Photocatalytic Oxidation of Pesticides with TiO₂-CeO₂ Thin Films Using Sunlight

Abstract

TiO₂ thin film coatings significantly improve catalyst separation in photocatalytic processes. They can be applied in heterogeneous photocatalysis under sunlight by mixing TiO₂ with other oxides, such as CeO₂, for the removal of pollutants in water. Here, TiO₂-CeO₂ thin films deposited on borosilicate slides were analyzed and applied in solar heterogeneous photocatalysis for the oxidation of pesticides. The films were synthesized by the sol-gel method with spin coating. The waste solutions from the synthesis were used to prepare TiO₂ and TiO₂-CeO₂ powders. These were analyzed by XRD and XPS to explain the behavior of the films. The thin films were characterized by UV-Vis spectroscopy with transmittance, UV-Vis spectroscopy with RDS, profilometry, AFM and SEM. The addition of CeO₂ to TiO₂ caused a decrease in the average crystal size and an increase in the strain index. The addition of a second layer made the TiO₂-CeO₂ thin films thinner. The CeO₂ created surface and electronic defects in the titania films, which enhanced their photocatalytic properties under sunlight in the mineralization of diuron and methyl parathion. The TiO₂-CeO₂-5.0% single-layer thin film samples were the most active in this study and will undoubtedly be applied in larger-scale reaction systems.

1. Introduction

Titanium oxide is the most widely used semiconductor in heterogeneous photocatalysis for the treatment of emerging contaminants in water and air due to its low cost, abundance, chemical–mechanical stability, low toxicity, and excellent textural, structural, and semiconductor properties [1,2]. Doping TiO₂ with noble metals, transition metals, metal oxides, lanthanides, and non-metals modifies the morphology, structure, and optical and electronic properties of this semiconductor, increasing its efficiency in the photocatalytic degradation of aqueous contaminants using sunlight or visible light [3]. However, it is crucial to note that many of these studies are conducted on photocatalysts in the form of extremely fine powders, which pose significant risks in occupational settings due to the potential for inhalation and epidermal exposure. The toxicity of TiO₂ depends on the exposure time and particle size, making it imperative to exercise caution when handling these materials [4]. Fine TiO₂ powders exhibit high photocatalytic activity because they can be almost homogeneously mixed with water, as shown by Degussa TiO₂ (P25). However, this also represents a significant drawback. It is extremely challenging to separate this solid from the aqueous medium for recovery, which increases the energy costs if it is to be reused [5]. Furthermore, it has been demonstrated that the depth of the light penetration in an aqueous medium is limited by the absorption of the TiO₂ particles and by the molecules present in the reaction [6].

These issues can be addressed through the utilization of TiO₂ coatings deposited on inert matrices, which facilitate the separation of the reaction medium. TiO₂ thin films are employed primarily as anti-reflective coatings, protective layers for optical objects, sensors, and self-cleaning anti-pollution surfaces [7]. The latter have been the subject of increased research activity because of the environmental benefits they generate. The creation of antimicrobial surfaces is a particularly promising avenue of research, with potential applications in hospitals and other settings where the elimination of pathogenic microorganisms is desirable [8]. Such coatings are also employed in the filtration of ventilation systems, with the objective of eliminating volatile organic compounds in the gaseous phase as contaminants in enclosed spaces. In the treatment of contaminated water, thin films have been deposited on inert substrates that are easily accessible and have a large contact area, using methodologies that are straightforward to reproduce and cost-effective. It can be reasonably deduced that methodologies based on the sol-gel process, such as dip coating [9] and spin coating [10], are optimal for achieving the above-mentioned conditions. The latter allows for effective control of the thickness of thin films by increasing or decreasing the rpm, thereby utilizing centrifugal forces. Furthermore, it is straightforward to regulate, as the number of precursors in the sol-gel method with suitable solvents can be varied, thus conferring considerable versatility for deposition in diverse matrices with extensive contact areas [10]. There is sufficient evidence in the literature to demonstrate the effectiveness of these coatings in the photocatalytic degradation of dyes in aqueous media [11]. It is currently the objective of researchers to enhance the efficiency of these coated materials when illuminated by visible or solar radiation, with the aim of reducing the costs involved in the process. This has been achieved by producing TiO₂ coatings doped with lanthanide ions [12] or with TiO₂ mixed with other photoactive inorganic oxides, such as ZnO [13], SiO₂ [14] and CeO₂ [15]. The combination of the aforementioned ideas leads to the conclusion that CeO₂ can act as both a dopant and a mixed oxide in TiO₂ depending on its concentration. Moreover, it has been demonstrated that the CeO₂ coating can also generate a self-healing effect in stainless steel as a thin film, thereby preventing corrosion [16]. The evaluation of thin films of TiO₂ doped with CeO₂ under UV light has demonstrated that these films can store and release oxygen, which causes surface vacancies. Furthermore, the effect of electron–hole pair recombination is reduced, thereby increasing the photocatalytic activity and generating a higher availability of hydroxyl radicals, which are responsible for the oxidation process [17]. Despite the wide band gap energies of TiO₂ and CeO₂, these materials exhibit an enhanced spectral response under UV light. However, when doped or formed into heterojunctions in the form of thin films, it has been demonstrated that they generate hydrophilic surfaces photoinduced by electron–hole pairs. Also, CeO₂ reduces the band gap energy of TiO₂, thereby increasing its absorption of visible light and improving the separation of photogenerated charges. It has previously been reported that the variation in the E_g of TiO₂-CeO₂ can be attributed to the formation of defect levels inside the forbidden energy levels of the system. There are five reasons why the E_g changes in TiO₂: (i) substitution of Ti⁴⁺ by Ce⁴⁺, (ii) vacancies created by Ti⁴⁺, (iii) the quantum size effect relating to the surface, (iv) oxygen vacancies created by Ti³⁺ and (v) the presence of Ce³⁺ on the grain boundaries [18]. This has significant implications for photocatalytic applications under visible light, as evidenced by previous studies [19,20,21]. The observations indicate a dearth of studies that have investigated the photocatalytic efficiency of TiO₂-CeO₂ thin films in the removal of aqueous pollutants under sunlight or visible light. The objective of this study is to analyze the effect of CeO₂ variations in TiO₂-CeO₂ thin films deposited on borosilicate slides for the solar photocatalytic oxidation of two pesticides in aqueous media. The coated borosilicate slides were prepared using the sol-gel method with spin coating, achieving up to two layers of deposition without the presence of fractures in the coating, with the ability to be continuously used and separated in different reaction cycles.

2. Results and Discussion

2.1. XRD

Figure 1 illustrates the XRD patterns of TiO₂, CeO₂-doped TiO₂ and CeO₂ powders. It can be observed that the diffraction peaks exhibited by the titania and doped titania samples are located at 25.3°, 37.1°, 48.2°, 53.1°, 55.1° and 62.2° on the 2θ scale. These peaks are associated with the anatase (A) crystalline phase planes of TiO₂ (101, 004, 200, 105, 211 and 204). The identification of these peaks was made using the Joint Committee on Powder Diffraction Standards (JCPDS) library. Moreover, no diffraction peak corresponding to CeO₂ was identified in the doped samples, which indicates the effective dispersion of CeO₂ in the titania. This is because concentrations lower than 10% by weight cannot be identified by XRD due to the small size of the agglomerates, which cannot be detected by this technique [22]. Furthermore, it can be observed that increasing the amount of CeO₂ reduces the intensity of the peak associated with plane 101. Specifically, in the sample with the highest CeO₂ content, a slight shift is observed, which consequently affects structural parameters such as the crystal size (D) and strain index (Ɛ). These are presented in

Table 1. Crystal size and strain index values for TiO₂ and CeO₂-doped TiO₂ powders.

Sample	Crystallite Size, D (nm)	Strain, Ɛ
TiO2	10.66	0.01549
TiO2-CeO2-0.3%	9.70	0.01702
TiO2-CeO2-0.5%	9.64	0.01713
TiO2-CeO2-1.0%	9.55	0.01728
TiO2-CeO2-3.0%	9.08	0.01818
TiO2-CeO2-5.0%	8.14	0.02027

, where the crystal size decreases from 10.66 nm to 8.14 nm, while the strain index increases from 0.01549 to 0.02027. These observations confirm that the incorporation of cerium cations in titania affects the particle size and demonstrate a good interaction as a dopant.

2.2. XPS

In order to analyze the surface chemical composition and oxidation states of the elements present in the powders generated by the precursors used to synthesize the TiO₂ and CeO₂-doped TiO₂ thin films, the samples were investigated by X-ray photoelectron spectroscopy (XPS). Figure 2a illustrates the high-resolution XPS spectra of pure TiO₂ and TiO₂-CeO₂-5% employed to normalize the binding energy shift resulting from the electrostatic charges, with the C 1s peak at 284.50 eV serving as the reference point. The XPS spectra of Ti 2p in Figure 2b demonstrate the presence of Ti 2p_2/3 and Ti 2p_1/2 characteristic signals at 458.6 eV and 464.2 eV, respectively. This indicates the presence of titanium in the form of Ti⁴⁺ in the tetragonal structures of both samples [23]. However, upon deconvolution of the Ti 2p_3/2 peak in both samples, a signal at 456.49 eV attributed to the Ti³⁺ species is observed, indicating the presence of active and surface adsorption sites in heterogeneous photocatalysis [24]. This species exhibits a notable increase in relative abundance in the CeO₂-doped sample, as evidenced in

Table 2. Relative abundance data obtained by XPS for the chemical species present in the TiO₂ and TiO₂-CeO₂-5.0% powders.

Sample	Ti 2p		O 1s			Ce 3d
	Ti4+	Ti3+	O2−	OH	H2O	Ce4+	Ce3+
TiO2	458.12	456.49	529.37	531.60	533.49	-	-
	97.1	2.9	67.3	26	6.7	-	-
TiO2-CeO2-5.0%	458.20	456.49	529.45	531.50	533.39	882, 888	880, 884
	95.8	4.2	73.1	23.7	3.3	49.20, 51.60	50.90, 48.40

. This observation suggests an imbalance of surface charges produced by the dopant. The fitting of the O 1s signal for TiO₂ reveals the presence of three bands at 529.37 eV, 531.60 eV and 533.49 eV. In the doped material, these bands manifest at 529.45 eV, 531.50 eV and 533.39 eV, as illustrated in Figure 2c and Table 2. The signal associated with a binding energy of 529 eV is related to oxygen in the form of oxide (O²⁻) present in the superficial Ti–O bonds [25,26]. The relative abundance of this signal increases when CeO₂ is present as a dopant, which could confirm the formation of the Ti–O–Ce bond and generate a greater chemical interaction at the surface level between TiO₂ and CeO₂. With regard to the signals observed at 531 eV and 533 eV, these have been primarily attributed to the oxygen present in the hydroxyl groups (OH) and to the surface water (H₂O), respectively, which result in the formation of oxygen defects in the surface network of TiO₂ [17,27]. Additionally, Table 2 illustrates that the relative abundance of both signals diminishes in the presence of CeO₂, substantiating the assertion that this dopant induces the creation of more oxygen defects on the surface of titania, thereby enhancing the material’s oxidative capacity and its ability to store and release oxygen [28]. In contrast, Figure 2d presents a detailed analysis of the XPS spectrum of the doped sample, with an amplification factor applied (4×). This reveals the Ce 3d_5/2 signals, with the peaks at 882 eV and 888 eV corresponding to Ce⁴⁺ (3d⁹4f²) O (2p⁴) and Ce⁴⁺ (3d⁹4f¹) O (2p⁵), respectively. The remaining signals at 880 eV and 884 eV are attributed to the Ce³⁺ (3d⁹4f²) O (2p⁵) and Ce³⁺ (3d⁹4f¹) O (2p⁶) mixture, as previously reported in references [25,29,30]. The height and width of each peak in the four signals exhibit close intensities, indicating a relative abundance value of approximately 1:1, as observed in Table 2. This confirms the presence of the Ce^3+/4+ ion mixture in the TiO₂-CeO₂ at 5.0%, whereby the Ce³⁺ ion is responsible for the reduction of Ti⁴⁺ to Ti³⁺ in titania, resulting in the formation of additional surface oxygen defects. This, in turn, enhances the photocatalytic properties of the system under visible light [31].

2.3. Transmittance UV-Vis Spectroscopy

The thickness of the analyzed thin films was calculated utilizing the UV-Vis transmittance spectra, as illustrated in Figure 3a,b. This was achieved through the application of the following equation [32]:

$$t=\left|{\displaystyle \frac{M{\lambda }_1{\lambda }_2}{2 \left(n_1{\lambda }_2-n_2{\lambda }_1\right)}}\right|,$$

where M represents the number of oscillations between the two extremum points (M = 1 between two consecutive maxima or minima). λ₁, n(_λ1) and λ₂, n(_λ2) are the corresponding wavelengths and index of refraction. It can be observed that the addition of CeO₂ to titania at the lowest concentration results in a reduction in the thickness of the thin film, whether one or two layers are present. This can be attributed to the fact that the incorporation of CeO₂ reduces the crystal size of titania, which in turn prevents crystal growth by the formation of Ti–O–Ce bonds. Furthermore, surface nucleation is also related to the ionic radius of Ce³⁺ (0.115 nm) and Ce⁴⁺ (0.101 nm), which are larger than the Ti⁴⁺ ion (0.062 nm). In some cases, Ce ions replace Ti ions, occupying interstitial positions [33,34]. Upon reaching a CeO₂ concentration of 1.0%, the minimum thickness of the various layers is attained. Subsequently, upon increasing the CeO₂ content to its highest concentration, the thickness of the samples increased, with the lowest value observed for the slides with two layers compared to those with one layer (Table 3). This behavior is primarily attributed to the presence of CeO₂, as evidenced by the finding that when increasing the number of layers in the thin films of pure TiO₂ on slides, under similar synthesis conditions, the thickness of the coating is directly proportional to the number of layers added [35].

2.4. UV-Vis Diffuse Reflectance Spectroscopy

The band gap energy (E_g) for all the samples analyzed was calculated using the Tauc method, plotting (αhν)² against the photon energy, as TiO₂ has been demonstrated to exhibit indirect transition E_g values using the following equation [19]:

$$\alpha ={\displaystyle \frac{A{(hv-E_g)}^n}{hv}},$$

where α represents the absorption coefficient, hν denotes the photon energy, A is a constant, E_g signifies the band gap energy and n is ½ for TiO₂. The UV-Vis spectra presented in Figure 3c,d were employed to estimate the E_g in eV value and light absorption edge in nm of the samples by extrapolating a line along the slope of each spectrum to its intercept with the x-axis. The E_g and light absorption edge values for all the samples are presented in Table 3. It can be observed that the slides with a coating layer have values between 3.36 eV (TiO₂-CeO₂ 0.3%) and 3.40 eV (TiO₂-CeO₂ 3.0%). However, for the films comprising two layers, the values were found to fall within the range of 3.35 eV (TiO₂-CeO₂ 5.0%) to 3.43 eV (TiO₂-CeO₂ 0.5%). In the case of the samples with the highest Ce content, an increase in the number of layers resulted in a reduction in the E_g value. This finding aligns with the observations reported by Sta et al. for pure TiO₂ thin films [35]. The value of the E_g in the TiO₂-CeO₂ system can be modified by a number of factors, including the preparation method, the choice of precursors, the application of heat treatment, and so forth. These factors give rise to the formation of mesoporous and macroporous structures, with the former displaying a lower E_g. Consequently, the reduction of this parameter can be attributed to the formation of defective levels situated between the E_g levels of TiO₂ and CeO₂; these levels are responsible for the observed reduction in the E_g. The interstitial substitution of Ti⁴⁺ ions by Ce³⁺ and Ce⁴⁺ ions results in the generation of surface oxygen vacancies. Furthermore, the presence of these ions at the grain boundaries leads to the formation of new energy levels situated within the electronic structure of TiO₂-CeO₂ [18]. In addition, it is established that a lanthanide-induced red shift occurs when the unoccupied 4f orbitals of Ce³⁺ permit an electronic transition to the 2p orbitals of oxygen (valence band), which then transfer electrons to the 3d orbitals of titanium (conduction band). In light of the aforementioned evidence, it can be posited that the 4f orbitals of Ce are occupied, thereby forming new electronic states within the TiO₂ band gap. This, consequently, reduces the charge transfer distance and the amount of photogenerated energy required to excite Ce-doped TiO₂ with light associated with longer wavelengths [18,36,37].

Table 3. Results of the thickness and bandgap energy of the thin films of TiO₂ and doped TiO₂-CeO₂ for the samples with one and two layers, respectively.

Sample	Thickness, t (nm)		Band Gap Energy, Eg (eV) [λ (nm)] *
	One layer	Two Layers	One Layer	Two Layers
TiO2	159.48	167.35	3.38 (366.86)	3.40 (364.70)
TiO2-CeO2-0.3%	136.97	134.39	3.36 (369.04)	3.41 (363.63)
TiO2-CeO2-0.5%	154.01	133.23	3.38 (366.86)	3.43 (361.51)
TiO2-CeO2-1.0%	125.66	125.42	3.39 (365.78)	3.42 (362.57)
TiO2-CeO2-3.0%	147.48	133.14	3.40 (364.71)	3.37 (367.95)
TiO2-CeO2-5.0%	230.03	182.18	3.37 (367.95)	3.35 (370.15)

* Light absorption edge [38].

Table 3. Results of the thickness and bandgap energy of the thin films of TiO₂ and doped TiO₂-CeO₂ for the samples with one and two layers, respectively.

Sample	Thickness, t (nm)		Band Gap Energy, Eg (eV) [λ (nm)] *
	One layer	Two Layers	One Layer	Two Layers
TiO2	159.48	167.35	3.38 (366.86)	3.40 (364.70)
TiO2-CeO2-0.3%	136.97	134.39	3.36 (369.04)	3.41 (363.63)
TiO2-CeO2-0.5%	154.01	133.23	3.38 (366.86)	3.43 (361.51)
TiO2-CeO2-1.0%	125.66	125.42	3.39 (365.78)	3.42 (362.57)
TiO2-CeO2-3.0%	147.48	133.14	3.40 (364.71)	3.37 (367.95)
TiO2-CeO2-5.0%	230.03	182.18	3.37 (367.95)	3.35 (370.15)

* Light absorption edge [38].

2.5. Profilometry and AFM

The thickness of the thin films with varying CeO₂ concentrations and layer numbers was confirmed by atomic force microscopy (AFM) and profilometry, as illustrated in Figure 4. The AFM images demonstrate that at lower CeO₂ content (a) (b), a rough surface is observed for both cases. It can be observed that the incorporation of the second layer increases the roughness, which is a characteristic of TiO₂ thin films deposited on glass [35]. However, the heat treatment temperature of the samples also influenced this behavior, as evidenced by the substantial increase in roughness when the treatment temperature rises from 400 to 600 °C in the TiO₂ thin films, due to the superficial growth of grains [39]. Furthermore, the presence of agglomerates and cracks is evident upon the addition of a second layer, with the thickness value approaching the optically calculated value, as illustrated in Table 3. Conversely, an increase in the Ce content in the samples resulted in a reduction in the roughness (c), which exhibited a slight increase upon deposition of a second layer (d). The incorporation of Ce as a dopant in TiO₂ thin films has been observed to enhance the surface uniformity, leading to the formation of stable agglomerates devoid of any indications of crack formation. These larger agglomerates may potentially be associated with a partial hydrolysis phenomenon that occurs during the synthesis process [19], as evidenced by the notable increase in the thickness of the thin films. Nevertheless, the structural development of the TiO₂-CeO₂ system with homogeneous and uniform characteristics can be explained by a mechanism that is determined at the solid solubility limit of Ce in titania, which can be expressed as follows. Up to the solid solubility limit, there is a homogeneous distribution of point defects that improves nucleation, recrystallization, and grain growth. At the Ce limit, the semiconductor properties of the system are optimized [40]. The reduction in thickness and the increase in roughness with the presence of cracks were also increased by the addition of another layer for the samples with high Ce content. The thickness of the thin films for the samples with lower and higher Ce content as a dopant was also determined by profilometry (e), and these results show values very close to those calculated theoretically in Table 3, which is the same for the thickness values obtained by AFM.

2.6. SEM

Figure 5 depicts the surface morphology of the TiO₂-CeO₂-5.0 wt% thin film with one layer (a). It can be observed that the surface exhibits a smooth appearance with the presence of uniformly distributed agglomerates at a magnification of 1500×. The cross-section view of the same sample (b) reveals that the thickness of the thin film was approximately 0.225 um (225 nm), which is in close agreement with the calculated value presented in Table 3. The EDS spectrum (c) and elemental distribution (d) are also illustrated in this figure, where the presence of Ce can be observed. The existence of elements such as Na, Mg, Si and Ca is attributed to the elemental content of the borosilicate slide, as previously reported [41].

2.7. Photocatalytic Test

The outcomes of the photocatalytic assessment of the degradation and mineralization of diuron (D) and methyl parathion (MP) using single- and double-layer TiO₂ and Ce-doped TiO₂ thin films under solar illumination are presented in

Table 4. Photocatalytic evaluation results. Reaction rate constant (kapp), half-life time (t_1/2) and % mineralization by TOC.

One Layer	kaap (min−1)	T1/2 (min)	TOC (%)	Two Layers	kaap (min−1)	T1/2 (min)	TOC (%)
Photolysis            D MP	0.0004	1732	4.45	Photolysis                   D MP	0.0004	1732	4.45
	0.0005	1386	1.22		0.0005	1386	1.22
TiO2                              D MP	0.0024	289	13.20	TiO2                    D MP	0.0023	301	19.25
	0.0021	330	28.88		0.0051	136	29.32
TiO2-CeO2 0.3%        D MP	0.0043	161	32.52	TiO2-CeO2 0.3%      D MP	0.0012	578	27.41
	0.0075	92	41.83		0.0059	117	37.57
TiO2-CeO2 0.5%D MP	0.0031	224	20.18	TiO2-CeO2 0.5%      D MP	0.0011	630	28.06
	0.0047	147	35.40		0.0050	139	37.91
TiO2-CeO2 1.0%D MP	0.0027	257	18.08	TiO2-CeO2 1.0%      D MP	0.0036	193	34.99
	0.0082	85	42.99		0.0049	141	40.17
TiO2-CeO2 3.0%D MP	0.0037	187	24.38	TiO2-CeO2 3.0%      D MP	0.0032	217	37.57
	0.0025	277	59.30		0.0083	84	57.95
TiO2-CeO2 5.0%      D MP	0.0044	158	35.93	TiO2-CeO2 5.0%      D MP	0.0050	138	38.55
	0.0087	80	69.46		0.0088	79	70.63

. The rate constants and half-life times for each reaction were employed as parameters to estimate the photodegradation using a pseudo-first-order reaction kinetics model. It can be observed that photolysis has no significant effect on the photodegradation and mineralization of both contaminants. Additionally, it was found that in the samples with one layer, the doped thin films exhibited enhanced activity compared to the thin films with pure TiO₂, thereby underscoring the importance of Ce doping. Conversely, the opposite trend was observed for thin films with a lower amount of dopant when a second layer was added. This reduction in degradation can be attributed to the generation of intermediate species, which possibly have absorbances close to λ = 275 nm for MP and λ = 250 nm for D, generated by the incorporation of the second layer and the effect of irradiation with sunlight. This is in line with the findings of Ameen et al., who reported that TiO₂-CeO₂ nanocomposites under visible irradiation generate intermediates from 20 min of reaction in the photodegradation of bromophenol, which lead to the mineralization of the dye [42]. Then, there is a great possibility that in this study, the decrease in the photocatalytic activity of some thin films of titania doped with cerium is due to the effect of sunlight. However, these intermediates play a crucial role in the reaction mechanisms leading to mineralization, a parameter used to determine the photocatalytic oxidation of pesticides.

In terms of the degradation of diuron, the thin films doped with a single layer that exhibited the greatest activity were TiO₂-CeO₂-0.3% (0.0043 min⁻¹) and TiO₂-CeO₂-5.0% (0.0044 min⁻¹). The latter remained the most active when the second layer was added, resulting in an increased rate constant of up to 0.0050 min⁻¹. In light of the aforementioned explanation, it can be posited that these thin films exhibit enhanced chemical stability, thereby preventing the generation of additional reaction intermediates. With regard to the photodegradation of methyl parathion, the thin films comprising one or two layers that exhibited the highest activity were also those doped with CeO₂ at 5.0% (0.0087 min⁻¹ and 0.0088 min⁻¹), respectively. Similarly, the samples with a lower Ce content exhibited reduced activity. In this instance, the thin films doped with Ce at 5.0% displayed a greater affinity for degrading methyl parathion than diuron, irrespective of the number of layers.

Table 4 also presents the data regarding the percentages of diuron and methyl parathion mineralization of all the thin films evaluated in the photocatalytic process under sunlight. Figure 6a,b depict the graphs that delineate the photocatalytic behavior of theTiO₂ and Ce-doped TiO₂ thin films with one and two layers in the diuron mineralization process. The findings illustrate that photolysis and pure TiO₂ thin films with one and two layers display a reduced capacity to mineralize both contaminants in comparison to the Ce-doped samples. Moreover, it was shown that the photocatalytic activity of pure titania thin films under sunlight is enhanced by increasing the thickness with respect to both contaminants. This is corroborated by reports indicating that the activity of TiO₂ thin films on slides under visible light is optimal when the thickness is less than 200 nm [43], which aligns with the values reported in Table 3 (159.48 nm and 167.35 nm).

The TiO₂-CeO₂-5.0% thin films with one and two layers demonstrated the highest mineralization percentage for both diuron and methyl parathion. It was observed that the mineralization of diuron in the films with the lowest Ce content, with a single coating, was found to be very similar to that of the films with the highest Ce content. However, the addition of a second layer resulted in a reduction of the mineralization. The mineralization is reduced, which is directly correlated with the decrease in the thickness of the thin film (136.97 nm to 134.39 nm). This is attributable to the presence of Ce, which serves to compact the thickness when a second layer is added. This is due to the fact that the addition of cerium reduces the crystal size, as observed in Table 1. This phenomenon does not occur with pure TiO₂ films. While the TiO₂-CeO₂-5.0% thin films with two layers exhibited the most favorable mineralization percentage for diuron, two samples demonstrated comparable values (TiO₂-CeO₂-3.0% with two layers and TiO₂-CeO₂-5.0% with one layer). The maximum difference observed was 2.7%, which prompted us to consider the possibility that the TiO₂-CeO₂-5.0% sample with one layer has a better possibility of being applied in a larger-scale reaction system for the mineralization of diuron. This approach would result in lower expense with respect to its photocatalytic efficiency. Figure 6c,d also demonstrate the mineralization of methyl parathion under sunlight. It can be observed that an increase in the amount of Ce in the thin films correlates with an enhancement of photocatalytic activity. This finding aligns with previous research on the utilization of Ce coatings on TiO₂ thin films for the photodegradation of dyes [42]. A comparable phenomenon is observed when the TiO₂-CeO₂ and TiO₂-CeO₂-ZrO₂ systems are evaluated under visible irradiation for the elimination of Rhb [44]. However, this trend is subject to an upper limit with regard to the optimal Ce content, given that exceeding this amount results in a decline in the photocatalytic activity. It has been estimated that the optimal amount of Ce for thin films of TiO₂ is 0.05 molar, used either under UV light or visible light, for the degradation of the MB dye [25]. Conversely, the TiO₂-CeO₂ powders indicate that the optimal Ce content is present at 5% wt when evaluated in the UV light degradation of 2,4-D. This is closely related to the crystal size, because as the optimal Ce content is exceeded, the crystal size increases. It is known that the photocatalytic activity increases inversely proportional to the average particle size [22]. Accordingly, the present study has revealed that the optimal Ce content is also manifested at 5% by weight, which is consistent with the range of 2–10 wt% established by Kumari et al. in their compilation [18]. Given these results, it may be proposed to perform future studies under the same conditions describing TiO₂-CeO₂ thin films with CeO₂ concentrations above 5.0 wt.% to establish a real optimum CeO₂ concentration in this type of system and to determine up to which amount of CeO₂ the reaction yield decreases.

The photocatalytic activity of Ce-doped TiO₂ thin films under solar irradiation can be explained by the presence of the Ce⁴⁺ and Ce³⁺ states of CeO₂, since the 4f orbitals of these species are able to absorb visible light (400–500 nm) by means of the charge transfer transition that occurs between the 4f electrons and the empty orbitals of Ti⁴⁺ [25]. Then, during the photocatalytic process, the photogenerated electrons are immediately transferred from the valence band to the conduction band of TiO₂ and CeO₂ (avoiding their recombination with holes), which migrate to the surface and are transferred to the absorbed oxygen molecules, which can combine to form superoxide radicals [45]. In this sense, the presence of Ce can also increase the catalytic activity of TiO₂ due to its ability to store oxygen on the surface and by reducing the particle size of the material, which has been previously studied by the working group [46].

The main photogenerated species involved in the photocatalytic oxidation mechanism of diuron and methyl parathion using CeO₂-doped TiO₂ thin films were developed through the implementation of photocatalytic evaluation tests conducted in the presence of scavengers, including isopropanol (for OH·), benzoquinone (for O₂²⁻·), EDTA (for h⁺) and potassium dichromate (for e⁻) [27,47], as shown in Figure 7a. The rate constants exhibited a notable decline when isopropanol was used, followed by benzoquinone, indicating that hydroxyl radicals and photogenerated superoxide radicals play a pivotal role in the oxidation (mineralization) of diuron and methyl parathion. Because of this, a schematic representation describing the reaction mechanism based on the formation of hydroxyl radicals (OH·) and superoxides (O₂²⁻·) is proposed in Figure 8, where the oxygen molecules (O₂) adsorbed on the surface of the titania react with the electrons (e⁻) to form O₂²⁻·, which inhibits the recombination of electron–hole pairs and interacts with the water (H₂O) to form hydrogen peroxide (H₂O₂), which also immediately reacts with e⁻ to generate OH·. At the same time, the H₂O adsorbed on the surface of the material reacts with the photogenerated holes (h⁺) to produce OH· [18]. Therefore, the production of hydroxyl radicals and superoxides is of great importance as they are responsible for mineralizing the diuron and methyl parathion present as aqueous pollutants. Section (b) of the same figure describes the five reuse cycles applied to the same sample, demonstrating the stability of the material after several consecutive reactions using the two pollutant molecules. It can be observed that the value for diuron mineralization varies within a narrow range of 35–41%, with a standard deviation (SD) of ±2.53, while for methyl parathion it varies between 63–70% (SD: ±2.96). This demonstrates that there are no significant alterations in the photocatalytic oxidation of the pollutants following reuse. With regard to the stability of the thin films in sections (c) and (d), SEM micrographs of the cross-section of the sample have been exanimated. It has been observed that there are no structural changes before and after the reaction, which confirms the stability of the thin films in different reaction runs. The sample TiO₂-CeO₂-5.0% with one layer shows great potential to be applied as a fixed bed in a large-scale continuous photocatalytic reaction system, which can be illuminated during the day by sunlight and at night by artificial visible light (LED light) to maintain the reaction for long times until the mineralization of the contaminants.

3. Materials and Methods

3.1. Thin Films Preparations

Borosilicate slides with dimensions of 50 × 20 × 1 mm were employed as substrates for the deposition of thin films. The slides were cleaned using a 1:1 water–acetone mixture in an ultrasonic bath for 10 min, after which they were dried in a vacuum chamber at 50 °C in order to remove any residual impurities. The precursors employed in the synthesis of the thin films via the sol-gel method were procured through the following means. For the synthesis of TiO₂, a solution of titanium butoxide (Sigma-Aldrich 97%, Ciudad de México, México) was prepared with n-butanol (Sigma-Aldrich 99% Ciudad de México, México) in a 1:4 ratio, with the reaction mixture kept under constant stirring for 30 min in a closed container with an inert atmosphere maintaining a pH of 5.25. The procedure for TiO₂-CeO₂ was modified slightly. Previously, a specific quantity of cerium nitrate hexahydrate (Sigma-Aldrich 99.99%, Ciudad de México, México) was dissolved in butanol, with the quantity varying according to the established weight percentage (0.3, 0.5, 1.0, 3.0 and 5.0%, respectively); here, the pH increased to 5.43 when the largest amount of Ce precursor was added. The deposition of the films on the substrates was conducted using a spin-coating apparatus (Laurell Technologies Corporation, Laurell WS-650MZ-23NPPB, PA, USA). Subsequently, the substrate was coated with the precursor solution at 500 rpm in a vacuum, and the speed was then increased to 1500 rpm. Subsequently, the films were subjected to a drying process at 120 °C for 15 min and a calcining procedure at 500 °C for 4 h, with a heating rate of 2 °C/min. This resulted in the formation of a highly transparent and fixed layer on the substrate. The deposition of successive layers without fractures and high transparency was only possible with the application of two layers. The second layer was deposited by repeating the aforementioned methodology after the first layer was dried at 120 °C. In total, six different types of thin films were produced with 10 replicates. Subsequently, the various precursors of the thin films, including the residual cerium nitrate dissolved in butanol, were subjected to the same heat treatment until the formation of powders was achieved. These powders were then analyzed using both DRX and XPS.

3.2. Sample Characterization

The obtained powders were characterized by XRD in order to identify the crystalline phases, as well as to estimate the average crystal size and deformation, using the Debye-Scherrer equation in the 101 plane of the anatase phase of TiO₂ [21]. The XRD spectra were obtained at room temperature using a Bruker D8 Advance powder diffractometer, with Cu-Kα radiation (λ = 0.154 nm) and a monochromatic graphite secondary beam in a range of 20 to 70 on the 2θ scale with a step of 0.05°, with a measurement time of 0.5 s per point. Furthermore, the powders were subjected to XPS analysis, which enabled the acquisition of data pertaining to the chemical and electronic environment at the surface level. This was achieved through the utilization of a VG Escalab II spectrometer, equipped with a K_α X-ray source (1486.6 eV) at 225 W (15 kV, 15 mA) and a hemispherical analyzer for electron detection. The XPS spectra were processed using Casa software, version 2.3.14, with carbon as the reference element at 285.0 eV in order to align the signals of the analyzed species. The thickness of the thin films was determined by measuring the optical transmittance spectra of each sample using the Manifacier equation [35] with a UV-Vis spectrophotometer (Agilent, Varian model Cary 300, VIC, Australia) in a wavelength range of 200 to 800 nm. However, the determination of the band gap energy was conducted using the same spectrophotometer but with an integration sphere to obtain diffuse reflectance spectra. These were then processed using the Tauc methodology [12]. The thickness of the layers and the roughness were confirmed only for the samples with higher photocatalytic activity using a surface profilometer (Bruker, Dektak-XT, Billerica, MA, USA) and an atomic force microscope (Ambios Technology Inc., Q-scope 250/400 Nomad, Santa Cruz, CA, USA). To analyze the morphology, thickness, microstructure and elemental analysis of the best thin film, a scanning electron microscope (JEOL, model JSM-6010LA, Tokyo, Japan) coupled to an energy-dispersive X-ray spectroscopy detector was used.

3.3. Photocatalytic Properties

Figure 9 depicts the reactor utilized for the assessment of the photocatalytic activity of the TiO₂ and TiO₂-CeO₂ thin films on slides. The reactor was situated within a solar simulator (Atlas Suntest CPS+ with a 15 kW Xe lamp and a specialized quartz filter that restricts light transmission) for the execution of the photocatalytic oxidation reactions of methyl parathion (MP = 25 mg/L, pH = 6.63) and diuron (D = 20 mg/L, pH = 6.87). Here, it was maintained at a constant temperature of 23 °C due to the constant recirculation of water through the inlets and outlets that the reactor has provided to maintain the reaction temperature constant. Prior to commencing the photocatalytic reactions for each pollutant, the system was stirred in the absence of light for a period of 30 min to achieve equilibrium between adsorption and desorption. The total reaction time was 180 min, with samples collected at 30 min intervals using filters with a pore size of 0.2 µm (Millex^®-GN, 25 mm, Millipore Nylon, Cork, Ireland). The photodegradation was monitored by UV-Vis spectroscopy, with the maximum absorbance peak of each pollutant (MP, λ = 275 nm; D, λ = 250 nm) being used as a reference point. The mineralization of MP and D was determined utilizing a total organic carbon analyzer (Shimadzu, UV-1700, Kyoto, Japan) with an experimental error of ±0.05 mg/L. The thin films exhibiting the most effective photocatalytic performance for MP and D were subjected to five consecutive evaluations. At the conclusion of each reaction, the coated slides were rinsed with deionized water and subjected to a 24 h drying process at 80 °C to facilitate reuse. Reactions with scavengers, including benzoquinone (O₂²⁻·), isopropanol (OH·), EDTA (h⁺) and potassium dichromate (e⁻), were conducted at a concentration of 10 mg/L to ascertain the primary photogenerated species engaged in the photocatalytic oxidation mechanism. Furthermore, the stability of the thin films was evaluated through the analysis of scanning electron microscopy (SEM) micrographs of the cross-sectional sample, both prior to and following the reaction.

4. Conclusions

Borosilicate slides were coated with Ce-doped TiO₂ thin films via the sol-gel spin-coating method. An increase in the amount of Ce resulted in a reduction in the crystal size and an enhancement of the strain index, which indicates an effective dispersion and interaction between TiO₂ and CeO₂; this was confirmed by XPS, which revealed the presence of Ti⁴⁺, Ti³⁺, Ce⁴⁺ and Ce³⁺ ions produced by the Ti–O–Ce bond. The Ce^3+/4+ ions facilitate the generation of oxygen vacancies, which in turn give rise to the formation of new energy levels in the TiO₂-CeO₂ electronic structure. The observed increase in sunlight absorption can be attributed to the presence of unoccupied 4f orbitals, which reduce the charge transfer distance and consequently decrease the photon energy, promoting an electron transfer from the valence band to the conduction band. On the other hand, the addition of a second layer increases the roughness but decreases with the higher CeO₂ content, causing greater thickness. The addition of CeO₂ enhances the photocatalytic activity of the material under sunlight due to the improvement in visible light absorption, oxygen storage and reduction in crystal size. TiO₂-CeO₂ thin films at 5.0 wt% with one layer exhibited the most favorable photocatalytic behavior under sunlight, displaying a higher affinity for oxidizing methyl parathion through photogenerated hydroxyl radicals and superoxides. In addition, future studies can be proposed to evaluate these thin films with an amount of CeO₂ higher than 5.0% to establish an optimal concentration. These coatings have the potential for large-scale application in the treatment of contaminated water, as they demonstrate stability and reactivity even after continuous use.