TiO₂ nanoparticulate-wire hybrids for highly efficient solid-state dye-sensitized solar cells using SSP-PEDOTs

Abstract

TiO₂ photoanodes for I₂-free solid-state dye-sensitized solar cells (ssDSSCs) were prepared from multifunctional new TiO₂ nanostructures to enhance light harvesting and charge collection efficiency in ssDSSCs using poly(3,4-ethylenedioxythiophene)s (PEDOTs). A new type of TiO₂ paste containing TiO₂ nanowires (TNW) was prepared and successfully transformed to TiO₂ nanoparticulate-wire hybrids (TNPW) with a large surface area of 61.4 m² g⁻¹ through thermal annealing. The thickness of the TNPW layer could be controlled up to 17 μm without cracks. As a as hole transport material, PEDOTs were infiltrated into the TNPW photoanode through in situ solid-state polymerization (SSP-PEDOTs) and N719 dyes were adsorbed to give ssDSSCs. The SSP-PEDOTs based ssDSSCs with TNPW photoanodes recorded a high cell efficiency (η) of 6.4% and short-circuit current (J_sc) of 14.3 mA cm⁻² without scattering particles, which were 30.6 and 22.2% higher than those of traditional TiO₂ nanoparticles (TNP) in the same conditions. Furthermore, liquid-state DSSCs with the TNPW photoanode attained a η of 8.4%, which was superior to that of a reference TNP cell (7.3%). The maximum η values were 7.1 and 9.9% for ssDSSCs and liquid type DSSCs, respectively, in the presence of additional scattering layers to support the importance of TNPWs. These enhanced photovoltaic performances of TNPW cells could be attributed to the unique TNPW structure that is advantageous for high charge collection with a long electron diffusion path and large surface area required for high dye adsorption efficiency.

---

Introduction

Interfacial charge transport in nanostructured metal oxides for photovoltaic conversion is highly sensitive to their size, shape, and surface structure.^1–3 In particular, the charge collection efficiency of a TiO₂ photoanode in DSSCs^4–9 is deeply related to the TiO₂ nanostructure,^10,11 which could critically affect the relatively low efficiency of solid-state DSSCs (ssDSSCs). The TiO₂ nanoparticles of the traditional paste with a size of 10–20 nm are advantageous for efficient dye adsorption resulting in high photocurrent generation. However, they limit the photoconversion efficiency of a cell because of the high charge recombination resulting from electron trapping events at the interfacial grain boundary of the nanoparticles.^12–18 To reduce such charge recombination and enhance charge transport through the TiO₂ nanostructure, a photoanode with one-dimensional (1D) nanomaterials has been suggested.^19,20 Although considerable efforts have been devoted to the fabrication of 1D TiO₂ nanostructures, including nanotubes (NTs),^21–24 nanowires (NWs),^25,26 and nanorods (NRs),^27,28 the relatively low surface area of such structures limits further enhancement in the photocurrent density.

Recently, liquid type DSSCs using a double-layered photoanode have been reported using oriented hierarchical TiO₂ nanowire arrays (HNWs) and nanoparticles (NPs). Compared to an NP-based cell, the TiO₂ HNW cell showed a lower electron recombination rate.²⁹ Nevertheless, the photoconversion efficiency (PCE) using the HNW/NP double layer was low, as expected from the limited interfacial electron transport in the NP layer and low dye adsorption onto the HNW layer. Clearly, the HNW/NP double layer system cannot be applied for ssDSSCs because the efficiency of leak-free ssDSSCs is generally lower than that of the liquid type DSSCs. Therefore, an important challenge that remains to be solved is the conflicting shortcoming between the surface area required for highly efficient dye adsorption and the 1D nanostructures necessary for enhanced electron transport.

To enhance not only electron transport but also the efficiency of dye adsorption in 1D nanostructures and ssDSSCs, we propose a facile synthesis of a new TiO₂ nanostructure with an enhanced electron diffusion length, large surface area, and light-harvesting property. Compared to other complicated methods,^30,31 including electrochemical lithography, photoelectrochemical etching, and template synthesis, we explore a paste coating process for the synthesis of 1D metal oxide nanostructures because it allows the precise controllability of the structure within a short reaction time.

The ssDSSCs in this work is based on a I₂-free, solid-state conjugated polymer as a hole transporting material, which is grown in situ inside TiO₂ nanopores, as reported previously.^32,33 Because the TiO₂ photoanode based on commercial TiO₂ nanoparticles showed limitations in increasing short-circuit current density (J_sc), open-circuit voltage (V_oc), fill factor (FF), and efficiency (η) of ssDSSCs due to the intrinsic problems of spherical TiO₂ nanoparticles, we explored a new TiO₂ nanostructure for high interfacial charge-collection capacity. Herein, we report the large-scale synthesis of the TNW paste and of an efficient nanostructure of TiO₂ nanoparticulate-wire hybrids (TNPW) as a high performance photoanode for ssDSSCs, possessing high charge collection capacity and efficient light harvesting.

Experimental

Materials

Ethanol, 1-butanol, isopropyl alcohol, acetonitrile, 1,2-dimethyl-3-propylimidazolium iodide, iodine, tert-butylpyridine, lithium iodide, lithium acetate dihydrate (LiAc·2H₂O), acetic acid, anhydrous terpineol, ethyl cellulose, titanium(IV) butoxide, N,N-dimethylformamide (DMF), and titanium bis(ethyl acetoacetate) were purchased from Aldrich. The TiO₂ paste (Dyesol 18NR-T) was acquired from Dyesol, LTD. (Australia) and N719 dye was purchased from Solaronix (Switzerland). All the solvents and chemicals were reagent grade and were used as received. DBEDOT (2,5-dibromo-3,4-ethylenedioxythiophene) was prepared according to the procedure from our previous report.³²

Synthesis of TiO₂ nanowires

TiO₂ nanowires were synthesized by a solvothermal method. Thus, 0.2 g of lithium acetate dihydrate and 2 mL of titanium(IV) butoxide (TB) were dissolved in 10 mL of solvent mixture consisting of DMF and acetic acid. The solution was transferred into a Teflon lined stainless steel autoclave, and heated in an oven at 200 °C for 20 h. Finally, the 1D nanowires were collected and thoroughly washed with ethanol several times, and dried overnight in a vacuum oven at 60 °C.

Preparation of TiO₂ paste and TNPW photoanode

The as-prepared TiO₂ nanowires (5 g) and acetic acid (1 mL) were mixed for 5 min and the mixture was grinded for 5 min after adding 5 mL of deionized water. All the liquid reagents were added drop by drop into the grinder. Then, 30 mL of ethanol was added and thoroughly mixed. Excess ethanol (100 mL) was added to the mixture, which was stirred at a rate of 500 rpm. Anhydrous terpineol (17 g) and 28 g of ethyl cellulose solution (10 wt% in ethanol) were added to the above mentioned mixture. After stirring and sonication, ethanol was removed using a rotary evaporator to give TiO₂ nanowire paste. The prepared paste was cast onto a compact TiO₂ layer-coated FTO glass using a doctor-blade technique and dried at 70 °C for 1 h, followed by successive sintering at 450 °C for 30 min and cooling to 30 °C for 8 h. The above mentioned conventional compact TiO₂ layer, with a 200 nm thickness, was prepared by spin coating a titanium bis(ethyl acetoacetate) solution (2 wt% in butanol) onto FTO (Pilkington, 8 Ω⁻¹) glass at 2000 rpm for 30 s, followed by calcination at 450 °C for 30 min. Separately, a commercially available TiO₂ paste (Dyesol 18NR-T) was used to prepare TNP photoanode by the same method for comparison.

Device fabrication

The prepared photoanodes were immersed in a dye solution containing N719 (0.3 mM in ethanol) for 24 h at room temperature. The Pt counter electrodes were prepared by drop-casting a H₂PtCl₆ solution (7 mM in isopropyl alcohol) onto a conductive FTO, followed by heating at 400 °C for 20 min and cooling to 30 °C for 8 h. The dye adsorbed TiO₂ electrodes and the Pt counter electrodes were facing each other in a sandwich-type cell using a hot-melt film (Surlyn, 25 μm) as spacer. The ssDSSCs were fabricated by drop-casting the HTM solution onto the photoanode and covering it with a Pt-coated counter electrode. DBEDOT was dissolved in ethanol. First, a few drops of dilute solution (1 wt%) of DBEDOT were dropped onto the TiO₂ photoanodes and dried under ambient conditions. Then, a few additional drops of a more concentrated solution (3 wt%) of DBEDOT in ethanol were directly cast onto the above mentioned photoanodes. After evaporating the solvent, the DBEDOT-incorporated photoanodes were thermally polymerized at 55 °C for 24 h in an oven to produce highly conductive polymer channels (Fig. 1c). A drop of the acetonitrile solution consisting of 1-methyl-3-propyl-imidazolium iodide (1.0 M), 4-tert-butylpyridine (0.2 M) and lithium bis(trifluoromethane)sulfonimide (0.2 M) was coated onto the photoanode as an electrolyte and dried before the assembly of ssDSSCs with the Pt counter electrode (Fig. 1c). For the preparation of liquid type DSSCs, the acetonitrile solution containing 0.6 M 1,2-dimethyl-3-propylimidazolium iodide, 0.1 M iodine, 0.5 M tert-butylpyridine, and 0.1 M lithium iodide was injected into the cell through a hole in the counter electrode. The active area of the cell was fixed at 0.16 cm².

Fig. 1 (a) Photographic images of the coated and calcinated photoanodes on FTO using a commercial TiO₂ nanoparticle paste (top) and TNWs paste (bottom). (b) The chemical structures of DBEDOT and solid-state-polymerized PEDOTs with 10 S cm⁻¹. (c) Schematic representation of the ssDSSCs structure that was fabricated with the TNPW/N719 dye photoanode.

Characterization

The morphology of TiO₂ was observed by a field emission scanning electron microscope (FE-SEM, JEOL Ltd, model JSM-7001F) and a transmission electron microscope (TEM, JEOL Ltd, model JSM-2010). X-ray diffraction (XRD) measurements were carried out using a Rigaku (Ultima IV) wide-angle goniometer using Cu-Kα radiation. The optical properties (transmittance and reflectance) of the TiO₂ film and the concentration of adsorbed dye were determined by using a UV/Vis spectrophotometer (PerkinElmer, Lambda 750). Photoelectrochemical and photovoltaic characteristics of the TNP and TNPW cells were determined using an electrochemical workstation (Keithley Model 2400) and a solar simulator (1000 W xenon lamp, Oriel, 91193). Light exposure was homogeneous across a 8 × 8 inch² area and was calibrated with a Si solar cell (Fraunhofer Institute for Solar Energy Systems, Mono-Si + KG filter, Certificate no. C-ISE269) to a sunlight intensity of 1 (100 mW cm⁻²). This calibration was confirmed with a NREL-calibrated Si solar cell (PV Measurements Inc.). IMVS and IMPS measurements were carried out on an electrochemical workstation equipped with a frequency response analyzer under a modulated green light-emitting diode (535 nm) driven by a source supply, which could provide both DC and AC illumination components (frequency range: from 10 000 to 0.01 Hz). IPCE spectra were obtained as a function of wavelength from 300 to 800 nm. The amount of adsorbed dye was determined by measuring dye desorption after the immersion of the dye adsorbed TiO₂ film into a solution of NaOH (5 mM) in an ethanol–water (1 : 2 v/v) mixture. The concentration of the desorbed dye was analyzed by using a UV/Vis spectrophotometer.

Results and discussion

A superior photoanode with TNPW was prepared by thermal treatment from a paste composed of TiO₂ nanowires (TNWs). The TNWs were synthesized by a simple solvothermal reaction in DMF and acetic acid according to a previously described method³⁴ (Fig. 1a). The pristine TNWs were obtained in a large scale with an average diameter of ∼15 nm and several hundred nanometers in length, as shown in Fig. 2a and b. A new paste containing the TNWs was prepared as a viscous yellowish solution, which was coated onto a FTO substrate that was pre-coated with a 200 nm thick compact TiO₂ layer.^32,35 Upon calcination at ∼450 °C for 30 min, the thin TNWs were aggregated and popped-up to yield TNPW, as shown in Fig. 2c and g. On the other hand, the nanowire structures were maintained when calcinated under mild conditions of 350 °C for 2 h. The length and diameter of the TNPW were approximately 200 nm and 40 nm, respectively, which were larger than those of the TiO₂ nanoparticles from the initial paste.

Fig. 2 SEM images of TiO₂ nanostructures: (a) after synthesis, (b) after annealing at 350 °C for 2 h, and (c) after calcination of the TNWs paste at 450 °C. TEM image of (d) and (e) TNWs, (g) and (h) TNPW and magnified HR-TEM image of (f) TNWs, (i) TNPW corresponding Fourier transform (FFT) pattern (inset).

The reformation of TNPW could be attributed to the aggregation and shrinkage of TNW bundle at high temperatures (450 °C), as schematically shown in Fig. S1a.† It is noteworthy that a 17 μm thick, large-area (3 cm × 2 cm) TNPW photoanode could be prepared without cracks from the TNWs paste. Moreover, the TNPW photoanodes showed a dense and well-organized mesoporous structure, as shown in Fig. S2.† The high resolution TEM (HR-TEM) images for the pristine TNWs (Fig. 2d–i) show a typical continuous lattice fringe of anatase TiO₂ with lattice spacings of 0.35 nm, which corresponded to the (101) crystalline plane. The phase purity and structure of pristine TNWs were examined further by X-ray diffraction (XRD) analysis in the 2θ range of 20° to 60° (Fig. S1b†). The sharp peaks at 25.3, 36.9, 37.8, 38.5, 47.9, 53.9, and 55° correspond to reflections from the (101), (103), (104), (112), (200), (105), and (211) crystal planes of anatase TiO₂, respectively (JCPDS no. 21-1272).

The lattice spacing of the (101) crystalline plane for the TNWs in the HR-TEM image (Fig. 2f) was well matched to that from XRD. The TNPW were shorter in length but thicker than TNW bundles, as compared in Fig. 2. Importantly, the continuous fringes that corresponded to the (101) planes in TNWs (Fig. 2f) were observed in TNPW (Fig. 2i). Moreover, TNPW showed the same XRD diffraction pattern as that of TNWs (Fig. S1b†), indicating that TNPW possess the crystal planes of anatase TiO₂ despite the high temperature treatment. This result suggests that the TNPW layer could be promising as a photoanode because of the popped-up particle structures that increase surface area and the tightly connected 1D TiO₂ structures that are useful for the extension of the electron transport path. Light harvesting is one of the important property to improve the cell efficiency in solar cell applications by maximizing the absorption of incident sunlight.³⁶ In view of a light harvesting technique, several conditions like light scattering by larger nanoparticles (∼100 nm),³⁷ light-trapping nanopatterned surfaces,³³ and plasmon light harvesting^38,39 are commonly fulfilled for the fabrication of solar cells. In particular, it has been necessary to add light-scattering layers into DSSCs. This requires extra engineering for the paste composition, such as careful control of the composition as well as distribution of the scattering particles, given that light scattering depends strongly on the size of the particles and the wavelength of the incident light. Because the size of the TNPW (∼200 nm long and 40 nm wide) is larger than that of the common TNP, the photoelectrode of the former exhibits a significantly higher reflectance compared to that of TNP, as illustrated in Fig. 3b and 1b. The high optical reflectance of TNPW was clearly observed in the visible-light wavelength range; it was 2-fold larger than that of the TNP for the entire visible range of 400 to 700 nm. The increased reflectance was sufficiently large to enhance light harvesting, improving the photocurrent density of DSSCs. Because of their aggregated 1D structures, the TNPW photoanodes exhibit a slightly reduced dye adsorption than the TNP photoanodes, as expected from the smaller surface area of the TNPW (Table S1†). On the other hand, the light-harvesting properties of the TNPW photoanode were improved compared with those of the TNP, as described below. Additionally, V_oc and FF of TNPW are higher than those of TNP. This enhancement could be originating from the molecular contact between the hole transport PEDOTs and the dye-adsorbed TNPW, which are long enough for extended charge transport and large enough for effective light harvesting. The reflectance of TNPW was much greater than that of TNP. Thus, the TNPW was more effective in utilizing the light absorption of dyes in a photoanode through enhanced light-scattering effects in visible regions, resulting in the improvement of current density. This result was consistent with the IPCE data. The IPCE spectra in Fig. 4b show that the quantum efficiency was improved in the visible wavelength region, mainly due to the light-harvesting property of TNPW. For TNPW with PEDOTs, the cell efficiency was enhanced by 30.6% probably due to the increase in J_sc, according to the reflectance increase. Furthermore, the recombination of electrons was well prevented due to the long length of TNPW, which could provide good interconnectivity and deep penetration of DBEDOT. Efficiency was increased up to 6.4% with TNPW, which is among the highest values for N719-dye-based ssDSSCs, and was greater than that of ssDSSCs with TNP.⁴⁰ The electron transport and charge recombination time for DSSCs were estimated from the intensity-modulated photocurrent spectroscopy (IMPS)/intensity-modulated photovoltage spectroscopy (IMVS) (Fig. 5c and d), according to eqn (1) and (2):

τ_d = 1/2πf_d:min (1)

τ_r = 1/2πf_r:min (2)

where f_d(f_r) is the characteristic frequency at minimum IMPS and IMVS imaginary components, and τ_d(τ_r) is the electron transport time (recombination time).⁴¹ The electron lifetime (τ_r) of TNPW is always longer than those of the TNP cells (Fig. 5c). Such a long τ_r of TNPWs can be attributed to the their unique structures (Fig. 2c and g–i). Inside the shape-transformed structures, the aggregated and shrunk TNWs are extensively interconnected, forming long charge transport paths (Fig. 5b and S1a†). This could reduce the charge recombination at the grain boundary and interface, compared to the spherical nanoparticles (diameter of ∼20 nm) with many gaps between the particles. Moreover, TNPWs possess a smaller surface area than TNPs, reducing the interfacial electron trapping sites for recombination with I³⁻ in the electrolyte. As shown in Fig. 4 and Table 1, the J–V curve of the ssDSSC fabricated with TNPWs and SSP-PEDOTs showed a J_sc of 14.3 mA cm⁻² and a η of 6.4%, which were much greater than those of the ssDSSCs with TNP.³² As a result, V_oc is higher for the TNPW than for the TNP (Table 1). This provides evidence that electrons in the TNPW are hindered from interfacial electron recombination with redox species, yielding a high V_oc. Furthermore, the electron diffusion coefficient (D_n) of the TNPW was higher than that of the TNP, as determined from eqn (3):

D_n = d²/(2.35τ_d) (3)

where d is the thickness of the photoanode.⁴² This indicates a fast electron transport in TNPW because of their well-organized 1D nanostructure (Fig. 5d). This, in turn, enhances the charge-collection capacity, resulting in an increase of the J_sc at the same dye adsorption level as the TNP. The effective electron-diffusion length (L_n) was determined from D_n and τ_r using eqn (4) as follows:

L_n = (D_nτ_r)^1/2 (4)

where L_n is the average path distance of the injected electrons through the photoanode before recombination.⁴¹ The diffusion length of the TNPW was longer than that of the TNP, and therefore, led to a higher charge collection efficiency and enhanced current density (Fig. S3†). With such high D_n, τ_r, and superior light scattering property, the TNPW photoanodes provide high open-circuit voltage, short-circuit current, and photoconversion efficiency despite the low dye loading of the TNPW cell compared to the TNP cell. Moreover, the TNPW cell showed a higher photoconversion efficiency than the TNWs/TNP double layer system,²⁹ possibly due to their unique structure prone to fast charge transport and to their light scattering structure. The TNPW application potential was examined in ssDSSCs containing a scattering layer (SL) and even in a liquid type DSSC. As shown in Fig. 4 and Table 1, the J–V curve of the ssDSSC fabricated with SL and SSP-PEDOTs showed a J_sc of 15.8 mA cm⁻² and a η of 7.1%, which were considerably greater than those of the ssDSSCs with TNP.^32,33 Furthermore, the liquid type DSSC with TNPW and SL showed a J_sc of 18.2 mA cm⁻² and a η of 9.9%. These results strongly indicate that the TNPW is useful not only in solid type DSSCs but also in liquid type cells.

Fig. 3 Optical properties of TNP and TNPW photoanodes: (a) comparison of transmittance spectra of bare FTO glass (black), TNP (red), and TNPW (blue) film on FTO glass, (b) reflectance spectra of bare FTO, TNP and TNPW films on the FTO substrate with a TiO₂ thickness of 10 μm for all the films. Inset: schematic representation of light trapping by scattering on the TNP and TNPW film.

Fig. 4 Cell performances and IPCE results: (a) J–V curves and (b) IPCE for ssDSSCs (filled) and liquid DSSCs (empty) with TNP (red) and TNPW (blue) without a scattering layer at 100 mW cm⁻². (c) J–V curves of DSSCs with the addition of a scattering layer at 100 mW cm⁻².

Fig. 5 (a) The electron lifetime (τ_r), (b) diffusion coefficient (D_n), and (c) electron diffusion length of ssDSSCs (filled) and liquid type DSSCs (empty) fabricated using TNP (red) and TNPW (blue), as determined from the IMPS/IMVS measurements.

Table 1 Performances of the DSSCs fabricated with various semiconductor nanoparticles at 100 mW cm⁻²

Sample^a	Thickness	HTM	Jsc (mA cm^−2)		Voc (V)	FF	η (%)	Dye loading^c (nmol cm^−2)
a The devices were fabricated with a transparent TiO2 layer without a scattering layer. The cells were investigated with a metal mask with an area of 0.16 cm^2.b The I2-free ssDSSCsdevices were fabricated with conducting polymer (PEDOTs) as HTM.c I2 based DSSCs. All devices contained theN719 dye.d The devices were fabricated with a scattering layer. Note: TNP devices were prepared from the commercially available Dyesol paste.
TNP-S^b	10 μm	PEDOTs	11.7	0.65		0.65	4.9	71.4
TNPW-S^b	10 μm	PEDOTs	14.3	0.66		0.67	6.4	60.8
TNP-L^c	11 μm	I^−/I^3−	14.1	0.73		0.72	7.3	78.3
TNPW-L^c	11 μm	I^−/I^3−	15.6	0.74		0.73	8.4	67.1
TNP-S-SL^b^,^d	8 + 3 μm	PEDOTs	13.5	0.66		0.65	5.8	61.2
TNPW-S-SL^b^,^d	8 + 3 μm	PEDOTs	15.8	0.67		0.67	7.1	54.1
TNP-L-SL^c^,^d	11 + 4 μm	I^−/I^3−	16.6	0.77		0.69	8.9	78.9
TNPW-L-SL^c^,^d	11 + 4 μm	I^−/I^3−	18.2	0.77		0.70	9.9	68.3
(TNP-S/ref. 43)	—	PPP-b-P3HT	8.8	0.81		0.65	4.65	—
(TNP-S/ref. 32)	11 μm	PEDOTs	14.2	0.64		0.60	5.4	—
(TNP-S/pat/ref. 33)	11 μm	PEDOTs	19.2	0.65		0.56	7.03	—

---

Conclusions

Shape-transformable TiO₂ nanowire (TNW) semiconductors were synthesized in large scale and used as a new paste to yield superior TNPW photoanodes with well-connected popped-up 1D structure without cracks, even for a 17 μm thick film. The TNPW photoanodes had a relatively high surface area (61.4 m² g⁻¹), which was comparable to the TNP. Their combined characteristics of fast charge transport and light-harvesting structures provided ssDSSCs of N719 with high V_oc (0.66 V) and J_sc (14.3 mA). This resulted in a high η of 6.4%, representing a 30.6% improvement in efficiency compared to the TNP of the same thickness, respectively. Given the prominent properties of TNPW, the η of a TNPW liquid-state cell reached 8.4%, which was higher than that of TNP at the same thickness. In the presence of scattering layers, the ssDSSCs showed a maximum η of 7.1%, indicating that TNPW photoanode could be further optimized by the addition of a SL. Moreover, the TNPW photoanode could be applied in liquid type DSSCs, to obtain a η of 9.9%. These high efficiency values could be attributed to the enhanced J_sc and V_oc values originating from the interconnected TNPW structure with large pores and high porosity, which were effective for both high dye adsorption and long charge transport. The new type of TNPW semiconductor structure with a relatively high surface area, efficient light harvesting, and long electron diffusion paths can be a possible candidate for quantum dot and perovskite solar cells and for other optoelectronic devices with high performance in the near future.

Acknowledgements

We acknowledge the financial support of the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIP) (no. 2007-0056091).※ MSIP: Ministry of Science, ICT & Future Planning.

Notes and references

1. J.-G. Yu, H.-G. Yu, B. Cheng, X.-J. Zhao, J. C. Yu and W.-K. Ho, J. Phys. Chem. B, 2003, 107, 13871–13879.
2. Z.-H. Lin, Y. Xie, Y. Yang, S. Wang, G. Zhu and Z. L. Wang, ACS Nano, 2013, 7, 4554–4560.
3. S. Seo, C. Allain, J. Na, S. Kim, X. Yang, C. Park, J. Malinge, P. Audebert and E. Kim, Nanoscale, 2013, 5, 7321–7327.
4. B. Oregan and M. Gratzel, Nature, 1991, 353, 737–740.
5. L. Passoni, F. Ghods, P. Docampo, A. Abrusci, J. Martí-Rujas, M. Ghidelli, G. Divitini, C. Ducati, M. Binda, S. Guarnera, A. Li Bassi, C. S. Casari, H. J. Snaith, A. Petrozza and F. Di Fonzo, ACS Nano, 2013, 7, 10023–10031.
6. M. Stefik, F. J. Heiligtag, M. Niederberger and M. Grätzel, ACS Nano, 2013, 7, 8981–8989.
7. P.-Y. Chen, X. Dang, M. T. Klug, J. Qi, N.-M. Dorval Courchesne, F. J. Burpo, N. Fang, P. T. Hammond and A. M. Belcher, ACS Nano, 2013, 7, 6563–6574.
8. P.-Y. Chen, R. Ladewski, R. Miller, X. Dang, J. Qi, F. Liau, A. M. Belcher and P. T. Hammond, J. Mater. Chem. A, 2013, 1, 2217–2224.
9. G. M. Lowman and P. T. Hammond, Small, 2005, 1, 1070–1073.
10. J. Nissfolk, K. Fredin, A. Hagfeldt and G. Boschloo, J. Phys. Chem. B, 2006, 110, 17715–17718.
11. M. Grätzel, Inorg. Chem., 2005, 44, 6841–6851.
12. N. Kopidakis, K. D. Benkstein, J. van de Lagemaat and A. J. Frank, J. Phys. Chem. B, 2003, 107, 11307–11315.
13. A. Solbrand, A. Henningsson, S. Södergren, H. Lindström, A. Hagfeldt and S.-E. Lindquist, J. Phys. Chem. B, 1999, 103, 1078–1083.
14. D. K. Roh, W. S. Chi, S. H. Ahn, H. Jeon and J. H. Kim, ChemSusChem, 2013, 6, 1384–1391.
15. J. Qu, G. R. Li and X. P. Gao, Energy Environ. Sci., 2010, 3, 2003–2009.
16. M. J. Bierman and S. Jin, Energy Environ. Sci., 2009, 2, 1050–1059.
17. L. Hu, S. Dai, J. Weng, S. Xiao, Y. Sui, Y. Huang, S. Chen, F. Kong, X. Pan, L. Liang and K. Wang, J. Phys. Chem. B, 2007, 111, 358–362.
18. T. Berger, T. Lana-Villarreal, D. Monllor-Satoca and R. Gómez, J. Phys. Chem. C, 2007, 111, 9936–9942.
19. F. Shao, J. Sun, L. Gao, S. Yang and J. Luo, J. Mater. Chem., 2012, 22, 6824–6830.
20. J.-Y. Liao, B.-X. Lei, H.-Y. Chen, D.-B. Kuang and C.-Y. Su, Energy Environ. Sci., 2012, 5, 5750–5757.
21. M. Ye, D. Zheng, M. Lv, C. Chen, C. Lin and Z. Lin, Adv. Mater., 2013, 25, 3039–3044.
22. T.-S. Kang, A. P. Smith, B. E. Taylor and M. F. Durstock, Nano Lett., 2009, 9, 601–606.
23. D. K. Roh, W. S. Chi, H. Jeon, S. J. Kim and J. H. Kim, Adv. Funct. Mater., 2014, 24, 379–386.
24. T. Stergiopoulos, E. Rozi, R. Hahn, P. Schmuki and P. Falaras, Adv. Energy Mater., 2011, 1, 569–572.
25. X. Feng, K. Zhu, A. J. Frank, C. A. Grimes and T. E. Mallouk, Angew. Chem., 2012, 124, 2781–2784.
26. X. Feng, K. Shankar, O. K. Varghese, M. Paulose, T. J. Latempa and C. A. Grimes, Nano Lett., 2008, 8, 3781–3786.
27. L. De Marco, M. Manca, R. Giannuzzi, M. R. Belviso, P. D. Cozzoli and G. Gigli, Energy Environ. Sci., 2013, 6, 1791–1795.
28. B. Liu and E. S. Aydil, J. Am. Chem. Soc., 2009, 131, 3985–3990.
29. W.-Q. Wu, J.-Y. Liao, H.-Y. Chen, X.-Y. Yu, C.-Y. Su and D.-B. Kuang, J. Mater. Chem., 2012, 22, 18057–18062.
30. S. Z. Chu, S. Inoue, K. Wada, S. Hishita and K. Kurashima, J. Electrochem. Soc., 2005, 152, B116–B124.
31. P. Hoyer, Langmuir, 1996, 12, 1411–1413.
32. J. K. Koh, J. Kim, B. Kim, J. H. Kim and E. Kim, Adv. Mater., 2011, 23, 1641–1646.
33. J. Kim, J. K. Koh, B. Kim, J. H. Kim and E. Kim, Angew. Chem., Int. Ed., 2012, 51, 6864–6869.
34. H. B. Wu, H. H. Hng and X. W. Lou, Adv. Mater., 2012, 24, 2567–2571.
35. J. Kim, J. K. Koh, B. Kim, S. H. Ahn, H. Ahn, D. Y. Ryu, J. H. Kim and E. Kim, Adv. Funct. Mater., 2011, 21, 4633–4639.
36. M. D. Kelzenberg, S. W. Boettcher, J. A. Petykiewicz, D. B. Turner-Evans, M. C. Putnam, E. L. Warren, J. M. Spurgeon, R. M. Briggs, N. S. Lewis and H. A. Atwater, Nat. Mater., 2010, 9, 239–244.
37. J.-W. Shiu, C.-M. Lan, Y.-C. Chang, H.-P. Wu, W.-K. Huang and E. W.-G. Diau, ACS Nano, 2012, 6, 10862–10873.
38. J. Qi, X. Dang, P. T. Hammond and A. M. Belcher, ACS Nano, 2011, 5, 7108–7116.
39. X. Dang, J. Qi, M. T. Klug, P.-Y. Chen, D. S. Yun, N. X. Fang, P. T. Hammond and A. M. Belcher, Nano Lett., 2013, 13, 637–642.
40. S. Hore, P. Nitz, C. Vetter, C. Prahl, M. Niggemann and R. Kern, Chem. Commun., 2005, 15, 2011–2013.
41. K. Yan, Y. Qiu, W. Chen, M. Zhang and S. Yang, Energy Environ. Sci., 2011, 4, 2168–2176.
42. L. Dloczik, O. Ileperuma, I. Lauermann, L. M. Peter, E. A. Ponomarev, G. Redmond, N. J. Shaw and I. Uhlendorf, J. Phys. Chem. B, 1997, 101, 10281–10289.
43. W. Chen, Y. Lee, C. Chen, K. Kau, L. Lin, C. Dai, C. Wu, K. Ho, J. Wang and L. Wang, ACS Nano, 2014, 8, 1254–1262.

---

Footnotes

† Electronic supplementary information (ESI) available: Detailed data containing SEM images of the TNPW film, surface area, pore volume and pore diameter results of devices. See DOI: 10.1039/c4ra08583c

‡ J. Na and J. Kim contributed equally to this work.

---