Na₂WO₄ was purchased from Aldrich. The detail procedure of preparing WO₃-modified TiO₂ sol is as follows. The preparation method was the same as synthesis of pure TiO₂ sol, as reported in the previous studies [18]. The only difference was that Na₂WO₄ and H₂O₂ were added in the heating step at 95˚C for 6 h under magnetic stirring. 6 h later, the transparent light yellow WO₃-TiO₂ sol was obtained. The molar ratio of TiO₂: H₂O₂ was 1: 6 and the weight ratio of WO₃: TiO₂ was 0.5:100, 1:100, 2:100 and 4:100, respectively. Ti(OH)₄ and Na₂WO₄ were converted to WO₃-TiO₂ crystallites at 95˚C in the presence of H₂O₂. It should be noted that there was no surfactant was added. The sol was very stable even after 2 years in stock.

WO₃-TiO₂ films were prepared by dip-coating method using the as-prepared WO₃-TiO₂ sols. The substrate was type 304 stainless steel. The total coating surface area of the substrate was 40 cm². Before coating, the substrate was cleaned by neutral abrasives and distilled water for several times, and then dried in an oven at 50˚C. The substrate was vertically soaked into the as-prepared sol for 2 min, then pull out with the constant speed of 60 cm/min for 6 times. The thickness of WO₃-TiO₂ films was kept around 120 nm. After coating, it was heated at 400˚C. The number in parentheses means the weight percentage of WO₃ in the sample (Figure 1).

The following methods were used to investigate the effect of WO₃-TiO₂ coating on anticorrosion capability of stainless steel 304.

WO₃-TiO₂ sol was light yellow transparent sol containing WO₃-TiO₂ nanoparticles dispersed in water. The yellow color was due to the presence of small amount of titanium peroxide [18]. If one decreased the amount of H₂O₂ in preparation, the color became light yellow. The pH values of TiO₂ and WO₃-TiO₂ sols are listed in Table 1. The pH value of TiO₂ sol was 8.7. Adding WO₃ in TiO₂ sol decreased the pH value very slightly because the tungsten precursor was acidic and its amount was very low.

The XRD patterns of all the samples are shown in Figure 3. The result in Figure 3 represents the anatase TiO₂ diffraction peaks located at 2θ = 25.4102˚, 37.9658˚, 48.1227˚, 62.7199˚, and 75.2245˚ corresponding to the anatase phase of (101), (004), (200), (204), and (215). The as-prepared TiO₂ had low intensity peaks for anatase, as shown in Figure 2(b), because the TiO₂ crystallites were very small, in agreement with literature data [14] [15] [16] [17] [18]. Figure 3 shows the XRD patterns of WO₃-TiO₂ samples. The diffraction peaks of these patterns coincide with the characteristic peaks of anatase TiO₂ and did not match any diffraction peak of WO₃. This is because the amount of WO₃ in WO₃-TiO₂ sample was very small. The characteristic peaks of WO₃．H₂O are at 2θ = 25.7164˚, 35.1076˚, and 52.7672˚. In order to check whether the diffraction peaks of WO₃ appeared in W-TiO₂ samples or not, we added more amount of W precursors in the sample. Figure 5 shows the XRD pattern of 1 wt% WO₃-modified TiO₂ using H₂WO₄ as the precursor. There were no WO₃ XRD peaks in Figure 5. Tatsuma et al. [3] reported that WO₃ was highly dispersed in the bulk phase of TiO₂ particles and a new solid such as WO₃ was not formed. Our results are in accord.

The structure of TiO₂ and WO₃-TiO₂ films were examined by SEM. The top-view surface structure and cross-section view of the as-prepared thin film are shown in Figure 4. Each sample had very uniform and smooth surface of film on the top of substrate, indicating that dip coating is a good way to prepare the thin film. The surface of TiO₂ film was smoother than those of WO₃-TiO₂ samples. All films were nonporous and dense films. Its hardness reached 2 H after drying at 100˚C, and reached 5 H after annealing at 400˚C.

The morphology of as-prepared WO₃-TiO₂ sols was analyzed by TEM and HRTEM. HRTEM image in Figure 5(a) shows that the TiO₂ particles were rhombus shape with the major axis and minor axis of 30-50 nm and 15-30 nm, respectively. Figure 5 also shows the morphology and particle size of the as-prepared WO₃-TiO₂. Comparing with pure TiO₂ sol, the particle sizes of WO₃-TiO₂ was smaller than that of TiO₂. It is in agreement with the literature data reported by Tryba et al. [19]. No WO₃ particles were observed and one can conclude that W atoms were incorporated into the structure of TiO₂, providing that the concentration of W precursor was low. In conclusion, we have successfully developed a method to prepare WO₃-TiO₂ sample.

The samples were loaded in the 5 wt% sulfuric acid solution to test their anticorrosion capabilities. The results are shown in Table 2 and Table 3. Table 2 shows the results tested for 24 h and Table 3 shows the results tested for 48 h.

The bare sample without coating had very low resistance to corrosion. One can see in Table 2 that sample WT (1) had the lowest weight loss by corrosion, indicating that it had the best anticorrosion performance among all the samples. Instead, WO₃ sample had low anticorrosion capability. Table 2 and Table 3 also show that the weight losses of all the samples coated with WO₃-TiO₂ films were almost the same. The difference was small because the short time test was used. The results indicated that TiO₂ has good anticorrosion property. Adding suitable amount of WO₃ in TiO₂ improved the anticorrosion property of TiO₂. WO₃ can store the electrons generated by TiO₂ under light irradiation [6]. It also can improve the separation of electron-hole pairs, resulting in high anticorrosion capability [10] [11] [15].

Anticorrosion property of the sample was examined by four-point probe to determine the conductivity of the test specimen. If the conductivity of the specimen is high, the oxidative capacity is high, and its ability of anticorrosion is low. By applying an electric potential (1 V) and current (0.4 A) to the sample, the conductivity was zero in each sample, except the bare 304 stainless steel as shown in Table 4, indicating that all samples with coating had high anticorrosion property in the presence of and absence of light illumination.

The corrosion resistance and the behavior of materials were evaluated by the potentiostatic and potentiodynamic polarization methods. Figure 6 shows that the I-V curves of all samples were very smooth in the range between −0.6 V and −0.05 V. It illustrates that no obvious electrochemical reactions occurred in this region. This result has been initially applied to explain that coating onto the substrate would retard the electrochemical reactions.

Figure 6 shows that WT (4) sample had the lowest current density and TiO₂ sample had the highest current density. High current density infers low anticorrosion property. Figure 6 shows that the best sample had the weight ratio of TiO₂: WO₃ of 100:4. The amounts of charge of all samples are listed in Table 5. The higher the amount of charge means the lower the anticorrosion capability.

The results show that all of the samples had better anticorrosion capability than the unmodified TiO₂.

“X” means that the conductivity was zero.