Abstract

In this study, tungsten trioxide (WO3) thin films were electrodeposited on indium tin oxide (ITO) glass to form WO3-coated glass. The electrodeposition (ED) time () and ED current () were varied to control the film thickness and morphology. Furthermore, the crystallization of the thin films was controlled by annealing them at 250°C, 500°C, and 700°C. The results showed that the thickness of the WO3 thin films increased with and . The as-deposited thin films and those annealed at 250°C were amorphous, whereas the WO3 thin films annealed at 500 and 700°C were in the anorthic phase. Moreover, the amorphous WO3-coated glass exhibited high transmittance in visible light and low transmittance in near-infrared light, whereas the anorthic WO3-coated glass had high transmittance in near-infrared light. An empirical formula for determining the thickness of WO3 thin films was derived through multiple regressions of the ED process parameters.

1. Introduction

Tungsten trioxide (WO3) is a transition metal oxide and an n-type semiconductor. Different forms of crystalline have unique optical, electrical, and magnetic properties and are widely used in various fields. WO3 has electrochromic properties and is widely used in reversible electrochromic elements, smart windows, and mirrors [17]. It also shows high catalytic performance in degrading contaminants, often in combination with TiO2, Pt, Ag, Na+, and other materials [811]. Furthermore, WO3 is used in gas sensors because of its sensitivity to various gases such as CO, ethanol gas, NH3, NOx, ozone, and toluene gas [1216]. In addition to the aforementioned applications, WO3 can be used in lithium-ion battery electrodes [1720] and for lubricating mechanical components [2123] to improve their operating performance. Therefore, WO3 has high potential for use in developing multifunctional materials. Because of their characteristics, WO3 materials can be employed to modify traditional materials and endow them with specific characteristics to improve their performance.

WO3 thin films are typically fabricated using such methods as chemical vapor deposition [12, 24], sputtering [5, 6, 16], screen printing [12, 14], and electrodeposition (ED) [13, 25, 26]. ED involves a simple setup, is inexpensive, does not require special atmospheric conditions, and can easily form thin films on irregularly shaped substrates. WO3 thin films can be fabricated by using peroxotungstic acid in the ED process, and their thickness and morphology can be controlled by varying the ED voltage (), current (), and time (). Generally, ED is commonly used along with a three-electrode electrochemical cell in preparing a WO3 thin film [13, 2527].

In this study, ED using a two-electrode method was employed to fabricate a WO3 thin film on an indium tin oxide (ITO) glass substrate. The ED of the two-electrode method has advantages of lower equipment cost and higher deposition rate than the three-electrode electrochemical cell. By adjustment of the and , the structure, thickness, and morphology of the thin films were controlled. Furthermore, the crystallization of the thin films was controlled by annealing them at various temperatures (). Finally, an ultraviolet-visible-near-infrared (UV-VIS-NIR) spectrometer was used to determine the spectral characteristics of the WO3-coated glass.

2. Preparation of WO3 Thin Films

The primary materials used for preparing peroxotungstic acid were tungsten metal powder (particle size: 12 μm; purity: 99.9%; Acros, Belgium) and a low concentration (34.5%–36.5%) of hydrogen peroxide (H2O2, Sigma-Aldrich, Germany). Dissolving 3.0 g of tungsten metal powder in 30 mL of H2O2 yielded a colorless peroxotungstic acid solution. The exothermic reaction was conducted by cooling the acid solution in an isothermal unit (P-10, YSC, Taiwan) between 5°C and 10°C. The reaction formula for peroxotungstic acid preparation is as follows [25, 27]:

When the exothermic reaction was completed, the solution was filtered using filter paper (pore size, 10-μm) to remove the unreacted tungsten powder. The obtained clear solution was combined with 30 mL of 99.7% glacial acetic acid (CH3COOH, Showa, Japan), and the resulting solution was refluxed at 55°C for 12 h to decompose excess H2O2 and acetylate the peroxotungstic acid. Refluxing the peroxotungstic acid strengthened the adhesion between the thin film and the substrate [1]. Finally, the refluxed peroxotungstic acid was mixed with 60 mL of anhydrous absolute ethanol to produce a pale yellow coating solution for the ED.

Figure 1 shows the experimental setup used in this study. To adjust the electrode distance (20 mm), a platinum plate and ITO glass substrate (7 Ω/cm2, 1737, Corning, USA) were fixed to a polypropylene holder as the positive and negative electrodes. The holder was placed in a glass tank containing a thermoelectric module and temperature controller. Prior to ED, the ITO glass was cleaned to remove contaminants and improve the adhesion between the ITO glass surface and WO3 thin film. The ITO glass was cleaned in acetone for 20 min in an ultrasonic bath (5510R-DTH, Branson, USA), rinsed with deionized water, dried in nitrogen (N2, 99.995%), and heated to 60°C for 15 min on a hot plate (PC-420D, Corning, USA).

The platinum plate and ITO glass electrodes measured 30 mm × 20 mm × 0.8 mm (L × W × Thickness) and 20 mm × 20 mm × 0.6 mm (L × W × Thickness), respectively; the length of the electrode immersed in the coating solution was 18 mm. Therefore, the effective area of platinum plate and ITO glass electrodes for ED was approximately 7.5 and 3.6 cm2, respectively. Through ED, a WO3 thin film was formed on the ITO glass (negative electrode). Because the coating solution temperature affects the electrochemical reaction, the thermoelectric module was necessary to maintain the coating solution temperature at 15°C to preserve the stability and reproducibility of the WO3 thin-film fabrication process.

This study used ED along with two-electrode equipment (programmable DC power supply) and a constant process for preparing a WO3 thin film on ITO glass. This approach is advantageous because it involves low equipment costs and a simple process and achieves a relatively stable coating rate. The reaction formula of ED with peroxotungstic acid for the WO3 thin-film preparation is as follows [25, 27]:

In the ED process, a programmable DC power supply (PSS-3203, GW Instek, Taiwan) was used to maintain a constant (3, 5, 7, and 9 mA) for various (30, 40, 50, and 60 s). The current density for 3, 5, 7, and 9 mA of on the ITO electrode was approximately 0.83, 1.39, 1.94, and 2.50 mA/cm2, respectively. After the thin films were fabricated, the WO3-coated glass was cleaned by rinsing it with deionized water, drying it in N2, and heating it to 60°C for 10 min on a hot plate. WO3-coated glass samples were selected on the basis of preferred ED process parameters, which were determined according to the morphology of the WO3 thin films; the morphology was observed using visually and optical microscope. The selected WO3-coated glass samples were annealed at 250°C, 500°C, and 700°C for one hour. The annealing temperatures of 250°C, 500°C, and 700°C for WO3-coated glass mainly referred to the literature [13, 28, 29]. In general, in the annealing temperature of 350–550°C, the monoclinic WO3 phase was naturally formed, while the formation of triclinic (anorthic) WO3 phase was induced in the annealing temperature ranging from 550°C to 750°C. Above 750°C, the WO3 phase can be changed to orthorhombic phase [6]. The annealed samples were then cleaned according to the aforementioned procedure. The ED process parameters are listed in Table 1. For each process parameter, three samples were prepared for testing to ensure the reproducibility and reliability of the relevant experiments.

3. Characterization

The surface morphologies of the WO3 thin films were analyzed using a high-resolution field-emission scanning electron microscope (FESEM, S-4800, Hitachi, Japan). Crystallization was analyzed using a multifunction high-power X-ray diffractometer (XRD, D8 Discover SSS, Bruker, the Netherlands) with Cu Kα radiation, and all peaks measured by XRD were assigned by comparing them with those in the data of the International Centre for Diffraction Data [29].

The thickness of the WO3 thin films was determined by the height of film cross section, which was measured using the FESEM. The high magnification of the film cross section caused the observation range to be very small, which limited the representative of the film thickness. Therefore, the WO3 film thickness was also measured with a thin-film analyzer (F10-RT, Filmetric, USA); the measurement principle in this instrument is based on light interference in thin films. The crystallization will affect the refractive index and surface roughness of WO3 thin film, thereby affecting the accuracy of the thickness of a WO3 thin film measured by a thin-film analyzer. Therefore, the relevant setting parameters of the thin-film analyzer were corrected according to the test results of cross sections of the WO3 thin film measured by a FESEM to reduce the effects of refractive index and surface and roughness of the WO3 thin film on thickness measurement. To minimize measurement deviations, each fabricated process parameter (for three samples) for the film thickness involved three measurements (one measurement of the cross section by using the FESEM and two measurements of the thickness with the thin-film analyzer; each process parameter consists of a total of nine measured data). The five most concentrated measurements were then averaged, and this value was considered the experimental value of the samples.

The optical characteristics of WO3-coated glass for various process parameters were measured using a UV-VIS-NIR spectrometer (V-670, Jasco, Japan) with an integrating sphere (ISN-723, Jasco, Japan) to measure transmittance and reflectivity changes at wavelengths between 300 and 2400 nm at room temperature.

4. Results and Discussions

Figure 2 displays a photograph of a WO3-coated glass sample, along with various process parameters. The lower edge of the sample shows discoloration or stripping when the is higher than 5 mA. Therefore, process parameters corresponding to the values of 3 and 5 mA (range of the red dotted line) were considered in the subsequent experiments.

Figure 3 shows FESEM images of a WO3 thin film, along with various process parameters. An increase in and increased the surface roughness of the as-deposited WO3 thin film (Figure 3(a)). This observation was attributed to the ED rate increasing with , which caused the surface of the WO3 thin film to become increasingly uneven. In addition, the circuit impedance increased with the deposited film thickness, resulting in an unstable ED rate at long . Generally, a constant current leads to a relatively stable ED rate. However, the circuit impedance at each point on the thin-film surface during ED is affected by the uniformity of the film thickness and can cause the ED rate to fluctuate, which further affects the surface roughness and uniformity of the deposited film. The relationship between the crystallization of the WO3 thin film and can be analyzed as follows. In Figure 3(b), at °C, the surface morphologies of the WO3 thin films become compact, but no notable crystallization occurs. When the were 500 and 700°C, crystallization and crystal growth were evident (Figures 3(c) and 3(d)).

The diffraction patterns in Figures 4 and 5 show only the effect of the crystallization at various at  s. The reason for selecting this is that changes in the crystallization are easier to measure accurately in thicker films, and there is less interference in the diffraction signals from the substrate (ITO glass). The WO3 thin films grown at the as-deposited temperature and at °C did not show any crystallinity and were amorphous [27]. However, after annealing at 500 and 700°C, the anorthic (triclinic) phase (PDF numbers: 20-1323) [2, 3] appeared; this was seen in the diffraction pattern, and the XRD and FESEM analyses yielded the same result at this . The intensity of the (2 0 0) peak increased with and , mainly because a higher leads to a thicker WO3 thin film at a given , and a high results in the formation of an almost perfectly crystalline WO3 thin film.

Figure 6 is a cross-sectional FESEM image of the WO3-coated glass annealed at 250, 500, and 700°C. The cross section shows a three-layer structure comprising WO3, ITO, and glass. Defects formed in the glass substrate at °C may result from some ingredients or impurities melted in the glass substrate. Defects could affect the experimental determination of the optical characteristics and film thickness, which were measured using a spectrometer and thin-film analyzer. Therefore, the WO3-coated glass annealed at 700°C was omitted from the subsequent measurements of the optical characteristics and film thickness.

Figure 7 shows the variation of WO3 film thickness with various process parameters. The WO3 film thickness increases with and . However, the substantially affected the WO3 film thickness. The thicknesses of the as-deposited WO3 thin films at values of 3 and 5 mA were 132.8 and 216.1 nm at  s, respectively, and the ED rate was considerably high. The moisture of the WO3 thin film was removed, and the film became compact at °C; thereby, its average film thickness decreased. As the annealing temperature was raised to 500°C, some parts of the crystals grew larger to make the WO3 thin film become thicker at these parts but thinner at other parts, exhibiting an uneven morphology of the thin-film surface (Figure 3(c)) and increased average film thickness of WO3. In addition, increase in the film thickness of WO3 may result from change of the structure in the process of crystalline phase transition for WO3 thin film. The multiple regressions for deriving an empirical formula for the film thickness showed that the (sec), (mA), and (°C) were correlated with the WO3 thin-film thickness. The empirical formulas for the thickness of an as-deposited (, nm) and annealed (, nm) WO3 thin film are expressed as (3) and (4), respectively. The corresponding values are 0.946 and 0.843:

Figures 810 show the transmittance, reflectivity, and absorptance of the WO3-coated glass and ITO glass for various values of the process parameters. Table 2 lists the average transmittance, reflectivity, and absorptance of the WO3-coated glass and ITO glass in VIS (400–760 nm) and NIR (760–2400 nm) region for various values of the process parameters. The transmittance, reflectivity, and absorptance of the WO3-coated glass and ITO glass show similar trends at a given . In Figures 8(a), 8(b), 9(a), and 9(b), and Table 2, the amorphous WO3-coated glass and ITO glass (°C) show high transmittance and low reflectivity in the VIS region. However, the transmittance decreases but the reflectivity increases as the wavelength increases in the NIR region. The experimental results for the transmittance of the amorphous WO3-coated glass are in accordance with trends reported in the literature [3, 6].

The oscillations of amorphous WO3-coated glass in the VIS region, which are associated with interferences between atomic layers, indicate the good quality of the WO3 thin films [6]. The experimental results demonstrate that values of average transmittance of the amorphous WO3-coated glass are 76.6% to 82.9% and 21.3% to 26.1% in the VIS and NIR regions, respectively; values of average reflectivity of the amorphous WO3-coated glass are 11.4% to 17.8% and 43.4% to 49.8% in the VIS and NIR regions, respectively.

In Figures 8(c) and 9(c) and Table 2, the transmittance of the WO3-coated glass annealed at 500°C in the VIS region is slightly lower than that of the amorphous WO3-coated glass, but the transmittance and reflectivity of WO3-coated glass (°C) in the NIR region are, respectively, much higher and lower than those of the amorphous WO3-coated glass, showing that anorthic WO3 thin films have high transparency in NIR light [2]. In addition, the average reflectivity of the anorthic WO3-coated glass is at least 17.8% lower than that of the amorphous WO3-coated glass in the NIR region. The experimental results demonstrate that values of average transmittance of the anorthic WO3-coated glass are 74.8% to 77.6% and 58.5% to 69.0% in the VIS and NIR regions, respectively; values of average reflectivity of the anorthic WO3-coated glass are 16.1% to 18.8% and 9.7% to 25.6% in the VIS and NIR regions, respectively.

In Figure 10 and Table 2, the average absorptance of the amorphous WO3-coated glass samples (as-deposited and annealed at 250°C) is less than 6.5% in the VIS region, and the absorptance gradually increases in oscillations at wavelengths in the NIR region. However, the average absorptance of the anorthic WO3-coated glass (°C) is less than 7.6% in the VIS region, and the absorptance increases in a relatively stable state with the wavelength in the NIR region, unlike the amorphous WO3-coated glass. The experimental results demonstrate that values of average absorptance of the amorphous WO3-coated glass are 3.7% to 6.5% and 26.8% to 32.8% in the VIS and NIR regions, respectively; values of average absorptance of the anorthic WO3-coated glass are 3.7% to 7.6% and 15.2% to 22.2% in the VIS and NIR regions, respectively.

Although the transmittance, reflectivity, and absorptance of the WO3-coated glass samples fabricated at various and with the same show similar trends, they also display considerable differences at the same wavelength. The main reason for these differences is that the film thickness and surface roughness of the multilayered WO3-coated glass samples varied with the and . The thickness of the thin film causes constructive or destructive interference of incident light, which affects both transmission and reflection spectrum. Moreover, surface roughness cause incident light scattering and thus affects both transmission and reflection spectrum.

5. Conclusions

WO3 thin films were successfully prepared on ITO glass substrates through ED with various , , and settings, and their characteristics were determined using appropriate instruments and test methods. The findings of this study are summarized as follows:(1)To prepare WO3-coated glass through ED, the optimal were 3–5 mA at various (30–60 s), and less than 500°C can avoid damage and defects in the substrate (ITO glass).(2)The WO3 film thickness increased with and . However, also notably affected the WO3 film thickness.(3)The as-deposited WO3 thin films and WO3 thin films annealed at 250°C were amorphous. By contrast, WO3 thin films annealed at 500 and 700°C were in the anorthic phase.(4)The amorphous WO3-coated glass had high transmittance in VIS light and low transmittance in NIR light. However, the anorthic WO3-coated glass had high transmittance in NIR light.(5)Multiple regression and ED process parameters were used to derive an empirical formula for determining the thickness of WO3 thin films.(6)The amorphous WO3 thin films are typically used in an electrochromic device. In addition, the amorphous and uneven crystalline WO3 thin films can also be applied to the fields of spectral selection, gas sensors, lithium-ion battery electrodes, antiwear coating, and catalysis according to its characteristics (e.g., uneven surface of WO3 thin film may benefit reaction with external substances).

Competing Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgments

The authors would like to thank National Science Council of the Republic of China, for their financial support to this research under Contract no. NSC 102-2221-E-003-006-.