Three-dimensional nanocomposite formed by hydrophobic multiwalled carbon nanotubes threading titanium dioxide as the counter electrode of enhanced performance dye-sensitized solar cells

Abstract

A novel strategy of fast solvent induced assembly is used to synthesize a three-dimensional (3D) nanocomposite of multiwalled carbon nanotubes (MWCNTs) and TiO₂, as the counter electrode (CE) of dye-sensitized solar cells (DSSCs). The structure and morphology studies show that the as-prepared nanocomposite possesses a porous structure, large specific surface area and abundant active sites. Electrochemical impedance spectroscopy and cyclic voltammetry analysis reveal that the developed structure facilitates a fast rate of reduction of I₃⁻ and promotes charge transportation. The DSSCs exhibit a remarkable enhanced power conversion efficiency (PCE) of 7.95%, as compared to 7.38% for the conventional Pt CE, after irradiation under standard conditions (AM 1.5G, 100 mW cm⁻²).

---

1. Introduction

With the fossil energy depletion and increasingly serious environmental issues, solar energy is treated more and more as an effective means to mitigate the problems of global energy shortage and environmental pollution. Solar energy is a kind of clean and renewable energy,¹ and photovoltaic (PV) technology is an efficient way to convert solar energy into electrical energy.² As typical representatives of third generation solar cells, the artificial DSSCs have attracted worldwide interest due to their excellent properties of low cost, simple fabrication, environmental friendliness, and relatively high conversion efficiency, etc.^3,4 Over the past 20 years, tremendous progress has been made with respect to the efficiency, reproducibility, and stability of DSSCs.⁵ So far, the record efficiency of DSSCs in the laboratory is 14.5%.⁶ In general, there are four key components that impact the performance of DSSCs, which are nanocrystalline photoanode, dye, electrolyte and CE.^5,7 Usually, researchers use semiconductors of metal oxides like ZnO or TiO₂ as a transparent photoanode material for an electron transportation layer in DSSCs.^8,9 Dye molecules are attached to the surface of the photoanode to capture the incident photons and generate electron/hole pairs while the DSSCs are under illumination,¹⁰ and the electrolyte containing redox pairs of triiodide/iodide (I₃⁻/I⁻) is filled between the photoanode and CE.¹¹ Compared with other parts of the DSSCs, the corresponding research progress on CE is much slower.^12–15 Usually, the conductive glass loaded with Pt is a popular CE material for DSSCs. However, the expensive Pt enhances the cost of the DSSCs and hinders their practical application.⁴ Therefore, it is necessary to investigate cost-effective CE material for DSSCs. As we know, the CE mainly plays two vital roles in a typical DSSC device: one is to collect the electrons from the external circuit and the other is to be an electrocatalyst for the reduction of the redox pairs in the electrolyte solution.^5,16 As a consequence, a perfect CE should possess the characteristics of good electrical conductivity and excellent catalytic activity at the same time. Additionally, it must be highly stable and of low-cost.

With the factors of low cost, high durability, excellent catalytic activity and electrical conductivity, carbonaceous materials such as carbon nanotubes (CNTs),^12,13 carbon nanosheets,¹⁴ carbon black,¹⁵ and graphite¹⁶ have been widely studied as effective alternative CEs in DSSCs over the past decades. Among them, carbon nanotubes CEs have attracted a great deal of interest because of their unique structure and outstanding performance.^12,17 Various groups have employed pure CNTs, along with other materials for CE of DSSCs; however, their conversion is insufficient due to their improper structures, irregular fabrication processes and lack of interpenetrative networks.^18–21

In this study, a kind of 3D nanocomposite has been synthesized, through the method of fast solvent induced assembly, for CE of DSSCs. A thin film of the counter electrode is prepared on fluorine doped tin oxide (FTO) via a relatively simple technique of doctor-blading. The 3D interconnected porous structure not only provides fast transport channels for charge carriers, but also sufficient catalytic active sites for iodine reduction, due to the large surface area. Notably, our as-prepared MWCNTs/TiO₂ nanocomposite based DSSCs have achieved a PCE of 7.95%, under standard illumination (AM 1.5G, 100 mW cm⁻²).

2. Experimental section

2.1. Materials and sample preparation

In our study, all reagents and chemicals were of analytical grade. Among them, ethanol (C₂H₆O), terpineol (C₁₀H₁₈O), titanium tetrachloride (TiCl₄), tetrabutyl titanate (C₁₆H₃₆O₄Ti), ethyl cellulose, and acetic acid (HAc) were purchased from Sinopharm Chemical Reagent Co. Ltd. FTO glass (sheet resistance, 7 Ω cm⁻² and thickness, 2.2 mm) was supplied by Dalian HeptaChroma SolarTech Co. Ltd. It was washed ultrasonically with detergent, deionized water, isopropanol, and ethanol for 30 min, respectively, then treated with O₂ plasma for 15 min. MWCNTs were purchased from Beijing Boyu Gaoke New Material Technology Co. Ltd. N719 (cis-bis (isothiocyanato)bis(2,2′-bipyridyl-4,4′-dicarboxylato)ruthenium(II)bis-tetrabutylammonium) was supplied by Solaronix. Moreover, lithium iodide (LiI, 99.999%), iodine (I₂, 99.99%), TiCl₄·3H₂O (99.99%), 1-methyl-3-propylimidazolium iodide (PMII, 98%), 4-tert-butylpyridine (4-TBP, 96%) and tert-butyl alcohol (99.5%) were obtained from Aladdin. Acetonitrile (99.8%) and valeronitrile (99%) were purchased from Alfa Aesar.

The mesoporous TiO₂ nanoparticles were synthesized by a hydrothermal process as reported in literature.^22–24 Briefly, to prepare TiO₂ nanoparticles, 10 mL tetrabutyl titanate and 15 mL ethanol were added into a 100 mL beaker under vigorous stirring for 2 h at room temperature. Then, 10 mL acetic acid and 40 mL deionized water were put into the above solution. After vigorous stirring for 1 h, the reacted solution was transferred into a 100 mL hydrothermal reactor in an air oven at 200 °C for 8 h. After hydrothermal treatment, the obtained solution was centrifuged and the precipitate was washed with deionized water and ethanol three times, respectively. Finally, the wet TiO₂ precipitate was dried at 80 °C in an air oven overnight.

To develop the 3D structure of the nanocomposite, we modified the synthetic process reported in literature.^19,25 The procedures for the nanocomposite formation are shown in Fig. 1. In brief, 2.72 g hydrophobic MWCNTs were mixed with 80 mL non-polar solvent (toluene) in a 200 mL beaker, followed by ultrasonication and stirring for 30 min. Then, 0.68 g of as-synthesized TiO₂ nanoparticles were dispersed in the above solution and stirred for 2 h. Next, to promote the formation of short 3D assemblies, 100 mL of polar solvent (methanol) were added to the mixture under fast magnetic stirring. Centrifugation was used for the solid/liquid reacted solution. The resultant wet precipitate was washed three times with methanol and dried at 70 °C in a vacuum oven for 8 h. Finally, the dried powder was ground for 30 min and then annealed in a muffle furnace at 400 °C for 30 min.

Fig. 1 Schematic illustration of the nanocomposite formation.

2.2. Preparation of CE

To form the nanocomposite slurry, 1 g of the as-prepared nanocomposite powder, 0.5 mL acetic acid, 3 mL terpineol, 0.8 g ethyl cellulose and 10 mL ethanol were added to a 25 mL beaker and stirred vigorously for 48 h. After agitation, the solution was ground in a mortar, and processed ultrasonically for 30 min. Subsequently, the 3D nanocomposite slurry was placed onto the FTO substrate via the doctor-blading method. The thickness of the film was controlled by magic scotch tape (thickness, 50 μm). Finally, the CE was sintered in an annealing furnace at 400 °C for 30 min. For a comparison study, we prepared a Pt CE according to literature.²⁶

2.3. Fabrication of photoanode for DSSCs

The photoanode for DSSCs was fabricated according to the reported literature procedure.²² In detail, 1.5 g of as-synthesized TiO₂ nanoparticles were mixed with 12 mL ethanol in a 50 mL beaker, and 0.4 mL of acetic acid were added dropwise under magnetic stirring. After 10 min, 0.8 g ethyl cellulose and 0.5 g terpineol as binder and surfactant, respectively, were added to the above solution and stirred continuously for 48 h at room temperature. The white slurry was then ground in a mortar and treated with ultrasonication for 30 min. To fabricate a compact layer of TiO₂, the O₂ plasma treated FTO glass was immersed in 40 mM of TiCl₄ aqueous solution and kept at 70 °C in an oven under air flow for 40 min. Next, the extremely transparent TiO₂ slurry (thickness controlled by one layer of magic scotch tape) was coated on the above treated substrate via doctor-blading, and then heated to the temperature of 130 °C. Another thin layer of TiO₂ was coated onto the treated substrate and sintered at 500 °C for 1 h. After thermal annealing, the as-prepared sample was again treated with TiCl₄ solution. Later, the sample was washed three times with ethanol and calcined at 520 °C for 30 min. Finally, for the process of dye-sensitization, the TiO₂ photoanode was heated to 80 °C and soaked in 0.5 mM of N719 dye in acetonitrile/tert-butanol solution (v : v/1 : 1) at room temperature for 18 h, followed by rinsing with ethanol to remove the physically adsorbed dye molecules and drying in air.

2.4. Measurement and characterization

The photocurrent density versus voltage (J–V) measurements were carried out under dark and standard illumination with a solar simulator (Newport, USA) under AM 1.5G (100 mW cm⁻², calibrated with Si-reference cell certificated by NREL) conditions. Electrochemical impedance spectroscopy (EIS) measurements were recorded using an Autolab electrochemical workstation (model AUT84315, the Netherlands). Cyclic voltammetry (CV) measurements were carried out in a three-electrode system using a CHI660C potentiostat, with a platinum electrode and Ag/AgCl electrode as the counter electrode and reference electrode, respectively, at a scan rate of 50 mV s⁻¹ in acetonitrile solution containing 1 mM I₂, 10 mM LiI, and 100 mM LiClO₄·3H₂O. The X-ray diffraction (XRD, PANalytical B.V, Netherlands, radiation wavelength of 1.54060 Å) system, furnished with Cu-Kα radiation, was used to study the crystalline structures and analyze the phase purity of the samples. The surface morphologies of the samples were observed by a field emission scanning electron microscope (FE-SEM, FEI NOVA NanoSEM 450). Energy-dispersive X-ray spectroscopy (EDS) along with FE-SEM was used for the determination of various elements in the prepared films. Transmission electron microscopy (TEM) and high-resolution transmission electron microscopy (HRTEM) images were observed using FEI Tecnai G² 20 UTwin. Raman spectroscopy studies were conducted using a HORIBA Jobin Yvon UV-VIS-NIR LabRAM (532 nm). The BET surface area of the samples were evaluated from a N₂ adsorption isotherm at 77 K with a BET analyzer (JW-BK122W, Beijing JWGB Sci. & Tech. Co., Ltd.).

3. Results and discussion

3.1. Material characteristics

The surface morphologies of the synthetic nanocomposite samples were observed by FE-SEM. Fig. 2(a) shows SEM images of as-prepared nanocomposites synthesized by the method of fast solvent induced assembly. It also presents a spherical morphology, where the sizes range from a few hundred nanometers to a few micrometers. A representative nanocomposite spherical assembly is presented in Fig. 2(b), shown in detail in Fig. 2(c) on an enlarged scale. The inset of Fig. 2(c), which is a high-magnification SEM image of the circled area, shows clearly that the TiO₂ nanoparticles are threaded by MWCNTs. Fig. 2(d) exhibits the transmission electron microscopic (TEM) image of TiO₂ nanoparticles. As shown in the high-resolution TEM image (HRTEM, inset in Fig. 2(d)), we know that the TiO₂ nanoparticles are highly crystallized.^27,28 A part of the typical nanocomposite spherical assemblies mentioned above is shown in Fig. 2(e), and the circled area is enlarged in Fig. 2(f). Thus, it is further confirmed that the 3D structure is formed by MWCNTs threading nanocrystal TiO₂ nanoparticles.

Fig. 2 (a) SEM image of the distribution of spherical nanocomposite assemblies (circled areas). (b) SEM image of typical spherical nanocomposite assemblies with crossed structure. (c) The magnified SEM of a representative spherical nanocomposite assembly (inset is the high resolution SEM image of a circled section, showing TiO₂ threaded by MWCNTs). (d) TEM image of TiO₂ nanoparticles (inset is the HRTEM image showing the (101) lattice plane). (e) TEM image of the nanocomposite and (f) HRTEM image of a selected area of the nanocomposite (the 3D crossed structure of the nanocomposite can be clearly distinguished from the image).

Fig. 3(a) compares the PXRD spectra of TiO₂ nanoparticles, MWCNTs and the as-synthesized nanocomposite. The peaks at 25.31° (101), 37.79° (004), 48.04° (200) and 53.88° (105) (PCPDFWIN file no. 78-2486) clearly represent the anatase phase of TiO₂ with a tiny amount of the (110) and (101) peaks of the rutile phase contribution marked by asterisks (PCPDFWIN file no. 65-0190) at 27.44° and 36.09°. Hence, the main phase of the synthesized TiO₂ nanoparticles is anatase.^19,29–31 The XRD peaks for MWCNTs are expected at 26.38° (002), 42.22° (100), 44.39° (101) and 77.26° (110) (PCPDFWIN file no. 41-1487) and match very well with the literature.^32,33 The XRD pattern of the nanocomposite in Fig. 3(a) reveals the presence of both TiO₂ and MWCNTs. Other peaks are not apparent, but can also be clearly illustrated by Raman analysis and will be discussed later. It also suggests that the reaction method leads to a real interaction across the interface, rather than the formation of a simple physical mixture. The EDX analysis in Fig. 3(b)–(d) shows that no other elements are detectable in the nanocomposite. In addition, the obtained MWCNTs/TiO₂ nanocomposite was analyzed by EDX mapping for the distribution of various elements, shown in Fig. S4 (ESI†).

Fig. 3 (a) XRD patterns of the nanocomposite, MWCNTs and anatase. EDX spectra of (b) the nanocomposite (c) anatase (carbon: sample holder line) (d) MWCNTs.

Raman spectroscopy is an effective method for characterizing the structural change of materials.¹⁹ The obtained Raman spectra of the nanocomposite and MWCNTs demonstrate two significant peaks shown in Fig. 4(a) and (b). The peaks at 1341 cm⁻¹ and 1574 cm⁻¹ correspond to peak D (disordered carbon) and peak G (graphitic carbon).^12,19 The relative intensity ratio of I_D/I_G, shows the degree of graphitization, found to be 1.03 and 0.91 for the nanocomposite and MWCNTs, respectively.³⁴ Obviously, with the addition of TiO₂ in the MWCNTs, a higher degree of graphitization than pure MWCNTs is exhibited. The results can also be attributed to some surface imperfections or defects associated with graphitic MWCNTs, resulting in the degree of disorder.¹⁹ The pore structure of the nanocomposite and the annealed MWCNTs can also be characterized by nitrogen adsorption–desorption isotherms, shown in Fig. 4(c), and the corresponding pore size distributions in Fig. 4(d). The results indicate that the nanocomposite exhibits better porous structure than the annealed MWCNTs.^35–39 The Brunauer–Emmett–Teller (BET) surface area and the Barrett–Joyner–Halenda (BJH) desorption cumulative volume of pores for the nanocomposite are 114.171 m² g⁻¹ and 0.603 cm³ g⁻¹, respectively, while the corresponding values for the annealed MWCNTs are 109.553 m² g⁻¹ and 0.546 cm³ g⁻¹.

Fig. 4 Raman spectra of (a) the nanocomposite and (b) MWCNTs. (c) Nitrogen adsorption–desorption isotherms and (d) pore-size distribution curves for the nanocomposite and annealed MWCNTs.

3.2. Electrocatalytic properties

The catalytic properties of different materials mainly impact the redox reaction on the CEs of DSSCs. In our study, three electrode cyclic voltammetry (CV) was carried out to estimate the electrocatalytic activity of various CEs for the reduction of triiodide to iodide.¹⁴ Significantly, the cathodic current density greatly relies on the scan rate of the CV measurements. Fig. 5(a) shows the CV curves of different CEs under the potential range from −0.4 V to 1.2 V with the scan rate of 50 mV s⁻¹. From the CV curves, we can observe two distinct redox pair peaks, the low potential range corresponding to the transformation from I⁻ to I₃⁻ and the high potential range corresponding to the transformation from I₃⁻ to I₂.¹² There are two critical parameters that influence the performance of CE for DSSCs. One is the peak current density and the other is the peak-to-peak separation (E_pp). Obviously, our as-prepared nanocomposite CE exhibits a higher redox peak current density and a lower E_pp value (shown in Table 1 & Fig. S5†), as compared to the MWCNTs CE.

Fig. 5 The characteristics of DSSCs with various CEs. (a) Cyclic voltammetry curves. (b) Electrochemical impedance spectroscopy (EIS) measurements. (c) Equivalent circuit for Nyquist plots. (d) Tafel polarization curves of dummy cells fabricated by nanocomposite, Pt and MWCNTs CEs.

Table 1 Photovoltaic test parameters for different CEs of DSSCs

CE	Jsc (mA cm^−2)	Voc (V)	FF	PCE (%)	Rs (Ω)	Rct1 (Ω)	Rct2 (Ω)	ZN (Ω)	Epp (V)
Composite	20.1	0.66	0.60	7.95	14.72	6.76	35.98	7.12	0.52
MWCNTs	19.0	0.65	0.55	6.82	12.98	7.74	36.43	7.38	0.63
Pt	19.7	0.66	0.57	7.38	14.05	7.14	36.10	7.24	0.60

---

Electrochemical impedance spectroscopy (EIS) is another well-known technique for characterizing the electrocatalytic performance of different CEs. The obtained Nyquist plots and the equivalent circuit are shown in Fig. 5(b) and (c), respectively. In a typical Nyquist plot, the intercept of the 1st semicircle within the high frequency range (10⁶ to 10⁵ Hz) on the real axis represents the series resistance (R_s), related to the sheet resistance of FTO glass, the contact resistance of the cell and the resistance of the external circuit. The charge transfer resistance (R_ct1) obtained from the first semicircle in the high frequency range is assigned to the interface of the CE and the electrolyte. The R_ct1 values of the three as-prepared CEs are in the order of nanocomposite (6.76 Ω) < Pt (7.14 Ω) < MWCNTs (7.74 Ω), implying an inverse order of catalytic activity, which is consistent with CV analysis. In general, the values of R_s and R_ct1 are adopted to evaluate the resistance and intrinsic catalytic activity of the electrodes.^12,40 R_ct2, obtained from the second semicircle in the mid frequency range, represents the charge transfer resistance among the interfaces of photoanode, dye and electrolyte. The R_ct2 order of our three as-prepared CEs is nanocomposite (35.98 Ω) < Pt (36.10 Ω) < MWCNTs (36.43 Ω). The lower R_ct2 indicates higher electron injection efficiency on the photoanode and more dye molecules in the excited state. Due to the cross-linked structure of the nanocomposite, our as-prepared film provides a fast ion transport pathway and a suitable degree of graphitization, which is advantageous for charge transportation. In addition, the Nyquist diffusion impedance (Z_N) of the I₃⁻/I⁻ redox couple in the pores of the electrode material and the electrolyte is another important parameter for the analysis of EIS. It is acquired by the diameter of the low-frequency semicircle on the real axis, and follows the order of nanocomposite (7.12 Ω) < Pt (7.24 Ω) < MWCNTs (7.38 Ω). By comparison, the nanocomposite CE shows the lowest value among our three as-prepared CEs, which can be attributed to the superior electrical conductivity and porous structure of the nanocomposite CE.⁴¹ In addition, the suitable synergistic effect between MWCNTs and TiO₂ in the nanocomposite can obviously limit the electron transfer resistance and reduce the recombination of charge carriers inside the DSSCs, yielding a proper enhancement of the performance of the DSSCs based on the nanocomposite CE.^42–44 The EIS analysis of DSSCs is in great agreement with the cyclic voltammetry experiments. To further explore the interfacial charge-transfer properties of the I₃⁻/I⁻ redox couple on the surface of CEs, we measured the Tafel polarization characterization of the three CEs, as shown in Fig. 5(d). From the Tafel polarization curves, it is found that the integration of TiO₂ into MWCNTs has enhanced not only the exchange current density (J₀), but also the limiting diffusion current density (J_lim). It is generally acknowledged that J₀ and J_lim can be used to assess the reducing ability of I⁻ for CEs and the diffusion capacity of I₃⁻. This result is in line with CV analysis.

3.3. Photovoltaic performance

There are three pivotal parameters used to assess the photovoltaic performance of DSSCs, which are short-circuit current density (J_sc), open-circuit voltage (V_oc) and fill factor (FF). The photovoltaic parameters of our as-prepared DSSCs are shown in Table 1. Fig. 6(a) exhibits the J–V (current density vs. voltage) curves of DSSCs with different CEs under dark and standard illumination. The J–V curves of all devices were measured, with active area of 0.15 cm². Generally, the I₃⁻/I⁻ reduction reaction occurring on the interface of the electrocatalyst can be influenced by three fundamental factors, which are diffusion of electrolyte, electron transport properties, and catalytic active sites. The excellent catalytic activity is the result of the optimal combination of these three factors. The DSSC with (MWCNTs/TiO₂) nanocomposite CE has shown the highest PCE (7.95%), in contrast to DSSCs of MWCNTs (6.82%) and Pt (7.38%) CEs. The improvement is mainly attributed to the synergistic effect between MWCNTs and TiO₂ nanoparticles on ion diffusion and electrocatalysis. The hydrophobic TiO₂ nanoparticles are threaded by MWCNTs to form a 3D crossed structure with the best electrical conductivity and excellent catalytic sites, thus the power conversion efficiency is enhanced. The open-circuit voltage is related to the Fermi level of the working electrode (TiO₂ film) and the potential energy of the redox potential of the mediating agent. The highest FF value for the as-prepared nanocomposite CE is attributed to the abundant available active sites for the redox reaction of DSSCs. Fig. 6(b) shows the incident photon-to-current conversion efficiency (IPCE) data for our three as-prepared CEs. The IPCE value of the nanocomposite exceeds 50% at a broad spectral range of 400 nm to 615 nm, and reaches its maximum around 80% at the wavelength of 532 nm, which is higher than that of the CEs of Pt and MWCNTs. The IPCE spectrum fades away at the end of 800 nm, indicating a broad spectral light harvesting scope. The obtained IPCE results are in agreement with the enhanced performance of the nanocomposite CE of DSSCs.

Fig. 6 (a) Typical photocurrent voltage (J–V) curves for DSSCs under standard illumination and dark current conditions. (b) IPCE spectra for DSSCs based on three different CEs.

4. Conclusion

In conclusion, we have successfully fabricated a 3D inter-connected structure nanocomposite (MWCNTs/TiO₂) for the CEs of high performance DSSCs, using a novel method of fast solvent induced assembly. The developed 3D nanocomposite provides short pathways and mesoporous channels for charge transportation, yielding high electrocatalytic activity. The obtained PCE (7.95%) of our fabricated device is comparable to that of the DSSC with Pt CE (7.38%), and is much superior to that of the annealed MWCNTs CE (6.82%). The obtained results demonstrate that the electrocatalytic activity of MWCNTs is significantly enhanced through the addition of suitable amounts of TiO₂. Our proposed novel method is low-cost and may be extendable for other portable energy storage devices.

Acknowledgements

This study is supported by the National Natural Science Foundation of China (11374110, 11204093), the Overseas Master Program (MS2011HZKJ043), and the Fundamental Research Funds for the Central Universities (HUST: 2014TS124). Y. H. G would like to thank Prof. Zhong Lin Wang for the support of experimental facilities in WNLO of HUST.

References

1. M. Grätzel, Acc. Chem. Res., 2009, 42, 1788–1798.
2. Z. Yu, F. Li and L. Sun, Energy Environ. Sci., 2015, 8, 760–775.
3. Y. Luo, D. Li and Q. Meng, Adv. Mater., 2009, 21, 4647–4651.
4. B. He, X. Meng and Q. Tang, ACS Appl. Mater. Interfaces, 2014, 6, 4812–4818.
5. S. Yun, A. Hagfeldt and T. Ma, Adv. Mater., 2014, 26, 6210–6237.
6. K. Kakiage, Y. Aoyama, T. Yano, K. Oya, J.-i. Fujisawa and M. Hanaya, Chem. Commun., 2015, 51, 15894–15897.
7. W. Ke, G. Fang, J. Wang, P. Qin, H. Tao, H. Lei, Q. Liu, X. Dai and X. Zhao, ACS Appl. Mater. Interfaces, 2014, 6, 15959–15965.
8. F. Xu and L. Sun, Energy Environ. Sci., 2011, 4, 818–841.
9. B. Liu and E. S. Aydil, J. Am. Chem. Soc., 2009, 131, 3985–3990.
10. G. Calogero, F. Bonaccorso, O. M. Maragò, P. G. Gucciardi and G. Di Marco, Dalton Trans., 2010, 39, 2903–2909.
11. K.-C. Huang, Y.-C. Wang, R.-X. Dong, W.-C. Tsai, K.-W. Tsai, C.-C. Wang, Y.-H. Chen, R. Vittal, J.-J. Lin and K.-C. Ho, J. Mater. Chem., 2010, 20, 4067–4073.
12. W. J. Lee, E. Ramasamy, D. Y. Lee and J. S. Song, ACS Appl. Mater. Interfaces, 2009, 1, 1145–1149.
13. M. R. Al-Bahrani, W. Ahmad, H. F. Mehnane, Y. Chen, Z. Cheng and Y. Gao, Nano-Micro Lett., 2015, 7, 298–306.
14. Z. A. Akbar, J.-S. Lee, H.-I. Joh, S. Lee and S.-Y. Jang, J. Phys. Chem. C, 2015, 119, 2314–2321.
15. H. Choi, H. Kim, S. Hwang, Y. Han and M. Jeon, J. Mater. Chem., 2011, 21, 7548–7551.
16. T. N. Murakami, S. Ito, Q. Wang, M. K. Nazeeruddin, T. Bessho, I. Cesar, P. Liska, R. Humphry-Baker, P. Comte and P. Péchy, J. Electrochem. Soc., 2006, 153, A2255–A2261.
17. S. I. Cha, B. Koo, S. Seo and D. Y. Lee, J. Mater. Chem., 2010, 20, 659–662.
18. G. r. Li, F. Wang, Q. w. Jiang, X. p. Gao and P. w. Shen, Angew. Chem., Int. Ed., 2010, 49, 3653–3656.
19. S. Muduli, W. Lee, V. Dhas, S. Mujawar, M. Dubey, K. Vijayamohanan, S.-H. Han and S. Ogale, ACS Appl. Mater. Interfaces, 2009, 1, 2030–2035.
20. P. Joshi, Y. Xie, M. Ropp, D. Galipeau, S. Bailey and Q. Qiao, Energy Environ. Sci., 2009, 2, 426–429.
21. C.-T. Hsieh, B.-H. Yang and J.-Y. Lin, Carbon, 2011, 49, 3092–3097.
22. D. Macwan, P. N. Dave and S. Chaturvedi, J. Mater. Sci., 2011, 46, 3669–3686.
23. S. Peng, L. Tian, J. Liang, S. G. Mhaisalkar and S. Ramakrishna, ACS Appl. Mater. Interfaces, 2011, 4, 397–404.
24. P. Yan, X. Wang, X. Zheng, R. Li, J. Han, J. Shi, A. Li, Y. Gan and C. Li, Nano Energy, 2015, 15, 406–412.
25. K. Woan, G. Pyrgiotakis and W. Sigmund, Adv. Mater., 2009, 21, 2233–2239.
26. P. Li, J. Wu, J. Lin, M. Huang, Z. Lan and Q. Li, Electrochim. Acta, 2008, 53, 4161–4166.
27. M. M. Gui, S.-P. Chai, B.-Q. Xu and A. R. Mohamed, Sol. Energy Mater. Sol. Cells, 2014, 122, 183–189.
28. J. Jiu, S. Isoda, F. Wang and M. Adachi, J. Phys. Chem. B, 2006, 110, 2087–2092.
29. C. W. Kim, D. K. Kim, W. J. Shin, M. J. Choi, Y. S. Kang and Y. S. Kang, Nano Energy, 2015, 13, 573–581.
30. Y. Ogomi, A. Morita, S. Tsukamoto, T. Saitho, N. Fujikawa, Q. Shen, T. Toyoda, K. Yoshino, S. S. Pandey and T. Ma, J. Phys. Chem. Lett., 2014, 5, 1004–1011.
31. G. Li, C. P. Richter, R. L. Milot, L. Cai, C. A. Schmuttenmaer, R. H. Crabtree, G. W. Brudvig and V. S. Batista, Dalton Trans., 2009, 10078–10085.
32. A. L. Mihalchik, W. Ding, D. W. Porter, C. McLoughlin, D. Schwegler-Berry, J. D. Sisler, A. B. Stefaniak, B. N. Snyder-Talkington, R. Cruz-Silva and M. Terrones, Toxicology, 2015, 333, 25–36.
33. A. Cao, C. Xu, J. Liang, D. Wu and B. Wei, Chem. Phys. Lett., 2001, 344, 13–17.
34. M.-H. Yeh, L.-Y. Lin, C.-L. Sun, Y.-A. Leu, J.-T. Tsai, C.-Y. Yeh, R. Vittal and K.-C. Ho, J. Phys. Chem. C, 2014, 118, 16626–16634.
35. E. Ramasamy and J. Lee, Chem. Commun., 2010, 46, 2136–2138.
36. D. Wu, F. Zhu, J. Li, H. Dong, Q. Li, K. Jiang and D. Xu, J. Mater. Chem., 2012, 22, 11665–11671.
37. D. Wang, B. Yu, F. Zhou, C. Wang and W. Liu, Mater. Chem. Phys., 2009, 113, 602–606.
38. B. Fang, S.-Q. Fan, J. H. Kim, M.-S. Kim, M. Kim, N. K. Chaudhari, J. Ko and J.-S. Yu, Langmuir, 2010, 26, 11238–11243.
39. J. S. Jang, D. J. Ham, E. Ramasamy, J. Lee and J. S. Lee, Chem. Commun., 2010, 46, 8600–8602.
40. S. Hwang, M. Batmunkh, M. J. Nine, H. Chung and H. Jeong, ChemPhysChem, 2015, 16, 53–65.
41. S. Yun, H. Zhang, H. Pu, J. Chen, A. Hagfeldt and T. Ma, Adv. Energy Mater., 2013, 3, 1407–1412.
42. P. Sudhagar, S. Nagarajan, Y.-G. Lee, D. Song, T. Son, W. Cho, M. Heo, K. Lee, J. Won and Y. S. Kang, ACS Appl. Mater. Interfaces, 2011, 3, 1838–1843.
43. Y. Yao, G. Li, S. Ciston, R. M. Lueptow and K. A. Gray, Environ. Sci. Technol., 2008, 42, 4952–4957.
44. X. Cai, Z. Lv, H. Wu, S. Hou and D. Zou, J. Mater. Chem., 2012, 22, 9639–9644.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra06705k

---