Journal of The Electrochemical Society, 153 共2兲 E11-E16 共2006兲 0013-4651/2005/153共2兲/E11/6/$15.00 © The Electrochemical Society E11 Influence of Coloring Voltage and Thickness on Electrochromical Properties of e-beam Evaporated WO3 Thin Films R. Azimirad, O. Akhavan, and A. Z. Moshfeghz Department of Physics, Sharif University of Technology, Tehran, Iran In this investigation, the effect of coloring voltage and thickness on optical and also electrochromical properties of WO3 thin films has been studied. The WO3 thin films were grown on glass and indium tin oxide coated conducting glass substrates by e-beam evaporation at different thicknesses of 200, 400, and 700 nm. Optical properties of the deposited samples were characterized in the ultraviolet-visible range 共300–1100 nm兲. The optical bandgap energy of the WO3 was obtained in a range of 3.3–3.5 eV showing its increase by decreasing the film thickness. The refractive index of the WO3 films was measured around 2 in the visible range. Surface chemical states of the films were studied by X-ray photoelectron spectroscopy, which showed the stoichiometry of our deposited tungsten oxide thin films is acceptable. Atomic force microscopy was used for studying surface morphology of the deposited films. The electrochromic properties of the WO3 films were characterized using a lithium-based electrolyte. It was shown that there is an optimum coloring voltage for each film thickness, which maximizes the change in optical density during electrochromic process. The coloration efficiency of the samples at the optimum voltage was linearly improved by increasing the film thickness at a constant wavelength 共500 nm兲. © 2005 The Electrochemical Society. 关DOI: 10.1149/1.2137653兴 All rights reserved. Manuscript submitted June 20, 2005; revised manuscript received August 16, 2005. Available electronically December 14, 2005. Tungsten oxide 共WO3兲 thin film is one of the most important transparent metal oxides due to its potential technological applications in selective catalysts,1,2 electrochromic devices,3-19 gas sensors,20-22 gaschromics,23 and optical recording.24 These films are typically made by thermal evaporation,25 sputtering,9 solution spray,7 pulsed laser ablation,26 chemical vapor deposition,27 sol-gel coating,28 and electrodeposition.11,19 Electrochromic thin films are of considerable technological interest due to their potential applications in “smart-window”,29 antiglare rear view mirrors of automobiles,30,31 nonemissive displays,32,33 and recently, hybrid organic/inorganic memory devices.34 The electrochromic process is a reversible change in optical properties of a particular material that occurs when the material is electrochemically oxidized 关loss of electron共s兲兴 or reduced 关gain of electron共s兲兴, and is of great academic and commercial interests. Tungsten oxide is the most important inorganic material for electrochromic applications after the first report on its electrochromic property in 1969. This material remains the most promising candidate for large-scale use in electrochromic devices. Quantitative parameters of the electrochromic properties of tungsten oxide 共efficiency, reversibility, and kinetics of colorationbleaching process兲 strongly depend on its structural, morphological, and compositional characteristics, and therefore on the deposition techniques and deposition parameters. There are some reports on the effect of the deposition parameters such as substrate temperature,35 oxygen pressure,4,15 influence of substrate,12 proton insertion,13 water content of the film,14,16 film thickness,17 work temperature,18 and annealing temperature11 on the electrochromic properties of WO3 films. In this investigation, we present the results on variation of film thickness and coloring voltage on the electrochromic properties of e-beam evaporated WO3 thin films. In addition, optical properties and surface structure of the WO3 films, with different thicknesses, have been studied. Experimental Thin films of WO3 were deposited on microscope slide glass and indium tin oxide 共ITO兲 coated glass 共Isfahan Optic Industries Co., commercial ITO ⬃1 ␮m thick and Rs ⬃ 100 ⍀/䊐兲 using the e-beam evaporation method at room temperature. The deposition system was evacuated to a base pressure of ⬃4 ⫻ 10−3 Pa. The evaporation source was composed of a compressed WO3 powder z E-mail: moshfegh@sharif.edu with 99.9% purity. The distance between the substrate and the source was 36 cm. The rate of evaporation was about 0.7 nm/s, which was controlled by a quartz crystal thickness monitor. In this work, the thickness of deposited films was considered about 200, 400, and 700 nm measured by the stylus and optical techniques. Optical transmission and reflection measurements of the deposited films were performed in a range of 300–1100 nm wavelength using a Jascow V530 ultraviolet 共UV兲 visible spectrophotometer with a resolution of 1 nm. X-ray photoelectron spectroscopy 共XPS兲 having a Specs EA 10 Plus hemispherical analyzer with an Al K␣ anode at an energy of 1486.6 eV was employed to study the atomic composition and chemical state of the tungsten oxide thin films. The pressure in the ultrahigh-vacuum surface analysis chamber was less than 1 ⫻ 10−7 Pa. All binding-energy values were determined by calibration and fixing the C 共1s兲 line to 285.0 eV. The XPS data analysis and deconvolution were performed by SDP 共version 4.0兲 software. Surface topography of the deposited films at nanoscale was investigated by Thermo Microscope Autoprobe CP-Research atomic force microscopy 共AFM兲 in air with a silicon tip of 10 nm radius in contact method. The electrochromic properties of the deposited WO3 thin films on ITO/glass substrates were investigated. The tungsten oxide film, as a working electrode, was electrochemically cycled in a 1 M LiClO4 in propylene carbonate 共PC兲 electrolyte in a glass test vessel, using a bare ITO film as the counter electrode. Since this electrochromic structure can be applied in the smart windows working with two electrodes, here we did not use the three-electrode setup for study of the coloration/bleaching process. Although without using a three-electrode arrangement the coloration data are qualitative, our arrangement can also be used to assess relative trends between coloration behaviors for the films with the same and different thicknesses. In addition, the films of different thicknesses colored at the same potential and time can be reasonably compared. The transmittance of the WO3 thin films was studied in two states. In the first state, it was measured as a function of time at a constant wavelength of 500 nm 共in which the eye has a high sensitivity兲 at different coloring voltages, and correspondingly, negative voltages applied for bleaching the films. In addition, during the electrochromic process, the magnitude of the current between the two electrodes was recorded. In the second state, it was measured by the spectrophotometer in a range of 300–1100 nm wavelength for the films colored after a constant time 共90 s兲 at the different coloring voltages. E12 Journal of The Electrochemical Society, 153 共2兲 E11-E16 共2006兲 Figure 1. Optical transmittance 共T兲 and reflectance 共R兲 spectra of the WO3 thin films deposited on glass substrate with different thicknesses. Figure 3. Spectral dependence of the refractive index of the WO3 films with different thicknesses. Results and Discussion Optical properties.— The optical transmission and reflection spectra of the WO3 films, deposited with the different thicknesses on glass substrate, have been shown in Fig. 1, in a range of 300–1100 nm wavelength. It is seen that the transmittance of all the films with the different thicknesses in the visible range varies from about 80 up to nearly 100% 共without considering the substrate contribution兲. Correspondingly, the maximum value of the reflectance for both the film and the substrate is about 20% 共the reflectance from the bare glass substrate is measured about 10%兲. The maxima and minima in the optical spectra of thin films arise due to differences between the refractive index of films and substrates, and so interference of multiple reflections originated from their surfaces. The sharp reduction in the transmittance spectra at the wavelength of ⬃350 nm is due to the fundamental absorption edge that was also reported previously in Ref. 6 and 36. The optical spectra are similar for the different thicknesses of the WO3 thin films, however reduction in the thickness leads to reduction in the number of maxima and minima fringes. The optical transmittance of the WO3 films strongly depends on the oxygen content of the films. In fact, nonstoichiometric films with composition of WO3−x show a blue tinge for x ⬎ 0.03.37 All the prepared tungsten oxide films were highly transparent with no observable blue coloration, under our experimental conditions. The optical bandgap 共Eg兲 was evaluated from the absorption coefficient 共␣兲 using the standard relationship 共␣h␯兲1/2 = A共h␯ − Eg兲, for indirect allowed bandgap of semiconductors. Here, ␣ was determined near the absorption edge using the simple relationship ␣ = ln⌊共1 − R兲2 /T⌋/d, where d is the thickness of the film. More about the optical bandgap calculation is reported in Ref. 38. Therefore, the plot of 共␣h␯兲1/2 vs h␯ can show a linear behavior and the intercept on the h␯ axis gives the optical bandgap energy 共Fig. 2兲. Using the above-mentioned procedure, a value of Eg for 200, 400, and 700 nm thicknesses of WO3 layer was obtained about 3.5, 3.4, and 3.3 eV, respectively. Therefore, by increasing the thickness of the WO3 thin films, the value of Eg decreased. For evaporated WO3 films, one has found 2.7 ⬍ Eg ⬍ 3.5 eV.4,6,36,38 The refractive index 共n兲 of the thin films can be determined by using the transmittance and reflectance spectra. It is well known that when the refractive index of a film is higher than the refractive index of its substrate, the maxima and the minima of the reflectance spectrum occur at the wavelength ␭ related to film thickness according to the equations of d = 共m − 1/2兲␭/2n and d = m␭/2n, respectively, where m is a positive integer.39 Therefore, by knowing the film thickness, one can obtain the refractive index. The spectral dependence of n for the different film thicknesses is presented in Fig. 3. In general, values of the refractive index are within 1.9–2.5 in the spectral range of 300–1000 nm. Meanwhile, the refractive index is approximately constant in wavelengths higher than 500 nm. In addition, the thickness of the films does not change n significantly. It is seen that variation of n is larger at the shorter wavelengths. A similar behavior for the refractive index of WO3 thin films was also observed by others, using different techniques and/or relations.4,6,36 XPS analysis.— The stoichiometry of the WO3 films was studied by XPS. The only elements detected on the film surface were tungsten and oxygen. A small trace amount of carbon impurity was also observed. Figure 4a shows the XPS spectrum of the WO3 thin film with 200 nm thickness in the binding energy range of 41-32 eV. In this range, two pronounced peaks including W 共4f7/2兲 at 35.85 eV and W 共4f5/2兲 at 38.0 eV appear that show the existence of WO3 Figure 2. Plots of 共␣h␯兲1/2 vs h␯ for the WO3 thin films with different thicknesses: 共a兲 200, 共b兲 400, and 共c兲 700 nm. Journal of The Electrochemical Society, 153 共2兲 E11-E16 共2006兲 E13 Figure 4. Deconvoluted XPS spectra of 共a兲 W 共4f7/2,5/2兲 and 共b兲 O 共1s兲 photoelectron lines of the WO3 thin film with 200 nm thickness. composition.4,40 The area ratio of these two peaks is 0.75, which is supported by the spin-orbit splitting theory of 4f levels. Moreover, the structure is shifted by ⬃5 eV toward higher energy relative to the metal. It is thus clear that the main peaks in our XPS spectrum represent the W6+ state on the surface. Figure 4b shows the XPS spectrum of O 共1s兲 peaks at room temperature for the WO3 thin film with 200 nm thickness. This peak has been deconvoluted into two components after a Shirley background subtraction. The full width at half maximum 共fwhm兲 of the deconvoluted peaks was considered to be 1.7 eV, and they were constructed using a mixed 80% Gaussian-20% Lorentzian character. The binding energy of the first component at 531.1 eV is related to the oxygen in WOx and the other one at 532.9 eV corresponds to the oxygen bound of water molecules in the film structure or adsorbed on the surface.41 The stoichiometry of the evaporated WO3 film is determined based on the intensity ratio of the core-level binding-energy peaks, namely, W 共4f兲 and O 共1s兲 peaks. By using only the lower binding energy of the O 共1s兲 component, the stoichiometry of the films grown at room temperature was found to be x = 3.1 ± 0.05. The excess measured oxygen, relative to a stoichiometric WO3, can be related to the OH binding in the thin film. This quantitative analysis reveals that the stoichiometry of our deposited tungsten oxide thin films is very close to the stoichiometry of the transparent one, i.e., WO3. Moreover, we have found that the measured stoichiometry of the films is independent of the thickness. AFM surface morphology.— To study the surface morphology of the e-beam evaporated WO3 thin films and its thickness dependence, AFM analysis was utilized. Figure 5a shows a 3D view of the surface topography of the WO3 film with 200 nm thickness in a scale of 1⫻1 ␮m. It can be seen that the texture of the film is relatively continuous and homogeneous with an amorphous structure, as also reported by other investigators.42,43 In fact, the value of Figure 5. 共a兲 3D-AFM image of the WO3 thin film with 200 nm thickness and 共b兲 grain size distribution on the film surface. the root mean square of the surface roughness for this sample was measured about 1 nm. The surface contains very small polycrystalline and nanostructure grains. Figure 5b presents a surface distribution of the grain size showing that a considerable amount of the surface grains 共about 50%兲 have a size of around 30 nm. A very similar surface topography has been also observed for the surfaces of the WO3 thin films with the other thicknesses 共not shown here兲 indicating an independence of surface roughness and grain size on the film thickness. Electrochromical characterizations.— The coloration-bleaching kinetics and the optical modulation of the WO3 films were followed by putting the ITO and WO3 electrodes in the prepared electrolyte, and then applying different coloring voltages to them. Figure 6 illustrates the optical transmission variation of the WO3 films with the different thicknesses as a function of time at the constant wavelength 共500 nm兲 and the various applied coloring voltages. By applying the voltages to the electrodes at t = 10 s and after 2 s delay, the transmittance of the films continuously decreased and they were colored. For all the samples, the considered coloring voltages were disconnected at t = 90 and 10 s later the polarity of the applied voltages was inverted. By changing the polarity, the transmittance of the films was increased and they were bleached with a rate faster than the coloration process. The coloration and bleaching processes of WO3 films are associated with the intercalation and deintercalation of Li+ ions and electrons in the films according to the following reaction3,44 E14 Journal of The Electrochemical Society, 153 共2兲 E11-E16 共2006兲 Figure 7. Variation of ⌬OD vs the applied coloring voltage at ␭ = 500 nm for the WO3 thin films with different thicknesses. Figure 6. Optical transmission variation of the WO3 films in 1 M LiClO4 + PC electrolyte as a function of time, at ␭ = 500 nm and for various coloring voltages, with different film thicknesses: 共a兲 200, 共b兲 400, and 共c兲 700 nm. WO3共bleached兲 + xLi+ + xe− ↔ LixWO3共colored兲 关1兴 Thus, the films can be reversibly made transparent by electrochemical oxidation and colored by reduction in a Li+ containing solution. An equal transmittance of the films before coloration and after bleaching shows that the electrochromic reaction of the WO3 films for that coloring voltage is reversible. It is observed from Fig. 6 that by increasing the coloring voltage up to its optimum value, the transmittance of the colored films is decreased and the rate of the coloration is increased. But after applying a higher voltage, the transmittance of the bleached films did not show the reversible process. This can be due to an increase of the x value in LixWO3 so that the reversibility of Reaction 1 cannot be valid.4,44,45 Moreover, it can be seen that the coloration time of the WO3 films at the optimum voltage is reduced to 10 s, which is the minimum coloration time for our experimental conditions. To understand the effect of thickness and the coloring voltage on the film coloration, we have calculated the change of optical density 共⌬OD兲 using the following relation: ⌬OD = ln关T共t1,␭兲/T共t2,␭兲兴, where t1 = 12 s and t2 = 92 s are selected. Figure 7 shows variation of ⌬OD as a function of the applied coloring voltage for the different film thicknesses. It is evident from Fig. 7 that by increasing the coloring voltage, ⌬OD is increased, and at the optimum value of the coloring voltage, ⌬OD has its maximum amount. In addition, we have found that by increasing the thickness of the films, ⌬OD has been increased. It is clear that for the WO3 film with our minimum thickness, i.e., 200 nm, the variation of ⌬OD is very sensitive to the coloring voltage near its optimum value, so that increasing the voltage by 2 V resulted in reduction of the ⌬OD to one-half of its maximum value. Figure 8a shows variation of the electrochromic current between the two electrodes as a function of time for the WO3 film with 400 nm thickness. As can be seen, just after applying the coloring voltages, a sharp increase of the electrochromic current was observed. Then, the current was decreased slightly and continuously. When the polarity was changed, the current behaved similarly to the coloration process but with negative values. In general, it has been found that by increasing the coloring voltage up to its optimum value, the current is increased. A similar behavior was also observed for the other film thicknesses 共not shown here兲. During the electrochromic process, the total charges transmitted through the electrodes can be measured using the area surrounded by the current-time graphs. Figure 8b illustrates the electrochromic charge density of the WO3 layer with different thicknesses vs the applied coloring voltage. It shows that by increasing the applied voltage up to the optimum value, the charge density was increased for all the film thicknesses, corresponding to an increase of the x value in LixWO3 and a decrease of the transmittance after the coloration process 共see also Fig. 7兲. However, by increasing the coloring voltage to the values higher than its optimum value, the charge density was reduced and this reduction is sharper for our lowest film thickness 共200 nm兲. It can be seen that the behavior of the charge density is similar to the behavior of ⌬OD, unless the maximum value of the charge density does not depend on the film thickness. An important parameter for electrochromic films is the coloration efficiency, CE共␭兲, which is defined as CE共␭兲 = ⌬OD共␭兲/Q, where ⌬OD共␭兲 represents change in the optical density at a wavelength of ␭, and Q is transfer of charge density in terms of C cm−2. The Journal of The Electrochemical Society, 153 共2兲 E11-E16 共2006兲 E15 Figure 10. 共a兲 Transmission spectra of the WO3 film with 400 nm thickness at various applied coloring voltages and 共b兲 the maximum change of optical density vs the WO3 film thickness at different wavelengths, after electrochromic process. Figure 8. 共a兲 Electrochromic current variation of the WO3 film with 400 nm thickness in 1 M LiClO4 + PC electrolyte as a function of time for various applied coloring voltages and 共b兲 variation of the charge density 共Q兲 vs the applied coloring voltage for the WO3 thin films with different thicknesses. coloration efficiency, which is also called electrochromic efficiency 共EE兲, depends strongly on the sample preparation so that it may vary from 40 to 100 cm2 C−1.3,4,44 Figure 9 shows CE共␭ = 500 nm兲 for the different thicknesses of the WO3 films for Figure 9. Coloration efficiency 共CE兲 of the WO3 films as a function of film thickness at ␭ = 500 nm. the maximum ⌬OD共␭兲 observed at the optimum voltage. It can be seen that CE共␭兲 increased linearly with increasing the film thickness in our considered range. So, it seems that the electrochromic WO3 films with larger thicknesses yield better coloration efficiency. However, it was previously shown that the reversibility of WO3 electrochromic films during cycles reduced as film thickness increased.46 Therefore, selecting an optimum thickness of WO3 film for different electrochromic applications depends on both the coloration efficiency and the number of cycles considered in that application. The transmission spectra of the WO3 thin films with 400 nm thickness, colored at the various voltages after the constant time of 90 s, in the range of 300–1100 nm wavelength have been shown in Fig. 10a. Before applying the coloring voltage, the transmittance of the WO3 /ITO/glass structure is very similar to the WO3/glass structure 共see Fig. 1兲. It is seen that after applying the voltage, the transmittance of the film was decreased, which shows Li+ insertion to the film and thus its coloration. The reduction of the transmission is stronger at the long wavelengths. Therefore, the transmittance of the colored films at the near-IR range is less than the transmittance in the visible range. By increasing the coloring voltage, the films show more coloration and so further reduction in the transmission. In addition, the optical fringes disappear, which shows reduction of the interference effect by increasing absorption of the films. After coloration of the films at different applied voltages, the maximum transmittance occurred at a wavelength of about 400 nm, because the color of LixWO3 is blue. A similar trend was also observed for the other film thicknesses 共not shown here兲. But, by increasing the film thickness, the transmission was reduced at the optimum coloring voltage. Moreover, the transmittance of the maximum colored WO3 films with 700 nm thickness in the range of the near-IR is E16 Journal of The Electrochemical Society, 153 共2兲 E11-E16 共2006兲 nearly zero and in the visible range is less than 10%. Using the optical spectra shown in Fig. 10a and ⌬OD relation, Fig. 10b presents the change of the optical density of the WO3 films as a function of the thickness, at the maximum coloration of the films 共optimum coloring voltage兲 and for the different wavelengths. The change in the optical density is approximately independent of the film thickness in the wavelengths near the absorption edge. But at the longer wavelengths, the thickness dependence is more pronounced and at the visible range it shows a linear behavior. The maximum changes in ⌬OD have been observed at the wavelength of 700 nm and similarly at the higher wavelengths. Conclusions Tungsten oxide thin films were grown by the e-beam evaporation technique with different thicknesses including 200, 400, and 700 nm. By using UV-visible spectrophotometery, it was shown that the deposited films are highly transparent 共about 90% in the visible range兲. The Eg of the WO3 films was obtained about 3.5, 3.4, and 3.3 eV for 200, 400, and 700 nm thicknesses, respectively. In addition, the mechanism of the optical transition of the films was found to be indirect allowed. The value of the refractive index of the films was determined around 2 at ␭ = 500–1000 nm. By decreasing the wavelength, the value of n was increased to about 2.5 at 300 nm wavelength for the different film thicknesses. XPS analysis showed that the deposited films are pure and stoichiometric. Using AFM analysis, the surface roughness of all the films was measured about 1 nm with a nanometric grain size of around 30 nm showing a nanostructure surface. The electrochromic properties of the WO3 thin films grown at the different thicknesses were measured in 1 M LiClO4 + PC electrolyte. For each thickness, we have found an optimum coloring voltage minimizing the transmittance of the films. At this optimum voltage, the coloring time was reduced to its minimum value 共⬃10 s兲. The CE of the WO3 film with 700 nm thickness was measured about 60 cm2 C−1 at ␭ = 500 nm, after applying its optimum coloring voltage. Moreover, it decreases linearly with decreasing the film thickness. Meanwhile, the ⌬OD of the films with the minimum thickness 共200 nm兲 shows a high sensitivity to the applied coloring voltages. The transmittance of the films was reduced by applying different coloring voltages, and this reduction was more pronounced at the near-IR region. For the WO3 with 700 nm thickness, we have observed a maximum coloration so that transmission of the colored films was less than 10% in the wavelength range of 300–600 nm and it was about zero in the range of 600–1100 nm. Acknowledgments The authors thank the Research Council of Sharif University of Technology and the High Technology Organization 共Ministry of Industries and Mines兲 for financial support of the project. Partial support by the Third World Academy of Science 共TWAS兲—Iran chapter is also appreciated. The assistance of A. Zavvarian, H. Heidari, S. Rafiei, A. Babapour, and S. Vaseghinia for various stages of the project is greatly acknowledged. References 1. D. Li and H. Haneda, J. Photochem. Photobiol., A, 160, 203 共2003兲. 2. S. Djerad, L. Tifouti, M. Crocoll, and W. Weisweiler, J. Mol. Catal. A: Chem., 208, 257 共2004兲. 3. P. M. S. Monk, R. J. Mortimer, and D. R. Rosseinsky, Electrochromism: Fundamentals and Applications, VCH, Weinheim 共1995兲. 4. C. G. Granqvist, Handbook of Electrochromic Materials, Elsevier, Amsterdam 共1995兲. 5. P. K. Biswas, N. C. Pramanik, M. K. Mahapatra, D. Ganguli, and J. Livaga, Mater. Lett., 57, 4429 共2003兲. 6. K. Gesheva, A. Szekeres, and T. Ivanova, Sol. Energy Mater. Sol. Cells, 76, 563 共2003兲. 7. P. S. Patil, S. B. Nikam, and L. D. Kadam, Mater. Chem. Phys., 69, 77 共2001兲. 8. N. S. Gaikwad, G. Waldner, A. Brüger, A. Belaidi, S. M. Chaqour, and M. Neumann-Spallart, J. Electrochem. Soc., 152, G411 共2005兲. 9. T. Pauporté, M. C. Bernard, Y. Soldo-Olivier, and R. Faure, J. Electrochem. Soc., 151, H21 共2004兲. 10. E. Syrrakou, S. Papaefthimiou, and P. Yianoulis, Sol. Energy Mater. Sol. Cells, 85, 205 共2005兲. 11. M. Deepa, M. Kar, and S. A. Agnihotry, Thin Solid Films, 468, 32 共2004兲. 12. A. Siokou, S. Ntais, S. Papaefthimiou, G. Leftheriotis, and P. Yianoulis, Surf. Sci., 566/568, 1168 共2004兲. 13. H. Kamal, A. A. Akl, and K. Abdel-Hady, Physica B, 349, 192 共2004兲. 14. D. L. Sun, S. Heusing, J. Puetz, and M. A. Aegerter, Solid State Ionics, 165, 181 共2003兲. 15. E. Masetti, M. L. Grilli, G. Dautzenberg, G. Macrelli, and M. Adamik, Sol. Energy Mater. Sol. Cells, 56, 259 共1999兲. 16. G. Leftheriotis, S. Papaefthimiou, and P. Yianoulis, Sol. Energy Mater. Sol. Cells, 83, 115 共2004兲. 17. P. S. Patil, P. R. Patil, S. S. Kamble, and S. H. Pawar, Sol. Energy Mater. Sol. Cells, 60, 143 共2000兲. 18. J. P. Matthews, J. M. Bell, and I. L. Skryabin, Electrochim. Acta, 44, 3245 共1999兲. 19. T. Pauporté, J. Electrochem. Soc., 149, C539 共2002兲. 20. I. Jiménez, M. A. Centeno, R. Scotti, F. Morazzoni, A. Cornet, and J. R. Morante, J. Electrochem. Soc., 150, H72 共2003兲. 21. J. Guérin, K. Aguir, M. Bendahan, and C. Lambert-Mauriat, Sens. Actuators B, 104, 289 共2005兲. 22. R. Ionescu, A. Hoel, C. G. Granqvist, E. Llobet, and P. Heszler, Sens. Actuators B, 104, 124 共2005兲. 23. A. Georg, W. Graf, R. Neumann, and V. Wittwer, Thin Solid Films, 384, 269 共2001兲. 24. Y. Takeda, N. Kato, T. Fukano, A. Takeichi, and T. Motohiro, J. Appl. Phys., 96, 2417 共2004兲. 25. M. Gillet, A. Al-Mohammad, and C. Lemire, Thin Solid Films, 410, 194 共2002兲. 26. G. J. Fang, K.-L. Yao, and Z.-L. Liu, Thin Solid Films, 394, 64 共2001兲. 27. P. Tagtstrom and U. Jansson, Thin Solid Films, 352, 107 共1999兲. 28. A. Patraa, K. Auddya, D. Gangulib, J. Livagec, and P. K. Biswas, Mater. Lett., 58, 1059 共2004兲. 29. H. N. Cui, M. F. Costa, V. Teixeira, I. Porqueras, and E. Bertran, Surf. Sci., 532/535, 1127 共2003兲. 30. C. M. Lampert, Mater. Today, 7, 28 共2004兲. 31. G. Liu and T. J. Richardson, Sol. Energy Mater. Sol. Cells, 86, 113 共2005兲. 32. P. Bonhôte, E. Gogniat, F. Campus, L. Walder, and M. Grätzel, Displays, 20, 137 共1999兲. 33. W. Lu, A. G. Fadeev, B. Qi, and B. R. Mattes, J. Electrochem. Soc., 151, H33 共2004兲. 34. S. Möller, C. Perlov, W. Jackson, C. Taussig, and S. R. Forrest, Nature (London), 426, 166 共2003兲. 35. J. Wang and J. M. Bell, Sol. Energy Mater. Sol. Cells, 43, 377 共1996兲. 36. E. E. Khawaja, S. M. A. Durrani, and M. A. Daous, J. Phys.: Condens. Matter, 9, 9381 共1997兲. 37. K. D. Lee, Thin Solid Films, 302, 84 共1997兲. 38. A. Z. Moshfegh, R. Azimirad, and O. Akhavan, Thin Solid Films, 484, 124 共2005兲. 39. M. Ohring, Materials Science of Thin Films, Academic Press, New York 共2002兲. 40. B. V. Crist, Handbook of Monochromatic XPS Spectra: The Elements and Native Oxides, Vol. 1, John Wiley & Sons Ltd., Chichester 共2000兲. 41. G. Leftheriotisa, S. Papaefthimioua, P. Yianoulisa, A. Siokoub, and D. Kefalas, Appl. Surf. Sci., 218, 275 共2003兲. 42. M. D. Antonik, J. E. Schneider, E. L. Wittman, K. Snow, J. F. Vetelino, and R. J. Lad, Thin Solid Films, 256, 247 共1995兲. 43. A. Al-Mohammad and M. Gillet, Thin Solid Films, 408, 302 共2002兲. 44. P. R. Somani and S. Radhakrishnan, Mater. Chem. Phys., 77, 117 共2002兲. 45. M. Denesuk and D. R. Uhlmann, J. Electrochem. Soc., 143, L186 共1996兲. 46. K. D. Lee, J. Korean Phys. Soc., 38, 33 共2001兲.