Nanoparticle-based hierarchical zinc oxide chains for enhanced efficiency of dye-sensitized solar cells

Abstract

One-dimensional ZnO chains composed of nanoparticles were prepared using a facile method. Inspired by the possibility of achieving unique structures for the working electrodes of dye-sensitized solar cells (DSSCs), hierarchical ZnO chains can take the advantages of both one-dimensional (1D)-structures and nanoparticles. A superior light scattering ability, slower electron recombination rate and faster electron transport rate together enhanced the photoelectric conversion performance and showed a superior conversion efficiency compared to ZnO 1D-based DSSCs and ZnO nanoparticle-based DSSCs.

---

Introduction

Zinc oxide (ZnO) is considered to be the best potential alternative to TiO₂ used as the photoanode of dye-sensitized solar cells (DSSCs) due to the similarity of its physical properties compared to TiO₂, such as higher electron mobility and ease of control over material morphologies.^1–3 In the past decades, considerable effort has been devoted to increase the conversion efficiency (η) of ZnO-based DSSCs through the synthesis of ZnO with varied architectures and morphologies attributed to increased dye loading and/or light-harvesting efficiency, faster electron transport, and prolonged electron lifetime.^4–9

In DSSCs, the photoelectrode film must possess a large surface area to adsorb sufficient dye molecules to favor the capture of incident photons. Porous nanocrystalline films can satisfy this requirement, and they have been extensively studied.^10–13 However, such photoanodes have been found to restrict further increase in the DSSC performance due to a higher charge-recombination rate and slower electron transport.^14–16 Therefore, a variety of functional one-dimensional (1D) nanostructured ZnO photoelectrodes, such as nanotubes,^17,18 nanowires,^14,19 and nanobelts,²⁰ have been developed to tackle this issue. 1D nanostructure photoelectrodes can provide a direct transport channel for rapid electrons transport, decreasing the possibility of charge recombination and avoiding the high resistance existing in randomly oriented nanoparticles. However, the cell performance of simple 1D nanostructures kept at a relatively low level due to their deficient dye loading resulting from insufficient surface area.^14,21 So, a further improved η value can be expected by making photoanodes possessing both a high surface area and fast electron transport.²² Thus, various hierarchical ZnO structures combining multi-scale building blocks have been employed to meet these harsh requirements, including branch structures,²³ microspheres on wires,²⁴ tetrapods,²⁵ nanoflowers,^26,27 and composite nanowires/nanoparticles.²⁸ The η values of these cells were largely increased, compared to those of pure ZnO nanoparticle (NP) or 1D nanostructure DSSCs. However, these DSSCs either obtained η less than 4%, or were synthesized via complicated synthetic steps.

In this paper, nanoparticle-based hierarchical ZnO chains were synthesized and applied as the photoanode for DSSCs. The unique architecture has a hierarchical structure: the chain is several micrometers in length; the chain consists of interconnected ZnO nanoparticles. The chains serve as a direct pathway for fast electron transport, causing a slower electron recombination rate and faster electron transport rate. Meanwhile, the nanoparticles constituting the chains can provide a larger surface area than one-dimensional ZnO rods with similar lengths to improve the photocurrent. DSSCs fabricated from the ZnO chains gave an η of 5.06%, as a result of taking the advantages of both chains and nanoparticles.

Experimental

Preparation of the electrode

12 mmol of Zn(CH₃COO)₂·2H₂O and 36 mmol of CO(NH₂)₂ were added to 30 mL of H₂O. The mixture was further stirred and then transferred to a 50 mL Teflon-lined stainless steel autoclave. The synthesis was performed at 120 °C for 3 h in an electric oven. The white precipitate was filtered, washed with distilled water and dried in air naturally. Finally, the powder was calcined at different temperatures and times in air for various samples. The corresponding working electrodes were fabricated according to our previous paper.²⁷

ZnO rod electrodes were prepared using a precursor template method. The clean fluorine-doped tin oxide (FTO) glass was dipped in a 5 mmol L⁻¹ ethanol solution of zinc acetate for 30 s, taken out, and then sintered at 350 °C for 20 min. 4.5 mmol of Zn(NO₃)₂·6H₂O and 1.5 mmol of C₆H₁₂N₄ were added to 30 mL of H₂O with continuous stirring for 10 minutes in a 50 mL stainless Teflon-lined autoclave. The FTO leaned against the inner wall of the autoclave. After hydrothermal treatment at 100 °C for 3 h in an electric oven, the FTO was took out and washed with deionized water.

The electrodes were immersed into N719 ethanol solution for 90 min. A magnetron sputter Pt mirror served as the counter electrode. The FTO substrate, film, and counter electrode constituted a sandwich-like open cell. A drop of the electrolyte was injected into the cell, which was composed of 30 mM I₂, 0.5 M tert-butylpyridine, 1.0 M 1-butyl-3-methylimidazolium iodide (BMIMI) and 50 mM LiI, in a mixed solvent of valeronitrile and acetonitrile (v/v, 15 : 85).

Characterization

The morphology was characterized using transmission electron microscopy (TEM, JEOL 3010, Japan) and scanning electron microscopy (SEM, FEI NOVA NanoSEM230, USA). Thermogravimetric (TG) analysis was performed on an American TA SDT Q600 analyzer at a temperature from 25 °C to 700 °C. X-ray diffraction (XRD) data were measured using an X-ray diffractometer with Cu-Kα radiation at 40 kV and 200 mA. To quantify the amount of adsorbed dye, N719 was desorbed into a 0.1 M NaOH aqueous solution, and the optical absorption spectra of the solution were collected using a UV-vis spectrophotometer. A Shimadzu UV-2550 UV-vis spectrometer was used to investigate the diffuse reflectance spectrum and optical absorption of the powder. The incident-photon-to-current conversion efficiency (IPCE) spectra were recorded on PEC-S20 (Peccell Technology). The photocurrent–voltage characteristics (J–V) were measured under a sunlight simulator with an active area of 0.132 cm² (Oriel 92251A-1000, AM 1.5 globe, 100 mW cm⁻²). Intensity-modulated photovoltage/photocurrent spectra (IMVS/IMPS) were measured using the electrochemical workstation (Zahner, Zennium).

Results and discussion

The SEM image of the precursor shows that its morphology is a sheet with smooth surfaces (see ESI, Fig. S1a†). The XRD peaks of the precursor presented in Fig. S2a (ESI†) are attributed to hydrozincite Zn₅(OH)₆(CO₃)₂ (JCPDS #72-1100). ZnO samples with different morphologies were generated after annealing from the precursor at different treating temperatures and times. When the precursor was calcined at 450 °C for 30 minutes, porous ZnO nanosheets were fabricated. The sheet morphology is maintained, but numerous nanopores are formed in the sheet as shown in Fig. S1b.† This could be ascribed to the release of H₂O and CO₂ during the heat treatment of Zn₅(OH)₆(CO₃)₂ (see ESI, Fig. S3†). When the calcination temperature increases to 500 °C, the as-synthesized powder displays a chain feature with a diameter of 150–200 nm and length of 10–15 μm, as shown in Fig. 1a (labeled as P1). The enlarged SEM image reveals that the building blocks for the chain are particle structures with a diameter of 50–70 nm (Fig. 1b). It is noteworthy that thicker chains with a sub-micrometer diameter are composed of simple chains (marked by an arrow in Fig. 1c) which could generate an efficient scattering center to enhance the light-harvesting capability. The TEM images in Fig. 2a and b further show that the single ZnO chain and thicker sub-micrometer chains are constructed of interconnected nanoparticles with a diameter of 50–70 nm. Fig. 2c shows the HRTEM image taken from the space between two nanoparticles. A lattice spacing of 0.26 nm could be attributed to the (101) plane of hexagonal ZnO. For comparison, the ZnO nanoparticles with the diameter of 50–70 nm were observed after calcination in air at 550 °C for 3 h (labeled as P2, see ESI, Fig. S1c and d†). Fig. S1f (ESI†) shows a cross-sectional SEM image of the synthesized ZnO rod photoanodes (labeled as P3), showing vertically oriented ZnO arrays grown on the substrate with 12 μm in average length and 500–600 nm in diameter. The XRD patterns of all samples correspond well to hexagonal ZnO (a = b = 0.3249 nm, c = 0.5205 nm, α = β = 90°, γ = 120°, JCPDS card file no. 89-0511) (see ESI, Fig. S2b†).

Fig. 1 SEM images of: (a–c) the ZnO chains at different magnifications. (d) Cross section of the ZnO chain photoanode. (e) Top view of the ZnO chain photoanode. (f) High magnification view from the squared region of part (d). Inset: the scale bar is 200 nm in (b).

Fig. 2 (a and b) TEM and (c) HRTEM images of the ZnO chains.

Fig. 3 compares the effect of the ZnO morphologies on performance and their corresponding characteristics, open-circuit voltage (V_oc), fill factor (FF), short-circuit current (J_sc) and η, are tabulated in the inset table. The DSSCs assembled with ZnO chains exhibit an η of 5.06% with a J_sc of 14.53 mA cm⁻², a V_oc of 0.57 V, and a FF of 60.92%, which is higher than the η of ZnO nanoparticles (P2, 3.62%) and ZnO rods (P3, 2.78%). The remarkable improved η is the result from the larger J_sc, compared with that obtained from P2 (10.97 mA cm⁻²) and P3 (8.27 mA cm⁻²). In the given ZnO/dye/electrolyte system, the variation in the J_sc could be attributed to the dye adsorption amount or/and light-harvesting efficiency.²⁹ The surface areas of P1, P2 and P3 are 18.26 m² g⁻¹, 27.29 m² g⁻¹ and 7.32 m² g⁻¹; then, the corresponding amount of dye adsorption is 4.6 × 10⁻⁸ mol cm⁻², 6.2 × 10⁻⁸ mol cm⁻² and 2.2 × 10⁻⁸ mol cm⁻², respectively. The nanoparticle constituted chains can provide chains with a larger surface area than ZnO rods with a similar length, and are responsible for the improved photocurrent. Since the dye absorption amount on P1 is less than that of P2, the enhanced J_sc value for P1 should be due to the light-harvesting efficiency. To analyze in more detail the light-harvesting efficiency, diffuse-reflection spectroscopy was conducted (Fig. 4a). It is apparent that the intensity increases in the order of P3, P2 and P1 at wavelengths from 450 to 800 nm, which can be explained due to the different structures. Firstly, it is believed that resonant scattering can occur when the medium size is comparable to the wavelength of incident light.³⁰ Herein, the thicker chains with a sub-micrometer diameter in P1 films could act as an efficient scattering center to improve the light-harvesting efficiency.³¹ Secondly, prior studies have demonstrated that admixing large particles into nanocrystalline films could enhance the light-harvesting capability of photoelectrodes.^32,33 The P1 films assembled by staggered thick and thin chains can achieve the same effect (Fig. 1e).³⁴ Thirdly, the high magnification cross-sectional view of a 12 μm-thick chain film photoanode indicates a multilayer morphology aligned horizontally (Fig. 1f). The incident light goes through the photoanode layer by layer and, at the same time, scatters in them, which increases the optical path length inside the DSSC and the opportunities for light absorption; furthermore, to confirm our conclusions, IPCE spectra were collected (Fig. 4b). The increased IPCE value at the 370–600 nm wavelengths could be attributed to a higher dye loading amount. At long wavelengths (600–750 nm), the higher IPCE values could be ascribed to the improved scattering effect of the film.³⁵ The IPCE of ZnO chain-based films is higher at 550–750 nm than that of nanoparticles, suggesting that the J_sc increment of ZnO chains-based films is attributed to the improved light scattering effect.

Fig. 3 J–V curves for cells based on different ZnO electrodes. Photoanode thickness: 12 μm.

Fig. 4 (a) Diffuse reflectance spectra and (b) IPCE curves of the different ZnO electrodes.

To investigate the electron transfer properties and charge recombination reaction, IMPS/IMVS measurements were carried out under an irradiation intensity from 30 to 150 W m⁻² (Fig. 5). The electron transport time (electron lifetime) can be estimated from the equations τ_d = 1/2πf_d (τ_r = 1/2πf_r), where f_d (f_r) is the characteristic frequency at the minimum of the IMPS (IMVS) imaginary component (Fig. 5a and b).³⁶ Obviously, the τ_d of P1-based DSSCs is similar to the value of the P3-based DSSCs and considerably shorter than that of the P2-based DSSCs under diverse light intensities, implying a faster transport rate than that of P2. This result verified that the ZnO chains are in favour of the electron transport compared to the disordered ZnO nanoparticles. The τ_r based on P1 photoanode is longer than that of P2, indicating the longer electron lifetime and is responsible for the higher V_oc of the former.³⁷ This result is attributed to fewer nanocrystalline boundaries and electron trapping sites for ZnO chains compared with ZnO nanoparticles.¹⁴ It is generally known that the surface charge trap-site density could have a great impact on charge recombination. More trap sites would cause faster charge recombination. Compared with P2, P1 has relatively fewer grain boundaries and crystal defects, which results in a smaller trap site density, so reduced charge recombination can be expected.³⁸ Furthermore, the electron diffusion coefficient (D_n = d²/(4τ_d), d: film thickness) is highest in P3 and lowest in P2 (Fig. 5c). The higher the value of D_n, the faster the transmission speed of the photogenerated electrons to the anode contact.³⁹ Electrons in the ZnO chain structure are slightly lower than those in ZnO rods but faster than those in ZnO nanoparticles. This is because the ZnO rod and chain photoanodes can provide a direct transport pathway for the rapid collection of electrons.^32,39 In contrast, ZnO nanoparticles would offer more traps in the nanocrystalline boundaries, in which the electron could be trapped during the transmission to some extent.¹⁴ The effective electron diffusion length (L_n = (D_nτ_r)^1/2) of P1 is longer than that of P2 at different light intensities, which suggests that chain-based DSSCs are obviously superior to nanoparticle-based DSSCs for a given recombination loss, associated with a high η.⁴⁰ Although ZnO rod-based DSSCs demonstrate a superior electron transport velocity compared to ZnO chain-based DSSCs, the insufficient amount of dye loading limits the generation of a J_sc, leading to a relatively lower η; despite having larger surface areas than ZnO chain-based devices, ZnO nanoparticle-based devices show lower efficiencies, which are ascribed to the inferior light scattering capacity for boosting the light-harvesting efficiency compared to that of ZnO chains. Having considered all of the factors above, the significant improvement of η for the nanoparticle-based hierarchical ZnO chains can be ascribed to it taking the advantages of both the chains and nanoparticles. A superior light scattering ability, slower electron recombination rate, faster electron transport rate are responsible for the enhanced η.

Fig. 5 (a) Transport time, (b) lifetime, (c) electron diffusion coefficients and (d) the effective electron diffusion length based on the different photoelectrodes.

Conclusions

In summary, hierarchical ZnO chains composed of nanoparticles were prepared as photoanodes for DSSCs via a facile method. Experimental research showed that ZnO samples with different morphologies derive from different treating temperatures and times. ZnO chains with the maximum energy conversion efficiency were formed at 500 °C for 0.5 h by annealing the precursor. The investigations revealed that the nanoparticle constituted chains can provide chains with a larger surface area than the 1D ZnO rods with a similar length, which was responsible for the improved photocurrent. Meanwhile, the chains can provide direct electrical pathways for the collection of photogenerated electrons and present a slower electron recombination rate and faster electron transport rate than nanoparticles due to the presence of far fewer surface defects and grain boundaries than the latter. The DSSCs fabricated from such hierarchical structures give a conversion efficiency of 5.06%, far higher than 2.78% for ZnO rod-based DSSCs, and 3.62% for ZnO nanoparticle-based DSSCs, as a result of taking the advantages of both chains and nanoparticles.

Acknowledgements

This work was supported by the 973 Programs (No. 2011CB933303, 2014CB239302, 2013CB632404), NSFC (No. 21301101 and 11174129), Natural Science Foundation of Jiangsu Province (No. BK2012015 and BK2011056), Jiangsu Technical support plan – industrial parts (BE2012089), Kunshan New industries multiplication plan science and technology special Fund (KX1202), Natural Science Foundation of Nanyang Normal University (No. ZX2014040 and QN2015010).

Notes and references

1. Q. Zhang, C. S. Dandeneau, X. Zhou and G. Cao, Adv. Mater., 2009, 21, 4087.
2. H. Pan, N. Misra, S. H. Ko, C. P. Grigoropoulos, N. Miller, E. E. Haller and O. Dubon, Appl. Phys. A, 2009, 94, 111.
3. J. A. Anta, E. Guillén and R. Tena-Zaera, J. Phys. Chem. C, 2012, 116, 11413.
4. Q. Zhang and G. Cao, J. Mater. Chem., 2011, 21, 6769.
5. Z. Dong, X. Lai, J. E. Halpert, N. Yang, L. Yi, J. Zhai, D. Wang, Z. Tang and L. Jiang, Adv. Mater., 2012, 24, 1046.
6. Q. Zhang, T. Chou, B. Russo, S. A. Jenekhe and G. Cao, Adv. Funct. Mater., 2008, 18, 1654.
7. H. Dong, L. Wang, R. Gao, B. Ma and Y. Qiu, J. Mater. Chem., 2011, 21, 19389.
8. C. Wu and J. Wu, J. Mater. Chem., 2011, 21, 13605.
9. Y. Lai, C. Lin, H. Chen, J. Chen, C. Kung, R. Vittal and K. C. Ho, J. Mater. Chem., 2010, 20, 9379.
10. S. B. Ambade, R. S. Mane, S.-H. Han, S.-H. Lee, M.-M. Sung and O.-S. Joo, J. Photochem. Photobiol., A, 2011, 222, 366.
11. G. Pérez-Hernández, A. Vega-Poot, I. Pérez-Juárez, J. M. Camacho, O. Arés, V. Rejón, J. L. Peña and G. Oskam, Sol. Energy Mater. Sol. Cells, 2012, 100, 21.
12. K. Keis, E. Magnusson, H. Lindstrom, S. E. Lindquist and A. Hagfeldt, Sol. Energy Mater. Sol. Cells, 2002, 73, 51.
13. T. Yoshida, J. B. Zhang, D. Komatsu, S. Sawatani, H. Minoura, T. Pauporte, D. Lincot, T. Oekermann, D. Schlettwein, H. Tada, D. Wohrle, K. Funabiki, M. Matsui, H. Miura and H. Yanagi, Adv. Funct. Mater., 2009, 19, 17.
14. M. Law, L. E. Greene, J. C. Johnson, R. Saykally and P. D. Yang, Nat. Mater., 2005, 4, 455.
15. K. Zhu, N. R. Neale, A. Miedaner and A. J. Frank, Nano Lett., 2007, 7, 69.
16. D. Chen, H. Zhang, S. Hu and J. H. Li, J. Phys. Chem. C, 2008, 112, 117.
17. M. Guo, P. Diao and S. M. Cai, Chem. Lett., 2004, 15, 1113.
18. A. B. F. Martinson, J. W. Elam, J. T. Hupp and M. J. Pellin, Nano Lett., 2007, 7, 2183.
19. G. Chen, K. Zheng, X. Mo, D. Sun, Q. Meng and G. Chen, Mater. Lett., 2010, 64, 1336.
20. X. D. Wang, Y. Ding, C. J. Summers and Z. L. Wang, J. Phys. Chem. B, 2004, 108, 8773.
21. J. B. Baxter and E. S. Aydil, Appl. Phys. Lett., 2005, 86, 053114.
22. B. O’Regan and M. Grätzel, Nature, 1991, 353, 737.
23. S. H. Ko, D. Lee, H. W. Kang, K. H. Nam, J. Y. Yeo, S. J. Hong, C. P. Grigoropoulos and H. J. Sung, Nano Lett., 2011, 11, 666.
24. X. Kang, C. Jia, Z. Wan, J. Zhuang and J. Feng, RSC Adv., 2015, 5, 16678.
25. W. Chiu, C. H. Lee, H. Cheng, H. Lin, S. C. Liao, J. M. Wu and W. F. Hsieh, Energy Environ. Sci., 2009, 2, 694.
26. İ. Şişman, M. Can, B. Ergezen and M. Biçer, RSC Adv., 2015, 5, 73692.
27. Z. Li, Y. Zhou, J. Zhang, W. Tu, Q. Liu, T. Yu and Z. Zou, Cryst. Growth Des., 2012, 12, 1476.
28. S. Yodyingyong, Q. Zhang, K. Park, C. S. Dandeneau, X. Zhou, D. Triampo and G. Cao, Appl. Phys. Lett., 2010, 96, 073115.
29. J. Qian, P. Liu, Y. Xiao, Y. Jiang, Y. Cao, X. Ai and H. Yang, Adv. Mater., 2009, 21, 3663.
30. P. E. Wolf and G. Maret, Phys. Rev. Lett., 1985, 55, 2696.
31. S. Hore, P. Nitz, C. Vetter, C. Prahl, M. Niggemann and R. Kern, Chem. Commun., 2005, 2011.
32. Q. Zhang, T. P. Chou, B. Russo, S. A. Jenekhe and G. Cao, Angew. Chem., 2008, 120, 2436.
33. S. Nishimura, N. Abrams, B. A. Lewis, L. I. Halaoui, T. E. Mallouk, K. D. Benkstein, J. van de Lagemaat and A. J. Frank, J. Am. Chem. Soc., 2003, 125, 6306.
34. Y. Rui, Y. Li, H. Wang and Q. Zhang, Chem.–Asian J., 2012, 7, 2313.
35. M. Grätzel, Inorg. Chem., 2005, 44, 6841.
36. H. X. Wang, P. G. Nicholson, L. Peter, S. M. Zakeeruddin and M. Grätzel, J. Phys. Chem. C, 2010, 114, 14300.
37. Y. Wang, W. Zhao, X. Li and D. Li, Electrochim. Acta, 2015, 151, 399.
38. I. Shiyanovskaya and M. Hepel, J. Electrochem. Soc., 1999, 146, 243.
39. Y. Cui, L. Zhang, K. Lv, G. Zhou and Z. Wang, J. Mater. Chem. A, 2015, 3, 4477.
40. M. Kim, K.-W. Lee, L. Jang, D. Jeon, J. Ju, H. Yun and I. Lee, Electrochim. Acta, 2014, 133, 610.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra21847k

---