Hydrogenated black TiO₂ nanowires decorated with Ag nanoparticles as sensitive and reusable surface-enhanced Raman scattering substrates

Abstract

Recently, surface-enhanced Raman scattering (SERS) has been the subject of great interest because of its ultrasensitive, rapid and nondestructive analysis, detection, and imaging. Highly sensitive, stable and recyclable substrates are of great importance in fulfilling the practical potential of the technique in various fields, such as trace analysis, diagnosis and criminal investigation. In this work, we introduce wafer-scale Ag-modified hydrogenated black TiO₂ nanowires as active SERS substrate by hydrogenation. The Ag/H–TiO₂ NWs substrates deliver a relative enhancement factor of up to ∼10² compared with the substrate before hydrogenation and a detection limitation improved by at least three orders of magnitude compared with R6G detection. Moreover, hydrogenated TiO₂/Ag also exhibits outstanding stability, and is recyclable by self-cleaning under UV irradiation. The excellent SERS performance is ascribed to the abundant surface electronic states induced by the hydrogenated amorphous layers and trivalent titanium, and more photogenerated charge carriers may contribute to the chemical and electromagnetic enhancement. The relative standard deviation (RSD) value for Raman signals less than 0.15 confirms the reproducibility of the hydrogenated composite SERS substrate. Therefore, the hybrid SERS substrates have great advantages in self-cleaning, reproducibility and stability, making them promising candidates for sensitive and recyclable SERS detection.

---

1. Introduction

Surface-enhanced Raman scattering (SERS) has been extensively explored in analysis, chemistry, and biology as a fast and ultrasensitive technique for nondestructive acquisition of structural information from chemical and biological molecules.^1–7 Noble metals (mainly Au and Ag) are well-known SERS-active substrates and enhance the Raman scattering resulting from the localized surface plasma resonance significantly. However, metal substrates have disadvantages, such as high cost, low stability, and lack of reusability, that have greatly limited their applications for SERS sensors. Moreover, most of the synthetic techniques, such as vacuum evaporation, nanosphere lithography, and electron beam lithography, that are used to fabricate metal nanostructures with sufficiently high SERS activity to meet practical demands are technologically challenging, expensive and produce disposable sensors.⁸

Recent studies on semiconductor-noble metal nanocomposites showed that they are effective for creating highly stable, recyclable, cheap and easy to prepare substrates. In particular, semiconductors such as TiO₂ and ZnO have received intense interest because they are photocatalytically self-cleaning, which makes the SERS substrate recyclable.^9–13 In addition, it is generally accepted that semiconductors themselves also contribute to the SERS activity by the charge-transfer (CT) mechanism, namely semiconductor-induced chemical enhancement related to the surface state energy level.^12,14–17 However, the chemical enhancement of semiconductor-based SERS substrates is still considerably weaker than that of metals, and there are inefficient hot spots ascribed to the large gaps in the noble metal nanomaterial. Recently, to overcome these problems, more attention has been paid to the modification and modulation of the morphology, structure, electronic state, and pattern of semiconductors.^2,10,18,19

In this work, we present SERS-active substrates with TiO₂ nanowires (NWs)/Ag nanoparticle (NP) composites on a Ti mesh scaffold. Hydrogenation was used to modulate the electronic states as well as the photoresponse of TiO₂. Although a wide variety of techniques are available to achieve hydrogenated TiO₂,^20–22 the controllable solid-state reaction of NaBH₄ and crystalline TiO₂ is a facile general hydrogenation method for large-scale production.²³ Hydrogenated TiO₂ nanomaterials exhibit enhanced light absorption and significantly improved photocatalysis resulting from disordered surface layers,^20,24 and even trivalent Ti ions or oxygen vacancies in the TiO₂ lattice, which produce extra electronic states near the conduction band minimum (CBM) and valence band maximum (VBM).^21,25 These abundant and varied surface states of hydrogenated TiO₂ are favorable for TiO₂-to-molecule CT and TiO₂-to-Ag NP CT,¹⁹ as illustrated in Fig. 1. Electrons are excited from the valence band to surface states near the conduction band in TiO₂ by laser irradiation, and they then migrate to organic molecules or Ag NPs leading to increased contributions to chemical and electromagnetic enhancement. As a result, a larger increase in the Raman signal can be achieved on composites of hydrogenated TiO₂ NWs and Ag NPs. Moreover, because the standard electrode potential for the Ti⁴⁺/Ti³⁺ balanced half reaction is −0.092 V, which is much more negative than that for Ag₂O/Ag (+0.342 V), the Ag NPs would be chemically protected by Ti³⁺ and not be degraded, prolonging the substrate lifetime. Furthermore, SERS substrates are usually disposable and non-recyclable;^26–28 therefore, it is desirable to develop a reusable SERS substrate to save money and conserve resources. Here, we will present a hybrid SERS substrate for SERS sensors in practical applications that exhibits great advantages in sensitivity, recyclability, and stability.

Fig. 1 Schematic diagrams of TiO₂-to-molecule and TiO₂-to-Ag NP transitions.

2. Experimental section

2.1. Materials and instruments

The 100 mesh titanium scaffold (250 μm thick) was purchased from Shanghai Youlan Electronic Technology Co., Ltd. Analytical grade anhydrous glucose, NaOH, AgNO₃, NaBH₄, R6G and titanium nanocrystalline powders were supplied by Aladdin Co., Ltd and were used as received without further purification. The as-prepared substrates were characterized with a FEI Sirion 200 field emission scanning electron microscope (FESEM) with an accelerating voltage of 10.0 kV. Transmission electron microscopy (TEM) and high-resolution TEM images were obtained on a JEOL JEM-2100F field emission source transmission electron microscope with an accelerating voltage of 200 kV. UV-vis diffuse reflectance spectra were recorded on a PE lambda 950 spectrometer. Photoluminescence emission (PL) spectra were measured on a Fluoromax-4 (Horiba Jobin Yvon) spectrofluorometer. A Bruker EPR ELEXSY 500 spectrometer was used to obtain the electron paramagnetic resonance (EPR) spectra. SERS spectra were collected on an inVia Raman microscope (Renishaw) with an excitation wavelength of 514 nm at a laser power of 1% (full power of 30 mW) and a thermal dispersive spectrometer equipped with a 532 nm laser at a laser power of 2 mW. The photocatalysis experiments were conducted by exposing the substrate to a 500 W high-pressure mercury lamp, and the distance from the substrate to the light source was about 10 cm.

2.2. Synthesis of TiO₂ NWs on a Ti mesh scaffold

Commercial Ti wire mesh is covered with a dense carbon coating, and must be washed alternately with dilute hydrofluoric acid and sodium hydroxide solution to acquire a clean, burnished Ti mesh. A single-crystalline TiO₂ scaffold on the Ti mesh scaffold was synthesized by a literature method.^29,30 Typically, a 2 × 2 cm piece of Ti mesh was placed on the bottom of a 50 mL Teflon-lined pressure vessel with aqueous 2.5 M NaOH (30 mL). After the vessel was heated at 150 °C for 15 h, sodium titanate nanowires spread over the Ti mesh skeleton. The Ti mesh covered with nanowires was immersed in aqueous 1 M HCl for 1 h to exchange Na⁺ with H⁺ and convert the sodium titanate nanowires to hydrogen titanate nanowires. Finally, the mesh was rinsed with deionized (DI) water and heated at 450 °C for 2 h to obtain TiO₂ nanowires on the Ti mesh scaffold.

2.3. Deposition of Ag NPs on TiO₂ NWs

The Ag NPs were deposited on the TiO₂ nanowires by the reduction of [Ag(NH₃)₂]⁺ with glucose. First, dilute ammonia solution was added dropwise to AgNO₃ solution until it turned transparent. The final concentration of Ag⁺ was adjusted to 0.1 M. Second, a piece of scaffold substrate was immersed in this solution for 30 min to reach the adsorption–desorption equilibrium of [Ag(NH₃)₂]⁺ ions. The substrate was dried under a flow of N₂, and then dipped in DI water for 30 s. The substrate was placed in glucose solution (30 mL) and heated in a water bath at 50 °C for 1 h to obtain TiO₂ NWs decorated with Ag NPs (TiO₂ NWs/Ag NPs) on a Ti mesh scaffold.

2.4. Hydrogenation of Ag NPs/TiO₂ NW scaffold

The substrate was cut into two pieces, and one of them was sealed in a quartz tube under vacuum with NaBH₄ (0.5 g) and was calcinated at 300 °C for 3 h. The as-obtained substrate was rinsed thoroughly with DI water three times to remove excess NaBH₄, and allowed to dry under ambient conditions. Two substrates (with or without hydrogenation) were compared to investigate the effect of hydrogenation on the SERS properties.

3. Results and discussion

SEM images were used to determine the morphology of TiO₂ NWs derived from Ti mesh grid with or without hydrogenation. Fig. 2a shows the Ti mesh grid treated by the alkaline hydrothermal method. Each titanium wire with a length of ∼100 μm is woven into a net and covered with a fuzzy layer of TiO₂ NWs. The tough irregular surface shown in Fig. 2b suggests a 3D stacked structure of nanowires, which is favorable for loading more Ag nanoparticles and improving sensitivity of the substrates in trace analysis. Fig. 2c reveals uniform nanowires with an average diameter of about 20 nm. Hydrogenation thickened the TiO₂ NWs from 30 to 40 nm in diameter, as shown in Fig. 2d.

Fig. 2 SEM images: (a) Ti mesh scaffold coated with TiO₂ NWs, (b) 3D TiO₂ NW nanostructures, (c) TiO₂ NWs, and (d) hydrogenated TiO₂ NWs.

To explore the morphology and structural evolution during hydrogenation treatment, high-resolution TEM (HRTEM) images were obtained (Fig. 3a and b). Before hydrogenation, the as-prepared TiO₂ NWs with a diameter of about 15 nm exhibited excellent crystallinity and a clean surface, as shown by the unambiguous fringes. The distance between the adjacent lattice planes was 0.376 nm, indicating a (101) facet of the anatase phase. However, the hydrogenation generated a unique structure of a crystalline core coated with an amorphous shell. The fringes were indistinct and a thick amorphous layer (∼4 nm) covered the TiO₂ NWs. The phase of the single crystal core was still anatase TiO₂, as shown by the spacing of the exposed crystallographic plane. The diameter of the nanowire increased to ∼27 nm, probably because of the deformation and destruction of the crystal lattice and the appearance of the amorphous layer.

Fig. 3 High-resolution TEM (HRTEM) images of (a) TiO₂ and (b) hydrogenated TiO₂ (H–TiO₂) NWs. (c) UV-vis-NIR absorption spectra of TiO₂ nanowires and H–TiO₂ nanowires. The inset shows the corresponding photographs. (d) Electron paramagnetic resonance (EPR) spectra of TiO₂ and H–TiO₂ NWs.

For convenience, to characterize the physical properties of the TiO₂ NWs, a large amount of TiO₂ NWs were prepared from titanium nanopowders by alkaline hydrothermal treatment, where the reaction conditions and subsequent processing procedure were exactly the same as that of Ti mesh grids. Fig. 3c presents the UV-vis-NIR absorption spectra of hydrogenated and non-hydrogenated TiO₂ NWs. The original TiO₂ NWs possess an absorption edge at 400 nm, corresponding to 3.1 eV, which is the typical bandgap absorption edge of nano anatase TiO₂. However, the hydrogenated TiO₂ (H–TiO₂) NWs have an absorption edge extended to ∼500 nm and significantly increased light absorption from 400 nm to the near infrared region. This is in good agreement with the color change of the nanowires from white to black, as indicated by the inset photographs in Fig. 3c. Highly sensitive electron paramagnetic resonance (EPR; detection limit up to 10⁻⁸ M) was used to examine unpaired electrons. Fig. 3d shows the EPR signals of TiO₂ NWs and the corresponding hydrogenated product. No resonance peaks in TiO₂ were observed, indicating no detectable Ti³⁺ species. After hydrogenation with NaBH₄, an intense EPR signal of H–TiO₂ at g = 1.998 appears. The g values are attributed to the paramagnetic oxygen vacancies and Ti³⁺ species originating from the reduction of Ti⁴⁺ by H atoms.²³ Many reports have shown that both disordered surface layers and trivalent titanium ions increase light absorption for the following reasons. (1) Hydrogen atoms in the amorphous layers can interact strongly with the Ti 3d and O 2p electrons, leading to a considerable contribution to the mid-gap electronic states between the valence band maximum (VBM) and conduction band minimum (CBM).^31,32 (2) The surface lattice disorder caused by hydrogenation induces a tail state band and causes an upward shift of the VBM to narrow the bandgap of TiO₂.^20,24,33 (3) Ti³⁺ and oxygen vacancies bring in CBM tails from Ti³⁺ 3d¹ and VBM tails from O 2p states.³⁴ These abundant surface states lead to a longer life time for light-excited e–h pairs and promote an effective electron transfer to adsorbed molecules or Ag nanoparticles, which can be verified by the photoluminescence (PL) measurements (Fig. S1†).

The morphology and structure of Ag-modified TiO₂ NWs scaffold were also investigated by SEM and TEM. The SEM image in Fig. 4a shows that Ag NPs coated the TiO₂ nanowires, similar to many reports of metal NP deposition on nanoarrays.^10,35,36 Discrete Ag nanoparticles with a low distribution density cannot create hot spots generated by closely adjacent nanoparticles. However, we focused on the effect on the SERS activity of the hydrogenation of substrate. Fig. 4b shows a TEM image of a single Ag-modified NW remover from the Ti mesh scaffold substrate by ultrasonic oscillation, and small nanoparticles are closely attached to the surface of TiO₂ NWs. Fig. 4c gives a high-resolution TEM (HRTEM) image of an Ag NP, which belonged to a (111) facet of cubic-phase Ag, based on the spacing of adjacent fringes.

Fig. 4 (a) SEM image of the Ag/TiO₂ NW composite. (b) TEM image of a single Ag-modified TiO₂ NW removed from the substrate (non-hydrogenated). (c) HRTEM image of one Ag nanoparticle attached to the nanowire.

The widely used efficient bio-indicator, R6G, was selected as a probe molecule to evaluate the SERS-activity of the as-prepared substrates.³⁷ To achieve sufficient coverage of a monolayer on the Ag and TiO₂ surface, the substrates were immersed in R6G ethanol solution of different concentrations for 30 min, and then washed and dried under ambient conditions. Fig. 5a presents the SERS spectra of Ag/TiO₂ NWs and Ag/H–TiO₂ NWs collected under the same test conditions. For the Ag/TiO₂ NW substrate, only several sporadic peaks located at 610, 1120, 1361 and 1501 cm⁻¹ were distinguished. Most of the characteristic vibration bands were drowned in the fluorescent background of R6G, which indicates only a small proportion of the adsorbed R6G molecules on the limited number of hot spots contributed to the SERS. However, the hydrogenated Ag/H–TiO₂ substrate displayed significantly enhanced Raman peaks for the R6G molecule with a better signal-to-noise ratio. The relative enhancement factor (EF) of Ag/H–TiO₂ to Ag/TiO₂ was calculated to be ∼10². In the literature, the absolute EF describes an important aspect of the SERS substrate performance for comparison with other SERS substrates. Here, the absolute EF was also evaluated to be ∼1 × 10⁸, and details of the calculation method based on Fig. S2 are presented in the ESI.† The SERS performance of the H–TiO₂/Ag composite substrate exceeded that of other recent results,³⁸ where hot spots generated by closely packed Ag NPs are predominant. Fig. 5b shows the SRES spectra of R6G adsorbed on Ag/H–TiO₂ NWs from ethanol solutions with various concentrations. The Raman peaks of R6G were still clearly visible, even when the analyte solution was diluted to 10⁻⁹ M. Meanwhile, Fig. S3† shows a dilute solution with a concentration as low as 10⁻⁷ M. None of the Raman peaks of R6G were observed on the Ag/TiO₂ NW substrate. Therefore, the hydrogenated TiO₂ NWs/Ag scaffold demonstrated greatly improved sensitivity and increased detection limits. The hot spots originating from single Ag nanoparticles discretely grown on TiO₂ NWs are inadequate for making the TiO₂/Ag scaffold a sensitive SERS substrate; however, hydrogenation is an effective method for activating the limited hot spots by adding more photogenerated electrons to the Fermi sea of Ag NPs, enabling powerful chemical enhancement by promoting more electron-to-molecule migration. The enhancement of the CT-dominated Ag/H–TiO₂ substrates is comparable or even superior to that of the reported Ag/TiO₂ substrates resulting from the nano-gaps between Ag NPs inducing hot spots in the literature.^39,40

Fig. 5 (a) SERS spectra of R6G collected on Ag/TiO₂ NWs and Ag/H–TiO₂ NWs. R6G concentration, 10⁻⁶ M; excitation, 514 nm; power, 1% of P_out, ∼0.3 mW; data collection, extended scan for 10 s. (b) SERS spectra collected on Ag/H–TiO₂ NWs exposed to different R6G concentrations. For 10⁻⁶ and 10⁻⁸ M: excitation, 514 nm; power, ∼0.3 mW; data collection, 10 s. For 10⁻⁹ M: excitation, 532 nm; power, 2 mW; data collection, 10 s.

As most metal-based SERS substrates are non-recyclable, searching for reusable substrates is important for reducing consumption of resources and costs for practical applications. Hydrogenated TiO₂ NWs, which have excellent photocatalytic properties, in our SERS substrate, endow our substrates with better recyclability. After the substrates were analyzed by SERS, they were immersed in deionized water and irradiated with a 500 W high-pressure mercury lamp for 10 min to accomplish self-cleaning. The cleaned substrate could be used again to detect dye molecules. Fig. 6 shows the results of the reversible SERS behavior of the Ag/H–TiO₂ composite NWs over three cycles. After UV light irradiation, R6G was detected from the Ag/H–TiO₂ substrate. Because the degradation products are small molecules, including CO₂ and H₂O, which can be removed from the substrate by washing, the SERS substrate was free of contamination. Fig. S4† indicates that although the substrate was reused at least 16 times to detect R6G molecules, no obvious signal loss of the Raman spectra was detected.

Fig. 6 Raman spectra of R6G recorded over three cycles indicating the reversible SERS behavior of the Ag/H–TiO₂ NW substrate.

The mechanism of the photocatalytic process is illustrated in Fig. S5.† Under UV irradiation, electron–hole pairs are generated in TiO₂ NWs by the transition of the valence electrons to the conduction band, and the photogenerated electrons migrate to Ag NPs through the Ag/H–TiO₂ interface.^41,42 When electrons are scavenged by molecular oxygen to produce reactive oxygen radicals and holes are captured by H₂O molecules to form hydroxyl radicals (HO˙), organic molecules will react with these oxidizing agents and be degraded to small inorganic molecules like CO₂ and H₂O. The large number of surface states induced by the amorphous layer and trivalent titanium in the Ag/H–TiO₂ substrate allow the separation and transfer of electron–hole pairs. Furthermore, the strong interaction between Ag nanoparticles and TiO₂ NWs (Fig. 4) would inhibit the recombination of the excited electrons and holes, and increase the photocatalytic performance of the substrate.

Reproducibility is another crucial factor for evaluating SERS performance. To confirm that the composite substrates of hydrogenated TiO₂ and Ag NPs show reproducible and stable SERS performance, two separate substrates (substrates 1 and 2) prepared via the same treatment were measured. Fig. 7a shows the Raman spectra of R6G molecules, randomly collected from three different locations on each substrate. The relative standard deviation (RSD) of the major Raman peaks is an effective parameter for evaluating the reproducibility of the SERS signals, and Fig. 7b shows the corresponding RSD spectrum derived from these Raman spectra in Fig. 7a.⁴³ All the calculations are based on the raw SERS spectra and no data preprocessing, including background removal, was observed. The maximum RSD value for both of the major SERS peaks and spectrum background was less than 0.15, indicating that the Ag/H–TiO₂ substrates also have good reproducibility and stability for SERS applications. In addition, we also compared the SERS activity of composite substrates prepared at least 6 months ago with that of the newly prepared substrate (Fig. S6†). The SERS spectra of 10⁻⁶ M R6G on the freshly prepared sample and on the sample stored for at least 6 months showed that no significant change in Raman intensity occurred in the SERS spectra obtained on the substrate stored for 6 months, which confirms the stability of our SERS substrates.

Fig. 7 (a) SERS spectra of R6G molecules randomly collected from different locations on two separate Ag/H–TiO₂ NW substrates, respectively. Substrates 1 and 2 were prepared by the same process, and three randomly selected positions for each substrate were measured with the Raman spectrometer. R6G concentration, 10⁻⁶ M; excitation, 514 nm; power, 1% of P_out, ∼0.3 mW; data collection, extended scan for 10 s. (b) The relative standard deviation (RSD) spectrum calculated from all the Raman spectra collected on substrates 1 and 2.

4. Conclusions

Wafer-scale TiO₂ nanowires on a Ti mesh scaffold deposited with dispersive spherical Ag nanoparticles were synthesized. Hydrogenation by a facile and versatile NaBH₄ method was used to obtain a hydrogenated TiO₂ NWs/Ag SERS substrate. The hydrogenated TiO₂ NWs/Ag scaffold suggests a new strategy to realize the preparation of a SERS substrate with greatly improved sensitivity. Hydrogenation is an effective method for boosting the light harvesting efficiency, including visible and near infrared light, by introducing amorphous shells and trivalent titanium, which generates abundant mid-gap electronic states between the VBM and CBM. More electrons are photo-excited and transferred to molecules or Ag NPs to promote both chemical and electromagnetic enhancement. The as-prepared Ag/H–TiO₂ NW substrates deliver a relative enhancement factor of up to ∼10² compared with the non-hydrogenated substrate, and the detection limit was improved by at least three orders of magnitude. Moreover, our experiments have confirmed that Ag/H–TiO₂ composites also have self-cleaning properties and are stable and give reproducible results. Therefore our hybrid SERS substrates have superior sensitivity, recyclability, reproducibility and stability as SERS sensors for practical applications.

Acknowledgements

This study was supported by a fund from the National Natural Science Foundation of China (NSFC, Contract no. 51471182).

Notes and references

1. D. Qi, L. Lu, L. Wang and J. Zhang, J. Am. Chem. Soc., 2014, 136, 9886–9889.
2. J. Ni, R. J. Lipert, G. B. Dawson and M. D. Porter, Anal. Chem., 1999, 71, 4903–4908.
3. X. Qian, J. Li and S. Nie, J. Am. Chem. Soc., 2009, 131, 7540–7541.
4. W.-H. Hsiao, H.-Y. Chen, Y.-C. Yang, Y.-L. Chen, C.-Y. Lee and H.-T. Chiu, ACS Appl. Mater. Interfaces, 2011, 3, 3280–3284.
5. Y. Yang, Z.-Y. Li, K. Yamaguchi, M. Tanemura, Z. Huang, D. Jiang, Y. Chen, F. Zhou and M. Nogami, Nanoscale, 2012, 4, 2663–2669.
6. Y. Yang, M. Tanemura, Z. Huang, D. Jiang, Z.-Y. Li, Y.-P. Huang, G. Kawamura, K. Yamaguchi and M. Nogami, Nanotechnology, 2010, 21, 325701.
7. Y. Yang, J. Shi, T. Tanaka and M. Nogami, Langmuir, 2007, 23, 12042–12047.
8. Z. Dai, G. Wang, X. Xiao, W. Wu, W. Li, J. Ying, J. Zheng, F. Mei, L. Fu and J. Wang, J. Phys. Chem. C, 2014, 118(39), 22711–22718.
9. X. Li, H. Hu, D. Li, Z. Shen, Q. Xiong, S. Li and H. J. Fan, ACS Appl. Mater. Interfaces, 2012, 4, 2180–2185.
10. E.-Z. Tan, P.-G. Yin, T.-T. You, H. Wang and L. Guo, ACS Appl. Mater. Interfaces, 2012, 4, 3432–3437.
11. X. Jiang, X. Li, X. Jia, G. Li, X. Wang, G. Wang, Z. Li, L. Yang and B. Zhao, J. Phys. Chem. C, 2012, 116, 14650–14655.
12. G. Sinha, L. E. Depero and I. Alessandri, ACS Appl. Mater. Interfaces, 2011, 3, 2557–2563.
13. Z. Mao, W. Song, L. Chen, W. Ji, X. Xue, W. Ruan, Z. Li, H. Mao, S. Ma and J. R. Lombardi, J. Phys. Chem. C, 2011, 115, 18378–18383.
14. X. Wang, W. Shi, G. She and L. Mu, Phys. Chem. Chem. Phys., 2012, 14, 5891–5901.
15. L. Yang, X. Jiang, W. Ruan, B. Zhao, W. Xu and J. R. Lombardi, J. Phys. Chem. C, 2008, 112, 20095–20098.
16. A. Musumeci, D. Gosztola, T. Schiller, N. M. Dimitrijevic, V. Mujica, D. Martin and T. Rajh, J. Am. Chem. Soc., 2009, 131, 6040–6041.
17. L. Yang, X. Jiang, W. Ruan, J. Yang, B. Zhao, W. Xu and J. R. Lombardi, J. Phys. Chem. C, 2009, 113, 16226–16231.
18. X. Hu, G. Meng, Q. Huang, W. Xu, F. Han, K. Sun, Q. Xu and Z. Wang, Nanotechnology, 2012, 23, 385705.
19. L. Yang, Y. Zhang, W. Ruan, B. Zhao, W. Xu and J. R. Lombardi, J. Raman Spectrosc., 2010, 41, 721–726.
20. X. Chen, L. Liu, Y. Y. Peter and S. S. Mao, Science, 2011, 331, 746–750.
21. A. Naldoni, M. Allieta, S. Santangelo, M. Marelli, F. Fabbri, S. Cappelli, C. L. Bianchi, R. Psaro and V. Dal Santo, J. Am. Chem. Soc., 2012, 134, 7600–7603.
22. X. Jiang, Y. Zhang, J. Jiang, Y. Rong, Y. Wang, Y. Wu and C. Pan, J. Phys. Chem. C, 2012, 116, 22619–22624.
23. H. Tan, Z. Zhao, M. Niu, C. Mao, D. Cao, D. Cheng, P. Feng and Z. Sun, Nanoscale, 2014, 6, 10216–10223.
24. X. Chen, L. Liu, Z. Liu, M. A. Marcus, W.-C. Wang, N. A. Oyler, M. E. Grass, B. Mao, P.-A. Glans and Y. Y. Peter, Sci. Rep., 2013, 3, 1510.
25. W. Wang, Y. R. Ni, C. H. Lu and Z. Z. Xu, RSC Adv., 2012, 2, 8286–8288.
26. W. Smith, Chem. Soc. Rev., 2008, 37, 955–964.
27. I. Yoon, T. Kang, W. Choi, J. Kim, Y. Yoo, S.-W. Joo, Q.-H. Park, H. Ihee and B. Kim, J. Am. Chem. Soc., 2008, 131, 758–762.
28. H. Chen, Y. Wang and S. Dong, J. Raman Spectrosc., 2009, 40, 1188–1193.
29. X. Peng and A. Chen, Adv. Funct. Mater., 2006, 16, 1355–1362.
30. J. M. Baik, M. H. Kim, C. Larson, X. Chen, S. Guo, A. M. Wodtke and M. Moskovits, Appl. Phys. Lett., 2008, 92, 242111.
31. Z. Wang, C. Yang, T. Lin, H. Yin, P. Chen, D. Wan, F. Xu, F. Huang, J. Lin and X. Xie, Adv. Funct. Mater., 2013, 23, 5444–5450.
32. J. Lu, Y. Dai, H. Jin and B. Huang, Phys. Chem. Chem. Phys., 2011, 13, 18063–18068.
33. L. Liu, Y. Y. Peter, X. Chen, S. S. Mao and D. Shen, Phys. Rev. Lett., 2013, 111, 065505.
34. Y. Zhao, T. Hou, Y. Li, K. S. Chan and S.-T. Lee, Appl. Phys. Lett., 2013, 102, 171902.
35. E. Galopin, J. Barbillat, Y. Coffinier, S. Szunerits, G. Patriarche and R. Boukherroub, ACS Appl. Mater. Interfaces, 2009, 1, 1396–1403.
36. S. Deng, H. Fan, X. Zhang, K. P. Loh, C. Cheng, C. Sow and Y. Foo, Nanotechnology, 2009, 20, 175705.
37. Y. Peng, L. Qiu, C. Pan, C. Wang, S. Shang and F. Yan, Electrochim. Acta, 2012, 75, 399–405.
38. H. Fang, C. X. Zhang, L. Liu, Y. M. Zhao and H. J. Xu, Biosens. Bioelectron., 2015, 64, 434–441.
39. L. He, D. Riassetto, P. Bouvier, L. Rapenne, O. Chaix-Pluchery, V. Stambouli and M. Langlet, J. Photochem. Photobiol., A, 2014, 277, 1–11.
40. Y. Zhao, L. Sun, M. Xi, Q. Feng, C. Jiang and H. Fong, ACS Appl. Mater. Interfaces, 2014, 6, 5759–5767.
41. Y. Zhou, J. Chen, L. Zhang and L. Yang, Eur. J. Inorg. Chem., 2012, 3176–3182.
42. Y. Zheng, L. Zheng, Y. Zhan, X. Lin, Q. Zheng and K. Wei, Inorg. Chem., 2007, 46, 6980–6986.
43. Z. Y. Bao, D. Y. Lei, R. Jiang, X. Liu, J. Dai, J. Wang, H. L. Chan and Y. H. Tsang, Nanoscale, 2014, 6, 9063–9070.

---

Footnote

† Electronic supplementary information (ESI) available: Photoluminescent spectra, SERS spectra collected on Ag/TiO₂ NWs, schematic diagram of the photocatalytic mechanism, and EF calculation method. See DOI: 10.1039/c5ra04352b

---