4246
Langmuir 2004, 20, 4246-4253
Morphological and Photoelectrochemical Characterization of Core-Shell Nanoparticle Films for Dye-Sensitized Solar Cells: Zn-O Type Shell on SnO2 and TiO2 Cores N.-G. Park,* M. G. Kang, K. M. Kim, K. S. Ryu, and S. H. Chang Basic Research Laboratory, Electronics and Telecommunications Research Institute (ETRI), Daejeon 305-350, South Korea
D.-K. Kim Department of Chemistry, Kyungpook National University, Daegu 702-701, South Korea
J. van de Lagemaat, K. D. Benkstein, and A. J. Frank* National Renewable Energy Laboratory, Golden, Colorado 80401 Received November 11, 2003. In Final Form: February 17, 2004 Core-shell type nanoparticles with SnO2 and TiO2 cores and zinc oxide shells were prepared and characterized by surface sensitive techniques. The influence of the structure of the ZnO shell and the morphology of nanoparticle films on the performance was evaluated. X-ray absorption near-edge structure and extended X-ray absorption fine structure studies show the presence of thin ZnO-like shells around the nanoparticles at low Zn levels. In the case of SnO2 cores, ZnO nanocrystals are formed at high Zn/Sn ratios (ca. 0.5). Scanning electron microscopy studies show that Zn modification of SnO2 nanoparticles changes the film morphology from a compact mesoporous structure to a less dense macroporous structure. In contrast, Zn modification of TiO2 nanoparticles has no apparent influence on film morphology. For SnO2 cores, adding ZnO improves the solar cell efficiency by increasing light scattering and dye uptake and decreasing recombination. In contrast, adding a ZnO shell to the TiO2 core decreases the cell efficiency, largely owing to a loss of photocurrent resulting from slow electron transport associated with the buildup of the ZnO surface layer.
Introduction Photoelectrochemical dye-sensitized solar cells based on wide-band-gap nanocrystalline semiconductors are potential low-cost alternatives to conventional solid-state devices.1 The active component of the conventional cell is a several micrometer thick film of a porous nanocrystalline metal oxide, which is deposited onto a conducting glass substrate. The surface of the nanocrystalline film is covered with a monolayer of a sensitizing dye, and the porous film structure is interpenetrated with a liquid redox electrolyte. The counter electrode is a conducting glass plate coated with platinum. Because of the large internal surface area of the nanoparticle film, the amount of adsorbed dye is sufficient to absorb a significant fraction of the solar spectrum. Solar-to-electrical energy conversion efficiencies of more than 10% at AM 1.5 light intensity have been reported for a solar cell consisting of a nanoparticle anatase TiO2 film, a ruthenium complex as the sensitizer, and I-/I3- in acetonitrile as the redox electrolyte.2 To understand the factors that control the cell performance, studies of the morphology and optical properties of nanocrystalline oxide films,3-6 the photochemical properties of molecular sensitizers,2,7-16 and the electro* To whom correspondence should be addressed. (1) O’Regan, B.; Gra¨tzel, M. Nature 1991, 353, 737. (2) Nazeeruddin, M. K.; Pe´chy, P.; Renouard, T.; Zakeeruddin, S. M.; Humphry-Baker, R.; 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. (3) Kavan, L.; Gra¨tzel, M.; Rathousky, J.; Zukal, A. J. Electrochem. Soc. 1996, 143, 394.
chemical properties of redox electrolytes17-20 have been carried out. A relatively recent topic in dye-sensitized solar cell research has been core-shell nanoparticle films consisting of mixed oxides. The short-circuit photocurrent and open-circuit photovoltage of dye-sensitized solar cells (4) Barbe´, C. J.; Arendse, F.; Comte, P.; Jirousek, M.; Lenzmann, F.; Shklover, V.; Gra¨tzel, M. J. Am. Ceram. Soc. 1997, 80, 3157. (5) Papageorgiou, N.; Barbe´, C. J.; Gra¨tzel, M. J. Phys. Chem. B 1998, 102, 4156. (6) Park, N.-G.; van de Lagemaat, J.; Frank, A. J. J. Phys. Chem. B 2000, 104, 8989. (7) Fillinger, A.; Parkinson, B. A. J. Electrochem. Soc. 1999, 146, 4559. (8) Hara, K.; Sugihara, H.; Singh, L. P.; Islam, A.; Katoh, R.; Yanagida, M.; Sayama, M.; Murata, S.; Arakawa, H. J. Photochem. Photobiol., A 2001, 145, 117. (9) Laschewsky, A.; Ouari, O.; van Cleuvenbergen, P. Chem. Mater. 2001, 13, 3888. (10) Ehret, A.; Stuhl, L.; Spitler, M. T. Electrochim. Acta 2000, 45, 4553. (11) Aranyos, V.; Grennberg, H.; Tingry, S.; Lindquist, S.-E.; Hagfeldt, A. Sol. Energy Mater. Sol. Cells 2000, 64, 97. (12) Hara, K.; Horiguchi, T.; Kinoshita, T.; Sayama, K.; Sugihara, H.; Arakawa, H. Sol. Energy Mater. Sol. Cells 2000, 64, 115. (13) He, J.; Zhao, J.; Shen, T.; Hidaka, H.; Serpone, N. J. Phys. Chem. B 1997, 101, 9027. (14) Ruile, S.; Kohle, O.; Pe´chy, P.; Gra¨tzel, M. Inorg. Chim. Acta 1997, 261, 129. (15) Heimer, T. A.; D’Arcangelis, S. T.; Frazad, F.; Stipkala, J. M.; Meyer, G. J. Inorg. Chem. 1996, 35, 5319. (16) Kay, A.; Gra¨tzel, M. J. Phys. Chem. 1993, 97, 6272. (17) Ren, Y.; Zhang, Z.; Gao, E.; Fang, S.; Cai, S. J. Appl. Electrochem. 2001, 31, 445. (18) Nusbaumer, H.; Moser, J.-E.; Zakeeruddin, S. M.; Nazeeruddin, M. K.; Gra¨tzel, M. J. Phys. Chem. B 2001, 105, 10461. (19) Montanari, I.; Monkhouse, R.; Nelson, J.; Durrant, J. R. J. Phys. Chem. B 2001, 105, 7517. (20) Liu, Y.; Hagfeldt, A.; Xiao, X.-R.; Lindquist, S.-E. Sol. Energy Mater. Sol. Cells 1998, 55, 267.
10.1021/la036122x CCC: $27.50 © 2004 American Chemical Society Published on Web 04/03/2004
Core-Shell Nanoparticle Films
made up of a SnO2 and ZnO mixture are reported to be higher than those of cells composed of either ZnO or SnO2 nanoparticle films.21 Improvements are also observed for cells made up of a ZnO and TiO2 mixture.22 Such improvements are attributed to an improved charge-injection efficiency resulting from a downward shift of the conduction band edge and a decreased recombination rate owing to a reduced number of recombination sites.22 Improved current density and voltage characteristics have also been reported for nanocrystalline SnO2 and TiO2 covered with metal oxides, such as ZnO,23 SrO,24 Nb2O5,25 and various other oxides.26-30 Faster transport has been invoked to explain the effect of the bilayer structure on the solar cell characteristics.21,31,32 Other studies have, however, attributed the improved efficiency to slower transport,26 decreased recombination,25,26,29,30 or band-edge movement.27,28 The morphologies of nanoparticle core-shell materials and their effect on the photoelectrochemical properties of dye-sensitized solar cells are still not well understood. In this paper, we report on the morphological and photoelectrochemical properties of mesoporous (2-50 nm) and macroporous (>50 nm) core-shell nanoparticle films, involving a ZnO-like shell on SnO2 and TiO2 cores. The films were characterized by several techniques, including X-ray diffraction (XRD), scanning electron microscopy (SEM), X-ray absorption near-edge spectroscopy (XANES), and extended X-ray absorption fine structure spectroscopy (EXAFS). The relationship between the morphology of the bilayer materials and the PV properties of dyesensitized solar cells is also discussed. Experimental Section Preparation of ZnO-Covered Nanocrystalline SnO2 and Unmodified ZnO. Nanocrystalline SnO2 colloids were prepared by the precipitation of hydrated tin oxide from an aqueous solution of 0.04 M SnCl4 (Aldrich, 99%) and 1.2 M urea (Aldrich, 99%).33 The solution was subsequently hydrolyzed at 85 °C for 4 h, filtered, and then washed with deionized water. The wet gel was suspended in an aqueous acetic acid solution (pH 2; 240 mL) and hydrothermally treated at 200 °C for 12 h. The resulting colloidal suspension was condensed to a final SnO2 concentration of 12 wt % using a rotary evaporator. ZnO-coated SnO2 colloids were prepared by the addition of 0.087 g (3.98 × 10-4 mol of Zn2+) and 0.437 g (1.99 mmol of Zn2+) of Zn(CH3COO)2‚2H2O to 5 g of 12 wt % SnO2 solution (3.98 mmol of Sn4+), which corresponds to Zn2+/ Sn4+ ratios of 0.1 and 0.5, respectively. The viscosity of (21) Tennakone, K.; Kottegoda, I. R. M.; De Silva, L. A. A.; Perera, V. P. S. Semicond. Sci. Technol. 1999, 14, 975. (22) Wang, Z.-S.; Huang, C.-H.; Huang, Y.-Y.; Hou, Y.-J.; Xie, P.-H.; Zhang, B.-W.; Cheng, H.-M. Chem. Mater. 2000, 13, 678. (23) Park, N.-G.; Kang, M. G.; Chang, S. H. In W1 Dye-Sensitized Solar Cells, 14th International Conference on Photochemical Conversion and Storage of Solar Energy (IPS-14), Sapporo, Japan, 2002. (24) Yang, S.; Huang, Y.; Huang, C.; Zhao, X. Chem. Mater. 2002, 14, 1500. (25) Chen, S. G.; Chappel, S.; Diamant, Y.; Zaban, A. Chem. Mater. 2001, 13, 4629. (26) Kay, A.; Gra¨tzel, M. Chem. Mater. 2002, 14, 2930. (27) Palomares, E.; Clifford, J. N.; Haque, S. A.; Lutz, T.; Durrant, J. R. J. Am. Chem. Soc. 2003, 125, 475. (28) Diamant, Y.; Chen, S. G.; Melamed, O.; Zaban, A. J. Phys. Chem. B 2003, 107, 1977. (29) Chappel, S.; Chen, S.-G.; Zaban, A. Langmuir 2002, 18, 3336. (30) Kumara, G. R. R. A.; Tennakone, K.; Perera, V. P. S.; Konno, A.; Kaneko, S.; Okuya, M. J. Phys. D 2001, 34, 868. (31) Tennakone, K.; Bandara, J. Appl. Catal., A 2001, 208, 335. (32) Tennakone, K.; Bandara, J.; Bandaranayake, K. M. P.; Kumara, G. R. R. A.; Konno, A. Jpn. J. Appl. Phys. 2001, 40, L732. (33) Baik, N. S.; Sakal, G.; Miura, N.; Yamazoe, N. Sens. Actuators, B 2000, 63, 74.
Langmuir, Vol. 20, No. 10, 2004 4247
the zinc acetate-containing SnO2 colloidal solution was controlled by adding 60 wt % hydroxypropyl cellulose (MW 80 000) with respect to the SnO2 weight. The resulting slurry was spread onto a transparent conducting glass substrate (Libbey-Owens-Ford; F-doped SnO2 overlayer; 80% transmittance in the visible; 5% haze; 8 Ω/square) by the doctor-blade technique, dried at ambient temperature, and subsequently heated in air for 1 h at 500 °C. The thickness of the SnO2 and Zn-coated SnO2 films was about 5 µm as measured with a Tencor Alpha-Step 500 profiler and cross-sectional scanning electron microscopy. ZnO films were prepared using a commercial ZnO powder (Cerac, ∼200 mesh, 99.999%). The slurry was made by mixing ZnO powder and hydroxy propyl cellulose in distilled water. The ZnO content was adjusted to 12 wt %. Preparation of ZnO-Covered Nanocrystalline TiO2. Twenty nanometer diameter anatase TiO2 colloids were prepared by hydrolyzing titanium tetraisopropoxide (Aldrich, 98%) in the presence of distilled acetic acid followed by autoclaving at 220 °C for 12 h as described elsewhere.34 After the hydrothermal treatment, the TiO2 colloidal solutions were condensed to a final TiO2 concentration of 12% with a rotary evaporator. Zn-coated TiO2 colloids were prepared by adding 0.016, 0.04, 0.08, and 0.16 g of Zn(CH3COO)2‚2H2O to 5 g of 12% TiO2 colloid solution, which corresponds to respective Zn2+/ Ti4+ molar ratios of 0.01, 0.025, 0.05, and 0.1. The viscosity of the resulting slurry was controlled by adding 40 wt % hydroxypropyl cellulose (MW 80 000) with respect to the TiO2 weight. Titania films were prepared and their thicknesses measured as described for the SnO2 films. In the case of both Zn-modified SnO2 and Zn-modified TiO2, the pH of the deposition solution, which is about 2, is too low to form ZnO in the solution phase due to the presence of acetic acid. It is, however, expected that a significant fraction of Zn2+ ions adsorbs to the colloid surface.35 Also, when the films are fired in air, the acetic acid and acetate anions evaporate or decompose thermally. During the annealing phase, the ZnO-like shells are created. Because no zinc species is particularly volatile, all of the zinc ions end up, in some form, in the final film at the same molar ratio as that of the original deposition solution. Preparation of Dye-Sensitized Solar Cells. The core-shell nanoparticle films were immersed in absolute ethanol containing 3 × 10-4 M Ru[LL′(NCS)2] (L ) 2,2′bipyridyl-4,4′-dicarboxylic acid, L′ ) 2,2′-bipyridyl-4,4′ditetrabutylammoniumcarboxylate) (N719 dye) for 24 h at room temperature. The dyed electrodes were subsequently rinsed with absolute ethanol and dried under an N2 stream. Platinum counter electrodes were prepared by spreading a drop of 5 mM hexachloroplatinic acid (Fluka, purum) in 2-propanol on a piece of conducting glass followed by heating at 400 °C for 30 min in air. The electrodes were sealed with 30-µm-thick Surlyn (Dupont, grade 1702) at a pressure of 200 kPa/cm2 and a temperature of about 100 °C. The redox electrolyte, which consisted of 0.6 M 1,2-dimethyl-3-hexyl-imidazolium iodide (C6DMI), 0.2 M LiI, 40 mM I2, and 0.2 M 4-tertbutylpyridine in acetonitrile, was introduced into the cell through one of two small holes drilled in the counter electrode. The holes were then covered and sealed with small squares of microscope objective glass and Surlyn. (34) Zaban, A.; Ferrere, S.; Sprague, J.; Gregg, B. A. J. Phys. Chem. B 1997, 101, 55. (35) Pelet, S.; Moser, J.-E.; Gra¨tzel, M. J. Phys. Chem. B 2000, 104, 1791.
4248
Langmuir, Vol. 20, No. 10, 2004
Characterization. X-ray diffraction data were collected with a Rigaku D/MaxRC X-ray diffractometer using Cu KR radiation (λ ) 1.5406 Å). The film morphology was investigated by SEM using a HITACHI S-800 microscope. A Hitachi U-3501 UV-vis spectrophotometer was used to determine the adsorbed dye concentration by measuring the absorbance of dye at 507 and 382 nm desorbed from the surface of the oxides in basic solution. In particular, the dye-covered films (electrodes) were immersed in a warm (60 °C) 5 mL 0.1 M NaOH aqueous solution for about 5 min to desorb the dye. The concentrations were evaluated spectrally with respect to prepared standards. The Zn K-edge X-ray absorption spectra were recorded at the EXAFS3C1 beam line at the Pohang Accelerator Laboratory (PAL), operating at 2.5 GeV with a stored current of 100-150 mA. A Si(311) double crystal monochromator was employed to collect high-resolution XANES spectra. The absorption energy was calibrated by simultaneously measuring the spectrum of Zn metal foil. All XANES spectra were recorded in transmission mode. Data analyses of the experimental spectra were performed using the following standard procedure:36 The inherent background in the data was removed by fitting a polynomial to the preedge region and extrapolating through the entire spectrum from which it was subtracted. The resulting spectra, µ(E), were normalized to an edge jump of unity for direct comparison of the XANES spectra of the samples. The absorption spectrum for the isolated atom, µ0(E), was approximated by the same cubic spline. The EXAFS function, χ(E), was obtained from the relation: χ(E) ) {µ(E) - µ0(E)}/µ0(E). Further analysis was performed in k space, where the photoelectron wavevector k is defined as k ) [(8π2m/h2)(E - E0)]1/2, where m is the electron mass and E0 is the threshold energy of a photoelectron at k ) 0. The resulting EXAFS spectra were k3-weighted and Fourier transformed in the range of ∼3 Å-1 e k e ∼13 Å-1 with a Hanning apodization function to reduce the truncation effect. Short-circuit photocurrent (Jsc) and open-circuit voltage (Voc) were measured with a Keithley model 2400 source measure unit. The light source was a 1000-W xenon lamp (Oriel, 91193). The light was homogeneous up to 8 × 8 in., and its radiant power was adjusted with respect to Si reference solar cell (Fraunhofer Institute for Solar Energy System; Mono-Si + KG filter; Certificate No. C-ISE269) to about one-sun light intensity (100 mW/cm2).
Park et al.
Figure 1. X-ray diffraction patterns of SnO2, Zn-modified SnO2, and ZnO films. Arrows indicate the peak positions of crystalline wurtzite ZnO.
Morphology of Zn-O Type Shells on SnO2 and TiO2 Cores. Figure 1 shows the XRD patterns of annealed nanoparticle films of unmodified SnO2 and ZnO and Zn-covered SnO2 at various atomic Zn/Sn ratios. The sample with a Zn/Sn ratio of 0.1 exhibits only the SnO2 phase, implying that the Zn treatment does not change significantly the bulk crystal structure of SnO2 and that any Zn-O overlayer is either amorphous or too thin to be detected. Also, at the 0.1 Zn/Sn ratio, there is no XRD pattern indicative of crystalline ZnO. However, the presence of a small amount of crystalline ZnO cannot be excluded because the scattering efficiency of ZnO is much lower than that of SnO2, owing to Zn having a lower atomic weight than Sn. In contrast, when a large quantity of zinc acetate is mixed with the SnO2 colloid (0.5 Zn/Sn ratio), crystalline ZnO is clearly observable. It is also seen in Figure 1 that adding Zn to SnO2 increases the full-width at half maximum (fwhm) of the XRD peaks, indicating
that SnO2 particles become smaller. The crystalline particle size D can be estimated from the Scherrer equation,37 D ) 0.9λ/β cos θ, where λ is the wavelength of the X-ray source, β is the fwhm, and θ is the Bragg angle of the specific diffraction peak. The particle size of the pure (unmodified) SnO2 sample is estimated to be 6.5 nm from the peak at 2θ ) 26.6°. The respective particle sizes of the 0.1 and 0.5 Zn/Sn ratio samples are, on the other hand, about 4.8 and 4.0 nm. The apparent smaller size of the SnO2 particles can be explained by the formation of a mixed Zn/Sn oxide layer (SnxZnyOz) at the interface between the two materials, disrupting the crystalline SnO2 lattice at the surface. In addition to the mixed oxide interface, the XRD of the 0.5 Zn/Sn ratio sample indicates ZnO nanocrystals of 28 nm diameter for 2θ ) 31.8°. Figure 2 shows SEM micrographs of SnO2 and Znmodified SnO2 films. Surface modification of SnO2 nanoparticles is seen to have a significant effect on film morphology. The pure SnO2 film (Figure 2a) is compact, mesoporous, and consists of nanometer-sized particles. In contrast, the 0.1 Zn/Sn ratio film (Figure 2b) consists of 500 nm diameter clusters, which, based on XRD results, is likely made up of small (ca. 5 nm) SnO2 particles with a mixed surface oxide layer. For the 0.5 Zn/Sn ratio film, the SEM micrograph (Figure 2c) indicates further change in morphology. SEM micrographs of Zn-modified samples at low magnification (panels d and e of Figure 2) show that the films have macropores associated with the presence of the 500 nm sized clusters. Furthermore, along with the formation of large clusters in Zn-modified SnO2 electrodes, the samples become more opaque with higher Zn/SnO2 ratios, indicating that adding Zn to SnO2 enhances the light scattering of the film. Figure 3 shows Zn K-edge XANES spectra of films of Zn-modified SnO2, ZnO, and Zn(OH)2. XANES spectra are sensitive to the local surroundings of an atom and provide information about chemical species. The XANES spectral shapes of the 0.1 and 0.5 Zn/Sn ratio samples differ from that of crystalline ZnO. The XANES of the low Zn/Sn ratio samples indicates the presence of zinc in the film. The features of the XANES spectra of the zincmodified samples are quite different from those in the pure ZnO, although in the case of the 0.5 Zn/Sn ratio sample, crystalline ZnO is detected by XRD. These results imply that the largest amount of zinc is present in an amorphous compound, which is likely on the SnO2 surface. The XANES spectrum for pure ZnO displays a strong
(36) Rehr, J. J.; Mustre de Leon, J.; Zabinsky, S. I.; Albers, R. C. J. Am. Chem. Soc. 1991, 91, 5135.
(37) Cullity, B. D., Elements of X-ray Diffraction, 2nd ed.; AddisonWesley: Boston, MA, 1977.
Results and Discussion
Core-Shell Nanoparticle Films
Langmuir, Vol. 20, No. 10, 2004 4249
Figure 2. Scanning electron micrographs of (a) SnO2 film and Zn-modified SnO2 films with Zn/Sn ratios of (b) 0.1 and (c) 0.5. Micrographs in (d) and (e) are enlargements of those displayed in (b) and (c), respectively.
Figure 3. Zn K-edge XANES spectra of Zn-modified SnO2 film and reference samples of zinc oxide and zinc hydroxide (Zn(OH)2). The curve labeled Zn(OH)2-500 °C represents a heattreated Zn(OH)2 film at 500 °C for 1 h.
absorption at 9669 eV with a small shoulder at 9662 eV, which is ascribed to a 1s f 4p electronic transition of Zn2+ ions having tetrahedral symmetry.38 Corresponding peaks are observed for Zn(OH)2 but at a slightly lower energy. Similar peaks are observed in the XANES spectra of the Zn-modified SnO2 samples at 9664 and 9668 eV, which indicates that the valence state of zinc is also +2. The positions of these peaks correlate with those for Zn(OH)2, which suggests that Zn coordinates with water to form hydroxide species. However, inasmuch as there are no bulk Zn(OH)2 films after they are annealed at 500 °C, it follows that Zn is present in a surface layer at least in the case of the 0.1 Zn/Sn ratio film. Furthermore, when the spectrum of the 0.5 Zn/Sn ratio film is fitted to a linear combination of the ZnO and Zn(OH)2 spectrum, it is found
that a combination of 63 ( 2% of Zn(OH)2 and 37 ( 2% of ZnO, comprising a total of 100% of zinc atoms, fits the spectrum very well. This means that about 63% of the Zn atoms are located in Zn(OH)2-like environments as a hydrated shell layer on SnO2 and that about 37% of the Zn atoms are located in ZnO nanocrystals, consistent with XRD and SEM data. The positions of the edge peaks and the features at high energy are also reproduced in the fit of the spectrum of the 0.5 Zn/Sn ratio sample. In contrast, the spectrum of the 0.1 Zn/Sn ratio sample cannot be fitted by such a linear combination, indicating the absence of a significant amount of crystalline ZnO, which is consistent with XRD results (Figure 1). Furthermore, these results imply that the zinc in the 0.1 and 0.5 Zn/Sn ratio samples is a Zn-coordinated surface hydroxide species, which is not identical but similar to Zn(OH)2 (see discussion above). Figure 4 shows the amplitude of the Fourier transformed Zn EXAFS spectra for Zn-modified SnO2 samples as a function of Zn content. The peaks in this spectrum are due to reflections from atoms in the Zn vicinity. The first peak in the spectrum of the pure ZnO sample at 1.5 Å is due to reflection of the evanescent electron waves from the electron density surrounding the nearest neighboring atoms. In the case of wurtzite ZnO, zinc is tetrahedrally coordinated with oxygen at a Zn-O bond length of about 1.98 Å.39,40 The second major peak in the spectrum of the pure ZnO at 2.7 Å is attributed to reflection from the 12 zinc neighbors in the second shell at a Zn-Zn distance of 3.2 Å. It can also be seen in the spectrum of ZnO that there is a large amount of long-range order, corresponding to a crystalline structure, as is evident by the relatively intense peaks at 4, 5, and 6 Å. It can also be seen in Figure 4 that the positions of the first shell peak of the 0.1 and 0.5 Zn/Sn ratio samples at 1.5 Å coincide with that of ZnO and have approximately the same intensity, indicating that the Zn atoms are also (38) Rose, J.; Moulin, I.; Masion, A.; Bertsch, P. M.; Wiesner, M. R.; Bottero, J.-Y.; Mosnier, F.; Haehnel, C. Langmuir 2001, 17, 3658. (39) Kihara, K.; Donnay, G. Can. Mineral. 1985, 23, 647. (40) Sabine, T. M.; Hogg, S. Acta Crystallogr. 1969, B52, 2254.
4250
Langmuir, Vol. 20, No. 10, 2004
Figure 4. Fourier transformed Zn K-edge EXAFS spectra of ZnO and Zn-modified SnO2 films.
Figure 5. X-ray diffraction patterns of TiO2 and Zn-modified TiO2 films.
tetrahedrally coordinated to oxygen atoms. Furthermore, for the 0.5 Zn/Sn ratio sample, a large peak is found for the second shell that is about 1/2 of the intensity of that in the pure ZnO spectrum. Also, for the 0.5 Zn/Sn ratio sample, the positions of the peaks at larger distances correspond to those in the pure ZnO spectrum. These observations are consistent with the interpretation of the XRD and XANES data that about 37% of the Zn is present as large-grained crystalline ZnO. The remaining Zn atoms are, however, still tetrahedrally coordinated with oxygen, which is consistent with the presence of a zinc oxide or hydroxide on the surface of the SnO2 particles. Zinc cannot be substituted for Sn in the SnO2 rutile lattice, where the cations are surrounded by six anions. It is, however, possible that Zn occupies tetrahedral sites in the SnO2 lattice. At a 0.1 Zn/Sn ratio, the low intensity scattering (about 1/10 of that for ZnO) at distances of 3, 4, 5, and 6 Å indicates that a small amount of crystalline ZnO is also present. At such a low Zn/Sn ratio, as mentioned above, it is not possible to observe ZnO by XRD. Figure 5 shows XRD patterns of ZnO-modified TiO2 films annealed at 500 °C for 1 h. At all Zn/TiO2 ratios, the TiO2 phase is anatase, regardless of the amount of added Zn. No crystalline ZnO is evident in the XRD spectra. Also, adding Zn is seen to have no effect on the width of the XRD peaks, indicating that the particle size does not change appreciably. From the diffraction peak at 2θ ) 25.38°, a particle size of 20 nm is estimated. Figure 6 compares SEM micrographs of pure TiO2 and Zn-modified TiO2 films. In contrast to SnO2 films, Zn
Park et al.
modification of nanocrystalline TiO2 has no significant influence on the film morphology. Figure 7 shows the Zn K-edge XANES spectra of Zn-modified TiO2 films. The near-edge spectra of Zn-modified TiO2 samples with Zn/ Ti ratios lower than 0.1 differ from those of polycrystalline ZnO and Zn-modified SnO2 samples (Figure 3). Bulk crystalline ZnO shows two main absorption peaks separated by an energy of 7 eV. The peak at higher energy is normally more intense than the one at lower energy.41 When ZnO loses crystallinity and long-range atomic order, the peak no longer splits into two. The peak becomes narrower, while simultaneously undergoing an inversion of peak intensity.42 The spectral shape of the 0.1 Zn/Ti ratio sample is almost the same as that of the 0.1 Zn/Sn ratio sample, which shows two main absorption peaks of similar intensity and a peak separation of 4 eV. The samples with Zn/Ti ratios of 0.01-0.05 exhibit only a single intense peak at 9664 eV with a shoulder at 9675 eV and a broad feature centered at 9720 eV. The shape of the absorption spectra is similar to that for Zn2+ ions intercalated into V2O5 xerogels in which Zn exists in a four-coordinate environment with oxygen in a coplanar geometry.43 In the latter case, however, the first and most intense peak is observed at lower energy (9660 eV) with a shoulder at 9670 eV and a large broad feature centered at about 9705 eV. Because of the large energy difference between the spectral features of Zn2+ intercalated V2O5 xerogels and the 0.01-0.05 Zn/Ti ratio samples, we conclude that zinc ions are not coplanar with oxygen in the latter samples. We also investigated the Ti K-edge XANES spectra (not shown) and find no spectral change before and after Zn modification, indicating that electronic and geometric characteristics of TiO2 are not significantly influenced by the addition of Zn. Figure 8 shows the amplitude of the Fourier transformed Zn-EXAFS spectra for pure ZnO and Zn-modified TiO2 samples as a function of Zn content. The peak at 1.5 Å is reproduced in all of the spectra with an approximately equal intensity, indicating that in all cases Zn2+ ions are coordinated to four oxygen ions. The second peak at 3 Å is present in all samples but is most intense in the 0.1 Zn/Ti ratio sample. However, there are no peaks beyond 3 Å in the Zn-modified TiO2 samples that correlate with those in the ZnO sample, indicating the absence of longrange order around the zinc ions in the Zn-modified TiO2, which signifies that no significant amount of crystalline ZnO is formed. Assuming that a compact layer of ZnO having a density of 5.61 g/cm3 is formed around the TiO2 particles upon adding zinc, it is calculated that the respective thicknesses of the ZnO shells for Zn/Ti ratios of 0.01, 0.025, 0.05, and 0.1 are 0.2, 0.6, 1.1, and 2.3 Å. In view of the fact that the unit axes of wurtzite ZnO are a ) 3.25 and c ) 5.21 Å, none of these shell thicknesses is enough to form crystalline ZnO. A thickness of 2.3 Å at the Zn/Ti ratio of 0.1 is, however, sufficient to form a single Zn-O monolayer, which correlates with the observation of a strong second reflection in the EXAFS spectra. This result is also consistent with XRD (Figure 5) and XANES (Figure 6) measurements, where there is no evidence for crystalline ZnO. It follows that the zinc oxide is present as isolated islands at Zn/Ti ratios of
