Highly uniform indicator-mediated SERS sensor platform for the detection of Zn²⁺

Abstract

An indicator-mediated surface-enhanced Raman scattering sensor platform with highly uniform SERS sensitivity was fabricated by modifying 4-mercaptopyridine molecules onto the surface of Ag nanoparticles that were anchored onto a silicon wafer. The present SERS sensor platform realized quantitative detection of Zn²⁺.

---

1. Introduction

Zinc ion (Zn²⁺) has attracted a great deal of attention ascribed to its biological significance. Zn²⁺ is believed to be involved in some pathological processes, such as Alzheimer's disease, epilepsy, ischemic stroke, and infantile diarrhea.^1–3 Some methodologies have been established to determine Zn²⁺ over the last decades, including atomic absorption spectrophotometry, fluorescence spectroscopy, mass spectrometry and electrochemical analysis.^3–6 However, these techniques require extensive sample preparation.^7–11 So, it is required to develop a facile technique for the determination of Zn²⁺.

Surface-enhanced Raman scattering (SERS) as a prospective analytical method has attracted more and more attentions owing to its advantages such as low cost, high selectivity, ultrasensitivity, long-term stability and good compatibility with aqueous environments.^12–15 However, as a molecular spectroscopy technique, SERS could only reveal the vibrational, rotational and other low-frequency modes in a system. In principle, the SERS technique cannot be directly utilized to analyze the atomic ions, which are lack of the vibrational signatures.^12,16 In order to use the SERS to detect the atomic ions, a so-called indicator-mediated SERS strategy was developed. In this strategy, molecule indicators are modified onto the plasmonic surfaces of metal.^16–20 When the indicator-mediated SERS sensor is used to detect the atomic ions, the interaction between the molecule indicator and atomic ions could influence the Raman band intensity and shift of the indicator. Such variations of Raman spectroscopy could be utilized as the spectral characteristics for the detection of atomic ions.²¹ Chen et al. used the variation of the Raman band intensity of molecule indicator functionalized silver nanoparticles for the highly sensitive detection of Hg²⁺ and As³⁺.^22,23 The variation of the Raman band shift was also used for the detection of some ions.^24–26 The indicator-mediated SERS sensor based on colloidal nanoparticles had been used for the detection of Zn²⁺.¹⁷ However, this colloidal nanoparticles-based sensor is not adequate for quantitative detection and the real time and in situ monitoring of Zn²⁺. A highly uniform plasmonic platform is more promising for the application of the indicator-mediated SERS strategy. Moreover, in the previously reported SERS sensor, 4-(2-pyridylazo) resorcinol and 1-(2-pyridylazo)-2-naphthol have been used as colorimetric reagent for Zn²⁺ since they can form very stable and highly colored complexes with Zn²⁺.^17,27 However, the similar Raman spectra of the complexes make it difficult to distinguish the Zn²⁺ from the other ions. In order to promote the selectivity of the indicator-mediated SERS sensor for Zn²⁺, an appropriate molecule indicator owning specific interaction with Zn²⁺ should be developed. Such molecule indicator should have distinct Raman characteristics, and could be efficiently and rationally decorated onto the plasmonic surfaces of metals. The interaction between molecule indicator and target Zn²⁺ should also be investigated to promote the indicator-mediated SERS strategy. Accordingly, selecting a special molecule indicator, taking insight into the interaction between molecule indicator and target Zn²⁺ and developing a highly uniform plasmonic platform for detecting Zn²⁺ would be a key issue to develop high performance indicator-mediated SERS sensor for the detection of Zn²⁺.

In this study, 4-mercaptopyridine (4-MPY) was chosen as the special molecule indicator to be modified onto the surface of the Ag nanoparticles, and a highly uniform indicator-mediated SERS sensor platform for Zn²⁺ was fabricated. The SERS sensor platform can distinguish Zn²⁺ from various common metal ions. Especially, it can distinguish Zn²⁺ from Cd²⁺ that is usually difficult for fluorescent sensor and SERS sensor.^9,18 Quantitative detection of Zn²⁺ in aqueous solution is realized by employing the linear dependence of area ratios between two distinctive Raman bands (the shifted Raman band 1027 cm⁻¹ and the unshifted band 1102 cm⁻¹ of 4-MPY) on the logarithmic concentrations of Zn²⁺. Moreover, the interaction between Zn²⁺ and 4-MPY had been investigated by the means of Raman, FTIR and XPS. The current work represents an effective technique for detection of Zn²⁺ in aqueous solution, which is prospective for life sciences.

2. Experimental details

2.1 SERS-active substrate preparation

The Ag nanoparticles-anchored Si (AgNP–Si) substrate was prepared by immersing the cleaned Si wafer into an aqueous solution consisting of 5 mM AgNO₃ and 4.8 M HF for 10 seconds. After rinsing thoroughly with deionized water, the AgNP–Si substrate was soaked in the 1 mM 4-MPY ethanol solution for 1 hour to assembling 4-MPY molecules onto the surface of AgNPs. Then the substrate was washed with ethanol to remove the physically adsorbed 4-MPY molecules and the SERS-active substrate was obtained.

2.2 Metal ions detection

The as-prepared SERS-active substrates were immersed respectively in aqueous solutions of metal chlorides containing various metal ions, including Zn²⁺, Mn²⁺, Ca²⁺, K⁺, Ni²⁺, Fe²⁺, Cd²⁺ and Cr³⁺. After the immersion, the substrates were rinsed with deionized water and dried in air. SERS spectra were collected from the SERS-active substrates treated with above metal ions.

2.3 Instrumentation

The SERS measurements were performed using a Renishaw Raman system-inVia-Reflex with an excitation laser wavelength of 532 nm, excitation intensity of 0.5 mW, and data acquisition time of 10 s throughout these experiments. The laser beam was focused on a spot about 3 μm in diameter by a 50× microscope objective. The Raman band of a silicon wafer at 520 cm⁻¹ was employed to calibrate the spectrometer. FTIR spectra of all the samples were recorded using Excalibur 3100 with a resolution of 4 cm⁻¹. The morphologies of the substrates were characterized by field emission scanning electron microscopy (FE-SEM, Hitachi S-4800) and atomic force microscopy (Bruker Multimode 8) under tapping-mode. The kind of tip for AFM images is Bruker RTESPA. The X-ray photoelectron spectroscopy (XPS) was recorded by a PHIQUANTERA-IISXM at a voltage of 15 kV.

3. Results and discussion

By immersing the cleaned Si wafer into an aqueous solution consisting of AgNO₃ and HF for a desired time, Ag deposition could take place on the surface of the Si wafer and the Ag nanoparticles-coated Si (AgNP–Si) substrate was obtained.²⁸ The morphology of the as-prepared AgNP–Si substrate is shown in Fig. 1 and S1.† The 2D and 3D AFM topographical images of AgNP–Si substrate are identical with the SEM images, as shown in Fig. S2.† The spherical AgNPs have an average diameter of 70 nm, which is appropriate for SERS enhancement.²⁹

Fig. 1 SEM image of AgNP–Si substrate.

The substrate was considered to be chemically modified with 4-MPY (molecule indicator) after being treated with 4-MPY ethanol solution. Fig. 2a shows the Raman spectra of the 4-MPY molecules-modified AgNP–Si substrate (SiAg–4MPY). The measured Raman spectrum of 4-MPY shows a relatively strong Raman peak intensity at the concentration of 1 mM, which is hard to be detected in aqueous solution by the means of Raman spectrum. The AgNP–Si substrate exhibits a significant enhancement on Raman signal of 4-MPY. It implies that the AgNP–Si substrate is appropriate for SERS enhancement. Moreover, to examine the uniform SERS sensitivity of the substrate, the SERS spectra of 4-MPY form 441 points on the substrate were obtained, as shown in Fig. S3a.† The relative standard deviation of the Raman counts from the 441 points at 1102 cm⁻¹ is 6.2%, which indicates that the AgNP–Si substrate has high uniformity of SERS sensitivity. The kind of Ag-based SERS substrates could decrease SERS activity due to the adsorption of chloride ions or the formation of AgCl on the Ag surface.³⁰ To test the SERS activity of the substrate from the Cl⁻ attack, the SERS substrates have been treated by KCl with different concentrations for 5 hours. The results have shown that the SERS spectra of the sensor did not change even when the concentration of KCl was as high as 500 mM, as shown in Fig. S4a.†

Fig. 2 (a) the FT-Raman spectra of pure 4-MPY, the SERS spectra of SiAg–4MPY and SiAg–4MPY–Zn²⁺. (b) SERS spectra of 4-MPY (curve a) and SiAg–4MPY be treated with metal ions: Zn²⁺ (curve b), Mn²⁺ (curve c), Ca²⁺ (curve d), K⁺ (curve e), Ni²⁺ (curve f), Fe²⁺ (curve g), Cd²⁺ (curve h) and Cr³⁺ (curve i).

The SiAg–4MPY was used for the detection of an important metal ion, i.e. Zn²⁺. After treating the SiAg–4MPY with Zn²⁺ solution (10 mM) (SiAg–4MPY–Zn²⁺), new Raman bands located at 1027 cm⁻¹ and 1594 cm⁻¹ were observed, as shown in Fig. 2a. The appearance of new Raman bands makes it possible to utilize the SiAg–4MPY as an indicator-mediated SERS sensor for the detection of Zn²⁺. To demonstrate the selectivity of the SERS sensor for Zn²⁺, the response of the SERS sensor to various metal ions was measured, which results are shown in Fig. 2b. By careful comparing the SERS spectra of SiAg–4MPY which are treated with 8 kinds of metal ions, Mn²⁺, Ca²⁺, K⁺, Ni²⁺, Fe²⁺, Cd²⁺ and Cr³⁺ do not induce the two specific Raman bands of 4-MPY as Zn²⁺ does. And the two new Raman bands could be employed as the spectral characteristics to selectively detect Zn²⁺. Furthermore, the Raman spectra response of SiAg–4MPY toward Zn²⁺ in the presence of the other cations was also investigated (Fig. S5†). The results showed that Mn²⁺, Ca²⁺, K⁺, Ni²⁺ and Fe²⁺ have ignorable influences on the detection of Zn²⁺, but the presence of Cd²⁺ and Cr³⁺ will inhabit the SERS sensor for detection of Zn²⁺. It implies that the presented SERS sensor platform can distinguish Zn²⁺ from the other 7 cations separately, but it can't detect Zn²⁺ when Cd²⁺ or Cr³⁺ is present. It should be mentioned that although the presented SERS sensor platform can't detect Zn²⁺ when Cd²⁺ is present, but it can distinguish Zn²⁺ from Cd²⁺ separately which is also an improvement for SERS sensor as the previous report obtained indistinguishable SERS spectra when detect Zn²⁺ and Cd²⁺.¹⁸ The results showed that the SERS sensor exhibited excellent selectivity for Zn²⁺ over various common metal ions. The anti-interference ability of the sensor should be further improved, and the follow-up optimization work aiming to improve the sensitivity and the anti-interference ability of the sensor is under way.

The intensities of the two new Raman bands varied with the immersing times, as shown in Fig. S6.† To obtain the quantitative information of the sensor platform, we investigate the detection of Zn²⁺ with concentrations range from 0.1 mM to 10 mM. For this detection, the SiAg–4MPY substrates were immersed in Zn²⁺ solutions for 20 hours. Fig. S7† shows that the intensity of the two new bands of 1027 cm⁻¹ and 1594 cm⁻¹ are enhanced with increase of the Zn²⁺ concentration. The quantitative information regarding the concentration of Zn²⁺ was obtained by comparing the band area ratios between the shifted Raman band 1027 cm⁻¹ and the unshifted band 1102 cm⁻¹ of 4-MPY. As the overlaps with its neighbor peak, the peak of 1027 cm⁻¹ should be separately first. The areas under the bands 1008 cm⁻¹ and 1027 cm⁻¹ were deconvoluted assuming a Lorentzian shape, where the band position and the full width at half maxima were fixed (Fig. 3a).¹⁶ And the area under the band 1102 cm⁻¹ was also obtained by fitting the band with a Lorentzian shape. Applying this process to all the spectra, the band area ratio S₁₀₂₇/S₁₁₀₂ between band 1027 cm⁻¹ and band 1102 cm⁻¹ can be obtained for each spectrum. We plotted the area ratios as a function of the Zn²⁺ concentrations. Fig. 3b reveals a linear relation between the area ratios and the logarithmic concentrations. It has been discussed in many previous reports that the SERS intensity of the molecular probe linearly increases with the logarithm of the concentration of target.^31–34 It is believed that the logarithm law is related with the adsorption behaviors of Zn²⁺. This linear curve provides a calibration for quantitative detection of Zn²⁺. Moreover, the SERS spectra from the sensor indicate the high uniformity of the sensor, as shown in Fig. S3b.† The relative standard deviation of the Raman counts from the 441 points at 1027 cm⁻¹ is 6.5%. The influence of Cl⁻ on the SERS detection of Zn²⁺ was investigated by adding KCl (10–500 mM) to 10 mM ZnCl₂ solution. The results shown in Fig. S4b† reveal no obvious changes in the SERS spectra of theses samples, implying that the influence of Cl⁻ on the SERS sensor platform is negligible. The results indicate that SiAg–4MPY could be used as an effective and uniform SERS sensor platform.

Fig. 3 (a) SERS spectra of SiAg–4MPY used to detect Zn²⁺ at different concentrations. Bands where the deconvolution was carried out are highlighted (black, experimental spectra; yellow, resulting spectra of coadding peak 1, red, peak 2, magenta and peak 3, gray). (b) The linear relationship between the ratios S₁₀₂₇/S₁₁₀₂ and the logarithmic concentrations.

To improve the detection performance of this sensor platform, we have optimized the structure of the SERS substrate by depositing more Ag nanoparticles. The immersing time for depositing Ag nanoparticles had been extended to be 60 s, 180 s and 300 s. The SEM images of as-optimized SERS substrate in Fig. S8† show that Ag nanoparticles were getting bigger and getting more. The optimized sensor platforms were used to detect Zn²⁺ and the results are shown in Fig. S9.† The optimized sensor platforms exhibit stronger band intensity as compared with the un-optimized platform at the same concentration of Zn²⁺. It implies that the sensor platform can be optimized to improve its detection performance.

To understand the selectivity of the SERS sensor for Zn²⁺, it is necessary to investigate the origin of two new Raman bands in the case of Zn²⁺. In order to explore the origin of two new Raman bands, the Raman spectrum of the sensor response to Zn²⁺ (SiAg–4MPY–Zn²⁺) was compared with that of the composite of zinc and 4-MPY (i.e., Zn–4MPY). The Zn–4MPY was synthesized by the reaction of 4-MPY and ZnCl₂ in aqueous solution with vigorous stirring. The reaction flask was wrapped with Al foil to minimize exposure to ambient light. After 1 h reaction, the resultant white precipitate (Zn–4MPY) was collected via filtration and washed with deionized water and then dried in air at room temperature. Fig. 4a shows the FT-Raman spectrum of Zn–4MPY. It has to be mentioned that, according to some reports, the organic–inorganic material might be reacted under laser illumination and thus led to the change of Raman spectrum.³⁵ In order to clarify that the spectrum obtained in our experiment is still the Raman spectrum of Zn–4MPY, the vibrational spectroscopic technique (FTIR) was used to study Zn–4MPY, as shown in Fig. 4a. The FTIR spectrum band frequencies of Zn–4MPY are in accordance with Raman Spectrum band frequencies of Zn–4MPY shown in Fig. 2a. The FTIR spectrum reveals the intrinsic vibrational and rotational modes of Zn–4MPY. The accordance of the FTIR spectrum band frequencies and the Raman spectrum band frequencies implies that the FT-Raman spectrum obtained in Fig. 4a should be the Raman spectrum of Zn–4MPY without any change.

Fig. 4 (a) the Raman spectrum and FTIR spectrum of Zn–4MPY, the SERS spectrum of SiAg–4MPY–Zn²⁺. Inset shows the 4-MPY molecule structure diagram. (b) XPS results of N 1s region for Zn–4MPY and 4MPY molecules.

The FT-Raman spectrum of Zn–4MPY was compared with the SERS spectrum of SiAg–4MPY–Zn²⁺, as shown in Fig. 4a. The two new bands observed in the SERS spectrum of SiAg–4MPY–Zn²⁺ correspond well to the two most intensive bands of Zn–4MPY. It reveals that the origin of two new Raman bands for SiAg–4MPY–Zn²⁺ came from Zn–4MPY. The ring breathing observed at 988 cm⁻¹ in the case of 4-MPY molecules shifted to 1021 cm⁻¹ in the case of Zn–4MPY and 1027 cm⁻¹ in the case of SiAg–4MPY–Zn²⁺. The C–C stretching observed at 1618 cm⁻¹ in the case of 4-MPY molecules shifted to 1593 cm⁻¹ in the case of Zn–4MPY and 1594 cm⁻¹ in the case of SiAg–4MPY–Zn²⁺.^19,36 The shifted ring breathing and C–C stretching implies the interaction of functional group in 4-MPY with Zn²⁺. 4-MPY is a typical aromatic thiol compound. It consists of a thiol group in the para position to N atom in the pyridine ring. This structure may allow it to form well-ordered self-assembled monolayers on the Ag surface through the thiol group, leaving the pyridyl nitrogen to interact with Zn²⁺.^25,37 Moreover, the nitrogen atom contains the relatively high-energy non-bonding electron pair which makes the nitrogen atom easy to be bound by coordination of Zn²⁺.¹ The electron donation from nitrogen orbitals to Zn²⁺ leads to the charge redistribution in the pyridine ring of 4-MPY and the charge redistribution results in the variation of static polarizabilitities of the ring breathing and C–C stretching. Accordingly, the Raman shifts of the two vibrational modes are observed.³⁸

X-ray photoelectron spectroscopy (XPS) is a surface-sensitive quantitative spectroscopic technique that can be used to study the chemical state and electronic state of the N element of 4-MPY before and after the interaction with Zn²⁺. Xue et al. fabricated TiO₂/4-MPY/Ag sandwich structure to enhance Raman signal and investigated the interaction between 4-MPY and AgNPs by means of XPS.³⁹ The results indicated that the Ag–S bond was more stable than the Ag–N and the AgNPs were mainly assembled through the S atom of 4-MPY.³⁹ In regarding of this, the XPS analysis was performed mainly for the N 1s region. We have performed the experiment to investigate the N element of 4-MPY before and after the interaction with Zn²⁺. But unfortunately, none XPS spectrum of N 1s region was obtained for the samples. This might be due to that the very little amount of 4-MPY molecules modified on the substrate and the molecules cannot be recognized by XPS. The SERS spectrum of SiAg–4MPY–Zn²⁺ have the band vibration according to Zn–4MPY, as can be seen in Fig. 4a. So the interaction between 4-MPY molecule and Zn²⁺, which induced the Raman band accordance, is same in SiAg–4MPY–Zn²⁺ and Zn–4MPY. Therefore, to investigate the interaction between 4-MPY molecule and Zn²⁺, an XPS analysis was performed for the 4-MPY molecules and Zn–4MPY, as shown in Fig. 4b. The N 1s signal of 4-MPY molecules is centered at 399.9 eV.^39,40 It is noticed that one peak located at 398.8 eV was observed for Zn–4MPY in Fig. 4b. The shift in binding energy of N 1s orbitals from 4-MPY molecules to Zn–4MPY implies that pyridyl nitrogen had interacted with Zn²⁺, which is accordance with the analyzed results of the as discussed Raman spectral data.

4. Conclusion

A highly uniform SERS sensitivity indicator-mediated SERS sensor platform of 4-MPY molecules modifying Ag nanoparticles based on silicon wafer (SiAg–4MPY) was proposed for the selective and quantitative detection of Zn²⁺. The SERS spectrum of the molecule indicator used to detect Zn²⁺ was characterized by features of the two new Raman bands observed at 1027 cm⁻¹ and 1594 cm⁻¹. Through the Raman spectra and XPS analysis, it was proved that 4-MPY molecules bind on Ag surface through the thiol group and interact with Zn²⁺ through the pyridyl nitrogen. It can be determined that the interaction between pyridyl nitrogen of 4-MPY molecule and Zn²⁺ influences the ring breathing and C–C stretching of the molecule, and results in the two new Raman bands. The indicator-mediated SERS sensor platform allows for selective and quantitative detection of Zn²⁺ in aqueous solution. We anticipate this kind of sensor to be of key importance in the study of environmental protection, food safety applications and life sciences.

Acknowledgements

This work was supported by Chinese Academy of Sciences (Grant KGZD-EW-T02), NSFC (Grants 51272302, 51272258 and 91333119), National Basic Research Program of China (973 Program) (Grant 2012CB932400) and Youth Innovation Promotion Association CAS.

Notes and references

1. Z. Xu, J. Yoon and D. R. Spring, Chem. Soc. Rev., 2010, 39, 1996–2006.
2. N. Singla, A. Tripathi, M. Rana, S. Kishore Goswami, A. Pathak and P. Chowdhury, J. Lumin., 2015, 165, 46–55.
3. J. E. Kwon, S. Lee, Y. You, K. H. Baek, K. Ohkubo, J. Cho, S. Fukuzumi, I. Shin, S. Y. Park and W. Nam, Inorg. Chem., 2012, 51, 8760–8774.
4. D. W. Domaille, E. L. Que and C. J. Chang, Nat. Chem. Biol., 2008, 4, 168–175.
5. Z. Xu, K. H. Baek, H. N. Kim, J. Cui, X. Qian, D. R. Spring, I. Shin and J. Yoon, J. Am. Chem. Soc., 2010, 132, 601–610.
6. P. Wang, L. Liu, P. Zhou, W. Wu, J. Wu, W. Liu and Y. Tang, Biosens. Bioelectron., 2015, 72, 80–86.
7. D. L. Giokas, G. Z. Tsogas, A. G. Vlessidis and M. I. Karayannis, Anal. Chem., 2004, 76, 1302–1309.
8. Q. W. He, E. W. Miller, A. P. Wong and C. J. Chang, J. Am. Chem. Soc., 2006, 128, 9316–9317.
9. X. J. Peng, J. J. Du, J. L. Fan, J. Y. Wang, Y. K. Wu, J. Z. Zhao, S. G. Sun and T. Xu, J. Am. Chem. Soc., 2007, 129, 1500.
10. L. Xue, C. Liu and H. Jiang, Org. Lett., 2009, 11, 1655–1658.
11. L. Ding, W. Cheng, X. J. Wang, S. J. Ding and H. X. Ju, J. Am. Chem. Soc., 2008, 130, 7224.
12. J. F. Li, Y. F. Huang, Y. Ding, Z. L. Yang, S. B. Li, X. S. Zhou, F. R. Fan, W. Zhang, Z. Y. Zhou, Y. Wu de, B. Ren, Z. L. Wang and Z. Q. Tian, Nature, 2010, 464, 392–395.
13. P. L. Stiles, J. A. Dieringer, N. C. Shah and R. P. Van Duyne, Annu. Rev. Anal. Chem., 2008, 1, 601–626.
14. J.-C. Bian, Z. Li, Z.-D. Chen, H.-Y. He, X.-W. Zhang, X. Li and G.-R. Han, Appl. Surf. Sci., 2011, 258, 1831–1835.
15. A. J. Frank, A. McEneny-King, N. Cathcart and V. Kitaev, RSC Adv., 2015, 5, 73919–73925.
16. D. Tsoutsi, J. M. Montenegro, F. Dommershausen, U. Koert, L. M. Liz-Marzan, W. J. Parak and R. A. Alvarez-Puebla, ACS Nano, 2011, 5, 7539–7546.
17. L. Szabó, L. F. Leopold, B. I. Cozar, N. Leopold, K. Herman and V. Chiş, Cent. Eur. J. Chem., 2011, 9, 410–414.
18. J. Yin, T. Wu, J. Song, Q. Zhang, S. Liu, R. Xu and H. Duan, Chem. Mater., 2011, 23, 4756–4764.
19. Y. Zhao, J. N. Newton, J. Liu and A. Wei, Langmuir, 2009, 25, 13833–13839.
20. S. J. Lee and M. Moskovits, Nano Lett., 2011, 11, 145–150.
21. R. A. Alvarez-Puebla and L. M. Liz-Marzan, Angew. Chem., 2012, 51, 11214–11223.
22. L. Chen, N. Qi, X. Wang, L. Chen, H. You and J. Li, RSC Adv., 2014, 4, 15055.
23. J. Li, L. Chen, T. Lou and Y. Wang, ACS Appl. Mater. Interfaces, 2011, 3, 3936–3941.
24. Z. Jurasekova, A. Torreggiani, M. Tamba, S. Sanchez-Cortes and J. V. Garcia-Ramos, J. Mol. Struct., 2009, 918, 129–137.
25. L. G. Crane, D. X. Wang, L. M. Sears, B. Heyns and K. Carron, Anal. Chem., 1995, 67, 360–364.
26. A. Burneau and B. Teiten, Vib. Spectrosc., 1999, 21, 97–109.
27. L. Szabo, K. Herman, N. E. Mircescu, A. Falamas, L. F. Leopold, N. Leopold, C. Buzumurga and V. Chis, Spectrochim. Acta, Part A, 2012, 93, 266–273.
28. K. Q. Peng, Y. J. Yan, S. P. Gao and J. Zhu, Adv. Funct. Mater., 2003, 13, 127–132.
29. C. J. Orendorff, A. Gole, T. K. Sau and C. J. Murphy, Anal. Chem., 2005, 77, 3261–3266.
30. T. Murphy, H. Schmidt and H. D. Kronfeldt, Appl. Phys. B: Lasers Opt., 1999, 69, 147–150.
31. B. Teiten and A. Burneau, J. Raman Spectrosc., 1997, 28, 879–884.
32. J. Wang, X. Wu, C. Wang, N. Shao, P. Dong, R. Xiao and S. Wang, ACS Appl. Mater. Interfaces, 2015, 7, 20919–20929.
33. C. Song, L. Min, N. Zhou, Y. Yang, S. Su, W. Huang and L. Wang, ACS Appl. Mater. Interfaces, 2014, 6, 21842–21850.
34. T. W. Lin, H. Y. Wu, T. T. Tasi, Y. H. Lai and H. H. Shen, Phys. Chem. Chem. Phys., 2015, 17, 18443–18448.
35. Y. M. Lee, K. Kim and K. S. Shin, J. Phys. Chem. C, 2008, 112, 10715–10720.
36. W. H. Do, C. J. Lee, D. Y. Kim and M. J. Jung, J. Ind. Eng. Chem., 2012, 18, 2141–2146.
37. A. Calloni, A. Brambilla, G. Berti, G. Bussetti, E. V. Canesi, M. Binda, A. Petrozza, M. Finazzi, F. Ciccacci and L. Duò, Langmuir, 2013, 29, 8302–8310.
38. S. Liu, Y. Li, X. Zhao, X. Liu and M. Chen, Spectrochim. Acta, Part A, 2011, 82, 205–212.
39. X. Xue, D. Xu, W. Ruan, L. Chen, L. Chang and B. Zhao, RSC Adv., 2015, 5, 64235–64239.
40. H. Hu, W. Song, W. Ruan, Y. Wang, X. Wang, W. Xu, B. Zhao and Y. Ozaki, J. Colloid Interface Sci., 2010, 344, 251–255.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra28041a

---