Droplet-Confined Electroless Deposition of Silver Nanoparticles on

Jun 5, 2017 - Copyright © 2017 American Chemical Society. *E-mail: [email protected]. Cite this:ACS Appl. Mater. Interfaces 9, 25, 21548-21553 ...
0 downloads 0 Views 6MB Size
Research Article www.acsami.org

Droplet-Confined Electroless Deposition of Silver Nanoparticles on Ordered Superhydrophobic Structures for High Uniform SERS Measurements Daren Xu, Fei Teng, Zhongshun Wang, and Nan Lu* State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University, Changchun 130012, P. R. China S Supporting Information *

ABSTRACT: Surface-enhanced Raman scattering spectroscopy (SERS) is a nondestructive testing technique. To increase reproducibility of the SERS measurement is the key issue for improving the performance of SERS. In this article, we demonstrate an efficient method to improve the reproducibility, using confined silver nanoparticles (AgNPs) as a substrate. The AgNPs are formed uniformly on the tops of the prepared nanopillars by droplet-confined electroless deposition on the hydrophobic Si nanopillar arrays. The AgNPs present an excellent reproducibility in Raman measurement; the relative standard deviation is down to 3.40%. There exists a great linear correlation between the concentration of Rhodamine 6G (R6G) and the Raman intensity in the log−log plot; R2 is 0.998, indicating that this SERS substrate can be applied for the quantitative SERS analysis. Meanwhile, the minimum detection concentration is down to 10−11 M on the hydrophobic substrate, with R6G as a probe molecule. KEYWORDS: AgNPs, droplet-confined electroless deposition, ordered superhydrophobic structures, reproducibility, SERS

1. INTRODUCTION Surface-enhanced Raman scattering (SERS) is a high-efficiency and nondestructive technique for chemical and biological detection due to its real-time response and ultrahigh detection sensitivity. Many efforts have been made to decrease the detection limit and increase the reproducibility.1,2 There exist two effects that contribute to the SERS enhancement. One effect is chemical enhancement, which is introduced by charge transfer between the analytes and the SERS substrate; the other is electromagnetic enhancement, which plays the dominant role in SERS enhancement.3,4 It is known that metal nanogaps and nanotips can lead to strong electromagnetic fields, with an evanescent character, which is termed hot spot in SERS measurements.5−7 It is worth noting that Song et al. successfully fabricated nanomushrooms between the gold head and silver cap, with the DNA strands as the interlayer, resulting in the nanogaps down to 1−2 nm, which presented a strong enhancement of SERS8 and a great potential for practical application in the area of biological sensing.9 Fang et al. have demonstrated that only 0.7863% of the total number of molecules on the hot spots can contribute to 69% of the overall Raman signal on the silver-film-over-nanospheres substrate.10 The result indicates that making analytes on hot spots is extremely important for improving the performance of SERS measurements. David et al. revealed that a significant enhancement could be obtained when the nanogap was less than 10 nm by investigating the effect of the gap size between two nanobowls on the Raman enhancement factor.11 Recently, © XXXX American Chemical Society

Wu et al. reported a method to create a nanogap within 10 nm by a two-step chemical deposition process to form highly packed silver nanosheets; this method is both time and cost effective.12 An ideal SERS substrate should present a low detection limit and a good reproducibility, which requires a uniform distribution of hot spots on a SERS substrate.13,14 Therefore, to form uniformly distributed and accessible hot spots is an efficient approach to improve the reproducibility.15,16 Physical vapor deposition and chemical reduction reaction methods are usually applied to create nanogaps on substrates. The physical vapor deposition method allows for regulating the size and morphology of the metal nanostructures, which benefits the improvement of the performance of SERS.17 However, these methods need specialized equipments. The chemical methods mainly include synthesis of nanoparticles,18 electrodeposition,19 and electroless deposition.20 Among them, electroless deposition has attracted more attention due to its simplicity and efficiency in fabricating metal nanostructures. For example, Cheng et al. created the SERS substrate by generating silver nanoparticles (AgNPs) on a glass slide by the electroless deposition method,21 but a random distribution of the particles may lower the reproducibility of measurements. To improve reproducibility, a template-directed electroless Received: March 25, 2017 Accepted: June 5, 2017 Published: June 5, 2017 A

DOI: 10.1021/acsami.7b04240 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 1. Schematic of fabricating AgNPs on a hydrophobic Si nanopillar array by DCED. (A) Decreasing the surface energy of the Si nanopillar array by assembling fluoroalkylsilane molecules on its surface. (B) Electroless depositing of AgNPs on the modified Si nanopillar array by adding a droplet of AgNO3 aqueous solution on its surface. (C) Obtaining AgNPs on top of the nanopillars by removing the excessive solution and rinsing with deionized water. system (Oxford Instrument Co., U.K.) to obtain Si nanopillars. The RIE treatment was set as follows: 6 sccm SF6, 45 sccm CHF3, radio frequency power of 25 W, inductively coupled plasma power of 100 W, and pressure of 8 mTorr. Tetrahydrofuran was used to dissolve the remaining PS spheres. Then, the samples were placed in an airtight glass dish, with 20 μL of trimethoxy (1H,1H,2H,2H-nonafluorohexyl) silane. After being kept at 250 °C in the heating stage for two hours, the samples were cooled down to the room temperature 2.3. Creating AgNPs. AgNO3 (0.01 M) solution and HF solution (46 mM) were prepared separately. They were mixed with a volume ratio of AgNO3/HF/H2O = 3:0.15:2.85. A drop of the 10 μL mixed solution was dropped on a prepared Si nanopillar array carefully. After a certain time of reaction, the substrate was rinsed with deionized water and then dried with nitrogen. For comparison, a prepared Si nanopillar array was dipped into the reaction solution to form AgNPs on its surface. After reacting for a certain time, the substrate was taken out and rinsed with deionized water and then dried with flow nitrogen. 2.4. Characterization. All of the scanning electron microscopy (SEM) images were taken with a Hitachi SU8020 field emission scanning electron microscope, operated at 3.0 kV. An optical fiber portable Raman spectrometer (B & W Tek Inc.) was used for collecting Raman spectra. The laser wavelength was 532 nm, and the laser intensity was set at 7.92 mW. The integration time of the Raman spectrum was 1000 ms.

deposition has been proposed. Bai et al. fabricated Ag nanorod arrays by electroless deposition of AgNPs on a Si nanowire, which exhibited high sensitivity in SERS detection. 22 Esmaielzadeh et al. demonstrated a facile soft chemical approach for fabricating ZnO/Ag nanorod arrays on the Si wafers. The uniform ZnO/Ag nanoarrays offered an appropriate platform for SERS sensing. The relative standard deviation (RSD) of the Raman intensity was down to 13.2%.23 Chen et al. created the SERS substrates by depositing AgNPs on polydopamine-coated pillars, which led to a good reproducibility (RSD < 8.00%).24 However, in these works, AgNPs were randomly distributed on the surface of the nanostructures, which led to a random distribution of hot spots. Some hot spots were formed on locations which can hardly be reached by the probe molecules (such as side wall and bottom of the nanostructures). Even if a few probe molecules could reach the randomly distributed hot spots, the enhancement factors of the hot spots are not the same, leading to a poor reproducibility. Herein, we propose an efficient method to fabricate uniformly distributed and accessible hot spots on the substrate. The SERS substrate is created by taking a droplet-confined electroless deposition (DCED) on a hydrophobic Si nanopillar array. The AgNPs are generated only on top of the Si nanopillars with this method. The hot spots formed by AgNPs are more accessible to the probe molecules, which results in an excellent reproducibility for SERS measurements. The RSD is down to 3.40%. There exists a great linear correlation between the concentration of Rhodamine 6G (R6G) and the Raman intensity in the log−log plot; R2 is 0.998, indicating that the SERS substrate can be applied for quantitative SERS analysis. Meanwhile, the minimum detection concentration is down to 10−11 M on the hydrophobic substrate, with R6G as a probe molecule.

3. RESULTS AND DISCUSSION Figure 1 shows the schematic illustration of DCED. The hydrophobic Si nanopillar array was created by nanosphere lithography, as shown in Figure 2A. According to the Cassie theory,25 the height and diameter of the Si nanopillars were set at 510 and 180 nm, respectively, to create a hydrophobic Si nanopillar array. After modification of fluoroalkylsilane, the contact angle (CA) of the array was increased to 139.5 from 55.6° (Figure S1). Then, the hydrophobic Si nanopillar array was used as a substrate to obtain AgNPs. The mixed solution (10 μL) of AgNO3 and HF was dropped on the hydrophobic substrate. The concentration of AgNO3 was 0.01 M, which could affect the deposition rate of AgNPs. The deposition rate increased with increase in the AgNO3 concentration, as shown in Figures S2 and S3. After 50 s of the reaction, the substrate was rinsed with deionized water and dried with nitrogen; the AgNPs were formed on the substrate. Figure 2B shows that the AgNPs are formed only on top of the Si nanopillars. This result might be introduced by the hydrophobicity of the surface, which prevents the mix solution from coming in contact with the surface except for the top of the Si nanopillars. Therefore, no AgNPs can be formed on the side wall and bottom of the Si nanopillars. To investigate the influence of hydrophobicity on the creation of the AgNPs, arrays with different CAs were used for creating the AgNPs, as shown in Figure S4. After 10 μL of the mixed solution being

2. EXPERIMENTAL SECTION 2.1. Chemicals and Materials. One-side-polished n-type (100)oriented Si wafers were purchased from Youyan Guigu, Beijing, China. Polystyrene (PS) spheres of 290 nm diameter, with less than 5% of diameter variation were obtained from Wuhan Tech Co., Ltd. Acetone, chloroform, ethanol, tetrahydrofuran, and hydrofluoric acid were purchased from Beijing Chemical Works. Trimethoxy (1H,1H,2H,2Hnonafluorohexyl) silane, AgNO3, and R6G were purchased from Sigma Aldrich Co. The solvents and chemicals were used without further purification. 2.2. Fabrication of Hydrophobic Si Nanopillar Arrays. The Si substrates were sonicated in acetone, chloroform, ethanol, and deionized water for 5 min, respectively, to remove organic contaminants. PS sphere monolayers were prepared on the Si substrates, as described elsewhere.19 A reactive ion etching (RIE) process was carried out on a Plasmalab Oxford 80 plus (ICP 65) B

DOI: 10.1021/acsami.7b04240 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 2. (A) SEM image of the created Si nanopillars, with diameter, height, and period of 180, 510, and 290 nm, respectively. (B, C) Side and top view SEM images of AgNPs formed on the Si nanopillars, respectively. (D) SEM image in high resolution of the AgNPs formed on the Si nanopillars.

dropped on the substrate with different CAs, the area of the spot increases from 3.8 to 16.6 mm2, with the CA reducing from 139.5 to 75.2°. It can be observed that the AgNPs are generated only on the top of the Si nanopillars when the CA is greater than 116.4° (Figures S5A,S5B), but when the CA is less than 116.4°, the AgNPs are generated over the bottom, the side wall, and top of the Si nanopillars (Figure S5C−E). It means that the transition from Cassie’s mode to Wenzel’s mode occurs at a CA of about 116.4° on this substrate. Figure 2C shows that the AgNPs are distributed uniformly on top of the Si nanopillars when the CA is 139.5°. It can be speculated that the droplet is kept in the Cassie state on the hydrophobic surface, which only contacts with the top of the Si nanopillars due to the air chamber formed under the droplet. Therefore, each top of the Si nanopillars under the droplet is a reaction unit, which benefits the uniform formation of the AgNPs on each Si nanopillar, as shown in the high-resolution SEM image in Figure 2D. For investigating the selectivity of formation of the AgNPs, we created the AgNPs by dipping the hydrophobic Si nanopillar array into the solution. Figure S6 shows that the AgNPs are distributed randomly on top of the Si nanopillars because the hydrophobicity makes nonuniform solution contact with the Si nanopillars. Therefore, the DCED method guarantees uniform distribution of the AgNPs. The statistical mean diameter of the AgNPs is about 45 nm and the statistical mean gap size of AgNPs on one Si nanopillar is about 6 nm, which are shown in Figures S7A,S7B. The area of the spot formed by the AgNPs is about 3.8 mm2 (Figure S8). After formation of the AgNPs, the CA of the Si nanopillars is still exhibited as 138°, as shown in Figure S9. As presented in Figure 3A, taking R6G as a probe molecule for the Raman measurement, the minimum detectable concentration is 10−11 M, but it is only 10−9 M on the substrate, with a CA of 75.2°, as shown in Figure S10. The much lower detectable concentration on the hydrophobic substrate might be contributed by the narrow nanogaps

Figure 3. (A) Raman spectra of R6G, with concentrations ranging from 10−8 to 10−11 M (10 μL). The laser of 532 nm is 7.92 mW. The integration time is 1000 ms. (B) Raman spectra of R6G (10−9 M) randomly collected on 20 random spots. (C) Linear correlation of the concentration of R6G and Raman intensity in the log−log plot.

between the AgNPs and the concentrated effect of the hydrophobicity of the substrate. Then, the Raman spectra were randomly recorded from 20 different areas on the hydrophobic substrate for evaluating the reproducibility. As shown in Figure 3B, for R6G (10−9 M), the RSD of the peak at 611 cm−1 is 3.40%. Figure 3C demonstrates the linear dependence relationship of Raman intensity of the peak at C

DOI: 10.1021/acsami.7b04240 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 4. Distribution of the electromagnetic field in the XY plane with the same number of AgNPs and nanogaps between the AgNPs of (A) 40, (B) 35, (C) 30, (D) 25, (E) 20, (F) 15, (G) 10, and (H) 5 nm, which is calculated using FDTD simulations.

611 cm−1 and the concentration of R6G. R2 is 0.998, which is up to the standard for quantitative SERS analysis compared to that of the other substrates (Table S1), but the RSD on the substrate, created by dipping a hydrophobic Si nanopillar array in solution, is up to 17.7% (Figure S11) because of the random distribution of AgNPs, resulting in different Raman intensities. The Raman signal intensity is too low to evaluate reproducibility on the hydrophilic Si nanopillars (with a CA of 75.2°, as shown in Figure S5E) when the concentration of R6G is 10−9 M. So we took R6G with a higher concentration (10 −8 M) as the probe molecule for evaluating the reproducibility, as shown in Figure S12. The RSD of the substrates, with CAs of 95.9 and 75.2° (as shown in Figures S5D,S5E) is 8.91 and 10.64%, respectively, which is higher than 3.74% obtained on the hydrophobic substrate (with CA of 139.5°). It can be speculated that the AgNPs formed on top of the Si nanopillars are more accessible for probe molecules than that formed on other locations, which results in a higher reproducibility. To further investigate the effect of size and distance of the AgNPs on Raman signal intensity, AgNPs with different nanogaps were fabricated by varying the electroless deposition time from 20 to 70 s, as shown in Figure S13. It is obvious that the AgNPs are preferentially formed around the top edge of the Si nanopillars because the density of the surface energy at the edge is larger than that in other areas.26 The nanogaps between the AgNPs gradually decrease from 32.17 to 5.95 nm when the deposition time is extended from 30 to 50 s; the statistical graph is shown in Figure S14. When the deposition time is more than 60 s, the adjacent AgNPs aggregate together and the nanogaps disappear. Actually, the diameters of the AgNPs increase when their nanogaps decrease gradually. Therefore, in our experiment, the nanogaps are not the only factor that affects the electromagnetic field intensity; the size of the AgNPs

could also affect the electromagnetic field intensity, so we chose to use electroless deposition time as the main factor to evaluate the Raman signal. The Raman signals of the samples, with electroless deposition time ranging from 20 to 70 s, are recorded and shown in Figure S15A. The statistics of the absolute signal strength at 611 cm−1 Raman shift (as shown in Figure S15B) shows that the Raman signal is gradually enhanced when the electroless deposition time is increased from 20 to 50 s. But the Raman signal decreases when the deposition time is more than 60 s, which should be caused by disappearing of the nanogaps between the AgNPs. To further study the effect of size and distance of the AgNPs on distribution of the electromagnetic field, the finite difference time domain (FDTD) software was used for simulating the distribution of the electromagnetic field. In this work, the size of the nanogaps and number and diameter of the AgNPs change simultaneously. Therefore, according to the experimental results, we built up two types of models to investigate the effect of nanogaps on distribution of the electromagnetic field. One is to change the size of nanogaps but keep the number of the AgNPs constant on a single Si nanopillar; another is to change the number of AgNPs on a single Si nanopillar but keep the size of the nanogaps constant. First, we simulated the distribution of the electromagnetic field using the models of type one by changing the size of the nanogaps from 40 to 5 nm and keeping 8 AgNPs on a single Si nanopillar. As shown in Figure 4, the hot spots are located between the AgNPs on a single Si nanopillar, and the intensity of the electromagnetic field increases obviously when decreasing the nanogaps to less than 20 nm; the (E/E0)4 is up to 5.1 × 104, with the nanogaps down to 5 nm. Then, we simulated the distribution of the electromagnetic field when the number of AgNPs was reduced from 20 to 4 on D

DOI: 10.1021/acsami.7b04240 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 5. Distribution of the electromagnetic field in the XY plane, with the same nanogaps, and the number of AgNPs on a single nanopillar of (A) 20, (B) 12, (C) 8, and (D) 4, which is calculated using FDTD simulations.



a single Si nanopillar, with 10 nm nanogaps. Figure 5 indicates that the intensity of the electromagnetic field is enhanced distinctly with reduction of the number of AgNPs. However, the experimental result shows that the nanogaps between AgNPs will disappear eventually when increasing the size of the AgNPs. In this case, we build up a continuous Ag ring model instead of the AgNPs. As revealed in Figure S16, the intensity of the electromagnetic field is much weaker than the models with nanogaps. In summary, strong electromagnetic fields can be obtained when the number of AgNPs is from 4 to 8 on a single Si nanopillar and the size of nanogaps between the AgNPs is less than 10 nm. The simulated results agree well with the experimental results. Figure S17 indicates that the hot spots are located only on top of the Si nanopillars, which are more accessible than other locations, resulting in an excellent reproducibility. To test the practical application of our substrate, we used melamine as a practical analyte to test the sensitivity and enrichment power of the created substrate. We measured the Raman signal of melamine aqueous solutions, with concentrations of 10−4, 10−5, 10−6, and 10−7 M, respectively. The Raman spectra are shown in Figure S18. The peak at 675 cm−1 caused by ring breathing vibration and the peak at 984.6 cm−1 caused by C-N-C bending vibration are usually regarded as the Raman marker bands of melamine.27 These marker bands can be obviously observed in Figure S18. The lowest detection concentration of melamine is 10−7 M with the substrate. The results indicate that the fabricated SERS-active substrates could be applied in practical detection.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.7b04240. CAs of the Si nanopillar array before and after the modification of fluoroalkylsilane; SEM images of the AgNPs fabricated with different concentrations of AgNO3 by DCED; SEM images of the AgNPs formed by taking different reaction times and concentrations of AgNO3; CAs of the arrays after taking oxygen plasma etching for different times and the corresponding spots formed by the AgNPs fabricated with DCDE; AgNPs created on the Si nanopillar array with different CAs; SEM images of the AgNPs fabricated by the immersing method; the statistical diameter and nanogaps of the AgNPs; SEM image of the spot of AgNPs formed by DCDE; CA of the Si nanopillar array with AgNPs, Raman spectra of R6G, with different concentrations on the substrate, with a CA of 75.2°; Raman spectra of R6G (10−9 M) recorded on the substrate fabricated by the immersing method; Raman spectra of R6G (10−8 M) randomly recorded on 20 random spots on the substrates, with different CAs; SEM images of the AgNPs formed by reacting for different durations; Statistical nanogaps of the AgNPs created by reacting for different times; Raman spectra collected on the substrates prepared by depositing for different durations and the absolute signal strength at 611 cm−1 Raman shift; distribution of the electromagnetic field in the XY plane of the connected AgNPs; distribution of the electromagnetic field in the XZ plane on a single pillar; Raman spectra of melamine with different concentrations (PDF)

4. CONCLUSIONS



We demonstrate an efficient method, DCED, to create uniformly distributed hot spots on ordered hydrophobic Si nanopillar arrays. By taking this method, the hot spots are formed uniformly on top of the hydrophobic Si nanopillars, which are more accessible for the analyte molecules. The substrate performs an excellent reproducibility in SERS measurements, the RSD is down to 3.40%, and the linear dependence relationship is up to the standard for quantitative SERS analysis. Meanwhile, the minimum detection concentration is down to 10−11 M on the hydrophobic substrate with Rhodamine 6G as a probe molecule. Meanwhile, the minimum detection concentration is down to 10−11 M on the hydrophobic substrate, with R6G as a probe molecule.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Nan Lu: 0000-0002-2988-6963 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (No. 21673096 and 21273092). E

DOI: 10.1021/acsami.7b04240 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces



(20) Krishnan, J. N.; Kim, I. T.; Ahn, S. H.; Kim, Z. H.; Cho, S. H.; Kim, S. K. Electroless Deposition of SERS Active Au-Nanostructures on Variety of Metallic Substrates. BioChip J. 2013, 7, 375−385. (21) Cheng, M.-L.; Yang, J. Seed-Mediated Growth Method for Electroless Deposition of AgNPs on Glass Substrates for Use in SERS Measurements. J. Raman Spectrosc. 2009, 41, 167−174. (22) Bai, F.; Li, M.; Fu, P.; Li, R.; Gu, T.; Huang, R.; Chen, Z.; Jiang, B.; Li, Y. Silicon Nanowire Arrays Coated with Electroless Ag for Increased Surface-Enhanced Raman Scattering. APL Mater. 2015, 3, No. 056101. (23) Kandjani, A. E.; Mohammadtaheri, M.; Thakkar, A.; Bhargava, S. K.; Bansal, V. Zinc Oxide/silver Nanoarrays as Reusable SERS Substrates with Controllable ‘Hot-Spots’ for Highly Reproducible Molecular Sensing. J. Colloid Interface Sci. 2014, 436, 251−257. (24) Chen, J.; Li, Y.; Huang, K.; Wang, P.; He, L.; Carter, K. R.; Nugen, S. R. Nanoimprinted Patterned Pillar Substrates for SurfaceEnhanced Raman Scattering Applications. ACS Appl. Mater. Interfaces 2015, 7, 22106−22113. (25) Cassie, A. B. D.; Baxter, S. Wettability of Porous Surfaces. Trans. Faraday Soc. 1944, 40, 546−551. (26) Gentile, F.; Laura Coluccio, M.; Candeloro, P.; Barberio, M.; Perozziello, G.; Francardi, M.; Di Fabrizio, E. Electroless Deposition of Metal Nanoparticle Clusters: Effect of Pattern Distance. J. Vac. Sci. Technol., B: Nanotechnol. Microelectron.: Mater., Process., Meas., Phenom. 2014, 32, No. 031804. (27) Meier, R. J.; Maple, J. R.; Hwang, M. J.; Hagler, A. T. Molecular Modeling Urea- and Melamine-Formaldehyde Resins. 1. A Force Field for Urea and Melamine. J. Phys. Chem. 1995, 99, 5445−5456.

REFERENCES

(1) Du, Y.; Shi, L.; He, T.; Sun, X.; Mo, Y. SERS Enhancement Dependence on the Diameter and Aspect Ratio of Silver-Nanowire Array Fabricated by Anodic Aluminium Oxide Template. Appl. Surf. Sci. 2008, 255, 1901−1905. (2) Yi, Z.; Xu, X.; Li, X.; Luo, J.; Wu, W.; Tang, Y.; Yi, Y. Facile Preparation of Au/Ag Bimetallic Hollow Nanospheres and its Application in Surface-Enhanced Raman Scattering. Appl. Surf. Sci. 2011, 258, 212−217. (3) Le Ru, E. C.; Meyer, M.; Blackie, E.; Etchegoin, P. G. Advanced Aspects of Electromagnetic SERS Enhancement Factors at a Hot Spot. J. Raman Spectrosc. 2008, 39, 1127−1134. (4) Xia, L.; Chen, M.; Zhao, X.; Zhang, Z.; Xia, J.; Xu, H.; Sun, M. Visualized Method of Chemical Enhancement Mechanism on SERS and TERS. J. Raman Spectrosc. 2014, 45, 533−540. (5) Xu, H.; Aizpurua, J.; Kall, M.; Apell, P. Electromagnetic Contributions to Single-Molecule Sensitivity in Surface-Enhanced Raman Scattering. Phys. Rev. E 2000, 62, 4318−4324. (6) Lim, D. K.; Jeon, K. S.; Hwang, J. H.; Kim, H.; Kwon, S.; Suh, Y. D.; Nam, J. M. Highly Uniform and Reproducible Surface-Enhanced Raman Scattering from DNA-Tailorable Nanoparticles with 1-nm Interior Gap. Nat. Nanotechnol. 2011, 6, 452−460. (7) Lindquist, N. C.; Nagpal, P.; Lesuffleur, A.; Norris, D. J.; Oh, S. H. Three-Dimensional Plasmonic Nanofocusing. Nano Lett. 2010, 10, 1369−1373. (8) Shen, J.; Su, J.; Yan, J.; Zhao, B.; Wang, D.; Wang, S.; Li, K.; Liu, M.; He, Y.; Mathur, S.; Fan, C.; Song, S. Bimetallic Nano-Mushrooms with DNA-Mediated Interior Nanogaps for High-Efficiency SERS Signal Amplification. Nano Res. 2014, 8, 731−742. (9) Su, J.; Wang, D.; Norbel, L.; Shen, J.; Zhao, Z.; Dou, Y.; Peng, T.; Shi, J.; Mathur, S.; Fan, C.; Song, S. Multicolor Gold-Silver NanoMushrooms as Ready-to-Use SERS Probes for Ultrasensitive and Multiplex DNA/miRNA Detection. Anal. Chem. 2017, 89, 2531− 2538. (10) Fang, Y.; Seong, N. H.; Dlott, D. D. Measurement of the Distribution of Site Enhancements in Surface-Enhanced Raman Scattering. Science 2008, 321, 388−392. (11) Fromm, D. P.; Sundaramurthy, A.; Schuck, P. J.; Kino, G.; Moerner, W. E. Gap-Dependent Optical Coupling of Single “Bowtie” Nanoantennas Resonant in the Visible. Nano Lett. 2004, 4, 957−961. (12) Wu, Y.; Hang, T.; Yu, Z.; Gu, J.; Li, M. Quasi-Periodical 3D Hierarchical Silver Nanosheets with Sub-10 nm Nanogap Applied as an Effective and Applicable SERS Substrate. Adv. Mater. Interfaces 2015, 2, No. 1500359. (13) Kanipe, K. N.; Chidester, P. P.; Stucky, G. D.; Moskovits, M. Large Format Surface-Enhanced Raman Spectroscopy Substrate Optimized for Enhancement and Uniformity. ACS Nano 2016, 10, 7566−7571. (14) Li, W.; Camargo, P. H. C.; Lu, X.; Xia, Y. Dimers of Silver Nanospheres: Facile Synthesis and Their Use as Hot Spots for SurfaceEnhanced Raman Scattering. Nano Lett. 2009, 9, 485−490. (15) Diebold, E. D.; Peng, P.; Mazur, E. Isolating Surface-Enhanced Raman Scattering Hot Spots Using Multiphoton Lithography. J. Am. Chem. Soc. 2009, 131, 16356−16357. (16) Shen, W.; Lin, X.; Jiang, C.; Li, C.; Lin, H.; Huang, J.; Wang, S.; Liu, G.; Yan, X.; Zhong, Q.; Ren, B. Reliable Quantitative SERS Analysis Facilitated by Core-Shell Nanoparticles with Embedded Internal Standards. Angew. Chem., Int. Ed. Engl. 2015, 54, 7308−7312. (17) Oh, M.-K.; Shin, Y.-S.; Lee, C.-L.; De, R.; Kang, H.; Yu, N. E.; Kim, B. H.; Kim, J. H.; Yang, J.-K. Morphological and SERS Properties of Silver Nanorod Array Films Fabricated by Oblique Thermal Evaporation at Various Substrate Temperatures. Nanoscale Res. Lett. 2015, 10, 259. (18) Leng, W.; Vikesland, P. J. MGITC Facilitated Formation of AuNP Multimers. Langmuir 2014, 30, 8342−8349. (19) Ma, R.; Lu, N.; Liu, L.; Wang, Y.; Shi, S.; Chi, L. Fabrication of Single Gold Particle Arrays with Pattern Directed Electrochemical Deposition. ACS Appl. Mater. Interfaces 2012, 4, 3779−3783. F

DOI: 10.1021/acsami.7b04240 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX