.4?μm composite SERS substrate. (f) The calibration curves of Raman intensity versus R6G concentration (10⁻⁴ − 10⁻⁹?M) at 1,654?cm⁻¹ on the 8.4?μm composite SERS substrate.

In order to quantitatively characterize the enhancement performance of the SERS substrate, the enhancement factor of SERS substrate was calculated using formula as follows:

Here, the I_SERS represents the intensity of 10⁻⁹?M R6G obtained from the SERS spectrum at 1,362?cm⁻¹, while I_normal denotes the intensity of the normal Raman spectrum of 10⁻¹?M R6G at 1,362?cm⁻¹. C_SERS and C_normal refer to concentrations in SERS and conventional Raman measurements, respectively. The results indicated that the maximum enhancement factor of the SERS substrate with a height of 8.4?μm was found to be 4.2 × 10⁸.

In addition to enhancing the effect, the SERS active substrate also requires good uniformity and reproducibility of SERS signal [40]. To more accurately evaluate the uniformity of Raman signal distribution, a 20?μl 10⁻⁵?M R6G solution was added to the composite substrate and air-dried at room temperature. SERS spectra of 10 random points were collected under a unified detection environment. The probe focus can move freely on the surface of the substrate, and the spectral trend and peak shape of all points are consistent in Figure 9(a). The uniformity of spectral data reflects the distribution and performance of nanoparticles on the substrate. The RSD?of the characteristic peak intensity at the corresponding point 1,654?cm⁻¹ is 5.26?% (Figure 9(b)). The Raman mapping taken from a 250?μm × 250?μm area was used to evaluate the signal repeatability. As shown in Figure 9(c), the uniform color of the mapping spectrum at 1,654?cm⁻¹ using 10⁻⁴?M R6G as the probe molecule indicates the excellent signal homogeneity of the Au@Ag NSs/PPSi substrate over a large microscopic area in the detection. Figure 9(d) exhibits the signal intensity deviation of the peak at 1,654?cm⁻¹ calculated by using the spectra from the SERS mapping shown in Figure 9(c). More than 70?% of the sites exhibit intensities within the range of I_average ± σ, while over 95?% of the sites have intensities falling within the range of I_average ± 2σ. The RSD?of SERS signals is evaluated to be 8.6?%, displaying a good spectral uniformity of the sensor in the microscopic region. The good uniformity of the SERS signal can be attributed to the periodic PPSi array structure and the well-distributed Au@Ag NSs. Homogeneous distribution of nanoparticles on the surface of PPSi enables uniform distribution of hot spots, allowing for effective oscillation by incident laser in creating high-density three-dimensional active hot spots within PPSi troughs. The Au@Ag NSs/PPSi composite structure provided a uniform spatial distribution for analyte molecular detection at designated hot spots.

Figure 9:

Spectral uniformity characterization of SERS substrate. (a) Randomly select 10 different positions on the SERS substrate to record the SERS signal of 10⁻⁵?M R6G. (b) The RSD?value of the corresponding signal calculated from the characteristic peak of 1,654?cm⁻¹. (c) SERS mapping at 1,654?cm⁻¹ for 10⁻⁴?M R6G recorded from the Au@Ag NSs/PPSi substrate. (d) SERS intensity deviation at the 1,654?cm⁻¹ calculated with the SERS mapping.

3.4 Application in the detection of pathogenic bacteria

S. aureus samples were utilized to validate the potential for pathogen detection and the efficacy of Raman enhancement. In comparison to the control group, a distinct peak at 1,145?cm⁻¹ corresponding to S. aureus was identifiable [4]. As the concentration of S. aureus increased from 10⁴?CFU/mL to 10⁷?CFU/mL, there was a significant rise in SERS signal intensity with increasing bacterial solution concentration, as depicted in Figure 10(a). Notably, the characteristic peak of S. aureus remained detectable even at a concentration of 10⁴?CFU/mL. The SERS dependence on bacterial concentration between 10⁴?CFU/mL and 10⁷?CFU/mL exhibited strong linearity, with a high linear regression fitting R² value of 99.7?% (Figure 10(b)).

Figure 10:

SERS spectra of S. aureus and calibration curves of Raman intensity versus S. aureus. (a) SERS spectra with S. aureus concentration of 10⁴–10⁷?CFU/mL on the composite SERS substrate. (b) The calibration curve of Raman intensity versus S. aureus concentration (10⁴–10⁷?CFU/ml) at 1,145?cm⁻¹ on the composite SERS substrate.

3.5 FDTD simulation

To better understand the enhancement mechanism and behavior of the Au@Ag NSs/PPSi composite SERS substrate, finite difference time domain (FDTD) simulation was conducted using Ansys Lumerical software to calculate the local enhanced electric fields of Au@Ag NSs/Si and Au@Ag NSs/PPSi. The size of the micro–nano structure in all simulations was determined based on statistical results from experimental measurements. For all array calculations, the background refractive index was set to the refractive index of air. Periodic boundary conditions were applied to the x and y axes, while PML boundary conditions were used for the z axis. The light source utilized was a plane wave with a wavelength of 785?nm.

Au@Ag NSs with a diameter of 85?nm and a spacing of 2?nm placed on a Si plate was simulated for comparison in Figure 11(a), while the electric field properties at the top, middle, and bottom of the PPSi attached Au@Ag NSs (diameter of 85?nm, spacing of 2?nm) with the height of 8.4?μm were simulated and recorded as the Figure 11(b) shows. It can be seen clearly that local electric field enhancement of Au@Ag NSs/Si is located between nanospheres, where EM enhancement |E/E₀|⁴ was as high as 10⁶. As shown in Figure 11(c), the incident laser generates effective oscillations within the PPSi, which exhibit a strong localized electric field enhancement within the PPSi valley, presenting as a band-like distribution in the x − z view. Regarding the enhancement mechanism of the 3D composite SERS substrate, a uniform and locally enhanced electromagnetic field hot spot is generated between Au@Ag NSs. Additionally, multiple oscillations of the incident laser within the PPSi valleys result in optical field amplification, creating a high-density three-dimensional active hot spot. The EM enhancement |E/E₀|⁴ reached as high as 10⁸, which was mutually corroborated with the measured results of AEF. The Au@Ag NSs/PPSi composite SERS substrate provided a uniform spatial distribution for the detection of analyte molecules within the specified hotspot areas.

Figure 11:

Au@Ag NSs/PPSi composite substrate FDTD simulation model and electromagnetic field distribution. (a) Au@Ag NSs/Si simulation model and electromagnetic field distribution. The intensity of the electromagnetic field |E/E₀|⁴ was represented by color scale. (b) Au@Ag NSs/PPSi SERS substrate FDTD simulation model. (c) Electromagnetic field distribution of composite SERS substrate with a height of 8.4?μm.