High-performance flexible dye-sensitized solar cells by using hierarchical anatase TiO₂ nanowire arrays

Abstract

Pioneering products in wearable technologies, such as Apple Watch, Google Glass or Nike smart sneakers, are leading a new trend in electronic devices. However, the fast developing field of wearable electronics urgently demands lightweight and flexible/bendable energy supplying devices. Here, we present thin, lightweight and flexible dye-sensitized solar cells (DSSCs) using three different kinds of TiO₂ nanostructures grown on Ti meshes as photoanodes. The flexible DSSCs based on hierarchical anatase TiO₂ nanowire arrays exhibit a higher power conversion efficiency (1.76%) than those based on TiO₂ nanotube arrays (1.52%) or TiO₂ nanowire arrays (0.85%). This enhancement can be attributed to (1) the large surface area and (2) the low recombination rate of electrons and the redox electrolyte. Moreover, commercial ink has been identified as a promising alternative to fabricate a flexible counter electrode and the corresponding DSSCs achieve a considerable conversion efficiency of 1.59%. The as-fabricated mesh based DSSC retains 88% of its PCE after bending for 300 cycles, demonstrating its good flexibility and fatigue resistance. In addition, the as-fabricated sealed DSSCs have been integrated to light LEDs, which demonstrates that our flexible DSSCs are promising candidates for wearable/flexible energy supplying devices in everyday applications.

---

Introduction

Flexible and wearable electronic devices are becoming predominant in modern electronics.^1–3 Given this, lightweight and flexible energy supplying devices, such as flexible lithium-ion batteries, supercapacitors or solar cells, have attracted extensive scientific and social research interest in recent years.^4–10 Compared with other kinds of solar cells, dye-sensitized solar cells (DSSCs) have many advantages, including low manufacturing costs, great stability and high power conversion efficiency (PCE).^11–16 Flexible DSSCs have been previously studied using conductive plastic films,^17–20 metal wire and foil,^21–26 carbon fibers²⁷ or commercial thread²⁸ as the flexible substrate. Among them, indium-doped tin oxide/poly(ethylene terephthalate) (ITO/PET) is one of the most popular candidates for a photoanode substrate.²⁰ However, the TiO₂ photoanode films have to be annealed at under only 150 °C due to the heat resistance limitation of the ITO/PET substrate, which results in low crystallization and poorly interconnected TiO₂ nanoparticles (NPs), leading to a low PCE. For the requirements of good electrical conductivity and heat-resistance ability, metal foils have been developed as substitutes for flexible DSSCs. As is well known, metal foils are opaque and a back-side illumination is needed, therefore this structure suffers energy loss while light passes through the counter electrode and redox electrolyte. Besides, the low flexibility of metal foils also restricts the application of flexible DSSCs. Therefore, a novel fiber-shaped DSSC using metal wire as the substrate, which has advantages including being lightweight, easy encapsulation and good flexibility, has been proposed.^29–32 The fiber-shaped DSSCs are generally assembled into a double wire structure: one wire coated with a nanostructured TiO₂ film as the working electrode, while another wire (e.g., Pt) used as the counter electrode. The fiber-shaped DSSCs can be woven into flexible solar cell textiles for practical applications. However, fiber-shaped DSSCs are still facing many challenges, such as connecting devices in serial/parallel and large scale weaving.

Recently, several attempts have been made to produce flexible/bendable DSSC textiles based on woven metal mesh electrodes.^33–35 Compared to other types of flexible DSSCs, mesh based DSSCs exhibit competitive superiority since they possess excellent heat-durability, great transparency and flexibility. Generally, micrometer-sized metal wires are woven together to form the mesh body. TiO₂ nanotube (NT) arrays or NP films are then coated on the metal wire surface, which is used as the working electrode and assembled with the mesh counter electrode to form the desired flexible DSSC. Although some previous works have demonstrated that mesh based DSSCs have been fabricated and achieved considerable PCE, they mostly lack the complete seal of devices and are far from practical applications. Moreover, many challenges remain to fabricate mesh based flexible DSSCs, such as producing low-cost, high-performance, bending-resistant, long-term-stable photoanodes and counter electrodes.

Herein, thin and bendable mesh based DSSCs are developed by using hierarchical anatase TiO₂ nanowire (NW) arrays as the photoanode. A two-step hydrothermal method is used to synthesize the TiO₂ hierarchical nanowire (HNW) arrays, which achieve higher PCE within DSSCs than TiO₂ NT arrays and TiO₂ NW arrays. This improvement due to the large surface area and low recombination rate was investigated in detail. The fabricated flexible DSSCs have excellent bending resistance and can be integrated to light up a group of LEDs, which demonstrates that the mesh based DSSC is a promising candidate for wearable energy supply devices.

Experimental

Pre-treatment

The Ti meshes (diameter: 100 μm, size: 100 meshes) were purchased from Sheng Zhuo Mesh Products Co., Ltd. The surface of the Ti wire was coated with graphite, which was used as a lubricant for weaving into the square Ti mesh, which was ultrasonically washed with concentrated hydrochloric acid to remove the graphite layer. Then, the Ti meshes were ultrasonically cleaned in acetone, ethanol and deionized water in sequence.

Fabrication of highly aligned TiO₂ NT arrays on the Ti mesh

The Ti mesh was electrochemically anodized at a constant potential of 60 V in an ethylene glycol solution containing 0.3 wt% NH₄F and 2 vol% deionized water for 6 h. The Ti mesh and Pt wire were used as the working electrode and counter electrode, respectively. The as-prepared samples were annealed in air at 500 °C for 30 min and then soaked in a 0.1 M aqueous TiCl₄ solution at 70 °C for 30 min, followed by annealing at 450 °C for 30 min.

Fabrication of vertically TiO₂ NW and HNW arrays on the Ti mesh

A piece of Ti mesh was put into a 100 mL Teflon-lined stainless steel autoclave filled with 30 mL 1 M NaOH aqueous solution. After that, the sealed autoclave was kept in an electric oven at 220 °C for 24 h. The as-prepared sodium titanate NW arrays on the Ti mesh were immersed in a 1 M HCl solution for 30 min, and H₂Ti₂O₅·H₂O NW arrays were obtained. TiO₂ NW arrays were obtained after annealing the H₂Ti₂O₅·H₂O NW arrays in air at 500 °C for 60 min. TiO₂ HNW arrays were obtained through a two-step hydrothermal reaction. Specifically, the TiO₂ NW arrays on the Ti mesh were immersed in a 20 mL solution containing 0.35 g potassium titanium oxide oxalate dihydrate (C₄K₂O₉Ti·2H₂O), 2.5 mL H₂O and 17.5 mL diethylene glycol, and kept at 180 °C for 6 h. The as-prepared TiO₂ NW and HNW arrays were soaked in a 0.1 M aqueous TiCl₄ solution at 70 °C for 30 min, and then annealed in air at 450 °C for 30 min.

Preparation of flexible counter electrode

The dip-coating method was adopted to prepare ink coated Ti mesh, as follows: the cleaned Ti meshes were dipped into commercial carbon ink for 1 min, taken out, and heated in air at 300 °C for 20 min. The platinum coated Ti mesh was fabricated by thermal deposition of H₂PtCl₆ onto Ti mesh in air at 420 °C for 30 min.

Device fabrication

The as-synthesized photoanodes were immersed in alcohol containing 0.5 mM ruthenium dye N719 and sensitized for 24 h. A flexible DSSC was fabricated by stacking the flexible dye-sensitized photoanode and counter electrode. A piece of PET with multiple tiny holes (20 μm in thickness) was used as a separator inserted between the photoanode and counter electrode. Then, the DSSC was sealed between two pieces of the PET film (60 μm in thickness) by pressing with surlyn film at 110 °C. The electrolyte, consisting of 0.6 M 1-butyl-3-methylimidazolium iodide (BMII), 0.03 M I₂, 0.06 M LiI, 0.5 M 4-tert-butyl-pyridine (4-TBP) and 0.1 M guanidinium thiocyanate, was injected into the space between the two PET films.

Characterization

A field emission scanning electron microscope (FE-SEM, ZEISS ULTRA 55), an X-ray diffraction (XRD) analyzer (Rigaku, MiniFlex600, Cu Kα, λ = 0.15406 nm, 40 kV) and a transmission electron microscope (TEM, JEOL 2100F, 200 kV) were used to characterize the morphologies and nanostructure of the samples. The photocurrent density–voltage (J–V) curves of the DSSCs were obtained by using an AM 1.5 G simulator (Abet Model 11000, 100 mW cm⁻²) with the incident light intensity calibrated using a standard silicon solar cell. The amount of dye loaded on the TiO₂ photoanodes (desorbed in a 0.02 M NaOH solution) was measured with an UV-visible spectrophotometer (Shimazu UV-2550). The electrochemical impedance spectra (EIS) of the samples were recorded using a VersaSTAT 3-400 (Princeton Applied Research) electrochemical workstation, which was carried out in dark conditions with the frequency ranging from 10 mHz to 1 MHz.

Results and discussion

The highly aligned one-dimensional TiO₂ nanostructures (NT, NW and HNW) on the flexible Ti mesh substrate were used as photoanodes in DSSCs. Fig. S1† shows a large-scale (75 × 90 mm²) flexible Ti mesh, which has good flexibility and high bending fatigue resistance. The diameter of the Ti wires in the mesh is ∼100 μm; the distance between two adjacent wires is 100–300 μm (Fig. S2†). Fig. 1a shows schematically the synthetic process of the anatase TiO₂ NT, NW and HNW arrays. Vertically aligned TiO₂ NTs were grown on the Ti wire surface using electrochemical anodization (Fig. S3†). The TiO₂ NWs were fabricated with an alkali hydrothermal treatment and a subsequent ion-exchange reaction. After a further hydrothermal reaction, lots of TiO₂ nanorod branches had grown on the TiO₂ NW trunks and the TiO₂ HNWs were obtained. Fig. 1b shows photos of TiO₂ NT arrays, TiO₂ HNW arrays, and dye-sensitized TiO₂ NT arrays and TiO₂ HNW arrays. It can be noted that the NT/Ti sample is slightly more transparent than the HNW/Ti sample.

Fig. 1 (a) Schematic illustration of the synthetic process of anatase TiO₂ NT, NW and HNW arrays on a Ti mesh substrate. (b) Photographs of the Ti mesh, TiO₂ NT/Ti mesh, TiO₂ HNW/Ti mesh, dye-sensitized TiO₂ NT/Ti mesh and dye-sensitized TiO₂ HNW/Ti mesh.

A SEM image of the Ti mesh after anodization at 60 V for 6 h is shown in Fig. 2a. It can be observed that the TiO₂ NT arrays are grown around the Ti wires closely and uniformly. The diameter of the TiO₂ NT/Ti wires is ∼110 μm. Fig. 2b and c are side-view SEM images of TiO₂ NT arrays. The length and outside diameter of TiO₂ NTs determined from the SEM images are ∼35 μm (Fig. 2b) and 100–150 nm (Fig. 2c), respectively.

Fig. 2 SEM images of the as-prepared TiO₂ NT, NW and HNW arrays. (a) Ti mesh covered with TiO₂ NT arrays. Cross-sectional SEM images of TiO₂ NT arrays at (b) low and (c) high magnifications. (d) Ti mesh covered with TiO₂ NW arrays. Cross-sectional SEM images of TiO₂ NW and HNW arrays at (e and g) low and (f and h) high magnifications, respectively. (i) More detailed view of the TiO₂ HNWs.

The anatase TiO₂ NW arrays on the Ti mesh were prepared via an alkali hydrothermal treatment and subsequent ion-exchange reaction. At first, the sodium titanate (Na₂Ti₂O₅·H₂O) NW arrays on the Ti substrate surface were synthesized through an alkali hydrothermal reaction. Then, the samples were immersed in a HCl aqueous solution. The H₂Ti₂O₅·H₂O NW arrays were obtained when Na⁺ was replaced completely by H⁺, which could be converted to anatase TiO₂ NW arrays after annealing at 500 °C for 30 min. The XRD patterns of Na₂Ti₂O₅·H₂O, H₂Ti₂O₅·H₂O and TiO₂ NW arrays are shown in Fig. S4.† Fig. 2d is a typical SEM image of a Ti mesh covered with TiO₂ NW arrays. It can be observed that the diameter of the TiO₂ NW/Ti wires is ∼125 μm. The side-view SEM images (Fig. 2e and f) further show that the TiO₂ NWs are ∼16 μm in length and ∼100 nm in diameter. The growth of the TiO₂ NW arrays can be attributed to the reactions below:

The as-fabricated TiO₂ NW arrays/Ti mesh undergo a further hydrothermal reaction in the solution containing C₄K₂O₉Ti, diethylene glycol and H₂O, leading to TiO₂ HNW arrays composed of numerous short nanorod branches on the surface of the NW trunks. Fig. 2g–i show the cross-sectional SEM images of the TiO₂ HNW arrays at different magnifications. The length of the TiO₂ HNWs is ∼17 μm, which agrees with that of NWs (Fig. 2g). As shown in the enlarged views (Fig. 2h and i), a large number of short nanorods (50–100 nm in length) are growth on the TiO₂ NW trunks.

The morphologies and crystal structures of anatase TiO₂ NW and HNW arrays are further characterized by XRD and TEM analysis. Fig. 3a displays the XRD patterns of the as-prepared H₂Ti₂O₅·H₂O NW arrays, TiO₂ NW arrays, TiO₂ HNW arrays and TiO₂ NT arrays on the Ti mesh. The bottom curve is the diffraction pattern of the pristine Ti mesh. The diffraction peaks of the H₂Ti₂O₅·H₂O NW arrays prepared through the hydrothermal treatment and ion-exchange reaction are in agreement with the H₂Ti₂O₅·H₂O phase according to its PDF card (JCPDS 47-0124). Notably, the (101) and (200) diffraction peaks of the anatase TiO₂ phase (JCPDS 21-1272) can be observed at 2θ = 25.3° and 48.1° for the TiO₂ NW, HNW and NT arrays, respectively. As is well known, the anatase TiO₂ nanostructures are more outstanding in their photovoltaic performance for DSSC applications. Fig. 3b is a typical TEM image of an anatase TiO₂ NW of ∼100 nm in diameter, which is in line with the SEM observations. Fig. 3b combined with the SEM image (Fig. 2f) confirm that the TiO₂ NW is a cylindrically shaped nanostructure. Fig. 3c displays a high-resolution transmission electron microscopy (HRTEM) image of the edge area of the TiO₂ NW (marked in a red color in Fig. 3b), which confirms that the NW is single crystalline. The inter-planar spaces of (200) and (101) are 0.19 nm and 0.36 nm, respectively, which are in excellent agreement with the anatase phase. Furthermore, Fig. 3c shows that the axis of TiO₂ NW is along the [100] direction.^24,36,37 TEM results (Fig. 3d and e) of the TiO₂ HNWs clearly show that a number of short TiO₂ nanorod branches (∼10 nm in diameter and 20–80 nm in length) are grown directly on the long NW trunks. From the HRTEM investigation (Fig. 3f), it is found that the short nanorods are single crystalline structures and grow along the [010] direction, which is perpendicular to the axial direction [100] of the NW. The inter-planar spaces of (200) and (020) are both equal to 0.19 nm. This lattice structure greatly matches the boundary structure of the NW trunk (as shown in Fig. 3c and f). These results suggest that the nanorod branches are well perpendicular to the NW.

Fig. 3 (a) XRD patterns of Ti mesh, and H₂Ti₂O₄(OH)₂ NW arrays, TiO₂ NW arrays, TiO₂ HNW arrays and TiO₂ NT arrays on Ti mesh. (b) TEM image and (c) HRTEM image of a TiO₂ NW. (d and e) TEM images of TiO₂ HNWs. (f) HRTEM image of the connecting region of the nanorod branch and NW trunk.

The as-synthesized TiO₂ NW, HNW and NT arrays were further explored as photoanode materials to fabricate mesh based flexible DSSCs. As depicted in Fig. 4a, the dye-sensitized TiO₂ photoanode, separator and Pt/Ti mesh counter electrode were sandwiched together to assemble the cell. The flexible device is filled with a redox electrolyte and sealed with two pieces of PET film. For the mesh based TiO₂ photoanode, the conductivity of the entire Ti mesh should be considered. The formation of TiO₂ films on the Ti wire surface may break the connection between the interlaced horizontal and longitudinal Ti wires. Therefore, all of the horizontal and longitudinal Ti wires coming out of the sealed device were attached to a copper foil electrode for electron collection (Fig. 4b). To avoid the short circuit between the photoanode and counter electrode while bending the device, a thin PET film with multiple tiny holes was used as a separator, inserted between the photoanode and counter electrode. The as-fabricated mesh based DSSC is flexible and can be bent at different angles without fracture (Fig. 4c).

Fig. 4 (a) Schematic diagram of the mesh based flexible DSSCs. (b) Photograph of a flexible DSSC device. (c) Photograph showing the flexibility of the device.

Fig. 5a shows the typical J–V curves of flexible DSSCs based on the three different photoanodes (NT, NW and HNW). Similar to a planar DSSC device, the effective area for calculating the PCE and dye adsorption amounts is the product of the length and width of the whole Ti mesh (1 cm × 1 cm), as shown in Fig. 5a. The resultant photovoltaic parameters of different DSSCs including the open-circuit voltage (V_oc), short-circuit current density (J_sc), fill factor (FF) and PCE (ƞ) are summarized in Table 1. The PCE of the NT, NW and HNW cells are 1.52, 0.85 and 1.76%, respectively. The normalized PCE are calculated using the projected area of the TiO₂/Ti mesh photoanodes. The projected area of the NT, NW and HNW cells are 0.68, 0.75 and 0.75 cm², respectively; and their normalized PCEs are 2.21, 1.13 and 2.39%, respectively. The J_sc of these cells are 7.04, 3.36 and 7.24 mA cm⁻². Actually, the J_sc values of the DSSCs are closely related to the amounts of dye adsorbed on the photoanode, which is determined by the surface area of the TiO₂ nanostructures. The amounts of dye adsorbed on different photoanodes are calculated using the Beer–Lambert Law according to the UV-vis absorption measurement (Fig. S5†). The dye adsorption amounts of NT, NW and HNW arrays are 72.3, 56.7 and 112.9 nmol cm⁻², respectively, as shown in Table 1. This result indicates that the surface areas of the different TiO₂ nanoarrays are ascending in the sequence of NW < NT < HNW. Hence, the increased dye loading amounts of the larger surfaces definitely contributed to the enhanced photocurrents and thus higher efficiencies. Specifically, despite the fact that the TiO₂ HNW arrays are significantly higher in dye loading amounts than the NT arrays (112.9 and 72.3 nmol cm⁻², respectively), the J_sc values of them are almost identical. This is owing to the fact that the NT arrays are more competent for light-harvesting through light scattering and trapping than the NW and HNW arrays. It also can be seen in Fig. 1b, the TiO₂ NT arrays/Ti mesh sample is darker than TiO₂ HNW arrays/Ti mesh sample.

Fig. 5 (a) J–V curves of flexible DSSCs based on the TiO₂ NT, NW and HNW arrays, which are measured by using an AM 1.5 G solar simulator (100 mW cm⁻²). (b) EIS spectra of TiO₂ NT, NW and HNW array based DSSCs.

Table 1 Detailed photovoltaic performance parameters (J_sc, V_oc, FF and ƞ) of DSSCs based on the different photoanodes

	Adsorbed dye (nmol cm^−2)	Rs (Ω)	R1 (Ω)	R2 (Ω)	Jsc (mA cm^−2)	Voc (V)	FF	ƞ (%)	Normalized ƞ (%)
NT	72.3	13.5	6.6	192.0	7.04	0.68	0.32	1.52	2.21
NW	56.7	12.9	6.1	152.1	3.36	0.61	0.47	0.85	1.13
HNW	112.9	17.4	4.2	294.7	7.24	0.72	0.34	1.76	2.39

---

Table 1 shows that the V_oc values of the NT, NW and HNW array based cells are 0.68, 0.61 and 0.72 V, respectively. It is well known that the V_oc value is closely related to the recombination resistance and electron lifetime in the conduction band of TiO₂.^38,39 To obtain electrochemical behavior information within these cells, EIS measurements were performed. In Fig. 5b, the Nyquist plots of the DSSCs for the different types of TiO₂ photoanode are presented. Generally, two semicircles are observed in a Nyquist diagram of a DSSC. The first semicircle in the high frequency region represents the charge transfer resistance at the electrolyte/counter electrode interface (R₁), while the second semicircle in the low frequency region represents the recombination resistance at the photoanode/electrolyte interface (R₂). The impedance spectra can be analyzed using Z-view software using an equivalent circuit and the corresponding parameters (internal ohmic resistance (R_s), R₁ and R₂) are obtained by fitting the equivalent circuit. As shown in Table 1, these cells possess similar R_s and R₁ values, which is due to using the same electrolyte and counter electrode in each case. However, the R₂ value of the NW, NT and HNW cells are 152.1, 192 and 294.7 Ω, respectively. It is worth noting that the HNW cells exhibit greatly increased R₂ values compared to the NW cells. This may be because a compact TiO₂ layer was formed at the bottom of the HNW arrays while the short nanorod branches were grown on the surface of the TiO₂ NW trunks, which is conducive to preventing the recombination of photo-generated electrons with I₃⁻ at the electrolyte/Ti substrate interface.⁴⁰ The increase of the recombination resistance (NW < NT < HNW) accounts for the decline in the recombination reaction rate (increase in the electron lifetime), which will lead to increased V_oc values. This result is in concurrence with the above photovoltaic data.

Recently, mesh based flexible DSSCs have received immense attention due to their great potential for portable and wearable electronic applications. Here, the anatase TiO₂ HNW arrays/Ti mesh photoanode and the Pt/Ti mesh counter electrode are used to fabricate a complete device (1 cm × 1 cm), which shows excellent flexibility and bending stability. As shown in Fig. S6,† a DSSC device was bent 300 times. The J_sc, V_oc, FF and PCE (ƞ) values as functions of bend cycle number are shown in Fig. 6. During the bending cycles, these photovoltaic parameters values varied slightly. The TiO₂ HNW arrays/Ti mesh based DSSC has long-term and repeatable flexibility, in which case the PCE decreased less than 12% over 300 bending cycles (1.33, 1.33, 1.22 and 1.17% for 0, 100, 200 and 300 cycles, respectively, as shown in Fig. S6b†). A detailed comparison to other works is summarized in Table S1.† Our TiO₂ HNW/Ti mesh based DSSCs achieved a relatively high PCE compared to other works (1.49% for Fan et al.,²² 1.23% for Liu et al.,³⁵ 3.67% for Pan et al.⁵ and 0.93% for Li et al.³³). In Pan’s work, the TiO₂ NT/Ti mesh was used as the photoanode, and the PCE of the DSSC device decreased ∼10% after bending only 100 times. Our completely sealed flexible DSSC device has good bendability and stability, which can maintain an unchanged PCE after bending 100 times and maintain ∼88% of PCE after bending 300 times. Our work demonstrated that the anatase HNW based DSSCs have better bending stability than NT based DSSCs, probably due to the larger spaces and higher compression tolerance of HNW (as shown in Fig. 2b and g, for comparison).

Fig. 6 Photovoltaic parameters (J_sc, V_oc, FF and ƞ) of the flexible DSSCs as a function of bending cycle number up to 300 cycles.

A flexible, inexpensive and effective counter electrode material is also a research hotspot in flexible DSSCs.^41,42 Here, commercial ink used as a counter electrode material to replace Pt has also been investigated. Cyclic voltammetry (CV) was performed to examine the catalytic activity of the commercial ink counter electrode in the DSSCs (Fig. 7a). The electrolyte is an acetonitrile solution containing 10 mM LiI, 1 mM I₂ and 0.1 M LiClO₄. Two pairs of oxidation and reduction peaks are observed for Pt and ink in the CV scan, which are attributed to the following reactions:

left: I₃⁻ + 2e⁻ ↔ 3I⁻ (4)

right: 3I₂ + 2e⁻ ↔ 2I₃⁻ (5)

Fig. 7 (a) Cyclic voltammograms of the Pt/Ti mesh and ink/Ti mesh electrodes measured at a scan rate of 50 mV s⁻¹ (b) J–V curves of the DSSCs based on different counter electrodes (Pt and ink).

Generally, the counter electrode serves as the electrocatalyst to reduce I₃⁻ to I⁻. The reduction peak current density (J_red) and peak-to-peak separation voltage (V_pp) are used to gauge the catalytic activities of counter electrodes. As the ink electrode material, with a high specific area (consisting of ∼20 nm carbon NPs, as shown in Fig. S7†), shows a high J_red, it is more reasonable to gauge their catalytic activities by comparing the V_pp values. The Pt/Ti mesh electrode exhibits a higher catalytic performance than the ink/Ti mesh due to a lower V_pp (0.7 and 0.82 V for Pt and ink, respectively). In Fig. 7b, a higher PCE of the DSSC device using the Pt/Ti mesh counter electrode is presented. This result indicates that the commercial ink is a great choice for cost-efficiency and widespread production consideration, but Pt still represents the optimal component within the DSSCs.

As expected, the mesh based flexible DSSCs could be integrated and applied to provide power for various electronic devices (Fig. 8a and b). Fig. 8c shows the J–V characteristics of a flexible solar panel implemented by cascading four mesh based DSSCs. The V_oc value has been increased from 0.72 to 2.4 V. As a result, these four series-connected solar cells can efficiently power a JNU logo consisting of 32 green-LEDs. These results suggest that the Ti mesh based flexible DSSC device possesses promising potential for application as the lightweight and flexible energy supply in wearable and portable electronics.

Fig. 8 (a and b) Photographs of 32 green-LEDs lit up by four series-connected DSSCs (note: due to the strong background brightness from the solar simulator, the LEDs do not look as bright as they actually are). (c) J–V curve of the four mesh based flexible DSSCs in series.

Conclusion

In summary, thin, lightweight and flexible mesh based DSSCs have been developed by using three different anatase TiO₂ nanostructures on Ti meshes as efficient flexible photoanodes. A two-step hydrothermal method is used to synthesize the TiO₂ HNW arrays, which achieved a higher PCE of 1.76% than electrochemical anodized TiO₂ NT arrays (1.52%). This improvement can be attributed to the large surface area and low recombination rate of electrons and redox electrolyte. Moreover, the commercial ink has been identified as a promising candidate to fabricate flexible counter electrodes to replace Pt and achieve a considerable PCE of 1.59%. The as-fabricated mesh based DSSC retains 88% of PCE after 300 bending cycles, demonstrating its good flexibility and fatigue resistance. These sealed DSSCs have been integrated to light up a group of green-LEDs, which demonstrated that mesh based DSSCs are promising candidates for wearable/bendable energy supply devices in everyday applications.

Acknowledgements

W. J. M. and M. M. Wu are thankful for the financial support from the National Natural Science Foundation of China (Grants 21376104 and 21271190) and the Natural Science Foundation of Guangdong Province, China (Grants S2013010012876, 2014A030306010 and S2012020011113).

Notes and references

1. H. Nishide and K. Oyaizu, Science, 2008, 319, 737.
2. H. Li, Q. Zhao, W. Wang, H. Dong, D. Xu, G. Zou, H. Duan and D. Yu, Nano Lett., 2013, 13, 1271.
3. Y.-H. Lee, J.-S. Kim, J. Noh, I. Lee, H. J. Kim, S. Choi, J. Seo, S. Jeon, T.-S. Kim, J.-Y. Lee and J. W. Choi, Nano Lett., 2013, 13, 5753.
4. L. Du, P. Yang, X. Yu, P. Liu, J. Song and W. Mai, J. Mater. Chem. A, 2014, 2, 17561.
5. S. Pan, Z. Yang, P. Chen, J. Deng, H. Li and H. Peng, Angew. Chem., 2014, 126, 6224.
6. M. Armand and J. M. Tarascon, Nature, 2008, 451, 652.
7. S. Pan, H. Lin, J. Deng, P. Chen, X. Chen, Z. Yang and H. Peng, Adv. Energy Mater., 2014, 5, 1401438.
8. S. Lee, Y. Lee, J. Park and D. Choi, Nano Energy, 2014, 9, 88.
9. Y. Fu, H. Wu, S. Ye, X. Cai, X. Yu, S. Hou, H. Kafafy and D. Zou, Energy Environ. Sci., 2013, 6, 805.
10. P. Yang, Y. Ding, Z. Lin, Z. Chen, Y. Li, P. Qiang, M. Ebrahimi, W. Mai, C. P. Wong and Z. L. Wang, Nano Lett., 2014, 14, 731.
11. B. O’Regan and M. Grätzel, Nature, 1991, 353, 737.
12. S. Mathew, A. Yella, P. Gao, R. Humphry-Baker, F. E. CurchodBasile, N. Ashari-Astani, I. Tavernelli, U. Rothlisberger, K. NazeeruddinMd and M. Grätzel, Nat. Chem., 2014, 6, 242.
13. S. Ito, T. N. Murakami, P. Comte, P. Liska, C. Grätzel, M. K. Nazeeruddin and M. Grätzel, Thin Solid Films, 2008, 516, 4613.
14. Z. Chai, J. Gu, P. Qiang, X. Yu and W. Mai, CrystEngComm, 2015, 17, 1115.
15. X. Tan, P. Qiang, D. Zhang, X. Cai, S. Tan, P. Liu and W. Mai, CrystEngComm, 2014, 16, 1020.
16. P. Qiang, P. Yang, Z. Liang, Y. Luo, J. Yu, Y. Lan, X. Cai, S. Tan, P. Liu and W. Mai, J. Alloys Compd., 2014, 583, 300.
17. D. Zhang, T. Yoshida, T. Oekermann, K. Furuta and H. Minoura, Adv. Funct. Mater., 2006, 16, 1228.
18. T. Yamaguchi, N. Tobe, D. Matsumoto, T. Nagai and H. Arakawa, Sol. Energy Mater. Sol. Cells, 2010, 94, 812.
19. X. Chen, Q. Tang, Z. Zhao, X. Wang, B. He and L. Yu, Chem. Commun., 2015, 51, 1945.
20. H. C. Weerasinghe, F. Huang and Y.-B. Cheng, Nano Energy, 2013, 2, 174.
21. S. Ito, N.-L. C. Ha, G. Rothenberger, P. Liska, P. Comte, S. M. Zakeeruddin, P. Pechy, M. K. Nazeeruddin and M. Grätzel, Chem. Commun., 2006, 4004.
22. X. Fan, Z. Chu, L. Chen, C. Zhang, F. Wang, Y. Tang, J. Sun and D. Zou, Appl. Phys. Lett., 2008, 92, 113510.
23. X. Fan, Z. Z. Chu, F. Z. Wang, C. Zhang, L. Chen, Y. W. Tang and D. C. Zou, Adv. Mater., 2008, 20, 592.
24. J. Liao, B. Lei, H. Chen, D. Kuang and C. Su, Energy Environ. Sci., 2012, 5, 5750.
25. X. Wang, Q. Tang, B. He, R. Li and L. Yu, Chem. Commun., 2014, 51, 491.
26. J. Liang, G. Zhang, Y. Yang and J. Zhang, J. Mater. Chem. A, 2014, 2, 19841.
27. X. Cai, H. Wu, S. Hou, M. Peng, X. Yu and D. Zou, ChemSusChem, 2014, 7, 474.
28. S. Hou, Z. Lv, H. Wu, X. Cai, Z. Chu, Yiliguma and D. Zou, J. Mater. Chem., 2012, 22, 6549.
29. W. Zeng, L. Shu, Q. Li, S. Chen, F. Wang and X. Tao, Adv. Mater., 2014, 26, 5310.
30. M. Peng, X. Yu, X. Cai, Q. Yang, H. Hu, K. Yan, H. Wang, B. Dong, F. Zhu and D. Zou, Nano Energy, 2014, 10, 117.
31. Z. Yang, J. Deng, X. Sun, H. Li and H. Peng, Adv. Mater., 2014, 26, 2643.
32. Z. Yang, J. Deng, H. Sun, J. Ren, S. Pan and H. Peng, Adv. Mater., 2014, 26, 7038.
33. H. Li, Q. Zhao, H. Dong, Q. Ma, W. Wang, D. Xu and D. Yu, Nanoscale, 2014, 6, 13203.
34. W. He, J. Qiu, F. Zhuge, X. Li, J.-H. Lee, Y.-D. Kim, H.-K. Kim and Y.-H. Hwang, Nanotechnology, 2012, 23, 225602.
35. Z. Liu, V. Subramania and M. Misra, J. Phys. Chem. C, 2009, 113, 14028.
36. J. E. Boercker, E. Enache-Pommer and E. S. Aydil, Nanotechnology, 2008, 19, 095604.
37. B. Liu, J. E. Boercker and E. S. Aydil, Nanotechnology, 2008, 19, 505604.
38. M. Wang, P. Chen, R. Humphry-Baker, S. M. Zakeeruddin and M. Grätzel, ChemPhysChem, 2009, 10, 290.
39. W. Wu, H. Rao, Y. Xu, Y. Wang, C. Su and D. Kuang, Sci. Rep., 2013, 3, 1892.
40. H. Yu, S. Zhang, H. Zhao, G. Will and P. Liu, Electrochim. Acta, 2009, 54, 1319.
41. L. Qiu, Q. Wu, Z. Yang, X. Sun, Y. Zhang and H. Peng, Small, 2014, 11, 1150.
42. S. Yun, A. Hagfeldt and T. Ma, Adv. Mater., 2014, 26, 6210.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra17294b

---