Journal of Sol-Gel Science and Technology 35, 127–136, 2005 c 2005 Springer Science + Business Media, Inc. Manufactured in The Netherlands.


Photo-Induced Hydrophilicity of TiO2 Films Deposited on Stainless Steel via Sol-Gel Technique S. PERMPOON∗ Laboratoire de Thermodynamique et de Physico-Chimie M´etallurgique, ENSEEG-INPG, BP 75, Domaine Universitaire, 38402 Saint Martin d’H`eres, France; Laboratoire des Mat´eriaux et de G´enie Physique, ENSPG-INPG, BP 46, Domaine Universitaire, 38402 Saint Martin d’H`eres, France [email protected]

M. FALLET Laboratoire des Mat´eriaux et de G´enie Physique, ENSPG-INPG, BP 46, Domaine Universitaire, 38402 Saint Martin d’H`eres, France ´ B. BAROUX AND J.C. JOUD G. BERTHOME, Laboratoire de Thermodynamique et de Physico-Chimie M´etallurgique, ENSEEG-INPG, BP 75, Domaine Universitaire, 38402 Saint Martin d’H`eres, France M. LANGLET Laboratoire des Mat´eriaux et de G´enie Physique, ENSPG-INPG, BP 46, Domaine Universitaire, 38402 Saint Martin d’H`eres, France Received August 3, 2004; Accepted March 31, 2005

Abstract. The photo-induced hydrophilicity of TiO2 films deposited on stainless steel substrates and silicon wafers using two different sol-gel routes has been investigated. The results indicate that crystalline titanium oxide films with excellent hydrophilic properties can be obtained on silicon wafer with both routes. XPS and XRD data reveal that films deposited on stainless steel exhibit crystallization features similar to those of films deposited on silicon wafers, and only differ by their oxidation degree owing to a TiO2 reduction process associated to a diffusion of iron ions during deposition of the acidic sol and/or high temperature post-treatment. Consequently, hydrophilic properties of films deposited on stainless steel are inhibited. The deposition of a SiOx barrier layer at the film/substrate interface allows preventing such a detrimental substrate influence. A low temperature deposition route of the TiO2 film associated to the presence of a barrier layer yields best results in preventing iron contamination of the films. Keywords:

1.

sol-gel process, photo-induced hydrophilicity, TiO2 films, stainless steel

Introduction

It is well known that, when exposed to UV radiation, titanium oxide (preferentially in its anatase polymorphic ∗ To

whom all correspondence should be addressed.

form) shows photocatalytic properties. Since Fujishima and Honda [1] discovered in 1972 that a TiO2 electrode can decompose water to hydrogen and oxygen by photo-illumination, various applications have been demonstrated, which are based on the photo-induced activity of TiO2 . The self-cleaning property is one of

128

Permpoon et al.

the most fascinating applications. It is based on two kinds of photo-induced mechanisms, namely a photocatalytic oxidative decomposition and/or a photoinduced hydrophilicity, which arise both from the photo-generation of electron-hole pairs. The former is ◦ caused by active oxidizing radicals (O− 2 , OH ) able to decompose the organic matter adsorbed at the TiO2 surface. These radicals are formed by redox reactions with photoelectrons and holes generated by the titanium oxide under UV irradiation. The latter is based on the trapping of adsorbed water on surface oxygen vacancies (O2 ), which are created by a photo-reduction of TiO2 (Ti4+ + e− → Ti3+ and 2O2− + 2h+ → O2 ), resulting in the formation of a superhydrophilic surface [2, 3]. Several works have shown that TiO2 films deposited on glass possess photocatalytic and photo-induced hydrophilic properties [4–8], resulting in a self-cleaning functionality. These properties are also particularly attractive in the domain of stainless steels, which are widely used in many application fields such as architectural, food, and pharmaceutical industry. Besides, Ohko et al. [9] revealed that TiO2 films deposited on stainless steel by a spray pyrolysis technique exhibited a cathodic photopotential, which was more negative than the corrosion potential of the bare metal, leading to a protection of the metal against corrosion. However, very few studies investigated the self-cleaning, and especially the photo-induced hydrophilic properties, of TiO2 films on stainless steel. Some works were devoted to TiO2 films prepared by an electrophoretic deposition technique [10–12]. These works pointed out that the diffusion of cations (Cr3+ , Fe3+ ) from the stainless steel substrate into the TiO2 films during high temperature heat treatment deteriorated the photocatalytic activity. They also suggested that these cations were dissolved in the lattice of titania and constituted electron-hole recombination centres, which inhibited the photocatalytic activity. Zhu et al. [13] mentioned that Cr and Ni still existed in the metallic state and only Fe was oxidized during TiO2 film annealing, owing to the great sensitivity of iron toward oxygen. Diffusing Fe ions formed an interlayer of iron oxide in the rhombohedral Fe2 O3 form at the interface between TiO2 film and steel substrate. Iron species also diffused into the TiO2 film, inhibiting thus the formation of a photoactive anatase TiO2 phase. On the other hand, Yu et al. [14] indicated that the amount of diffused Fe3+ and the ratio of Fe3+ to Fe2+ ions had a contradictory influence on the photocatalytic activity and photoinduced

hydrophilicity. They reported that, in optimal calcinations conditions, a critical amount of diffused Fe3+ ions enabled enhanced photocatalytic activity and photoinduced hydrophilicity, owing to an increasing amount of charge trapping sites induced by Fe dopants. Then, more photogenerated holes can oxidize the lattice O2− anions, resulting in more oxygen vacancies. On the contrary, Fe2+ ions retarded photogenerated hole transfers toward the film surface, leading to decreasing photocatalytic activity and photoinduced hydrophilicity. Hence, avoiding the detrimental effects of Fe ions diffusion during heat treatment at high temperature appears to be a challenge of primary importance to obtain photo-active TiO2 films on steel substrates. The deposition of an interfacial SiOx barrier layer may provide a first solution to this problem. Such a barrier layer has already been successfully used to prevent the diffusion of Na+ ions from soda-lime glasses into TiO2 films during high temperature calcinations [15]. Besides, Hattori et al. [16] indicated that a SiO2 barrier layer containing a small amount of fluorine issued from the soda-lime glass substrate improved the crystallinity of anatase TiO2 films. A second possibility for preventing iron contamination would rely on the preparation of anatase films at a temperature low enough to prevent ion diffusions. The sol-gel process is a versatile wet chemistry method that allows to answer such a need. In the present work, two titanium precursor sols have been investigated for the deposition of crystalline anatase films on stainless steel. A first sol was prepared using a standard sol-gel procedure [17], which required a high temperature post-treatment to yield crystalline TiO2 films. A second sol consisted of a suspension of anatase nano-crystallites, which was prepared using a specific sol-gel route [18]. In a previous work, we showed that this suspension was compatible with the room temperature deposition of optical grade photocatalytic films on polymer substrates. In this article, photo-induced hydrophilic properties of TiO2 films deposited on stainless steel from both sols have been studied and are discussed with respect to the heat treatment temperature and presence or not of a barrier layer. 2. 2.1.

Experimental Preparation of TiO2 Thin Films

Titanium oxide thin films were prepared using two different sol-gel routes. A classical method was based on the deposition of a mother solution (MS), which was

Photo-Induced Hydrophilicity of TiO2 Films Deposited on Stainless Steel

prepared by mixing tetraisopropyl orthotitanate (TIPT) with deionized water, hydrochloric acid (HCl), and absolute ethanol as a solvent [17]. The TIPT concentration in the solution was 0.4 M, and the TIPT/H2 O/HCl molar composition was 1/0.82/0.13. The solution was aged at room temperature for 2 days before deposition. The second method relies on the preparation of a liquid suspension (LS) of TiO2 nano-crystallite in absolute ethanol. This suspension was prepared from the mother solution using a multistep procedure. The mother solution was first diluted in an excess of deionized water (H2 O/TIPT molar ratio of 90) and then autoclaved at 130◦ C for 6 hours. Autoclaving yielded the crystallization of TiO2 particles in the liquid phase. An exchange procedure was then performed in order to remove water from the sol and to form a crystalline suspension in absolute ethanol. The final TiO2 concentration in ethanol was 0.24 M. For more data, the whole procedure has been described in a previous paper [18]. The final sol was composed of TiO2 nanocrystallites of about 6 nm in size, which were aggregated in the form of polycrystalline grains of 50–100 nm in size. Previous works showed that both MS and LS preparation conditions gave rise to very stable sols, which indicated that no gelation took place in MS sols, while no significant crystal aggregation took place in LS sols. Consequently, these sols could be used for several weeks in reproducible film deposition conditions. Three different substrates were used in this work: (100) silicon wafer as a reference substrate, AISI 304 stainless steel, and AISI 304 stainless steel coated with a SiOx barrier layer (100 nm in thickness), which was deposited via plasma enhanced chemical vapour deposition (CVD) or physical vapour deposition (PVD) technique. All uncoated and SiOx -coated stainless steel samples were provided by Arcelor Company. SiOx deposition conditions are not presented here for confidentiality reasons. Prior TiO2 film deposition, the substrates were ultrasonically cleaned in an ethanol/acetone mixture for 3 minutes, then rinsed with distilled water and dried with air spray. Then, thin films with a thickness of 90 or 60 nm were deposited by spin-coating (20 µL of sol, spin-speed of 3000 rpm) from MS (0.4 M) or LS (0.24 M) sols, respectively. After deposition, MS films underwent a room temperature sol-gel transformation through the traditional hydrolysis/polycondensation route, which led to a solid amorphous xerogel consisting of Ti O Ti chains with chain-end alkoxy and hydroxyl groups. In the case of LS films, room temperature solvent (ethanol) evapora-

129

tion took place, leading to crystalline TiO2 films. The films were subsequently heat-treated for 2 hours at temperatures ranging up to 500◦ C for MS films, and at 110◦ C for LS films. Heat-treatments were performed in air and the samples were directly introduced in the pre-heated oven. After heat-treatment, the films were cooled to room temperature under ambient condition. In the case of MS films, the heat-treatment in air was expected to complete the polycondensation or oxidation of remaining alkoxy and hydroxyl groups and to induce crystallization of pure TiO2 films. For LS films, the low temperature heat-treatment was only performed to complete solvent evaporation.

2.2.

Characterizations

The films were characterized by Fourier transform infrared (FTIR) transmission spectroscopy in the range of 4000–250 cm−1 with a resolution of 4 cm−1 using a Bio-Rad FTS-165 spectrometer. Spectra were recorded in room atmosphere without any purging and consisted of 300 scans. The spectra were analysed after subtraction of the bare substrate spectrum. The crystalline phase of TiO2 films was studied by X-ray diffraction (XRD) with Cu Kα radiation using a Siemens D5000 diffractometer. A 2-theta range of 15◦ to 60◦ was scanned with a step of 0.05◦ using an integration time of 15 seconds. The film thickness and refractive index were measured using a Gaertner L116B ellipsometer at 635 nm wavelength. Surface analysis was performed by X-ray photoelectron spectroscopy (XPS) using a XR3E2 apparatus from Vacuum Generator employing an Mg Kα (1253.6 eV). The X-ray source was operated at 15 kV for a current of 20 mA in an ultra high vacuum chamber (10−10 mbar). Photoelectrons were collected by a hemispherical analyzer at two different take-off angles of 30◦ and 90◦ in order to distinguish superfacial layer and deeper layer features, respectively. All spectra were calibrated with the Ti2 p line of the major component Ti4+ species at 458.5 eV. The photo-induced hydrophilicity was evaluated by the in-time variation of the water contact angle under UV light irradiation provided by a 100 W black light bulb with a light intensity of 21 mW/cm2 (B-100AP, Ultraviolet Product Co. Ltd.). A 365 nm filter was used and the distance between the lamp and the sample was fixed at 5 cm. Real-time experiments were performed under constant temperature (20◦ C) by using a KRUSS G 10 goniometer connected with a video camera. Water

130

Permpoon et al.

droplets of 0.5 µL volume were spread on the sample and water contact angle variations were measured at different points of the thin film surface.

3. 3.1.

Results and Discussions TiO2 Films Deposited on Silicon Wafer

Silicon wafer was used as a model substrate because, owing to its native SiO2 surface layer, no film/substate interaction can alterate the photoactive properties of the film. Figure 1 shows FTIR spectra of MS TiO2 films annealed at temperatures up to 500◦ C and a LS film heat-treated at 110◦ C. All MS TiO2 films heattreated in the range (110–350◦ C) exhibited a broad and weakly intense band around 400 cm−1 (Fig. 1(a)). MS films heat-treated at 400◦ C or more show two resolved bands at 430 and 260 cm−1 , which correspond to transverse optical (TO) vibration modes of anatase (Fig. 1(b)–(d)) [19]. We previously indicated that band intensities and widths are closely related to the crystallization degree and crystalline size [18]. For amorphous films, aforementioned bands are very broad and overlap, constituting a single band around 400 cm−1 . As crytallization occurs, they increase in intensity and sharpen, thus forming two separate bands. Results illustrated in Fig. 1(a)–(d) indicate, therefore, that MS films heat-treated below 350◦ C are amorphous and crystallization starts above this temperature. Moreover, the two bands are observed to sharpen and grow when increasing the thermal treatment temperature (Fig. 1(b)– (d)), implying an increase of the crystallization degree and/or crystallite size. The LS films heat-treated at 110◦ C exhibit two rather broad but well resolved

Figure 1. FTIR spectra of MS TiO2 films heat-treated at (a) 350◦ C, (b) 400◦ C, (c) 450◦ C, (d) 500◦ C, and (e) LS film heat-treated at 110◦ C.

bands at 430 and 300 cm−1 (Fig. 1(e)). The 430 cm−1 band corresponds to the TO mode of anatase, while the second band at 300 cm−1 corresponds to a combination of LO mode at 330 cm−1 and TO mode of anatase at 265 cm−1 [18]. Previous transmission electron microscopy (TEM) studies showed that the LS sol is constituted of nanocrystallites of about 6 nm in diameter. Such a small size would explained the rather broad and weak bands observed in Fig. 1(e). Bands illustrated in Fig. 1(b)–(d) suggest that larger crystallites are formed from MS films heat-treated at high temperature. XRD was used to investigate the crystalline phase constituting TiO2 films. Results in Fig. 2 confirm that MS TiO2 films heat-treated at 400◦ C or more are constituted of a single anatase phase and MS films heat-treated at 350◦ C or less are amorphous. Moreover, an observation of the (101) reflection of anatase indicates that the XRD peaks slightly sharpen and increase in intensity when increasing the temperatures. This result is consistent with FTIR data, indicating that crystallization and crystal growth are promoted by a temperature increase. The crystallite size was determined from the full width at half maximum (FWHM) of the (101) reflection of anatase (2θ = 25.4◦ ) using Scherrer’s formula (Table 1). The size increases from 26 to 33 nm when the temperature is varied from 400 to 500◦ C, confirming that crystalline MS films are composed of larger crystallites than LS films. XPS analysis allowed investigating the surface chemical state of TiO2 films. XPS spectra of films deposited on silicon wafer indicated that the films contained Ti, O, and C elements (not shown here), whose relative amounts depended on the experimental conditions (sol formulation, heat-treatment temperature).

Figure 2. XRD patterns of MS TiO2 films heat-treated at (a) 350◦ C, (b) 400◦ C, (c) 450◦ C, and (d) 500◦ C. Main anatase reflections are referenced.

Photo-Induced Hydrophilicity of TiO2 Films Deposited on Stainless Steel

131

Table 1. Crystallite sizes of TiO2 films deposited on silicon wafer and stainless steel and heat-treated at different temperatures. Size was determined from XRD data, except for the LS film (TEM measurements [18]). Crystalline sizes (nm) Heat treatment temperature

TiO2 on silicon wafer

TiO2 on stainless steel

MS film, 400◦ C

26.7

26.9

MS film, 450◦ C

28.0

28.7

MS film, 500◦ C

33.5

33.9

LS film, 110◦ C

6

–

Despite the use of a large HCl quantity in the precursor sols, no Cl− could be detected at the surface of MS films, and only traces were detected at the surface of LS films. Because the post-autoclaving exchange procedure of LS sols was performed in open atmosphere, it is likely that Cl− ions were partially vaporized in the form of Cl2 . In the case of MS films, Cl− ions were presumably released in the form of Cl2 during liquid film deposition and post-deposition heat-treatment. The Ti, O, and C XPS lines were fitted with a 10% Lorentzian/Gaussian ratio (L/G), as shown in Table 2. Spectra of the Ti2 p peaks may be decomposed into two contributions, which correspond to the different oxidation states of titanium (insert of Fig. 3). The peak situated at a binding energy E b (Ti2P3/2 ) = 458.5 eV corresponds to Ti4+ cations. The minor component of Ti3+ cations is situated at E b (Ti2 p3/2 ) = 457.5 eV. These energies are consistent with values reported in the literature [6], and indicate that our films are partially reduced. Concentration variations of reduced Ti3+ species are presented in Fig. 3 for TiO2 films heattreated at different temperatures. A significant amount of Ti3+ was detected at the surface of as-deposited MS Table 2.

Figure 3. Atomic percentage of Ti3+ species for MS TiO2 films heat-treated at different temperatures (full symbol) and a LS film heat-treated at 110◦ C (open symbol). Data were deduced from XPS measurements performed at 90◦ take-off angle (
, ), and 30◦ takeoff angle (, ). The insert shows Ti3+ and Ti4+ components of the XPS Ti2 p3/2 peak.

films or MS films heat-treated below 350◦ C (around 12 at% for a take-off angle of 30◦ ). The similarity between the Ti3+ concentration in these films and the HCl/Ti ratio in MS sols (13 mol%) and the absence of Cl− at the surface of MS films suggest that Cl2 formation and release from the film implies the low temperature reduction of Ti4+ (2Ti4+ + 2Cl− → 2Ti3+ + Cl2 ). Because Cl2 release proceeds from the outer surface of the film, this mechanism is efficiently completed in the surface layers of the film and occurs less efficiently in the deeper layers. This would in turn explain why Ti4+ reduction is weaker in the deeper layers (Fig. 3, take-off angle of 90◦ ). Figure 3 shows that the degree of oxidation sharply increases when MS films are heat-treated in the 350–400◦ C range. This thermal range corresponds to the onset of crystallization, which would indicate that MS film oxidation and crystallization proceed simultaneously. An optimum oxidation degree was measured on well crystallized MS films

Ti2p, C1s and O1s XPS peak fitting data.

Ti2p Binding energy (eV)

Ti4+

Ti3+

458.5 ± 0.2

457.5 ± 0.2

1.4 ± 0.1

1.5 ± 0.1

C1s

C H/C C

C O

O C O/C O

O C O

Binding energy (eV)

284.6 ± 0.2

285.8 ± 0.2

287.3 ± 0.2

288.5 ± 0.2

1.7 ± 0.1

1.7 ± 0.1

1.7 ± 0.1

1.7 ± 0.1

Ti O

Ti OH

C O

H2 O

529.9 ± 0.2

531.4 ± 0.2

532.5 ± 0.2

533.3 ± 0.2

1.6 ± 0.1

1.6 ± 0.1

1.6 ± 0.1

1.6 ± 0.1

FWHM

FWHM O1s Binding energy (eV) FWHM

132

Permpoon et al.

Figure 4. Dependence of the water contact angles on the UV irradiation time for MS TiO2 films heat-treated at (a) 350◦ C, (b) 400◦ C, (c) 450◦ C, (d) 500◦ C, and (e) for a LS film heat-treated at 110◦ C. Figure 4(f) shows the same dependence for a film without UV illumination.

heat-treated at 500◦ C, for which the Ti3+ concentration was about 2 at% both at the surface (30◦ take-off angle) and in the deeper layers (90◦ take-off angle). It is worthwhile mentioning that well crystallized LS films heat-treated at 110◦ C showed an oxidation degree similar to that of MS films crystallized at 500◦ C. The photo-induced hydrophilicity properties of films deposited on silicon wafer are illustrated in Fig. 4. For amorphous MS films heat-treated at 350◦ C or less, the water contact angles do not change upon UV irradiation (Fig. 4(a)), and remain similar to the value measured without UV illumination (Fig. 4(f)). MS films crystallized at 400◦ C or more exhibit a bi-regime behavior. The contact angle remains stable over a first period and then sharply decreases after some minutes of UV irradiation (Fig. 4(b)–(d)). Moreover, the hydrophilizing rate increases with increasing the heat-treatment temperature, i.e. the duration of the first regime decreases. For a MS film crystallized at 500◦ C, this first regime can no longer be observed and the contact angle continuously decreases, reaching a value of about 3◦ after 5 min UV irradiation. These data demonstrate that superhydrophilic properties primarily require well crystallized films. On the other hand, data illustrated in Fig. 4(b)– (d) would suggest that an increase of crystallization degree, crystallite size, and/or surface oxidation degree, beneficially influences photohydrophilic properties. Further studies are neccessary to better understand the respective influence of each parameter. The LS film heat-treated at 110◦ C exhibits an intermediary behavior between MS films crystallized at 400 and 500◦ C, i.e. a very short first regime followed by a fast con-

tact angle decrease. Hence, it can be concluded that crystallized TiO2 films with good hydrophilic properties can be prepared using a low temperature sol-gel procedure. To explain the slightly better performances of a MS film heat-treated at 500◦ C, it should be noted that the main difference with the LS film heat-treated at 110◦ C relies on the crystallite size, which was 33 nm for the former and 6 nm for the latter. This observation suggests that a larger crystallite size might induce better photohydrophilic properties. This feature has recently been studied and will be the subject of a future article. Besides, LS films are deposited from a crystallite suspension in ethanol it is not excluded that, in as deposited LS film and LS film heat treated at 110◦ C, crystallite are coated with an ethanol solvation layer, which might in turn influence hydrophilic properties.

3.2.

TiO2 Films Deposited on Stainless Steel

The XPS spectra of MS TiO2 films deposited on stainless steel are illustrated in Fig. 5. The photoelectron peaks of Ti2 p, O1s and C1s appear at a binding energy identical to that of TiO2 films deposited on silicon wafer. In addition, the photoelectron peak of Fe2 p is detected at a binding energy of 710 eV, which corresponds to oxidized Fe cations. This peak appears after a heat-treatment at 450◦ C and its intensity increases after annealing at 500◦ C. The iron percentage was estimated by XPS to be around 5 at% in the superficial layers (30◦ incidence) and around 9 at% in deeper layers (90◦ incidence). These results show that a thermally activated diffusion takes place from the stainless steel substrate into the TiO2 film. Because this diffusion arises from

Figure 5. XPS spectra of MS TiO2 films deposited on stainless steel and heat treated at (a) 400◦ C, (b) 450◦ C, (c) 500◦ C, and (d) for a LS film heat-treated at 110◦ C.

Photo-Induced Hydrophilicity of TiO2 Films Deposited on Stainless Steel

the substrate and XPS only probes the film surface, it is likely that ion species are already present in the deeper layers of the film at lower temperatures, i.e. diffusion starts below 450◦ C. Let us note that Cr and Ni, which are other constituents of the substrate, could not be detected in the films at any temperature. This result is consistent with data reported in the literature [13] indicating that no oxidation of Cr and Ni takes place and only Fe is oxidized during the film heat-treatment. It is therefore inferred that the diffusion of Fe into the film is associated to an oxidation process. As for films on silicon, the XPS Ti2 p peak could be decomposed in two Ti4+ and Ti3+ components. The atomic percentage of reduced Ti3+ species is illustrated in Fig. 6 for MS TiO2 films deposited on stainless steel. The results indicate that films annealed below 400◦ C show a tendency similar to that observed on silicon wafer. For films annealed above 400◦ C, the oxidation degree is observed to sharply decrease with increasing temperature. This titanium reduction takes place in a thermal range that corresponds to the detection of iron species at the film surface, which suggests that the Ti4+ reduction might be associated to the diffusion of iron into the film. However, it is also observed that the Ti3+ amount is greater at the surface of a MS film heattreated at 500◦ C than in the deeper layers of this film. This observation does not correlate a thermal diffusion mechanism of iron species from the substrate, which would imply a greater amount of Ti3+ in the deeper layers of the films. It is therefore inferred that the high temperature heat-treatment of MS films deposited on stainless steel substrates involves a complex oxydoreduction mechanism, which has not been elucidated yet.

Figure 6. Atomic percentage of Ti3+ species for MS TiO2 films deposited on stainless steel and heat-treated at different temperatures: 90◦ take-off angle (
) and 30◦ take-off angle ().

133

XRD peak identification and intensity measurements indicated that pure anatase MS and LS TiO2 films could be formed on stainless steel, with a crystallization degree similar to that of films deposited on silicon (not shown here). For a same heat-treatment temperature, the crystallite size of films deposited on silicon and stainless steel was similar (Table 1). This result implies that the diffusion of Fe ions into the TiO2 films does not impede the formation of anatase. To summarize, XPS and XRD data suggest that films deposited on silicon and stainless steel exhibit similar crystallization degrees and crystallite sizes, and only differ by their oxidation degree and iron presence at the film surface. Figure 7 shows the UV irradiation time dependence of water contact angles for MS and LS TiO2 films deposited on stainless steel. As for films on silicon, amorphous MS films do not show any photo-hydrophilicity (Fig. 7(a)). On the other hand, crystalline MS films exhibit a very slow contact angle decrease upon UV illumination, which does not significantly varies with crystallization temperature (Fig. 7(b) and (c)). This is presumably due to the detrimental effect of Fe ions diffused into the film during heat treatment, which act as electron-hole recombination centres and inhibit the creation of oxygen vacancies by photo-induced holes. Ti3+ ions issued from the reduction of Ti4+ can also trap photo-induced holes, and act as recombination centres. It has to be noted that MS films crystallized at 400◦ C also show a very weak photohydrophilicity, despite the lack of detection of Ti3+ and iron at the film surface by XPS. This suggests that a very weak amount of such species, which would be under the XPS sensitiv-

Figure 7. Variations of the water contact angles upon UV irradiation for MS TiO2 films deposited on stainless steel and heat-treated at (a) 350◦ C, (b) 400◦ C, (c) 500◦ C, and (d) for a LS film deposited on stainless steel and heat-treated at 110◦ C. Figure 7(e) shows the same dependence for a film without UV illumination.

134

Permpoon et al.

ity, might be sufficient to inhibit photo-hydrophilicity. According to previous results, it was inferred that the room temperature deposition of a TiO2 crystalline suspension could be an interesting alternative to prevent the thermally activated iron diffusion. However, water contact angle variations measured on a LS TiO2 film heat treated at 110◦ C showed the same trend as for crystalline MS films (Fig. 7(d)). This is presumably due to the fact that the stainless steel surface was corroded by the solution during liquid film deposition at room temperature. Corrosion is all the more likely to take place as the solution contains Cl− species. Though Cl− ions were partially released during the exchange procedure of LS sols, it is likely that a certain amount of Cl− remained in the sols. Accordingly, Cl− traces could be detected at the surface of LS films. During LS sol deposition, iron species can react with Cl− ions and undergo rapid migration in the liquid phase. The Fe amount detected by XPS for a LS film heat-treated at 110◦ C (2 at%) was weaker than for a MS film annealed at 500◦ C but presumably sufficient to inhibit photohydrophilic properties. Besides, this amount was similar at the film surface and in deeper layers. This similarity confirms that iron contamination occurred through a room temperature migration process in liquid phase rather than a thermally activated solid state diffusion from the substrate. Let us note that no iron could be detected at the surface of MS films heat-treated at low temperature. Contrary to LS films deposition, which only involves solvent evaporation and crystalline grain packing, MS film deposition proceeds through a solgel transformation, which rapidly yields the formation of an amorphous and rather dense xerogel film. This film presumably acts as a diffusion barrier for heattreatments at low temperature, but cannot prevent iron diffusion at high temperature. The deposition of a SiOx barrier layer at the film/substrate interface was thus considered in order to check whether iron diffusion was the actual cause of the photohydrophilicity lack in our films. A SiOx barrier layer was deposited by CVD plasma or PVD technique, after what MS and LS films were deposited on this layer and heat-treated at 500 and 110◦ C, respectively. XPS analysis reveals that the concentration of diffused Fe ions into TiO2 films sharply decreases in the presence of a SiOx barrier layer (Fig. 8). Figure 8 shows that the combination of a low temperature solgel procedure (LS films) with the presence of a SiOx barrier layer yields best results for preventing iron diffusion. The results also suggest that a PVD method

Figure 8. Atomic percentage of diffused Fe ions in MS (solid line) and LS (dashed line) TiO2 films on stainless steel in the absence or presence of a SiOx barrier layer. Data were deduced from XPS measurements performed at 90◦ take-off angle (
, ), and 30◦ takeoff angle (, ). The MS and LS films were heat-treated at 500 and 110◦ C, respectively.

might promote better barrier performances. This feature is out of the topics of this work and has not been studied. When considering the XPS Ti2 p peak, it appears that the atomic percentages of Ti3+ species in TiO2 films exhibit a good correlation with the percentages of iron, i.e. they decrease in the presence of a barrier layer and, in the case of a CVD barrier, are weaker in LS films heat-treated at 110◦ C than in MS films heattreated at 500◦ C (Fig. 9). This correlation confirms that the titanium reduction is presumably related to the film contamination with Fe species. However, MS films deposited on SiOx -coated substrates and heat-treated at 500◦ C exhibited same features as films on bare substrates, i.e. a greater amount of Ti3+ was detected at the film surface (30◦ take-off angle) than in deeper layers (90◦ take-off angle). Figures 10 and 11 show the time dependence of water contact angles under UV illumination for MS and LS TiO2 films, respectively, in the presence or not of a SiOx (PVD or CVD) barrier layer. The MS and LS films were heat-treated at 500 and 110◦ C, respectively, and were also compared to similar films deposited on silicon. Figures 10(b), (c), 11(b) and (c) indicate that the water contact angle of MS and LS TiO2 films on a SiOx (PVD or CVD) barrier layer sharply decreases after some minutes of UV irradiation, similarly to what is observed on a silicon wafer (Figs. 10(d) and 11(d)). Let us note that, MS TiO2 films on a SiOx barrier layer exhibit an excellent photo-hydrophilicity though some amounts of Fe ions could be detected at their surface. This suggests that iron diffusion is not the only factor that influences superhydrophilic properties of TiO2

Photo-Induced Hydrophilicity of TiO2 Films Deposited on Stainless Steel

135

films on stainless steel. Further studies are needed to better understand such a feature.

4.

Figure 9. Atomic percentage of Ti3+ species in MS (solid line) and LS (dashed line) TiO2 films on stainless steel in the absence or presence of a SiOx barrier layer. Data were deduced from XPS measurements performed at 90◦ take-off angle (
, ), and 30◦ takeoff angle (, ). The MS and LS films were heat-treated at 500 and 110◦ C, respectively.

Figure 10. Variations of the water contact angles upon UV irradiation for MS TiO2 films deposited on stainless steel and heat-treated at 500◦ C (a) without SiOx barrier layer, (b) with a SiOx CVD barrier layer (c) with a SiOx PVD barrier layer, and (d) for a MS TiO2 film deposited on silicon wafer and heat-treated at 500◦ C.

Conclusions

Crystalline titanium oxide films with excellent hydrophilic properties can be prepared on silicon wafer using two different sol-gel routes: a standard procedure, which requires a high crystallization temperature post-treatment, and a low temperature procedure based on the deposition of a liquid suspension of TiO2 crystallites. The results suggest that hydrophilic properties primarly require a good film crystallization degree. Crystallite size and surface oxidation degree are other factors which seem to influence the photo-induced hydrophilicity. XPS and XRD results indicate that films deposited on stainless steel exhibit crystallization features similar to those of films deposited on silicon wafer, and only differ by their oxidation degree, which is influenced by the diffusion of Fe ions during a high temperature heat-treatment and by the migration of iron species during the deposition of the acidic solution. The presence at the film surface of iron and Ti3+ species promotes a very poor photohydrophilicity on stainless steel. On the other hand, the deposition of a SiOx barrier layer at the film/substrate interface prevents the thermal diffusion or migration in liquid phase of Fe ions, resulting in excellent photo-induced hydrophilic properties of TiO2 films on stainless steel. The combination of a low temperature sol-gel route with the presence of a SiOx barrier layer yields best results in preventing iron contamination and titanium reduction. However, best photohydrophilic properties are obtained for TiO2 films deposited on a SiOx barrier layer and heat-treated at high temperature. This study yields promising results toward the fabrication of self-cleaning stainless steel. References

Figure 11. Variations of the water contact angles upon UV irradiation for LS TiO2 films deposited on stainless steel and heat-treated at 110◦ C (a) without SiOx barrier layer, (b) with a SiOx CVD barrier layer (c) with a SiOx PVD barrier layer, and (d) for a LS TiO2 film deposited on silicon wafer and heat-treated at 110◦ C.

1. A. Fujishima and K. Honda, Nature 238, 37 (1972). 2. R. Wang, N. Sakai, and A. Fujishima, J. Phys. Chem. B 103, 2188 (1999). 3. N. Sakai, A. Fujishima, and T. Wanatabe, J. Phys. Chem B 105, 3023 (2001). 4. T. Wanatabe, A. Nakajima, and R. Wang, Thin Solid Films 351, 260 (1999). 5. X.P. Wang, Y. Yu, and X.F. Hu, Thin Solid Films 371, 148 (2000). 6. J. Yu, X. Zhao, and Q. Zhao, Mater. Chem. Phys. 69, 25 (2001). 7. R. Sun, A. Nakajima, and A. Fujishima, J. Phys. Chem B 105, 1984 (2001).

136

Permpoon et al.

8. J.C. Yu, J. Yu, H.Y. Tang, and L. Zhang, J. Mater. Chem. 12, 81 (2002). 9. Y. Ohko, S. Saitoh, and T. Tatsuma, J. Electrochem. Soc. 148, B24 (2001). 10. A. Fernandez, G. Lassaletta, and V.M. Jimenez, Apply. Catal. B: Environ. 7, 49 (1995). 11. J.M. Hermann, Catalysis Today 54, 131 (1999). 12. H.C. Chan, J.F. Porter, and J.P. Barford, J. Mater. Res. 17, 1758 (2002). 13. Y. Zhu, L. Zhang, and L. Wang, J. Mater. Chem. 11, 1864 (2001).

14. J. Yu, W. Ho, and J. Lin, Environ. Sci. Tech. 37, 2296 (2003). 15. T. Wanatabe, S. Fukayama, and M. Miyauchi, J. Sol-Gel Sci. Tech. 19, 71 (2000). 16. A. Hattori, K. Shimoda, and H. Tada, Langmuir 15, 5422 (1999). 17. M. Langlet, M. Burgos, and C. Coutier, J. Sol-Gel Sci. Tech. 22, 139 (2001). 18. M. Langlet, A. Kim, and M. Audier, J. Mater. Sci. 38, 3945 (2003). 19. R.J. Gonzales, R. Zallen, and H. Berger, Phys. Rev. B. 55, 7014 (1997).