Ce³⁺/Ce⁴⁺-Modified TiO₂ Nanoflowers: Boosting Solar Photocatalytic Efficiency

Abstract

Cerium-doped titania nanoflowers are obtained by hydrothermal synthesis, with different amounts of cerium (0.3, 0.5, and 1.0 at%). Both undoped nanoflowers (TNF) and Ce-doped TNF (Cex) are tested as photocatalysts in the degradation of the target pollutant (metronidazole) under simulated solar light. The samples are rutile polymorphs with high crystallinity and present a nanoflower-like morphology of about 1 µm in diameter and are made up of nanoscale petals (in the range of 100–300 nm). EDX spectroscopy was coupled with SEM and performed on the Ce-doped samples to determine the elemental composition of the catalysts and the Ce distribution in each sample. Optical and electronic spectroscopies reveal that Ce loading narrows the band gap from 3.0 to 2.8 eV, extending light absorption into the visible range of the spectrum and thus enhancing the photocatalytic activity. The best sample, Ce1, achieved 67% degradation of metronidazole after 360 min under solar irradiation at pH 4, compared to bare TNF, which reached 35%. Reusability tests confirm the chemical stability and photocatalytic efficiency of Ce1 over three cycles, and free-radical trapping experiments confirmed ·O₂⁻ and ·OH as major active species in metronidazole degradation. This study highlights the synergistic impact of morphology and doping on solar-driven organic pollutant degradation.

1. Introduction

Globally, water scarcity has been intensified by the combined effects of climate change and population growth [1]. According to the IPCC (Intergovernmental Panel on Climate Change) report, about 380 billion m³ of wastewater is produced worldwide every year, and only 15% of it is adequately treated [2]. Furthermore, in recent decades, and particularly after the COVID-19 pandemic, the production and use of pharmaceuticals have rapidly increased. The widespread use and misuse of this class of substances have resulted in higher levels of pollutants in wastewater, due to the absence of or insufficient treatment before their release into the ecosystem. The contamination of aquatic ecosystems by the presence of these compounds, potentially hazardous even at low concentrations (<1 µg/L), has toxic effects on all organisms [3]. This class of compounds is known as persistent organic pollutants (POPs) because of their intrinsic difficulty in being degraded, and for this reason, classical treatments are not sufficient for their complete removal. Among them all, one of the most commonly detected antibiotics is metronidazole (MDZ), used to treat infections caused by microorganisms in the skin, digestive tract, and reproductive system [4,5]. Advanced oxidation processes (AOPs), a group of chemical treatment processes, have attracted great interest [6]. More specifically, heterogeneous photocatalysis exploits the interaction of light with a semiconductor, in which electron-hole (e⁻-h⁺) pairs are created, leading to the formation of reactive oxygen species (ROS), responsible for organic molecule mineralization [7,8].

An efficient and widely used semiconductor is titanium dioxide (TiO₂, titania), due to its many properties, namely stability, low toxicity, high photocatalytic activity, and ease of synthesis [9]. Despite these properties, titania shows some limitations. Firstly, the high recombination rate prevents or delays redox reactions that degrade organic pollutants [10]. Additionally, the wide band gap between 3.0 and 3.2 eV, depending on the crystalline phase, enables interactions only with high-energy radiation, thus in the ultraviolet (UV) region. UV radiation accounts for a small fraction of the solar spectrum (~5%), meaning that titania is poorly efficient in the rest of the wavelength range, limiting its overall application under solar light [11,12].

Many strategies have been studied to increase the photocatalytic activity of pure titania in the visible (Vis) and near-infrared (NIR) ranges. Among all the possibilities, doping with metallic and/or non-metallic elements appears to be promising [13]. Specifically, the introduction of cerium (Ce) ions appears to enhance the oxygen atom mobility, promoted by the Ce³⁺/Ce⁴⁺ redox pair. Ce ions enable the formation of a band near the conduction band (CB), which causes a decrease in the band gap energy value, a red shift in absorption, and a drop in the charge recombination rate, thus improving the overall photocatalytic activity of the semiconductor in the visible range [13,14].

In the present work, we propose the synthesis of titania nanoflowers (TNFs), coupled with the introduction of novel features through doping engineering, to optimize optoelectronic properties and enhance photocatalytic activity. The synthesis of TNF follows a hydrothermal approach, and Ce-doped TNFs are obtained by adding cerium precursor during the synthesis. The Ce concentration is varied in atomic percentage (at%), with a range of 0.3–1.0 at%, to investigate the optimal content by photocatalytic tests under UV light and simulated solar light (SSL) irradiation. X-ray diffraction (XRD), scanning electron microscopy (SEM), and optical spectroscopy were used to study the structural and electronic properties of the synthesized materials. Photocatalytic activity was evaluated based on the photooxidation of MDZ as the target molecule, in aqueous solution, at room temperature, atmospheric pressure, under SSL irradiation, and in two different pH conditions.

The main goal was to prove that synergy between morphology control and doping engineering significantly enhances the catalysts’ optical absorption capability and overall photocatalytic efficiency for the photooxidation of complex molecules such as antibiotics. Our work paves the way for the development of efficient Ce-doped TiO₂ photocatalysts in the degradation of persistent organic pollutants in wastewater.

2. Results and Discussion

2.1. Morphology and Structure

Hierarchical submicrometric bare and Ce-doped TiO₂ nanoflowers were obtained using a hydrothermal approach (Figure 1a). To find the optimal Ce content, three different concentrations were chosen: 0.3, 0.5, and 1.0 at% Ce in TiO₂. The samples were identified as TNFt for bare TiO₂ nanoflowers, where t stands for the reaction temperature (200 and 150 °C), and Cex for Ce-doped TNF, where x stands for the Ce content in at%.

Figure 1. (a) Reaction scheme for the preparation of bare and Ce-doped TNF. (b) X-ray diffraction patterns of TNF samples with TiO₂ rutile pdf #21-1276 (fuchsia lines), brookite pdf #0005160 (light blue lines), and ceria pdf #00-034-0394 (yellow lines) as references. FE-SEM images of (c) TNF200, (d) TNF150, and (e) Ce1 at different magnifications.

The surface morphology of all samples was analyzed by Field-Emission gun Scanning Electron Microscopy (FE-SEM). As can be observed in Figure 1c,d, the hydrothermal temperature change does not affect the morphology of the particles, presenting a nanoflower-like morphology of about 1 µm in diameter and comprising nanoscale petals (in the range of 100–300 nm). Figure 1e and Figure S1b,c report the SEM micrographs of the Ce-doped samples. A slight change in morphology is observed in the doped materials. Specifically, the addition of Ce-ions seems to increase the nuclei growth for the rods composing each nanoflower and the average dimensions (from about 1 µm up to about 4.5 µm for Ce05). EDX spectroscopy was coupled to SEM and performed on the Ce-doped samples to determine the elemental composition of the catalysts and the actual concentration of Ce in each sample. The results are reported in Figures S2–S4, where a difference between the theoretical at% and the actual at% can be observed, with a maximum of 0.5 at% Ce in the Ce1 sample.

The crystal structure and phase identification were carried out by X-ray diffraction (XRD) analysis (Figure 1b). According to the JCPDS card number 21-1276, all samples present the characteristic peaks of the rutile phase at 2θ (°) 27.4, 36.1, 39.2, 41.2, 44.1, 54.3, 56.6, 69.0, and 69.8. After the doping process, the appearance of ceria (CeO₂, ICDD pdf number 00-034-0394) and TiO₂ brookite phase (AMCSD pdf number 0005160) is noticed. The transition from rutile to brookite has been reported in the literature [15] and can eventually occur when low concentrations of Ce are added to the sample. As a matter of fact, as the content of cerium increases, the peaks related to brookite decrease (25.4, 30.9, 32.9, and 37.4° of 2θ).

2.2. Optical Properties

The effect of Ce doping on the optical and electronic properties of TNF has been studied by diffuse reflectance UV-visible-near-infrared spectroscopy (DRUV-Vis-NIR). The reflectance spectra of all samples, shown in Figure 2a, present the main absorption feature of TiO₂ optical bandgap (lying at ~3.1 eV), which induces a strong absorption edge at about 390 nm, corresponding to the transition O²⁻ (2p) → Ti⁴⁺ (3d). For all Ce-doped samples, a clear red-shift towards higher wavelengths of the main band can be observed, compared to bare TNF (Figure 2b). Ce doping allows higher absorbance in the 390–490 nm region, compared to undoped samples. Moreover, increased oxygen mobility, promoted by the Ce³⁺/Ce⁴⁺ redox couple, can contribute to narrowing the band gap of bare TiO₂. It can also be noticed that the onset of absorption in doped samples is localized at higher wavelengths (~500 nm). This can be induced by the presence of additional localized states within the bandgap structure of TiO₂ (for instance, Ce 4f states below TiO₂ CB), which leads to an overall bandgap energy reduction (Figure 2c,d) [13].

Figure 2. (a) DRUV-Vis-NIR spectra, (b) F(R) spectra calculated through the Kubelka–Munk function, (c) Tauc plots including an example of the intercept method used to determine the band gap values (inset graph, linear fittings in dashed red lines), and (d) band gap energy values for the bare and Ce-doped samples.

Room-temperature steady-state emission spectra and Time-Correlated Single Photon Counting (TCSPC) measurements were recorded for all TNF150 and Ce-doped TNF samples, exciting at 375 nm with a picosecond-pulsed laser. Since the emission intensities were very low, the area under the ns-decay curves (Figure 3a) was integrated to quantitatively evaluate the variations among samples. The integrated values reveal a progressive decrease in emission intensity from bare TNF to Ce-doped TNF, suggesting a decrease in charge carrier recombination in the presence of Ce ions (Figure 3b and Table 1). This behavior suggests that Ce doping introduces additional non-radiative pathways or trap states that favor charge separation over radiative recombination. This effect is beneficial for photocatalytic applications, as it enhances the lifetime of photogenerated carriers, thus improving the probability of surface redox reactions.

Figure 3. (a) TCSPC decays of bare and Ce-doped TNF150 evaluated at 450 nm (maximum wavelength of emission spectra), excited at 375 nm with a ps-pulsed laser (PicoQuant, Berlin, Germany) using a cut-off filter at 430 nm (430 LP). The corresponding instrument response function (IRF, black curve) and fittings (dashed red lines) for TNF150 and Ce1 are also reported. (b) Intensities of the bare and Ce-doped TNF calculated as absolute area of the ns-decay, including bar errors.

Table 1. Intensity calculated as absolute area and intensity averaged ns-lifetimes <τ> and relative standard deviations.

Sample	Intensity (a.u.)	ns-Lifetime <τ> (ns)
TNF150	2.0 × 10−6 ± 4.8 × 10−7	2.48 ± 0.09
Ce03	4.1 × 10−7 ± 1.2 × 10−7	--
Ce05	4.6 × 10−7 ± 1.3 × 10−7	--
Ce1	6.2 × 10−7 ± 1.9 × 10−7	1.45 ± 0.05

During hydrothermal synthesis, Ce³⁺/Ce⁴⁺ ions partially substitute Ti⁴⁺ sites within the TiO₂ lattice or are incorporated near the surface as Ce–O–Ti species. Because Ce³⁺ has a slightly larger ionic radius than Ti⁴⁺, this substitution introduces lattice distortion and creates oxygen vacancies to maintain charge balance. Both the oxygen vacancies and the Ce³⁺/Ce⁴⁺ redox pair contribute to trap electronic states that promote visible-light absorption and enhance charge separation. These structural and electronic effects collectively lead to improved photocatalytic performance.

2.3. Photocatalytic Activity Evaluation

Photocatalytic activity was investigated by monitoring the degradation efficiency of a target organic pollutant. Tests were carried out under simulated solar light (SSL) irradiation in aqueous solution at room temperature and under atmospheric pressure. As mentioned above, a variety of organic substances produced mainly by human activity, known as emerging contaminants (ECs), are regularly detected in water. ECs belong to different classes of compounds (i.e., dyes, pharmaceuticals, pesticides, surfactants, polycyclic aromatic chemicals, and others), which may cause dramatic effects to human and ecosystem health when released into the natural environment [16,17]. Among ECs, pharmaceuticals are widely distributed in aquatic environments [16], particularly antibiotics derived from industrial waste and pharmacological use for the treatment of human and animal infections. Hence, it is important to assess the efficiency of the synthesized photocatalysts on these contaminants to investigate their practical application.

Specifically, photocatalytic activity was investigated by monitoring the degradation efficiency of metronidazole (MDZ). Tests were carried out under UV (Figure S5) and simulated solar light (SSL, Figures S6 and S7) irradiation in aqueous solution, at room temperature and under atmospheric pressure. Photodegradation was monitored for a reaction time of 180 or 360 min, following the maximum absorbance peak of MDZ (319 nm). Photodegradation tests were performed both on bare (TNF150) and Ce-doped titania samples. The TNF150 sample was selected instead of TNF200 as the bare TiO₂ reference because the crystalline structure and morphology remain unchanged between the two synthesis temperatures. Consequently, from an energy-saving and circular economy perspective, the lower-temperature sample was preferred. Moreover, to investigate the influence of pH on the photocatalytic activity, the efficiency of the Ce-doped samples was assessed under SSL irradiation both at the solution’s pH (pH 6, Figure S6) and in acidic conditions (pH 4, Figure S7).

All samples presented low adsorption during the dark period, except Ce05 at pH 4, which reached ~18% of its adsorption capacity. High adsorption in dark conditions is detrimental for photodegradation tests as it interferes with accurately determining the true photocatalytic activity of the material. When a too high percentage of the target pollutant adsorbs onto the catalyst surface prior to illumination, a significant portion of the pollutant may be removed through adsorption rather than the actual photodegradation process. Therefore, controlling and minimizing dark adsorption is essential to ensure that the degradation observed under light exposure is due to photocatalytic reactions, not simple physical adsorption.

The degradation efficiencies showed an increment when acidifying the reaction solution, confirming its influence on the efficiency. As a matter of fact, lowering the pH of the solution can enhance photodegradation efficiency due to several factors. In acidic conditions, the surface of TiO₂ becomes more positively charged (point of zero charge, PZC_TiO2 ≈ 6), promoting stronger electrostatic attraction with negatively charged metronidazole molecules (pk_a ≈ 2.6), which improves adsorption and facilitates subsequent photocatalytic reactions. In addition, in the case of Ce-doped TiO₂, the acidic environment can further stabilize the Ce³⁺/Ce⁴⁺ redox couple, enhancing charge separation and overall photocatalytic activity [18]. As a result, acidic conditions generally lead to higher degradation rates and more efficient pollutant removal.

The activity of the doped samples under the two different pH conditions is summarized in Figure 4. Ce1 was the best-performing sample, considering the average performance in the photodegradation reaction of MDZ under both UV and SSL irradiation, at pH 6 and pH 4. The kinetic evaluation of Ce-doped samples and bare TNF under all the conditions is reported in Figures S5 and S7, confirming the higher constants k for Ce1 under the different reaction conditions tested.

Figure 4. Comparison of degradation efficiencies for pH 6 and pH 4 solutions for Ce-doped TNF samples exposed to MDZ. Adsorption capacities for each catalyst (reported in gray) are included for reference.

Free-radical trapping experiments were carried out for Ce1 in order to determine the main active species responsible for degradation (Figure 5). This test plays a crucial role in photodegradation studies as it helps identify and confirm the reactive radical species generated during light-induced degradation processes, introducing radical scavengers into the reaction environment. Specifically, the photocatalytic tests were performed in the presence of the target molecule, maintaining the same operating conditions (atmospheric pressure, room temperature, pH 4) but adding oxalic acid (OA, 0.6 mM), tert-butanol (tBuOH, 1 mM), and N₂ (g) as holes, ·OH and ·O₂⁻ scavengers, respectively. The MDZ degradation is inhibited by adding N₂ and tBuOH to the system, while it is not affected by the addition of OA. Thus, the major active species in drug degradation are hydroxyl and superoxide radicals. Based on these considerations, we propose a plausible mechanism for photodegradation, as summarized in Figure 5b. It can be noticed that the degradation efficiency in the presence of OA is slightly higher than that of the bare MDZ solution, which may be attributed to a comparable amount of OA relative to that of the catalyst, suggesting full hole scavenging and an overall increase in the reactivity of the two other radical species.

Figure 5. (a) Scavenger experiments of the active species for the degradation of MDZ in presence of Ce1 at pH 4. (b) Proposed photodegradation mechanism of MDZ by means of Ce-doped TNF under simulated solar light irradiation at room temperature and atmospheric pressure.

Finally, the stability and reusability test was performed on the best-performing sample (Ce1) (Figure S8). The reusability test is an important step in photodegradation experiments as it evaluates the stability and recyclability of a photocatalyst over multiple uses. By repeatedly testing the same material under identical conditions, its photocatalytic efficiency can be assessed. Consistent performance across cycles indicates good reusability and long-term applicability, confirming that no structural changes or loss of active sites have occurred. As can be observed, the degradation efficiency of MDZ is stable, ranging from 56 to 52% from the first to the third cycle, suggesting a good stability of the synthesized photocatalyst after use.

3. Materials and Methods

3.1. Materials

The following commercial reagents were used in all the experimental phases without any further purification: deionized water (H₂O); titanium(IV) butoxide (TBOT, 97%, Merck, Darmstadt, Germany); oxalic acid (OA, H₂C₂O₄, 98%, Thermo Scientific, Waltham, MA, USA); cerium(III) nitrate hexahydrate (Ce(NO₃)₃ · 6H₂O, Acros Organics, Geel, Belgium); hydrochloric acid (HCl, ≥37%, Sigma-Aldrich, St. Louis, MO, USA); metronidazole (MDZ, Sigma-Aldrich, St. Louis, MO, USA); tert-Buthyl alcohol (t-BuOH, (CH₃)₃COH, ≥99%, Merck, Darmstadt, Germany); nitrogen gas (N₂).

3.2. Synthesis of Bare and Ce-Doped TiO₂ Nanoflowers

The synthesis of bare TiO₂ nanoflowers (TNF) followed a hydrothermal approach. TBOT (5.9 mmol) and OA (22 mmol) were added to deionized water, and the solution was stirred for 30 min until a transparent solution was obtained. The solution was transferred to a 100 mL Teflon-line and the reaction was carried out at two different temperatures, 200 and 150 °C, for 24 h. The sample was collected through centrifugation, washed several times with deionized water, and then dried at 60 °C overnight. The dried powder was calcined at 450 °C (ramp rate 5 °C/min) for 2 h.

The same procedure was used to obtain Ce-doped TNF, with doping carried out during the initial solution preparation and the hydrothermal reaction occurring at 150 °C. The lower temperature reaction condition (150 °C instead of 200 °C) was selected for the synthesis of doped samples to optimize energy efficiency and align with circular economy principles. More specifically, cerium(III) nitrate hexahydrate (Ce(NO₃)₃ · 6H₂O, solution 0.01 M) was used as the Ce precursor and added to the TBOT and OA solutions, varying the at% to determine the optimal concentration for photocatalytic performance. In particular, three different concentrations were selected, at 0.3, 0.5, and 1.0 at%, of Ce in TiO₂. The samples were referred to as TNFt for bare TiO₂ nanoflowers, where t stands for the reaction temperature, and Cex for Ce-doped TiO₂ nanoflowers, where x stands for the Ce content in at%.

3.3. Material Characterization

Field-emission scanning electron microscopy (FESEM) was performed with a FEI Magellan 400 FEG-SEM instrument (FEI, Hillsboro, OR, USA). All micrographs were obtained with the electron gun operated at an accelerating voltage of 3 kV. Elemental analysis was carried out by Energy Dispersive X-ray Spectroscopy (EDS), coupled with FESEM, to map the elemental distribution.

The porosity of pure THS was studied through N₂ physisorption at −196 °C, and the specific surface area of the nanoparticles was determined by applying the Brunauer–Emmett–Teller (BET) equation [19,20]. Micromeritics Tristar II plus (Micromeritics, Norcross, GA, USA) was used to obtain the isotherms and the corresponding value of surface area.

X-ray diffraction (XRD) measurements were performed on a PanAnalytical Empyrean diffractometer (Malvern PanAnalytical, Malvern, Worcestershire, UK) equipped with Cu Kα radiation. Data were collected over the 2θ range of 10–90° at a scan rate of 2° min⁻¹. Phase identification was achieved by comparing the diffraction patterns with those in the JCPDS, ICDD, and AMCSD databases.

Diffuse reflectance UV-visible-near-infrared (DRUV-Vis-NIR) spectra were recorded using a PerkinElmer Lambda 1050+ UV–Vis–NIR (Perkin Elmer, Waltham, MA, USA) spectrophotometer equipped with an integrating sphere, covering the wavelength range of 200–1500 nm. The Kubelka–Munk method was applied to estimate the band gap (Equation (1)) [21].

$$\mathrm{F}\left({\mathrm{R}}_{\infty }\right) ={\displaystyle \frac{{(1 - {\mathrm{R}}_{\infty })}^2}{2{\mathrm{R}}_{\infty }}}={\displaystyle \frac{\mathrm{K}}{\mathrm{S}}} ,$$

(1)

Here, F(R_∞) denotes the Kubelka–Munk function, R_∞ is the diffuse reflectance, K represents the absorption coefficient, and S corresponds to the scattering coefficient. The band gap energy evaluation for TiO₂ semiconductors can be expressed using Equation (2), exploiting the Tauc method [22]:

$${(\mathrm{F}\left({\mathrm{R}}_{\infty }\right)\mathrm{h}\nu )}^{{\displaystyle \frac{1}{2}}} = \mathrm{B}(\mathrm{h}\nu -{\mathrm{E}}_{\mathrm{g}}),$$

(2)

where h represents Planck’s constant, ν is the frequency of the light, and B denotes a constant. A Tauc plot is generated by plotting F(R_∞)^1/2 as a function of photon energy (eV). The band gap energy of the sample can then be estimated from the x-intercept of the linear portion of the Tauc plot [22].

Steady-State and Time-Correlated Single-Photon Counting (TCSPC) measurements were performed exciting with a vertically polarized 370 nm ps-pulsed laser (PicoQuant, Berlin, Germany). Single decays were acquired at 450 nm. The integration time was 900 s for all samples. The analysis of time-resolved data was performed with FluoFit v.4.6 software from PicoQuant. All decays were fitted with a mono- or bi-exponential reconvolution model, including the IRF (instrument response function). The intensity-averaged decay time <τ> was calculated as the intensity-weighted average decay time at the selected emission wavelength [23].

3.4. Photocatalytic Evaluation

The photocatalytic performance of the synthesized materials was assessed by monitoring the degradation of an aqueous solution of the drug metronidazole (MDZ), selected as a model pollutant. All photocatalytic tests were carried out under both UV light and simulated solar light (SSL) irradiation, at atmospheric pressure (P_atm) and room temperature (r.t.). The test was carried out both at the solutions’ pH (pH~6) and at a lower pH (pH~4), adding HCl 0.1 M. The UV photocatalytic reactions took place in a reactor system using a 125 W high-pressure mercury lamp, operating at wavelengths between 180 and 420 nm with a maximum peak at 366 nm and at a distance from the reactor system of 5 cm. The photocatalytic experiments were performed using a Sunlight Solar Simulator equipped with an AM 1.5G filter and a 100 W xenon arc lamp (Model 10500, Low Cost Solar Simulator, Abet Technologies, Milford, CT, USA). During irradiation, the levels of solar simulated irradiance were set at 153 mA. The photon flux was calibrated using a Si-based reference solar cell (#15150, Abet Technologies, Milford, CT, USA) placed against the outer wall of the photoreactor filled with pure water. A 100 mL cylindrical concentric Pyrex–quartz reactor was used for the tests. The starting concentration of the target molecule was 8.56 mg/L. The amount of photocatalyst was fixed at 62.5 mg/L (5 mg of catalyst in 80 mL of solution), as prior studies have shown that this amount achieves optimal photodegradation efficiency [12]. Before irradiation, all the targeting solutions were kept in dark conditions for 60 min to establish the adsorption/desorption equilibrium. At given time intervals (−60, 5, 10, 20, 30, 40, 60, 120, and 180 min for UV and −60, 0, 5, 10, 20, 30, 40, 60, 120, 180, 240, 300, and 360 min for solar simulator analyses), ~1 mL of solution was collected separating the catalyst from the system by centrifugation (centrifuge 5415D, Eppendorf, Milan, Italy) and analyzed through UV–Vis spectroscopy (Perkin Elmer Lambda 1050+ UV-Vis-NIR spectrophotometer, Perkin Elmer, Waltham, MA, USA). The development of the photo-oxidation reaction and degradation efficiency was monitored by following the maximum absorbance of MDZ, at 319 nm. After determining the concentration at the selected wavelength using the Lambert–Beer law, the degradation efficiency was evaluated by calculating the ratio C/C₀, where C is the concentration after the monitored time t and C₀ is the initial concentration of the solution in the absence of catalyst (−60 min sample). Kinetic analyses of the photodegradation process (0–360 min) were performed by applying a pseudo-first-order reaction model (Equation (3)):

$$\ln {\displaystyle \frac{\mathrm{C}}{{\mathrm{C}}_0}} =-\mathrm{kt},$$

(3)

where k represents the pseudo-first-order rate constant (min⁻¹), calculated by plotting −ln(C/C₀) versus t (min) and considering the slope of the linear fitted data.

To assess the stability and reusability of the photocatalysts, a three-cycle reusability test using the best-performing sample (Ce1) in the aqueous solution was performed. Ce1 was collected by centrifugation, washed with deionized water several times, and dried overnight after each photocatalytic cycle. The main active species in the photodegradation of the mixture were investigated by performing free-radical trapping experiments. Briefly, three different photocatalytic tests were carried out using the same conditions as before but adding tert-Butanol (tBuOH, 1 mM), oxalic acid (OA, 0.6 mM), and N₂ (g) as ·OH-, holes, and ·O₂⁻ scavengers, respectively, to the reaction system.

4. Conclusions

In this work, a hydrothermal method was selected to obtain bare TiO₂ nanoflowers (TNF) and Ce-doped TiO₂ nanoflowers (Cex). The content of the dopant was optimized by testing different Ce atomic percentages (0.3, 0.5, and 1 at%) to maximize the efficiency of the photocatalytic process. SEM and XRD analyses confirmed the morphology and the transition from rutile phase to brookite, typical of Ce-doped TiO₂ materials. Additionally, the appearance of CeO₂ was noticed. EDX analysis validated the presence of homogeneously dispersed Ce-ions in the Cex samples, while the amount of Ce increased with higher doping. These TNF/Cex were evaluated as photocatalysts for metronidazole (MDZ) photodegradation in water under UV radiation and simulated solar-light irradiation at room temperature and atmospheric pressure. Moreover, the efficiency of Ce-doped samples under SSL irradiation was assessed at both solution pH and pH 4. The best-performing sample was Ce1, showing the highest degradation efficiency at both pH, UV and SSL, and the lowest adsorption in the dark period. Thus, the combination of morphology and doping engineering improved the photo-response of Ce-doped TNF (Cex), where the highest doping (1 at%) induced a decrease in the band gap compared to the bare sample, also visible from the redshift of the main absorption edge. Steady-state and TCSPC measurements confirmed that the presence of Ce ions reduces charge carrier recombination, likely due to the formation of Ce-induced trap states that promote non-radiative pathways and charge separation, thus improving carrier lifetimes and enhancing the probability of surface redox reactions. Thanks to free-radical trapping experiments performed on the Ce1 sample, we demonstrated that the main active species in MDZ photodegradation are ·OH and ·O₂⁻ radicals. Furthermore, the reusability test performed on Ce1 proved the very good stability of Ce-doped TNF after three cycles of MDZ photodegradation under SSL irradiation. Overall, these results demonstrate that Ce-doped TiO₂ nanoflowers, particularly at 1 at% doping, are efficient, stable, and reusable photocatalysts, making them promising candidates for environmental remediation applications.