Introduction

Research efforts in the last decade have shown titanium dioxide (TiO2) to be the best candidate for self-cleaning applications. TiO2 can simultaneously decompose organic contamination and become superhydrophilic when irradiated with UV light. A superhydrophilic TiO2 film favors the fast and complete spreading of water droplets (e.g., rainwater) over the surface, which can wash off contaminant and dust.1,2

Numerous methods have been deployed to produce TiO2 thin films, including chemical-vapor deposition,3 sol–gel deposition,4 electron beam evaporation,5 pulsed laser deposition,6 molecular beam epitaxy,7 biomimetic approaches,8 and magnetron sputtering. Magnetron sputtering is an important method for a wide range of industrial applications.9,10 Compared with other methods, magnetron sputtering of TiO2 thin films reduces the use of elevated temperature in the deposition process and produces homogenous films on a larger scale.11,12

In the literature, we can find different approaches that yield high-performance anatase TiO2, yet many are unsuitable for industrial applications because they require a high-temperature process or produce thick films. Substrate heating and annealing are necessary to form anatase or rutile crystalline films,13,14,15,16,17 and the film thickness exceeds 100 nm.18 For a long time, the deposited films were found to be generally amorphous when the substrate was not heated during the deposition.19,20,21,22,23 Amorphous TiO2 films are not in practical use, and their characteristics have rarely been studied.24,25,26

In this study, we used DC magnetron sputtering to deposit TiO2 films with thicknesses of 20–60 nm on unheated substrates; the total pressure was controlled to influence the film structure. Amorphous films were produced under different conditions. The aim of this work was to study the structure, the photoinduced surface wettability, and the photocatalytic activity of amorphous TiO2 thin films with different thicknesses deposited at different total pressures.

Experimental details

Sample preparation

The TiO2 films were deposited on microscopic glass slides using DC magnetron sputtering without external heating. The depositions were performed in a chamber equipped with two confocal guns as sputtering sources. The distance between the targets and the substrate was 13 cm. Samples were rotated during the deposition at a speed of 25 rpm. The base pressure of the chamber was 10−8 Torr. A titanium target (99.99%) was sputtered with a plasma (15 W/cm2) of argon (99.99%, flow 12 sccm), and reactive oxygen (99.99%) was introduced between the plasma and the substrate. For this work, we prepared samples at nine different pressures (4–8, 10, 15, 20 and 25 mbar) and three deposition times, corresponding to equivalent film thicknesses of 20, 40, and 60 nm.

A precise determination of the oxygen flow is required in order to obtain a titanium oxide with the appropriate stoichiometry, i.e., TiO2. Therefore, we measured the evolution of the cathode voltage with the oxygen flow rate at each of the pressures investigated. To ensure that TiO2 was the only species being deposited, all subsequent experiments were performed using an oxygen flow rate of 6 sccm.

The sample-to-sample variation in the TiO2 film thickness was monitored in situ using a quartz crystal monitoring system (QCM), and the absolute value of the TiO2 film thickness was determined using optical white light interferometry at the University of Applied Sciences Western Switzerland. The precision of the thickness measurements was <1 nm. The deposition rate variation as a function of the total pressure is shown in Fig. 1.

Fig. 1
figure 1

Growth rate at different working pressures

Full size image

Contact angle measurements

Photoinduced modification of the surface hydrophilicity was evaluated by measuring the contact angle of a water drop (5 µl) with a goniometer (ramé-hart Model 190). Samples were exposed to UVA light in a BS02 radiation chamber (Opsytec Dr. Gröbel GmbH). Four different irradiances were used during these tests (2, 4, 6, and 8 mW/cm2). Contact angle measurements were performed on each sample before and after UV light exposure, and after storage in the dark for 24 h. Each measurement was performed five times in order to take into account possible sample nonhomogeneity. The precision of the contact angle measurements was <5°.

Photocatalytic tests

Photocatalytic tests were performed by observing the photodegradation of methylene blue (MB) under UV irradiation at 6 mW/cm2, with stirring. Samples were exposed to UVA light (315–400 nm) in a controlled UV chamber (BS02—Opsytec Dr. Groebel GmbH). Eight ultraviolet fluorescent tubes of 15 Watts each were installed. These tubes irradiated light in the UVA range of the solar spectrum. Figure 2 shows the spectrum range for these UVA tubes. The integrated radiometric sensors accurately measured the intensity of the radiation over UVA spectral range (315–400 nm) with RM-12 radiometer (RM-12—Opsytec Dr. Groebel).

Fig. 2
figure 2

Spectrum for UVA tubes

Full size image

Microscopic glass samples were immersed in beakers containing 50 ml of methylene blue solution (1 mg/l). Samples (1 ml) were periodically taken from the mixture and analyzed using UV–visible absorption spectroscopy (UV 1800, Shimadzu Corp.) to assess the MB degradation efficiency. Peaks observed between 600 and 700 nm and were assigned as the absorption of the system.27 According to the Beer–Lambert Law, MB’s concentration is directly proportional to its absorbance, which makes it possible to determine the MB degradation efficiency.

X-ray photoelectron spectroscopic analysis

After deposition, the TiO2 thin films were placed in an ultra-high vacuum chamber and kept at 10−6 mTorr for several hours before X-ray photoelectron spectroscopic (XPS) analysis. XPS measurements were performed with a Mg Kα (hν = 1254 eV) X-ray source, using an EAC 2000-SPHERA photoelectron spectrometer with a hemispherical photoelectron analyzer. Broadband excitation was used for full spectra (step, 1.0 eV; pass energy, 44 eV; energy resolution, 1.5 eV) and for chemical analysis (step, 0.1 eV; pass energy, 22 eV; energy resolution, 1.0 eV). Spectral analysis included a Shirley background subtraction and peak separation using mixed Gaussian–Lorentzian functions.

Characterization

The crystalline structure of the TiO2 films was analyzed using grazing incidence X-ray diffraction (XRD). A Bruker D8 DiscoverTM instrument (copper tube, incident angle of 1°, step scan mode with a 0.04° step size) was used. The surface morphological features of the samples were observed using scanning electron microscopy (SEM) with a field emission SEM (FE-SEM, JEOL JSM-7600F).

Results and discussion

TiO2 film deposition

The evolution of the cathode voltage with the oxygen flow rate at each of the pressures investigated is displayed in Fig. 3. The curves reveal the expected features. First, the discharge voltage increases when the oxygen flow rate increases from zero to a threshold. Between these two values, the films deposited on the substrate are mainly a mix of the metal and its oxides with low oxidation states. The threshold occurs around 3 sccm when the pressure is above 8 mTorr and around 4 sccm for pressures below 10 mTorr. When the threshold is reached, the cathode voltage shows an abrupt increase due to the full coverage of the target by the oxide. Above the threshold, the target voltage decreases to a stable value. It is well known that the required titanium stoichiometry (TiO2) is obtained after the transition region.

Fig. 3
figure 3

Cathode voltage vs oxygen flow rate: (a) for pressures between 4 and 8 mTorr, (b) for pressures between 10 and 25 mTorr

Full size image

Figure 1 shows the curve of the growth rate of TiO2 films deposited at different working pressures. When the working pressure increased from 4 to 8 mTorr, the growth rate of the film increased gradually from 1.4 to 1.7 nm/min. However, when the working pressure increased from 10 to 25 mTorr, the growth rate decreased from 0.8 to 0.6 nm/min.

Structural characterization

X-ray diffraction (XRD) analyses were performed to investigate the crystallographic structure of the TiO2 films. The XRD pattern (Fig. 4) shows a broad peak at small angle. The peaks characteristic of the crystalline phases are absent, indicating the presence of an amorphous material.

Fig. 4
figure 4

XRD pattern of amorphous TiO2: (a) deposited at 10 mTorr, (b) deposited at 25 mTorr

Full size image

XPS analyses on all TiO2 samples were identical. Therefore, a comparison between the lowest (4 mTorr) and the highest pressure (25 mTorr) is shown (Fig. 5). In addition to the expected Ti 2p and O1s peaks of TiO2 (Fig. 5a), in some cases, some carbon (C1s peak) appeared at the surface of the TiO2 film. This carbon may be related to contamination of the film during growth. Decomposition of the Ti2p XPS spectra (Fig. 5b) reveals the presence of two components. The Ti 2p 1/2 and Ti 2p 3/2 spin-orbital splitting photoelectrons are located at binding energies of 465.2 and 459.4 eV, respectively. The separation between the Ti 2p 1/2 and Ti 2p 3/2 peaks is 5.92 eV. These peak positions and their separation closely match the reported values for Ti4+ in bulk TiO2.28,29,30 Thus, these peaks can be assigned to TiO2. The XPS spectrum of O1s (Fig. 5c) shows a two-band structure. The dominant peak at 530.6 eV is characteristic of metallic oxides and is in agreement with the O1s electron binding energy for TiO2.31 There is also a shoulder located toward the side of higher binding energies. This secondary peak was assigned to OH species. The binding energy difference of 71.28 eV between the observed peak positions of Ti 2p 3/2 and O1s (oxide) is also in excellent agreement with reported literature values of 72.9–71.2 eV.32,33,34,35,36

Fig. 5
figure 5

XPS analysis: (a) survey (vertically shifted for clarity), (b, c) high-resolution Ti2p and O1s spectra, respectively, and (d) curve resolved O1s signal

Full size image

SEM images recorded for films deposited at different working pressures are shown in Fig. 6. It is difficult to accurately determine the mean diameter of the TiO2 particles using image analysis software because they are in close contact with one another. As an initial evaluation, films deposited at 10 mTorr and above have particles with a mean diameter smaller than 10 nm. Those deposited at pressures below 10 mTorr have particles with a mean diameter larger than 12 nm. We can, therefore, conclude that the TiO2 particle size decreases as working pressure increases.

Fig. 6
figure 6

SEM images of amorphous TiO2 deposited at: (a) 4 mTorr, (b) 25 mTorr, and (c, d) 10 mTorr with two different magnifications

Full size image

Photoinduced superhydrophilicity

Excellent surface wettability is crucial for the self-cleaning function of the film. The surface hydrophilicity of the TiO2 films was quantified using water contact angle measurements performed in the absence and presence of UV light at 8 mW/cm2. Figure 7a shows the results obtained with films deposited at pressures below 10 mTorr. Although they show a decrease in their contact angles upon UV light illumination, none of them is superhydrophobic, that is to say, none of them is characterized by a contact angle close to zero. Only samples deposited at 10 mTorr and above (Fig. 7b) become superhydrophilic upon UV light illumination. Water droplets spread out on these films, resulting in a contact angle of about 0° (below 5°). Figure 8 shows the water droplet shape on the TiO2 films before and after UV light illumination. Prior to illumination, the TiO2 films yielded water contact angles between 45° and 60°. This indicates that these TiO2 films have hydrophilic surfaces. Upon illumination with UV light, the water droplet spread out on the film, resulting in a contact angle of about 0° for all the films prepared with total pressure ≥10 mTorr. The observed phenomenon can be understood in terms of the TiO2 nanoparticle size. As explained in a previous report,37 the density of surface hydroxyl groups increases dramatically when the size of the nanoparticle is reduced. This can lead to a substantial increase in the reorganization energy, hence the hydrophilicity of the nanoparticles. This observation is supported by our XPS spectra analysis. For the smaller particle size coatings deposited at 25 mTorr, XPS data reveal an increase in the secondary peak of the O1s spectrum attributed to OH species (Fig. 5c), indicating a higher concentration of surface hydroxyl groups.

Fig. 7
figure 7

Hydrophilicity of TiO2 films upon UV irradiation at: (a, b) 8 mW/cm2, (c) 6 mW/cm2 and (d) 4 mW/cm2

Full size image
Fig. 8
figure 8

Photographs of water droplet shape on TiO2 surface: (a) before and (b) after UV irradiation

Full size image

Samples displaying superhydrophilicity upon UV light irradiation at 8 mW/cm2 were tested with lower UV irradiances (6, 4, and 2 mW/cm2). UV irradiation at 6 mW/cm2 rendered all surfaces superhydrophilic, with water contact angles smaller than 5° (Fig. 7c). Films with a thickness of 20 nm needed 3 h of UV irradiation to become superhydrophilic, while films with a thickness of 40 or 60 nm needed only 45 min, even with UV irradiation at 4 mW/cm2. However, the surfaces of these films show superhydrophilic behavior after 3 h of UV irradiation at 2 mW/cm2. As expected, increasing the film thickness increases the amount of UV light absorbed; therefore, thick films generate photoelectrons. These electrons reduce surface Ti4+ sites to the Ti3+ state. Then, oxygen vacancies are generated through the oxidation of bridging O2− species to oxygen via the photogenerated holes. Subsequently, dissociated water adsorption on the vacancy sites creates hydrophilic OH groups on the surface. This mechanism of the photoinduced hydrophilic conversion has been extensively studied.38,39,40,41,42,43,44,45

After the UV irradiated films were placed in the dark for 24 h, the wettability of their surfaces was measured again. During storage in the dark, the superhydrophilic films reverted to their original states, except those deposited at 10 mTorr and thicker than 20 nm, which showed a low contact angle near 20°. The increase in water contact angle of the stored TiO2 film may be due to the adsorption of contaminants during the storage or the release of water molecules. Reversible surface wettability conversion reactions were achieved by alternate UV irradiation and storage in the dark on both these films.

Photocatalytic activity

Photocatalytic activity was evaluated in terms of the rate at which an aqueous solution of MB underwent photodegradation upon UV irradiation. The same experiment was repeated without UV irradiation to estimate the adsorption of MB on the TiO2 surfaces, which is estimated at 6% after 4 h (Fig. 9a). For accurate comparison, the adsorption percentage has been subtracted from the decomposition ratios.

Fig. 9
figure 9

Absorption spectral changes due to MB adsorption on TiO2 films: (a) in the dark, and (b–d) due to MB degradation upon UV irradiation of films of different thicknesses

Full size image

The height of the MB absorbance peak, which indicates the concentration of MB in solution, decreases with increasing illumination duration as a result of photochemical reaction. It can be seen from Fig. 10 that the working pressure strongly influences the MB degradation rate of amorphous TiO2 films. The MB degradation rate of TiO2 films increases from 16% to 35% when the working pressure rises from 10 to 25 mTorr (Fig. 11a). This result can be explained as follows. The photocatalytic activity of the TiO2 films is considered to increase as the TiO2 particle size decreases. The SEM images confirm that the TiO2 particle size decreased with increasing working pressure; thus, the degradation rate for the MB solution increased.

Fig. 10
figure 10

Absorption spectral changes due to MB photodegradation for different pressures

Full size image
Fig. 11
figure 11

MB Photodecomposition ratios for: (a) 20 nm films deposited at different pressures and (b) films of different thickness deposited at 10 mTorr

Full size image

The absorption spectra (Fig. 9) are consistent with the results obtained in similar studies on TiO2 thin films: the MB decomposition ratio strongly depends on film thickness. This ratio increases with each increment of thickness. Fifty percent of the MB was degraded with 60-nm-thick films deposited at 10 mTorr (Fig. 11b). The higher photocatalytic activity originates from the greater amount of photogenerated electron–hole pairs, which are transferred from the bulk to the surface of the TiO2 films.

Conclusion

The self-cleaning properties of amorphous TiO2 have been studied as a function of deposition parameters. Amorphous TiO2 deposited at a minimum working pressure of 10 mTorr has a superhydrophilic surface induced by UV light irradiation irrespective of its crystalline nature. These results are consistent with those of other research studies.46 A reversible hydrophilicity to superhydrophilicity transition was observed and was nicely controlled by alternation of UV light illumination and dark storage.

This work shows that amorphous TiO2 films of high photocatalytic activity can be obtained using a direct current pulse magnetron sputtering system with no intentional heating. The working pressure is a key deposition parameter that influences the size of TiO2 nanoparticles. As the working pressure increases, the size of the resulting TiO2 particles decreases, improving the catalytic activity of amorphous TiO2. This surprising result requires further investigation since it has never been reported before. Samples prepared at 10 mTorr and thicker than 20 nm are the best because they absorb more light than thinner films. They require only 1 h to become superhydrophilic despite the UV irradiance. They not only preserve a small contact angle (20°) even after 24 h storage in the dark, but also decompose 50% of the MB in solution.