Influence of Al₂O₃ nanoparticles embedded-TiO₂ nanofibers based photoanodes on photovoltaic performance of a dye sensitized solar cell

Abstract

Different weight percentages (1, 3 and 5 wt%) of the Al₂O₃ nanoparticles (NPs) embedded TiO₂ nanofibers (NFs) were prepared by the electrospinning technique. A mixture of titanium isopropoxide, acetic acid, Al₂O₃ NPs and PVP in ethanol were used. X-ray diffraction studies confirmed the formation of anatase and mixed phase (anatase/rutile) for the pure TiO₂ NFs and Al₂O₃ NPs embedded TiO₂ NFs, respectively. SEM and TEM studies showed the morphologies of Al₂O₃ NPs embedded TiO₂ NFs and their average diameter was found to be ∼83 nm. The chemical composition of TiO₂ NFs was confirmed by EDX analysis. The effects of Al₂O₃ NPs embedded TiO₂ nanofibers on the adsorption of dye and the charge transfer resistances were studied using UV-vis spectroscopy and electrochemical AC-impedance studies. Dye sensitized solar cells were fabricated using different weight percentages of Al₂O₃ NPs embedded TiO₂ NFs as photoanodes and Pt as counter electrode and 0.5 M 1-butyl-3-methylimidazolium iodide, 0.5 M LiI, 0.05 M I₂, 0.5 M 4-tertbutylpyridine in acetonitrile as the electrolyte. Among the three photoanodes tried, the cell fabricated using 3 wt% of Al₂O₃ NPs embedded TiO₂ NFs as photoanode had an improved photocurrent efficiency (PCE) of 36.2% compared to other similar systems in the literature. This might be due to an improvement in dye adsorption by the addition of a basic oxide (Al₂O₃) and also the formation of a rutile phase (15.7%), thereby retarding the charge recombinations inside the DSSC and in turn increasing the V_oc and the J_sc of the cell.

---

1. Introduction

Dye sensitised solar cells (DSSCs) are presently considered to be a promising candidate to replace the traditional p–n junction solar cells, which harvest clean and inexhaustible energy of the sun and convert it into electricity. Low cost fabrication and better solar conversion efficiency are the key factors involved in attracting researchers for this 3^rd generation solar cell.^1–3 The most crucial part of a DSSC is the wide band gap semiconductor metal oxides, which perform two important roles as being carrier for dye molecules and a transporter for the injected electrons. It has been found that the performance of DSSC depends mostly on the photoanode material, because it influences both the photocurrent and the photovoltage of the solar cell. Among the various reported photoanode materials (TiO₂, ZnO, SnO₂, etc.), TiO₂ is found to exhibit high power conversion efficiency owing to its unique optoelectronic properties.^4–7 There are two ways to optimise DSSC photoanodes, which will give an optimal light to electricity conversion.

The first one is to enhance the charge transport and the other is minimizing charge recombination which can be achieved by changing the morphology, porosity and crystallinity of the photoanodes. One dimensional (1D) nanostructured materials, such as nanofibers, nanotubes, nanobelts, nanowires and nanorods, showed better electron transport properties compared to nanoparticles based DSSCs.^8,9 In the case of one dimensional material, electron transport is faster because of the presence of lower grain boundaries, high connectivity and high surface area as observed in nanofibers. The nanofibers provide direct electric pathways to ensure rapid collection of charge carriers generated in the device.

The second one is to maximize the light harvesting capability of the device in the photoanode. This is done by using high surface area and high porosity, to adsorb maximum amount of dye. The nanofibers allow better filling of electrolyte in their pores, which improve the contact with the semiconductor and also help in regeneration of oxidized dyes, leading to high energy conversion efficiency.^10,11

Surface modification of TiO₂ can be done by thin insulating oxides, such as Y₂O₃, MgO, Nb₂O₅, ZnO and Al₂O₃, or using high band gap semiconductors to increase dye adsorption and thus to improve the photocurrent efficiency.^12–16 This difference is ascribed to the more basic nature of insulating oxides as compared to TiO₂, favouring the dye attachment through its carboxylic acid groups. The voltage and fill factor of the cell are strongly enhanced due to the suppression of electron back transfer from TiO₂ to the redox electrolyte in the presence of the insulating oxide.

In the present investigation, we prepared for the first time, Al₂O₃ NPs embedded TiO₂ NFs by an electrospinning technique and studied the influence of different wt% of Al₂O₃ NPs embedded TiO₂ nanofibers on the photovoltaic performances of DSSC.

2. Experimental details

2.1 Preparation of Al₂O₃ NPs embedded TiO₂ NFs

Different weight percentages (1, 3 and 5 wt%) of Al₂O₃ NPs embedded TiO₂ NFs were prepared by an electrospinning technique (Fig. 1) using a precursor solution containing 0.01 M titanium(IV) isopropoxide (TiP, Aldrich) and 0.04 M acetic acid, different wt% of Al₂O₃ NPs, and 5 wt% polyvinylpyrrolidone (PVP: M_W = 1 300 000, Sigma Aldrich) in ethanol. After stirring for 6 h, the solution was loaded into a syringe equipped with a 27G stainless steel needle. The spinning rate (0.5 mL min⁻¹) was controlled using a syringe pump. The electric field of 15 kV was applied between a metal orifice and an electrically grounded flat collector. The collector was placed at 12 cm below the tip of the needle. The humidity level inside the electrospinning chamber was maintained below 35% at ∼25 °C. The electrospun nanofibrous mats were removed from the collector and calcinated in a muffle furnace (Technico, India) at an increasing temperature rate of 5 °C min⁻¹ from 30 °C to 500 °C to get different wt% of Al₂O₃ NPs embedded TiO₂ NFs.

Fig. 1 Schematic illustration for electrospinning of Al₂O₃ NPs embedded TiO₂ nanofibers.

2.2 Characterization

The structural properties of different weight percentages of Al₂O₃ NPs embedded TiO₂ NFs were characterized by X-ray diffraction studies (Rigaku, Ultima IV) over a range of 2θ angles from 20° to 80°. The scanning electron microscopy (Hitachi, Model: S-4200) coupled with EDX (energy dispersive X-ray spectrometer) was employed to study surface morphology and the composition of different wt% of Al₂O₃ NPs embedded TiO₂ NFs. The morphologies of Al₂O₃ NPs embedded TiO₂ NFs were also probed using a transmission electron microscope (FEG-TEM, JSM-7600F). The absorption spectra were recorded by UV-vis spectrophotometer (PerkinElmer, Model: L-650) to study the influence of different wt% of Al₂O₃ NPs embedded TiO₂ NFs on adsorption of the dye.

2.3 Fabrication of DSSC

FTO glass plates were cleaned with the help of detergent solution for 10 min, followed by cleansing in de-ionised water, acetone, and ethanol for 5 min each using an ultrasonic bath, and then they were dried in hot air. The pure TiO₂ NFs and Al₂O₃ NPs embedded TiO₂ NFs paste was prepared by the following method: as-prepared pure TiO₂ NFs/Al₂O₃–TiO₂ NFs with PVP were stirred vigorously for 15 minutes followed by addition of 2 mL ethanol and 0.3 mL acetylacetone to avoid aggregation. Subsequently, 0.2 mL acetic acid was added. Finally 4-octylphenol polyethoxylate (Triton X-100) was added to act as a dispensing agent followed by 1 mL ethanol. PVP was used to increase the viscosity and adhesion property of the paste and is easily decomposed at 450 °C during sintering.¹⁷ The paste were coated with a thickness of 10–12 μm on FTO glass plates by the doctor blade technique and sintered at 450 °C for 30 min. The active area of the cell was found to be 0.2 cm². After cooling to 80 °C, the TiO₂ electrodes were immersed into the purified 0.3 mM N719 dye solution [di-tetrabutyl ammonium cis-bis(isothiocyanato)bis(2,2′-bipyridyl-4,4′-dicarboxylato)ruthenium(II)] for 12 h at 30 °C.¹⁸ After completion of the dye adsorption, the films were rinsed with pure ethanol to remove the excess dye. In the counter electrode, a hole was drilled to later pass the electrolyte and subsequently cleaned in an ultrasonic bath with acetone and ethanol for 5 min. Then, a Pt paste was applied using the doctor blade technique, followed by sintering in a muffle furnace at 450 °C for 30 min. A 60 μm Surlyn spacer (SX 1170-60 from Solaronix) was sandwiched between the photoelectrode and the standard platinum coated FTO counter electrode. The space of each cell was filled with 0.5 M LiI, 0.05 M I₂, 0.5 M 4-tertbutylpyridine in acetonitrile as the electrolyte by vacuum backfilling through the hole in the counter electrode. Finally, the hole was sealed with a hot-melt sheet and covered with a thin glass (0.1 mm).

2.4 Cell studies

The photovoltaic performances were measured using a computer controlled digital source meter. A Xe lamp with AM 1.5 filter (Oriel) was used to illuminate the DSSC at 100 mW cm⁻². Electrochemical AC-impedance measurements were carried out at room temperature in the frequency range of 1 mHz to 100 kHz with the AC amplitude of 10 mV to determine charge transfer resistance between the conducting oxide (FTO) and Al₂O₃ NPs embedded TiO₂ NFs and the interface between Al₂O₃ NPs embedded TiO₂ NFs |dye| electrolyte. The illumination condition for the impedance measurements was the same as that of J–V measurements.

3. Results and discussion

3.1 X-ray diffraction studies

Fig. 2(a–d) represents the X-ray diffraction (XRD) patterns of pure TiO₂ NFs and Al₂O₃ NPs embedded TiO₂ NFs. Fig. 2(a) shows that the TiO₂ NFs calcinations at 500 °C has a single anatase phase. The diffraction peaks representing the crystal planes (101), (004), (200), (105), (116) and (215) corresponding to the anatase phase, matches and is confirmed by the JCPDS data (Card nos 89-4920 and 89-4921). A small rutile phase along with the anatase phase has a synergistic effect on the efficient electron transport from rutile to anatase than a single anatase phase.^19–21 An increase in wt% of Al₂O₃ NPs embedded in TiO₂ NFs increased the percentage of rutile phase in TiO₂ NFs, which is shown in Fig. 2(b–d). But, no peaks corresponding to Al₂O₃ NPs were observed. The percentage of anatase/rutile was calculated from the X-ray diffractogram using the Spurr equation:
where, I_A is the intensity of (101) peak and I_R is the intensity of (110) peak. The effect of different wt% of Al₂O₃ NPs embedded TiO₂ NFs on the formation of anatase and rutile phases is presented in Table 1. The formation of rutile phase is because of geometrical effects and hence the overall number of potential surface nucleation sites per unit volume increased with a decrease in grain size which increase the rate of phase transformation.

Fig. 2 XRD patterns of TiO₂ Nanofibers and Al₂O₃ NPs embedded TiO₂ NFs.

Table 1 Effect of different wt% of Al₂O₃ NPs embedded TiO₂ NFs on the formation of anatase and rutile phases and their absorption band gap

Weight% of Al2O3 NPs in TiO2 NF's	% of anatase phase	% of rutile phase	Absorption band gap (eV)
0	100.0	0.0	3.20
1	91.6	8.4	3.14
3	84.6	15.7	3.06
5	82.8	17.2	3.00

---

3.2 EDX analysis

The X-ray diffraction study of Al₂O₃ NPs embedded TiO₂ NFs reveals no peaks for crystalline Al₂O₃ due to its very small wt% present in the samples. Table 2 shows EDX (energy dispersive X-ray spectrometer) characterisation results. The amount of Al₂O₃ NPs embedded TiO₂ NFs and their corresponding EDX spectra are shown in Fig. 3(a–d).

Table 2 EDX elemental map showing distribution of Ti, Al, and O, atoms in different wt% of Al₂O₃ NPs embedded TiO₂ NFs

Composition of NFs	Titanium wt(%)	Aluminum wt(%)	Oxygen wt(%)
Pure TiO2 NFs	23.01	0.00	76.99
1 wt% Al2O3 in TiO2 NFs	64.76	0.59	34.65
3 wt% Al2O3 in TiO2 NFs	42.83	0.92	56.25
5 wt% Al2O3 in TiO2 NFs	39.98	0.97	59.05

---

Fig. 3 EDX spectra of (a) pure TiO₂ NFs (b) 1 wt% Al₂O₃ NPs embedded TiO₂ NFs (c) 3 wt% Al₂O₃ NPs embedded TiO₂ NFs (d) 5 wt% Al₂O₃ NPs embedded TiO₂ NFs.

3.3 SEM analysis

Fig. 4(a) and (b) show low resolution SEM images of as-spun TiO₂ NFs and calcinated TiO₂ NFs. It was seen that the diameter of the nanofibers after calcination at 500 °C showed an overall reduction of ∼64.7% in their average diameter with respect to as spun nanofibers, which was calculated as ∼235 nm and ∼83 nm for as-spun and calcinated TiO₂ NFs, respectively. It was also observed that the continuous nature of TiO₂ NFs after calcination remained unchanged without any loss to their continuity; however, some breakage was observed in the nanofibers. Fig. 5(a–d) shows SEM images of pure TiO₂ NFs and different wt% of Al₂O₃ NPs embedded TiO₂ NFs, which were taken to further investigate for any change in the morphology of TiO₂ NFs. It was observed that no change in the morphologies occurred with the addition of Al₂O₃ NPs in TiO₂ NFs and the diameter of TiO₂ NFs also remains unchanged. The presence of Al₂O₃ NPs on TiO₂ NFs was not observed because of the low contrast. Hence, further analysis was carried out using high resolution transmission electron microscopy (HRTEM). Fig. 6(a) shows a representative TEM image of 3 wt% of Al₂O₃ NPs embedded TiO₂ NFs. It can be observed that the 3 wt% of Al₂O₃ NPs embedded over TiO₂ NFs as-depicted by black dots. The TEM image of the interface between TiO₂ and Al₂O₃ NPs is displayed in Fig. 6(b). The presence of the anatase TiO₂, rutile TiO₂ phase, and Al₂O₃ NPs, can be observed clearly in Fig. 6(c). In Fig. 6(d), the selected-area electron-diffraction (SAED) pattern confirms that the TiO₂ Nanofibers are composed of a polycrystalline phase; i.e., the Debye–Scherrer concentric rings of (101), (004), (200), (105), (204), (220), and (215) planes. The SAED pattern is in good agreement with the XRD results. Similarly, 1 wt% and 5 wt% of Al₂O₃ NPs embedded TiO₂ NFs show the same microstructures with no appreciable differences.

Fig. 4 SEM images of (a) as-spun TiO₂ NFs and (b) calcinated TiO₂ NFs.

Fig. 5 SEM images of different wt% of Al₂O₃ NPs embedded TiO₂ NFs (a) 0 wt% (b) 1 wt% (c) 3 wt% and (d) 5 wt% after calcination at 500 °C.

Fig. 6 TEM images of the 3 wt% of Al₂O₃ NPs embedded TiO₂ NFs: (a) bright field image, (b) & (c) high resolution images at different magnifications, and (d) SAED pattern.

3.4 UV-vis spectral studies

The influence of different weight percentage of Al₂O₃ NPs embedded TiO₂ NFs on dye adsorption was carried out. TiO₂ NF paste was prepared using the following method; TiO₂ NFs with PVP were stirred in the presence of ethanol and acetylacetone for 15 minutes, and acetic acid was added, subsequently. Finally, Triton X-100 was added as dispensing agent, followed by ethanol, and then the solution was concentrated by heating to obtain a viscous paste. These pastes were applied onto glass substrates by the doctor blade technique using adhesive tape as a frame and spacer. For each Al₂O₃ NPs embedded TiO₂ film, two identical photoanodes were fabricating and kept in air for 1 h and then placed inside a muffle furnace (Technico India) at 450 °C for 30 minutes. After sintering, the glass substrates with different wt% of Al₂O₃ NPs embedded TiO₂ NFs based films, were placed in a dye bath containing 0.3 mM cis-bis(isothiocyanato)bis-(2,2′bypiridyl-4,4′ dicarboxylato)ruthenium(II)bistetrabutylammonium (N719) dye in ethanol for 24 h, maintaining the sensitisation time the same for all the films such that we could avoid differences on the basis of time of sensitisation. The excess dye was removed by rinsing the surface of each glass substrates with ethanol. UV-vis spectral studies were done in solid mode for the dye adsorbed on different wt% of Al₂O₃ NPs embedded TiO₂ NFs (Fig. 7). The increase in light absorbance by Al₂O₃ NPs was explained on the basis of increase in dye adsorption. The amount of dye adsorbed on the TiO₂ NFs significantly increased with the addition of Al₂O₃ NPs from 0 wt%. to 3 wt%. The adsorption enhancement was related to an increase in Al₂O₃ NPs in TiO₂ NFs, which made the surface of TiO₂ NFs to become more basic in nature in view of high isoelectric point of Al₂O₃ nanoparticles (∼9.2). This helped to form strong chemisorption bonds with dye molecules via the carboxylic ends of the N719 dye.²² This improved the solar cell performances.^21–24 But, in the addition to the 5 wt% of Al₂O₃ NPs, the adsorption of dye on TiO₂ NFs decreased. This might be due to increase in the amorphous nature on the surface of TiO₂ NFs, which reduced the overall surface area by blocking the porosity on the surface of the nanofibers, thereby reducing the dye adsorption.^19,25 To know about the effect on the absorption band gap with increase in wt% of Al₂O₃ NPs embedded in TiO₂ NFs, we measured the absorption band gap from the UV-vis spectra using Origin Pro-9 and their values are given in Table 1. It can be observed that there was a reduction in band gap from 3.20 eV to 3.00 eV when increasing the wt% of Al₂O₃ NPs from 1 wt% to 5 wt% in TiO₂ NFs. The reduction in band gap is due to the addition of less valence cation (Al³⁺⁾ than that of TiO₂ (Ti⁴⁺), which increased the concentration of oxygen vacancies, such that the charge balance was obtained. This change introduced a defect in the system that reduced the overall grain size, which lead to a contraction in the system, thereby reducing the bandgap.

Fig. 7 UV-vis spectra of different wt% of Al₂O₃ NPs embedded TiO₂ NFs on adsorption of N719 dye.

Table 3 EIS parameters of R_s, R_ct1 and R_ct2 values for DSSCs fabricated using pure TiO₂ NFs and Al₂O₃ NPs embedded TiO₂ NFs

Photoanode	Rs	Rct1 (Ω cm^−2)	Rct2 (Ω cm^−2)
Pure TiO2 NFs	15.06	6.89	23.84
1 wt% Al2O3–TiO2 NFs	14.85	7.20	21.54
3 wt% Al2O3–TiO2 NFs	14.70	6.17	18.42
5 wt% Al2O3–TiO2 NFs	14.00	7.10	20.50

---

3.5 Electrochemical impedance studies

The electrochemical impedance study was carried out to study the effect of different wt% of Al₂O₃ NPs embedded TiO₂ NFs on charge transfer resistance. Nyquist plot consisted of two semicircles (Fig. 8 and Table 3), corresponding to the charge transport resistance at counter electrode/electrolyte (R_ct1) in higher frequencies (kilohertz range) and at electrical contacts between FTO/TiO₂ NFs or among TiO₂ NFs. The larger semicircle at lower frequencies corresponded to charge transfer resistance at TiO₂ NFs/dye/electrolyte interface (R_ct2) and the Warburg diffusion process of I⁻/I₃⁻ redox couple in electrolyte (Z_w) at open circuit voltage under darkness.^26–28 Though R_ct2 virtually overlapped Z_w, the identical semicircles at higher frequency suggested that Al₂O₃ NPs embedded in TiO₂ NFs did not affect the charge transfer at counter electrode/electrolyte interface. R_ct2 value decreased with the increase in Al₂O₃ NPs embedded in TiO₂ NFs up to 3 wt% and then increased with further increase in Al₂O₃ NPs embedded in TiO₂ NFs. The variations in R_ct2 value were consistent with the J_sc value of respective cell, i.e., when the R_ct2 value decreased, the J_sc value increased and vice versa. Increase in the Al₂O₃ NPs embedded in TiO₂ NFs from 3 wt% to at 5 wt%, led to increase in V_oc on expense of the J_sc value. This is due to the reduction in dye adsorption and less electrolyte uptake, which led to decrease in overall device power conversion efficiency (PCE).

Fig. 8 Nyquist Plots of DSSCs fabricated with different wt% of Al₂O₃ NPs embedded TiO₂ NFs.

3.6 Photovoltaic measurements

Table 4 summarizes the photovoltaic parameters of DSSCs based on different wt% of Al₂O₃ NPs embedded TiO₂ NFs. Fig. 9 represents the current density–voltage (J–V) characteristics of DSSCs fabricated with various wt% Al₂O₃ NPs embedded TiO₂ NFs as photoanodes. The DSSC fabricated using 3 wt% Al₂O₃ NPs embedded TiO₂ NFs as photoanode exhibited an improved power conversion efficiency of 36.2% with respect to DSSC fabricated using pure TiO₂ NFs. The obtained results were compared with the results reported for Al₂O₃ coated TiO₂ nanoparticles based DSSCs (Table 5). The Al₂O₃ NPs embedded TiO₂ NFs based DSSC exhibited higher percentage in PCE than other similar systems reported in the literature.^23,29,30 The higher efficiency of the cell is due to reduction in the back carrier recombination by the addition of Al₂O₃ NPs, which have higher conduction band edge and higher band gap (7 eV) than pure TiO₂ NFs (3.2 eV) allowing only a minor carrier flow to the oxidised dye molecules or in other words to the oxidised redox couple leading to an increase in conduction band electrons as reflected by increase in J_sc. In the case of 5 wt% of Al₂O₃ NPs embedded TiO₂ NFs, excess of Al₂O₃ decreased the dye adsorption and electrolyte uptake and increased the % of rutile phase, thereby decreasing the PCE.

Fig. 9 Photocurrent density–voltage characteristics of DSSCs fabricated with different wt% of Al₂O₃ NPs embedded TiO₂ NFs.

Table 4 Photovoltaic parameters of DSSCs fabricated with different wt% of Al₂O₃ NPs embedded TiO₂ nanofibers

Photoanode	Jsc (mA cm^−2)	Voc (V)	FF	η (%)
Pure TiO2 NFs	10.3	0.64	73	4.81
1% Al2O3–TiO2 NFs	10.8	0.66	75	5.35
3% Al2O3–TiO2 NFs	11.7	0.70	80	6.55
5% Al2O3–TiO2 NFs	11.0	0.72	76	5.85

---

Table 5 Comparison of photovoltaic performances of DSSCs fabricated by using Al₂O₃ coated TiO₂ as photoanodes

Photoanode	Voc (mV)	Jsc (%)	FF (%)	Cell efficiency (%)	% of improvement in PCE (%)	Ref.
TiO2/Al2O3 core shell	705.6	19.0	63	8.40	35.5	23
Al2O3 coated nanoporous TiO2 films	750	10.9	65	5.00	30.0	29
Al2O3 coated nanoporous TiO2	470	9.5	52	2.60	33.5	30
Al2O3 NPs embedded TiO2 NFs	750	11.7	80	6.55	36.2	Present work

---

4. Conclusions

DSSC fabricated using 3 wt% of Al₂O₃ NPs embedded TiO₂ NFs exhibited an overall percentage of improvement in PCE of 36.2% compared to standard DSSC made from pure TiO₂ NFs as photoanode. This improvement is due to the presence of an optimum rutile phase (15.7%) which helped in the synergetic effect to easily improve the electron transport from rutile to anatase phase. This improved the dye adsorption which in turn improved the photovoltaic performance of the cell by increasing both the V_oc and J_sc up to 3 wt% of Al₂O₃ NPs embedded in TiO₂ NFs. But in 5 wt% of Al₂O₃ NPs embedded in TiO₂ NFs, the dye adsorption lowered and offered high resistance to flow of electrons across TiO₂ NFs| Al₂O₃ |dye| electrolyte interface. An increase in J_sc of the cell was in agreement with the decrease in resistance up to 3 wt% of Al₂O₃ NPs embedded in TiO₂ NFs. With further increase of Al₂O₃ NPs embedded in TiO₂ NFs, the resistance to the flow of electrons across the TiO₂ NFs/dye/electrolyte (R_ct2) increased. Because of the increase in the tunnelling length by the large number of Al₂O₃ NPs on the surface of TiO₂ that hindered the electron injection, this leads to a decrease in J_sc and also the efficiency of the cell. Thus 3 wt% of Al₂O₃ NPs embedded TiO₂ NFs with 15.7% rutile phase could be used as an effective photoanode for DSSC to improve the efficiency of the cell.

Acknowledgements

The authors gratefully acknowledge the CSIR ref. no. 01(2359)/10/EMR-II, New Delhi for the financial support.

References

1. B. O. Regan and M. Gratzel, Nature, 1991, 353, 737–740.
2. M. Gratzel, Nature, 2001, 414, 338–344.
3. A. Hagfeldt and M. Gratzel, Acc. Chem. Res., 2000, 33, 269–277.
4. Y. G. Guo, Y. S. Hu, W. Sigle and J. Maier, Adv. Mater., 2007, 19, 2087–2091.
5. S. A. Akbar, L. B. Younkman and P. K. Dutta, ACS Symp. Ser., 1998, 690, 161–167.
6. M. K. Nazeeruddin, F. De Angelis, S. Fantacci, A. Selloni, G. Viscardi, P. Liska, S. Ito, B. Takeru and M. G. Gratzel, J. Am. Chem. Soc., 2005, 127, 16835–16847.
7. A. Fujishima and K. Honda, Nature, 1972, 238, 37–38.
8. A. C. Fisher, L. M. Peter, E. A. Ponomarev, A. B. Walker and K. G. U. Wijayantha, J. Phys. Chem. B, 2000, 104, 949–958.
9. J. Bisquert and V. S. Vikhrenko, J. Phys. Chem. B, 2004, 108, 2313–2322.
10. K. Onozuka, B. Ding, Y. Tsuge, T. Naka, M. Yamazaki, S. Sugi, S. Ohno, M. Yoshikawa and S. Shiratori, Nanotechnology, 2006, 17, 1026–1031.
11. M. Y. Song, Y. R. Ahn, S. M. Jo, D. Y. Kim and J.-P. Ahn, Appl. Phys. Lett., 2005, 87, 113113.
12. E. Palomares, J. N. Clifford, S. A. Haque, T. Lutz and J. R. Durrant, J. Am. Chem. Soc., 2003, 125, 475–482.
13. A. Kay and M. Gratzel, Chem. Mater., 2002, 14, 2930–2935.
14. Z.-S. Wang, C.-H. Huang, Y.-Y. Huang, Y.-J. Hou, P.-H. Xie, B.-W. Zhang and H.-M. Cheng, Chem. Mater., 2001, 13, 678–682.
15. S. G. Chen, S. Chappel, Y. Diamant and A. Zaban, Chem. Mater., 2001, 13, 4629–4634.
16. H. Alarcon, G. Boschloo, P. Mendoza, J. L. Solis and A. Hagfeldt, J. Phys. Chem. B, 2005, 109, 18483–18490.
17. S. N. Karthick, K. V. Hemalatha, C. Justin Raj, A. Subramania and H.-J. Kim, Thin Solid Films, 2012, 520, 7018–7021.
18. A. Subramania, E. Vijayakumar, N. Sivasankar, A. R. Sathiya Priya and K.-J. Kim, Ionics, 2013, 19, 1649–1653.
19. T. K. Yun, S. S. Park, D. Kim, J.-H. Shim, J. Y. Bae, S. Huhc and Y. S. Won, Dalton Trans., 2012, 41, 1284–1288.
20. J. Ding, Y. Li, H. Hu, L. Bai, S. Zhang and N. Yuan, Nanoscale Res. Lett., 2013, 8, 5–9.
21. G. H. Li, C. P. Richter, G. W. Brudvig and V. S. Batista, Dalton Trans., 2009, 45, 10078–10085.
22. X. T. Zhang, I. Sutanto, T. Taguchi, K. Tokuhiro, Q. B. Meng, T. N. Rao, A. Fujishima, H. Watanabe, T. Nakamori and M. Uragami, Sol. Energy Mater. Sol. Cells, 2003, 80, 315–326.
23. V. Ganapathy, B. Karunagaran and S.-W. Rhee, J. Power Sources, 2005, 195, 5138–5143.
24. C. P. Lee, P. Y. Chen, R. Vittal and K. C. Ho, Phys. Chem. Chem. Phys., 2010, 12, 9249–9255.
25. E. Barajas-Ledesma, M. L. Garcia-Benjume, I. Espitia-Cabrera, M. Ortiz-Gutierrez, F. J. Espinoza-Beltran, J. Mostaghine and M. E. Contreras-Garcia, Mater. Sci. Eng., B, 2010, 174, 71–73.
26. L. Han, N. Koide, Y. Chiba and T. Mitate, Appl. Phys. Lett., 2004, 84, 2433–2435.
27. Z. Zhang, S. M. Zakeeruddin, B. O. Regan, R. Humphry-Baker and M. Gratzel, J. Phys. Chem. B, 2005, 109, 21818–21824.
28. R. Kern, R. Sastrawan, J. Ferber, R. Stangl and J. Luther, Electrochim. Acta, 2002, 47, 4213–4225.
29. E. Palomares, J. N. Clifford, S. A. Haque, T. Lutz and R. James, Chem. Commun., 2002, 14, 1464–1465.
30. X.-T. Zhanga, I. Sutantoa, T. Taguchia, K. Tokuhiroa, Q.-b. Menga, T. N. Raoa, A. Fujishimaa, H. Watanabeb, T. Nakamorib and M. Uragami, Sol. Energy Mater. Sol. Cells, 2003, 80, 315–326.

---