Formation of Efficient Dye-Sensitized Solar Cells by Introducing an

Langmuir 2008, 24, 13225-13230

13225

Formation of Efficient Dye-Sensitized Solar Cells by Introducing an Interfacial Layer of Long-Range Ordered Mesoporous TiO2 Thin Film Yong Joo Kim,† Yoon Hee Lee,† Mi Hyeon Lee,† Hark Jin Kim,† Jia Hong Pan,† Goo Il Lim,† Young S. Choi,† Kyungkon Kim,‡ Nam-Gyu Park,*,‡ Chongmu Lee,§ and Wan In Lee*,† Nano Materials and DeVices Laboratory, Department of Chemistry, Inha UniVersity, Incheon 402-751, Korea, Solar Cell Research Center, Materials Science & Technology DiVision, Korea Institute Science & Technology (KIST), Seoul 136-791, Korea, and Department of Materials Science and Engineering, Inha UniVersity, Incheon 402-751, Korea ReceiVed July 21, 2008. ReVised Manuscript ReceiVed September 1, 2008 Long-range ordered cubic mesoporous TiO2 films with 300 nm thickness were fabricated on fluorine-doped tin oxide (FTO) substrate by evaporation-induced self-assembly (EISA) process using F127 as a structure-directing agent. The prepared mesoporous TiO2 film (Meso-TiO2) was applied as an interfacial layer between the nanocrystalline TiO2 film (NC-TiO2) and the FTO electrode in the dye-sensitized solar cell (DSSC). The introduction of Meso-TiO2 increased Jsc from 12.3 to 14.5 mA/cm2, and Voc by 55 mV, whereas there was no appreciable change in the fill factor (FF). As a result, the photovoltaic conversion efficiency (η) was improved by 30.0% from 5.77% to 7.48%. Notably, introduction of Meso-TiO2 increased the transmittance of visible light through the FTO glass by 23% as a result of its excellent antireflective role. Thus the increased transmittance was a key factor in enhancing the photovoltaic conversion efficiency. In addition, the presence of interfacial Meso-TiO2 provided excellent adhesion between the FTO and main TiO2 layer, and suppressed the back-transport reaction by blocking direct contact between the electrolyte and FTO electrode.

Introduction Recently, dye-sensitized solar cells (DSSCs) based on molecular dyes adsorbed on the surface of a thick nanocrystalline TiO2 film (NC-TiO2) have attracted extensive attention due to their low production cost of electricity and relatively high energy conversion efficiency.1-5 One of the highest photovoltaic conversion efficiencies so far reported for DSSCs was about 11%,6-8 and further improvement is expected, considering the energy gap of the Ru-based dye molecules. One of the key issues in increasing DSSC efficiency is the strategic design and tailoring of the porous TiO2 structure at a nanoscale level in order to increase the adsorption amount of molecular dye and to expedite the transport of both electrons and electrolytes.9 * Corresponding author. Tel:+82-32-863-1026; fax :+82-32-867-5604; e-mail: [email protected] (W.I.L.) Tel: +82-2-958-5365; fax: +82-2-9585309; e-mail: [email protected] (N.-G.P.). † Department of Chemistry, Inha University. ‡ Korea Institute Science & Technology. § Department of Materials Science and Engineering, Inha University.

(1) O’Regan, B.; Gra¨tzel, M. Nature 1991, 353, 737. (2) Kay, A.; Gra¨tzel, M. Sol. Energy Mater. Sol. Cells 1996, 44, 99. (3) Gra¨tzel, M. Nature 2001, 414, 338. (4) Pettersson, H.; Gruszecki, T. Sol. Energy Mater. Sol. Cells 2001, 70, 203. (5) (a) Dai, S. Y.; Wang, K. J.; Weng, J.; Sui, Y. F.; Huang, Y.; Xiao, S. F.; Chen, S. H.; Hu, L. H.; Kong, F. T.; Pan, X.; Shi, C. W.; Guo, L. Sol. Energy Mater. Sol. Cells 2005, 85, 447. (b) Yamaguchi, T.; Tobe, N.; Matsumoto, D.; Arakawa, H. Chem. Commun. 2007, 45, 4767. (6) Nazeeruddin, M. K.; Pechy, P.; Renouard, T.; Zakeeruddin, S. M.; Baker, R. H.; Comte, P.; Liska, P.; Cevey, L.; Costa, E.; Shklover, V.; Spiccia, L.; Deacon, G. B.; Bignozzi, C. A.; Gra¨tzel, M. J. Am. Chem. Soc. 2001, 123, 1613. (7) Gra¨tzel, M. Inorg. Chem. 2005, 44, 6841. (8) (a) Wang, Q.; Ito, S.; Gra¨tzel, M.; Fabregat-Santiago, F.; Mora-Sero, I.; Bisquert, J.; Bessho, T.; Imai, H. J. Phys. Chem. B 2006, 110, 25210. (b) Chiba, Y.; Islam, A.; Watanabe, Y.; Komiya, R.; Koide, N.; Han, L Jpn. J. Appl. Phys., Pt. 2 2006, 45, 638. (9) (a) Wei, M.; Konishi, Y.; Zhou, H.; Yanagida, M.; Sugihara, H.; Arakawa, H. J. Mater. Chem. 2006, 16, 1287. (b) Yoon, J. H.; Jang, S. R.; Vittal, R.; Lee, J. W.; Kim, K. J. J. Photochem. Photobiol. A: Chem. 2006, 180–184. (c) Park, J. H.; Lee, T. W.; Kang, M. G. Chem. Commun. 2008, 25, 2867.

For the construction of DSSCs, TiO2 nanoparticles sized about 20 nm are typically deposited as a nanoporous layer with a thickness of 10-15 µm on a transparent conductive oxide (TCO).10-12 In this NC-TiO2, transferring electrons through the conduction band (CB) of TiO2 would not be efficient, because the electrons generated from the dye molecules have to pass through numerous grain boundaries in order to reach the TCO. In addition, the transport of electrolytes is inefficient because of the irregularity of the generated pores. In this regard, the tailoring of TiO2 nanostructures is a crucial aspect of increasing the current photovoltaic efficiency of DSSCs. Long-range ordered mesoporous TiO2 film (Meso-TiO2) is considered to be a promising candidate as a nanoporous electrode of DSSC, because of its high surface area, few grain boundaries, and uniform pore structure with excellent connectivity of mesopores. Thus the efficient transfer of electrons and the diffusion of electrolytes are expected to be achieved by this long-range ordered Meso-TiO2. In general, the mesopores are prone to collapse during the heat-treatment process to obtain sufficiently crystallized structure, but this drawback was considerably overcome by the development of a evaporation-induced self-assembly (EISA) process.13-18 Nonetheless, it is still difficult to utilize Meso-TiO2 films as nanoporous electrodes of DSSC, since the typical thickness of mesoporous films derived from the EISA process is only about 300 nm. Several research groups (10) (a) Nakade, S.; Saito, Y.; Kubo, W.; Kitamura, T.; Wada, Y.; Yanagida, S. J. Phys. Chem. B 2003, 107, 8607. (b) Nazeeruddin, M. K.; Kay, A.; Rodicio, I.; Humphry-Baker, R.; Mueller, E.; Liska, P.; Vlachopoulos, N.; Gra¨tzel, M. J. Am. Chem. Soc. 1993, 115, 6382. (11) Kalyansundaram, K.; Gra¨tzel, M. Coord. Chem. ReV. 1998, 77, 347. (12) Gra¨tzel, M. J. Photochem. Photobiol. A 2004, 164, 3. (13) Yang, P. D.; Zhao, D. Y.; Margolese, D. I.; Chmelka, B. F.; Stucky, G. D. Nature 1998, 396, 152. (14) Yang, P. D.; Zhao, D. Y.; Margolese, D. I.; Chmelka, B. F.; Stucky, G. D. Chem. Mater. 1999, 11, 2813. (15) Brinker, C. J.; Lu, Y.; Sellinger, A.; Fan, H. AdV. Mater. 1999, 11, 579. (16) Coakley, K. M.; Liu, Y.; McGehee, M. D.; Frindell, K. L.; Stucky, G. D. AdV. Funct. Mater. 2003, 13, 301.

10.1021/la802340g CCC: $40.75  2008 American Chemical Society Published on Web 10/16/2008

13226 Langmuir, Vol. 24, No. 22, 2008

have tried to increase the thickness of Meso-TiO2 films, but only limited success has been achieved thus far.19-22 Commercial fluorine-doped tin oxide (FTO) glass used as the TCO layer looks hazy with a rough surface appearance and considerable light scattering on its surface. The formation of a uniform TiO2 layer over the rough surface of the FTO layer will therefore be difficult. In general, nanoparticles are agglomerated to form large colloids in suspensions or pastes in order to reduce the surface energy. For example, for Degussa P25 (particle size: 25 nm), which is utilized in the formation of NC-TiO2, the colloidal particle size in the paste is as large as several hundred nanometers.23 When the thick TiO2 layer is formed by the doctor blade method, therefore, the rough FTO surface cannot be uniformly covered by the large secondary TiO2 particles without the formation of some voids in its interface. Presumably, some part of the FTO surface is not covered by TiO2, and can directly contact the electrolytes, thereby deteriorating the adhesion between the TCO and TiO2 layer and, as a result, decreasing Voc by the electron back-transport reaction [FTO (2e-) + I3- f FTO + 3I-]. Therefore, controlling the interface between the TCO and TiO2 layer is crucial for the formation of efficient DSSCs. Thus far, however, only a few studies have been reported on the interfacial control between TCO and nanoporous TiO2 layers.24-26 In this work, a long-range ordered 300 nm-thick Meso-TiO2 was formed as an interfacial layer between the main TiO2 and TCO. This introduction of Meso-TiO2 improved η by 30.0%. The mechanistic role of the interfacial Meso-TiO2 thin layer in enhancing η of the DSSC was also investigated.

Experimental Section Preparation of Meso-TiO2. The thin layer of Meso-TiO2 in cubic Im3m mesophase was formed on the FTO substrate by the EISA process.27 The preparation method was based on our previous report on Meso-TiO2 films. In a typical experiment, Pluronic F127 (Aldrich) was dissolved in ethanol to form a clear solution. In another beaker, titanium tetraisopropoxide (TTIP) was dissolved and stabilized in an aqueous HCl solution while vigorously stirring. After aging for 15 min, the yellowish Ti-precursor solution was added to the F127 ethanol solution dropwise. The molar ratio of the compositions for the prepared Ti-sol was TTIP/F127/HCl/H2O/EtOH ) 1:0.005:1.7: 10:24. The initial sol was aged under a mild stirring condition at room temperature for 3 h, and was then spin-coated on an FTO glass (Pilkington, TEC-8, 8 Ω/sq), which was precleaned by the RCA method. The spin-speed was raised with an acceleration of 100 rpm/ s, and then maintained at 600 rpm for 20 s. All the deposited films were aged for 5 days in a humid chamber controlled to a relative humidity (RH) of 70% in 20 °C. The removal of F127, used as structure-directing agent, was carried out by heating at 400 °C for 4 h (ramping rate: 1 °C/min) in air. (17) (a) Grosso, D.; Soller-Illia, G. J. A. A.; Babonneau, F.; Sanchez, C.; Albouy, P. A.; Brunneau, A. B.; Balkenende, A. R. AdV. Mater. 2001, 13, 1085. (b) Crepaldi, E. L.; de Soler-Illia, G. J.; Grosso, D.; Cagnol, F.; Ribot, F.; Sanchez, C. J. J. Am. Chem. Soc. 2003, 125, 9770. (c) Etienne, M.; Grosso, D.; Boissiere, C.; Sanchez, C.; Walcarius, A. Chem. Commun. 2005, 36, 4566. (18) Yun, H.; Miyazawa, K.; Zhou, H.; Honma, I.; Kuwubara, M. AdV. Mater. 2001, 13, 1377. (19) Zukalova, M.; Zukal, A.; Kavan, L.; Nazeeruddin, M. K.; Liska, P.; Gra¨tzel, M. Nano Lett. 2005, 5, 1789. (20) Wei, M.; Wang, K.; Yanagida, M.; Sugihara, H.; Morris, M. A.; Holmes, J. D.; Zhou, H. J. Mater. Chem. 2007, 17, 3888. (21) Chen, W.; Sun, X.; Cai, Q.; Weng, D.; Li, H. Electrochem. Commun. 2007, 9, 382. (22) Zukalova´, M.; Procha´zka, J.; Zukal, A.; Yum, J. H.; Kavan, L. Inorg. Chim. Acta 2008, 361, 656. (23) Chae, S. Y.; Park, M. K.; Lee, S. K.; Kim, T. Y.; Kim, S. K.; Lee, W. I. Chem. Mater. 2003, 15, 3326. (24) Kang, T.-S.; Moon, S.-H.; Kim, K.-J. J. Electrochem. Soc. 2002, 149, E155. (25) Cameron, P. J.; Peter, L. M. J. Phys. Chem. B 2003, 107, 14394. (26) Ahn, K.-S.; Kang, M.-S.; Lee, J.-W.; Kang, Y. S. J. Appl. Phys. 2007, 101, 084312.

Kim et al. Fabrication of DSSCs. For the preparation of the NC-TiO2, a viscous paste was prepared by the following procedure. A 0.2 g portion of ethyl cellulose, 0.05 g of lauric acid, and 5 g of terpineol were added to an ethanol suspension containing 2.0 g of TiO2 nanoparticles (Degussa P25), and the solvent was then evaporated by a rotary evaporator in room temperature to obtain viscous pastes. The prepared TiO2 paste was coated on the bare FTO or on the Meso-TiO2-coated FTO layer with the doctor blade method. The coated films were baked at 150 °C for 30 min and subsequently calcined at 500 °C for 15 min. For the dye adsorption, the fabricated TiO2 film was immersed in anhydrous ethanol containing 5 × 10-4 M N719 dye (Ru[LL′-(NCS)2], L ) 2,2′-bypyridyl-4,4′-dicarboxylic acid; L′ ) 2,2′-bypyridyl-4,4′-ditetrabutylammonium carboxylate; Solaronix Co.), and kept for 24 h at room temperature. The Ptcoated FTO, used as a counter electrode, was prepared by dropping a 0.7 mM H2PtCl6 solution on an FTO glass followed by heating at 400 °C for 20 min in air. The electrolyte consisted of 0.6 M 1-hexyl-2,3-dimethyl-imidazolium iodide (HDMII, Merck Co.), 0.05 M I2, 0.1 M lithium iodine (LiI, Aldrich Chemical Co.), and 0.5 M 4-tert-butylpyridine (TBP, Aldrich Chemical Co.) in methoxyacetonitrile (Aldrich Chemical Co.). The active area of the dye-coated TiO2 film was 0.160 cm2. Characterizations. Photocurrent-voltage measurements were performed using a Keithley model 2400 source measurement unit. A 1000 W xenon lamp (Spectra-Physics) was used as the light source, and the light intensity was adjusted using an NREL-calibrated Si solar cell equipped with a KG-5 filter for approximating AM 1.5G one sun light intensity. The incident photon-to-current efficiency (IPCE) spectra was measured as a function of wavelength from 400 to 900 nm using a specially designed IPCE system (PV Measurements, Inc.) for DSSCs. Electrical impedance spectra were measured with an impedance analyzer (Compactstat, IVIUM Tech.) at open-circuit potential under AM 1.5G one sun light illumination, with frequency ranging from 10-1 Hz to 106 Hz. The magnitude of the alternative signal was 10 mV. Impedance parameters were determined by fitting of impedance spectra using Z-view software. The electron diffusion coefficient and lifetime were measured by pulsed-laser-induced current and voltage transients with a Nd:YAG laser (532 nm) under continuous irradiation of a diode laser (635 nm).28-33 Small-angle X-ray diffraction (SAXRD) patterns for the TiO2 films were obtained by using a Rigaku Multiflex diffractometer with a monochromated high-intensity Cu KR radiation operated at 40 kV, 20 mA. The diffraction patterns were scanned at a rate of 0.1°/min over the 2θ region of 0.7-5.0°. For the observation of the prepared mesoporous thin films by transmission electron microscopy (TEM, Philips CM30 transmission electron microscope operated at 200 kV), the films were scratched off from the substrate, and the collected flakes were gently dispersed in methanol. The suspension was then dropped on a holey amorphous carbon film deposited on a Ni grid (JEOL, Ltd.). The morphology of the TiO2 mesoporous films was examined by a field-emission scanning electron microscope (FESEM, Hitachi S4200) and an atomic force microscope (AFM, Digital Instrument Nanoscope Multimode IVa). The optical transmissions of the films were recorded by a UV-visible spectrophotometer (Perkin-Elmer Lambda 40) in the wavelength range of 300-900 nm.

Results and Discussion Characterization of Interfacial Meso-TiO2. Figure 1a shows TEM images for the Meso-TiO2 formed on the FTO substrate. (27) Pan, J. H.; Lee, W. I. New J. Chem. 2005, 29, 841. (28) Benkstein, K. D.; Kopidakis, N.; van de Lagemaat, J.; Frank, A. J. J. Phys. Chem. B 2003, 107, 7759. (29) Yanagida, S.; Kitamura, T.; Kohmoto, M. Electrochem. 2002, 70, 399. (30) Kopidakis, N.; Benkstein, K. D.; van de Lagemaat, J.; Frank, A. J. J. Phys. Chem. B 2003, 107, 11307. (31) Nakade, S.; Kubo, W.; Saito, Y.; Kanzaki, T.; Kitamura, T.; Wada, Y.; Yanagida, S. J. Phys. Chem. B 2003, 107, 14244. (32) Nakade, S.; Kanzaki, T.; Kubo, W.; Kitamura, T.; Wada, Y.; Yanagida, S. J. Phys. Chem. B 2005, 109, 3480. (33) Kopidakis, N.; Neale, N. R.; Frank, A. J. J. Phys. Chem. B 2006, 110, 12485.

Efficient DSSC by Introducing Meso-TiO2 Thin Film

Langmuir, Vol. 24, No. 22, 2008 13227

Figure 1. TEM images for the Meso-TiO2 deposited on FTO (a) and Pyrex (b). (c) Cross-sectional SEM image for the Meso-TiO2 formed on an FTO layer.

Figure 2. SAXRD patterns for Meso-TiO2 on FTO and Pyrex.

Mesopores with a cavity diameter of about 7 nm were aligned periodically, suggesting the formation of a long-range ordered mesoporous structure. When fabricated on the Pyrex substrate under the same conditions, the mesopores are more uniformly organized, as shown in Figure 1b. The mesophase of the film is clearly shown to be analogous to SBA-16, which consists of a cubic array of mesopores in the Im3m space group.34 Figure 1a,b also shows that the mesopores formed on the FTO or Pyrex substrate are oriented to the (100) plane. Figure 1c shows the cross-section of the Meso-TiO2 deposited on the FTO layer. The Meso-TiO2 with thickness of about 300 nm was so tightly adhered to the 600 nm-thick FTO layer that its boundary was not clearly seen. SAXRD patterns for the Meso-TiO2 deposited on the FTO layer and Pyrex substrate are shown in Figure 2. Both films show the characteristic peak at 1.9°, indexed to the (200) plane, even though the peak of the film deposited on FTO was considerably broader than that formed on the Pyrex substrate. This suggests that the mesopores in the surface of both films were oriented to the (100) direction, which is compatible with the orientation observed by TEM, and the Meso-TiO2 formed on the FTO layer was less uniformly organized. Presumably, the appreciable deterioration of the long-range ordering was closely related to the rough surface of the FTO layer. The AFM image in Figure 3a and 3b indicates that the root-mean-square (rms) roughness of the FTO surface was as high as 34 nm, compared to only 0.9 nm for the bare Pyrex glass. The formation of highly ordered mesoporous structure would have been difficult on such a rough surface, which explains why the long-range ordering was deteriorated in the film formed on the FTO substrate. The (34) Zhao, D.; Huo, Q.; Feng, J.; Chmelka, B. F.; Stucky, G. D. J. Am. Chem. Soc. 1998, 120, 6024.

Figure 3. AFM images for the surfaces of bare Pyrex glass (a), FTO layer (b), and Meso-TiO2/FTO layer (c).

Figure 4. N2 adsorption-desorption isotherms for the Meso-TiO2 formed on Pyrex and an FTO layer.

mesoporosity of the Meso-TiO2 formed on the FTO or Pyrex substrate was analyzed by N2 adsorption-desorption isotherms, and the results are shown in Figure 4. Both films showed a notably narrow pore size distribution with a pore diameter of about 7 nm, which is compatible with the result obtained from the TEM analysis. This suggests that the pores in the Meso-TiO2 on the FTO substrate formed a typical mesoporous structure

13228 Langmuir, Vol. 24, No. 22, 2008

Kim et al.

Figure 5. Cross-sectional SEM image of the NC-TiO2/Meso-TiO2/FTO applied to a DSSC.

Figure 7. IPCE spectra for the DSSCs with and without interfacial Meso-TiO2.

Figure 6. I-V curves for the DSSCs with and without interfacial MesoTiO2.

with uniform pore diameter, even though the long-range ordering was relatively degenerated. Characterization of DSSCs. In the present work, the 300 nm-thick Meso-TiO2 was applied as the interfacial layer between the FTO and the main TiO2 layer. The main NC-TiO2 was coated over the Meso-TiO2/FTO substrate by the doctor blade technique with paste based on Degussa P25 (25 nm-sized TiO2 nanoparticle). The thickness of the NC-TiO2 after annealing at 450 °C for 30 min was 11.5 µm, as shown in the cross-sectional SEM image of Figure 5. Figure 6 indicates the I-V characteristics for the fabricated DSSCs. The cell derived from the NC-TiO2 with Meso-TiO2 interfacial layer (NC-TiO2/Meso-TiO2) demonstrated considerably improved short circuit current (Jsc) and open circuit voltage (Voc), compared to that of the bare NC-TiO2. That is, Jsc was increased from 12.3 to 14.5 mA/cm2, and Voc was increased from 0.680 to 0.735 V, whereas there is no appreciable change in the fill factor (FF). As a result, η was improved by 30.0% from 5.77% to 7.48% by introducing the interfacial Meso-TiO2 thin layer. Figure 7 presents the IPCE spectra for the DSSCs with and without interfacial Meso-TiO2. The introduction of the interfacial Meso-TiO2 appreciably increased the quantum efficiency over the whole spectral range of incident light. That is, the quantum efficiency at 550 nm was increased from 58% to 70%. Especially, in shorter wavelength region such as 400-550 nm, the increment of quantum efficiency was relatively higher. The introduction of a thin interfacial mesoporous layer only slightly increased the thickness of the overall TiO2 layer from 11.5 to 11.8 µm, which will not appreciably influence the

Figure 8. Plan-view SEM images for the bare FTO (a), and MesoTiO2/FTO (b). (c) Transmittance spectra for FTO/Pyrex and MesoTiO2/FTO/Pyrex.

conversion efficiency.35 The question therefore remains regarding the role of the interfacial Meso-TiO2 in increasing the photovoltaic conversion efficiency. Several factors might be involved in the observed increase in Jsc and Voc. First of all, more light can be transmitted through the FTO glass because of the antireflective capability of the thin Meso-TiO2. The surface of the commercial FTO film coated on Pyrex presents a hazy appearance due to its rough surface, as shown in the SEM photograph of Figure 8a and the AFM image of Figure 3b. By depositing the Meso-TiO2 on the FTO layer, however, the film surface gains a much more uniform appearance, as shown in Figure 8b. The AFM image in Figure 3c also indicates that the rms roughness of the MesoTiO2/FTO layer is only 3 nm, suggesting that the Meso-TiO2 successfully levels the rough surface of the FTO layer. (35) Huang, C.-Y.; Hsu, Y.-C.; Chen, J.-G.; Suryanarayanan, V.; Lee, K.-M.; Ho, K.-C. Sol. Energy Mater. Sol. Cells 2006, 90, 2391.

Efficient DSSC by Introducing Meso-TiO2 Thin Film

Effect of Meso-TiO2 on Optical Transmittance. The transmittance spectra for the bare FTO and Meso-TiO2/FTO substrate are shown in Figure 8c. Strikingly, the Meso-TiO2/ FTO on Pyrex showed considerably higher transparency than the bare FTO/Pyrex did. Especially, the Meso-TiO2/FTO showed higher transmittance in the relatively shorter wavelength region of visible light. Obviously, this is caused by the improved uniformity of the film surface with the formation of the MesoTiO2 on FTO. It is well-known that light of a shorter wavelength is relatively more scattered on a rough surface than that of a longer wavelength, since the scattering efficiency of light is proportional to 1/λ4, where λ is the wavelength of incident light.36 The IPCE spectra in Figure 7 indicate that the improvement of quantum efficiency achieved by introducing Meso-TiO2 is relatively higher in a shorter wavelength region (400-550 nm), whereas there is no significant increase in the longer wavelength region. This observation is compatible with the trend of light transmittance as a function of wavelength (Figure 8c). The N719 dye with an energy gap of 1.6 eV between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) can absorb light with a wavelength shorter than 750 nm. The overall increase of the transmitted light in the wavelength region shorter than 750 nm, which can be absorbed by the N719 dye, was calculated as 23% from the transmittance spectra shown in Figure 8c. To evaluate the effect of light intensity in increasing the cell efficiency, we measured the I-V curve for the bare NC-TiO2 cell under a 23%increased light intensity. That is, the 1.23-fold increase of the AM 1.5G light intensity was obtained by reducing the distance from the light source to the cell. It was found that the cell efficiency was increased by 20.5%, and this was solely achieved by the increase of Jsc without any change in Voc or FF. Therefore, it is deduced that about 2/3 of the total increment of η (30.0%) was caused by the antireflective role of the Meso-TiO2 film. Effect of Meso-TiO2 on Adhesion and Back-Transport Reaction. It was found that the Meso-TiO2 demonstrated excellent adhesion to FTO glass. In most cases, Meso-TiO2 films formed by the EISA process are very difficult to strip off without damaging the FTO layer. Quantitatively, the Meso-TiO2 film was unaffected without scratching by abrasion with a 4H pencil. It also exhibited good adhesion to the main NC-TiO2. Thus it is deduced that the contact between the FTO layer and NC-TiO2 was greatly improved by introducing thin interfacial Meso-TiO2. No clear boundary between Meso-TiO2 and FTO, shown in Figure 1c, also suggests a tight contact between these two layers. Furthermore, the back-transport of electrons in the interface of FTO electrode would be suppressed by introducing MesoTiO2, since it blocks the direct encounter between the electrolyte and FTO electrode. I- ions migrate from the counter electrode to the nanopores of TiO2, transfer the electrons to the HOMO of the dye anchored on the TiO2, and are then converted to I3ions. In general, nanoparticles are agglomerated to form large colloids in suspensions or pastes in order to reduce the surface energy. In the case of Degussa P25 (particle size: 25 nm), utilized in the formation of NC-TiO2 in this work, the colloidal particle size in the paste was larger than 500 nm. Hence the large secondary TiO2 particles cannot perfectly cover the rough FTO surface without void formation, when the main TiO2 layer is formed by the doctor blade method. As a result, in the interface of the FTO layer and the porous TiO2 electrode, some of I3- ions can encounter the FTO electrode. Then, the electrons would be backtransported from the FTO electrode to the I3- ions, as described in Figure 9. However, the presence of the Meso-TiO2 between (36) Rayleigh, L.; Strutt, J. W. Philos. Mag. 1899, 47, 375.

Langmuir, Vol. 24, No. 22, 2008 13229

Figure 9. Scheme for the back-transport reaction, which may have occurred on the FTO/NC-TiO2 interface.

Figure 10. Impedance spectra for the DSSCs with and without interfacial Meso-TiO2.

FTO and NC-TiO2 in the DSSC can block the direct contact between the I3- ions and FTO electrode, thereby effectively preventing the back-transport of electrons from the FTO electrode to the I3- ions, and thus increasing both Jsc and Voc.25,37,38 Figure 10 shows the impedance spectra for the NC-TiO2 and NC-TiO2/Meso-TiO2 cells. The first hemisphere represents the resistance applied to the TiO2 electrode and to the FTO/TiO2 interface.39,40 The relatively smaller hemisphere for the NCTiO2/Meso-TiO2 cell indicates that the resistance on the FTO/ TiO2 interface was appreciably decreased, considering that the resistance applied at NC-TiO2 was the same for these two cells. This result clearly demonstrated that the Meso-TiO2 layer induces excellent contact between the FTO and NC-TiO2 layer. The electron diffusion coefficient and the electron lifetime during the DSSC operation were analyzed by transient photocurrent and photovoltage spectroscopy. The electron diffusion coefficient (De) was determined by the following equation: De ) L2/(2.35 × τc), where L and τc are the film thickness and time constant, respectively.32 The time constant (τc) can be obtained by fitting a decay of transient photocurrent as a function of time (t) with a single-exponential function, exp(-t/τc). On the other hand, the electron lifetime (τe) can also be obtained by fitting (37) Zhua, K.; Schiff, E. A.; Park, N.-G.; van de Lagemaat, J.; Frank, A. J. Appl. Phys. Lett. 2002, 80, 685. (38) Frank, A. J.; Kopidakis, N.; van de Lagemaat, J. Coord. Chem. ReV. 2004, 248, 1165. (39) Kern, R.; Sastrawan, R.; Ferber, J.; Stangl, R.; Luther, J. Electrochim. Acta 2002, 47, 4213. (40) Hoshikawa, T.; Yamada, M.; Kikuchi, R.; Eguchi, K. J. Electrochem. Soc. 2005, 152, E68.

13230 Langmuir, Vol. 24, No. 22, 2008

Kim et al.

layer, as shown in Figure 11b. The increased τe value suggests the retardation of the recombination or back reaction during the transport of electrons through the main TiO2 layer. Therefore, the role of the interfacial Meso-TiO2 is considered to be blocking the back-transport reaction at the FTO/TiO2 interface. Voc is known to be strongly dependent on the recombination or back reactions taking place on the TiO2 and electrolyte interface, and a larger Voc value can be obtained by suppressing those reactions.25,41 The introduction of Meso-TiO2 increased Voc by 55 mV, as shown in Figure 6. This clearly shows that the backtransport of electrons from the FTO electrode to the I3- ions was suppressed by the introduction of Meso-TiO2.

Conclusions

Figure 11. Electron diffusion coefficients (a) and electron lifetimes (b) in the DSSCs with and without interfacial Meso-TiO2.

a decay of transient photovoltage with exp(-t/τe). Figure 11 shows the electron diffusion coefficient as a function of Jsc, and the electron lifetime as a function of Voc for the DSSC with and without Meso-TiO2. Both cells showed similar De values, as shown in Figure 11a, suggesting that the electron diffusion through the main TiO2 layer was not influenced by the introduction of interfacial Meso-TiO2 layer. On the contrary, the electron lifetime was appreciably increased by the introduction of Meso-TiO2

By introducing a long-range ordered cubic Meso-TiO2 film between the NC-TiO2 and FTO layer in a DSSC, Jsc and Voc were increased from 12.3 to 14.5 mA/cm2 and from 0.680 to 0.735 V, respectively, thus improving the photovoltaic conversion efficiency by 30.0%. First, the surface roughness of the FTO substrate was dramatically reduced by depositing interfacial MesoTiO2. The antireflective capability of Meso-TiO2 increased the light transmitted through the cell by 23%, and thereby considerably increased Jsc. Second, the Meso-TiO2 provided excellent adhesion between the FTO substrate and NC-TiO2 layer, and blocked direct contact between the FTO substrate and electrolyte. Thus the resistance at the FTO/NC-TiO2 interface was considerably decreased, while the electron lifetime in CB of TiO2 was appreciably increased. In addition, Voc was increased by 55 mV as a result of the retardation of back-transport reaction in the FTO/NC-TiO2 interface. Acknowledgment. This work was supported by a KIST internal project, the Korea Science and Engineering Foundation (KOSEF R01-2006-000-10956-0), and the Ministry of Information and Communication of Korea under the ITRC support program (IITA-2008-C109008010030). LA802340G (41) Cameron, P. J.; Peter, L. M. J. Phys. Chem. B 2005, 109, 7392.