Optimization of Ag coated hydrogen silsesquioxane square array hybrid structure design for surface-enhanced Raman scattering substrate

Hong Wang; Zihang Huo; Zihao Zhang; Shaofei Chen; Shaoji Jiang

doi:10.1364/OE.26.001097

1. Introduction

Raman spectroscopy is a powerful analytical tool that offers molecular vibrational fingerprints of analytes but commonly suffers from a combination of weak signals with low sensitivity. When compared with the inefficient traditional Raman spectrum, surface-enhanced Raman scattering (SERS) can achieve single-molecule-level detection because of its enormous near-field enhancement property. SERS has found important applications in areas including optical sensing [1, 2], and biomedical [3, 4], and chemical analysis [5, 6], and has attracted the attention of many research groups. It is therefore imperative to find a rational method for SERS substrate design according to the actual application requirements.

Over the past few decades, there have been many theoretical and experimental works in the SERS research field, and some SERS substrate design rules are now generally accepted. It is well known that SERS enhancement mechanisms are predominantly the result of chemical mechanism (CM) and electromagnetic mechanism (EM) enhancements [7]. The EM enhancement plays a major role in intensity enhancements. The main source of EM enhancement is that generated by localized surface plasmon resonance (LSPR), which can be adjusted by varying properties including nanostructure size, composition, morphology and dielectric environment, excitation wavelength [8–12]. Additionally, hotspots are also often used to amplify the electromagnetic field and are generated near the sharp tips [13, 14] and in the tiny gaps between adjacent nanostructures [15, 16]. The SERS enhancement mechanism has attracted considerable research interest. The enhancement mechanism is becoming increasingly clear and therefore provides a theoretical guarantee for rational SERS substrate design according to application requirements.

SERS substrate fabrication technologies are divided into two main categories: bottom-up and top-down approaches. Each approach has its own strengths and weaknesses. Of the two categories, top-down approaches [17–21], such as focused ion beams, electron-beam lithography (EBL), interference lithography and nanoimprint lithography, were used for sample preparation by the research groups in the literatures. Among the technologies, EBL is a powerful tool for use in micro- and nanomanufacturing processes. It provides more versatile surface morphologies and offers good homogeneity and replicability, but is expensive and requires long sample preparation times. Some recent works have reported noble metal deposition on pre-defined surfaces formed by EBL [12, 22–26]. These patterns include circles, triangles, squares, rectangles, stars and bowtie shapes. Methods that enable accurate implementation of the designed nanostructure effectively provide a technical guarantee for rational SERS substrate design.

In this work, we proposed a design process for a novel SERS substrate structure and verified the resulting structure through both experiments and simulations. First, to ensure the required LSPR wavelength position, we used the particle swarm optimization (PSO) algorithm to obtain optimal design parameters. On this basis, periodic hydrogen silsesquioxane (HSQ) nanopillar arrays of different sizes were fabricated using EBL after silver was sputtered onto the HSQ nanostructure using a magnetron sputtering system. The Raman spectra of the samples were measured for experimental analysis. Rhodamine 6G (R6G) was used as the probing molecule under laser excitation at 514.5 nm. The optical properties of each sample were also calculated by the finite-difference time-domain (FDTD) method. The simulation results, which are the product of the electric field intensity enhancements at excitation and Raman scattering wavelengths, corresponded well with the experimental results and thus validated the accuracy of the simulation model. The effects of particle size on the SERS measurements were also researched systematically and the field enhancement mechanisms were observed through calculating the absorption spectra, the charge distribution, and the electric field distribution. To the best of our knowledge, this is the first introduction of the PSO algorithm to SERS calculations. Many previous research works in this field involving optimization have simply found the best value for one variable. In our work, multiple parameters are optimized simultaneously, including the thickness of HSQ array, the thickness of the deposited Ag film and the width of the square. As this optimization algorithm becomes increasingly mature, it will also become a powerful tool that will help researchers to determine the best possible SERS substrate for their specific applications.

2. Methodology

2.1 Sample fabrication

HSQ nanopillar arrays were manufactured by EBL (EBPG5000 plus; Raith). A 65-nm-thick negative resist HSQ film was spin-coated on cleaned Si substrates at 4000 rpm for 60 s and then baked at 90°C for 5 min. The resist was exposed at an accelerating voltage of 100 kV with an area dose of 1400 μC cm⁻², a beam current of 2 nA, a beam step size of 2 nm and resolution of 1 nm. The proximity effect was corrected in the process. After exposure, the substrates were immersed in tetramethyl ammonium hydroxide for 2 min and finally were rinsed using deionized water for 1 min.

The metal film was deposited using magnetron sputtering (MSP-300; Jinshengweina Technology). A 20-nm-thick Ag layer was deposited on the patterned resist substrates at a deposition rate of 2 Ås⁻¹. Simultaneously, a smooth Ag film of the same thickness was deposited on a blank silicon substrate under the same conditions for comparison with the patterned substrate in terms of their SERS performances. A schematic diagram of the sample fabrication flow is shown in Fig. 1. The letters a, d, h and P represent the Ag film thickness, the width of the square HSQ nanopillars, the height of the HSQ nanopillars and the array period, respectively.

Fig. 1 Schematic diagram of the samples fabrication flow.

Download Full Size | PDF

2.2 Characterization of substrates and SERS measurements

The nanostructure patterns were characterized using a focused ion beam scanning electron microscope (Auriga; Zeiss). The height of the HSQ nanopillars and the deposited Ag film were measured by a profilometer (AMBIOS XP-200; Ambios Technology Inc.).

The Raman measurements were performed using a confocal Raman microscope (inVia Reflex + Simple-Tau152-DX; Renishaw). To prepare the substrates for SERS, drops of 10⁻⁶ M R6G solution were applied to the samples and then allowed to dry naturally at room temperature; round spots subsequently appeared on these samples. The Raman signals were obtained using a 50 × objective equipped with a 514.5 nm excitation laser under the conditions that the laser power impinging on the samples was 0.001 mW, the spot size was approximately 1 μm, and the integration time was 10 s. Simultaneously, we used a smooth Ag film coated with R6G molecules at a concentration of 10⁻² M under a laser power intensity of 0.01 mW as a reference sample; the other measurement parameters were the same as those of the previous sample above. Average Raman intensities were calculated from the measured values of ten random spots on the samples.

2.3 Numerical simulations

Commercial FDTD software (version 8.17, Lumerical Solutions, Inc.) was used to simulate the optical properties of the samples. A coordinate system was defined in the simulation, and the center of the HSQ square at the Si-HSQ nanopillars interface (x-y plane) was considered to be the coordinate origin. The light source was set as a normally incident plane wave traveling in the z-direction with its electric field polarized along the x-axis. Perfectly matched layer boundary conditions were used in the z-direction, while periodic boundary conditions were used in the other two directions. The mesh size was set at 1 nm. The Ag and Si dispersion models were selected from the materials database. The relative dielectric constants of the background and the HSQ were assumed to be constant, with values of 1 and 1.4, respectively. The absorption spectrum, the charge distribution and the electric field distribution were obtained from the simulation model. To ensure the accuracy of the SERS calculations, a complex three-dimensional analysis group was built for the electric field that was used to calculate the electric field intensity near the interface between the Ag film and the air, rather than calculate the electric field intensity across the entire simulation area. The analysis group then added up all grid values that were located at the outer interface and were the product of the electric field intensity enhancements at the excitation and Raman scattering wavelengths.

Based on the results from the three-dimensional analysis group, PSO was used to optimize the structural parameters. A set of N = 200 sample points (maximum number of generations = 20, generation size = 10) was built, and three variables were scanned in continuous parameter spaces; the HSQ array thickness range was set to vary from 40 nm to 70 nm, while the thickness range for the deposited Ag film was set to vary from 10 nm to 25 nm, and the width range of the square was set to vary from 40 nm to 200 nm. The electric field enhancement that was calculated by the analysis group above was then set as the target function.

3. Results and discussion

3.1 Optimization

Optimization algorithms have found a relatively wide range of applications in various fields. The lastest review discussed the problems of identification and optimization materials using genetic algorithms, and found that machine learning models can speed up the identification of useful and novel materials [27]. Due to the chief importance of SERS in plasmonics, the reseachers chose to optimize the near field enhancement ( $I_{S E R S} \approx {| α_{s} |}^{2} {| g (λ_{0}) |}^{4} I_{0}$ ) to design nanoparticles automatically based on coupling of an efficient global optimization algorithm, then the inverse-designed metal (Au, Ag, Al) nanoparticles with optimal g were obtained [28]. Offering extinction cross section as an object function, a nanolinear programming optimizer was used to search for the best morphological parameters of the structure nano-sinusoid, and EBL is implemented to fabricate the proposed pattern [29]. A new antenna geometry was presented by fully automatic design using an evolutionary algorithm, setting the near-field intensity enhancement as the optimized conditions [30]. To date, SERS substrates design has largely been reliant on empirical observations and few research works have addressed multi-parameters optimization of SERS substrates. PSO, which was developed by Kennedy and Eberhart in 1995 [31], is an evolutionary optimization technique. It offers population-based random solutions that are inspired by the social behavior of flocks of birds [32]. A binary version of the PSO algorithm was used to design 100nm diameter Ag nanospheres array geometry which can achieve broadband field enhancements in the spectral range from 400 to 900 nm, and their results showed that the disorder array was higher than the periodic array [33]. In our work, we used PSO algorithm to search for the best solution for the SERS substrates, taking the excitation and Raman frequency into consideration.

The enhancement factor (EF) is an important figure for expression of SERS performance and has several definitions [34]. It can be expressed using the following formula [35, 36], which is normally used to compare average SERS performances among different substrates.

E F = \frac{I_{S E R S} / N_{S u r f}}{I_{b u l k} / N_{b u l k}}

Here,

I_{S E R S}

and

I_{b u l k}

are the Raman signal intensities at the scattering wavelengths that are detected from the probe molecules that are adsorbed on the nanostructures and from the bulk probe molecules, respectively, and

N_{s u r f}

and

N_{b u l k}

are the number of probe molecules that are adsorbed on the nanostructures and the bulk probe molecules that are exposed to the laser spot, respectively.

The current theoretical understanding indicates that the total enhancement of SERS can also be written as follows [37]:

\frac{I_{S E R S}}{I_{0}} = 〈 {| \frac{E_{l o c} (ω_{e x})}{E_{i n c} (ω_{e x})} |}^{2} 〉 〈 {| \frac{E_{l o c} (ω_{s})}{E_{i n c} (ω_{s})} |}^{2} 〉

Here,

ω_{e x}

and

ω_{s}

are the frequencies of the incident and scattered radiation, respectively, while the intensities

I_{S E R S}

and

I_{0}

refer to the coverage of identical probe molecules when illuminated with the same intensity, and deposited on the island film and the inert reference substrate, respectively.

E_{l o c} (ω_{e x})

and

E_{l o c} (ω_{s})

are the enhanced local electric fields at the laser excitation frequency and at the scattering frequency, respectively.

E_{i n c} (ω_{e x})

is the incident electric field at the excitation frequency, and

E_{i n c} (ω_{s})

is the electric field that is radiated by the molecules at the scattering frequency. When Eq. (1) is compared with Eq. (2), the following formula can be obtained, where EF is proportional to the product of the squared the electric field enhancement at the incident and Raman scattered frequencies [38, 39].

E F \propto 〈 {| \frac{E_{l o c} (ω_{e x})}{E_{i n c} (ω_{e x})} |}^{2} 〉 〈 {| \frac{E_{l o c} (ω_{s})}{E_{i n c} (ω_{s})} |}^{2} 〉

Different objective functions could be set depending on the specific application. Obviously, the best SERS substrate has the maximum near-field EF on the surface. Therefore, our optimization criterion is to find the maximum EF. Previous studies have shown that the resonance peak obtained the greatest enhancement when it is at a frequency between the excitation and Raman frequency [38, 40]; so whether the Raman frequency is taken into account has an impact on calculating the position of the LSPR. In our work, in order to calculate the EF more practically and accurately, both the excitation and Raman frequency are taken into consideration. Because the specific analyte R6G and the EF at 562 nm were studied in the experiments, the evaluation function was then

M e r i t = \max {(\int {| \frac{E_{l o c} (ω_{514})}{E_{i n c} (ω_{514})} |}^{2} {| \frac{E_{l o c} (ω_{562})}{E_{i n c} (ω_{562})} |}^{2} d A)}_{i}, (i = 1, 2, \dots 200) .

We then obtained the following results. The best parameters were d = 93.1 nm, h = 67.4 nm, and a = 20.9 nm. A highest EF of

5.65 \times 10^{9}

was obtained based on these best parameters. Our goal was to assess samples with parameters of h = 67 nm and a = 21 nm, while d varied in size; however, the experimentally measured data differed slightly from the anticipated data.

In our work, only three variables were optimized to find the best possible solution, but more parameters can be added the optimization process based on the design requirements, including refractive index and periods, but more sample points will be required to obtain the global maximum, along with massive calculation resources, and the process will be more time-consuming.

3.2 Geometrical characterization of samples

HSQ nanopillars of various sizes (d $\approx$ 60, 105, 145, 165, 185, and 205 nm, which are named S1–S6) with a fixed height (h $\approx$ 65 nm), and a fixed period (P = 300 nm) were first fabricated by EBL on silicon substrates. Then, the patterned substrates were coated with the 20-nm-thick Ag films. Figures 2(a)-2(f) show the top-view SEM images of S2, S5 and S3 before and after metal deposition. Figures 2(g)-2(h) show the tilted-view (54°) images of S3 before and after metal deposition. To confirm the Ag film deposition on the HSQ sidewalls, the sample S3 was cut by the focused ion beam, then a cross-sectional image of S3 after metal deposition can be obtained, as shown in Fig. 2(i), from top to bottom the four layers are the Pt protection layer, Ag deposition layer, HSQ nanopillars, and Si substrate, respectively. It is easy to see that the HSQ nanopillars that have been fabricated by EBL are both highly uniform and reproducible. The sizes of the HSQ structure were obtained from the SEM images, while the thicknesses of the HSQ and Ag films were derived from profilometer measurements. These data were then used in the FDTD simulations for comparison with the experiment results.

Fig. 2 (a) and (b) The top-view SEM images of S2 before and after metal deposition. (c) and (d) The top-view images of S5 before and after metal deposition. (e) and (f) The top-view images of S3 before and after metal deposition. (g) and (h) The tilted-view (54°) images of S3 before and after metal deposition. (i) The cross-sectional image of S3 after metal deposition. The scale bars represent 200 nm. In (a)–(f), the signals were collected by detector Inlens. In (g)-(i), the signals were collected by detector SE2.

Download Full Size | PDF

3.3 Comparison of experimental and simulated results

In the experiments, the Raman spectra of a series of samples (S1–S6) were measured at 0.5 μL of R6G (10⁻⁶M) under illumination at an excitation wavelength of 514.5 nm, and the 20-nm-thick smooth silver film (called S7) was measured at a concentration of 10⁻² M for reference under the same experimental conditions, except for the laser intensity, which had been increased by ten times. Figure 3(a) shows the Raman spectra of S3 and S7; large differences in intensity can be seen between the two samples and eight obvious peaks are shown in the Raman spectra that correspond to the various molecular vibrations of R6G. The Raman intensity at 562 nm (Raman shift of 1649 cm⁻¹) was studied in both our experiments and our simulations. Figure 3(b) plots a two dimensional Raman mapping at the Raman peak of 1649 cm⁻¹ on S3. The scanning area is 900 μm², the scan step is 1 μm and the total number of imaging pixels is 961 (31 × 31). The relative standard deviation is 6.84%, and these results show that the sample homogeneity is high. The Raman spectra of samples S1–S6 are shown in Fig. 4(a). The second sample shows the highest enhancement factor, and the influence of the nanopillar size on the field enhancement is remarkable. According to Eq. (1), when we consider that the smooth Ag film was selected, $N {}_{b u l k}$ and $I_{b u l k}$ can be replaced by $N_{r e f}$ and $I_{r e f}$ , respectively. The number of molecules on the surface perpendicular to the z-axis could be the product of the concentration and the volume, but the number of molecules on the sidewall due to chemisorption is low and may be negligible. The solution volumes of the R6G solutions applied to the patterned and smooth substrates are considered to be approximately equal. Therefore, the approximate number of molecules can be calculated using the following: $N_{s u r f} / N_{r e f} = (C_{s u r f} \times V) / (C_{r e f} \times V) = 10^{- 4}$ . Then, the EFs of the six samples at 562 nm were calculated using S7 as a reference, with results as drawn in Fig. 4(b). The experimental EF of sample S2 is $4.73 \times 10^{5}$ .

Fig. 3 (a) Raman scattering spectra of samples S3 and S7. (b) SERS intensity mapping of sample S3 over a large area of 900 μm² at the Raman signal of 1649 cm⁻¹. The integration time of single spectra was 10 s.

Download Full Size | PDF

Fig. 4 (a) Raman scattering spectra of samples S1–S6. (b) EF comparison of experimental and simulated results at 562 nm.

Download Full Size | PDF

For further clarification of the size dependence of the SERS performance on the samples, FDTD simulations corresponding to the experimental data were performed. Using Eq. (3), the EFs of the six samples at 562 nm were calculated and the results are also drawn in Fig. 4(b). The simulated EF of sample S2 is $8.48 \times 10^{8}$ . It is normal for there to be differences between the experimental and theoretical estimation results, and a similar phenomenon was described in [35]. Figure 4(b) shows that the trends of the two types of results are similar, thus confirming the accuracy of the simulation model.

3.4 Origin of the field enhancement

To explore the enhancement mechanism described above, the absorption spectra were calculated for each sample. As shown by the absorption spectra in Fig. 5, the LSPR peak varies with the width of the square of each sample, and the second peak is closest to the 514–562 nm wavelength regions. Additionally, as the size increases, the resonance wavelength shows a red shift, the width of the peak broadens, and the peak height becomes lower than that of the second sample. To explain these phenomena, the charge distribution is calculated at the resonance peak for each sample. The charge distributions at the main resonance peaks of S2 ( $λ$ = 524 nm) which were located at 22, 28, 38, 50, 60, 70, 80 and 83 nm along the z-axis were calculated, with results as shown in Fig. 6. From these figures, we see that the charge distributions decrease rapidly with the decreasing height; compared with the Fig. 6(g) and 6(h), the height dropped by only 3nm, but the charge distribution fell by half; compared with the Figs. 6(a) and 6(b), the charge distributions increase slightly at the bottom of the nanopillar. Therefore, the charge distributions on the upper surfaces have the main effect on the formation of the resonance peaks. Then, the charge distributions at the main resonance peaks of S1 ( $λ$ = 483 nm), S2 ( $λ$ = 524 nm), S3 ( $λ$ = 565.5 nm) and S5 ( $λ$ = 628.5 nm) were selected to be shown for illustration in Figs. 7(a)-7(d). The charge distributions of minor resonance peaks of S2 ( $λ$ = 366 nm), S3 ( $λ$ = 385.5 nm), S5 ( $λ$ = 408 nm) and S5 ( $λ$ = 470.5 nm) were selected to be shown in Figs. 7(e)-7(h). These peaks are all located at 83 nm along the z-axis. Obviously, the main resonance peaks are generated by the dipole mode, but the minor resonance peaks, which become increasingly apparent, are generated by multipolar modes. Therefore, because of the phase retardation effects that occur in the larger particles, enhancement of the multipolar plasma resonance modes can lead to broadening and red shifts in the dipole mode.

Fig. 5 Simulated absorption spectra of samples S1–S6.

Download Full Size | PDF

Fig. 6 Charge distributions of sample S2 under the major resonance peaks located at (a) 22 nm, (b) 28 nm, (c) 38 nm, (d) 50 nm, (e) 60nm, (f) 70nm, (g) 80nm and (h) 83nm. The maximum and minimum color bar values represent the maximum and minimum values of the extracted data, respectively.

Download Full Size | PDF

Fig. 7 (a)-(d) Charge distributions of samples S1, S2, S3 and S5, respectively, at their main resonance peaks. (e)-(h)Charge distributions of samples S2, S3 and S5 at their minor resonance peaks. For clarity, the data were magnified 10 times and a uniform color bar was used.

Download Full Size | PDF

The near-field distributions of samples S1, S2, S3, and S5 were also calculated at the wavelengths of 514 and 562nm, with results as shown in Figs. 8(a)–8(d) and Figs. 8(e)–8(h) respectively, while Figs. 8(i) and 8(j) show the side-view near-field distributions of samples S2 and S5 at 514 nm. The color bar values correspond to decibels of the electric field intensity. As shown in Fig. 8, the field enhancement is distributed in the forms of sharp corners and sharp edges. At the excitation wavelength, the second sample shows the highest enhancement; however, at the scattering wavelength, the third sample shows the highest enhancement.

Fig. 8 (a)–(d) Near-field distributions in the x-y plane of samples S1, S2, S3 and S5 at the wavelength of 514 nm, respectively. (e)–(h) Near-field distributions in the x-y plane of samples S1, S2, S3 and S5 at the wavelength of 562 nm, respectively. (i), (j) Near-field distributions in the x-z plane of samples S2 and S5 at the wavelength of 514 nm, respectively. (a)–(h) are all located at 83 nm along the z-axis. (i) and (j) are located at 0 nm along the y-axis.

Download Full Size | PDF

4. Conclusions

In this work, we have mainly shown how to set an appropriate evaluation function based on the actual demand of the application and then obtain the best possible solutions within the predefined scope. A highest EF of $5.65 \times 10^{9}$ was obtained based on these best parameters, which were obtained during the optimization process. A series of SERS substrates composed of Ag thin films deposited on patterned substrates using EBL technology were fabricated and the SERS performances of these substrates were measured in the experiments. Good agreement is demonstrated between the experimental results and the simulation results, thus confirming the accuracy of the simulation model. To explain the field enhancement mechanism, the absorption spectra, the charge distribution and the electric field distribution were all calculated based on FDTD simulations. It was shown that the second sample obtains the greatest enhancement; the locations of the resonance peaks in the absorption spectra could be tuned by adjusting the fabrication parameters; the charge distributions on the upper surfaces have the main effect on the formation of the resonance peaks; enhancement of the multipolar plasma resonance modes can lead to broadening and red shifts in the dipole mode. Our work provides a method for rational design of SERS substrates according to their application requirements.

Funding

National Natural Science Foundation of China (NSFC) (61275159); Special Program for Applied Research on Super Computation of the NSFC-Guangdong Joint Fund (the second phase) under Grant No.U1501501.

References and links

1. J. N. Anker, W. P. Hall, O. Lyandres, N. C. Shah, J. Zhao, and R. P. Van Duyne, “Biosensing with plasmonic nanosensors,” Nat. Mater. 7(6), 442–453 (2008). [CrossRef] [PubMed]

2. G. Bodelón, V. Montes-García, V. López-Puente, E. H. Hill, C. Hamon, M. N. Sanz-Ortiz, S. Rodal-Cedeira, C. Costas, S. Celiksoy, I. Pérez-Juste, L. Scarabelli, A. La Porta, J. Pérez-Juste, I. Pastoriza-Santos, and L. M. Liz-Marzán, “Detection and imaging of quorum sensing in Pseudomonas aeruginosa biofilm communities by surface-enhanced resonance Raman scattering,” Nat. Mater. 15(11), 1203–1211 (2016). [CrossRef] [PubMed]

3. G. Lu, H. De Keersmaecker, L. Su, B. Kenens, S. Rocha, E. Fron, C. Chen, P. Van Dorpe, H. Mizuno, J. Hofkens, J. A. Hutchison, and H. Uji-i, “Live-cell SERS endoscopy using plasmonic nanowire waveguides,” Adv. Mater. 26(30), 5124–5128 (2014). [CrossRef] [PubMed]

4. R. A. Alvarez-Puebla and L. M. Liz-Marzán, “SERS-based diagnosis and biodetection,” Small 6(5), 604–610 (2010). [CrossRef] [PubMed]

5. J. Zheng and L. He, “Surface-Enhanced Raman Spectroscopy for the Chemical Analysis of Food,” Compr. Rev. Food Sci. Food Saf. 13(3), 317–328 (2014). [CrossRef]

6. Y. Yang, J. Liu, Z. W. Fu, and D. Qin, “Galvanic replacement-free deposition of Au on Ag for core-shell nanocubes with enhanced chemical stability and SERS activity,” J. Am. Chem. Soc. 136(23), 8153–8156 (2014). [CrossRef] [PubMed]

7. P. L. Stiles, J. A. Dieringer, N. C. Shah, and R. P. Van Duyne, “Surface-enhanced Raman spectroscopy,” Annu. Rev. Anal. Chem. (Palo Alto, Calif.) 1(1), 601–626 (2008). [CrossRef] [PubMed]

8. K. L. Kelly, E. Coronado, L. L. Zhao, and G. C. Schatz, “The optical properties of metal nanoparticles the influence of size, shape, and dielectric environment,” J. Phys. Chem. B 107(3), 668–677 (2003). [CrossRef]

9. H. Ko, S. Singamaneni, and V. V. Tsukruk, “Nanostructured surfaces and assemblies as SERS media,” Small 4(10), 1576–1599 (2008). [CrossRef] [PubMed]

10. R. A. Álvarez-Puebla, “Effects of the Excitation Wavelength on the SERS Spectrum,” J. Phys. Chem. Lett. 3(7), 857–866 (2012). [CrossRef] [PubMed]

11. S. Wei, M. Zheng, Q. Xiang, H. Hu, and H. Duan, “Optimization of the particle density to maximize the SERS enhancement factor of periodic plasmonic nanostructure array,” Opt. Express 24(18), 20613–20620 (2016). [CrossRef] [PubMed]

12. W. Q. Li, G. Wang, X. N. Zhang, H. P. Geng, J. L. Shen, L. S. Wang, J. Zhao, L. F. Xu, L. J. Zhang, Y. Q. Wu, R. Z. Tai, and G. Chen, “Geometrical and morphological optimizations of plasmonic nanoarrays for high-performance SERS detection,” Nanoscale 7(37), 15487–15494 (2015). [CrossRef] [PubMed]

13. A. Gopalakrishnan, M. Chirumamilla, F. De Angelis, A. Toma, R. P. Zaccaria, and R. Krahne, “Bimetallic 3D Nanostar Dimers in Ring Cavities: Recyclable and Robust Surface-Enhanced Raman Scattering Substrates for Signal Detection from few molecules,” ACS Nano 8(8), 7986–7994 (2014). [CrossRef] [PubMed]

14. E. M. Perassi, C. Hrelescu, A. Wisnet, M. Döblinger, C. Scheu, F. Jäckel, E. A. Coronado, and J. Feldmann, “Quantitative Understanding of the Optical Properties of a Single, Complex-Shaped Gold Nanoparticle from Experiment and Theory,” ACS Nano 8(5), 4395–4402 (2014). [CrossRef] [PubMed]

15. S. Kim, J. Jin, Y. J. Kim, I. Y. Park, Y. Kim, and S. W. Kim, “High-harmonic generation by resonant plasmon field enhancement,” Nature 453(7196), 757–760 (2008). [CrossRef] [PubMed]

16. H. Im, K. C. Bantz, N. C. Lindquist, C. L. Haynes, and S. H. Oh, “Vertically oriented sub-10-nm plasmonic nanogap arrays,” Nano Lett. 10(6), 2231–2236 (2010). [CrossRef] [PubMed]

17. T. Liu, Y. Shen, W. Shin, Q. Zhu, S. Fan, and C. Jin, “Dislocated double-layer metal gratings: an efficient unidirectional coupler,” Nano Lett. 14(7), 3848–3854 (2014). [CrossRef] [PubMed]

18. S. J. Tan, L. Zhang, D. Zhu, X. M. Goh, Y. M. Wang, K. Kumar, C. W. Qiu, and J. K. Yang, “Plasmonic color palettes for photorealistic printing with aluminum nanostructures,” Nano Lett. 14(7), 4023–4029 (2014). [CrossRef] [PubMed]

19. Y. Shen, J. Zhou, T. Liu, Y. Tao, R. Jiang, M. Liu, G. Xiao, J. Zhu, Z. K. Zhou, X. Wang, C. Jin, and J. Wang, “Plasmonic gold mushroom arrays with refractive index sensing figures of merit approaching the theoretical limit,” Nat. Commun. 4(4), 2381 (2013). [PubMed]

20. N. A. Cinel, S. Bütün, G. Ertaş, and E. Ozbay, “‘Fairy Chimney’-shaped tandem metamaterials as double resonance SERS substrates,” Small 9(4), 531–537 (2013). [CrossRef] [PubMed]

21. W.-D. Li, F. Ding, J. Hu, and S. Y. Chou, “Three-dimensional cavity nanoantenna coupled plasmonic nanodots for ultrahigh and uniform surface-enhanced Raman scattering over large area,” Opt. Express 19(5), 3925–3936 (2011). [CrossRef] [PubMed]

22. N. A. Hatab, C. H. Hsueh, A. L. Gaddis, S. T. Retterer, J. H. Li, G. Eres, Z. Zhang, and B. Gu, “Free-standing optical gold bowtie nanoantenna with variable gap size for enhanced Raman spectroscopy,” Nano Lett. 10(12), 4952–4955 (2010). [CrossRef] [PubMed]

23. M. Chirumamilla, A. Toma, A. Gopalakrishnan, G. Das, R. P. Zaccaria, R. Krahne, E. Rondanina, M. Leoncini, C. Liberale, F. De Angelis, and E. Di Fabrizio, “3D nanostar dimers with a sub-10-nm gap for single-/few-molecule surface-enhanced raman scattering,” Adv. Mater. 26(15), 2353–2358 (2014). [CrossRef] [PubMed]

24. A. Dowd, M. Geisler, S. Zhu, M. L. Wood, and M. B. Cortie, “Role of multipolar plasmon resonances during surface-enhanced Raman spectroscopy on Au micro-patches,” RSC Advances 6(116), 115284 (2016). [CrossRef]

25. K. Hasna, A. Antony, J. Puigdollers, K. R. Kumar, and M. K. Jayaraj, “Fabrication of cost-effective, highly reproducible large area arrays of nanotriangular pillars for surface enhanced Raman scattering substrates,” Nano Res. 9(10), 3075–3083 (2016). [CrossRef]

26. M. Zheng, X. Zhu, Y. Chen, Q. Xiang, and H. Duan, “Three-dimensional donut-like gold nanorings with multiple hot spots for surface-enhanced raman spectroscopy,” Nanotechnology 28(4), 045303 (2017). [CrossRef] [PubMed]

27. T. C. Le and D. A. Winkler, “Discovery and Optimization of Materials Using Evolutionary Approaches,” Chem. Rev. 116(10), 6107–6132 (2016). [CrossRef] [PubMed]

28. C. Forestiere, Y. He, R. Wang, R. M. Kirby, and L. Dal Negro, “Inverse Design of Metal Nanoparticles’ Morphology,” ACS Photonics 3(1), 68–78 (2016). [CrossRef]

29. D. Mortazavi, A. Kouzani, and M. Kalani, “Design and fabrication of nano-sinusoid LSPR devices,” Opt. Express 22(16), 18889–18903 (2014). [CrossRef] [PubMed]

30. T. Feichtner, O. Selig, M. Kiunke, and B. Hecht, “Evolutionary optimization of optical antennas,” Phys. Rev. Lett. 109(12), 127701 (2012). [CrossRef] [PubMed]

31. J. Kennedy and R. Eberhart, “Particle swarm optimization,” in Proceedings of IEEE International Conference on Neural Networks (IEEE, 2002), pp. 1942–1948.

32. R. C. Eberhart and Y. Shi, “Particle Swarm OptimizationDevelopments, Applications and Resources,” in Proceedings of IEEE Congress on Evolutionary Computation (IEEE, 2002), pp. 81–86.

33. C. Forestiere, M. Donelli, G. F. Walsh, E. Zeni, G. Miano, and L. Dal Negro, “Particle-swarm optimization of broadband nanoplasmonic arrays,” Opt. Lett. 35(2), 133–135 (2010). [CrossRef] [PubMed]

34. E. C. Le Ru, E. Blackie, M. Meyer, and P. G. Etchegoin, “Surface Enhanced Raman Scattering Enhancement Factors: A Comprehensive Study,” J. Phys. Chem. C 111(37), 13794–13803 (2007). [CrossRef]

35. Z. Zhu, B. Bai, H. Duan, H. Zhang, M. Zhang, O. You, Q. Li, Q. Tan, J. Wang, S. Fan, and G. Jin, “M-shaped grating by nanoimprinting: a replicable, large-area, highly active plasmonic surface-enhanced Raman scattering substrate with nanogaps,” Small 10(8), 1603–1611 (2014). [CrossRef] [PubMed]

36. S. Zhu, C. Fan, J. Wang, J. He, E. Liang, and M. Chao, “Surface enhanced Raman scattering of 4-aminothiophenol sandwiched between Ag nanocubes and smooth Pt substrate: The effect of the thickness of Pt film,” J. Appl. Phys. 116(4), 044312 (2014). [CrossRef]

37. H. G. Bingler, H. Brunner, A. Leitner, F. R. Aussenegg, and A. Wokaun, “Interference enhanced surface Raman scattering of adsorbates on a silver-spacer-islands multilayer system,” Mol. Phys. 85(3), 587–606 (1995). [CrossRef]

38. N. Félidj, J. Aubard, G. Lévi, J. R. Krenn, M. Salerno, G. Schider, B. Lamprecht, A. Leitner, and F. R. Aussenegg, “Controlling the optical response of regular arrays of gold particles for surface-enhanced Raman scattering,” Phys. Rev. B 65(7), 075419 (2002). [CrossRef]

39. Y.-P. Zhao, S. B. Chaney, S. Shanmukh, and R. A. Dluhy, “Polarized Surface Enhanced Raman and Absorbance Spectra of Aligned Silver Nanorod Arrays,” J. Phys. Chem. B 110(7), 3153–3157 (2006). [CrossRef] [PubMed]

40. A. D. McFarland, M. A. Young, J. A. Dieringer, and R. P. Van Duyne, “Wavelength-Scanned Surface-Enhanced Raman Excitation Spectroscopy,” J. Phys. Chem. B 109(22), 11279–11285 (2005). [CrossRef] [PubMed]

Optimization of Ag coated hydrogen silsesquioxane square array hybrid structure design for surface-enhanced Raman scattering substrate

Abstract

1. Introduction

2. Methodology

2.1 Sample fabrication

2.2 Characterization of substrates and SERS measurements

2.3 Numerical simulations

3. Results and discussion

3.1 Optimization

3.2 Geometrical characterization of samples

3.3 Comparison of experimental and simulated results

3.4 Origin of the field enhancement

4. Conclusions

Funding

References and links

Cited By

Figures (8)

Equations (4)

Optics Express