English

• Titanium oxide (TiO₂), known by its high stability and electronic properties, constitutes a versatile platform in the fields of photocatalysis, energy storage, and biomedical applications [1-3]. Operando monitoring of the reactions on a TiO₂ surface is extremely important for investigating the yields of TiO₂-based photocatalysts. The full understanding of the reactions at the water/TiO₂ interface is crucial for further optimization of TiO₂ based photocatalysts or reactors [4]. However, in-situ monitoring of the reactions on the TiO₂ surface, especially on-site reactions, under operando conditions is still a challenge. In that regard, surface-enhanced Raman spectroscopy (SERS) is an ideal analytical technology for ultrasensitive surface chemical analysis [5-8]. For example, SERS was successfully employed by Tian et al. to investigate the photoinduced reaction on the solid/gas interface of single-crystal TiO₂ [9]. As a matter of fact, most common TiO₂-based catalysts are generally low-cost polycrystalline composites and TiO₂ nanomaterials are usually applied in a liquid environment. Meanwhile, compared with strong signals from a solid/gas interface, the largely decayed Raman signals in the liquid medium mean that the use of semiconductor-based SERS substrates in practice is difficult [10]. Therefore, improving the SERS activity of semiconductor-based substrate in liquid medium has become an urgent task.

  For semiconductor-based SERS substrates, it is generally agreed that the tuning of photoinduced charge-transfer (PICT) through band energy alignment between substrate and molecule plays a dominant role [11]. Till now, lots of strategies have been developed to promote PICT efficiencies, such as n-/p- doping and stoichiometry [12-14]. Especially, injecting oxygen vacancies (V_O) as an important method in stoichiometry has been demonstrated to be useful for providing some additional defect levels in the bandgap of semiconductors, thus facilitating the exciton resonance in semiconductors, as well as the charge transfer (CT) between semiconductors and molecules [15-17]. Recently, the SERS activities of self-organized TiO₂ nanotubes have been discovered by Weidinger and coworkers [18]. Nevertheless, the Raman activity of TiO₂ nanotube arrays is significantly inferior to that of layered two-dimensional (2D) semiconductor materials, which have been demonstrated to possess a large number of structural defects [19-21]. Besides the structural defects induced V_O, Zhao and coworkers also utilized electrochromic technology to introduce V_O into the WO₃ substrate [22]. Inspired by the intrinsic electrochromic capacity, feasibility, and widespread applicability of TiO₂ nanopore films (TiO₂NPs) prepared by electrochemical anodization [23, 24], we hypothesis that V_O would be also inserted by an electrochromic way, thus enabling TiO₂NPs to be employed as alternative SERS substrates.

  For this purpose, we developed a facial and effective approach to prepare a large-scale porous semiconductor substrate with good electrochromic performance. The SERS measurements were carried out by employing bis(tetrabutylammonium) dihydrogen bis(isothi ocyanate) bis(2,2′-bipyridyl-4,4′-dicarboxylate) ruthenium(II) (N719) placed onto TiO₂/WO₃/TiO₂NPs. The generation of tunable V_O via electrochromic processing of the substrate resulted in the noticeable increase of the Raman signals of N719, and the PICT-mediated signal enhancement was attributed to the V_O induced by electrochromic effect.

  The multi-walled NPs were constructed on the self-ordered TiO₂NPs that were prepared by electrochemical anodization. The scanning electron microscopy (SEM) images reveal the uniform nanotube structure of the as-formed TiO₂NPs with the inner diameter of ~100 nm and the length of 300 nm (Fig. 1A and Fig. S1 in Supporting information). The W and Ti nanoparticles were subsequently applied onto the NPs via sputtering in sequence. This enabled one to increase the wall thickness and roughness of the NPs (Figs. 1B and C). The final NPs exhibited a smaller inner diameter. The uniform distributions of W and Ti elements across the tube surface and walls were confirmed via the morphological and elemental characterization (Fig. S2 in Supporting information), and the structural uniformity of NPs was crucial for obtaining the SERS substrates with good signal reproducibility. The annealing was carried out at 300 ℃ to achieve the high degree of crystallinity of the hybrid NPs. The X-ray diffraction (XRD) experiments were carried out to demonstrate the annealing-induced crystalline phase transition. After annealing, the peaks attributed to a monoclinic WO₃ phase with preferential (002) orientation (2θ = 23.0°) and anatase TiO₂ with (101) orientation (2θ = 25.3°) could be detected (Fig. S3 in Supporting information). The morphology of TiO₂/WO₃/TiO₂NPs was further characterized by transmission electron microscope (TEM) and high resolution (HR)-TEM in Figs. 1D-F and Fig. S4 (Supporting information). The crystal lattices of WO₃ (002) and TiO₂ (101) (004) can be identified at the wall of the resulted sample. These crystalline phases possess high conductivity and cyclic voltammetry stability [24]. The current response and electrochromic switching ability were investigated within a potential window between +0.8 V and −0.8 V (Fig. 1G). The amounts of inserted and extracted electrons during the voltage scanning were calculated based on the current-time curves. The TiO₂/WO₃/TiO₂NPs exhibited the charge density Q_cathodic of −95.9 mC/cm², which was much higher than those of the samples without WO₃ coating (−37.8 mC/cm² for TiO₂NPs and −68.5 mC/cm² for TiO₂/TiO₂NPs). Noticeably, the insertion of H⁺ protons from the electrochromic materials via the application of negative bias voltage was accompanied by a reflectance change (Fig. S5 in Supporting information), which could be attributed to the lower valence states of W- and Ti-ions [25]. Clearly, TiO₂/WO₃/TiO₂NPs underwent the more pronounced reflectance changes than the other two samples under applying the same bias voltage, and the electrochromic performance can be simply tuned by applying different voltages (Fig. S6 in Supporting information).

  Figure 1

  

  Figure 1.  SEM images of (A) TiO₂NPs and (B, C) TiO₂/WO₃/TiO₂NPs. (D) TEM, HAADF-STEM (inset), and (E, F) HR-TEM images of TiO₂/WO₃/TiO₂NPs. (G) Current density vs. time curves of TiO₂NPs (I), TiO₂/TiO₂NPs (II), and TiO₂/WO₃/TiO₂NPs (Ⅲ) acquired during potential pulse cycling in a voltage range from –0.8 V to +0.8 V. (H) SERS spectra of TiO₂/WO₃/TiO₂NPs after applying different bias voltage. (I) EPR spectra of TiO₂/WO₃/TiO₂NPs and V_O-TiO₂/WO₃/TiO₂NPs.

  DownLoad: Full-Size Img PowerPoint

  The SERS spectra of TiO₂/WO₃/TiO₂NPs exposed to various negative bias voltages (labeled as V_O-TiO₂/WO₃/TiO₂NPs) are plotted in Fig. 1H. It is worth mentioning that the Raman peaks of WO₃ broaden after the negative bias voltage (−0.5 V and −0.8 V) is applied, which can be explained by the generation of V_O in the substrate [26]. The introduction of V_O in TiO₂/WO₃/TiO₂NPs by applying with negative bias voltages was also verified by electron paramagnetic resonance (EPR) spectrum (Fig. 1I), XRD analysis (Fig. S7 in Supporting information), Electrochemical impedance spectroscopy (Fig. S8 in Supporting information), and in-situ generation of metal silver (Fig. S9 and Table S1 in Supporting information). It is important to note that plenty of V_O in the SERS substrates are conducive thus facilitating the PICT between the probe molecules and the substrates [15]. For semiconductors, the introduction of V_O can bring about a defect state that overlaps with the conduction band (CB), leading to the emergence of a band tail (Fig. S10 in Supporting information) [27]. It is expected that the CT between the substrate and the adsorber would thus be facilitated, and leading to a preferable SERS activity.

  As a proof-of-concept, the Raman spectra of N719 onto TiO₂/WO₃/TiO₂NPs were in-situ collected in HCl electrolyte by using a custom-built electrochemical cell (Fig. 2A). In this study, the NPs samples were exposed to a negative potential for 30 s to introduce V_O in the substrate, and the Raman signals were then acquired at the open circuit potential. Considering the energy match between the incident photons and the absorption spectra of N719 (Fig. S11 in Supporting information), a 532 nm laser was chosen as the excitation light source for the Raman measurements. The orientation of N719 on the TiO₂ surface was investigated using an angle X-ray photoelectron spectroscopy (XPS) at the take off angles θ of 10° and 70° (Fig. S12 in Supporting information). As shown in Fig. 2B and Table S2 (Supporting information), the O 1s (−COOH), N 1s, and S 2p signals exhibit the higher atomic concentrations at 70° (bulk models) than those at 10°. Meanwhile, the atomic population on the N 1s signals at 70° is obviously higher than those on the O 1s (−COOH) and S 2p signals. These results suggest that the −SCN groups are oriented far away from the Raman substrates, and N719 molecules have been anchored onto the TiO₂ surface in the form of a bridged configuration via the two carboxyl groups. Such bidentate coordination was considered to provide a favorable way for the CT between the substrate and adsorbed molecules [28]. Compared to the faint signals collected on the pristine samples (see the dot lines in Fig. 2C), the prominent signals of N719 molecules at 1474, 1545, and 1612 cm⁻¹, corresponding to a typical 2,2′-bipyridyl(bpy) ring stretching mode [29], were acquired on the V_O-containing samples (the solid lines in Fig. 2C). In Fig. 2C, V_O-TiO₂/WO₃/TiO₂NT exhibits a higher SERS activity than V_O-TiO₂/TiO₂NPs. This phenomenon can be ascribed to the WO₃ thin layer. Owing to the large optical modulation capacity, long-term durability, and memory effect of WO₃, the coating of WO₃ onto TiO₂ can efficiently improve the electrochromic efficiency (Fig. 2D and Fig. S13 in Supporting information). Furthermore, an excessively thick WO₃ film could result in a rough surface and the higher charge-transfer resistance (Figs. S14 and S15 in Supporting information) [30]. Therefore the W-sputtering thickness in this study was optimized at 5 nm (Figs. S16 and S17 in Supporting information).

  Figure 2

  

  Figure 2.  (A) Schematic of the electrochemical-SERS setup used in this study. (B) Atomic concentrations of N719 modified TiO₂/WO₃/TiO₂NPs at take-off angles of 10° and 70° (inset: orientation of N719 on the sample). (C) SERS spectra of N719 (3 × 10⁻³ mol/L) on pristine and V_O-contained TiO₂/WO₃/TiO₂NPs and TiO₂/TiO₂NPs. (D) Potential recovery of TiO₂/TiO₂NPs and TiO₂/WO₃/TiO₂NPs after applying the negative bias voltage of −0.8 V for 30 s.

  DownLoad: Full-Size Img PowerPoint

  The role of V_O on the SERS signal of the substrates was further investigated by repeatedly applying positive and negative voltages cycles. In Fig. 3A, an enhanced N719 signal can be observed when applying the voltage of −0.8 V. In turn, switching the voltage to +0.8 V causes the weak Raman signals. The repeatable SERS activities were achieved by switching the voltages back to −0.8 V, verifying the V_O can be electrochemically tuned with satisfactory feasibility and reproducibility. Meanwhile, these results also allow one to conclude whether the weakened SERS activity is derived from the vanishing of V_O rather than from the shedding or desorption of N719 molecules from the substrates.

  Figure 3

  

  Figure 3.  (A) SERS spectra of N719 adsorbed on TiO₂/WO₃/TiO₂NPs after exposed to the bias voltages of −0.8 V and +0.8 V alternately. (B) Raman intensity (at 1545 cm⁻¹) of N719 on TiO₂/WO₃/TiO₂NPs measured in air and in liquid. (C) SERS spectra of N719 collected from 20 randomly selected positions on V_O-TiO₂/WO₃/TiO₂NPs. (D) Intensity variation of the Raman peak at 1545 cm⁻¹ in these 20 selected positions.

  DownLoad: Full-Size Img PowerPoint

  To evaluate the application advantage of the proposed strategy in aqueous solutions, the Raman signals at 1545 cm⁻¹ (ν_(C=C)(bpy), the stretching of the C=C bond in bpy) were collected at different N719 concentrations on TiO₂/WO₃/TiO₂NPs in air and in aqueous solution (Fig. 3B). Apparently, the Raman signals recorded in aqueous solution are much weaker than those acquired in air. Impressively, a 39-fold increase in the Raman signals was achieved from the SERS substrates with embedded V_O under a negative bias voltage. Notably, the Raman signals recorded on such a V_O-rich substrate in aqueous solution even exceeded those obtained in air. Moreover, the Raman signal was still conspicuous in the solution when the N719 concentration decreased to 10⁻⁶ mol/L (Fig. S18 in Supporting information and Fig. 3B), indicating high sensitivity of the V_O-rich substrate. The enhancement factor (EF) of V_O-TiO₂/WO₃/TiO₂NPs was calculated to be 8.6 × 10⁴ (Fig. S19 in Supporting information). Additionally, Fig. 3C shows the Raman signals of N719 at twenty randomly selected regions on V_O-TiO₂/WO₃/TiO₂NPs. The relative standard deviation (RSD) of the Raman peak at 1545 cm⁻¹ is determined to be 6.65% (Fig. 3D), indicating excellent reproducibility of the V_O-rich substrates.

  For a comprehensive understanding of the mechanism of V_O-induced SERS activity, the Raman spectra of N719 were investigated under different voltages (Fig. 4A). Owing to the increase in the V_O content with increasing negative bias voltage, the Raman signal intensity from the substrates became substantially higher [26]. The band structural analysis for V_O-TiO₂/WO₃/TiO₂NPs and N719 specimens was also carried out at different voltages (Figs. S20 and S21 and Table S3 in Supporting information). Compared to pristine samples, V_O-TiO₂/WO₃/TiO₂NPs exhibited the narrower bandgaps. The Mott-Schottky analysis also demonstrates that the charge-carrier density (N_d) calculated for V_O-TiO₂/WO₃/TiO₂NPs (3.45 × 10¹⁹ cm⁻³) is larger than that of the pristine sample (1.84 × 10¹⁸ cm⁻³), which is beneficial for the acceleration of PICT in V_O-TiO₂/WO₃/TiO₂NPs. Meanwhile, the contribution of CT in V_O-TiO₂/WO₃/TiO₂NPs to the SERS signal of N719 was quantifiably determined based on the following equation [16]:

  (1)

  Figure 4

  

  Figure 4.  (A) SERS spectra of N719 on TiO₂/WO₃/TiO₂NPs before and after exposure to different bias voltages. (B) Schematic energy level diagram and CT pathways in the N719-V_O-TiO₂/WO₃/TiO₂NPs system.

  DownLoad: Full-Size Img PowerPoint

  According to the SERS spectra in Fig. S22 (Supporting information), the degree of CT (ρ_CT) of N719 was calculated as 0.41 and 0.69 for TiO₂/WO₃/TiO₂NPs and V_O-TiO₂/WO₃/TiO₂NPs, respectively (details were provided in supporting Information). The key role of CT in our system was further demonstrated via attaching other Raman probes onto TiO₂/WO₃/TiO₂NPs by chemical bonding (4-MBA) or physical absorption (Ru(bpy)₃²⁺ and 2,2′-Bipyridine) (Fig. S23 in Supporting information). Compared to 4-MBA, the SERS signals of Ru(bpy)₃²⁺ and 2,2′-Bipyridine were poor and didnot exhibit obvious enhancement on V_O-TiO₂/WO₃/TiO₂NPs. Moreover, the signal of N719 was remarkably dropped when a thin layer of SiO₂ film was covered onto TiO₂/WO₃/TiO₂NPs before N719 modification to block CT between substrate and molecules (Figs. S24 and S25 in Supporting information). All these results indicate that CT plays an important role in the improved SERS signals after applying the negative bias.

  Considering the location of the highest occupied molecular orbital (HOMO, −5.34 eV) and the lowest unoccupied molecular orbital (LUMO, −3.01 eV) of N719 molecules [31], the PICT could take place in following pathways (Fig. 4B): (1) PICT from the HOMO of N719 to the CB of TiO₂ and WO₃. The electrons can be directly excited from the HOMO of N719 to CB of TiO₂ (−4.21 eV) and WO₃ (−5.24 eV) [32] by an incident light of 532 nm (2.33 eV). (2) PICT from the HOMO of N719 to the defect levels (TiO_2-x and WO_3-x). In general, the defect levels induced by V_O are usually located at ~0.5−1.0 eV below the minimum value of CB [33]. As a result, PICT from HOMO of N719 to defect levels (TiO_2-x and WO_3-x) can provide more available PICT pathways than the sample that only contains TiO₂ and WO₃, which can further lead to a magnification of the Raman scattering cross section, thus greatly magnifying the polarization tensor of N719 molecules. (3) Electrons in the LUMO of N719 transfer from the photo-excited N719 molecules to CB of TiO₂. In addition, because of the existence of the defect levels (WO_3-x), the band gap of WO₃ can be narrowed as 1.7−2.2 eV [33]. Therefore, electrons can be excited from VB to WO_3-x (μ_ex) by a 532-nm laser and the photogenerated holes in WO_3-x subsequently transfer to VB of TiO₂. Meanwhile, excited electrons in CB of TiO₂ transfer to CB of WO₃. Therefore, a built-in electric field at the interface of TiO₂ and WO₃ with the direction of the electric field pointing from WO₃ to TiO₂ is formed. Driven by the built-in electric field, photogenerated electron-hole pairs can be effectively separated, realizing spatial charge separation and prolongating the lifetime of charge carriers, and thus enhancing the SERS activity [17, 34-37].

  To summarize, the electrochromic properties of the semiconductors were utilized to successfully produce the highly active SERS substrates that could be applied in aqueous electrolytes. Experiment data and theoretical calculation revealed that the abundant V_O induced by the electrochromic process facilitates the CT between the substrate and adsorbed molecules, thus enhancing their SERS activity. Especially, the as-proposed substrates were largely scalable, and their Raman signals were reproduced by controlling the applied bias voltage, thus providing an easily assessable and low-cost platform for in-situ monitoring of the reactions at the solid/liquid interfaces.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was supported by National Natural Science Foundation of China (Nos. 21874013, 22074013 and 22073030), the Fundamental Research Funds for the Central Universities (Nos. N2105018 and N2005027), and the China Postdoctoral Science Foundation (No. 2019M661109). The CPU time was supported by the Supercomputer Centre of East China Normal University (ECNU Public Platform for Innovation No. 001). Special thanks are due to the instrumental or data analysis from Analytical and Testing Center, Northeastern University.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.03.011.

  
  
• 1. [1]

     A. Fujishima, K. Honda, Nature 238 (1972) 37–38. doi: 10.1038/238037a0

  2. [2]

     K. Lee, A. Mazare, P. Schmuki, Chem. Rev. 114 (2014) 9385–9454. doi: 10.1021/cr500061m

  3. [3]

     C. Zhao, L. Zhou, Z. Zhang, et al., J. Phys. Chem. Lett. 11 (2020) 9931–9937. doi: 10.1021/acs.jpclett.0c02927

  4. [4]

     X.X. Han, L. Chen, U. Kuhlmann, et al., Angew. Chem. Int. Ed. 126 (2014) 2514–2517. doi: 10.1002/ange.201310123

  5. [5]

     F. Kruse, A.D. Nguyen, J. Dragelj, et al., J. Am. Chem. Soc. 143 (2021) 2769–2776. doi: 10.1021/jacs.0c10767

  6. [6]

     L. Liu, X. Du, Chin. Chem. Lett. 32 (2021) 1942–1946. doi: 10.1016/j.cclet.2020.12.061

  7. [7]

     D. Buhrke, P. Hildebrandt, Chem. Rev. 7 (2020) 3577–3630.

  8. [8]

     A.U.R. Bacha, I. Nabi, Z. Fu, et al., Chin. Chem. Lett. 30 (2019) 2225–2230. doi: 10.1016/j.cclet.2019.07.058

  9. [9]

     S.P. Zhang, J.S. Lin, R.K. Lin, et al., Chem. Sci. 11 (2020) 6431–6435. doi: 10.1039/D0SC00588F

  10. [10]

      L. Yang, C. Zhao, J. Xu, et al., Chem. Commun. 56 (2020) 10333–10336. doi: 10.1039/D0CC04396F

  11. [11]

      S. Schlücker, Angew. Chem. Int. Ed. 53 (2014) 4756–4795. doi: 10.1002/anie.201205748

  12. [12]

      X. Fan, M. Li, Q. Hao, et al., Adv. Mater. Interfaces 6 (2019) 1901133. doi: 10.1002/admi.201901133

  13. [13]

      L. Lan, Y. Gao, X. Fan, et al., Front. Phys. 16 (2021) 43300. doi: 10.1007/s11467-021-1047-z

  14. [14]

      I. Alessandri, J.R. Lombardi, Chem. Rev. 116 (2016) 14921–14981. doi: 10.1021/acs.chemrev.6b00365

  15. [15]

      X. Wang, W. Shi, S. Wang, et al., J. Am. Chem. Soc. 141 (2019) 5856–5862. doi: 10.1021/jacs.9b00029

  16. [16]

      G. Song, W. Cong, S. Gong, Z. Zhao, Angew. Chem. Int. Ed. 60 (2021) 5505–5511. doi: 10.1002/anie.202015306

  17. [17]

      J. Seo, J. Lee, Y. Kim, et al., Nano Lett. 20 (2020) 1620–1630. doi: 10.1021/acs.nanolett.9b04645

  18. [18]

      I.H. Oer, C.J. Querebillo, C. David, et al., Angew. Chem. Int. Ed. 57 (2018) 7225–7229. doi: 10.1002/anie.201802597

  19. [19]

      Y. Yang, W. Wei, P. He, et al., Chin. Chem. Lett. 33 (2022) 2600–2604. doi: 10.1016/j.cclet.2021.09.098

  20. [20]

      Z. Yu, L. Huang, Z. Zhang, G. Li, Chin. Chem. Lett. 33 (2022) 3853–3858. doi: 10.1016/j.cclet.2021.10.045

  21. [21]

      C. Hess, Chem. Soc. Rev. 50 (2021) 3519–3564. doi: 10.1039/D0CS01059F

  22. [22]

      S. Cong, Z. Wang, W. Gong, et al., Nat. Commun. 10 (2019) 678. doi: 10.1038/s41467-019-08656-6

  23. [23]

      Y.C. Nah, A. Ghicov, D. Kim, S. Berger, P. Schmuki, J. Am. Chem. Soc. 130 (2008) 16154–16155. doi: 10.1021/ja807106y

  24. [24]

      Y.Y. Song, Z.D. Gao, J.H. Wang, X.H. Xia, R. Lynch, Adv. Funct. Mater. 21 (2011) 1941–1946. doi: 10.1002/adfm.201002258

  25. [25]

      K. Chang, X. Hai, J. Ye, Adv. Energy Mater. 6 (2016) 1502555. doi: 10.1002/aenm.201502555

  26. [26]

      G. Wang, Y. Ling, H. Wang, et al., Energy Environ. Sci. 5 (2012) 6180–6187. doi: 10.1039/c2ee03158b

  27. [27]

      X. Chen, L. Liu, P.Y. Yu, S.S. Mao, Science 331 (2011) 746–750. doi: 10.1126/science.1200448

  28. [28]

      T. Pan, G. Song, S. Cong, et al., J. Phys. Chem. C 124 (2020) 20530–20537. doi: 10.1021/acs.jpcc.0c04935

  29. [29]

      Z. Mao, Y. Ye, H. Lv, et al., Angew. Chem. Int. Ed. 59 (2020) 10780–10784. doi: 10.1002/anie.201915824

  30. [30]

      M. Altomare, N.T. Nguyen, S. Hejazi, P. Schmuki, Adv. Funct. Mater. 28 (2018) 1704259. doi: 10.1002/adfm.201704259

  31. [31]

      X. Yang, Y. Chen, P. Liu, et al., Adv. Funct. Mater. 30 (2020) 2001557. doi: 10.1002/adfm.202001557

  32. [32]

      H. Zhu, Q. Yang, D. Liu, et al., J. Am. Chem. Soc. 143 (2021) 9236–9243. doi: 10.1021/jacs.1c04631

  33. [33]

      Z. Zheng, S. Cong, W. Gong, et al., Nat. Commun. 8 (2017) 1993. doi: 10.1038/s41467-017-02166-z

  34. [34]

      M. Li, X. Fan, Y. Gao, T. Qiu, J. Phys. Chem. Lett. 10 (2019) 4038–4044. doi: 10.1021/acs.jpclett.9b00972

  35. [35]

      M. Li, Y. Wei, X. Fan, et al., Nano Res. 15 (2022) 637–643. doi: 10.1007/s12274-021-3537-2

  36. [36]

      X. Fan, P. Wei, G. Li, et al., ACS Appl. Mater. Interfaces 13 (2021) 51618–51627. doi: 10.1021/acsami.1c11977

  37. [37]

      J. Seo, J. Lee, Y. Kim, et al., Nano Lett. 20 (2020) 1620–1630. doi: 10.1021/acs.nanolett.9b04645

• 

  Figure 1  SEM images of (A) TiO₂NPs and (B, C) TiO₂/WO₃/TiO₂NPs. (D) TEM, HAADF-STEM (inset), and (E, F) HR-TEM images of TiO₂/WO₃/TiO₂NPs. (G) Current density vs. time curves of TiO₂NPs (I), TiO₂/TiO₂NPs (II), and TiO₂/WO₃/TiO₂NPs (Ⅲ) acquired during potential pulse cycling in a voltage range from –0.8 V to +0.8 V. (H) SERS spectra of TiO₂/WO₃/TiO₂NPs after applying different bias voltage. (I) EPR spectra of TiO₂/WO₃/TiO₂NPs and V_O-TiO₂/WO₃/TiO₂NPs.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (A) Schematic of the electrochemical-SERS setup used in this study. (B) Atomic concentrations of N719 modified TiO₂/WO₃/TiO₂NPs at take-off angles of 10° and 70° (inset: orientation of N719 on the sample). (C) SERS spectra of N719 (3 × 10⁻³ mol/L) on pristine and V_O-contained TiO₂/WO₃/TiO₂NPs and TiO₂/TiO₂NPs. (D) Potential recovery of TiO₂/TiO₂NPs and TiO₂/WO₃/TiO₂NPs after applying the negative bias voltage of −0.8 V for 30 s.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (A) SERS spectra of N719 adsorbed on TiO₂/WO₃/TiO₂NPs after exposed to the bias voltages of −0.8 V and +0.8 V alternately. (B) Raman intensity (at 1545 cm⁻¹) of N719 on TiO₂/WO₃/TiO₂NPs measured in air and in liquid. (C) SERS spectra of N719 collected from 20 randomly selected positions on V_O-TiO₂/WO₃/TiO₂NPs. (D) Intensity variation of the Raman peak at 1545 cm⁻¹ in these 20 selected positions.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (A) SERS spectra of N719 on TiO₂/WO₃/TiO₂NPs before and after exposure to different bias voltages. (B) Schematic energy level diagram and CT pathways in the N719-V_O-TiO₂/WO₃/TiO₂NPs system.

  下载: 全尺寸图片 幻灯片
•