Removal of Contaminants of Emerging Concern from a Wastewater Effluent by Solar-Driven Heterogeneous Photocatalysis: A Case Study of Pharmaceuticals

The occurrence of emerging micropollutants (pharmaceuticals, pesticides, personal care products, industrial compounds, etc.) in the environment is considered a major threat to human health and aquatic ecosystems. These micropollutants enter the environment through anthropogenic actions and have been identified in surface, ground, waste, and even in drinking water, in quantities ranging from ng L−1 to µg L−1. Currently, the pollution of the global water cycle with persistent organic pollutants remains one of the major challenges of the twenty-first century. Most of these organic substances are only partially removed by conventional wastewater treatment plants. Particularly, considerable amounts of pharmaceuticals are used in human and veterinary medicine, which are not efficiently removed during conventional wastewater treatments and subsequently continuously enter freshwater systems and even agricultural crops. Accordingly, we have evaluated the effectivity of TiO2 as a photocatalyst in tandem with Na2S2O8 as an oxidant for the treatment of a wastewater effluent polluted with pharmaceutical (atenolol, carbamazepine, clarithromycin, erythromycin, irbesartan, and ketoprofen) residues. Results show that the use of solar heterogeneous photocatalysis by means of band-gap semiconductor materials, especially TiO2 in combination with a strong oxidant such as Na2S2O8, significantly enhances their disappearance from the wastewater effluent. However, the selected pharmaceuticals show a slow degradation in wastewater effluent compared to pure water indicating that the occurrence of dissolved salts and organic carbon in wastewater effluent noticeably slows down the efficiency of the treatment. A single first-order model satisfactorily explains the photocatalytic degradation of the compounds studied for both, pure and wastewater. In the case of wastewater effluent, the highest DT50 values were observed for macrolides (13 and 16 min for erythromycin and clarithromycin, respectively), while the other compounds studied showed DT50 values below 10 min. This methodology has a notorious interest in some areas of the Mediterranean basin with water shortage, such as SE of Spain, where more than 3000 h of sunlight per year are recovered.

pollutants (EPs), also called contaminants of emerging concern (CECs), in different aquatic environments, including wastewater (WW) (Gogoi et al., 2018;Noguera-Oviedo & Aga, 2016;Tang et al., 2019). CECs, including their transformation products, are chemicals without regulatory evaluation and whose effects on the environment and humans are poorly understood (Deblonde et al., 2011). The occurrence of these pollutants (mostly of organic in nature) and their harmful impact on aquatic and terrestrial ecosystems, as well as on human health, is now a matter of concern among the governmental institutions, scientific community, and citizens. CECs are not necessarily new substances. These compounds are present for a long time in environmental compartments, although their presence and involvement are currently being clarified. Despite guidelines established by national and local authorities, unregulated discharges often occur due to the lack of specific legislation and ecotoxicity data of CECs (Parida et al., 2021). Many of them are resistant to conventional treatments or have slow biodegradation rates (Choi et al., 2017). Uncontrolled release of CECs into the aquatic environment causes problems such as high persistence, ecotoxicity, and likely harm to human and animal health. Recent scientific studies and reports on WW composition have focussed attention on the presence in the environment of many chemicals derived from anthropogenic activities from urban, agricultural, and industrial sources (Pal et al., 2010;Deblonde et al., 2011;Archer et al., 2017;Wilkinson et al., 2017). The global occurrence of CECs in developed and developing countries shows the urgent need to address this type of contaminants. A compilation of worldwide occurrence of CECs detected in the environment has been recently published by Lee et al. (2021).
Although CECs are often present in the environment at low levels ranging from ng L −1 or µg L −1 , it is still unclear whether their concentrations in the aquatic environment can cause endocrine disruption in wildlife and humans. A lot of them are persistent in water, putting pressure on the wastewater treatment plants (WWTPs) for their operative removal (Archer et al., 2017). According to the NORMAN database, several thousand chemicals from 21 types have been identified in the European aquatic environment in recent decades (Norman, 2022). In addition, based on the number of chemical compounds listed by the European Chemical Agency, between 3 × 10 4 and 5 × 10 4 industrial chemicals are in daily use, and a large number are catalogued as potential CECs owing to their release into the environment (ECHA, 2022). Among them, pesticides also called plant protection products (herbicides, insecticides), personal care products (fragrances, parabens, UV filters), pharmaceuticals (analgesics, antibiotics, legal drugs, betablockers, steroids), drugs of abuse (cannabinoids, opioids) life-style compounds (nicotine, caffeine), or industrial additives and by-products (chlorinated solvents, plasticizers, polyaromatic hydrocarbons) among others have been found in WW worldwide over the last years (Tang et al., 2019).
Concretely, a pharmaceutical (PhMC), also called a medicine, pharmaceutical product, medication, or medicinal product, is "any substance or combination of substances which may be used in or administered to human beings either with a view to restoring, correcting or modifying physiological functions by exerting a pharmacological, immunological or metabolic action, or to making a medical diagnosis" (Vogler & Zimmermann, 2016). PhMCs include the active substance (a substance that, alone or in combination with one or more ingredients, is considered to fulfill the intended activity of a medicinal product) and excipients (substances, different from the active substance, that have been adequately evaluated for safety and are included in a medicine delivery system to protect, support, or enhance stability, bioavailability, or patient acceptability, among other functions). Many of them are regularly detected in high levels in the aquatic environment, possibly due to their unceasing release from WWTPs, which is meaningfully faster than their elimination rates. Consequently, they are considered as a group of pseudo-persistent contaminants with slow transformation rate, as pointed out by laboratory and field studies (Arnold et al., 2013;Bu et al., 2016). Even though PhMCs have been ubiquitous in aqueous media for a long time, the levels found in the environment have only recently begun to be monitored and recognized as potentially harmful to environmental ecosystems. This is due to the advance of new analytical techniques (mainly LC-MS/MS) that allow the identification and quantitation at very low levels (ng L −1 ) of these compounds in waste, surface, and groundwater (Rivera-Utrilla et al., 2013). PhMCs have become an important public health issue as environmental pollutants over the last years (Kummerer, 2010). After ingestion, PhMCs are partially excreted unchanged and may subsequently reach the WWTPs via the sewer network.
Water shortage and the fitful terrestrial distribution of rainfall are disturbing concerns in arid and semiarid areas, where water management strategies promote the reuse of WW effluents for agricultural purposes because of climate change. Some papers have recently reviewed the occurrence and fate of a large number and variety of CECs likely to occur in agroecosystems (Boxall, 2012;Snow et al., 2020). Thus, PhMCs can enter in the agricultural environment directly (via therapeutic use in livestock and fisheries), as well as indirectly, through the cumulative use of reclaimed WW and the application of municipal biosolids and manure. Currently, WW is reused worldwide, particularly in semiarid areas like SE of Spain. Bearing in mind the extensive variety of CECs (many of them still unassessed) entering WWTPs, many compounds (transformed or not) can end up in agricultural soils with the consequent risk of plant intake. For this reason, the EU has focused on this issue with the reconsideration of the minimum requirements for water reuse according to Regulation 2020/741/EU (EC, 2020). The purpose of this regulation is to guarantee that reclaimed water is harmless for agricultural reuse, supporting resilience to climate change, promoting the circular economy, and contributing to the objectives of the Water Framework Directive (EC, 2000) by controlling water scarcity and the consequent pressure on water resources. Regarding this initiative, PhMCs are highlighted in the EU Strategic Approach on Pharmaceuticals in the Environment (EC, 2019).
In many cases, the conventional (biological) treatments are ineffective for their removal (Michael et al., 2013;Pérez-Lucas et al., 2022a). Consequently, there is a clear requirement to curb this problem by means of innovative and environmentally friendly technologies established in WWTPs to effectively remove these CECs. Membrane technology (nano, ultra, and microfiltration, reverse osmosis, adsorption dynamics on carbon nanotubes or graphene oxide, aerobic granular sludge, or gravity-driven) has been extensively applied during the last years to isolate micropollutants from wastewater in reactors (closed system) and in natural water system (open system). However, during its long-term process, the pollutants gradually accumulate into the adsorption materials, until they reached saturation, then it became inactive (Campo et al., 2021;Chen et al., 2019;Luo et al., 2022Luo et al., , 2023Pronk et al., 2019), so that it is usually combined with ozonation, activated carbon, photodegradation, etc. (Vasilachi et al., 2021). In this context, relatively new treatment methods such as light-driven advanced oxidation processes (AOPs) need to be addressed (Rivera-Utrilla et al., 2013;Teixeira et al., 2015;He et al., 2016;Awfa et al., 2018;Almomani et al., 2018;Kanakaraju et al., 2018, Vasilachi et al., 2021. Lightbased processes can be categorized into (i) UV/oxidant (H 2 O 2 , S 2 O 8 2− ), (ii) UV/O 3 , (iii) photo-Fenton (Fe 2+ ) or photo-Fenton like (Fe 3+ ), and (iv) heterogeneous photocatalysis (with semiconductor materials, SC), each of which has a range of UV wavelengths and factors affecting their operation. Heterogeneous photocatalysis has special interest, which consists of the acceleration of a chemical reaction (photoreaction) by the action of a photocatalyst involving the combination of photochemistry and catalysis (Augugliaro et al., 2019). Both light (direct photolysis) and photocatalyst are needed to accelerate a chemical reaction. PhMCs are oxidized by highly reactive oxidant species (ROS), mainly hydroxyl radicals (HO • , E 0 = 1.9-2.7 V vs. normal hydrogen electrode, NHE) including others such as superoxide anion (O 2 •− ) and hydridodioxygen (HO 2 • ). Other radicals such as sulfate radical anion (SO 4 •− , E 0 = 2.6-3.1 V vs. NHE) may also be implicated when using persulfate (S 2 O 8 2− ) to avoid electron (e − )/hole (h + ) recombination . Commonly, light-driven heterogeneous photocatalysis is affected by different factors such as pH, temperature, UV/Visible absorbance and source, nature of catalyst and loading, nature and concentration of pollutant, and water matrix components (Malato et al., 2009). The impact of water composition especially affects the performance of light-driven processes because dissolved organic and inorganic compounds can adsorb on the SC surface reducing the number of active sites on it, can inhibit light penetration, and can act as ROS scavengers creating less powerful oxidants as is the case for some anions such as HCO 3 − , SO 4 2− , and/or Cl − (Pérez-Lucas et al., 2022b;Ribeiro et al., 2019).
The main advantage of these processes is that they attain the removal or at least the reduction of PhMCs by mineralization, rather than transferring them from one place to another as with conventional processes (Miklos et al., 2018). Among various SCs, TiO 2 is the most widely used for water treatment, mainly due to its non-toxicity, photo-stability, corrosion resistance, availability, biological and chemical inertness, and chemical and thermal stability over a wide pH range (Kanakaraju et al., 2014). With this aim, we have assessed the effectivity of TiO 2 as photocatalyst in combination with Na 2 S 2 O 8 as oxidant under natural sunlight for the treatment of a wastewater effluent (WWe) polluted with six PhMCs, atenolol and irbesartan (anti-hypertensives), clarithromycin and erythromycin (antibiotics), carbamazepine (anti-epileptic), and ketoprofen (anti-inflammatory) commonly used worldwide.

Experimental Setup
Photocatalytic tests were performed in Pyrex glass vessels (110 mm length × 80 mm i.d.) exposed to sunlight during September 2021 in Murcia, SE Spain. In all cases, 500 mL of water were spiked with PhMCs at 100 µg L −1 of each compound. Previously, the solution was homogenized for 20 min in darkness. Then, the appropriate amount of catalyst (250 mg L −1 of TiO 2 ) was added to the reaction solutions.
The mixture was maintained for 30 min in the dark prior to illumination to achieve maximum adsorption of the PhMCs on the semiconductor surface as previously tested (Pérez-Lucas et al., 2022a). Subsequently, Na 2 S 2 O 8 (250 mg L −1 ) was added, and the samples were exposed to natural sunlight for 240 min (10-14 h). The mean temperature measured during the photoperiod was 31.6 ± 2.2 °C. UV-A and UV-B radiation values during the experiment varied from 21.2 to 26.3 W m −2 and 0.8-1.5 W m −2 , respectively. Three replicates were performed in each case, and several samples (50 mL) were taken during the photoperiod.

Analytical Determinations
The aqueous samples (50 mL) were adjusted to pH ≈ 3 with H 3 PO 4 by adding 1 g of NaCl to increase the retention of the compounds on the adsorbent and the ionic strength, respectively. After homogenization, samples were passed through an Oasis® HLB 60 µm HLB extraction cartridge purchased from Waters (Milford, MA, USA) under vacuum, at a flow rate of ≈ 3 mL min −1 . Previously, the extraction cartridges were conditioned with CH 3 CN (5 mL) and equilibrated with Milli-Q water (5 mL). Once the sample had passed through the cartridge, it was washed with Milli-Q water (5 mL), discarding the eluate and drying the cartridge with air. Finally, the target compounds were eluted with CH 3 CN (5 mL) at a flow rate of 2 mL min −1 , collecting CH 3 CN in a graduated tube and recording the total volume. After completion of SPE process, 2 mL were filtered through a nylon filter (20 µm). Chromatographic analyses were performed on an Agilent 1100 Series HPLC system (Santa Clara, CA, USA) comprising a reversed phase C8 analytical column (150 mm × 4.6 mm and 5 μm particle size) (Zorbax Eclipse XDB-C8) coupled to an Agilent G6410A triple quadrupole mass spectrometer operating in ESI + ion mode. A primary study of optimal selected reaction monitoring (SRM) transitions was carried out for each compound.

Disappearance Kinetics
A simple comparison of the main important physicochemical parameters reveals substantial differences between the studied compounds (except macrolides) with different structures, functions, and therapeutic activities, as shown in Table 1. These compounds represent a wide range of PhMCs commonly present in WWTPs, including β-blockers, anti-epileptics, antibiotics, and antiinflammatory drugs (Helwig et al., 2013;Lin et al., 2010;Yu et al., 2006). An earlier test carried out under laboratory conditions using a photochemical reactor equipped with light emitting diode (LED) lamps (data not shown) demonstrated higher efficiency of TiO 2 compared to ZnO (with the same photocatalyst loading) in the photooxidation of the studied PhMCs. TiO 2 and ZnO are the most widely used photocatalysts for environmental applications. However, TiO 2 is a stable SC, while ZnO dissolves at acidic pH because it is not resistant to anodic photocorrosion (Pérez-Lucas et al., 2022a, b). Many photocatalytic studies using TiO 2 have investigated the effects of operating parameters, such as TiO 2 type and loading, initial pollutant concentration, solution pH, wavelength/light intensity, and water matrix composition, on the degradation kinetics of PhMCs (Carbajo et al., 2016). Thus, in the study carried out by Carabin et al. (2015), the most efficient TiO 2 type for carbamazepine degradation was P90, followed by P25, with removal efficiencies of 69% and 60%, respectively, after testing different TiO 2 materials (Hombikat UV 100, PC500, P25, P90, and ST01), which was attributed to the presence of a mixture of anatase and rutile in P90 and P25 materials. Dimitrakopoulou et al. (2012) examined the effect of eight different TiO 2 samples (Hombikat UV 100, Millennium PC50, Millennium PC100, Millennium PC105, Millennium PC500, Degussa P25, Tronox AK1, and Aldrich Anatase AA), on the degradation of the antibiotic amoxicillin showing P25 as the most suitable for its removal, which is due to its slower e − /h + recombination rate and favorable proportion ratio of anatase and rutile phases. Another study carried out by Bianchi et al. (2017) examined the disappearance of paracetamol/aspirin mixtures in DW using TiO 2 P25 and micro-sized TiO 2 K1077 concluding that TiO 2 P25 achieved complete degradation and a mineralization efficiency around 90% (after 6 h) in contrast to TiO 2 K1077, which confirmed lower activity and poor mineralization (40% after 4 h).
In our study, the selected PhMCs show a slow degradation in WWe compared to DW, where the degradation was significantly very faster as shown in Fig. 1. After 5 min of sunlight exposure, the concentrations of all compounds strongly decreased in DW to values below 5% of their initial amounts. These results indicate that the occurrence of dissolved salts and organic carbon in WWe noticeably slows down the efficiency of the treatment.
It is well known that the point of zero charge (PZC) of TiO 2 is about 6.25 (Kosmulski, 2018). Therefore, when water pH is below it, the surface of TiO 2 is positively charged. Contrarily, when the pH is above it, the surface is negative. In our case, no significant differences (p < 0.05) were found in the degradation of the studied compounds in the pH range 6-8. Depending on pH, SO 4 •− /HO • radicals can be present individually or simultaneously in the persulfate oxidation system (Liang & Su, 2009). Thus, at pH < 7, SO 4 •− is predominant; at pH = 9, SO 4 •− and HO • are present; and at pH > 9, HO • is the predominant radical. At circumneutral pH (pH of WWe used was 7.5), SO 4 •− can be more reactive than HO • , since the E 0 of SO 4 •− is detailed to be higher than that of HO • (E 0 = 2.6-3.1 V and 1.9-2.7 V, respectively) (Oh et al., 2016). Ismail et al. (2017) found that at pH 7, HO • and SO 4 •− radicals were involved in the degradation of sulfaclozine, whereas at pH 11, no contribution of SO 4 •− was observed. In addition, SO 4 •− can generate more HO • under neutral or basic conditions according to the following equations (Eqs. 1-3): The high solubility and non-toxic properties of persulfate are helpful for WW treatment, although the main disadvantage observed when it is used is the increase in SO 4 2− concentration generated by the reaction between SO 4 •− and PhMCs. According to Liu et al. (Liu et al., 2014), the single first-order (SFO) kinetic model is regularly appropriate to describe the photooxidation rate of many organic pollutants using SC materials at low substrate concentration (Eq. 4): (1) where t is the reaction time, C 0 is the initial concentration of PhMCs, C t is the residual concentration at time t, and k is the rate constant. From the above equation, the time required for X % disappearance of PhMCs (disappearance time) from the water can be calculated according to Eq. 5: Consistent with the data shown in Tables 3 and 4, the SFO model satisfactorily explains the photocatalytic degradation of the compounds studied. As can be seen, R 2 values are very close to 1 in DW with a standard error of estimation (S y/x ) ≤ 0.005. In the case of WWe, PhMCs also fitted realistically to the exponential decay curve with R 2 ≥ 0.98 in all cases and S y/x < 0.05. In DW, DT 50 (time required to reduce the concentration of a compound to exactly half of its original value) was in all cases < 1 min. In the case of WWe, the highest DT 50 values were observed for macrolides, 13 and 16 min for erythromycin and clarithromycin, respectively, while the other compounds studied showed DT 50 values below 10 min.
In the study carried out by Georgaki et al. (2014), photocatalytic conversion of carbamazepine was found to be generally slower, relative to ibuprofen using TiO 2 P25 and ZnO as photocatalyst under UV/ Visible light irradiation, with TiO 2 showing better photocatalytic efficiency in the degradation of both pharmaceuticals compared to ZnO. Photocatalytic treatment of WWe using TiO 2 and high-intensity Xenon discharge lamps (55 W) to simulate solar light resulted in high removal efficiencies for poorly biodegradable PhMCs in WWe (100% for propranolol, 100% for diclofenac, and 76% for carbamazepine) after 96 h (He et al., 2016). In all cases, photodegradation followed SFO kinetics, and the rate constant of photocatalysis was much higher than that of photolysis in absence of TiO 2 . Similar results were obtained by Eskandarian et al. (2016), where photocatalytic decomposition using TiO 2 and LED lamps was also much more significant than photolytic decomposition for different PhMCs such as acetaminophen, diclofenac, ibuprofen, and sulfamethoxazole.

Impact of Water Matrix Composition
On the other hand, the influence of the water composition on the photocatalytic process is crucial to assess its suitability for real WW treatment, and its effect can be complex. The surface charge of TiO 2 and the rate of ROS formation can be strongly affected by the components dissolved in the water matrix. The generated negative effect, mainly at high concentrations, is generally attributed to the occupancy sites on the photocatalyst surface by inorganic and organic components, the scavenging of HO • , and/or the aggregation of TiO 2 particles when the ionic strength is high Table 3 Kinetic parameters obtained following an SFO model for photocatalytic degradation of PhMCs in DW *** p < 0.001; S y/x standard error of estimation  (Ahmed et al., 2011;Náfrádi et al., 2022 (Ahmed et al., 2011;Bi et al., 2016;Ribeiro et al., 2019Náfrádi et al., 2022).  (Bi et al., 2016 . Above pH = 10.3, CO 3 2− is the predominant specie, but at pH below 8.3, all CO 3 2− has been converted to HCO 3 − (Manaham, 2010). As pH decreases, HCO 3 − also decreases and dissolved CO 2 increases. The relationship between CO 3 2− / HCO 3 − and solution pH is represented as follows (Eq. 6): The high concentration of HCO 3 − strongly scavenges HO • , generating CO 3 •− , a selective and less reactive reaction partner with a lower electrode potential (E 0 = 1.6 V) than HO • (E 0 = 2.7 V).
The effect of NO 3 − on the photooxidation of pollutants in water is contradictory . On the one hand, it can absorb UV light by acting as an inner filter. On the other hand, some authors have reported that NO 3 − can generate HO • in aqueous media, incrementing pollutant degradation. However, NO 2 − exhibits a notorious effect on the photooxidation process, which should be attributed to a group of complex reactions as specified in Fig. 2. However, in our case, the effect of NO 3 − /NO 2 − could be considered negligible due to their low concentrations in WWe.
Finally, both inhibitory and synergistic effects of dissolved organic carbon (DOC) on the degradation of different groups of CECs in water have been demonstrated as a function of the concentration and type of organic compounds involved, mainly humic (HA) and fulvic (FA) acids (Aliste et al., 2021;Tung et al., 2019;Yang et al., 2019). The presence of DOC may reduce PhMCs photodegradation, since the sensitizing (synergistic) effect is masked by quenching (inhibitory), a strong filter effect. DOC could act as an important sunlight absorber, suffering photolysis under UV-Visible light and consequently dissolved compounds can also react quickly with HO • and SO 4 •− . Aliste et al. (2021) found that a high DOC content considerably decreased the reaction rate constants observed for the photodegradation of some insecticides and their main intermediates in leaching water. However, in our case, the DOC content (3.3 mg L −1 ) in WWe is very low, and consequently, its effect on the photooxidation of the target compounds should be negligible.

Conclusions
Currently, wastewater pollution constitutes a major environmental concern. Therefore, different wastewater treatment methods and techniques have been proposed to curb the problem, of which heterogeneous photocatalysis has proven to be one of the most effective and promising method standing out for its numerous advantages over other techniques. Solar-driven heterogeneous photocatalytic processes using photocatalyst materials such as TiO 2 have been widely applied for pharmaceutical drug degradation and have proven to be efficient, eco-friendly, and cost-effective methodologies to remove emerging micropollutants in wastewater. The present study shows the effect of the tandem TiO 2 / Na 2 S 2 O 8 under natural sunlight on the degradation rate of different PhMCs in both WWe and DW. The rate is much faster in deionized water reaching 100% degradation in a few minutes compared to wastewater due to the matrix effect. Degradation experiments of the 6 investigated pharmaceutical compounds, consisting of atenolol, carbamazepine, clarithromycin, erythromycin, irbesartan, and ketoprofen, showed that the time required for 90% disappearance of PhMCs (DT 90 ) in WWe ranged from 10 to 53 min for atenolol and clarithromycin, respectively. Therefore, solar heterogeneous photocatalysis using TiO 2 in tandem with Na 2 S 2 O 8 constitutes a valuable tool for wastewater remediation, particularly in those areas that receive a large number of sunshine hours per year such as the Mediterranean basin.
Acknowledgements The authors are grateful to the Ministry of Science and Innovation of Spain for financial support (Project PID2019-106648RB-I00/AEI/https:// doi. org/ 10. 13039/ 50110 00110 33) in the framework of the State Programs of Generation of Knowledge and Scientific and Technological Strengthening oriented to the Challenges of Society.
Funding Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. For open access charge: University of Murcia Data Availability The datasets analyzed during the current study are available from the corresponding author on reasonable request.

Competing Interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.