AOPS: RECENT ADVANCES TO OVERCOME BARRIERS IN THE TREATMENT OF WATER, WASTEWATER AND AIR Photo-Fenton degradation of the pharmaceuticals ciprofloxacin and fluoxetine after anaerobic pre-treatment of hospital effluent João A. de Lima Perini1 & Beatriz Costa e Silva1 & Adriano L. Tonetti2 & Raquel F. Pupo Nogueira1 Received: 29 January 2016 /Accepted: 4 August 2016 /Published online: 15 August 2016 # Springer-Verlag Berlin Heidelberg 2016 Abstract This work evaluated the photo-Fenton degradation of two pharmaceuticals extensively used in human medicine, ciprofloxacin (CIP), and fluoxetine (FLU) when present in an anaerobic pre-treated hospital effluent (HE) at low concentra- tion (100 μg L−1). Operational parameters such as concentra- tion of hydrogen peroxide, iron, and initial pH as well as the effect of iron citrate complex were evaluated considering the degradation of the pharmaceuticals. Iron citrate complex (Fecit) influenced significantly FLU degradation at pH 4.5 achieving 80 % after 20 min, while with iron nitrate only 36 % degradation was obtained after the same time. However, only a slight effect was observed on CIP degrada- tion, achieving 86%with Fecit and 75%with Fe(NO3)3, after 20 min. Samples of HE used in this work were previously treated in an anaerobic reactor followed by sand filtration; however, the presence of pharmaceuticals was detected. Degradation of both FLU and CIP was significantly hindered when present in HE, due to the relatively high content of organic (39.6 mg L−1) and inorganic (12.5 mg L−1) carbon, which may have consumed ·OH in side reactions. However, the iron cycle reduction was not affected by the matrix in the presence of citrate. Despite the recalcitrance of the matrix (no total organic carbon removal), it was possible to achieve over 50 % degradation of both pharmaceuticals after 90 min. Keywords Hospital effluent . Antibiotic . Antidepressant . Citrate . Iron complex Introduction Due to the increasing use of pharmaceutical products, residues of these compounds have often been found at concentrations ranging from ng L−1 to μg L−1 in wastewaters, surface waters, and even in drinking water (Rodil et al. 2012; Vázquez et al. 2013; Kosjek et al. 2013). After ingestion, pharmaceuticals are partially converted in- to metabolites or excreted unchanged (Heberer 2002). Once excreted, these compounds are released directly into the envi- ronment, or pass through a wastewater treatment plant (WWTP), main route of input of pharmaceuticals waste in the aquatic environment due to inefficient elimination of these compounds (Martínez-Bueno et al. 2007). Hospitals sewage network is often directly connected to the municipal WWTP, increasing the concentration of pharmaceuticals and metabo- lites in the effluent to be treated, which is considered an inad- equate solution (Pauwels et al. 2006; Vieno et al. 2007). Furthermore, some compounds present in hospital effluents can inhibit microbial biomass and consequently reduce the efficiency of WWTP (Verlicchi et al. 2010). In Brazil, anaerobic systems can be employed for the treat- ment of hospital effluents and other health care establish- ments, which could be decentralized, to treat sewage on site. Anaerobic filter combined with sand filter is one of the sys- tems suggested by the Brazilian standards. Several studies demonstrated its robustness and ability to generate a good effluent quality, with significant reduction of chemical oxygen demand (COD), total suspended solids, coliform bacteria, and viruses (Gross and Mitchell 1990; Emerick et al. 1999). Responsible editor: Philippe Garrigues * Raquel F. Pupo Nogueira nogueira@iq.unesp.br 1 Department of Analytical Chemistry, Institute of Chemistry of Araraquara, UNESP—Univ Estadual Paulista, P.O. Box 355, Araraquara, SP 14801-970, Brazil 2 School of Civil Engineering, Architecture and Urban Design—FEC, UNICAMP—University of Campinas, P.O. Box 6021, Campinas, SP 13083-852, Brazil Environ Sci Pollut Res (2017) 24:6233–6240 DOI 10.1007/s11356-016-7416-4 http://crossmark.crossref.org/dialog/?doi=10.1007/s11356-016-7416-4&domain=pdf However, the literature does not present any study of its action on removal of pharmaceuticals. Fluoxetine (FLU) is a selective serotonin reuptake inhibitor widely used for treatment of patients with depression (Chu and Metcalfe 2007). FLU was found in university hospital effluent in concentrations in the range of 34.8–105 ng L−1 in Portugal (Santos et al. 2013) and 21 ng L−1 in an effluent from a psychiatric hospital in China (Yuan et al. 2013). It has been reported that exposure of medaka fish to 1–5 μg L−1 concen- trations of FLU during 4 weeks affected egg fertilization (Foran et al. 2004). Increase of mosquito fish lethargy has been also observed when exposed to FLU concentrations be- tween 0.05 and 5 μg L−1, although survival was not affected (Henry and Black 2008). Ciprofloxacin (CIP) is a fluoroquinolone broad- spectrum antibiotic effective against gram-positive and gram-negative bacteria used in the treatment of diseases in humans and animals (Bongaerts and Hoogkamp- Korstanje 1993). CIP is often detected in hospital effluent at concentrations in the range of 32–99 μg L−1, while in secondary wastewater and surface water ng L−1 levels have been reported (Martins et al. 2008; Santos et al. 2013). Fenton process has attracted considerable attention for the degradation of non-biodegradable and/or toxic compounds including pharmaceuticals due to the high efficiency in ·OH production from a mixture of H2O2 and Fe2+ in acid medium (Eq. 1). Degradation can be accelerated with UV-Vis irradia- tion due to the photo-reduction of Fe3+ to Fe2+, establishing an iron cycle besides of generating extra ·OH (Eq. 2) (Pignatello 1992). Fe2þ þ H2O2→ Fe3þ þ ˙OHþ OH− ð1Þ Fe OHð Þ2þ þ hν→ Fe2þ þ ˙OH ð2Þ The enhancement of iron photo reduction can be achieved by the use of organic ligands such as oxalic or citric acid which shows much higher quantum yield of Fe(II) generation than iron aqua complexes besides the higher molar absorption coefficients in the UV-Vis region (Trovó et al. 2008; Klamerth et al. 2013). In addition, organic ligands extend the pH range that can be used in the Fenton reaction (optimum pH 2.5–3.0), allowing application at near neutral pH values (Silva et al. 2007; Perini et al. 2013; Soares et al. 2015). The aim of this work was to study the photo-Fenton deg- radation of CIP and FLU at μg L−1 levels present in an anaer- obic pre-treated hospital effluent. Firstly, the influence of some parameters such as pH, iron, H2O2 concentrations, and iron complexation on pharmaceuticals degradation was eval- uated in distilled water. Secondly, photo-Fenton process was evaluated when these pharmaceuticals were present in an an- aerobic pre-treated hospital effluent. Material and methods Reagents Ciprofloxacin hydrochloride monohydrate (99 %) (C17H18FN3O3·HCl·H2O; MW = 385.82 g mol−1) and fluox- e t ine hydroch lor ide (99 %) (C17H18F3NO·HCl ; MW = 345.79 g mol−1) were obtained from Pharma Nostra (São Paulo, Brazil). Fe(NO3)3·9H2O (Mallinkrodt, Paris, KY, USA)was used to prepare aqueous 0.25M iron stock solution. H2O2 30 % (w/w) was used (Synth, São Paulo, Brazil). Citric acid (Synth, São Paulo, Brazil) was used as iron ligand. 2,2′- bipyridyl and peroxidase (type II-A from horseradish, 1500 units/mg solid) were purchased from Sigma-Aldrich (St. Louis, MO, USA). N,N-diethyl-1,4-phenylene-diamine (DPD) was obtained from Fluka (Steinheim, Germany). 1,10-phenanthroline was obtained from Vetec (Rio de Janeiro, Brazil). A 1 M H2SO4 (Chemis, São Paulo, Brazil) solution was used for pH adjustment. Methanol, formic acid, and acetic acid (HPLC grade) were purchased from J.T. Baker (Xalostoc, Mexico). Ultrapure water (DG 500UF, Gehaka, São Paulo, Brazil) was used for dilutions and for HPLC analysis. Hospital effluent Effluent from a university hospital was firstly treated in an anaerobic process. As described by Tonon et al. (2015), the system consists of an upflow anaerobic filter filled with coconut shells (C. nucifera) followed by a sand filter. The hydraulic retention time (HRT) of the anaerobic filter was 9 h, and the hydraulic loading rate of the sand filter was 200 L m−2 day−1. The effluent from sand filters, named hereafter as hospital effluent (HE), was collected and used to carry out the photo- Fenton degradation of the pharmaceuticals after determi- nation of pH using a pH meter (1100 series, Oakton, Vernon Hills, IL, USA), total organic and inorganic car- bon concentration using a total organic carbon (TOC) analyzer (TOC-5000A-Shimadzu, Kyoto, Japan), turbid- ity (Turbidimeter-Quimes Q279P, São Paulo, Brazil), conductivity and total dissolved solids (pHtek-pH8b, São Paulo, Brazil), and chemical oxygen demand (COD) using COD digestion kit (HACH, Loveland, CO, USA) and then analyzed in a COD photometer (Macherey-Nagel model PF-3, Düren, Germany). The concentration of iron present in HE was measured using ICP-OES Optima 8000 spectrometer from Perkin Elmer (Waltham, MA, USA) after H2O2/HNO3 digestion. The effluent was spiked with 100 μg L−1 of each pharma- ceutical (CIP and FLU) to evaluate the efficiency of their degradation when present in this matrix. 6234 Environ Sci Pollut Res (2017) 24:6233–6240 Experimental degradation procedures Photo-Fenton experiments were carried out in an upflow photoreactor previously described (Nogueira and Guimarães 2000). The source of irradiation was a 15 W black-light lamp with maximum emission at 365 and 410 nm. The irradiated volume of the reactor was 280 mL, and a total volume of 500 mL of pharmaceutical solution was recirculated at a flow rate of 90 mL min−1 using a peristaltic pump (Masterflex 7518-12, Vernon Hills, IL, USA). The iron complex was pre- pared in situ by the addition of citric acid to iron nitrate solu- tion at 1:1 M ratio. The solution pH was then adjusted to the desired value by addition of 1 M H2SO4. Appropriate volume of H2O2 was then added to the solution under magnetic stir- ring and immediately pumped into the reactor. The lamp was only switched on once the reactor was completely filled, when the time started to be monitored. The initial concentration of pharmaceuticals in distilled water (DW) or HE was 100 μg L−1 (0.26 μM for CIP and 0.29 μM for FLU), and concentrations of iron and hydrogen peroxide were 1 and 50 μM, respectively, unless otherwise stated. Solid phase extraction Solid phase extractions (SPEs) were carried out for pre- concentration of the pharmaceuticals, which was necessary for quantification at μg L−1 levels. In this procedure, Fenton reaction is quenched since pharmaceuticals are retained in the cartridge while aqueous phase containing iron ions and resid- ual H2O2 is discharged, thus interrupting the degradation. Sep- Pak-C18 (360mg) cartridges were used for FLU extraction. In the case of CIP, recovery values with this cartridge were very low, probably due to the strong interaction of residual silanol groups with the acid group of CIP molecule. Therefore, Oasis HLB (60 mg) cartridges (Waters, Milford, MA, USA) were used for CIP extraction. Sep-Pak-C18 cartridges were previ- ously conditioned with 5 mL methanol followed by 3 mL water. Then 15 mL of sample was percolated through the cartridge and recovered with 5 mL methanol. In the case of CIP, Oasis HLB cartridges were used after conditioning with 1 mL methanol followed by 1 mL water. Then 10 mL of sample was percolated through the cartridge and recovered with 1 mL methanol/formic acid (50:50) solution. Average recovery percentages for CIP were 105.5 ± 12.8 % in DW and 94.6 ± 5.5 in HE, while for FLU recoveries were 108.8 ± 20.2 in DW and 113.0 ± 8.5 in HE. Chemical analysis The decay of pharmaceuticals concentration during the exper- iments was determined using reversed-phase HPLC (LC 20AT Prominence, Shimadzu, Kyoto, Japan) coupled to a flo- rescence detector Shimadzu (FL-20A). A C-18 column (EVO Kinetex, 5 μm, 150 × 4.6 mm, Phenomenex, Torrance, CA, USA) was used. The mobile phase was a mixture of methanol and acetate buffer (pH = 3.6) (50:50) for FLU and methanol and formic acid (15:85) for CIP, both at 0.5 mL min−1 flow rate of in isocratic mode and 40 μL injection volume. The excitation and emission wavelengths for CIP detection were 278 and 445 nm and for FLU detection 225 and 310 nm, respectively. Under these conditions, retention time was 6.8 min for FLU and 7.3 min for CIP, and the limit of quan- tification was 300 ng L−1 for FLU and 75 ng L−1 for CIP. The samples were filtered through 0.45-μm nylon membrane sy- ringe filter (Millipore, Bedford, MA, USA) before HPLC analysis, and no decrease of pharmaceuticals concentration was observed. Residual hydrogen peroxide concentration during photo- Fenton experiments was determined by measuring the absor- bance at 551 nm after a peroxidase-catalyzed reaction with DPD (Bader et al. 1988). Concentration of ferrous ions gen- erated during experiments was determined by measuring the absorbance at 510 nm after reaction with 1,10-phenanthroline (Fortune and Mellon 1938). A Shimadzu UV mini-1240 (Kyoto, Japan) spectrophotometer was used in both cases. Results and discussion Effect of iron source, pH, concentration of H2O2 and Fecit Considering that the experiments were carried out with rela- tively low concentration of the pharmaceuticals (100 μg L−1), low concentrations of iron (1 μM) and hydrogen peroxide (50 μM) were also used. Experiments were firstly carried out in DW to verify the effect of main conditions for pharma- ceuticals degradation and then further applied in HE. The iron speciation is very important in photo-Fenton deg- radation, as each target compound may exhibit different inter- actions with iron and thus interfering on Fenton reaction (Nogueira et al. 2005). Therefore, degradation of target com- pounds was compared using free iron (Fe(NO3)3), and iron citrate complex (Fecit) in photo-Fenton process at initial pH 4.5. FLU degradation was strongly influenced by the iron species with faster reaction in the presence of Fecit achieving 80 % after 20 min, while only 36 % were degraded with iron nitrate. On the other hand, only a slight difference was ob- served for CIP degradation achieving 75 and 86 % of antibi- otic removal after 20 min, in the absence and presence of Fecit, respectively (Fig. 1). The higher degradation efficiency of pharmaceuticals, principally FLU, in the presence of Fecit in relation to Fe(NO3)3 can be attributed to the higher quantum efficiency in the generation of Fe2+ at pH 4.0 (ϕFe(II) = 0.45, for Fecit) (Abrahamson et al. 1994), much higher compared to the hydroxylated species of iron(III) (ϕFe(II) = 0.12 at pH 4.0; Faust and Hoigné 1990). Environ Sci Pollut Res (2017) 24:6233–6240 6235 The pseudo-first-order rate constants (kobs) for the removal of pharmaceuticals using Fe(NO3)3 and Fecit were determined by linear regression of ln(C/C0) (C = concentration of phar- maceutical at time t; C0 = initial concentration of pharmaceu- tical) as a function of reaction time. All curves were linear with correlation coefficients higher than 0.95. Overall, a higher pseudo-first-order rate constant (kobs) of pharmaceuti- cal degradation in the presence of Fecit was observed, especially for FLU (kobs = 0.097 min−1), when compared to Fe(NO3)3 (kobs = 0.023 min−1). On the other hand, the pseudo- first order rate constant of CIP degradation was similar with both iron sources, kobs = 0.085 and 0.067 for Fecit and Fe(NO3)3, respectively (Table 1). The consumption of hydrogen peroxide can be used as an indirect measure of the target compound degradation efficiency. So, the initial consumption of H2O2 during CIP and FLU degradation with both iron species was mea- sured. Lower consumption of H2O2 occurred using Fe(NO3)3, especially in the case of FLU degradation, reaching only 28 % after 90 min reaction. On the other hand, higher hydrogen peroxide consumption was ob- served with Fecit, achieving around 69 % after the same reaction time for both pharmaceuticals, CIP and FLU (Fig. 1a, b, open symbol), due to the higher Fe2+ genera- tion from this organic iron complex for reaction with H2O2. The photo-Fenton process depends strongly on pH, with maximum degradation efficiency in a narrow pH range (2.5–3.0) (Pignatel lo 1992; Nogueira and Guimarães 2000). To evaluate the pH effect, experiments were performed at three different initial pH values (3.4, 4.5, and 5.6) using Fecit complex as a source of iron. At initial pH 3.4, CIP and FLU degradation was higher than 97 % after only 20 min, with kobs = 0.20 and 0.18 min−1, respectively (Fig. 2) and decreased slightly at pH 4.5. On the other hand, when the initial pH was 5.6, FLU deg- radation was significantly lower than CIP, achieving only 56 % ,while 88 % CIP were removed within 90 min (Fig. 2). This different kinetics is probably due to the degra- dation of citrate in reaction medium at pH 5.6 (40 % TOC removal from citrate after 30min; data not shown), which lead to iron precipitating Fe(III), and thus hindering FLU degrada- tion. On the other hand, CIP can complex Fe(III) forming Fe(CIP)2 and FeCIP complexes (Eldin et al. 1996), thereby providing soluble iron for Fenton reaction, which may have favored CIP degradation compared to FLU at pH 5.6. 0 20 40 60 80 100 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 b Fe(NO 3 ) 3 Fecit Fe(NO 3 ) 3 Fecit F L U /F L U 0 Treatment time (min) C IP /C IP 0 0 20 40 60 80 100 a H 2 O 2 c o n s u m p ti o n ( % ) 0 20 40 60 80 100 H 2 O 2 c o n s u m p ti o n ( % ) Fig. 1 Influence of iron species on degradation (solid symbols) and H2O2 consumption (open symbols) of CIP (a) and FLU (b) during photo-Fenton process in distilled water. Experimental conditions: [CIP] = [FLU] = 100 μg L−1, [Fe(NO3)3] = [Fecit] = 1 μM, [H2O2] = 50 μM, initial pH = 4.5 Table 1 Kinetic parameters obtained for CIP and FLU degradation under different experimental conditions Matrix pH iron (μM) H2O2 (μM) kCIP (min −1) t1/2CIP (min) kFLU (min−1) t1/2FLU (min) DW 4.5 1a 50 0.067 10 0.023 30 DW 3.6 1b 50 0.20 3.4 0.18 3.9 DW 4.5 1b 50 0.085 8.2 0.097 7.7 DW 5.6 1b 50 0.045 15 0.022 32 DW 4.5 1b 100 0.13 5.3 0.16 4.3 DW 4.5 2b 50 0.44 1.6 0.26 2.8 HE 4.5 2b 100 0.022 32 0.0075 92 HE 4.5 2b 100 + 100c 0.020 35 0.0073 95 a Fe(NO3)3 b Fecit c Second addition of H2O2 after 40 min 6236 Environ Sci Pollut Res (2017) 24:6233–6240 At pH 5.6, there is a predominance of less photoactive species as Fe(OH)(cit)− (Faust and Zeep 1993; Abrahamson et al. 1994; Ou et al. 2008). Using the public domain program Visual MINTEQ 3.1, it was calculated that at pH 4.5, 83 % of the Fe(III) is in the form of iron citrate complex, while 66% of the iron concentration is complexed at pH 5.6, favoring deg- radation at lower pH values. Consequently, pharmaceutical degradation efficiency using Fecit is practically independent of pH up to pH 4.5, decreasing considerably at pH 5.6, espe- cially in the case of FLU, showing the importance of using complexed iron in detriment to free iron, due to the low solubility-product constant of Fe(III) (Ksp = 2 × 10−39). The effect of H2O2 and iron concentration is another im- portant parameter for photodegradation of contaminants, since the excess or lack of these reagents may reduce the efficiency of the process. Firstly, experiments were carried out at pH 4.5 with two different H2O2 concentrations (50 and 100 μM), while the concentration of Fecit was maintained at 1 μM, resulting in a H2O2/Fe molar ratio of 50 and 100, respectively. The use of 100 μM hydrogen peroxide lead to an increase in process efficiency achieving more than 95% degradation of CIP and FLU after 20 min (Fig. 3). In this case, the H2O2 concentration was approximately eight times the stoichiometric amount (13 μM) for total degradation of 0.26μMCIP and 0.29μMFLU (100μg L−1) into CO2, water, and inorganic ions. Moreover, when a lower concentration of the oxidant was used (50 μM, approximately four times the theoretical stoichiometric H2O2 concentration), total degrada- tion of the pharmaceuticals was not achieved until 90 min (Fig. 3). As can be seen in Table 1, the rate constants of CIP and FLU degradation increased with the increment of hydro- gen peroxide concentration. The influence of iron concentration in pharmaceuti- cals degradation was assessed using two different Fecit concentrations (1 and 2 μM) and initial H2O2 concen- tration fixed at 50 μM, which corresponds to a H2O2/Fe molar ratio of 50 and 25, respectively. The degradation efficiency of pharmaceuticals was increased with in- creasing initial Fecit concentration from 1 to 2 μM (Fig. 3). The kinetic constant for CIP degradation en- hanced 5.2 times and FLU 2.7 times, with a reduction of half-life time from 8.2 to 1.6 min for CIP and 7.7 to 2.8 min for FLU (Table 1). Therefore, using 2 μM Fecit concentration resulted in higher degradation rates in comparison to the higher concen- tration of H2O2 (100 μM), which shows that the iron 0 20 40 60 80 100 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 Fecit/H 2 O 2 = 1/50 (µM) Fecit/H 2 O 2 = 1/100 (µM) Fecit/H 2 O 2 = 2/50 (µM) b F L U /F L U 0 Treatment time (min) a Fecit/H 2 O 2 = 1/50 (µM) Fecit/H 2 O 2 = 1/100 (µM) Fecit/H 2 O 2 = 2/50 (µM) C IP /C IP 0 Fig. 3 Influence of H2O2 and Fecit concentration onCIP (a) and FLU (b) degradation during photo-Fenton process in distilled water. Experimental conditions: [CIP] = [FLU] = 100 μg L−1, initial pH = 4.5 0 20 40 60 80 100 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 pH = 3.4 pH = 4.5 pH = 5.6 F L U /F L U 0 Treatment time (min) b pH = 3.4 pH = 4.5 pH = 5.6 C IP /C IP 0 a Fig. 2 Influence of pH on CIP (a) and FLU (b) degradation during photo-Fenton process in distilled water. Experimental conditions: [CIP] = [FLU] = 100 μg L−1, [Fecit] = 1 μM, [H2O2] = 50 μM Environ Sci Pollut Res (2017) 24:6233–6240 6237 concentration is more important to the oxidation of pharma- ceuticals than the concentration of hydrogen peroxide. Photo-Fenton degradation of CIP and FLU inHE samples The analysis of Table 2 shows that the anaerobic system ap- plied as pre-treatment of the hospital effluent was able to gen- erate an effluent with a similar quality to those generated from more complex systems (Tonon et al. 2015). This demonstrates the viability of onsite treatment of hospital sewage, with respect to environmental parameters traditionally evaluated. However, in relation to pharmaceutical compounds and its metabolites, the anaerobic treatment was not able to remove it completely. A qualitative analysis of the HE (after anaerobic treatment and sand filter), two pharmaceuticals, CIP, and diclofenac was de- tected with high intensity in LC-MS/MS analysis in the effluent pre-concentrated 40 times using Sep-Pak-C18 cartridges. The quantification of these pharmaceuticals and others in treated and raw HE will be discussed in a future work. The concentra- tion of total carbon (TC) in HE at natural pH 5.6 is 52.1 mg L−1 of which 12.5 mg L−1 due to inorganic carbon (IC) corresponds to approximately 24 % of the total carbon and a COD of 100 mg L−1 (Table 2). However, the pH adjustment to 4.5 and magnetic stirring during 30 min in all experiments with HE reduced the TC and IC content to 45.2 and 7.3, respectively, due to evolution of CO2 (Table 2). The experiments with HE were performed using higher Fecit and H2O2 concentrations, 2 and 100 μM, respectively, due mainly to the IC content of HE, which even after pH adjustment could reduce the process efficiency by consump- tion of ·OH in side reactions. Although the calculated stoichio- metric amount of H2O2 necessary for complete oxidation of the HE is 6.25 mM, lower H2O2 concentration was used con- sidering that the objective of the photo-Fenton treatment pro- posed in this work was the degradation of pharmaceuticals present at low concentrations (100 μg L−1) and not the min- eralization of the effluent. Lower efficiency of CIP and FLU degradation (Fig. 4a, b) was observed in HE at pH 4.5, achieving 69 and 47 % after 90 min, with half-life time of 32 and 92 min (Table 1), respec- tively, even using twice the concentration of Fenton reagents compared to DW. Considering the relatively high TOC and IC content of the effluent when compared to low H2O2 added for the pharmaceuticals degradation, an experiment was carried out with a second addition of H2O2 (100 μM) after 40 min reaction. However, this procedure did not improve the remov- al of CIP and FLU, reaching the same removal level when compared to a single addition of oxidant, i.e., 72 and 40 % for CIP and FLU, respectively after 90 min (Fig. 4a, b), indi- cating that the hydrogen peroxide was not the limiting factor of the degradation reaction. No significant decrease of TOC was observed during 90 min reaction. However, despite the low concentration of hydrogen peroxide (100 μM), only 25 % was consumed dur- ing photo-Fenton degradation in HE (Fig. 4c), much lower consumption compared to DW (69%). Accordingly, the effect of HE on H2O2 consumption and, consequently, on pharma- ceutical degradation, could be associated to the presence of recalcitrant compounds in this effluent. As reported by Table 2 Main parameters determined for pre-treated hospital effluent (HE) Parameters Before pH adjustment After pH adjustment Total carbon (mg L−1) 52.1 45.2 Inorganic carbon (mg L−1) 12.5 7.3 Total organic carbon (mg L−1) 39.6 37.9 pH 5.6 4.5 Dissolved Fe (mg L−1/μM) 3.72/66.6 nd Turbidity (nephelometric units) 6.06 nd Total dissolved solids (mg L−1) 493 nd COD (mg O2 L −1) 100 nd Conductivity (μS cm−1) 1054 nd COD chemical oxygen demand, nd not determined 0 20 40 60 80 100 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 Fecit/H 2 O 2 = 2/100 (µM) (HE) T O C /T O C 0 Treatment time (min) c b Fecit/H 2 O 2 = 1/50 (µM) (DW) Fecit/H 2 O 2 = 2/100 (µM) (HE) Fecit/H 2 O 2 = 2/100+100 (µM) (HE) H 2 O 2 addition F L U /F L U 0 a Fecit/H 2 O 2 = 1/50 (µM) (DW) Fecit/H 2 O 2 = 2/100 (µM) (HE) Fecit/H 2 O 2 = 2/100+100 (µM) (HE) C IP /C IP 0 H 2 O 2 addition 0 20 40 60 80 100 H 2 O 2 c o n s u m p ti o n ( % ) Fig. 4 Influence of matrix on CIP (a) and FLU (b) degradation. TOC removal (solid symbols) and H2O2 consumption (open symbols) (c) during photo-Fenton process. Experimental conditions: [CIP] = [FLU] = 100 μg L−1, initial TOC = 37.8 mg L−1, initial pH = 4.5. DW distilled water, HE hospital effluent 6238 Environ Sci Pollut Res (2017) 24:6233–6240 Ruppert et al. (1993), both mineralization and hydrogen per- oxide consumption are greatly affected by the recalcitrance of target compounds. Probably, in the case of HE, recalcitrant products formed during anaerobic degradation process (low molecular weight carboxylates and alcohols) can inhibit the Fenton reaction (Michael-Kordatou et al. 2015). Furthermore, the presence of IC (carbonate and bicarbonate) in HE can also scavenge OH radicals (Eqs. 3 and 4): Ä nO�Hþ HCO3 −→ H2Oþ CO3 − ð3Þ Ä nO�Hþ CO3 2−→ HO− þ CO3 �− ð4Þ It was also considered that the compounds present in HE could be suppressing the generation of Fe2+ due to its com- plexation with organic compounds in the sample and conse- quently, decreasing the Fenton reaction efficiency. Therefore, experiments were performed to compare the generation of Fe2+ in DW and HE during irradiation. Initial iron (III) con- centration was 116 μM in DW and HE (50 + 66 μM already present in the effluent, totaling 116 μM) (Table 2), and the citrate added was 50 μM. It was observed that the generation of Fe2+ in HE was higher than in DW reaching 60 μM in the presence of Fecit within the first 10 min, while in DW it reached 45 μM after the same time (Fig. 5). This indicates that HE did not hindered Fe3+ reduction and recalcitrance of its components, and the inorganic carbon present is the main cause for the lower degradation of pharmaceuticals in this matrix. However, in the absence of Fecit, the generation of Fe2+ was significantly lower in both DW and HE, achieving approximately 7 μM after 10 min, showing the importance of organic complex in iron reduction cycle. Despite the recalcitrance of organic compounds present in the HE studied in this work, it was possible to achieve more than 50 % degradation of CIP and FLU after 90 min at pH 4.5 using very low concentration of Fenton additives. Moreover, from the chromatograms before and after 90 min photo-Fenton degrada- tion, it can be observed that the intensity of the main peaks decreased considerably indicating that photo-Fenton process was able to degraded compounds already present in HE (Fig. 6). Conclusions The photo-Fenton process was efficient for the degradation of the pharmaceuticals even when using very low concentrations of reagents (Fe(III) and H2O2). The FLU degradation was strongly favored with iron citrate (Fecit) when compared to iron nitrate, while CIP degradation was only slightly improved with iron citrate. The pharmaceutical degradation rates were higher at more acid pH (3.4 and 4.5)when compared to pH5.6. TheHE strongly hindered the degradation of pharmaceuticals in relation to DW, since in this case, the high ratio of carbon toCIP and FLU content resulted in a consumption of hydroxyl radical in side reactions. However, the redox iron cycle was not influenced by the HE content during photo-Fenton process. Despite the recalcitrance 0 20 40 60 80 100 0 5 10 30 40 50 60 70 F e 2 + c o n c e n tr a ti o n ( µM ) Treatment time (min) Fe(NO 3 ) 3 Fecit Fig. 5 Generation of Fe2+ during irradiation of Fecit and Fe(NO3)3 in distilled water (solid symbols) and hospital effluent (open symbols). Experimental conditions: [Fe(NO3)3] = 166 μM, [citrate] = 50 μM, initial pH = 4.5 0 2 4 6 8 10 0.0 5.0x10 5 1.0x10 6 1.5x10 6 2.0x10 6 2.5x10 6 3.0x10 6 3.5x10 6 0.0 5.0x10 5 1.0x10 6 1.5x10 6 2.0x10 6 2.5x10 6 3.0x10 6 3.5x10 6 0 min 90 min 0 min 90 min A r e a ( m U A ) Time (min) FLU b A r e a ( m U A ) a CIP Fig. 6 HPLC-FL chromatograms before and after 90 min photo-Fenton degradation of CIP (a) and FLU (b) by in pre-treated hospital effluent. Experimental conditions: [CIP] = [FLU] = 100 μg L−1, [Fecit] = 2 μM, [H2O2] = 100 μM, initial pH = 4.5 Environ Sci Pollut Res (2017) 24:6233–6240 6239 of the matrix, more than 50 % of pharmaceuticals were degraded within 90 min of experiment by photo-Fenton process. Acknowledgments The authors thank CNPq (Process 308649/2015-0) and FAPESP (2015/21732-5) for support of this work and for scholarship (Process 151022/2014-3) awarded to J.A.L. Perini and CAPES for the scholarship awarded to B. C. Silva. References Abrahamson HB, Rezvani AB, Brushmiller JG (1994) Photochemical and spectroscopic studies of complexes, of iron(III) with citric acid and other carboxylic acids. Inorg Chim Acta 226:117–127 Bader H, Sturzenegger V, Hoigné J (1988) Photometric method for the determination of low concentrations of hydrogen peroxide by the peroxidase catalyzed oxidation of N,N-diethyl-p-phenylenediamine (DPD). Water Res 22:1109–1115 Bongaerts GP, Hoogkamp-Korstanje JA (1993) In vitro activities of BAY Y3118, ciprofloxacin, ofloxacin and fleroxacin against gram- positive and gram-negative pathogens from respiratory tract and soft tissue infections. Antimicrob Agents Chemother 37:2017–2019 Chu S, Metcalfe CD (2007) Analysis of paroxetine, fluoxetine and norfluoxetine in fish tissues using pressurized liquid extraction, mixed mode solid phase extraction cleanup and liquid chromatography-tandem mass spectrometry. J Chromatogr A 1116:112–118 Eldin F, Suliman O, Sultan SM (1996) Talanta 43:559–568 Emerick RW, Manning J, Tchobanoglous G, Darby JL (1999) Impact of bacteria and dosing frequency on the removal of virus within inter- mittently dosed biological filters. Small Flows Q 1(1):36–41 Faust BC, Hoigné J (1990) Photolysis of Fe(III)-hidroxy complexes as sources of OH radicals in clouds, fog and rain. Atmos Environ 24: 79–89 Faust BC, Zeep RG (1993) Photochemistry of aqueous iron (III)- polycarboxylate complexes: role in the chemistry of atmospheric and surface water. Environ Sci Technol 27:2417–2522 Foran CM, Weston J, Slattery M, Brooks BW, Huggett DB (2004) Reproductive assessment of Japanese medaka (Oryzias latipes) fol- lowing a four-week fluoxetine (SSRI) exposure. Arch Environ Contam Toxicol 46:511–517 Fortune WB, Mellon MG (1938) Determination of iron with o- phenanthroline. Ind Eng Chem Anal Ed 10:60–64 Gross M, Mitchell D (1990) Virus removal by sand filtration of septic tank effluent. J Environ Eng 116:711720 Heberer T (2002) Occurrence, fate, and removal of pharmaceutical residues in the aquatic environment: a review of recent research data. Toxicol Lett 131:5–17 Henry TB, Black MC (2008) Acute and chronic toxicity of fluoxetine (selective serotonin reuptake inhibitor) in western mosquitofish. Arch Environ Contam Toxicol 54:325–330 Klamerth N, Malato S, Agüera A, Fernández-Alba A (2013) Photo-Fenton and modified photo-Fenton at neutral pH for the treatment of emerging contaminants in wastewater treat- ment plant effluents: a comparison. Water Res 47:833–840 Kosjek T, Perko S, Zigon D, Heath E (2013) Fluorouracil in the environ- ment: analysis, occurrence, degradation and transformation. J Chromatogr A 1290:62–72 Martínez-Bueno MJ, Agüera A, Gómez MJ, Hernando MD, García- Reyes JF, Fernández-Alba AR (2007) Application of liquid chromatography/quadrupole-linear ion trap mass spectrometry and time-of-flight mass spectrometry to the determination of pharmaceu- ticals and related contaminants in wastewater. Anal Chem 79:9372– 9384 Martins AF, Vasconcelos TG, Henriques DM, Frank CS, König A, Kümmerer K (2008) Concentration of ciprofloxacin in Brazilian hospital effluent and preliminary risk assessment: a case study. Clean 36:264–269 Michael-Kordatou I, Michael C, Duan X, He X, Dionysiou DD, Mills MA, Fatta-Kassinos D (2015) Dissolved effluent organic matter: characteristics and potential implications in wastewater treatment and reuse applications. Water Res 77:213–248 Nogueira RFP, Guimarães JR (2000) Photodegradation of dichloroacetic acid and 2,4-dichlorophenol by ferrioxalate/ H2O2 system. Water Res 34:895–901 Nogueira RFP, Silva MRA, Trovó AG (2005) Influence of the iron source on the solar photo-Fenton degradation of different clas- ses of organic compounds. Sol Energy 79:384–392 Ou X, Quan X, Chen S, Zhang F, Zhao Y (2008) Photocatalytic reaction by Fe(III)–citrate complex and its effect on the photodegradation of atrazine in aqueous solution. J Photochem Photobiol A197:382–388 Pauwels B, Ngwa F, Deconinck S, Verstraete W (2006) Effluent quality of a conventional activated sludge and a membrane bioreactor sys- tem treating hospital wastewater. Environ Technol 27:395–402 Perini JAL, Perez-Moya M, Nogueira RFP (2013) Photo-Fenton degra- dation kinetics of low ciprofloxacin concentration using different iron sources and pH. J Photochem Photobiol A 259:53–58 Pignatello JJ (1992) Dark and photoassisted Fe3+-catalyzed degradation of chlorophenoxy herbicides by hydrogen peroxide. Environ Sci Technol 26:944–951 Rodil R,Quintana JB,Concha-GrañaE, López-Mahía P,Muniategui-Lorenzo S, Prada-Rodríguez D (2012) Emerging pollutants in sewage, surface and drinkingwater inGalicia (NWSpain). Chemosphere 86:1040–1049 Ruppert G, Bauer R, Heisler G, Novalic S (1993)Mineralization of cyclic organic water contaminants by the photo-Fenton reaction— influence of structure and substituents. Chemosphere 27:1339–1347 Santos LHMLM, Gros M, Rodriguez-Mozaz S, Delerue-Matos C, Pena A, Barceló D,MontenegroMCBSM (2013) Contribution of hospital effluents to the load of pharmaceuticals in urban wastewaters: iden- tification of ecologically relevant pharmaceuticals. Sci Total Environ 461-462:302–316 Silva MRA, Trovó AG, Nogueira RFP (2007) Degradation of the herbicide tebuthiuron using solar photo-Fenton process and ferric citrate complex at circumneutral pH. J Photochem Photobiol A: Chem 191:187–192 Soares PA, Batalha M, Souza SMAGU, Boaventura RAR, Vilar VJP (2015) Enhancement of a solar photo-Fenton reaction with ferric-organic ligands for the treatment of acrylic-textile dyeing wastewater. J Environ Manag 152:120–231 Tonon D, Tonetti AL, Coraucci FB, Bueno DAC (2015) Wastewater treatment by anaerobic filter and sand filter: hydraulic loading rates for removing organic matter, phosphorus, pathogens and nitrogen in tropical countries. Ecol Eng 82:583–589 Trovó AG, Melo SAS, Nogueira RFP (2008) Photodegradation of the pharmaceuticals amoxicillin, bezafibrate and paracetamol by the photo-Fenton process— application to sewage treatment plant effluent. J Photochem Photobiol A: Chem 198:215–220 Vázquez MMP, Vázquez PP, Galera MM, García MDG, Uclés A (2013) Ultrasound-assisted ionic liquid dispersive liquid-liquid microextraction coupled with liquid chromatography-quadrupole- linear ion trap-mass spectrometry for simultaneous analysis of phar- maceuticals in wastewaters. J Chromatogr A 1291:19–26 Verlicchi P, Galletti A, PetrovicM, Barceló D (2010) Hospital effluents as a source of emerging pollutants: an overview of micropollutants and sustainable treatment options. J Hydrol 389:416–428 VienoN, Tuhkanen T, Kronberg L (2007) Elimination of pharmaceuticals in sewage treatment plants in Finland. Water Res 41:1001–1012 Yuan S, Jiang X, Xia X, Zhang H, Zheng S (2013) Detection, occurrence and fate of 22 psychiatric pharmaceuticals in psychiatric hospital and municipal wastewater treatment plants in Beijing, China. Chemosphere 90:2520–2525 6240 Environ Sci Pollut Res (2017) 24:6233–6240 Photo-Fenton degradation of the pharmaceuticals ciprofloxacin and fluoxetine after anaerobic pre-treatment of hospital effluent Abstract Introduction Material and methods Reagents Hospital effluent Experimental degradation procedures Solid phase extraction Chemical analysis Results and discussion Effect of iron source, pH, concentration of H2O2 and Fecit Photo-Fenton degradation of CIP and FLU in HE samples Conclusions References