Degradation of pharmaceuticals in sanitary effluent by the combination of oxidation and photo-oxidation processes

Abstract Most conventional sewage treatment systems are not able to fully remove micropollutants found in sewage. Thus, the simultaneous degradation of 11 pharmaceuticals identified in the effluent of a sewage treatment plant in Southern Brazil was herein investigated through advanced oxidation processes based on ozonation, ultraviolet radiation and hydrogen peroxide. For detection, samples were prepared through solid-phase extraction and pharmaceuticals were identified through ultra-high performance liquid chromatography tandem mass spectrometry. Active ingredients such as ciprofloxacin, oxytetracycline, paracetamol, sulfamethoxazole and trimethoprim had their concentrations increased for degradation analysis purposes. Trials were carried out on a bench at room temperature and neutral pH, with aliquots collected at 7.5 and 15 minutes. Two ozone doses (0.5 and 0.9 mg per mg of dissolved organic carbon), and combinations of the lowest ozone dose with photolysis (254 nm) and with 25 mg.L−1 of hydrogen peroxide were evaluated. Pharmaceuticals mineralization efficiency was assessed in a total organic carbon analyzer. The process combining ozone, hydrogen peroxide and ultraviolet radiation was the most efficient in the degradation of all pharmaceuticals detected in this study, since it enabled reducing oxytetracycline by 89.32%, caffeine by 96.79%, trimethoprim by 97.40%, ciprofloxacin by 97.75%, sulfamethoxazole by 99.79%, paracetamol by 99.96%, and clindamycin, ofloxacin, sulfadiazine, sulfathiazole and tylosin by 100%. This process also recorded the highest mineralization rate (60.52%), fact that confirmed the potential to decrease persistent pharmaceuticals found in conventional sewage treatment systems.


INTRODUCTION
Thousands of tons of pharmaceuticals are annually consumed to prevent and treat human and animal diseases, mainly in Asian, American, and European countries (JIANG et al., 2014;HU et al., 2018;MARTINI et al., 2018).However, most of these pharmaceuticals present high stability (VERLICCHI; AL AUKIDY; ZAMBELLO, 2012) and low assimilation in the body, since 50 to 90% of the administered doses are excreted as active substance in patients' urine (MARTINI et al., 2018;MONDAL;SAHA;SINHA, 2018).Thus, pharmaceuticals typically soluble in water arrive at sewage treatment plants (STP), where most of their compounds are not fully removed through conventional biological processes (BALAKRISHNA et al., 2017;GRAUMANS et al., 2022;KRISHNAN et al., 2021;LOZANO et al., 2022;LUO et al., 2014;KNOPP et al., 2016;MARSON et al. 2022;VERLICCHI;AL AUKIDY;ZAMBELLO, 2012;TAOUFIK et al., 2021).Consequently, recalcitrant substances such as antibiotics are released into the environment along with the final effluent (MARSON et al., 2022), fact that leads to (eco)toxicological effects on several non-target organisms, mainly on the aquatic ones (LIU et al., 2018;LOZANO et al., 2022).In addition, these substances can be biotransformed, bioaccumulated and biomagnified at different trophic levels (MONDAL; SAHA; SINHA, 2018).Therefore, it is necessary pursuing alternatives to enable the removal, or the maximum reduction, of several recalcitrant pharmaceuticals from municipal and industrial wastewaters.Luo et al. (2014) and Marson et al. (2022) reported the use of tertiary systems based on advanced oxidation processes (AOP), which stood out for the ease of implementation at real scale, and for the slight economic advantage, in comparison to the use of filtration membranes and activated carbon adsorption (HANSEN et al., 2016).In addition, the individual application of oxidizing agents such as ozone (O 3 ), hydrogen peroxide (H 2 O 2 ) and ultraviolet radiation (UV), mainly as disinfection systems to replace chlorinated agents in treatment plants, has increased (ARAÚJO et al., 2016).Accordingly, there has been increase in the number of studies about the use of AOP to enable the simultaneous removal of pharmaceuticals from sanitary effluents (AFONSO-OLIVARES et al., 2016;DE LA CRUZ et al., 2013;HANSEN et al., 2016;KNOPP et al., 2016;KRISHNAN et al., 2021;ZIMMERMANN et al., 2011), besides the ones addressing the individual degradation of compounds (MARTINI et al., 2018;MONDAL;SAHA;SINHA, 2018) or the reduced number of selected active ingredients (ALHARBI et al., 2017;KATSOYIANNIS;CANONICA;VON GUNTEN, 2011).
Besides conventional processes, AOP are a promising technique that can be used to remove a wide range of organic micropollutants from wastewater (ALHARBI et al., 2017;BRILLAS;MARTÍNEZ-HUITLE, 2015;TAOUFIK et al., 2021) due to the formation of highly-reactive species with high oxidizing power, such as the radical known as hydroxyl (OH • ) (E° = 2.80 V), which can enable the mineralization, or transformation, of several organic compounds into simpler substances (ARAÚJO et al., 2016).
O 3 is an extremely strong oxidant (E° = 2.07 V) that can be widely applied to disinfect water, as well as to remove taste, odor, and color from it (KATSOYIANNIS; CANONICA; VON GUNTEN, 2011;LIM et al., 2022).The transformation of organic compounds through O 3 application can happen in direct or indirect ways.The molecular ozone reacts to organic compounds in the direct mechanism via direct oxidation, whereas, in the indirect mechanism, there is the formation of OH • radicals, which react faster to organic compounds than O 3 (HANSEN et al., 2016;MAHMOUD;FREIRE, 2007;MARTINI et al., 2018).
Studies have described more than 70% reduction, and even the full removal, of some soluble pharmaceuticals that have reacted to ozone within the first 20 reaction min, at doses ranging from 0.5 to 0.9 mgO 3 mgDOC -1 (dissolved organic carbon) (HANSEN et al., 2016;KNOPP et al., 2016;ZIMMERMANN et al., 2011).However, drug removal through O 3 application can be limited by the short lifespan of the oxidizing agent and by compound adhesion to organic matter particles (KANAKARAJU; GLASS; OELGEMÖLLER, 2018).On the other hand, the combination between O 3 and UV-C radiation (200 to 360 nm) enabled O 3 photolysis and generated OH • radicals that potentiated the degradation, and mineralization, of organic compounds.Non-reactive OH • radicals, in their turn, recombined to each other and generated H 2 O 2 (ARAÚJO et al., 2016;MARTINI et al., 2018).
Although H 2 O 2 is a strong oxidant (Eº = 1.78 V), its oxidation potential is lower than that of O 3 and OH • .Thus, reactions based on H 2 O 2 are not efficient in degrading organic compounds due to their very slow decomposition in oxidizing radicals and to the strong influence of parameters such as temperature and pH (MARTINI et al., 2018).Therefore, H 2 O 2 is often used in combination with UV-C radiation (less than 280 nm), which breaks the O-O bond of the peroxide molecule and produces OH • (ALHARBI et al., 2017; ARAÚJO et al.,   2016; GRAUMANS et al., 2022).It is also possible making other combinations, such as H 2 O 2 and O 3 , in which H 2 O 2 works as homogeneous catalyst by accelerating ozone degradation to form non-selective OH • radicals (HANSEN et al., 2016;LIM et al., 2022).
Photolysis based on UV-C radiation (254 nm) is widely used to disinfect drinking water, besides being increasingly applied to slightly-turbid wastewater (ALHARBI et al., 2017).Although some pharmaceuticals are susceptible to degradation even at typical doses used in disinfection processes (AFONSO-OLIVARES et al., 2016), degradation is achieved when there is radiation enough to break the chemical bonds of the compounds, i.e., when the photon energy exceeds the bond energy (MARTINI et al., 2018).Drug degradation through UV radiation can happen both through direct and indirect ways.In the first case, molecules absorb radiation photons and break, whereas, in the second case, there is the generation of strong reactive species, such as OH • , from compounds naturally found in solution (ALHARBI et al., 2017;FATTA-KASSINOS;VAZQUEZ;KÜMMERER, 2011).Thus, it is recommended combining two, or more, AOP to form strong oxidants (OH • ) (MARSON et al., 2022;MARTINI et al., 2018;TAOUFIK et al., 2021), which can reduce the generation of toxic by-products during degradation processes (AFONSO-OLIVARES et al., 2016;ALHARBI et al., 2017;ARAÚJO et al., 2016;GRAUMANS et al., 2022).Therefore, the aim of the current study was to investigate the efficiency of simultaneous degradation, and mineralization, of pharmaceuticals in sanitary sewage post-treatment based on advanced oxidation processes comprising O 3 , UV radiation and photochemical associa-

Origin of the studied sanitary effluent
Post-treatment trials were performed with sanitary sewage from a sewage treatment plant (STP) located in the metropolitan region of Porto Alegre city, Rio Grande do Sul state (RS), Brazil.The mentioned system operates at mean flow Degradation of pharmaceuticals in sanitary effluent 2,250 ± 250 L•s -1 and serves approximately 600 thousand people.It comprises mechanical screening, de-sanding and biological treatment units in Upflow Anaerobic Sludge Blanket (UASB) reactors, as well as activated sludge, in the process called Unitank: cyclic operations in the aeration stage, which are followed by sedimentation and disinfection with H 2 O 2 .
Treated effluent samples were collected in the STP emissary in April 2018, from 7 a.m. to 12 p.m.; effluent aliquots were collected every hour until reaching 40 L volume.Non-toxic high-density polyethylene (HDPE) bottles were used, previously sanitized and rinsed with the samples at collection time.A 1-L aliquot of sample was collected in amber bottle to enable determining pharmaceuticals and hormones; the bottle was stored under refrigeration (± 4°C) for further analysis.

Detection of compounds in effluent samples
Effluent samples were subjected to tests focused on detecting 44 compounds (Table 1 supplementary material), including human and veterinary prescription pharmaceuticals and hormones, at the Pesticide Residue Analysis Laboratory (LARP) of the Federal University of Santa Maria (UFSM), RS, Brazil.
Solid analyte standards with purity ranging from 95.0 to 99.6% were purchased from the German companies Dr. Ehrenstorfer and Witega.The individual standard solution of each analyte was prepared at concentration 1,000 mg•L -1 in acetonitrile.Next, a solution containing the mixture of all pharmaceuticals at concentration 10 mg•L -1 in acetonitrile was prepared.These solutions were stored in amber bottles at 5 °C.
Analytes were quantified and identified by monitoring the selected reactions.

Sample characterization
Effluent samples were also subjected to physicochemical and microbiological characterization, before and after the post-treatment tests were applied to the most efficient process.The analyzed parameters were total organic carbon (TOC), total nitrogen (N Total ), alkalinity, turbidity, suspended solids (SS), temperature, pH, chemical oxygen demand (COD), biochemical oxygen demand (BOD 5 ), total phosphorus (P Total ) and thermotolerant coliforms, according to analytical methods set by the Standard Methods for the Examination of Water and Wastewater (APHA; AWWA; WEF, 2012).
The drug mineralization degree was determined through TOC removal in a high-sensitivity Shimadzu analyzer (model TOC-L CPH, TNM-L), based on the non-purgeable organic carbon (NPOC) fraction method.The mentioned equipment was also used to determine the dissolved organic carbon (DOC), which corresponded to the organic matter fraction permeated in a 0.45 μm filter, based on the difference between total and inorganic carbon (ZIMMERMANN et al., 2011) and N T .

Experimental procedure
After the effluent analysis, active ingredients such as ciprofloxacin, oxytetracycline, paracetamol, sulfamethoxazole and trimethoprim were added to the Table 1 -Concentrations of pharmaceuticals found through ultra-high-performance liquid chromatography coupled to tandem mass spectrometry in the sewage sample before and after the increase of active principles for post-treatment trials.
Note: LOD: limit of detection; LOQ: limit of quantification; n.d.: non-detected; *active principles that had their concentration increased for post-treatment trials.Source: elaborated by the authors.sample at concentrations higher than the ones predicted to have no effect (predicted no-effect concentrations -PNEC) on the most sensitive species, in order to enable the AOP trials.These compounds were chosen because they recorded the highest persistence and/or incidence rates in the herein investigated sample, as reported in other studies (DINH et al., 2017;HU et al., 2018;MARTÍN et al., 2012).In this way, the objective was to find a concentration capable of enabling a better comparison between different processes.

Degradation trial based on ozonation
The dissolved organic matter (measured in DOC) is the main matrix used to determine O 3 consumption in wastewater, regardless of the drug concentration (Hansen et al., 2016).Therefore, the herein adopted O 3 dose was determined based on the DOC concentration of the effluent, as explained by Zimmermann Trials were performed with two O 3 doses: the first (the lowest one) was 0.5 mgO 3 •mgDOC -1 , whereas the second (the highest one) was 0.9 mgO 3 •mgDOC -1 .
O 3 was generated in a Purizonium compact plant with production capacity of 600 mgO 3 •h -1 , based on atmospheric air suction, filtration in silica gel and electric discharge (Corona effect).The equipment outlet ducting was connected to a porous stone placed at the bottom of a 1000 mL gas scrubber-type reaction flask.
The output of the flask was connected to another 500 mL gas scrubber filled with 2% potassium iodide solution in order to transform residual ozone into oxygen (APHA; AWWA; WEF, 2012).The degradation test was carried out in batch for 15 minutes; 100 mL aliquots were collected after 7.5 min, and at the end of the reaction process, for chromatographic, TOC, COD, turbidity, and pH analyses.

Degradation trial based on ultraviolet radiation
The photolysis-based degradation experiment was performed in a cylindrical photoreactor with 76 mm diameter, 480 mm length, and 2.0 L useful volume.
The photoreactor was built in PVC and equipped in the center with a 15 W mercury OSRAM low UV-C pressure (λ = 253.7 nm) transparent quartz tube lamp; the effluent was placed in direct contact with the lamp, under slow agitation, to enable uniform exposure.The lamp was turned on 30 minutes before the photoreactor was powered on, in order to achieve maximum radiation.Next, the effluent was exposed to UV radiation for 15 minutes and aliquots were collected after 7.5 minutes, and at the end of the reaction process, under the same conditions adopted in the previous experiment.

Degradation trial based on ozonation and ultraviolet radiation
The ozonation experiment combined with UV radiation adopted the ozone dose 0.5 mgO 3 •mgDOC -1 in the photolysis test reactor, which was also used in the other UV experiments, under the same lamp preheating conditions.The porous stone of the O 3 generating station was placed inside the photoreactor to enable this experiment.After the system was filled with the effluent, O 3 started being injected in it in order to homogenize the solution and to enable uniform O 3 /effluent/UV contact.The same residual ozone destruction procedures in 2% potassium iodide solution were adopted at the reactor outlet.This experiment followed the sample collection pattern and the time adopted in the other experiments.

Degradation trial based on hydrogen peroxide and ultraviolet radiation
This experiment comprised the addition of 25 mgH 2 O 2 •L -1 (35%, Merck) effluent to the photoreactor, based on the study by Afonso-Olivares et al. (2016).
The effluent was placed in direct contact with the lamp, under slow agitation, to enable uniform exposure.Again, the lamp was turned on 30 minutes in advance and aliquots were collected after 7.5 and 15 reaction minutes for drug degradation, TOC, COD, turbidity, and pH analyses.

Degradation trial based on ozonation, hydrogen peroxide and ultraviolet radiation
This experiment was carried out under the same conditions adopted in the previous one; however, it evaluated the influence of the combination added with O 3 .The effluent containing 25 mgH 2 O 2 •L -1 (35%, Merck) was added to the system, after the hose with the porous stone was inserted in the photoreactor.The experiment adopted the dose 0.5 mgO 3 •mgDOC -1 , which enabled homogenization of, and uniformity in, the contact between the sample and the lamp.This system, like the others, was operated in batch; aliquots were collected halfway the reaction process, and at the end of it, to enable drug degradation, TOC, COD, turbidity, and pH analyses.
The photoreactor was washed between tests with running deionized water and homogenized with the sample to be post-treated.
The use of the lowest O 3 dosage (0.5 mgO

Analysis of advanced oxidation processes efficiency in degrading pharmaceuticals
The efficiency in degrading pharmaceuticals is shown in Equation 1-wherein E corresponds to efficiency (%), C 0 and C represent the initial and final concentration of interest, respectively -, which was used to evaluate drug removal/ reduction efficiency, TOC, and other effluent parameters. (1)

Degradation of pharmaceuticals in sanitary effluent
Chemical reaction kinetics principles were also taken into consideration in the herein adopted drug removal processes (ABARGUES et al., 2018); linear regression was used to determine the rate constants (K), correlation coefficient (R²) and the half-life time (t 1/2 ) of the pharmaceuticals in each process.
The degradation experiments were replicated in two additional times to statistically validate the processes.This validation was performed through ANOVA test based on the TOC concentration of the effluent at the beginning, and at the end, of the reaction.Results were subjected to Tukey test in order to compare the means (p ≤ 0.05), in compliance with procedures available in the R statistical package (R CORE TEAM, 2016).

RESULTS AND DISCUSSION
Thirteen pharmaceuticals out of the 44 evaluated compounds were found in the raw effluent of the herein investigated STP, 11 of them remained in the effluent after the treatment (Table 1), thus indicating the need of developing complementary techniques to enable their full removal or reduction.Table 1 also shows the concentration of active ingredients added to the effluent due to their persistence and incidence to enable better comparing the post-treatment processes.Enrofloxacin and metronidazole, which were found in the raw effluent at concentrations 0.037 and 0.023 μg•L -1 , respectively, were the only pharmaceuticals fully removed through the conventional system.although with the application of 50 mgO 3 •L -1 , i.e., 8.8 times the maximum application reported in the current study.The dose of 0.5 mgO 3 •mgDOC -1 did not enable full tylosin removal from the effluent, since the structural carbon, hydrogen and oxygen chain of this antibiotic is higher than that of the other herein evaluated compounds.
Sulfamethoxazole showed sensitivity to UV radiation and recorded high degradation rates in the first minutes of the process; however, the highest concentration reductions were observed in the O 3 /UV and O 3 /H 2 O 2 /UV processes -99.57and 99.79%, respectively (Figure 2).
On the other hand, clindamycin, oxytetracycline and trimethoprim were the antibiotics least affected by photolysis.This finding shows the limitation in the exclusive use of UV radiation in drug degradation processes, since sulfamethoxazole is co-administered with trimethoprim at ratio of 5:1 (LIN et al., 2018).UV radiation, in its turn, allowed reducing ciprofloxacin concentrations by more than 50% after the evaluation period.2018), most hybrid and integrated techniques perform better than single treatments.However, despite the significant reduction in the concentration of all pharmaceuticals, oxytetracycline and trimethoprim showed increased resistance to degradation at the end of the reaction process.
The drug removal efficiency shown in Figure 3 clearly indicates different responses of active principles to the evaluated oxidizing agents, fact that reinforces the importance of conducting studies with real samples in order to define the process capable of presenting the best results.
The high performance of the O 3 /H 2 O 2 /UV process was attributed to extensive OH • formation, due to photozonation.These radicals were amplified by H 2 O 2 addition (MARTINI et al., 2018).This process was mentioned by Jiang, Zhou, and Sharma (2013) as one of the main methods adopted for OH     evaluated by Mondal, Saha, and Sinha (2018).Ciprofloxacin removal rates increased from 27.50 to 62.40% as the O 3 dose increased from 0.5 to 0.9 mgO 3 •mgDOC -1 , whereas clindamycin removal rates increased from 6.58 to 40.85%, oxytetracycline removal rates increased from 15.94 to 88.40%, paracetamol removal rates increased from 80.91 to 99.34%, tylosin removal rates increased from 60.78 to 100%, and trimethoprim removal rates increased from 11.60 to 36.86%.Other studies also reported increased drug degradation efficiency as the O 3 dose increased (HANSEN et al., 2016;KATSOYIANNIS;CANONICA;VON GUNTEN, 2011;KNOPP et al., 2016;LEE et al., 2013;ZIMMERMANN et al., 2011).

Degradation of pharmaceuticals in sanitary effluent
The lowest removal efficiency rate was recorded for clindamycin at 0.5 mgO 3 •mgDOC -1 (6.58%), due to influence of the complex molecular structure of the compound.However, UV radiation was also poorly effective in removing clindamycin from the effluent, since it reduced the initial concentration of this antibiotic by 11.27%.Similar behavior was recorded for oxytetracycline (12.40%) and trimethoprim (9.11%) reduction under UV light.Alharbi et al.
(2017) reported trimethoprim reduction by 58.2% after 1 hour of exposure to UV radiation, which indicated low antibiotic reactivity to the process.
However, the application of UV radiation for 15 minutes enabled reducing ciprofloxacin concentrations by 54.73% in the present study; this outcome is very close to the 60% removal rate reported by Mondal, Saha, and Sinha (2018), who also applied direct photolysis to the same antibiotic, although for 120 minutes.Photolysis enabled reducing by more than 75% the concentrations of the other pharmaceuticals evaluated in the current study; this value was higher than the ones Afonso-Olivares et al. (2016) recorded for caffeine (10.99%) and ofloxacin (71%) after 45 reaction minutes; equal to the one recorded for sulfamethoxazole (> 99%); and lower to the ones recorded for ciprofloxacin (> 99%) and trimethoprim (18.80%).It is worth emphasizing that the mentioned authors adopted the ratio of 6.25 L sample per lamp (14 W), whereas the current study adopted the ratio of 2 L sample per lamp (15 W), although it used higher concentrations of pharmaceuticals.
The full drug degradation in wastewater is often compromised by the presence of organic matter and nitrogen, which interfere in the action of OH • radicals over target micropollutants (GRAUMANS et al., 2022;KATSOYIANNIS;CANONICA;VON GUNTEN, 2011).Accordingly, the reactivity of the effluent organic matter (EfOM) composition to hydroxyl was investigated by Dong, Mezyk, and Rosario-Ortiz (2010), who recorded reduced OH • radicals as the molecular weight of the organic fraction increased.Thus, the higher the residual EfOM concentration, the greater the reduction of OH • radicals.This outcome may explain the incomplete degradation of some pharmaceuticals at the end of the process, since the herein evaluated sanitary sewage had 42.94 mg•L -1 of COD, 10.69 mg•L -1 of TOC and 27.43 mg•L -1 of N Total .Therefore, higher O 3 and H 2 O 2 concentrations, as well as extended reaction times and analyses conducted under reduced organic matter and nitrogen concentrations, should be adopted in future studies.Furthermore, in certain cases, discharged metabolites are transformed back into their pristine structure and become bioactive again (GRAUMANS et al., 2022).

Kinetic analysis of the degradation
Overall, the degradation of antibiotics and other pharmaceuticals often follows first-order kinetics (KANAKARAJU; GLASS; OELGEMÖLLER, 2018).
However, degradation conditions enabled by AOP can induce second-order kinetics (JIANG; ZHOU; SHARMA, 2013; KATSOYIANNIS; CANONICA; VON GUNTEN, 2011;MONDAL;SAHA;SINHA, 2018), as recorded for most pharmaceuticals exposed to O 3 (0.9 mg), UV radiation and O 3 /UV combination in the current study.When these processes were analyzed in separate, only paracetamol degradation followed first-order kinetics, whereas the two processes combining the use of H 2 O 2 presented similar behavior towards degradation kinetics, whose parameters are shown in Table 2.
Based on the degradation kinetics of the herein investigated pharmaceuticals, it was possible seeing that the highest rate constant (k) values were recorded for the combination O 3 /H 2 O 2 /UV that, consequently, enabled the shortest halflife time to the active principles.This behavior reinforces the claims that drug removal happens in the early stage of the reaction, and that the process efficiency depends on operating conditions and on oxidizing agent concentrations.Other studies adopted neutral pH and recorded increased drug degradation at the beginning of reactions (HANSEN et al., 2016;KNOPP et al., 2016;MAHMOUD;FREIRE, 2007;RIVAS;BELTRÁN;ENCINAS, 2012;ZIMMERMANN et al., 2011)  Overall, oxytetracycline and trimethoprim were more resistant to degradation, whereas UV radiation enabled the lowest rate constants, except for sulfamethoxazole, which was highly sensitive to the UV light.This outcome confirmed the information about photolysis-based sulfamethoxazole degradation described by Alharbi et al. (2017).However, the mentioned authors reported increased solution toxicity, which indicated the formation of photolytic products that were more toxic than the original ones, fact that was not reported in the H 2 O 2 /UV combination.Luo et al. (2018) described pseudo first-order kinetics under direct photolysis (254 nm) applied to sulfamethoxazole, both at pH 3 and at 7.55, unlike results observed in the current study, although degradations depend on several factors such as photoreactor and effluent characteristics.This information was corroborated by Kanakaraju, Glass and Oelgemöller (2018), who pointed out that degradation reaction orders change from study to study and that there is no clear consensus about this aspect.
Mondal, Saha, and Sinha (2018) also found second-order kinetics in ciprofloxacin degradation in UV and H 2 O 2 /UV processes, thus corroborating the

Mineralization and chemical oxygen demand removal analysis
Drug mineralization through AOP was evaluated based on TOC reduction.
COD reduction was also evaluated, as shown in Figure 4.  were observed in combination processes.The limited TOC reduction may have been caused by residual organic matter found in the sewage, since it minimizes the action of OH • radicals in the pharmaceuticals, as already reported by Dong, Mezyk, and Rosario-Ortiz (2010) and by Katsoyiannis, Canonica, and Von Gunten (2011).TOC reduction results reported in the current study were higher than that described by Mondal, Saha, and Sinha (2018) in UV-based ciprofloxacin degradation (4.04%), similar to the one obtained by the same authors in the H 2 O 2 /UV process (35.41%) after 40 minutes of reaction, and to the maximum rate recorded (59.99%) in the modified photo-Fenton treatment.Márquez et al. (2014) also found 60% TOC reduction in a mixture of four pharmaceuticals (atenolol, hydrochlorothiazide, ofloxacin, and trimethoprim) in secondary effluent treated with solar photocatalysis and TiO 2 .There were no statistically significant differences in COD reduction among O 3 (0.9 mg•mgDOC -1 ), O 3 /UV, and O 3 /H 2 O 2 /UV processes.In addition, processes based on the lowest O 3 (0.5 mg•mgDOC -1 ), UV, and H 2 O 2 /UV concentrations showed statistically similar results.However, the O 3 /H 2 O 2 /UV combination was statistically similar to both groups ("A" and "B").This outcome can be attributed to COD increase caused by H 2 O 2 addition, since H 2 O 2 /UV showed the lowest COD reduction rate (8.17%), whereas the maximum reduction rate (30.69%) was observed in the highest O 3 dose.
According to Quinõnes et al. (2015), the synergy index (SI) can determine whether the combined processes are more, or less, efficient in drug mineralization than isolated processes.This index is calculated by dividing the mineralization efficiency value recorded for the combined process by the sum of the mineralization efficiency values recorded for isolated processes.SI higher than 1 means that the combined process has positive effect on efficiency, whereas SI lower than 1 indicates that the combined process is less efficient in drug mineralization.This evaluation was applied to O 3 / UV processes, such as O 3 /UV and O 3 /H 2 O 2 /UV, which resulted from O 3 and H 2 O 2 /UV combination.Both processes recorded synergy indices higher than 1-1.05 and 1.02, respectively; this outcome indicated that these processes were more efficient in drug mineralization than the isolated methods, as already recorded in the drug removal process.However, O 3 and H 2 O 2 concentrations, as well as the reaction time, may not have been enough to form the necessary OH • radicals for drug mineralization, as also observed by Martini et al. (2018).
Finally, the O 3 /H 2 O 2 /UV process enabled sanitary sewage disinfection (Escherichia coli elimination), mean turbidity reduction by 13.45%, BOD 5 by 15.32%, N Total by 28.94% and slight pH increase from 6.94 to 7.05.There were no significant changes in SS and P Total .
et al. (2011),Knopp et al. (2016) andHansen et al. (2016), and on the production capacity of the generating plant, which was evaluated according to the indirect titration iodometric method recommended by the International Ozone Association (APHA; AWWA; WEF, 2012).
Degradation tests were conducted through different AOP based on the knowledge about drug concentrations, DOC (6.31 mg•L -1 ) and TOC (10.69 mg•L -1 ) in the effluent.Results are shown in Figure1.The herein evaluated AOP recorded different drug removal performances, except for antibiotics such as ofloxacin, sulfadiazine and sulfathiazole, which were no longer found in the early reaction minutes, regardless of the degradation process.The degradation of these antibiotics may have been favored by low ofloxacin and sulfadiazine concentrations, which ranged from 0.025 to 0.078 μg•L -1 , respectively.However, although clindamycin was found at low concentration (0.071 μg•L -1 ) in the effluent, it was only removed through the H 2 O 2 /UV and O 3 /H 2 O 2 /UV processes.This drug was more sensitive to H 2 O 2 in combination with UV radiation, since O 3 and UV radiation, either in separate or in combination, did not enable eliminating the active principle due to its strongly complex molecular structure.Three other antibiotics -trimethoprim, ciprofloxacin, and sulfamethoxazole -also presented degradation rates lower than the ones recorded for other pharmaceuticals in processes evaluating the use of O 3 , in separate.This outcome may have been influenced by the oxidizing agent dose, sinceMartini et al. (2018) reported that sulfamethoxazole was more easily removed through a single ozonation, after 15 reaction minutes,
between the final (C) and initial (C 0 ) concentrations.Source: elaborated by the authors.

Figure 2 -
Figure 2 -Total ion chromatogram generated for sulfamethoxazole after effluent treatment based on different advanced oxidation processes.Source: elaborated by the authors.

Figure 4 -
Figure 4 -Efficiency in reducing total organic carbon and chemical oxygen demand through advanced oxidative processes.Note: TOC: total organic carbon; COD: chemical oxygen demand.Data represent the mean and standard error (±).Lowercase letters in TOC data and uppercase letters in COD data indicate significant difference between treatments (p ≤ 0.05).Source: elaborated by the authors.Studied processes (KNOPP et al., 2016;RIVAS;sed on O 3 ENCINAS, 2012;ZIMMERMANN et al., 2011)ch as O 3 /UV, H 2 O 2 /UV, and O 3 /H 2 O 2 /UV, were performed at laboratory bench scale, room temperature 20.6 ± 0.3°C and pH 6.94 ± 0.02 (adjusted with 0.1 N H 2 SO 4 ).Reaction time was set at 15 minutes, since drug degradation through AOP can happen in the early stage of the process(KNOPP et al., 2016;RIVAS;   BELTRÁN;ENCINAS, 2012;ZIMMERMANN et al., 2011), whereas longexposure periods would hardly be economically viable at large scale.Samples were not filtered to allow simulating real conditions.

Table 2 -
Kinetic parameters of drug degradation reactions, based on the different herein investigated advanced oxidative processes: order of reactions; kinetic constant; correlation coefficient and half-life time.Note: parameters of ofloxacin, sulfadiazine, sulfadiazine, sulfathiazole and tylosin were not presented because these antibiotics were fully removed in the first seconds of all processes, as it happened to clindamycin subjected to H 2 O 2 /UV and to O 3 /H 2 O 2 /UV; k constant: kinectic constant; t 1/2 : half-life time.Source: elaborated by the authors.
that micropollutant oxidations often follow second-order kinetics, mainly when there is O 3 action and combined conditions capable of favoring OH • formation.This kinetics is maintained as long as the concentration of micropollutants does not significantly affect the stability of the oxidant.Therefore, kinetic results recorded in the current study suggested the joint action of direct (due to the action of O 3 and UV radiation) and indirect (due to OH • formation) degradation.