In recent years, efforts have been carried out to improve the elimination of pharmaceutical compounds found in various water sources worldwide, including water inlets and outlets, streams, rivers, and groundwater. Pharmaceutical pollutants, persistent in aquatic environments, can have harmful effects on the human health, including kidney problems, skin diseases, and poisoning (Margot et al., 2015; Petrie et al., 2015). Non-steroidal anti-inflammatory drugs (NSAIDs) have gained particular attention as an often-detected class of pollutants because of their stability and resistance. NSAIDs discharge to the environment can originate from hospitals, pharmaceutical industry effluents, and domestic wastewater, which results from their frequent use, over-the-counter accessibility, and high excretion rates (Memmert et al., 2013; Simon and Evan Prince, 2017).

Many solutions have been proposed to reduce or eliminate organic pollutants in wastewater and improve its quality. Advanced oxidation processes (AOPs) have emerged as a viable option for treating a wide range of emerging contaminants, including pharmaceuticals (Stasinakis, 2008; Miklos et al., 2018; Anjali and Shanthakumar, 2019). The efficacy of AOPs is determined by the generation of reactive oxygen species, primarily hydroxyl radicals (OH•). Due to their non-selective nature, OH• can degrade various organic compounds present in water and wastewater, forming carbon dioxide, water, and mineral acids (Legrini et al., 1993). The most used AOPs include ozone-based processes (Abromaitis et al., 2022), Fenton reaction (Im et al., 2015a), advanced oxidation with persulfate (Ge et al., 2021), and photocatalysis (Ye et al., 2018). Among these, the photocatalytic process using a semiconductor material has been extensively researched for the elimination, degradation, and mineralization of organic compounds in aqueous systems (Li and Li, 2001; Choi et al., 2007).

One of the most widely investigated semiconductors is TiO₂, a very efficient photocatalyst used for the degradation of pharmaceutical contaminants in water (Mills et al., 1993; Mills and Le Hunte, 1997; Hashimoto et al., 2005; Ni et al., 2007). The crystallinity of TiO₂ plays a crucial role in photocatalysis (Tanaka et al., 1991; Li et al., 2012). For instance, the anatase phase exhibits superior activity when exposed to UV illumination, while the presence of mixed anatase-rutile crystalline phases has been associated with photoactivity at longer wavelengths in the visible light region (Collins-Martinez et al., 2007; Solís-Casados et al., 2017).

The application of TiO₂ in the photocatalytic oxidative remediation of pharmaceutically polluted water has already been reported (Molinari et al., 2006; Coleman et al., 2007). Different shapes of TiO₂ such as nanoparticles (Kanakaraju et al., 2015), nanofibers (Doh et al., 2008), and nanotubes (Albu et al., 2007; Macak et al., 2007; García-Valverde et al., 2014) can be used as a photocatalyst. The most popular ways to synthesize TiO₂-based nanomaterial include sol-gel (Akpan and Hameed, 2010), solsolvothermal (Zhou et al., 2010), hydrothermal (García-Valverde et al., 2014; Kasuga et al., 1999), and anodization (Macak et al., 2008). TiO₂ nanotube (TiO₂ NT) layers obtained via anodization have been particularly useful for the degradation of pharmaceuticals (Im et al., 2015a; Qian et al., 2021).

Among the NSAIDs, naproxen (NPX), which is widely prescribed for skeleton-muscle pain or inflammatory rheumatic disorders, is one of the most frequently detected pharmaceutical residues in water bodies (Todd and Clissold, 1990; Stovitz and Johnson, 2003). TiO₂ nanoparticles and TiO₂ nanotube (TiO₂ NT) layers have been reported in the elimination of NPX from various wastewater matrices (Im et al., 2015a; Kanakaraju et al., 2015). Different shapes of TiO₂ unmodified (Méndez-Arriaga et al., 2008b, 2008a; Jallouli et al., 2016a; Kanakaraju et al., 2016a) and modified for instance with Ce (Hao et al., 2023), MoS₂ (Sheydaei et al., 2022), and Cu-S (Amini et al., 2020) in the removal of different drugs show a positive impact on the degradation rates as shown in Supplementary Tables S1, S2. However, only one publication focused on TiO₂ NT layers as a catalyst using AOPs to determine their effectiveness in degrading NPX under ultrasonic radiation. In that study was demonstrated a remarkable 96.0% improvement in NPX degradation efficiency when TiO₂ NT layers were present, as opposed to their absence (Im et al., 2015).

To further enhance the photocatalytic performance of TiO₂-based nanomaterials and also to extend the light absorption in the visible spectral region, Fe₃O₄ nanoparticles (NPs) were incorporated into TiO₂-based nanomaterials (Liu et al., 2016; Bi et al., 2019; Yilmaz et al., 2020). Apart from the more favorable light absorption, this enhancement can be attributed to the efficient separation of h⁺ and e⁻, which occurs by reducing iron ions into Fe²⁺ or Fe⁰ (Li et al., 2017). Additionally, whenever a magnetic field can be applied to the material or system used, this modification facilitates and accelerates the separation of the liquid phase from the solid photocatalyst, using an external magnet, without the need for centrifugation or filtration. Recently, it was shown that TiO₂ single-tube (TiO₂ ST-NT) powders, derived from TiO₂ NT layers, present an excellent photocatalytic performance under UV light using methylene blue as a model dye (Beketova et al., 2020). The TiO₂ ST-NT powders were synthesized by anodization, followed by an etching process, and modification with Fe₃O₄ (magnetite) NPs, resulting in TiO₂ nanotube powders modified with Fe₃O₄ NPs (TiO₂ ST-NT@Fe₃O₄NPs) as a guidable photocatalyst (Beketova et al., 2020). The effect of the annealing temperature of TiO₂ ST-NT and TiO₂ ST-NT@Fe₃O₄NPs powders on the photocatalytic performance was evaluated, revealing that the best results were obtained after annealing at 600°C (Sepúlveda et al., 2023).

To the best of our knowledge, this study is the first to report on the photocatalytic performance of TiO₂ ST-NT and TiO₂ ST-NT@Fe₃O₄NPs powders for the degradation of NPX under visible light of different wavelengths. The aim of this study was to i) obtain hybrid photocatalytic materials based on TiO₂ ST-NT powders and Fe₃O₄ NPs, and ii) assess their photocatalytic activity. The TiO₂ ST-NT and TiO₂ ST-NT@Fe₃O₄NPs powders annealed at 600°C were tested for the photocatalytic degradation of NPX using visible light: the blue light (λ_max = 425 nm), the green light (λ_max = 525 nm), and the white light. The kinetics of degradation and the possible NPX degradation products were also discussed.

The X-ray diffraction (XRD) patterns, extensive scanning electron microscopy (SEM) evidence, specific surface areas, high-resolution scanning transmission electron microscopy (HRTEM-STEM-EDX) elemental maps exhibiting the distribution of Ti, Fe, and O, and pore size distributions of TiO₂ ST-NT and TiO₂ ST-NT@Fe₃O₄NPs powders annealed at 600°C have been reported in our previous work (Sepúlveda et al., 2023). To provide a new insight into the materials used, Figure 1 shows TEM images of TiO₂ ST-NT and TiO₂ ST-NT@Fe₃O₄NPs powders, both annealed at 600°C. As shown in Figure 1B, the surface of TiO₂ ST-NT@Fe₃O₄NPs powders exhibits an uniform distribution of Fe₃O₄ NPs on the surface of TiO₂ ST-NT powders.

The surface elemental composition of the TiO₂ ST-NT and TiO₂ ST-NT@Fe₃O₄NPs powders before the photocatalytic measurements was obtained by XPS analysis and it is shown in Supplementary Table S3. The TiO₂ ST-NT@Fe₃O₄NPs powders show a significantly increased amount of C compared to the TiO₂ ST-NT powders due to the Fe₃O₄ NPs modification using an oleic acid process approach. The high-resolution Ti 2p, O 1s, and C 1s XPS spectra of TiO₂ ST-NT powders are shown in Supplementary Figure S1. The Ti 2p spectrum exhibits its spin-orbit splitting doublet Ti 2p_3/2 and Ti 2p_1/2 fitted using six components that correspond to three different chemical species. The first doublet (red line) was assigned to Ti⁴⁺ from the TiO₂ (peaks centered at ∼458.7 and 464.4 eV) (Hoyos et al., 2017; Sopha et al., 2020). The second doublet (blue line) is related to Ti³⁺ in the TiO₂ lattice (peaks located at 457.5 and 463.2 eV) (Hoyos et al., 2017; Sopha et al., 2023). The third doublet (olive line) corresponds to non-stoichiometric TiO_x (peaks at 459.9 and 465.6 eV) (Sopha et al., 2020). The O 1s spectrum reveals the presence of four chemical states. The red peak was assigned to O^2- or TiO₂ at ∼529.9 eV (Hoyos et al., 2017; Sopha et al., 2020). The blue peak is related to –OH at ∼530.9 eV (Wu et al., 2017; Sopha et al., 2020), the olive peak corresponds to C-O species at 532.0 eV (Rouxhet and Genet, 2011) and the orange peak evidences the presence of (C=O)-OH at ∼533.1 eV (Rouxhet and Genet, 2011). The fitting of the C 1s spectrum was carried out with three different chemical states, adventitious carbon at ∼284.8 eV, C-O species at ∼286.4 eV and (C=O)-OH at ∼288.8 eV (Rouxhet and Genet, 2011).

Supplementary Figure S2 displays the Ti 2p, Fe 2p, O 1s, and C 1s XPS spectra of the TiO₂ ST-NT@Fe₃O₄NPs powders before the photocatalytic measurements. The fitting of the Ti 2p spectrum revealed one doublet suggesting the presence of one titanium oxidation state, Ti⁴⁺ (peaks located at ∼458.6 and ∼464.3 eV). In the case of the Fe 2p spectra its corresponding spin-orbit splitting Fe 2p_3/2 and Fe 2p_1/2 is evidenced. The peak fitting suggests the presence of two different oxidation states, in line with the expected states of the Fe₃O₄ NPs. The oxidation states for iron were Fe²⁺ (peaks at ∼710.3 and ∼723.9 eV), and Fe³⁺ (peaks at ∼712.3 and ∼725.9 eV) (Tang et al., 2018; Ai et al., 2019). O 1s and C 1s spectra reveal the same chemical species as the TiO₂ ST-NT powders, with the difference in the intensity and a small shifting in the binding energy ∼0.4 eV of –OH, C-O and (C=O)-OH species in the O 1s signal.

The photocatalytic performance of the powders was assessed in a series of NPX degradation tests, carried out using light sources of different main wavelengths. Experiments were conducted using either LED lamps emitting either blue (λ_max = 425 nm) or green (λ_max = 525 nm) light or using a solar simulator emitting white light. NPX degradation was observed in each experiment. Moreover, three main NPX transformation products were detected and monitored throughout the experiments: P1, P2, and P3 (marked according to their detection order, Supplementary Figure S3).

Figure 2A shows the photocatalytic performance of TiO₂ ST-NT powders activated using the blue light illumination. A 60% reduction of NPX initial concentration was noted within 3 h. At the same time, the concentration of all three transformation products, P1, P2, and P3, increased gradually. In the case of TiO₂ ST-NT@Fe₃O₄NPs powders, 92% of NPX was degraded within 3 h (Figure 2B). As for the NPX transformation products, only P1 and P3 appeared – P2 was not formed. At the end of the experiment, the levels of both P1 and P3 were higher than when unmodified ST-NT powder was used. Therefore, modification with Fe₃O₄ NPs enhanced the photocatalytic activity of the TiO₂ ST-NT powders under the blue light illumination.

Upon excitation of the powders with the green light illumination, only 20% NPX degradation was noted in both experiments – when either TiO₂ ST-NT or TiO₂ ST-NT@ Fe₃O₄NPs powders were used. In both tests, the NPX transformation products were detected at a very low level (∼1%). Figure 3A shows that all three transformation products appeared when TiO₂ ST-NT powders were used, whereas when TiO₂ ST-NT@ Fe₃O₄NPs powders were used, P2 was not detected (Figure 3B). Noteworthy, P2 was not found during both experiments using Fe₃O₄-modified materials, even if these photocatalysts were more active than bare TiO₂ ST-NTs. There are two possible explanations for this result. First, P2 could be adsorbed on the surface of TiO₂ ST-NT@Fe₃O₄NPs and removed via filtration prior to the HPLC analysis. Adsorption of various pharmaceuticals on iron oxides has already been described (Olusegun et al., 2023). Therefore, it is likely that in the present work, such a phenomenon exists between iron oxides and transformation products of pharmaceuticals, too. Alternatively, when Fe₃O₄-decorated powders were used, P2 could be degraded via photo-Fenton reactions typical for iron oxide-based materials, and thus it was not detected during the analysis (Punjabi et al., 2018; Olusegun et al., 2023). Two photodegradation mechanisms should be considered for iron oxides – regular excitation of the semiconducting materials and photo-dissolution of iron oxides – and P2 could be more susceptible to the latter, which led to its quick elimination from the system. It is also worth noting that the blue and green LED lamps had the same power. Consequently, the irradiance values differed because they depended on the irradiation wavelength. Although most studies on photocatalytic degradation provide little information on the light source used and this detail is rarely pointed out, it should be considered that differences in irradiance values can affect the photocatalytic degradation rates (Eskandarian et al., 2016; Zaveri et al., 2018).

Among several works on the photocatalytic degradation of pharmaceutical contaminants using TiO₂ combined with Fe-oxides (Amini et al., 2020; Pena-Velasco et al., 2021; Krakowiak et al., 2022) only a few reports show efficient removal rates under the visible light illumination (Sheydaei et al., 2022; Hao et al., 2023). High efficiency is often achieved when other processes, such as ozonation or photoelectrocatalysis, are additionally applied (Amini et al., 2020; Sheydaei et al., 2022). In the present study, we determined the influence of the visible irradiation of a given main wavelength, emitted from a low-power lamp on the TiO₂ ST-NT@Fe₃O₄NPs and noted comparable results. This indicates that the composites display high activity upon excitation with the visible light illumination. Recent studies on photocatalytic removal of commonly used medicines using TiO₂-iron oxide materials and visible light are listed in Supplementary Table S1.

Furthermore, the activity of TiO₂ ST-NT powders upon simulated sunlight excitation was assessed. In this case, PhotoCube™ was used – a reactor equipped with four LED panels emitting white light. 90% of NPX was degraded within 1 h, and a complete removal occurred within 2 h. Figure 4 shows that all three transformation products were detected during the photocatalytic experiment. Noteworthy, at the end of the experiment, not only NPX but also P2 were completely degraded.

NPX degradation under the white light (i.e., using simulated sunlight) occurred faster than under the other light sources, as compared in Table 1 (photocatalytic rate constants) and plotted in Supplementary Figure S3 (degradation curves). It is important to bear in mind that the tests conducted under the blue light and the white light (Figures 2–4) cannot be directly compared due to the different experimental setups and power of the illumination sources (Musial et al., 2023). Nevertheless, these results confirmed the high activity of unmodified TiO₂ ST-NT powder upon the white light illumination, comparable or higher to NPX degradation efficiencies shown in other photocatalytic studies, listed in Supplementary Table S2 (Uheida et al., 2019; Eslami et al., 2020; Kowalkińska et al., 2023). For instance, Eslami et al. achieved complete NPX removal within 2 h, using polycarbonate coated with N- and S-doped TiO₂, but employed a high-power (350 W) Xe lamp (Eslami et al., 2020). Similar results under comparable conditions (Xe lamp of 125 W) – complete degradation within 1.5 h – were reported by Uheida et al. (2019), using TiO₂ nanoparticles immobilized on polyacrylonitrile/multiwall carbon nanotubes composite nanofiber. However, in this study, only irradiation of a wavelength higher than 420 nm was used, whereas in the present work simulated sunlight conditions also include the participation of the UV-range irradiation in the solar light spectrum. Experiments using TiO₂ ST-NT@Fe₃O₄NPs powders and white light were not conducted due to technical obstacles, because the built-in magnetic stirring precluded using TiO₂ ST-NT@Fe₃O₄NPs powders without an unwanted influence on the degradation rates.

Previous studies have demonstrated that the two main NPX degradation routes are decarboxylation and hydroxylation (demethylation) as one can see in Figure 5 (Kanakaraju et al., 2016; Changanaqui et al., 2020; Jung et al., 2020). Noteworthy, these two routes can occur simultaneously or successively (Méndez-Arriaga et al., 2008b; Fan et al., 2019; Kowalkińska et al., 2023). Another route, dimerization, that leads to the formation of more complex compounds, was also described (Kowalkińska et al., 2023). Interestingly, the dimer formation was reported to depend on the facets exposition of the fluorinated TiO₂. The results of our analyses match previous reports and considerations. During high-performance liquid chromatography with photodiode-array detection (HPLC-DAD) analyses, it was noted that P1 was eluted earlier than NPX and both P2 and P3 later than NPX. This suggests that P1, is a more polar compound than NPX and is a product of NPX hydroxylation, whereas P2 and P3 are products of its decarboxylation. Furthermore, during the NPX degradation experiment using TiO₂ ST-NT powders and simulated sunlight, P2 was completely removed at the end of the experiment. This proves that not only the main tested contaminant but also its transformation products are susceptible to photocatalytic degradation using TiO₂ nanotube layers (Jallouli et al., 2016a).

Prolonged irradiation of the photocatalytic mixture may lead to the complete elimination of the drug transformation products. This is an important observation, as recent research on the toxicity of pharmaceutical pollutants focuses not only on the parent compound but also on the transformation products (Maculewicz et al., 2022). In vitro assays using bioluminescent bacteria (Allivibrio fischeri), showed that the transformation products displayed higher potential toxicity than the parent compound, but the toxicity of the whole post-degradation mixture was not considered as a major concern (Cazzaniga et al., 2020). In addition, phototransformation products of lower molecular weight were found to be more active towards A. fisheri. However, not all the NPX transformation products displayed higher toxicity than the parent compound, which was a proof of the stereostructure-activity relationship (DellaGreca et al., 2003). Other bioassays performed on algae, rotifers, and microcrustaceans proved that phototransformation products displayed higher acute and chronic toxicity than NPX. However, no genotoxic or mutagenic effects were found (Isidori et al., 2005). Moreover, mixtures of NPX and its transformation products proved to be more toxic toward southern toads (Anaxyrus terrestris) than NPX alone (Cory et al., 2019). In this regard the complete degradation of pharmaceuticals along with their transformation products during water treatment is necessary. Toxicity assessment of the wastewater effluents should be considered in the design of AOPs-based water remediation systems. All in all, as highly effective photocatalytic materials, TiO₂ ST-NTs powders hold great promise for successful implementation into wastewater treatment procedures.

TiO₂ ST-NT and TiO₂ ST-NT@Fe₃O₄NPs powders annealed at 600°C were used as catalysts for the degradation of naproxen (NPX) using the blue, green, or white light, for the first time. The photocatalytic performance of TiO₂ ST-NT and TiO₂ ST-NT@ Fe₃O₄NPs powders examined under the blue light showed a 60% and 92% reduction in NPX initial concentration within 3 h, respectively. This clearly indicates that the incorporation of Fe₃O₄ NPs enhanced the photocatalytic activity of TiO₂ ST-NT powders under the blue light. Under the green light, both TiO₂ ST-NT and TiO₂ ST-NT@Fe₃O₄4NPs powders only exhibited a 20% NPX degradation. Complete removal of NPX and one of its transformation products were observed in the experiment conducted using TiO₂ ST-NT powders and white light. Remarkably, 94% of NPX degraded within only 1 h. These findings demonstrate the immense potential of TiO₂ ST-NT powders for successful implementation in wastewater treatment.