Tunable structure of TiO₂ deposited by DC magnetron sputtering to adsorb Cr (VI) and Fe (III) from water

Abstract

TiO₂ thin films with mixtures of the anatase, rutile, and brookite phases were deposited on glass substrates via magnetron sputtering. Based on XRD and Raman results, the TiO₂-0.47 and TiO₂-3.47 films principally contained the brookite phase, while the TiO₂-1.27 and TiO₂-2.13 films were primarily anatase. The capacities of the TiO₂ films to adsorb heavy metals were tested with Cr(VI) and Fe(III) solutions, and the maximum Cr(VI) and Fe(III) adsorption capacities were realized with the TiO₂-0.47 film (334.5 mg/g) and TiO₂-3.47 film (271.3 mg/g), respectively. SEM‒EDS results revealed the presence of Cr and Fe on the surfaces of the films, thus corroborating the ability of the TiO₂ films to adsorb and remove heavy metals. They are strong candidates for use in wastewater treatment plants.

1. Introduction

TiO₂ is the most promising photocatalytic semiconductor due to its excellent physicochemical properties, high chemical stability, nontoxicity, long-term photoactivity, high stability under UV irradiation, etc. [[1], [2], [3], [4]]. As a photocatalyst, TiO₂ has mainly been used in dye degradation, and it was reported to degrade up to 92–95% of the methylene blue (MB) and 96% of rhodamine B (RhB) in solution [5,6]. High photocatalytic efficiency resulted from irradiation of TiO₂, during which it generated electron-hole pairs; the electrons had high reduction potentials and the holes at the semiconductor surface had high oxidation potentials, which enabled the breakdown of organic molecules by ·OH radicals [4]. In addition, TiO₂ has been used in other photocatalytic reactions, such as hydrogen generation and CO₂ reduction, which have gradually become more important due to the high efficiency of TiO₂ [7,8]. Based on these photocatalytic applications in which TiO₂ can be used, it is an excellent option for water remediation via adsorption of heavy metals. In an adsorption process, the metallic species present in an aqueous solution are adsorbed on the surface of a solid, and adsorption occurs through electrostatic charges [9]. Since heavy metals have high affinities for hydroxyl species (OH⁻), TiO₂ is used for adsorption of heavy metals; as noted above, TiO₂ generates abundant OH⁻ species, which increases the probability of adsorbing the metals and the efficiencies of photocatalytic reactions used for water remediation [9,10]. In this sense, TiO₂ adsorbs different heavy metals, such as Pb²⁺, Cu²⁺, Cd²⁺, Ni²⁺, and Cr⁶⁺ [[11], [12], [13], [14], [15], [16], [17], [18], [19], [20]]. However, most of the reported studies have used TiO₂ as a powder, which is a problem when the photocatalyst is to be recovered; alternatively, other semiconductors or organic compounds can be added with TiO₂ to increase the adsorption efficiency, as shown in Table S1.

In contrast, TiO₂ is a very attractive semiconductor due to the three crystalline phases anatase, rutile, and brookite, which influence the efficiency of TiO₂ in each photocatalytic process in different ways. Among the three crystalline phases of TiO₂, anatase is the most common phase for photocatalytic applications, followed by the rutile phase; the anatase phase exhibits lower absorption of solar light than the rutile phase, because the band gap of anatase is greater (3.2 eV) than that of the rutile phase (3.0 eV) [21,22]. However, synergy between these two phases results in higher photocatalytic efficiency for TiO₂, and the most efficient photocatalyst, Degussa TiO₂ powder P25, has a rutile–anatase ratio of 1:3; however, the efficiency of these phases requires efficient charge separation when the TiO₂ is irradiated. This efficiency has been associated with the migration of photoexcited electrons from the rutile phase to the conduction band of the anatase phase, while the photoexcited holes are left in the rutile phase [[22], [23], [24]]. The last crystalline phase, brookite, is the most difficult to synthesize due to its orthorhombic structure, which makes it more unstable than the other two phases [25,26]. The band gap reported for the brookite phase ranges from 3.1 to 3.4 eV, even though there are reports in which the band gap of this phase is 3.6 eV; therefore, the differences in the band gaps could be useful in photocatalytic processes [[27], [28], [29]].

Preparing a single TiO₂ phase is not easy, as the anatase phase changes to the rutile phase at approximately 600 °C, while brookite is converted into the rutile phase at approximately 800 °C [25,30]. However, as the three phases are used in photocatalytic processes, the percentages of the phases could determine the efficiency of each process. In this sense, an efficient way to use TiO₂ for heavy metal absorption is to support it on a substrate, which allows recovery and reuse. Among the methods used for depositing materials, physical techniques allow modification of the structural properties of the materials in a controlled way by changing the initial deposition conditions. Deposition of thin films via sputtering is a great option for modifying the structural properties of semiconductors, and it is possible to explore the depositional conditions to change the final properties of the film. With the sputtering technique, changing the amount of carrier gas inside the chamber, the gases used in the deposit, the power density, and the working pressure for deposition is feasible [[31], [32], [33], [34]]. Among these parameters, the working pressure is one of the best options for modifying the structural, optical, and morphological properties of thin films, as the ionization levels of the species inside the chamber increase with increasing working pressures [35,36]. The working pressure affects the path and energy through which the material species collide with the substrate inside the chamber, giving rise to preferential planes in the crystalline structure [35,[37], [38], [39]]. However, the presence of more than two TiO₂ phases is very common even with changes in the deposition conditions (see Table S2) [[32], [33], [34],[40], [41], [42], [43], [44], [45], [46], [47], [48], [49]].

For this reason, this work explores the influence of the working pressure on the deposition of TiO₂ thin films by the sputtering technique and the efficiency for adsorbing heavy metals such as Cr(VI) and Fe(III). The structural, optical, and morphological properties were studied to understand the effects of the working pressure during deposition.

2. Experimental

2.1. Deposition of TiO₂ thin films by magnetron sputtering

TiO₂ films were deposited using a Ti target (99.995% pure, 2.0000 diameter X 0.12500 thick, Kurt J. Lesker) over glass substrates. With a turbomolecular pump, the sputtering chamber was evacuated to a background pressure of $1.14x{10}^{-3}$ Pa, and then Ar gas (UHP) was introduced into the chamber with a flow rate of 15 sccm. The films were deposited using a current DC power (60 W) for 60 min. Each film has been deposited at different working pressures, such as 0.47, 1.27, 2.13, and 3.47 Pa, to obtain the different TiO₂ phases. The working pressures were chosen according to the equipment conditions. After deposition, the films were annealed at 400 °C for 60 min to form the oxide.

2.2. Physicochemical characterization of TiO₂ thin films

Structural, optical, and photocatalytic properties films were characterized by X -ray diffraction (PANalytical, radiation of 1.54 Å - Cu Ka, grazing angle), UV–Vis–NIR spectrophotometer (Cary 5000, range 200–800 nm), Surface morphology was studied by atomic force microscope (ASYLUM RESEARCH, model MFP3D-SA, in Tapping mode). X-ray photoelectron spectroscopy was used to determine chemical states and elemental composition (Thermo Scientific, Escalab 250Xi, equipped with an Al Ka monochromatic X-ray source, hv = 1486.7 eV).

2.3. Photocatalytic adsorption test

The photocatalytic test was performed using the TiO₂ films as photocatalytic reactors. The adsorption of metal ions was obtained by doing a calibration using five different concentrations of Cr and Fe. The calibration plots, the absorbance of each spectrum and the equation that describes the relation between the absorbance and the metal ion concentration are shown in the Supporting Information (Fig. S1). For the photocatalytic tests, in each reactor, 30 ml of a 2-ppm solution of Cr (VI) and Fe (III) are placed and are constantly stirring at room temperature. After 30 min, the sample reference was taken, and the reactor began to be irradiated with a commercial lamp of 90 Watts (Tecnolite/HEL-BLB). The tests were performed for 360 min and were monitored using a UV–Vis–NIR spectrophotometer. The results of these tests are presented in the supplementary information (Fig. S1).

3. Results and discussion

Fig. 1a shows the X-ray diffraction patterns of the TiO₂ films. The different diffraction patterns for each film indicate that the working pressure affected the crystalline structures. This was true for films deposited at the highest (3.47 Pa) and lowest (0.47 Pa) working pressures, which showed peaks at different 2θ positions. In contrast, the films obtained at pressures of 1.27 and 2.13 Pa had amorphous structures, which are common in sputtering deposits over glass substrates [32,42,49,50]. For this reason, a Rietveld refinement was performed to determine the lattice parameters and the phases resulting from the deposition conditions (Table S3). After the Rietveld refinement, the TiO₂-0.47 film showed rutile (tetragonal) and brookite (orthorhombic) phases. However, the brookite phase was no longer formed as the working pressure was increased (from 0.47 Pa to 2.13 Pa), and the TiO₂-1.27 and TiO₂-2.13 films had mixtures of the anatase and rutile tetragonal phases. In contrast, the three TiO₂ phases, anatase, rutile, and brookite, were present only in the TiO₂-3.47 film deposited at a high working pressure (3.47 Pa). The presence of the different TiO₂ phases and changes in the lattice parameters of each film are shown in the Supporting Information (Tables S3 and S4).

Fig. 1.

a) XRD patterns and b) Raman spectra of the TiO₂ films deposited at different working pressures. The black, blue, and red dots in the XRD patterns correspond to the anatase, rutile, and brookite phases, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

To complement the XRD measurements, Raman spectroscopy was used to determine the TiO₂ structure, and the Raman spectra of the TiO₂ films are shown in Fig. 1b. All the films exhibited the vibration modes of the anatase (E_g, A_1g, B_1g, E_u, and E_3g), rutile (A_1g, E_u, and E_3u), and brookite (A_1g and B_2g) phases. For the anatase phase, the principal vibrational modes (E_g, A_1g, B_1g) were those involving oscillations of the O atoms in O–Ti–O, O₂ pure vibrations, and Ti atom vibrations [51,52]. The A_1g and E_3u modes for the rutile phase involved atomic displacements of the oxygen sublattice and the IR-active polar modes, respectively [52,53]. Likewise, the A_1g and B_2g modes of the brookite phase have been used to identify this TiO₂ phase [54,55].

The transmittance percentages of the TiO₂ films are presented in Fig. 2a. As the pressure increased (from 1.27 Pa to 3.47 Pa), the transmittances increased to between 70 and 90%, except for the TiO₂-0.47 film (deposited at lower pressure 0.47 Pa), for which the transmittance was less than 5%. The optical band gap was calculated from the Tauc plot by extrapolating the linear portion, where α is the absorption coefficient, hν is the photon energy and E_g is the optical band gap energy (Fig. S2) [7]. As the pressure increased (from 0.47 Pa to 3.47 Pa), the band gap values were 3.74, 3.54, 3.32, and 3.6 eV for the TiO₂-0.47, TiO₂-1.27, TiO₂-2.13, and TiO₂-3.47 films, respectively. These values were consistent with the band gaps of the different TiO₂ phases [21,56]. The transmittance spectra showed that the adsorption edges of the films were shifted to higher wavelengths, and the bandgaps increased in the order TiO₂-2.13, TiO₂-3.47, and TiO₂-1.27. These shifts raised the Fermi levels above the conduction band and increased the band gaps [57]. This behavior is known as the Burstein–Moss effect, and it can be caused by dopants or structural defects that increase the concentrations of carriers and modify the absorption edge [57]. Optical properties such as ellipsometry angles, refractive index and PL spectra are shown in Figs. S3 and S4 (Supporting Information) [58,59].

Fig. 2.

a) Transmittance spectra and b) AFM images of the TiO₂ films deposited at different working pressures.

The surfaces of the films were analyzed with atomic force microscopy (AFM) over a 1.0 × 1.0 μm area. Fig. 2b shows that all films had the same grain morphologies because the grain sizes varied according to the working pressure used for deposition. The RMS, average height, and specific surface area were measured with AFM, and the values were 3.01, 3.12, 1.46, and 1.97 nm for the roughness; 7.68, 11.10, 4.93, and 5.93 nm for the average height; and approximately 1 nm² for the surface areas of TiO₂-0.47, TiO₂-1.27, TiO₂-2.13, and TiO₂-3.47, respectively.

The importance of surface defects in the photocatalytic performance of TiO₂ has already been identified, especially for CO₂ photocatalytic reduction [51,60]. In particular, oxygen vacancies serve as adsorption and active sites in heterogeneous photocatalysis and modify the electronic structure, charge transport, and surface properties [61,62]. For this reason, XPS was used to determine the oxidation states of titanium and the concentration of oxygen vacancies. High-resolution Ti 2p and O 1s spectra were fitted with a Shirley background. A comparison among the four samples is shown in Fig. 3. Different authors have claimed that the titanium in TiO₂ crystals exhibited both the 3⁺ and 4⁺ oxidation states [60]. However, in the Ti 2p spectra, only a doublet peak at 458.4 eV with a spin-orbital splitting of 5.7 eV was identified. This peak was attributed to Ti⁴⁺ bonded to O²⁻ [63]. The peak located at a higher binding energy (471.8 eV) was identified as a satellite peak. On the other hand, the O 1s spectrum contained five peaks. The peaks at 529.8, 530.8, 531.5, 532.4 and 533.7 eV were attributed to O^2–Ti⁴⁺ species in TiO₂, oxygen vacancies, chemisorbed species (hydroxyl groups), -O-C=O bonds and oxygen bonded to silicon, respectively [[63], [64], [65], [66]]. A small peak at 103.2 eV (Si–O bonds) was seen in the survey spectra of the TiO₂-0.47 and TiO₂-3.47 samples, which confirmed the peaks found in the oxygen spectrum [67].

Fig. 3.

High-resolution spectra for a) Ti 2p and b) O 1s core levels for TiO₂ films deposited at different working pressures.

The Ti 2p and O 1s peak intensities and their appropriate atomic sensitivity factors (ASFs) (obtained from the Scofield table) were used to determine the surface atomic compositions of the TiO₂ samples. The atomic percentages indicated by XPS for the TiO₂-0.47 and TiO₂-3.47 samples are shown in Table 1. A quantitative analysis indicated that the oxygen vacancy concentrations were highest in the TiO₂-1.27 and TiO₂-2.13 samples, which could be associated with better photocatalytic performance.

Table 1.

Atomic percentage of titanium and oxygen species in the TiO₂ thin films calculated from XPS spectra.

Sample	Ti 2p (Ti–O)	O 1s (O–Ti)	O 1s (Vo)	O 1s (OH−)	O 1s (O–C)	O 1s (O–Si)
M1	22.5	48.5	4.9	11.9	9.7	2.5
M2	11.5	22.5	14.0	42.9	9.2	0.0
M3	11.3	21.5	13.5	41.4	12.3	0.0
M4	15.2	32.5	7.9	14.9	20.3	9.3

Photocatalytic adsorption of the different samples are shown in the supplementary material(Figure S5, S6 and S7), and based on these results, we focused on the films that showed higher capacities for the adsorption of Cr (VI) and Fe (III) ions. Hence, the amounts of Cr⁶⁺ and Fe³⁺ ions adsorbed on the surfaces were determined for the TiO₂-0.47 and TiO₂-2.13 films, respectively. The relationship between the removal efficiency for a given interaction time and the adsorption capacity (q_t) of TiO₂ is shown in Fig. 4. The iron plot exhibited a constant linear trend over time, which provided information on the homogeneity of the films and the reactive sites at the surface. On the other hand, the chromium plot shows three levels, suggesting that physisorption had occurred; the first layer (ions physically bonded) was removed, followed by chemisorption at the surface. The different adsorption capacities for the two ions were related to the chemical nature of the solution. The sizes of the ions could have caused rearrangement at the surface and led to the adsorption of more Cr⁶⁺ (with the smaller radius: 58 p.m.) than Fe³⁺ (63 p.m.). Moreover, Cr⁶⁺ is more electropositive than Fe³⁺, which means that it is strongly attracted to negative species such as oxygen. The percentages of Cr(VI) and Fe(III) removal were 70 and 63%, respectively, after 360 min. These values were consistent with those reported in the literature for these ions [[68], [69], [70]].

Fig. 4.

Adsorption kinetics of TiO₂ samples adsorbing Cr and Fe in aqueous solution.

The absorption capacities for the films with each metal were different because heavy metals have high affinities for hydroxyl (OH⁻) ions [9,10]. Therefore, the presence of hydroxyls on a film surface increases the likelihood of adsorbing metals. Based on the XPS (Table 1) results, the TiO₂-2.13 film had a higher concentration of OH⁻ species on the surface, which increased the amount of Fe³⁺ adsorbed. In contrast, the TiO₂-0.47 film, with a major brookite component, did not have a high concentration of OH⁻ ions. However, as mentioned above, Cr⁶⁺ is an electropositive metal with a strong attraction for negative species such as oxygen, and the XPS results indicated that the TiO₂-0.47 film had a lower oxygen deficiency, which favored Cr⁶⁺ adsorption.

To confirm the presence of these ions on the surfaces of the TiO₂ thin films, TiO₂-0.47 and TiO₂-2.13 were analyzed via EDS and UV–Vis spectroscopy. Fig. 5 shows the SEM‒EDS maps of the 0.47-TiO₂-Cr (top) and 2.13-TiO₂-Fe (bottom) samples, with the different distribution of elements O, Ti and Fe/Cr obtained by EDS measurements (around 1–2 μm in depth). The colored elemental maps showed that O (red), Ti (blue), and Cr or Fe (green) were homogeneously distributed throughout the entire thin film. The atomic composition of the TiO₂ samples with Cr was O 60.4%, Ti 36.3%, and Cr 3.3%, while that with Fe was O 65.0%, Ti 33.8%, and Fe 1.2%. These results showed that the concentration of Cr in the TiO₂ was greater than that of Fe after 350 min of UV exposure, consistent with the adsorption capacities. Moreover, the UV–Vis spectra revealed decreased transmittance after adsorption of the metals, suggesting the formation of a thin layer of species at the surface. A decrease was also observed even for the TiO₂-0.47 sample, which exhibited approximately 2% transmittance at 300–500 nm, or approximately zero. These measurements indirectly confirmed the adsorption of species on the surfaces.

Fig. 5.

Scanning SEM-EDS mapping of TiO₂ thin films after adsorption of Cr (top) and Fe (bottom) and transmittance spectra for these samples.

4. Conclusions

TiO₂ thin films were deposited on glass substrates via magnetron sputtering to study the impact of the working pressure on the physicochemical properties. Combinations of the anatase, rutile and brookite phases were obtained by modifying the working pressures. XRD and Raman spectrometry revealed that the TiO₂-0.47 and TiO₂-3.47 samples were composed of brookite, rutile and anatase, while the TiO₂-1.27 and TiO₂-2.13 samples comprised anatase and rutile mixture. In the same way, the films deposited at higher pressures (1.27, 2.13, and 3.47 Pa) have the highest transmittance percentage (>70%), while the film deposit at 0.47 Pa has a transmittance percentage lower than 5%. Meanwhile, in AFM, results observed that the surface of the films was composed of little grains that agglomerated as the working pressure increased in each deposit (from 0.47 to 3.47 Pa). On the other hand, XPS results show that the principal contribution in the TiO₂ films is the Ti⁴⁺ bonded to O²⁺ species, as well as oxygen vacancies, chemisorbed species, and O–C=O bonds. According to the photocatalytic results, the TiO₂ samples preferably adsorbed Cr (VI) over Fe (III) and exhibited maximum adsorption capacities ranging from 53 to 334.5 mg/g for Cr (VI) and from 149.8 to 271.3 mg/g for Fe (III) ions in aqueous solutions, as shown with batch adsorption experiments. The higher adsorption capacity for Cr was presumably due to the small ionic radius, the electropositive nature of this element and the roughness of the TiO₂ thin film. The charge affinity and the metal electronegativity influenced the absorption capacity of each film, as Fe²⁺ adsorption depended on hydroxyl (OH⁻) species, and the presence of more oxygen increased the amounts of the electronegative metals adsorbed, as seen with Cr⁶⁺. After each reaction, the transmittance percentage and SEM-EDS measurements were done to confirm the presence of the metals over the film surface. SEM‒EDS maps of each film, revealed the presence of Cr and Fe at the surface, with contents of 3.3 and 1.2%, respectively. In contrast, the transmittance percentage of all films decreased, which confirmed the presence of the different metals over the film surface and the capacity of TiO2 to absorb different metals.

Data availability statement

The data will be available on request.

CRediT authorship contribution statement

F.A. Hernández-Rodríguez: Writing – original draft, Methodology, Investigation. R. Garza-Hernández: Writing – review & editing, Writing – original draft, Investigation, Formal analysis. M.R. Alfaro-Cruz: Writing – review & editing, Writing – original draft, Supervision, Investigation, Formal analysis, Conceptualization. Leticia M. Torres-Martínez: Writing – review & editing, Supervision, Resources, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: