Ultra-small TiO₂ nanowire forests on transparent conducting oxide for solid-state semiconductor-sensitized solar cells

Abstract

One-dimensional TiO₂ nanostructures that have large specific surface area have broadened their applications in solar cells and water splitting, but their synthesis still remains a challenge. In this report, vertically ordered rutile TiO₂ nanowires with an ultra-high coverage density of 2.4 × 10¹¹ cm⁻² and ultra-small width of barely ∼16 nm were synthesized on transparent conducting oxide by a facile solvothermal reaction using methanol and aqueous hydrochloride as solvent. The nanowires were fabricated with Sb₂S₃ into solid-state solar cells, which yielded a power conversion efficiency of 2.03%. Such a photoanode showed reduced electron recombination, due to moderate wire fusion at the bottom.

---

Vertically aligned one-dimensional (1-D) nanostructures, including nanowires (NWs), nanotubes and nanorods, when used as photoanodes in photovoltaics, are believed to significantly improve electron diffusion length, promote transport rate and increase the survivability of photogenerated electrons by providing a direct conduction pathway.^1–6 However, the insufficient internal surface area of 1-D nanostructures constrains the sensitizer loading, resulting in low light harvesting efficiency.⁷ Besides the growth of branches and leaves from the trunk of 1-D nanostructures to increase the internal surface area,^8–11 synthesis of hierarchical structures that have smaller diameters and larger coverage density is expected to be another effective way.

In this paper, we report a facile solvothermal method to grow large coverage density, vertically-aligned rutile TiO₂ nanowire arrays directly on transparent conducting oxides, using methanol and aqueous hydrochloride as solvent. Such nanostructures bear a coverage density of 2.4 × 10¹¹ cm⁻² which is one to two orders of magnitude larger than that in the previous reported literatures,^12–15 and width of barely ∼16 nm. So far, to the best of our knowledge, this is the smallest TiO₂ NW arrays grown perpendicularly on transparent conducting oxides. Moreover, these nanostructures were incorporated with Sb₂S₃ as sensitized film, P3HT as hole-transporting material, and gold as counter electrode to fabricate into solid-state solar cells (SSSCs) (see ESI† for experimental details). Such SSSCs yielded power conversion efficiency (PCE) of 2.03% and showed superior electron transporting due to the moderate wire fusion at the base to form a continuous underlayer which cut off the recombination path.

The synthesis of TiO₂ NWs is detailed in ESI.† The TiO₂ NWs were characterized using scanning electron microscopy (SEM), high resolution transmission electron microscopy (HRTEM), X-ray diffraction (XRD) and Raman. TiO₂ NW films with different growth durations were obtained. Fig. 1a and b and S1† are SEM images of the TiO₂ NW film with growth duration of 4 hours. Cross-sectional SEM images show that the vertically oriented dense NWs grow starting from the FTO substrate. The length of the NWs is about ∼1.9 μm and the width is narrowly distributed around 16 nm (inset in Fig. 1b). The 16 nm-width is the smallest width to date for vertical NWs grown on transparent conducting oxides. In addition, the coverage density reaches ∼2.4 × 10¹¹ cm⁻², which is one to two orders of magnitude larger than that reported in the literatures.^12–15 Generally, the high coverage density and small width of the NWs are beneficial to increase the specific surface area of the TiO₂ nanostructures for sensitizer loading. Notably, such good morphology only comes forth in the case of using methanol as solvent. When replacing methanol with ethanol or isopropanol while keeping other conditions constant, totally fused TiO₂ films are obtained (Fig. S2a and b†) (the fusion degree can be tuned by varying the ratio of alcohol to water, see ESI† for details). As the growth duration elongates, the thickness of the NWs film increases (Fig. 1c), accompanied with increasing wire fusion degree throughout the whole film (Fig. S3†). Fig. 1d presents the XRD patterns of such TiO₂ NW film. The peaks agree well with the standard rutile structure (JCPDS 086-0147). Notably, with respect to the intensity of the (110) peak, the (002) peak intensity at 62.8° increases from 6.7% in the standard pattern to 109.9% for the as-prepared NWs, indicating that the TiO₂ NWs grow preferentially in [001] direction. Fig. 1e shows an individual NW with width of ∼16 nm determined by HRTEM. The selected area electron diffraction (SAED) pattern and lattice fringes with interplanar spacings of d₁₁₀ = 3.2 Å and d₀₀₁ = 2.9 Å are also consistent with that of rutile TiO₂ (Fig. 1e).¹³ Moreover, the Raman peaks at 242, 446 and 610 nm are pertaining to the Raman shift characteristics of rutile TiO₂ (Fig. 1f).¹⁶

Fig. 1 (a) Cross sectional SEM images of the TiO₂ NWs synthesized with 4 hours (NWs-4 h). (b) Top-view SEM images of NWs-4 h (inset is the width distribution). (c) The relationship between the length and synthesis durations. (d) XRD patterns of NWs-4 h. (e) High resolution TEM image of a NW-4 h. (f) Raman spectra of NWs-4 h.

Due to less wire fusion and the promising length of ∼1.9 μm for the TiO₂ NW arrays at 4 h growth duration, the TiO₂ NWs were employed as photoanodes to fabricate photovoltaics. Sb₂S₃ as a sensitizer exhibited well photovoltaic behavior, showing PCE of 5.13% (ref. 17) and 4.5% (ref. 5) by using P3HT as HTM on TiO₂ nanoparticle and nanorods based devices, respectively. Here, Sb₂S₃ was synthesized as sensitizer due to its high absorption coefficiency (1.8 × 10⁵ cm⁻¹ at 450 nm) and suitable energy band gap (∼1.7 eV) for visible light harvesting.^17–19

The Sb₂S₃ deposition time was investigated for 1 h, 2 h, and 3 h. There is too little Sb₂S₃ deposited on the NWs when deposition time is 1 h (not shown) while too much when deposition time is 3 h (Fig. S4†). Fig. 2a and b present the SEM images of NWs after 2 h deposition of Sb₂S₃. It is found that the interspaces between NWs are partly filled with Sb₂S₃, which indicates that Sb₂S₃ managed to grow around the side surface of the NWs forest. The HRTEM image clearly shows that each NW is uniformly covered by a 7 nm-thick Sb₂S₃ layer (Fig. 2d). In addition, the lattice interplanar spacing of 3.56 Å matches well with that of (130) interplanar spacing of stibnite Sb₂S₃.²⁰ The XRD pattern in Fig. 2c is also indexed well with the peaks of stibnite Sb₂S₃ (JCPDS 042-1393). The peak at 45.9° belongs to Sb₂O₃ due to intentional oxidation when cooled in air after annealing in N₂ to retard recombination in the photovoltaic application.²¹

Fig. 2 (a) Cross-sectional SEM image of TiO₂ NWs after 2 h deposition of Sb₂S₃ [Sb₂S₃-NW (2 h)]. (b) Top view SEM image of Sb₂S₃-NW (2 h). (c) XRD pattern of Sb₂S₃-NW-FTO. (d) HRTEM image of a single NW deposited with Sb₂S₃.

After deposition of Sb₂S₃, P3HT and PEDOT:PSS were spin-coated onto the semiconductor-sensitized film. Before spin-coating of PEDOT:PSS, Energy dispersive X-ray spectroscopy (EDS) element mapping was performed to evaluate the filling of P3HT. As is shown in Fig. 3a, for a freshly exposed cross-sectional surface, elements of Sb and C are evenly distributed throughout the whole cross section of the film. Only slightly less C is observed in the area close to the substrate. Together with the individual TEM image of Sb₂S₃ coated NW in Fig. 2d, the results show that Sb₂S₃ and P3HT penetrated well through the dense NW film. Successful filling of P3HT could also be verified from the cross sectional SEM image of the device in Fig. 3c, where the smooth polymer could be clearly observed in the cross section SEM image.

Fig. 3 (a) EDS element mapping images of Sb₂S₃-NW-FTO after spin-coated with P3HT. (b) Illustration of band diagram of the device. (c) Cross-sectional SEM image of the device.

Fig. 3b demonstrates the band diagram of the device. The devices were fabricated using the NWs deposited with Sb₂S₃ for 1 h, 2 h, and 3 h (denoted as Device-1 h, Device-2 h, and Device-3 h). Fig. S5† presents the absorption spectra of the Sb₂S₃-NW films. The absorption of the visible light increases as the Sb₂S₃ deposition duration increases. However, the device performance does not follow the same trend. The current density–voltage curves (J–V curve) are shown in Fig. 4a, and the resultant characteristics are summarized in Table 1. All characteristics dramatically increase when the deposition time increases from 1 h to 2 h, especially for short circuit current (J_sc), fill factor (FF), and PCE. Such significant improvement can be explained by the increasing Sb₂S₃ nanocrystals deposited on the NWs harvesting more light as the deposition time increases. However, the performance deteriorates as the deposition time increased to 3 h. This is because the interspaces between the NWs are completely filled by excessive Sb₂S₃, and the continuous compact Sb₂S₃ layer on the top surface prevents P3HT from penetrating in (Fig. S4†). This leads to inefficient charge separation and poor photovoltaic performance. External quantum efficiency (EQE) as a function of wavelength in Fig. 4b shows that the average EQE of the best device in the wavelength range from 400 nm to 650 nm is up to 35%.

Fig. 4 (a) J–V characteristics of SSSCs using Sb₂S₃-NW-FTO as the photoanodes with different Sb₂S₃ deposition times. (b) External quantum efficiency (EQE) of the SSSC using Sb₂S₃-NW-FTO as the photoanodes with 2 h Sb₂S₃ deposition (target device). (c) Nyquist plots of the target device and the reference device at 600 mV forward bias. (d) Impedance modulus and Bode phase angle versus frequency plots.

Table 1 Photovoltaic characteristics of SSSCs that use Sb₂S₃-NW forest as the photoanodes with different deposition times of Sb₂S₃

Deposition time	Voc (mV)	Jsc (mA cm^−2)	FF (%)	PCE (%)
1 h	470.41	3.39	26.39	0.42
2 h	516.35	7.54	52.15	2.03
3 h	514.67	5.42	32.64	0.91

---

To better understand the advantage of using such NWs as photoanodes in photovoltaics, electrochemical impedance spectroscopy (EIS) of the device in this paper (target device) was compared with another photovoltaic device similar to the reported one⁵ (reference device, Fig. S6†) that using a vertically aligned TiO₂ NR arrays as photoanode (Fig. S2f, see ESI† for fabrication details). The performance of the target device reported here is inferior to that of the reference device, which might because less Sb₂S₃ is anchored onto the NWs due to that higher coverage density which spares less space for sensitizer loading. Fig. 4c shows the EIS spectra of the two devices under dark condition at 600 mV forward bias. Two semicircles are observed in the Nyquist plots. In the moderate frequency range from 0.1 to 10⁴ s⁻¹, both Nyquist plots shape a large semicircle. In general, the size of this semicircle represents the recombination resistance between electrons in the photoanode and holes in the electrolyte.^22–25 Surprisingly, the semicircle of the target device is twice bigger than that of the reference device. In addition, Bode plot of the target device in Fig. 4d also shows left-shift of the characteristic frequency. Straight forwardly, the impedance of the target device in the frequency range from 0.1 to 10³ s⁻¹ is much larger than that of the reference device. All these phenomena imply increased recombination resistance exists in the target device, which reduces the recombination. As also verified by the relative less P3HT distributed at the NW forest bottom (Fig. 3a), we attributed this to the relatively more seriously fused part at the bottom close to the FTO substrate (Fig. S1†), acting as recombination blocking layer, caused by the larger coverage density than that of the TiO₂ NRs for the reference device.

In conclusion, this paper overcame the relative small surface area drawback of 1-D materials by reporting an ultra-high coverage density and ultra-small width 1-D TiO₂ nanowire forests. Such vertically ordered rutile TiO₂ NW forests were synthesized on transparent conducting oxide by a facile solvothermal reaction by using methanol and aqueous hydrochloride as solvent. The width of the NWs is only 16 nm and length is 1.9 μm. The coverage density is up to 2.4 × 10¹¹ cm⁻². The morphology is synthesizing solvent dependant. By incorporating such NW arrays as photoanode, Sb₂S₃ as light absorber and P3HT as HTM, the photovoltaic device yielded a PCE up to 2.03%. Such nanostructures could also be used in photocatalysis and sensing application.

Acknowledgements

This work was supported by the State Scholarship Fund of China Scholarship Council and the National Science Foundation of China (21273241).

Notes and references

1. A. Kongkanand, K. Tvrdy, K. Takechi, M. Kuno and P. V. Kamat, J. Am. Chem. Soc., 2008, 130, 4007–4015.
2. W.-T. Sun, Y. Yu, H.-Y. Pan, X.-F. Gao, Q. Chen and L.-M. Peng, J. Am. Chem. Soc., 2008, 130, 1124–1125.
3. K. S. Leschkies, R. Divakar, J. Basu, E. Enache-Pommer, J. E. Boercker, C. B. Carter, U. R. Kortshagen, D. J. Norris and E. S. Aydil, Nano Lett., 2007, 7, 1793–1798.
4. D. R. Baker and P. V. Kamat, Adv. Funct. Mater., 2009, 19, 805–811.
5. J. C. Cardoso, C. A. Grimes, X. Feng, X. Zhang, S. Komarneni, M. V. Zanoni and N. Bao, Chem. Commun., 2012, 48, 2818–2820.
6. H. S. Kim, J. W. Lee, N. Yantara, P. P. Boix, S. A. Kulkarni, S. Mhaisalkar, M. Grätzel and N. G. Park, Nano Lett., 2013, 13, 2412–2417.
7. Q. Zhang, C. S. Dandeneau, X. Zhou and G. Cao, Adv. Mater., 2009, 21, 4087–4108.
8. L. Passoni, F. Ghods, P. Docampo, A. Abrusci, J. Martí-Rujas, M. Ghidelli, G. Divitini, C. Ducati, M. Binda, S. Guarnera, A. L. Bassi, C. S. Casari, H. J. Snaith, A. Petrozza and F. D. Fonzo, ACS Nano, 2013, 7, 10023–10031.
9. 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, 666–671.
10. Y. Qiu, K. Yan, H. Deng and S. Yang, Nano Lett., 2012, 12, 407–413.
11. F. Sauvage, F. D. Fonzo, A. Li Bassi, C. S. Casari, V. Russo, G. Divitini, C. Ducati, C. E. Bottani, P. Comte and M. Graetzel, Nano Lett., 2010, 10, 2562–2567.
12. W.-P. Liao and J.-J. Wu, J. Mater. Chem., 2011, 21, 9255–9262.
13. B. Liu and E. S. Aydil, J. Am. Chem. Soc., 2009, 131, 3985–3990.
14. X. Feng, K. Shankar, O. K. Varghese, M. Paulose, T. J. Latempa and C. A. Grimes, Nano Lett., 2008, 8, 3781–3786.
15. W.-P. Liao, S.-C. Hsu, W.-H. Lin and J.-J. Wu, J. Phys. Chem. C, 2012, 116, 15938–15945.
16. P. Sun, X. Zhang, C. Wang, Y. Wei, L. Wang and Y. Liu, J. Mater. Chem. A, 2013, 1, 3309.
17. J. A. Chang, J. H. Rhee, S. H. Im, Y. H. Lee, H. J. Kim, S. I. Seok, M. K. Nazeeruddin and M. Grätzel, Nano Lett., 2010, 10, 2609–2612.
18. S. H. Im, C. S. Lim, J. A. Chang, Y. H. Lee, N. Maiti, H. J. Kim, M. K. Nazeeruddin, M. Gratzel and S. I. Seok, Nano Lett., 2011, 11, 4789–4793.
19. M. Y. Versavel and J. A. Haber, Thin Solid Films, 2007, 515, 7171–7176.
20. G. Wang and C. L. Cheung, Mater. Lett., 2012, 67, 222–225.
21. Y. Itzhaik, O. Niitsoo, M. Page and G. Hodes, J. Phys. Chem. C, 2009, 113, 4254–4256.
22. J. Lagemaat, N. G. Park and A. J. Frank, J. Phys. Chem. B, 2000, 104, 2044–2052.
23. Q. Wang, J.-E. Moser and M. Grätzel, J. Phys. Chem. B, 2005, 109, 14945–14953.
24. M. Wang, C. Grätzel, S.-J. Moon, R. Humphry-Baker, N. Rossier-Iten, S. M. Zakeeruddin and M. Grätzel, Adv. Funct. Mater., 2009, 19, 2163–2172.
25. X. Xin, M. He, W. Han, J. Jung and Z. Lin, Angew. Chem., Int. Ed., 2011, 50, 11739–11742.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details, results obtained by using ethanol or isopropanol as synthesizing solvents, information about the reference device. See DOI: 10.1039/c4ra09134e

---