High-Efficiency Solid-State Dye-Sensitized Solar Cells Based on

Letter pubs.acs.org/NanoLett

High-Efficiency Solid-State Dye-Sensitized Solar Cells Based on TiO2Coated ZnO Nanowire Arrays Chengkun Xu, Jiamin Wu, Umang V. Desai, and Di Gao* Department of Chemical and Petroleum Engineering, University of Pittsburgh, Pittsburgh, Pennsylvania 15261, United States S Supporting Information *

ABSTRACT: Replacing the liquid electrolytes in dye-sensitized solar cells (DSCs) with solid-state hole-transporting materials (HTMs) may solve the packaging challenge and improve the long-term stability of DSCs. The efficiencies of such solid-state DSCs (ssDSCs), however, have been far below the efficiencies of their counterparts that use liquid electrolytes, primarily due to the challenges in filling HTMs into thick enough sensitized films based on sintered TiO2 nanoparticles. Here we report fabrication of high-efficiency ssDSCs using multilayer TiO2-coated ZnO nanowire arrays as the photoanodes. The straight channel between the vertically aligned nanostructures combined with a newly developed multistep HTM filling process allows us to effectively fill sensitized films as thick as 50 μm with the HTMs. The resulting ss-DSCs yield an average power conversion efficiency of 5.65%. KEYWORDS: Solid-state solar cells, photoanode, ordered nanostructure, ZnO nanowires, sensitized film effort has been made to improve the pore filling, to effectively fill thick nanoparticle-based mesoporous TiO2 films with HTMs remains challenging.16−18 One approach to solving the challenge of filling sensitized films with solid HTMs is to replace the sintered nanoparticlebased mesoporous TiO2 film with vertically ordered nanostructures. Recently, much effort has been devoted to developing vertically ordered nanostructures for DSCs.19−24 Such vertically ordered structures provide a direct pathway for electron transport and therefore reduce the probability of electron recombination. Moreover, the ordered structures provide a straight channel for filling the pores of the sensitized film with electrolytes. Despite the considerable amount of effort, synthesizing vertically ordered nanostructures with a sufficiently high internal surface area that can be used as DSC’s photoanodes remains challenging. Consequently, DSCs based on such photoanodes, to date, have suffered from low internal surface area and thus insufficient dye loading, resulting in low efficiencies.19−22,25,26 Recently, we have developed a process for synthesizing multilayer arrays of vertically aligned ZnO nanowires directly on transparent conductive oxides.27 The process involves alternate cycles of nanowire growth and self-assembled monolayer (SAM) coating processes. The SAM coating applied onto the ZnO nanowires between the different stages of nanowire growth makes the nanowire surface hydrophobic and prevents the reactants in the aqueous solution from getting into the space between the nanowires, thus protecting the underneath layers of nanowires from growing wider and fusing

W

ith increasing concern over the twin crises of fossil fuel depletion and environmental degradation, dye-sensitized solar cells (DSCs) have received considerable interest as one of the most promising devices for converting solar energy to electricity at low cost.1,2 Up to 12% of power conversion efficiency has been demonstrated for DSCs,3 but such an efficiency is obtained by using liquid electrolytes in the cells, which may suffer from electrolyte leakage and corrosion problems, posing a challenge to the long-term stability of the cells. One way to address these issues is to replace the liquid electrolyte with solid-state hole transporting materials (HTMs), such as spiro-OMeTAD [2,20,7,70-tetrakis-(N,N-di-pmethoxyphenylamine)9,90-spirobifluorene].4−9 At present, the power conversion efficiencies of solid-state DSCs (ss-DSCs) have been far below the efficiencies of their counterparts that use liquid electrolytes. The best performing ss-DSCs to date use spiro-OMeTAD as HTM and sintered nanoparticle-based mesoporous TiO2 films as photoanodes.10 The highest efficiency of such ss-DSCs is obtained when the thickness of the mesoporous TiO2 films is about 2 μm, 5 times thinner than the thickness needed for sufficient light absorption.5,7,11 Incomplete filling of the mesoporous TiO2 films with HTMs has been identified as a major factor that limits the performance of the ss-DSCs when the thickness of the TiO2 films is beyond 2 μm.6,12−15 For example, the pore filling fraction (i.e., volume fraction of the pores filled by the HTMs) of spiro-OMeTAD is more than 60% for a 2−3 μm thick film13,14 but drops to ∼20−40% as the film thickness increases to 8−11 μm.14 The ineffective filling of the pores in thick TiO2 films with spiro-OMeTAD causes low hole injection efficiency from the dye cation to spiro-OMeTAD, short recombination lifetime of charge carriers, and poor hole transport through spiro-OMeTAD.14,15 While considerable © 2012 American Chemical Society

Received: January 31, 2012 Revised: March 29, 2012 Published: April 9, 2012 2420

dx.doi.org/10.1021/nl3004144 | Nano Lett. 2012, 12, 2420−2424

Nano Letters

Letter

together. This approach is able to effectively circumvent the wire fusion problem that commonly occurs in the processes of growing vertically aligned long nanowires in aqueous solutions. We have demonstrated that the internal surface area of a fourlayer array produced through this approach is more than five times larger than that of a single-layer array.27 The resulting high internal surface area of the vertically aligned nanowire arrays opens up opportunities for us to fabricate high-efficiency ss-DSCs. In this paper, we report fabrication of high-efficiency ss-DSCs using multilayer TiO2-coated ZnO nanowire arrays sensitized with Z907 dye as the photoanode and spiro-OMeTAD as the solid-state HTM. The straight channel between the vertically aligned nanostructures combined with a newly developed multistep HTM filling process allows us to effectively fill the sensitized film as thick as 50 μm with spiro-OMeTAD. The resulting ss-DSCs yield an average power conversion efficiency of 5.65%. Figure 1 shows a scanning electron microscopy (SEM) image of a four-layer ZnO nanowire array. The entire array is about 50

Figure 2. TEM analysis of TiO2-coated ZnO nanowires. (a) STEM elemental mapping image of TiO2-coated ZnO nanowires. (b) Linescan EDX profile obtained along the yellow dotted line in (a).

Figure 1. SEM image of a four-layer ZnO nanowire array. Scale bar, 20 μm.

Figure 3. SEM image of a four-layer TiO2-coated ZnO nanowire array filled with spiro-OMeTAD by the single-step process. Scale bar, 20 μm.

μm thick, and the 4 individual layers are 11, 15, 14, and 10 μm thick, respectively, from bottom to top. The four-layer ZnO nanowire array possesses a roughness factor (RF, defined as the actual surface area per unit projection area of the array) of about 510. To achieve better DSC photovoltaic properties, the ZnO nanowires were coated with a thin layer of TiO2 by a chemical bath deposition method (see Supporting Information). Figure 2a,b presents a scanning transmission electron microscopy (STEM) elemental mapping image of a TiO2coated ZnO nanowire and a line-scan energy dispersive X-ray (EDX) spectrum across the width of the nanowire, respectively, showing that the ZnO nanowire is uniformly coated with a 20− 30 nm thick layer of TiO2. Although the four-layer TiO2-coated ZnO nanowire array provides us with straight channels for infiltration of the HTM solution, to effectively fill such high aspect ratio nanostructures with HTMs is still challenging. After the dye adsorption procedure, we first tried to load the HTM, spiro-OMeTAD, into the array through a literature-reported process that had

been used for loading spiro-OMeTAD into TiO2 nanoparticlebased films.6,13,14 It involved one application of 200 mg/mL spiro-OMeTAD solution on the film, followed by spinning off and drying. Figure 3 shows a representative cross-section SEM image of the nanowire array filled with spiro-OMeTAD by this process. It is seen that spiro-OMeTAD has infiltrated down to the bottom of the array, almost 50 μm underneath the top surface. However, there is a significant number of voids observed in Figure 3, especially in the bottom three layers, indicating that this HTM filling process is not effective enough for high aspect ratio nanostructures, such as our four-layer nanowire arrays. The literature-reported HTM filling procedure involves dropping the HTM solution in excess amount, spinning the substrate, and drying. Based on the pore filling mechanism proposed by H. J. Snaith,13 the excess solution on top of the porous film acts as a reservoir during the filling process. As the solvent evaporates, the concentration of spiro-OMeTAD in the reservoir increases, and more spiro-OMeTAD diffuses into the 2421

dx.doi.org/10.1021/nl3004144 | Nano Lett. 2012, 12, 2420−2424

Nano Letters

Letter

using this procedure. Compared to the result shown in Figure 3, it is evident that filling of spiro-OMeTAD is significantly improved, and the thickness of the overlayer is decreased to ∼1 μm. We fabricated two groups of ss-DSCs using the four-layer TiO2-coated ZnO nanowire arrays: the first group of cells (cell nos. 1−8) are filled with spiro-OMeTAD by the literaturereported single-step method and the second group of cells (cell nos. 9−15) are filled with spiro-OMeTAD by our multistep method. The short circuit current density (Jsc), open circuit voltage (Voc), fill factor (FF), and power conversion efficiency (η) of the first and the second groups of cells are summarized in Tables 1 and 2, respectively. Figure 5 shows the I−V

pores. However, the viscosity of the solution also increases with the evaporation of the solvent, which decreases the diffusion rate of spiro-OMeTAD.16 As a result, the solution on top of the porous film forms a solid overlayer during the subsequent drying process before enough HTM diffuses into the pores. Formation of the overlayer makes it extremely difficult for the HTM to effectively fill the pores even with additional solution applied onto the porous film. The SEM image shown in Figure 3 is a representative result of such a process, where a ∼7 μm thick overlayer is present on top of the 4-layer array, which prevents the HTM from effectively filling the space between the nanowires. Based on the above observation, we developed a multistep HTM filling process. Instead of applying an excess volume of the HTM solution in one step, we apply the solution (in the same concentration of 200 mg/mL spiro-OMeTAD) in multiple steps, and in each step, only a small volume, about one-half of the total space occupied by the sensitized film, is applied. By applying the solution in a volume smaller than the pore volume of the array in each step, the overlayer formation is to a large extent avoided. Figure 4 presents a four-layer TiO2coated ZnO nanowire array filled with spiro-OMeTAD by

Table 1. Photovoltaic Performance Data of ss-DSCs using Single-Step HTM Filling Process cell no.

Jsc (mA/cm2)

Voc (V)

FF

η (%)

1 2 3 4 5 6 7 8 Average

11.9 12.0 11.9 11.7 11.3 11.6 12.1 11.5 11.75

0.714 0.730 0.673 0.691 0.695 0.673 0.718 0.719 0.702

0.488 0.517 0.498 0.514 0.521 0.522 0.511 0.520 0.511

4.15 4.52 3.99 4.16 4.09 4.07 4.42 4.30 4.21

Table 2. Photovoltaic Performance Data of ss-DSCs using Multistep HTM Filling Process cell no.

Jsc (mA/cm2)

Voc (V)

FF

η (%)

9 10 11 12 13 14 15 Average

12.2 12.2 12.4 12.3 12.0 12.5 11.9 12.2

0.788 0.792 0.798 0.790 0.786 0.786 0.778 0.788

0.568 0.589 0.594 0.591 0.608 0.571 0.591 0.587

5.46 5.68 5.86 5.74 5.73 5.61 5.48 5.65

Figure 5. I−V characteristics of two representative ss-DSCs fabricated by filling the four-layer TiO2-coated ZnO nanowire arrays through single- (cell no. 4) and multistep (cell no. 10) processes. Inset, IPCE versus wavelength plot.

Figure 4. SEM image of a four-layer TiO2-coated ZnO nanowire array filled with spiro-OMeTAD by the multistep process. Scale bar, 20 μm. 2422

dx.doi.org/10.1021/nl3004144 | Nano Lett. 2012, 12, 2420−2424

Nano Letters

Letter

characteristics of two representative cells, one from the first group (cell no. 4) and the other from the second group (cell no. 10), with plots of the incident photon-to-current conversion efficiency (IPCE) versus wavelength shown in the inset. Significantly, Jsc of both groups of cells are about 12 mA/ cm2, which is greater than 9−11 mA/cm2 reported for the best performing nanoparticle-based ss-DSCs.7,10,28 Compared to the porous film made of sintered nanoparticles, the vertically aligned multilayer nanowire array allows us to use much thicker sensitized films and thus more dyes in the ss-DSCs, while still allowing penetration and infiltration of the sensitized films by HTMs. The increased dye loading leads to improved lightharvesting efficiencies and thus larger Jsc. Comparison between the two groups of ss-DSCs listed in Tables 1 and 2 reveals that the effectiveness of HTM filling has significant effect on the device performance. Compared to the literature-reported single-step method, our multistep method is able to fill the cell more effectively with HTM, leaving much fewer voids in the HTM. Consequently, Jsc, FF, and Voc have increased, on average, from 11.75 to 12.2 mA/cm2, 0.511 to 0.587, and 0.702 to 0.788 V, respectively. These improvements lead to a significant increase of the average η from 4.21% to 5.65%. To better understand the impact of HTM filling effectiveness on the device performance, we used electrochemical impedance spectroscopy (EIS) to examine the cells made by the two HTM filling processes. Figure 6 shows EIS spectra of two representative ss-DSCs (under illumination of AM 1.5G simulated sunlight) with their HTM filled by the single- and the multistep methods, respectively. In the frequency range of 0.1−1 × 104 s−1, both Nyquist plots of the two cells are in a shape of a large semicircle. In general, the semicircle on the Nyquist plot in this frequency range is associated with the charge transfer across the photoanode/electrolyte interface, and the size of the semicircle represents the recombination resistance between electrons in the photoanode and holes in the electrolyte.29−32 Apparently, the semicircle representing the cell made by the multistep HTM filling process is larger than the other one, indicating a larger recombination resistance in the cell. The increased recombination resistance as a result of the multistep HTM filling process leads to longer electron recombination lifetime, which is manifested by the left-shift of the characteristic frequency in the Bode phase plots (Figure 6b). In addition, the reduced overlayer thickness on top of the array, as a result of the multistep HTM filling process, should decrease the series resistance of the cell. Both the reduced series resistance and the long electron lifetime contribute to the improvements in Jsc, Voc, FF, and thus η of ss-DSCs. The development of the multilayer nanowire arrays and multistep HTM filling process offers an appealing promise of highly efficient ss-DSCs. Further improvement of ss-DSCs may be enabled by simultaneous use of a variety of advanced strategies, such as high-extinction coefficient dyes,28 cosensitization and energy relay dyes,3,8,33 new HTMs and dopants,10,17,34,35 and other novel techniques.36−41 Further thickening the nanowire arrays may remain as an option to increase light absorption. However, it should be noted that the probability of electron−hole recombination also increases with the nanowire length, although no significant decline in the electron collection efficiency is noticed for such arrays as the thickness increases from 10 to 50 μm in DSCs that use liquid electrolytes.27

Figure 6. EIS spectra under illumination of AM 1.5G simulated sunlight of two representative ss-DSCs fabricated by filling the fourlayer TiO2-coated ZnO nanowire arrays through single- and multistep processes. (a) Nyquist plot. (b) Bode phase angle versus frequency plot.

In summary, we have fabricated high-efficiency ss-DSCs using four-layer TiO2-coated ZnO nanowire arrays as the photoanodes and spiro-OMeTAD as the solid-state HTM. Although the vertically aligned nanostructure provides us with a straight channel for infiltration of HTM solution, previously reported procedures for HTM filling are found to be ineffective for filling pores with high-aspect ratios. We have developed a multistep HTM filling procedure, which is able to effectively fill spiro-OMeTAD into the nanowire arrays as thick as 50 μm. More effective HTM filling is found to be able to increase the electron recombination lifetime and reduce the series resistance in ss-DSCs, both of which contribute to improvements in Jsc, Voc, FF, and thus η. The resulting ss-DSCs yield an average power conversion efficiency of 5.65%.

■

ASSOCIATED CONTENT

S Supporting Information *

Experimental details. This material is available free of charge via the Internet at http://pubs.acs.org. 2423

dx.doi.org/10.1021/nl3004144 | Nano Lett. 2012, 12, 2420−2424

Nano Letters

■

Letter

(28) Cai, N.; Moon, S.-J.; Cevey-Ha, L.; Moehl, T.; Humphry-Baker, R.; Wang, P.; Zakeeruddin, S. M.; Grätzel, M. Nano Lett. 2011, 11, 1452. (29) van de Lagemaat, J.; Park, N. G.; Frank, A. J. J. Phys. Chem. B 2000, 104, 2044. (30) Wang, Q.; Moser, J.-E.; Grätzel, M. J. Phys. Chem. B 2005, 109, 14945. (31) Wang, M.; Grätzel, C.; Moon, S.-J.; Humphry-Baker, R.; Rossier-Iten, N.; Zakeeruddin, S. M.; Grätzel, M. Adv. Funct. Mater. 2009, 19, 2163. (32) Xin, X.; He, M.; Han, W.; Jung, J.; Lin, Z. Angew. Chem., Int. Ed. 2011, 50, 11739. (33) Hardin, B. E.; Hoke, E. T.; Armstrong, P. B.; Yum, J.-H.; Comte, P.; Torres, T.; Frechet, J. M. J.; Nazeeruddin, M. K.; Gratzel, M.; McGehee, M. D. Nat Photon 2009, 3, 406. (34) Wang, H.; Zhang, X.; Gong, F.; Zhou, G.; Wang, Z.-S. Adv. Mater. 2012, 24, 121. (35) Kim, J.; Koh, J. K.; Kim, B.; Ahn, S. H.; Ahn, H.; Ryu, D. Y.; Kim, J. H.; Kim, E. Adv. Funct. Mater. 2011, 21, 4633. (36) Jang, Y. H.; Xin, X.; Byun, M.; Jang, Y. J.; Lin, Z.; Kim, D. H. Nano Lett. 2011, 12, 479. (37) Li, Q.; Zhao, J.; Sun, B.; Lin, B.; Qiu, L.; Zhang, Y.; Chen, X.; Lu, J.; Yan, F. Adv. Mater. 2012, 24, 945. (38) Ye, M.; Xin, X.; Lin, C.; Lin, Z. Nano Lett. 2011, 11, 3214. (39) Hwang, D.; Jo, S. M.; Kim, D. Y.; Armel, V.; MacFarlane, D. R.; Jang, S.-Y. ACS Appl. Mater. Interfaces 2011, 3, 1521. (40) Wang, J.; Lin, Z. Chem. Mater. 2009, 22, 579. (41) Balis, N.; Makris, T.; Dracopoulos, V.; Stergiopoulos, T.; Lianos, P. J. Power Sources 2012, 203, 302.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Hardin, B. E.; Snaith, H. J.; McGehee, M. D. Nat. Photonics 2012, 6, 162. (2) Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Pettersson, H. Chem. Rev. 2010, 110, 6595. (3) Yella, A.; Lee, H.-W.; Tsao, H. N.; Yi, C.; Chandiran, A. K.; Nazeeruddin, M. K.; Diau, E. W.-G.; Yeh, C.-Y.; Zakeeruddin, S. M.; Grätzel, M. Science 2011, 334, 629. (4) Bach, U.; Lupo, D.; Comte, P.; Moser, J. E.; Weissortel, F.; Salbeck, J.; Spreitzer, H.; Gratzel, M. Nature 1998, 395, 583. (5) Schmidt-Mende, L.; Zakeeruddin, S. M.; Gratzel, M. Appl. Phys. Lett. 2005, 86, 013504. (6) Schmidt-Mende, L.; Grätzel, M. Thin Solid Films 2006, 500, 296. (7) Snaith, H. J.; Moule, A. J.; Klein, C.; Meerholz, K.; Friend, R. H.; Grätzel, M. Nano Lett. 2007, 7, 3372. (8) Yum, J.-H.; Hardin, B. E.; Moon, S.-J.; Baranoff, E.; Nüesch, F.; McGehee, M. D.; Grätzel, M.; Nazeeruddin, M. K. Angew. Chem., Int. Ed. 2009, 48, 9277. (9) Jiang, X.; Karlsson, K. M.; Gabrielsson, E.; Johansson, E. M. J.; Quintana, M.; Karlsson, M.; Sun, L.; Boschloo, G.; Hagfeldt, A. Adv. Funct. Mater. 2011, 21, 2944. (10) Burschka, J.; Dualeh, A.; Kessler, F.; Baranoff, E.; Cevey-Ha, N.L.; Yi, C.; Nazeeruddin, M. K.; Grätzel, M. J. Am. Chem. Soc. 2011, 133, 18042. (11) Snaith, H. J.; Schmidt-Mende, L. Adv. Mater. 2007, 19, 3187. (12) Kroeze, J. E.; Hirata, N.; Schmidt-Mende, L.; Orizu, C.; Ogier, S. D.; Carr, K.; Grätzel, M.; Durrant, J. R. Adv. Funct. Mater. 2006, 16, 1832. (13) Snaith, H., J; Robin, H.-B.; Peter, C.; Ilkay, C.; Shaik, M. Z.; Michael, G. Nanotechnology 2008, 19, 424003. (14) Ding, I. K.; Tétreault, N.; Brillet, J.; Hardin, B. E.; Smith, E. H.; Rosenthal, S. J.; Sauvage, F.; Grätzel, M.; McGehee, M. D. Adv. Funct. Mater. 2009, 19, 2431. (15) Melas-Kyriazi, J.; Ding, I. K.; Marchioro, A.; Punzi, A.; Hardin, B. E.; Burkhard, G. F.; Tétreault, N.; Grätzel, M.; Moser, J.-E.; McGehee, M. D. Adv. Energy Mater. 2011, 1, 407. (16) Ding, I. K.; Melas-Kyriazi, J.; Cevey-Ha, N.-L.; Chittibabu, K. G.; Zakeeruddin, S. M.; Grätzel, M.; McGehee, M. D. Org. Electron. 2010, 11, 1217. (17) Leijtens, T.; Ding, I. K.; Giovenzana, T.; Bloking, J. T.; McGehee, M. D.; Sellinger, A. ACS Nano 2012, 6, 1455. (18) Juozapavicius, M.; O’Regan, B. C.; Anderson, A. Y.; Grazulevicius, J. V.; Mimaite, V. Org. Electron. 2012, 13, 23. (19) Feng, X.; Shankar, K.; Varghese, O. K.; Paulose, M.; Latempa, T. J.; Grimes, C. A. Nano Lett. 2008, 8, 3781. (20) Liu, B.; Aydil, E. S. J. Am. Chem. Soc. 2009, 131, 3985. (21) Xu, C.; Shin, P.; Cao, L.; Gao, D. J. Phys. Chem. C 2010, 114, 125. (22) Xu, C.; Shin, P. H.; Cao, L.; Wu, J.; Gao, D. Chem. Mater. 2010, 22, 143. (23) Varghese, O. K.; Paulose, M.; Grimes, C. A. Nat. Nanotechnol. 2009, 4, 592. (24) Xin, X.; Wang, J.; Han, W.; Ye, M.; Lin, Z. Nanoscale 2012, 4, 964. (25) Wang, M.; Bai, J.; Le Formal, F.; Moon, S.-J.; Cevey-Ha, L.; Humphry-Baker, R.; Grätzel, C.; Zakeeruddin, S. M.; Grätzel, M. J. Phys. Chem. C 2012, 116, 3266. (26) Bendall, J. S.; Etgar, L.; Tan, S. C.; Cai, N.; Wang, P.; Zakeeruddin, S. M.; Gratzel, M.; Welland, M. E. Energy Environ. Sci. 2011, 4, 2903. (27) Xu, C.; Wu, J.; Desai, U. V.; Gao, D. J. Am. Chem. Soc. 2011, 133, 8122. 2424

dx.doi.org/10.1021/nl3004144 | Nano Lett. 2012, 12, 2420−2424