Enhanced photoelectric performance of CdS/CdSe co-sensitized TiO₂ nanosheets array films

Abstract

In this study, vertically aligned titanium dioxide nanosheets (TiO₂NSs) array films were employed for the fabrication of photoanodes for quantum-dot sensitized solar cells (QDSSCs). The TiO₂NSs array films were synthesized by a hydrothermal method. Then, cadmium sulfide (CdS) and cadmium selenide (CdSe) were assembled onto the TiO₂NSs array films through chemical bath deposition (CBD) and successive ionic layer adsorption and reaction (SILAR), respectively. Detailed characterization measurements were carried out to analyse the crystallinity, microstructure, composition and optical properties of the samples. The CdS and CdSe QDs were densely and uniformly deposited on both {001} and {101} facets of the TiO₂NSs array films. Moreover, the effect of CdS deposition time and number of CdSe deposition cycles (C) on the photoelectric performance was also investigated in detail. The best performance of 4.77% was achieved under AM 1.5 G illumination with J_sc = 18.22 mA cm⁻², V_oc = 0.51 V, and FF = 0.52 when the deposition time of CdS was 30 min and the deposition cycle of CdSe was 10C. The enhanced photoelectric performance mechanism of the co-sensitized device is also discussed in this paper.

---

1. Introduction

Over the past two decades, research on QDSSCs has grown significantly, and the technology has been considered one of the most cost-effective alternative to silicon and other semiconductor-based photovoltaic devices because of their high conversion efficiency, low-cost and simple fabrication process.^1–3 Nano-dimensional TiO₂ is one of the most attractive materials for this application and is widely used as a photoanode because of its appropriate energy band position, long term stability, non-toxicity and cost effectiveness.^4–7 Since the TiO₂ thin film is employed as an electron collecting layer, its morphology plays a significant role in the improvement of the power conversion efficiency (PCE) of solar cells. Recently, TiO₂ materials with different morphologies such as nanoparticles,^8–10 nanorods,¹¹ nanotubes,^12,13 and nanowires¹⁴ have been applied to the fabrication of QDSSCs. However, the efficiency of solar cells based on zero-dimensional mesoporous nanoparticles and one-dimensional nanostructured array films is unsatisfactory due to the formation of grain boundaries and insufficient specific surface areas. Therefore, considerable efforts have been made in the development of two-dimensional TiO₂ nanostructures. Compared to other nanostructures, vertically aligned TiO₂NSs array films with a large percentage of exposed {001} reactive facets present several extraordinary advantages, which can be summarized as follows. First, both theoretical and experimental studies found that the highly reactive {001} facet-dominated anatase TiO₂ has a substantial effect on photo-generated charge separation and transfer.^15–18 Second, the nanosheet array film structure with sufficient surface area facilitates the penetration of electrolytes and ensures the formation of well-filled nanoparticles.¹⁹ Third, the large nanosheets act as light-scattering centers to enhance the light absorption efficiency.²⁰ Lastly, the vertically grown nanostructure arrays provide direct paths for the fast motion of photo-generated electrons.²¹

It is well known to all because of its large band gap (3.2 eV), TiO₂ can only harvest ultraviolet light, which corresponds to only 5% of the entire solar spectrum. To extend light absorption into the visible light region, many efforts have been made to the development of high-performance sensitizers.^22–25 The versatile properties of inorganic narrow band gap semiconductor quantum dots such as multiple exciton generation, energy band gap tunability, and high absorption coefficient make them attractive candidates for QDSSCs.^26–29 CdSe is the most widely used sensitizer owing to its appropriate band gap (1.70 eV).^30,31 However, the large lattice mismatch between CdSe and TiO₂ hinders the growth of CdSe on the TiO₂ films.^32,33 Inserting a CdS (2.42 eV)^34,35 layer between CdSe and TiO₂ is considered as an effective interface modification because of the better lattice match between CdS and TiO₂. It is generally believed that the CdS layer on one hand assists the growth of CdSe on the TiO₂ films and on the other hand, it facilitates separation and transfer of photo-generated carriers by the realignment of the energy levels of CdS and CdSe.^36–39 Therefore, the CdS layer promotes the injection of photo-generated electrons from CdSe to TiO₂ due to the formed stepwise energy level structure. Recently, many researchers have focused on CdS/CdSe co-sensitized one-dimensional TiO₂ nanostructure array films such as nanorods,⁴⁰ nanotubes,^13,41,42 and nanowires.⁴³ However, the insufficient internal surface area exhibited by the dense one-dimensional nanostructure arrays limits the PCE to a relatively low level due to deficient QDs loading and light absorption. To the best of our knowledge, there are only few studies focusing on the fabrication of photoelectrodes based on vertically aligned TiO₂NSs array films. Hence, the fabrication of CdS/CdSe co-sensitized TiO₂NSs array films by a simple chemical solution method is of remarkable interest and challenging.

In this study, vertically aligned TiO₂NSs array films were employed in TiO₂/CdS/CdSe QDSSCs. The TiO₂NSs array film substrate was prepared by a hydrothermal method, and CdS and CdSe QDs were synthesized by the CBD and SILAR methods, respectively. The detailed synthesis process and characterization are discussed. Furthermore, the photoelectric performance as a function of CBD time and number of SILAR cycles was carefully investigated. It is found that the optimum CBD time for CdS is 30 min and the optimum number of SILAR cycles is 10C. The photoelectric performance analysis results indicate that the best efficiency of 4.77% was achieved under AM 1.5 G illumination with J_sc = 18.22 mA cm⁻², V_oc = 0.51 V, and FF = 0.52. In addition, it is also confirmed that the tuned band alignment in the TiO₂NSs/CdS/CdSe cascade structure has a substantial effect on the enhanced photoelectric performance.

2. Experimental section

First, the TiO₂NSs array films on the FTO substrates were synthesized by a hydrothermal growth method (Experimental section, ESI†). Then, CdS and CdSe were assembled onto the TiO₂NSs array films by the CBD and SILAR methods, respectively. More experimental details are provided in the ESI.†

3. Results and discussion

3.1 Sample characterization

3.1.1 Phase composition and phase structure. Fig. 1 shows the XRD patterns of (a) bare TiO₂NSs, (b) TiO₂NSs/CdS (30 min), (c) TiO₂NSs/CdSe (10C), and (d) TiO₂NSs/CdS (30 min)/CdSe (10C) samples. As shown in Fig. 1a, all the diffraction peaks are indexed to the standard anatase phase of TiO₂ (JPCDS no. 21-1272; space group: I4₁/amd). The significantly enhanced (101) diffraction peak at 25.34° indicates that the as-deposited films are highly oriented with respect to the FTO substrate. The XRD pattern of TiO₂NSs/CdS (30 min) exhibits new diffraction peaks at 43.97° and 52.08°, corresponding to the (220) and (311) planes of cubic phase CdS, which is consistent with the high-resolution transmission electron microscopy (HRTEM) images shown in Fig. 3d. In addition, the significantly enhanced diffraction peak at 26.51° also corresponds to the (111) plane of CdS. No new evident CdSe diffraction peaks are observed for TiO₂NSs/CdSe (10C) due to the minor amount of CdSe QDs and their high dispersion on the TiO₂ surface as shown in Fig. 2d. The TiO₂NSs/CdS (30 min)/CdSe (10C) sample exhibits new small diffraction peaks at 41.99°, which can be indexed to the (220) plane of cubic phase CdSe. It is noteworthy that the XRD patterns show no impurity phases, which indicate the stability and the high purity of the samples.

Fig. 1 XRD patterns of (a) bare TiO₂NSs, (b) TiO₂NSs/CdS (30 min), (c) TiO₂NSs/CdSe (10C) and (d) TiO₂NSs/CdS (30 min)/CdSe (10C) samples.

Fig. 2 Typical FESEM images: (a) and (b) are top view and cross-section view of bare TiO₂NSs array film, respectively; (c) and (d) are top view of TiO₂NSs/CdS (30 min) and TiO₂NSs/CdSe (10C), respectively; (e) and (f) are top view and cross-section view of TiO₂NSs/CdS (30 min)/CdSe (10C) sample, respectively.

3.1.2 FESEM and EDS characterization. The morphology and structure of the samples were investigated by field-emission scanning electron microscopy (FESEM). Typical FESEM images of the bare TiO₂NSs, TiO₂NSs/CdS (30 min), TiO₂NSs/CdSe (10C), and TiO₂NSs/CdS (30 min)/CdSe (10C) samples are shown in Fig. 2. Fig. 2a shows the top-section view of bare TiO₂NSs. It can be seen that the FTO substrate is covered with a dense and uniform nanosheets-array film. The magnified image in Fig. S1† shows that the nanosheets have a smooth surface and the thickness of a single nanosheet is about 260 nm. In addition, the FESEM image indicates that the well-faceted crystal structure of a single TiO₂ nanosheet possesses good geometrical symmetry; the two flat square surfaces are identified as {001} facets, while the other lateral surfaces are {101} facets.^15,44

The vertically aligned {001} facets and the non-overlapping structures of the TiO₂NSs are interlaced and interconnected. Fig. 1b shows the cross-section view of a TiO₂NSs array film, which shows the vertical growth structure, and the thickness of the array film is about 3 μm. Fig. 2c and d show the top-section view of TiO₂NSs/CdS (30 min) and TiO₂NSs/CdSe (10C), respectively. Both samples retain their original morphology, but become relatively rough and possess uniform nanoparticles, indicating that CdS and CdSe QDs have been successfully deposited on the surface of the TiO₂NSs array film. It is worth noting that the surface of the TiO₂NSs array film was entirely covered with CdS QDs, but CdSe QDs were only adsorbed on a small section of the TiO₂NSs surface, suggesting that the growth of CdSe on TiO₂NSs is more difficult than that of CdS. Fig. 2e and f show the top view and cross-section view of TiO₂NSs/CdS (30 min)/CdSe, respectively. A large number of CdS and CdSe QDs are deposited densely and uniformly on both the {001} and {101} facets of the TiO₂NSs array film. In contrast to the CdSe single-sensitized samples, more CdSe QDs can be observed. This behaviour may be attributed to the fact that better lattice match between TiO₂ and CdS assists the growth of CdSe, which is consistent with previous reports.³² Theoretical calculations have shown that the (110) plane of cubic CdS [(2 × 2) supercell parameters (11.65 × 8.24 Ų)] agrees best with the (101) plane of TiO₂ [(2 × 2) supercell parameters (10.87 × 7.73 Ų)] with a lattice mismatch smaller than 7%. However, the mismatch between the (110) plane of cubic CdSe [(2 × 2) supercell parameters (12.11 × 8.56 Ų)] and the (101) plane of TiO₂ is greater than 11%, suggesting that the growth of CdS on the TiO₂ film is easier, and the growth of CdSe is difficult. This observation is also confirmed by the XRD results obtained in this study.

The composition and elemental distribution of the composite array films were investigated by elemental energy-dispersive spectroscopy (EDS) and EDS mapping. The EDS mapping images corresponding to the FESEM images in Fig. 2c–e are shown in Fig. S3–S5,† respectively. A typical EDS spectrum (Fig. S5g†) shows that the TiO₂NSs/CdS (30 min)/CdSe (10C) sample is composed of Ti, O, Cd, S and Se elements. The EDS mapping images of Ti, O, Cd, S and Se elements are exhibited in Fig. S5b–f.† It can be seen that Ti, O, Cd, S and Se are homogenously distributed in the entire TiO₂NSs/CdS (30 min)/CdSe (10C) sample, suggesting that the CdS and CdSe QDs are grown homogeneously onto the TiO₂NSs array films. These results provide strong evidence for the successful coating of QDs on the surface of TiO₂NSs array films.

3.1.3 TEM and HRTEM characterization. The detailed microscopic characterization of the samples was carried out by TEM and HRTEM. Fig. 3a and S2† show the TEM and HRTEM images of bare TiO₂NSs array films. The HRTEM image displays that the lattice spacing is 0.352 nm, corresponding to the (101) plane of anatase TiO₂. Fig. 3b and c show the TEM images of the TiO₂NSs/CdS (30 min) and TiO₂NSs/CdS (30 min)/CdSe (10C) samples, respectively. As shown in the figures, compared with TiO₂NSs/CdS (30 min), the size of nanoparticles on TiO₂NSs/CdS (30 min)/CdSe (10C) was significantly enlarged. This proved once again the successful deposition of CdSe QDs on the surface of the TiO₂NSs/CdS array film. Fig. 3d shows the HRTEM image of the TiO₂NSs/CdS (30 min)/CdSe (10C) heterojunction; the distinct lattice fringes indicate that the sample is highly crystalline. The measured lattice spacings in Fig. 3d are consistent with the d-spacings of TiO₂, CdS and CdSe.

Fig. 3 TEM images of (a) bare TiO₂NSs, (b) TiO₂NSs/CdS (30 min), and (c) TiO₂NSs/CdS (30 min)/CdSe (10C). (d) HRTEM image of TiO₂NSs/CdS (30 min)/CdSe (10C).

Specifically, lattice fringes with interplanar spacing d (101) = 0.352 nm (Fig. 3d) is consistent with the anatase phase of TiO₂ [JCPDS no. 21-1272]. The spacings of d (220) = 0.206 nm and d (220) = 0.215 nm correspond to the cubic phase of CdS [JCPDS no. 75-1546] and to the cubic phase CdSe [JCPDS no. 19-191], respectively.

3.1.4 UV-vis diffuse reflectance absorption spectra. The light absorption capacity of the photoelectrodes was investigated by UV-vis diffuse reflectance absorption spectroscopy. It can be clearly seen that the photoresponse range is significantly extended and the light absorption intensity is enhanced upon CdS and CdSe QDs sensitization. As shown in Fig. 4a, the bare TiO₂NSs array film shows no significant absorbance of visible light due to its large energy gap (3.2 eV), which is also confirmed in Fig. 4c. The UV-vis spectrum of TiO₂NSs/CdS shows a broad absorption region between 330 and 540 nm. Moreover, it can be seen that the absorption intensity increases with the CdS deposition time, which indicates the simultaneous increase in CdS QDs. As shown in Fig. 4b, the TiO₂NSs/CdS/CdSe photoelectrodes have a broad photoresponse region from 330 nm to 740 nm. This result indicates that the deposition of CdSe QDs extended the light absorption region of the TiO₂NSs array film from the ultraviolet to nearly the entire visible light region. Furthermore, it can be seen that the absorption intensity increases with an increase in the number of CdSe deposition cycles, which can be attributed to the simultaneous increase in CdSe QDs. In addition, it can be seen that the TiO₂NSs/CdS (30 min)/CdSe (6–10C) samples exhibit much stronger light absorption than the TiO₂NSs/CdSe (10C) sample. These results indicate that the TiO₂NSs/CdS (30 min)/CdSe (6–10C) samples possess more CdSe QDs than the TiO₂NSs/CdSe (10C) sample. This could be because CdS as a seed layer is beneficial for the growth of CdSe on TiO₂NSs/CdS composite array films. Therefore, better coverage and larger number of CdSe QDs can be obtained on the surface of TiO₂NSs/CdS than on the bare TiO₂NSs. In addition, the photographs of the corresponding photoelectrodes are shown in Fig. 4e. The color of the films changes from white to yellow and finally to black-brown, suggesting the successful deposition of CdS and CdSe on the surface of the TiO₂NSs array film. The enhanced visible light absorption ability makes this type of QDs/TiO₂NSs composite array film a promising candidate for applications in photovoltaic devices.

Fig. 4 UV-vis diffuse reflectance absorption spectra of (a) bare TiO₂NSs array films and TiO₂NSs/CdS composite array films with different deposition times (10 min, 20 min, 30 min, 40 min) and (b) TiO₂NSs/CdS (30 min) array films and TiO₂NSs/CdS (30 min)/CdSe composite array films with different numbers of CdSe SILAR cycles (6C, 8C, 10C, 12C). (c) (αhv)²-Photon energy curves of bare TiO₂NSs array films and TiO₂NSs/CdS composite array films with different deposition times. (d) (αhv)²-Photon energy curves for TiO₂NSs/CdS (30 min)/CdSe composite array films with different numbers of CdSe SILAR cycles. (e) Photographs of the corresponding photoelectrodes.

It should be noted that slight red shifts were observed with an increase in CdS deposition time or number of CdSe deposition cycles, corresponding to the growth of the QDs. As known from the quantum confinement effect, when the particle size decreases to the quantum scale, the number of atoms in a particle will significantly decrease, causing a larger gap between energy levels. Hence, the energy levels will change from a quasi-continuous phase to a split phase, and the energy gap will become wider, leading to the blue shift of the absorption. Therefore, it is thought that the growth of QDs is responsible for the slight red shift of the absorption edge.

The optical band gap E_g was calculated according to eqn (1), where α and ν are the absorption coefficient and the frequency, respectively.

(αhν)² = c(hν − E_g) (1)

The E_g values of photoelectrodes can be obtained by extrapolating the linear segment of the plots of (αhν)²-photon energy at α = 0. As illustrated in Fig. 4c and d, the calculated E_g of the TiO₂NSs/CdS and TiO₂NSs/CdS/CdSe films gradually decrease with an increase in CdS deposition time or number of CdSe deposition cycles.

3.2 Photovoltaic performance of the photoelectrodes

The photocurrent–voltage (J–V) curves of these devices are presented in Fig. 5 and the corresponding photovoltaic parameters are summarized in Table S1.†Fig. 5a shows the J–V curves of the TiO₂NSs and TiO₂NSs/CdS electrodes. The results indicate that the PCE of the devices is enhanced by increasing the CdS deposition time, which is due to the improved absorbance of visible light. However, when the deposition time reaches 40 min, the excessive amount of deposited CdS reduces the photovoltaic performance. This phenomenon may be attributed to the fact that excess deposited CdS would act as recombination centers for photo-generated carriers. The optimal deposition time found for the CdS-sensitized TiO₂NSs photoelectrodes is about 30 min and the device exhibits an efficiency of 2.09%. The corresponding photovoltaic parameters are as follows: V_oc = 0.52 V, J_sc = 7.68 mA cm⁻², and FF = 0.52. We fabricated TiO₂NSs/CdS/CdSe based on the TiO₂NSs/CdS (30 min) photoelectrode and the corresponding J–V curves are shown in Fig. 4b. A similar phenomenon can be observed as previously discussed: the photocurrent density first increases and then decreases. When the number of SILAR cycles is 10C, the TiO₂NSs/CdS/CdSe photoelectrode exhibits the highest efficiency of 4.77%, and the corresponding photovoltaic parameters are V_oc = 0.51 V, J_sc = 18.22 mA cm⁻², and FF = 0.52. It is worth noting that the TiO₂NSs/CdSe (10C) sample shows much lower J_sc (6.42 mA cm⁻²) due to low coverage of CdSe on the TiO₂NSs array film. All the TiO₂NSs/CdS/CdSe photoelectrodes show higher efficiencies than the TiO₂NSs/CdS electrode. Several factors are responsible for the enhanced photoelectric performance. First, the vertically aligned TiO₂NSs array film retains both the advantage of sufficient surface area and photo-generated carrier transport path. Second, the significantly improved light absorption performance of the TiO₂/CdS/CdSe photoelectrode due to the dense and uniform coverage of QDs on the TiO₂NSs array film causes the increase in photocurrent density. Last, the energy level alignment at the TiO₂NSs/CdS/CdSe interfaces produces a stepwise band edge structure, which can effectively separate and transfer the photo-generated carriers and induce higher resistance to the transport of excitons back to the electrolyte as shown in Fig. 5c and d. We also observed that V_oc increased with CdS sensitization, but slightly decreased with the CdSe coating on TiO₂NSs/CdS. The trend can be qualitatively evaluated from the shift in the Fermi level (E_f) because of the contact between CdS and CdSe. The variation of the E_f after CdS and CdSe sensitization is shown in Fig. 5c and d. Compared with bare TiO₂NSs, the E_f shifts to more negative potentials upon sensitization with CdS for 30 min, which leads to an increase in V_oc from 0.37 V to 0.52 V. When CdSe was deposited on the TiO₂NSs/CdS (30 min) film, energy level realignment occurs in the TiO₂NSs/CdS (30 min)/CdSe composite film photoelectrodes⁴⁵ as shown Fig. 5d. However, the E_f of CdSe is less negative than that of CdS even after energy level realignment,^12,46 which leads to the shift in the E_f of the TiO₂NSs/CdS (30 min)/CdSe composite film to less negative values (from E_f1 to E_f2) and thus exhibits lower V_oc than that of the TiO₂NSs/CdS (30 min) composite film. In addition, the V_oc of TiO₂/CdSe with different layers is presented for better comparison. As shown in Fig. S6 and Table S1,† the V_oc of TiO₂/CdSe is lower than that of TiO₂NSs/CdS (30 min)/CdSe. This is because E_f of CdSe is less negative than that of the realigned energy levels of CdS/CdSe (E_f3). Because the E_f of CdS and CdSe are both more negative than that of bare TiO₂NSs, the V_oc of TiO₂NSs/CdS, TiO₂NSs/CdSe, and TiO₂NSs/CdS (30 min)/CdSe should also be larger than that of bare TiO₂NSs.

Fig. 5 . J–V curves of (a) TiO₂NSs/CdS photoelectrodes with different CdS deposition times (0 min, 10 min, 20 min, 30 min, 40 min); (b) TiO₂NSs/CdS (30 min), TiO₂NSs/CdSe (10C), and TiO₂NSs/CdS (30 min)/CdSe photoelectrodes with different numbers of CdSe SILAR cycles (6C, 8C, 10C, 12C). (c) Schematic representing the relative Fermi level (dashed lines) and band edge positions of TiO₂, CdS, and CdSe. (d) Schematic representing the proposed relative energy levels and charge transfer processes after E_f realignment in the TiO₂NSs/CdS/CdSe composite array film, E_f1 of TiO₂NSs/CdS (black dashed line), E_f2 of TiO₂NSs/CdS/CdSe (red dashed line), and E_f3 of CdS/CdSe (blue dashed line).

The monochromatic incident photon-to-electron conversion efficiency (IPCE) spectra (Fig. 6) clearly show that the photon-to-current responses of the TiO₂NSs/CdSe (10C) and TiO₂NSs/CdS (30 min)/CdSe (10C) devices exhibit significant red shifts compared with those of the TiO₂NSs/CdS (30 min) device. The integrated J_sc for the corresponding devices, namely, TiO₂NSs/CdS (30 min), TiO₂NSs/CdSe (10C), and TiO₂NSs/CdS (30 min)/CdSe (10C) are 7.34, 6.52 and 17.32 mA cm⁻², respectively, which is consistent with the experimentally obtained J_sc. These results indicate that TiO₂NSs/CdS (30 min)/CdSe (10C) photoelectrode not only has a broader light absorption range, but also has high IPCE value. These results also demonstrate that the stepwise band edge structure reduces charge recombination, resulting in more efficient photo-generated carrier separation and transfer. In addition, it is also demonstrated that the CdS QDs serve as a seed layer that promotes the growth of CdSe. The excellent IPCE value of the TiO₂NSs/CdS/CdSe device manifested its great potential for the construction of high-performance QDSSCs.

Fig. 6 IPCE spectra of (a) TiO₂NSs/CdS (30 min), (b) TiO₂NSs/CdSe (10C), and (c) TiO₂NSs/CdS (30 min)/CdSe (10C).

Long-term stability is very important in the performance of solar cells. To prove the long-term stability of the photoelectrode, the photostability of TiO₂NSs/CdS (30 min)/CdSe (10C) was measured for 1 h under continuous illumination. As shown in Fig. 7, a steady photocurrent is delivered by the device under 1 h-illumination with a drop in photocurrent of only 4% during this time. This result indicates that the TiO₂NSs/CdS (30 min)/CdSe (10C) photoelectrode exhibits excellent long-term stability.

Fig. 7 Photocurrent stability of TiO₂NSs/CdS (30 min)/CdSe (10C) under continuous illumination of 100 mW cm⁻².

4. Conclusion

In summary, a vertically aligned TiO₂NSs array film was applied in TiO₂/CdS/CdSe solar cells for the first time. We demonstrated that the amount of CdS and CdSe QDs has a great effect on the photoelectric performance of the devices. When the CdS deposition time was 30 min and the number of CdSe deposition cycles was 10C, the highest efficiency of 4.77% was achieved under AM 1.5 G illumination with V_oc = 0.51 V, J_sc = 18.22 mA cm⁻², and FF = 0.52. The improved photoelectric performance is attributed to the excellent light absorption property and the effective separation and transfer of photo-generated charge. Moreover, the excellent IPCE value of the devices further confirmed the superiority of the composite photoelectrode based on vertically aligned TiO₂NSs array films. The enhanced photoelectric performance demonstrates that the co-sensitized photoelectrode based on vertically grown TiO₂NSs array films will play an essential role in the photovoltaic field.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This study was financially supported by National Natural Science Foundation of China (No. 51272086). And the Technology Development Program of Jilin Province (Grant no. 20130206078GX).

Notes and references

1. M. Gratzel, Nature, 2001, 414, 338–344.
2. P. V. Kamat, J. Phys. Chem. Lett., 2013, 4, 908–918.
3. M. Ye, X. Gao, X. Hong, Q. Liu, C. He, X. Liu and C. Lin, Sustainable Energy Fuels, 2017, 1, 1217–1231.
4. X. Wang, Z. Li, J. Shi and Y. Yu, Chem. Rev., 2014, 114, 9346–9384.
5. Z. Chen, D. Pan, Z. Li, Z. Jiao, M. Wu, C.-H. Shek, C. M. L. Wu and J. K. L. Lai, Chem. Rev., 2014, 114, 7442–7486.
6. D. Fattakhova-Rohlfing, A. Zaleska and T. Bein, Chem. Rev., 2014, 114, 9487–9558.
7. W. J. Ong, L. L. Tan, S. P. Chai, S. T. Yong and A. R. Mohamed, Nanoscale, 2014, 6, 1946–2008.
8. Y. Zhao, H. Bala, W. Zhao, J. Chen, H. Li, W. Fu, G. Sun, J. Cao and Z. Zhang, Appl. Phys. Lett., 2014, 104, 013901.
9. E. Guo, L. Yin and L. Zhang, CrystEngComm, 2014, 16, 3403–3413.
10. Q. Wu, J. Hou, H. Zhao, Z. Liu, X. Yue, S. Peng and H. Cao, Dalton Trans., 2018, 47, 2214–2221.
11. Y. Chen, Q. Tao, W. Fu, H. Yang, X. Zhou, S. Su, D. Ding, Y. Mu, X. Li and M. Li, Chem. Commun., 2014, 50, 9509–9512.
12. S. Cheng, W. Fu, H. Yang, L. Zhang, J. Ma, H. Zhao, M. Sun and L. Yang, J. Phys. Chem. C, 2012, 116, 2615–2621.
13. Y. Lai, Z. Lin, D. Zheng, L. Chi, R. Du and C. Lin, Electrochim. Acta, 2012, 79, 175–181.
14. Q. Chen, R. Tong, X. Chen, Y. Xue, Z. Xie, Q. Kuang and L. Zheng, Catal. Sci. Technol., 2018, 8, 1296–1303.
15. H. G. Yang, C. H. Sun, S. Z. Qiao, J. Zou, G. Liu, S. C. Smith, H. M. Cheng and G. Q. Lu, Nature, 2008, 453, 638.
16. F. Li, J. Xu, L. Chen, B. Ni, X. Li, Z. Fu and Y. Lu, J. Mater. Chem. A., 2013, 1, 225.
17. D. Wu, S. Zhang, S. Jiang, J. He and K. Jiang, J. Alloys Compd., 2015, 624, 94–99.
18. L. Jiang, L. Sun, D. Yang, J. Zhang, Y.-J. Li, K. Zou and W.-Q. Deng, ACS Appl. Mater. Interfaces, 2017, 9, 9576–9583.
19. H. Zhou, L. Li, D. Jiang, Y. Lu and K. Pan, RSC Adv., 2016, 6, 67968–67975.
20. J. Yu, J. Fan and K. Lv, Nanoscale, 2010, 2, 2144–2149.
21. S. Hoang, S. P. Berglund, R. R. Fullon, R. L. Minter and C. B. Mullins, J. Mater. Chem. A., 2013, 1, 4307.
22. P. K. Santra and P. V. Kamat, J. Am. Chem. Soc, 2012, 134, 2508–2511.
23. Y. Ding, T. Zhang, C. Liu, Y. Yang, J. Pan, J. Yao, L. Hu and S. Dai, Sustainable Energy Fuels, 2017, 1, 520–528.
24. M. V. Vinayak, M. Yoosuf, S. C. Pradhan, T. M. Lakshmykanth, S. Soman and K. R. Gopidas, Sustainable Energy Fuels, 2018, 2, 303–314.
25. T. R. Rana, S. Kim, J. Kim, K. Kim and J. H. Yun, Sustainable Energy Fuels, 2017, 1, 1981–1990.
26. S. Kumar, M. Nehra, A. Deep, D. Kedia, N. Dilbaghi and K.-H. Kim, Renewable Sustainable Energy Rev., 2017, 73, 821–839.
27. J. H. Bang and P. V. Kamat, Adv. Funct. Mater., 2010, 20, 1970–1976.
28. Y. Lin, Y. Lin, Y. Meng, Y. Tu and X. Zhang, Opt. Commun., 2015, 346, 64–68.
29. J. Li, J. P. McClure, R. Fu, R. Jiang and D. Chu, Appl. Surf. Sci., 2018, 429, 48–54.
30. Z. Lu, J. Xu, X. Xie, H. Wang, C. Wang, S.-Y. Kwok, T. Wong, H. L. Kwong, I. Bello, C.-S. Lee, S.-T. Lee and W. Zhang, J. Phys. Chem. C, 2012, 116, 2656–2661.
31. J. Zhang, C. Sun, S. Bai, R. Luo, A. Chen, L. Sun and Y. Lin, Electrochim. Acta, 2013, 106, 121–126.
32. B. Zhang, J. Zheng, X. Li, Y. Fang, L.-W. Wang, Y. Lin and F. Pan, Chem. Commun., 2016, 52, 5706–5709.
33. X.-Q. Gong, A. Selloni, M. Batzill and U. Diebold, Nat. Mater., 2006, 5, 665.
34. S. S. Mali, R. S. Devan, Y.-R. Ma, C. A. Betty, P. N. Bhosale, R. P. Panmand, B. B. Kale, S. R. Jadkar, P. S. Patil, J.-H. Kim and C. K. Hong, Electrochim. Acta, 2013, 90, 666–672.
35. M. P. Genovese, I. V. Lightcap and P. V. Kamat, ACS Nano, 2012, 6, 865–872.
36. P. K. Santra, A. F. Palmstrom, J. T. Tanskanen, N. Yang and S. F. Bent, J. Phys. Chem. C, 2015, 119, 2996–3005.
37. C. She, A. Demortière, E. V. Shevchenko and M. Pelton, J. Phys. Chem. Lett., 2011, 2, 1469–1475.
38. J. M. Azpiroz, E. Ronca and F. De Angelis, J. Phys. Chem. Lett., 2015, 6, 1423–1429.
39. X. Jin, H. Li, S. Huang, X. Gu, H. Shen, D. Li, X. Zhang, Q. Zhang, F. Li and Q. Li, J. Colloid Interface Sci., 2018, 510, 376–383.
40. S. A. Pawar, D. S. Patil, H. R. Jung, J. Y. Park, S. S. Mali, C. K. Hong, J.-C. Shin, P. S. Patil and J.-H. Kim, Electrochim. Acta, 2016, 203, 74–83.
41. X. Ma, Y. Shen, G. Wu, Q. Wu, B. Pei, M. Cao and F. Gu, J. Alloys Compd., 2012, 538, 61–65.
42. Y. Zhu, R. Wang, W. Zhang, H. Ge and L. Li, Appl. Surf. Sci., 2014, 315, 149–153.
43. Z. Peng, Y. Liu, Y. Zhao, K. Chen, Y. Cheng and W. Chen, Electrochim. Acta, 2014, 135, 276–283.
44. H. Yao, W. Fu, H. Yang, J. Ma, M. Sun, Y. Chen, W. Zhang, D. Wu, P. Lv and M. Li, Electrochim. Acta, 2014, 125, 258–265.
45. C.-F. Chi, H.-W. Cho, H. Teng, C.-Y. Chuang, Y.-M. Chang, Y.-J. Hsu and Y.-L. Lee, Appl. Phys. Lett., 2011, 98, 012101.
46. Y.-L. Lee, C.-F. Chi and S.-Y. Liau, Chem. Mater., 2010, 22, 922–927.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8se00084k

---