Introduction
As a promising alternative to conventional silicon-based solar cell, dye-sensitized solar cells (DSSCs) have attracted considerable attention in the last two decades, because of their low manufacturing costs and relatively high energy-conversion efficiencies [1–4]. In DSSC components, the photoanode is a key part that plays an essential role in determining the dye loading and electron transportation, and hence the photon to electron conversion efficiency. To realize highly efficient DSSC, a nanostructured photoanode should possess several favorable intrinsic characteristics, such as large surface area to permit high dye loading, direct electron transport pathways for long electron diffusion lengths, and compatible energy levels to achieve high electron injection efficiency and high voltage. It is therefore highly desirable to develop a photoanode that meets the above requirements.
There have been many efforts to develop efficient photoanode materials such as TiO2, ZnO, SnO2, Nb2O5, and other composites. So far, TiO2 has been widely investigated as a photoanode material for DSSCs. With a band gap similar to that of TiO2, ZnO is another alternative photoanode material, which has attracted much attention due to its high mobility of about 115–155 cm2·V-1·s-1, which is much higher than that of TiO2 (10-5 cm2·V-1·s-1) [5,6]. Among the various morphologies, one-dimensional (1D) nanostructures such as nanorods, nanotubes, and nanowires, which can offer more superior electron transport pathways, have been attracting increasing attention. However, the power conversion efficiencies of DSSCs based on ZnO 1D nanostructures are still at relatively low levels, because of the insufficient internal surface area.
To overcome this shortcoming, it is necessary to design a novel core/shell photoanode by applying metal oxide shells such as SiO2 [7], ZrO2 [8], Nb2O5 [9–12], and Al2O3 [13] to the core nanostructures. The metal oxide shells synthesized on core nanostructures could increase the specific area of the electrode for dye loading, thereby enhancing the photocurrent density. Moreover, the layer formed by coating with these materials act as an energy barrier that decreases the electron recombination losses, shifts the conduction band downward, which increases the electron injection, and enhances the injection efficiency.
Nb2O5 is a promising metal oxide because it supports good N719 dye loading due to its basic character [14–17], and its conduction band level is more negative than that of TiO2 [18,19]. Here, we applied a Nb2O5 coating on ZnO nanorod to obtain ZnO/Nb2O5 core/shell nanorod arrays, which were used as the photoanodes in DSSCs. The results showed that improved efficiencies of 1.609% and 1.995% were obtained for DSSCs based on ZnO/Nb2O5 core/shell nanorod synthesized in-low concentration NbCl5 solution and in high-concentration NbCl5 solution, respectively, compared with the efficiency of 0.856% for the DSSC based on bare ZnO nanorod photoanode.
Experimental
Synthesis of ZnO nanorod arrays
The ZnO seed layer on fluorine-doped tin oxide (FTO) glasses was prepared using a previously published procedure [20]. Then, 0.455 g Zn(NO3)2·6H2O and 0.555 g NH4F were dissolved into 50 mL deionized water. After that, NH3·H2O was dropped into the above solution while stirring. In this process, the clear solution turned turbid when NH3·H2O was first dropped into the solution, and then gradually it turned clear with continued stirring. Subsequently, the prepared ZnO seed layer on FTO glass (with the FTO layer facing downwards) was placed into a Teflon-lined autoclave, which contained the above solution, and was then sealed. Finally, the Teflon-lined autoclave was heated at 70°C for 8 h to obtain ZnO nanorod arrays.
Synthesis of ZnO/Nb2O5 core/shell nanorod arrays
Figure 1 shows the process of growing ZnO/Nb2O5 core/shell nanorod arrays on FTO glass. First, 0.135 and 0.270 g NbCl5 were dissolved into 50 mL anhydrous ethanol. After that, the obtained ZnO nanorod arrays were placed into the two Teflon-lined autoclaves, each containing one of the above two NbCl5 solutions. Then, the Teflon-lined autoclaves were heated at 180°C for 24 h to obtain ZnO/Nb2O5 core/shell nanorod arrays with ZnO/Nb2O5(1) for 0.135 g NbCl5 and ZnO/Nb2O5(2) for 0.270 g NbCl5. Finally, the ZnO/Nb2O5 core/shell nanorod arrays were sintered at 500°C in air for 1 h.
Fabrication of DSSCs based on ZnO/Nb2O5 core/shell nanorod arrays
The ZnO/Nb2O5 core/shell nanorod arrays were dipped in a dye solution containing 0.5 mM†) N719 (Dyesol) dye for 24 h. Then, the ZnO/Nb2O5 electrode was scrapped to obtain an active area of 25 mm2. The counter electrode was prepared by coating a 0.6 mM H2PtCl6·6H2O solution in anhydrous ethanol onto the FTO substrate. After that, the ZnO/Nb2O5 electrode was assembled with the counter electrode by clamping a 25 mm thick polymeric film (Surlyn, DuPont). Then, an electrolyte solution was injected into the gap between the ZnO/Nb2O5 electrode and the counter electrode. The electrolyte solution contained 0.05 M LiI (Sigma-Aldrich), 0.03 M I2 (Aldrich), and 0.5 M 4-tert-butylpyridine (Aldrich) in a solution containing acetonitrile (Aldrich). Finally, the injecting hole was sealed with an adhesive tape to obtain the completed device.
Characterization
A scanning electron microscope (SEM, Quanta200, FEI, Netherlands) and a transmission electron microscope (TEM, JEOL TEM-2010) were used to measure the structure and morphology of the nanomaterials. The current–voltage characterization was performed using a Keithley 2400 source meter under simulated AM 1.5 sunlight illumination (100 mW·cm-2) provided by an Oriel solar simulator (Model 9119X, Newport Co.). The illuminated active area of the photovoltaic measurements was 0.16 cm2. The electrochemical impedance spectra (EIS) of the devices were tested at −0.6 V for the range from 1 MHz to 1 Hz, with an advanced electrochemical system (PAR2273) under dark conditions.
Results and discussion
Structural and morphological characterization of ZnO/Nb2O5 core/shell nanorod
Figures 2(a) and 2(b) show the top-view and cross-view SEM images of the ZnO nanorod arrays, respectively. Figure 2(a) reveals that the top of the nanorods is uniform. From the tilted-view SEM image, we can clearly see that the high-density ZnO nanorods grew vertically on the FTO substrate. Their length is about 5 mm. After the coating of the Nb2O5 shell, the core-shell structure was investigated by SEM. Figure 2(c) shows the top-view SEM image of the ZnO/ Nb2O5 core-shell nanorod film. It clearly reveals that the diameter of the core-shell nanorod increased and its surface became rough. The corresponding tilted-view SEM image shows that the core-shell structure has a similar shape as that of the ZnO nanorod shown in Fig. 2(d). To investigate the coating of Nb2O5, the samples were synthesized in a low-concentration NbCl5 solution (0.135 g NbCl5) and a higher-concentration NbCl5 solution (0.270 g NbCl5). The results are shown in Fig. 3. By comparison, we could find that upon increasing the concentration of NbCl5, the ZnO nanorods became rougher on the side surface and were fully covered with Nb2O5 nanoparticles on the top. This shows that a high concentration of NbCl5 could facilitate significantly more growth of Nb2O5 on the top of the nanorod than on its side.
Fig.3 Top-view SEM images of ZnO/Nb2O5 core/shell nanorod arrays synthesized in low concentration NbCl5 solution (0.135 g NbCl5): (a) low and (b) high magnification. The top-view SEM images of ZnO/Nb2O5 core/shell nanorod arrays synthesized in high concentration NbCl5 solution (0.270 g NbCl5): (c) low and (d) high magnification |
In the TEM image (Fig. 4(a)), obvious differences in terms of the contrast between the center and fringe parts of each individual rod-like structure are observed, which indicates that the rod-like structures are core-shell structures with core diameters of 150–200 nm and shell thicknesses of 30–40 nm. The corresponding sellected area electron diffraction (SAED) pattern obtained from the circled area in Fig. 4(a) is shown in Fig. 4(b). We can see single crystal diffraction spots in Fig. 4(b), corresponding to ZnO (103) and (203). In addition, Fig. 4(b) also shows an amorphous diffraction pattern, which is indicated by the dotted line. We measured the radius of the amorphous diffraction ring, and the spacing between the crystal planes ranged from 0.39 to 0.28 nm, which was consistent with the strong peak position of JCPDS (320711) Nb2O5. Therefore, we inferred that the amorphous diffraction ring was from the amorphous Nb2O5 wrapped on the surface of ZnO.
Characterization of photovoltaic performances for DSSCs
Figure 5 shows the optical absorption spectra of the ZnO and ZnO/Nb2O5 nanorod films. For the bare ZnO nanorod film, the onset of the band gap transition is at ~420 nm. The presence of Nb2O5 shell increases another adsorption edge at ~370 nm relative to Nb2O5. Compared with the ZnO nanorod arrays, the coating of the Nb2O5 moved the onset of absorption to a lower wavelength, and the light absorption intensity was enhanced. The larger band gap of Nb2O5 semiconductor is related to the more negative conduction band potential and the larger open-circuit photovoltage of the Nb2O5 cell. This result is consistent with the following photovoltaic performances.
The photocurrent density-voltage curves of DSSCs based on the photoanodes of bare ZnO nanorods, ZnO/Nb2O5(1) core/shell nanorod (0.135 g NbCl5), and ZnO/Nb2O5(2) core/shell nanorod (0.270 g NbCl5) are shown in Fig. 6. The photovoltaic performances of the three devices are listed in Table 1. The device based on the photoanode of bare ZnO nanorods showed poor performances with short-circuit current density (Jsc), open-circuit voltage (Voc), fill factor (FF), and power conversion efficiency (PCE) of 4.46 mA·cm-2, 537 mV, 0.357, and 0.856%, respectively. This efficiency was similar to the efficiencies of DSSCs based on bare ZnO nanorod arrays reported by other groups [21,22]. The device based on ZnO/Nb2O5(1) core/shell nanorod exhibited improved performances with Jsc, Voc, FF, and PCE of 5.52 mA·cm-2, 569 mV, 0.512, and 1.609%, respectively. The increased Jsc is mainly attributed to the increased surface area of ZnO/Nb2O5 core/shell nanorod, which was caused by the small Nb2O5 nanoparticles wrapped on the ZnO nanorod. The increased Voc is attributed to the fact that the Fermi level of ZnO/Nb2O5 core/shell nanorod is higher than that of bare ZnO nanorod, since Nb2O5 has a higher conduction band than ZnO. This increases the Voc value because the maximum Voc of DSSC is mainly decided by the difference between the Fermi level of the photoanode and the redox potential of I-/I3-. In addition to the increased Jsc and Voc, the device based on ZnO/Nb2O5(1) core/shell nanorod exhibited a higher FF of 0.512 than the device based on ZnO nanorod (0.357). This indicates that the Nb2O5 shell provided another charge-transporting channel other than the ZnO nanorod, resulting in decreased series resistance. As for the device based on ZnO/Nb2O5(2) core/shell nanorod, the Jsc, FF, and PCE were further increased to 5.95 mA·cm-2, 0.592, and 1.995%, respectively. Compared with the ZnO/Nb2O5(1) core/shell nanorod, the ZnO/Nb2O5(2) core/shell nanorod had a thicker diameter and had additional Nb2O5 particles on top of the nanorod. Such a core/shell nanorod structure makes the ZnO/Nb2O5 nanorod photoanode advantageous in the following two aspects. On the one hand, since the protons that are released from the dye molecules in the ethanolic solution dissolved ZnO to generate Zn2+-dye aggregates, the structure of the ZnO crystals was easily destroyed after loading the Ru-complex dyes. However, after being coated with Nb2O5, the recombination was suppressed by passivating its centers on the ZnO nanostructure surface. On the other hand, the electrolyte was suppressed presumably due to the energy barrier formed at the ZnO/ Nb2O5 interface. Thus, Jsc and Voc of the ZnO/ Nb2O5 nanorod photoanode are much higher than that of the bare ZnO photoanode.
Tab.1 Photovoltaic performance of DSSCs based on photoanodes of ZnO nanorods, ZnO/Nb2O5(1) nanorod and ZnO/Nb2O5(2) nanorod under AM1.5 conditions 100 mW·cm−2. The length of three nanorods in the devices are all around 5 µm |
| photoanode | Jsc/(mA·cm−2) | Voc/mV | FF | PCE/% |
|---|---|---|---|---|
| ZnO | 4.46 | 537 | 0.357 | 0.856% |
| ZnO/Nb2O5(1) | 5.52 | 569 | 0.512 | 1.609% |
| ZnO/Nb2O5(2) | 5.92 | 569 | 0.592 | 1.995% |
Conclusions
In summary, ZnO/Nb2O5 core/shell nanorods were successfully developed by solvothermal synthesis and applied as photoanodes for DSSCs. ZnO nanorod core was first synthesized by a hydrothermal process, and then Nb2O5 shell was directly synthesized on the ZnO nanorod core by solvothermal reaction in NbCl5 solution. SEM and TEM images revealed that the ZnO nanorods were uniformly wrapped by the Nb2O5 shell layers with a thickness of 30–40 nm. Photovoltaic characterization showed that improved efficiencies of 1.609% and 1.995% were obtained for DSSCs based on ZnO/Nb2O5 core/shell nanorod synthesized in low-concentration NbCl5 solution and high-concentration NbCl5 solution, respectively, compared with the efficiency of 0.856% for the DSSC based on bare ZnO nanorod photoanode. Our work offers a facile strategy to develop functional composite nanomaterials for photovoltaic devices.