Polarized SERS Controlled by Anisotropic Growth on Ordered Curvature Substrate

Colloidal lithography is an efficient and low-cost method to prepare an ordered nanostructure array with new shapes and properties. In this study, square-shaped and cone-shaped Au nanostructures were obtained by 70° angle deposition onto polystyrene bead array with the diameter of 500 nm when a space of 120 nm is created between the neighbor beads by plasma etching. The gaps between the units decrease when the Au deposition time increases, which leads to the polarized enhanced local field, in agreement with the surface-enhanced Raman scattering spectra (SERS) observations and finite-difference time-domain (FDTD) simulations. When the Au deposition time increased to 5 min, 5 nm gaps form between the neighbor units, which gave an enhancement factor of 5 × 109. The SERS chip was decorated for the detection of the liver cancer cell marker Alpha-fetoprotein (AFP) with the detection limit down to 5 pg/mL.


Introduction
Plasmonics is one of the important research fields with the focus on the interactions among photons, electrons, and metal nanostructures, in which many interesting phenomena are observed due to the regulated collective movements of the electrons excited by the incident light, such as the abnormal transmission of light [1][2][3], surface-enhanced spectroscopy [4,5], and localized heating on the nanometer scale, and so on [6]. Among these observations, the surrounding media, the chemical compositions and microstructures, and the nanostructure patterns play very important roles in the properties [7,8]. To drive the light to go along the expected route, the dielectrics and composition distributions of the surrounding conditions are designed by Zhu's group, and the interaction of the media materials confines the light movement in a certain area like a black hole [9]. When the metal nanostructures are prepared as the core coated by other materials, the surface plasmons can be confined around the core by the coverage layer, which tends to work as the controllable hotspot for high-sensitivity detections and is used in various fields for detection [10][11][12][13]. In addition to the composition and the surrounding conditions, the nanostructures with different shapes show some exciting properties and applications. To investigate the interesting behaviors of electrons and photons in metal nanostructures, the various shapes of metal nanostructures are designed in an ordered pattern, which are proposed to regulate the electron movements and the interactions [14][15][16][17]. In fact, the polarization effect caused by the anisotropic shape finds very important applications in plasmonic sensors for the self-reference effects [18,19]. In these patterned materials, the metal units give contributions to the enhanced local field besides the gaps, owing to the coupling surface plasmon excited by the light, which depends on the angle between the polarization direction of the electric field and the normal direction of the gap [20]. Among these techniques, nanosphere lithography is a simple process for ordered nanostructure fabrication, based on how many When Au film is deposited onto the PS array vertically, that is, along the normal direction, the Au nanocaps form on the PS array. For the PS etched for 3 min, the Au nanocaps showed the protruding parts due to the fast growth rate of Au film around the protruding parts induced by the large local strain, as shown in Figure 1d. For the PS array etched for 3.5 min, the Au nanocaps showed uniform growth along the PS surface, and no protruding parts were observed, as shown in Figure 1e. To get the nanostructures of the different shapes, the angle deposition was applied, which means the deposition deviated from the vertical direction, as shown in Figure S1. Generally, the curved surfaces induced a fast growth along the film deposition direction and a slow lateral growth, which led to the formation of nanorods for long-time depositions, as reported due to the preferential growth [23]. Although many hotspots were prepared between these nanorods, the beneath hotspots showed tiny contributions to the total SERS signals because they were far away from the sample surface and could not be detected. Therefore, the hotspots near the sample surface were preferred in practical applications. When the 500 nm PS array was etched for 3.5 min, the neighbor PS beads were separated completely, and the gaps between the neighbors were around 120 nm. Then the deposition angle was set as 70°, and the Au deposition was carried out.
During this deposition process, there were two kinds of possible nanostructures obtained on the PS array. One was the nanostructures of a square shape, and the other was a cone shape, as shown in figure 2. When the deposition was along the green line, as shown in figure 1b the film growth rate was fast along the deposition direction because the distance to the shadowing neighbor was very large, and the shadow effect was reduced greatly. Therefore, the film grew fast compared to the lateral growth. The lateral size was 400 nm, and the longitudinal size was 510 nm when the deposition time was 10 min, which led to the formation of the cone-shaped film, as shown in Figure 2a. When the film deposition continued to 15 min, the longitudinal size was measured as 510 nm, and the lateral size was 500 nm. The lateral growth decreased the gap sizes between the neighbor cones, which increased the shadow effects from the neighbors, and the longitudinal growth rate was reduced significantly. In comparison, the lateral growth increased, resulting in the structures of the hexagon-like shapes, as shown in Figure 2b. When the deposition was along the yellow line, as shown in Figure 1b, the film growth rates showed small differences between the deposition direction and the lateral direction because the similar When Au film is deposited onto the PS array vertically, that is, along the normal direction, the Au nanocaps form on the PS array. For the PS etched for 3 min, the Au nanocaps showed the protruding parts due to the fast growth rate of Au film around the protruding parts induced by the large local strain, as shown in Figure 1d. For the PS array etched for 3.5 min, the Au nanocaps showed uniform growth along the PS surface, and no protruding parts were observed, as shown in Figure 1e. To get the nanostructures of the different shapes, the angle deposition was applied, which means the deposition deviated from the vertical direction, as shown in Figure S1. Generally, the curved surfaces induced a fast growth along the film deposition direction and a slow lateral growth, which led to the formation of nanorods for long-time depositions, as reported due to the preferential growth [23]. Although many hotspots were prepared between these nanorods, the beneath hotspots showed tiny contributions to the total SERS signals because they were far away from the sample surface and could not be detected. Therefore, the hotspots near the sample surface were preferred in practical applications. When the 500 nm PS array was etched for 3.5 min, the neighbor PS beads were separated completely, and the gaps between the neighbors were around 120 nm. Then the deposition angle was set as 70 • , and the Au deposition was carried out.
During this deposition process, there were two kinds of possible nanostructures obtained on the PS array. One was the nanostructures of a square shape, and the other was a cone shape, as shown in Figure 2. When the deposition was along the green line, as shown in Figure 1b the film growth rate was fast along the deposition direction because the distance to the shadowing neighbor was very large, and the shadow effect was reduced greatly. Therefore, the film grew fast compared to the lateral growth. The lateral size was 400 nm, and the longitudinal size was 510 nm when the deposition time was 10 min, which led to the formation of the cone-shaped film, as shown in Figure 2a. When the film deposition continued to 15 min, the longitudinal size was measured as 510 nm, and the lateral size was 500 nm. The lateral growth decreased the gap sizes between the neighbor cones, which increased the shadow effects from the neighbors, and the longitudinal growth rate was reduced significantly. In comparison, the lateral growth increased, resulting in the structures of the hexagon-like shapes, as shown in Figure 2b. When the deposition was along the yellow line, as shown in Figure 1b, the film growth rates showed small differences between the deposition direction and the lateral direction because the similar shadow effects came from the shadowing neighbor on the right side, the left side, and the front side, as shown in Figure S2. When the deposition time was 10 min, the lateral size was 470 nm, and the longitudinal size was 480 nm along the deposition direction, as shown in Figure 2c, which means the shadow effects were similar around for this condition. For each Au cap, the front part also showed the decreased curvature of the surface due to the reduced shadow effect caused by the enlarged distances along the bead surface, which indicated that long-time growth led to a flat front surface. For the back part of Au caps, the self-shadow effect made the nanocap grow into a square-like shape, which means that the large distance led to fast growth and the small distance led to slow growth. The self-shadow effect led to square-shape nanostructures when the deposition time was 15 min, as shown in Figure 2d.
shadow effects came from the shadowing neighbor on the right side, the left side, and the front side, as shown in Figure S2. When the deposition time was 10 min, the lateral size was 470 nm, and the longitudinal size was 480 nm along the deposition direction, as shown in Figure 2c, which means the shadow effects were similar around for this condition. For each Au cap, the front part also showed the decreased curvature of the surface due to the reduced shadow effect caused by the enlarged distances along the bead surface, which indicated that long-time growth led to a flat front surface. For the back part of Au caps, the self-shadow effect made the nanocap grow into a square-like shape, which means that the large distance led to fast growth and the small distance led to slow growth. The self-shadow effect led to square-shape nanostructures when the deposition time was 15 min, as shown in Figure 2d. Raman spectra, respectively. In fact, EF means the ratio of the SERS signals to Raman signals for every single molecule on even. In our experiment, the laser spot was 1 μm in diameter and the penetration depth was 17 μm of the focused laser beam used; the density of 4-MBA was 1.345 g/cm −3 ; thus, the number of the 4-MBA molecule illuminated by the laser light was calculated to be 7.01 × 10 10 . NSERS was the number of surface-adsorbed molecules within the laser spot with the diameter (1 μm), and the surface area occupied by an adsorbed molecule could be adopted as 0.20 nm 2 /molecule. In comparison with the deposition along the normal direction, both Au square array and Au cone array showed an enhancement factor of 5 × 10 9 and 3 × 10 9 for the p polarization light, as shown in Figure  2, indicating the Au square array with excellent SERS performance due to the narrow gaps ( Figure 3). FDTD simulation was used to demonstrate the hotspot distribution in the Au square array structure, which proved the local coupling increased with the Au deposition of time when the gaps decreased between the neighbor units. In comparison with other nanostructures based on PS templates, the hotspot distributions in our observations were Raman spectra, respectively. In fact, EF means the ratio of the SERS signals to Raman signals for every single molecule on even. In our experiment, the laser spot was 1 µm in diameter and the penetration depth was 17 µm of the focused laser beam used; the density of 4-MBA was 1.345 g/cm −3 ; thus, the number of the 4-MBA molecule illuminated by the laser light was calculated to be 7.01 × 10 10 . N SERS was the number of surface-adsorbed molecules within the laser spot with the diameter (1 µm), and the surface area occupied by an adsorbed molecule could be adopted as 0.20 nm 2 /molecule. In comparison with the deposition along the normal direction, both Au square array and Au cone array showed an enhancement factor of 5 × 10 9 and 3 × 10 9 for the p polarization light, as shown in Figure 2, indicating the Au square array with excellent SERS performance due to the narrow gaps ( Figure 3). FDTD simulation was used to demonstrate the hotspot distribution in the Au square array structure, which proved the local coupling increased with the Au deposition of time when the gaps decreased between the neighbor units. In comparison with other nanostructures based on PS templates, the hotspot distributions in our observations were like those from the squared-shape noble nanostructures fabricated by conventional lithography with more hotspot distributions, as shown in Figure 4. When the deposition time increased from 10 min to 15 min, the shapes of the nanostructures were quite like the squared, and more hotspots were created. In addition, the local electromagnetic fields increased significantly due to the narrow gaps between the neighbor units when the deposition time increased. Compared to s polarization, p polarization resulted in more hotspots; therefore, the nanostructures by 15 min deposition showed the larger enhancement factor. FDTD simulations also indicated the SERS properties for Au deposition with the angle of 70 • showed the different symmetries and hotspot distributions affected by the template pattern and the film deposition parameters, indicating more possibilities to control the morphologies of the nanostructure array. Similar observations were also obtained for the cone-shaped samples, as shown in Figure S3.
like those from the squared-shape noble nanostructures fabricated by conventional lithography with more hotspot distributions, as shown in Figure 4. When the deposition time increased from 10 min to 15 min, the shapes of the nanostructures were quite like the squared, and more hotspots were created. In addition, the local electromagnetic fields increased significantly due to the narrow gaps between the neighbor units when the deposition time increased. Compared to s polarization, p polarization resulted in more hotspots; therefore, the nanostructures by 15 min deposition showed the larger enhancement factor. FDTD simulations also indicated the SERS properties for Au deposition with the angle of 70° showed the different symmetries and hotspot distributions affected by the template pattern and the film deposition parameters, indicating more possibilities to control the morphologies of the nanostructure array. Similar observations were also obtained for the cone-shaped samples, as shown in Figure S3.  The nanostructure with the square shapes was used as the SERS-active substrate for detecting AFP due to its excellent SERS when excited by p-polarization light. After 4-MBA decoration, the EDC-NHS mixed solution was applied to activate the carboxyl group for the generation of MBA-derived coupling agent, forming the SERS-active chips [24]. To The nanostructure with the square shapes was used as the SERS-active substrate for detecting AFP due to its excellent SERS when excited by p-polarization light. After 4-MBA decoration, the EDC-NHS mixed solution was applied to activate the carboxyl group for the generation of MBA-derived coupling agent, forming the SERS-active chips [24]. To avoid non-specific recognition after the anti-AFP modification, the chip was deactivated with a blocking buffer, which was to clear the cross-link active sites and avoid the reaction with any other proteins. SERS-active chips were immersed in AFP solutions of different concentrations for specific and saturated recognition. As reported, the frequency shift of the peak at 1075 cm −1 belonged to the ph-S-Au vibration mode, which could be used for quantitative analyses of the AFP concentration because the peak at 1075 cm −1 showed consistent shifts with AFP linkage amount [25]. For our sample with square-shaped units, the excitation light with p-polarization showed the significant SERS observations due to the significant hotspot distribution, and the peak at 1075 cm −1 showed continuous shifts when the AFP concentration increased from 5 pg to 5 ng/mL, confirmed by the magnified peak at 1075 cm −1 , as shown in Figure 5. Similar observations were also reached for the sample with the cone-shaped units. When the p polarization light was applied, the mean enhanced electromagnetic field was also very strong in comparison with the s polarization light, which means that both structures were suitable for AFP detection. These observations indicated that our preparation method was simple for good SERS signals based on 2D PS beads arrays.

Conclusions
Based on the 2D PS array template, Au arrays with different shapes were prepared by angle deposition. When the deposition angle was 70°, the square-shaped and coneshaped nanostructures were obtained on PS500 nm with a space of 120 nm after being etched. The increase in Au deposition time led to the Au caps showing a cube shape due to the similar shadow effects around. At the same time, the increased time of Au deposition results in the narrow gaps between Au caps, which made significant contributions to the enhanced local coupling electromagnetic field, was confirmed by significant SERS observations and FDTD simulations. The SERS chip was made for detecting the liver cancer cell marker AFP, and the detection limit of AFP in the range of 5 pg to 5 ng/mL was obtained, which provided a possible early detection of liver cancer.

Supplementary Materials:
The following are available online, Figure S1: deposition setting-up, Figure S2: different directions of the film depositions onto the PS array, Figure S3 and Figure

Conclusions
Based on the 2D PS array template, Au arrays with different shapes were prepared by angle deposition. When the deposition angle was 70 • , the square-shaped and cone-shaped nanostructures were obtained on PS500 nm with a space of 120 nm after being etched. The increase in Au deposition time led to the Au caps showing a cube shape due to the similar shadow effects around. At the same time, the increased time of Au deposition results in the narrow gaps between Au caps, which made significant contributions to the enhanced local coupling electromagnetic field, was confirmed by significant SERS observations and FDTD simulations. The SERS chip was made for detecting the liver cancer cell marker AFP, and the detection limit of AFP in the range of 5 pg to 5 ng/mL was obtained, which provided a possible early detection of liver cancer.

Supplementary Materials:
The following are available online, Figure S1: deposition setting-up, Figure S2: different directions of the film depositions onto the PS array, Figure S3 and Figure S4: different directions of the film depositions leading to the nanostructures with different shapes, Figure  S5: exciting light with P and S polarizations.
Author Contributions: Formal analysis and writing, Y.W.; Writing-original draft, A.Z.; Methodology, X.Z.; Writing-review and editing, Visualization, Y.Z. All authors have read and agreed to the published version of the manuscript.
Funding: National Natural Science Foundation of China (Nos. 61675090 and 61675080).

Data Availability Statement:
The data presented in this study are available on the request from the corresponding author.