Bilayered photoanode consisting of zinc oxide hollow spheres and urchin-like titanium dioxide microspheres enables fast electron transport and efficient light-harvesting for improved-performance dye-sensitized solar cells

Abstract

Derived from ZnO hollow spheres (ZHSs) as the underlayer and urchin-like TiO₂ spheres (UTSs) as the light scattering overlayer, a new bilayered photoanode (ZHS + UTS) is fabricated for use in dye-sensitized solar cells (DSSCs). The ZHSs which directly grow on FTO are synthesized via a one-pot hydrothermal method. The DSSC based on the ZHSs achieves an overall photoelectric conversion efficiency (PCE) of up to 4.84%, which is mainly ascribed to fast electron transport. On the other hand, three kinds of UTSs of different size and morphology are prepared by simply adjusting the titanium precursor dosage. The hierarchical UTSs composed of needle-like thin nanosheets have large specific surface areas, enabling sufficient dye adsorption. In particular, UTS-2 (synthesized using a TiCl₃ volume of 0.6 mL) shows strong light-scattering ability and hence is further applied as a scattering layer in the bilayered photoanode. The ZHS + UTS-2 based DSSC finally achieves the highest photocurrent density (J_sc = 18.13 mA cm⁻²) and therefore an enhanced PCE of 8.67%. The improved photovoltaic performance is attributed to the synergic effects of the ZHSs and UTSs, i.e. the fast electron transport favored by the highly crystallized ZHSs leading to a reduced interfacial charge recombination, the efficient light scattering effect due to the novel morphology of the UTSs, and the sufficient dye adsorption ensured by the large specific surface areas of the ZHSs and UTSs.

---

1. Introduction

During recent decades, the energy crisis has become a block to economic growth and human development. As the most important environmentally friendly and renewable power source, solar cells have become the key to solving this problem. Among the alternative solar cells, dye sensitized solar cells (DSSCs) with their low cost and high theoretical efficiency have attracted extensive interest.^1,2 Recently, M. Grätzel et al. achieved a photoelectric conversion efficiency (PCE) of up to 15% using an all solid-state cell structure.³ Compared with traditional DSSCs that use liquid electrolyte, solid-state DSSCs are more stable and less sensitive to photodegradation. As a well-known promising photoanode material for DSSCs, TiO₂ has been investigated considerably through extensive investigations conducted on the crystal structure, crystallite size and morphology.^4–6 As well as the photoanode material, the porous film structure also has a great influence on cell performance.^7–10

However, enhancement of the poor PCE of DSSCs is always hindered by the charge recombination in the photoanode film, especially for thick films.¹¹ To overcome the challenge of the recombination issue, nanostructured ZnO, with a wide direct bandgap (3.37 eV) and high electron mobility (115–155 cm² V⁻¹ S⁻¹), has recently been studied in the form of nanorods,^12,13 nanoflowers,¹⁴ nanospheres,¹⁵ etc. Particularly, ZnO nanorods are widely used as DSSC photoanode materials due to their fast electron transport which is favored by fewer grain boundaries; however, the small surface area of the nanorods leads to unfavorable scattering and reduced dye loading.^16,17 In contrast, ZnO hollow microspheres (ZHSs) have a marked electron conduction property with a relatively large specific surface area, and are satisfactory as photoanode materials and have good potential for DSSC application.¹⁸ Besides, bi-layer ZnO/TiO₂ or TiO₂/ZnO hybrid photoanodes in DSSCs have already been well reported and are proven to be an effective strategy to improve the photoelectric properties of DSSCs.^19,20

As is well known, bilayered photoanodes consisting of a light scattering overlayer and an electron transport underlayer have been demonstrated to be a promising strategy to improve DSSC performance.^21,22 The scattering layer can increase the optical path length within the photoanode film to improve the utilization rate of light, while the underlayer can facilitate electron transport and reduce charge recombination. Interestingly, although the scattering materials generally have a large surface area for sufficient dye loading, most of them exhibit poor electron transport properties.^23,24 Therefore, proper design of scattering materials in order to provide the dual function of light scattering and electron conduction is badly needed.

In our recent work, we developed a convenient ‘one-pot’ template-free synthesis method to fabricate ZHSs which grew directly on FTO glass. Compared with conventional templating methods, this hydrothermal method saves the trouble of a template removing process which not only consumes a lot of time and energy, but may also introduce impurities.^25,26 The ZHS based cell exhibits favorable light absorption and fast electron transport, as well as a low recombination rate. In addition, we also prepared UTSs via a facial hydrothermal method, and investigated the effect that the dosage of TiCl₃ has on the morphology and properties of the products. As a result, three kinds of UTSs (UTS-1, UTS-2 and UTS-3) were obtained. Interestingly, the monolayer UTS-2 DSSC achieves the best light scattering capability with a PCE of up to 8.28%, and hence is utilized as the scattering overlayer. Lastly, a bilayered ZHS + UTS-2 photoanode was fabricated for a DSSC. The ZHS + UTS-2 DSSC combines the advantages of both ZHSs and UTS-2, such as strong light scattering, fast electron transport and suppressed charge recombination, giving a high PCE of 8.67%.

2. Experimental section

2.1. Preparation of ZnO hollow spheres

ZHSs grown on FTO were prepared using a one-step hydrothermal method. Firstly, 0.1 mol of Zn(NO₃)₂·6H₂O and 10 mmol of sodium citrate were dissolved in 20 mL of distilled water. Then, 1 mL of NH₃·H₂O was added into the prepared aqueous solution. After being stirred for 0.5 h, the mixture was then transferred to a 45 mL Teflon-line autoclave. A piece of FTO glass (resistivity of 14 Ω per square, Nippon Sheet Glass, Japan), ultrasonically cleaned for 30 min in acetone and deionized water, respectively, was placed in the autoclave which was then sealed and kept at 175 °C for 24 h. After naturally cooling down to room temperature, the white film on the FTO was washed with deionized water and ethanol, and subsequently dried at 80 °C overnight. The porous ZHS film was finally obtained after annealing at 450 °C in air for 2 h at a heating rate of 2 °C min⁻¹.

2.2. Preparation of urchin-like TiO₂ spheres

In a typical synthesis of UTS-1, 0.2 mL of TiCl₃ solution (20 wt% of TiCl₃ in H₂O and HCl solution, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) was added into a brown glass bottle with an aqueous solution consisting of deionized water (10 mL), HCl (0.2 mL) and HAc (20 mL), and stirred for 20 min. The reaction solution was transferred into a 45 mL Teflon-lined autoclave and heated at 180 °C for 6 h. After being cooled to room temperature, the white precipitates were centrifuged repeatedly with deionized water and ethanol, and then dried at 80 °C overnight. To investigate the effect that the dosage of TiCl₃ has on the morphology and properties of the products, controlled experiments were also performed using the same process while altering the TiCl₃ volume to 0.6 mL and 1 mL, and the products were denoted as UTS-2 and UTS-3, respectively.

2.3. Characterization

The morphology and microstructure of the samples were observed using field-emission scanning electron microscopy (FESEM, JEOL JSM-7500F) and transmission electron microscopy (TEM, JEOL, EM-2100). X-ray diffraction (XRD; Rigaku TTRIII, with Cu Kα1 radiation) was employed to study the crystallographic information of the samples. Brunauer–Emmett–Teller (BET) surface area was characterized using a Micromeritics Gemini VII apparatus (Surface Area and Porosity System). The light-harvesting ability and dye adsorption amount of the film samples were determined using UV-vis absorption measurements (SHIMADZU 2550, equipped with an integrating sphere).

2.4. Fabrication and photovoltaic measurements of DSSCs

Viscous slurries of the UTSs were firstly prepared and then spread onto clean FTO using a doctor-blade method to form the films with a thickness of 41 μm. The active area of the film electrodes was about 25 mm². The bilayered photoanode (ZHS + UTS) was fabricated by covering 12 μm of UTS onto 29 μm of the ZHS bottom layer. After drying at 125 °C for 10 min, the resulting films were subsequently annealed at 450 °C for 30 min in air. The details about fabrication of slurries, sensitization of photoanodes and assembly of DSSCs have been fully described in our previous work.^27,28

Current–voltage (I–V) measurements of the cells were carried out using a Keithley 2400 Source Meter under the illumination of a xenon lamp solar simulator (Newport, model: 94023A) with a light intensity of 100 mW cm⁻² (AM 1.5). Incident photon-to-current conversion efficiency (IPCE) spectra were recorded on a Spectral Product Zolix DSC300PA. Electrochemical impedance spectroscopy (EIS) was performed using an electrochemical workstation (Solartron SI1287) in the frequency range of 10⁻¹ to 10⁵ Hz at open-circuit voltage.

3. Results and discussion

3.1. Characterization of materials

Fig. 1 displays the crystallinity of the four prepared samples. As can be seen in Fig. 1(a), all of the detectable peaks of the ZHSs can be indexed to a ZnO wurtzite structure, matching well with the standard data according to JCPDS (card no. 89-0511),²⁹ and no other peaks corresponding to impurities are detected, indicating a high crystallinity and purity for the ZHSs. The XRD patterns of UTS-1, UTS-2 and UTS-3 are presented in Fig. 1(b). In the three samples, the 2θ values at 25.32, 37.78, 48.11, 54.05, 55.08 and 62.57° can be indexed to distinct anatase TiO₂ (JCPDS no. 21-1272). It is favorable to prepare the anatase phase due to its better solar energy conversion than the rutile and brookite phases. Moreover, the average crystallite sizes of UTS-1, UTS-2 and UTS-3, calculated using the Scherrer formula: D = Kλ/β cos θ, are about 6.3, 8.0 and 10.2 nm, respectively.

Fig. 1 XRD patterns of (a) the as-synthesized ZHSs and (b) the UTSs (UTS-1, UTS-2, UTS-3).

The FESEM images were used to examine the morphologies and structures of the ZHSs and UTSs. Fig. 2(a) reveals that the size of the monodisperse ZHSs is about 3–4 μm, and that there are still some precursors which have not fully grown into the ZHSs on the surface of the film. The hollow structure is further confirmed from the image in the inset of Fig. 2(b). It can also be noted that a ZnO compact layer with a thickness of ∼1 μm grows on the FTO before the deposition process of the ZHSs, which can facilitate the interfacial electron collection. The side-view in Fig. 2(b) shows that the thickness of the ZHS film is about 28.2 μm. To be more convincing, TEM images of the ZHSs are also provided in Fig. 2(c). However, the hollow characteristic is not very obvious since the spherical shell is too thick (∼800 nm) to be penetrated by the electron beam. Fig. 2(d) shows a HRTEM image of a ZHS, revealing three distinct lattice spacings (d₁₀₀ = 0.28 nm, d₀₀₂ = 0.26 nm and d₁₀₁ = 0.25 nm) which coincide with the observation from the XRD patterns.

Fig. 2 (a and b) SEM images of the ZHSs. Top-view (a), side-view (b) and the corresponding high-magnification image (inset in b) of the ZHSs; (c) TEM and (d) HRTEM images of the ZHSs.

To understand the growth mechanism of the ZHSs, the effect of the hydrothermal time on the morphology of the products was investigated. However, except for the constantly increasing thickness of the ZHS film, no apparent structure distinction is observed between the SEM images of the ZHS samples at different reaction times, and there are always some uncompleted ZHSs distributed over the surface of the ZHS film, which implies that the ZHSs are formed in the solution and deposited on the FTO layer by layer. Fig. 3 shows SEM images of uncompleted ZHSs with different morphologies (Fig. 3(1–5)). According to the observation, a possible growth mechanism for completed ZHSs may be presented as follows: at the early stage, ZnO with a regular hexagonal prism morphology was formed (Fig. 3(1)), then this gradually grows into a polyhedron (Fig. 3(2)). By prolonging the reaction time, a ZnO sphere with a solid core was formed (Fig. 3(3 and 4)). Interestingly, the interior of the solid sphere will start to change obviously, which can be easily identified by the broken sphere shown in Fig. 3(5). At this stage, the solid interior is gradually hollowed out due to a corrosive effect. After the corrosive process is finished, the final product with a hollow structure can be obtained (Fig. 3(6)).

Fig. 3 SEM images of all kinds of uncompleted ZHSs with different morphologies.

The UTSs were synthesized by the hydrolysis of a titanium precursor in the presence of HCl and HAc. We mainly investigated the effect that the dosage of the precursor has on the morphology of the products. As shown in Fig. 4, the three kinds of spheres all exhibit urchin-like morphology, which is formed by irregularly arranged primary needle-like nanosheets. Fig. 4(a), (c) and (e) show that the diameters of these hierarchical spheres are about the same size (2–3 μm). However, with an increase of the dosage of TiCl₃, the density of the primary nanosheets improves (Fig. 4(b), (d) and (f)). The primary nanosheets are more densely arranged in UTS-2 than in UTS-1, and a further increase of TiCl₃ volume leads to much denser primary nanosheets which are closely packed with each other (Fig. 4(f)). Furthermore, the porous structure of the UTSs, created by the radial arrangement of the needle-like nanosheets can offer a good connection between adjacent spheres to facilitate electron transport, as well as improve the dye adsorption amount.

Fig. 4 SEM images of UTS-1 (a and b), UTS-2 (c and d) and UTS-3 (e and f) at different magnifications.

The BET surface areas and pore structures of the samples were investigated using N₂ adsorption–desorption measurements (Fig. 5), and the main surface parameters are summarized in Table 1. All of the samples show type IV isotherms with H3-type hysteresis loops, demonstrating the presence of slit-like mesopores (2–50 nm),³⁰ which coincides with the SEM images (Fig. 2 and 4). The BET surface areas of the ZHSs, UTS-1, UTS-2 and UTS-3 are estimated to be 57.8, 187.3, 126.3 and 100.3 m² g⁻¹, respectively. It is noteworthy that the large specific surface can ensure the sufficient dye adsorption amount, which is beneficial for improving the photocurrent. In addition, the inset of Fig. 5 shows the BJH pore size distribution curves, confirming the mesoporous nature of the porous ZHSs and UTSs. The ZHSs demonstrate two peaks located around 3.6 and 163 nm, evidencing the coexistence of mesopores and macropores. The most probable pore sizes for UTS-1 are located at around 20.1 and 139 nm, which are attributed to the outward arrangement of the primary needle-like nanosheets. While for UTS-2 and UTS-3, the pore size is centered at 9.8 nm and 5.5 nm, respectively, due to the high density and the quasi-parallel arrangement of the primary nanosheets.

Fig. 5 N₂ adsorption–desorption isotherms and corresponding BJH pore size distribution plots (inset) of the ZHSs, UTS-1, UTS-2 and UTS-3.

Table 1 Surface parameters of the ZHSs, UTS-1, UTS-2 and UTS-3

Sample	SBET (m^2 g^−1)	Pore volume (cm^3 g^−1)	Pore size (nm)
ZHS	57.82	0.45	29.6
UTS-1	187.28	1.17	37.6
UTS-2	126.31	0.60	10.2
UTS-3	100.29	0.46	12.1

---

3.2. Scattering effect of the materials

To investigate the scattering effect of the samples and find the most suitable scattering material, Fig. 6 compares the diffuse reflectance spectra of the photoanode composed of ZHSs and UTSs. Among the films tested, the UTS-2 film possesses the highest scattering capability in the visible and near-infrared regions, indicating that the incident light is significantly scattered within the film due to its micrometer-sized and nanosheet based sphere morphology. The ZHS film shows a relatively lower diffuse reflection because of the smooth spherical surface restraining the light scattering. In addition, the bilayered film applying UTS-2 as the scattering layer obviously improves the diffuse reflection capability. A detailed model of the light scattering for the three different kinds of film is shown in Fig. S1.† The novel rough surface of the UTSs which can extend the light path within the electrode is the main factor in enhancing the light scattering ability.

Fig. 6 UV-vis diffuse reflectance spectra of the ZHS and UTS photoanodes before dye loading.

3.3. Photovoltaic performance

Before the construction of the bilayered photoanode, we firstly studied the photovoltaic performance of the three kinds of UTSs (Fig. S2 and Table S1†), and found that the UTS-2 DSSC exhibited the highest PCE, at 8.28%, with a high fill factor (FF = 0.69). It is noteworthy that although UTS-1 possesses the highest specific area and a larger pore diameter that can facilitate the filling of electrolyte, as well as the highest J_sc (17.96 mA cm⁻²) when assembled in DSSCs, it shows lower FF and thus lower efficiency. We ascribe this to the smallest crystallite size of UTS-1 creating more grain boundaries that inhibit the electron transport. Fig. S3† shows the Nyquist plots of the UTS-1, UTS-2 and UTS-3 based DSSCs, and the corresponding simulative values of resistance (R₁, R₂ and R₃) from EIS spectra calculated using an equivalent circuit are presented in Table S2,† showing the relatively stronger electron transport property of the UTS-2 based DSSCs. Hence, UTS-2 was applied in the bilayered photoanode as the scattering layer. The I–V curves of the DSSCs based on the three different photoanodes (ZHS, UTS-2 and ZHS + UTS-2) under AM1.5G illumination are shown in Fig. 7(a), and the specific photovoltaic parameters are summarized in Table 2. As expected, the ZHS + UTS-2 DSSC exhibits the best cell performance, achieving an enhanced PCE of 8.67%. It is obvious that the photocurrent density (J_sc) and PCE are strongly related to the dye adsorption of the photoanode. The dye adsorption amounts of the three films were obtained by dye desorption experiments (Table 2). Hence, the UTS-2 DSSC shows a much higher J_sc than the ZHS DSSC as a result of a larger dye adsorption (1.70 × 10⁻⁷ mol cm⁻²). Moreover, the ZHS + UTS-2 DSSC shows the highest J_sc of 18.13 mA cm⁻², which is higher than that of the UTS-2 DSSC, while its dye adsorption is less than for the UTS-2 DSSC. This is because the J_sc of DSSCs is generally determined by the initial number of photogenerated carriers, the injection efficiency from dye molecules to the semiconductor, and the recombination rate between the injected electrons and the oxidized dye or redox species in the electrolyte. Therefore, even though the photogenerated carriers decrease due to the reduced dye adsorption, the ZHS underlayer can facilitate the electron injection and inhibit charge recombination, hence, a higher J_sc is obtained. Fig. 7(b) shows a structure diagram of the bilayered ZHS + UTS-2 photoanode, and the corresponding cross-sectional SEM image is shown in Fig. S4.†

Fig. 7 (a) I–V curves of the DSSCs based on ZHS, UTS-2 and ZHS + UTS-2 photoanodes. (b) Schematic illustration of the ZHS + UTS-2 bilayered photoanode structure. (c) Absorption spectra of the ZHS, UTS-2 and ZHS + UTS-2 films after the dye loading process. (d) IPCE spectra of the DSSCs based on the above three photoanodes.

Table 2 Photovoltaic parameters and experimental EIS data of the DSSCs based on ZHS, UTS-2 and ZHS + UTS-2 photoanodes

DSSC	Jsc (mA cm^−2)	Voc (V)	FF (%)	η (%)	R3 (Ω)	fmax (Hz)	τe (ms)	Dye adsorbed (×10^−7 mol cm^−2)
ZHS	13.80	0.56	62	4.84	58.3	1.68	94.7	1.42
UTS-2	17.34	0.73	69	8.28	103.8	12.16	13.1	1.70
ZHS + UTS-2	18.13	0.73	68	8.67	73.7	3.65	43.6	1.63

---

The existence of a light scattering layer can significantly extend the travelling distance of light within the photoanode film and thus intensively increase the probability of the incident photons interacting with the dye molecules, which will be directly reflected in the enhancement of the light harvesting efficiency. As shown in Fig. 7(c), the ZHS + UTS-2 film possesses the best light absorption ascribed to the utilization of UTS-2 which improves the dye adsorption and light scattering capability simultaneously. A schematic of the light-scattering effect favored by the UTS-2 layer is also illustrated in Fig. 7(b).

IPCE spectra for the three kinds of DSSCs are provided in Fig. 7(d) and are used to analyze in more detail the enhanced cell performance. The IPCE is determined by the light absorption efficiency, the quantum yield of electron injection, and the efficiency of collecting the injected electrons at the conducting glass substrate, which is also strongly affected by the morphology of the photoanode material.^31–33 The results are in good accordance with the I–V measurements and light absorption spectra. As shown in Fig. 7(d), the ZHS + UTS-2 DSSC reveals the highest IPCE values over the whole spectral region of 400–800 nm. At the maximum value of the IPCE spectra at 525 nm (the peak of the N719 adsorption spectrum), the IPCE of the ZHS + UTS-2 DSSC is approximately 62% and 3.6% higher than that of the ZHS DSSC and UTS-2 DSSC, respectively. The higher IPCE values can be attributed to the following reasons: (I) the large specific surface area (126.31 m² g⁻¹) of UTS-2 leads to excellent dye loading which further promotes light absorption; (II) the superior light scattering properties of UTS-2 enhances the light-harvesting capability; (III) the collaboration of the ZHS bottom layer with the dense single-crystalline nanosheet based UTS-2 provides more efficient electron transport pathways and reduces the charge recombination rate.

EIS is a powerful tool to study the electron transport phenomena and recombination characteristics within an electrochemical system. EIS measurements of the DSSCs based on the three different photoanodes were performed at a frequency range of 10⁻¹ to 10⁵ Hz under one-sun illumination at V_oc.^34,35 Fig. 8(a) shows the Nyquist plots, and the inset presents the equivalent circuit of the DSSCs obtained from ZView software fitting. Generally, the spectra may show three distinguishable semicircles, which are related to the electrochemical reaction at the electrolyte/Pt/FTO interface (R₂ in the kHz range), the electrochemical reaction at the metal oxide film/dye/electrolyte interface (R₃ in the 1–1000 Hz range) and the Warburg diffusion process of I⁻/I₃⁻ (<1 Hz range) from the left to the right circle, respectively.^36,37 However, the third semicircle is too weak to be observed in Fig. 8(a), implying very fast ion transport in the electrolyte. The three cells also show nearly equal values of R₁ (lumped series resistance for the cell) and R₂ due to the same counter electrode (Pt/FTO glass) and electrolyte being applied in this work. Furthermore, the extent of electron transport in the photoanode can be judged by the impedance (R₃), which is defined by the diameter of the middle semicircle. As shown in Table 2, the R₃ of the ZHS DSSC is much smaller than that of the UTS-2 DSSC, suggesting that the ZHS DSSC has lower interfacial charge recombination and faster electron transfer. It is noteworthy that UTS-2 with ZHS as an underlayer shows a marked decrease in R₃ compared with that of the UTS-2 DSSC, indicating that the ZHSs contribute to the effective charge transport at the photoanode.

Fig. 8 Nyquist plots (a) and Bode phase plots (b) of the ZHS, UTS-2 and ZHS + UTS-2 DSSCs. The inset in (a) illustrates the equivalent circuit simulated to fit the impedance spectra.

Fig. 8(b) is the corresponding Bode plot from which the electron lifetime (τ_e) in the photoanode film can be acquired (τ_e = 1/2πf_max). According to the EIS model developed by Kern et al.,³⁸ the f_max is the maximum frequency of the intermediate frequency regime. Therefore, a decrease in the f_max means a longer τ_e and thus, a slower recombination. As shown in Table 2, the f_max values of the different DSSCs increase in the order: ZHS (1.68 Hz) < ZHS + UTS-2 (3.65 Hz) < UTS-2 (12.16 Hz), while the order of their corresponding τ_e is reversed as ZHS (94.7 ms) > ZHS + UTS-2 (43.6 ms) > UTS-2 (13.1 ms). The EIS measurements are carried out at the V_oc condition, which means the electrons injected into the photoanode must recombine with I₃⁻ ions at the metal oxide film/dye/electrolyte interface. Hence, the increase in τ_e implies that the photo-induced electrons can diffuse further without trapping or interruption within the film. As a result, the ZHS + UTS-2 DSSC reveals a prolonged electron lifetime with a slower recombination rate, which can be attributed to the higher electron mobility through the highly single crystalline ZHSs and the compact ZnO layer beneath.

4. Conclusions

In summary, a new route toward ZnO hollow spheres is developed. The monolayer ZHS based cell exhibits an excellent electron lifetime (τ_e = 94.7 ms), indicating that the ZHSs have superior interfacial electron transport properties with a reduced charge recombination rate. In addition, three different kinds of UTSs with large specific surface areas are investigated. UTS-2, which is finally selected as the scattering material, shows the best light scattering ability and strong dye adsorption ability, thus improving the light harvesting efficiency and cell performance. A bilayered photoanode composed of a ZHS underlayer and a UTS-2 overlayer is fabricated for DSSCs, offering a PCE of 8.67% which is 4.7% higher than that of a UTS-2 DSSC, and far superior to that of a ZHS DSSC. The key factor contributing to the enhancement is that this facile structure strategy combines the advantages of the ZHSs and the UTSs.

Acknowledgements

This work was supported by National Nature Science Foundation of China (No. 61374218, 61134010, and 61327804) and Program for Chang Jiang Scholars and Innovative Research Team in University (No. IRT13018), National High-Tech Research and Development Program of China (863 Program, No. 2013AA030902 and 2014AA06A505) and the project development plan of science and technology of Jilin Province (20130521009JH).

Notes and references

1. B. Oregan and M. Grätzel, Nature, 1991, 353, 737.
2. M. K. Nazeeruddin, F. De Angelis, S. Fantacci, A. Selloni, G. Viscardi, P. Liska, S. Ito, B. Takeru and M. G. Grätzel, J. Am. Chem. Soc., 2005, 127, 16835.
3. J. Burschka, N. Pellet, S. J. Moon, R. Humphry-Baker, P. Gao, M. K. Nazeeruddin and M. Grätzel, Nature, 2013, 499, 316.
4. F. Zhu, D. P. Wu, Q. Li, H. Dong, J. M. Li, K. Jiang and D. S. Xu, RSC Adv., 2012, 2, 11629.
5. S. Y. Yang, P. N. Zhu, A. S. Nair and S. Ramakrishna, J. Mater. Chem., 2011, 21, 6541.
6. G. H. Tian, Y. J. Chen, W. Zhou, K. Pan, C. G. Tian, X. R. Huang and H. G. Fu, CrystEngComm, 2011, 13, 2994.
7. W. G. Wang, H. Y. Zhang, R. Wang, M. Feng and Y. M. Chen, Nanoscale, 2014, 6, 2390.
8. H. Liu, T. Liu, X. Dong and Z. Zhu, Mater. Lett., 2014, 134, 240.
9. W. Q. Wu, Y. F. Xu, H. S. Rao, C. Y. Su and D. B. Kuang, J. Am. Chem. Soc., 2014, 136, 6437.
10. J. J. Fan, S. W. Liu and J. G. Yu, J. Mater. Chem., 2012, 22, 17027.
11. J. Nissfolk, K. Fredin, A. Hagfeldt and G. Boschloo, J. Phys. Chem. B, 2006, 110, 17715.
12. L. B. Li, Y. F. Wang, H. S. Rao, W. Q. Wu, K. N. Li, C. Y. Su and D. B. Kuang, ACS Appl. Mater. Interfaces, 2013, 5, 11865.
13. C. J. Lin, S. J. Liao, L. C. Kao and S. Y. H. Liou, J. Hazard. Mater., 2015, 291, 9.
14. Q. Deng, H. B. Tang, G. Liu, X. P. Song, G. P. Xu, Q. Li, D. H. L. Ng and G. Z. Wang, Appl. Surf. Sci., 2015, 331, 50.
15. H. M. Sun, L. M. Wang, D. Q. Chu, Z. C. Ma, A. X. Wang, Y. J. Zheng and L. X. Wang, Ceram. Int., 2014, 40, 16465.
16. J. T. Xi, Q. F. Zhang, K. Park, Y. M. Sun and G. Z. Cao, Electrochim. Acta, 2011, 56, 1960.
17. Q. Zhang, C. S. Dandeneau, X. Zhou and C. Cao, Adv. Mater., 2009, 21, 4087.
18. L. Wang, C. Ma, X. L. Ru, Z. Guo, D. P. Wu, S. B. Zhang, G. H. Yu, Y. S. Hu and J. S. Wang, J. Alloys Compd., 2015, 647, 57.
19. C. K. Xu, J. M. Wu, U. V. Desai and D. Gao, Nano Lett., 2012, 12, 2420.
20. Y. Li, P. F. Lu, M. L. Jiang, R. Dhakal, P. Thapaliya, Z. H. Peng, B. Jha and X. Z. Yan, J. Phys. Chem. C, 2012, 116, 25248.
21. Z. Y. Gao, Z. L. Wu, X. M. Li, J. L. Chang, D. P. Wu, P. F. Ma, F. Xu, S. Y. Gao and K. Jiang, CrystEngComm, 2013, 15, 3351.
22. Y. C. Rui, Y. G. Li, Q. H. Zhang and H. Z. Wang, Nanoscale, 2013, 5, 12574.
23. Z. Y. Gao, Z. L. Wu, X. M. Li, J. L. Chang, D. P. Wu, P. F. Ma, F. Xu, S. Y. Gao and K. Jiang, CrystEngComm, 2013, 15, 3351.
24. Y. C. Rui, Y. G. Li, Q. H. Zhang and H. Z. Wang, CrystEngComm, 2013, 15, 1651.
25. H. Li, C. S. Ha and I. Kim, Langmuir, 2008, 24, 10552.
26. M. M. Titirici, M. Antonietti and A. Thomas, Chem. Mater., 2006, 18, 3808.
27. P. L. Zhao, P. F. Cheng, B. Wang, S. T. Yao, P. Sun, F. M. Liu, J. Zheng and G. Y. Lu, RSC Adv., 2014, 4, 64737.
28. P. L. Zhao, S. T. Yao, M. Wang, B. Wang, P. Sun, F. M. Liu, X. S. Liang, Y. F. Sun and G. Y. Lu, Electrochim. Acta, 2015, 170, 276.
29. J. Liu, A. X. Wei, Y. Zhao, K. B. Lin and F. Z. Luo, J. Mater. Sci.: Mater. Electron., 2014, 25, 1122.
30. K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J. Rouquerol and T. Siemieniewska, Pure Appl. Chem., 1985, 57, 603.
31. S. Ito, P. Chen, P. Comte, M. K. Nazeeruddin, P. Liska, P. Pechy and M. Grätzel, Prog. Photovolt. Res. Appl., 2007, 15, 603.
32. S. Hore, C. Vetter, R. Kern, H. Smit and A. Hinsch, Sol. Energy Mater. Sol. Cells, 2006, 90, 1176.
33. Y. Chiba, A. Islam, R. Komiya, N. Koide and L. Y. Han, Appl. Phys. Lett., 2006, 88, 223505.
34. R. Jose, V. Thavasi and S. Ramakrishna, J. Am. Ceram. Soc., 2009, 92, 289.
35. M. Wu, X. Lin, A. Hagfeldt and T. Ma, Chem. Commun., 2011, 47, 4535.
36. K. S. Kim, H. Song, S. H. Nam, S. M. Kim, H. S. Jeong, W. B. Kim and G. Y. Jung, Adv. Mater., 2012, 24, 792.
37. Q. Wang, J. E. Moser and M. Grätzel, J. Phys. Chem. B, 2005, 109, 14945.
38. R. Kern, R. Sastrawan, J. Ferber, R. Stangl and J. Luther, Electrochim. Acta, 2002, 47, 4213.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra25225c

---