wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 1 3 9 4e1 4 0 2 Avai lab le at www.sc iencedi rect .com journa l homepage : www.e lsev ier . com/ loca te /wat res Degradation of the antibiotic amoxicillin by photo-Fenton process e Chemical and toxicological assessment Alam G. Trovó a,1, Raquel F. Pupo Nogueira a,*, Ana Agüera b, Amadeo R. Fernandez-Alba b, Sixto Malato c aUNESP e Univ Estadual Paulista, Instituto de Quı́mica de Araraquara, CP 355, 14801-970, Araraquara, SP, Brazil b Pesticide Residues Group, University of Almerı́a, 04120 Almerı́a, Spain cPlataforma Solar de Almerı́a, CIEMAT, P.O. Box 22, 04200 Tabernas, Almerı́a, Spain a r t i c l e i n f o Article history: Received 10 May 2010 Received in revised form 12 August 2010 Accepted 25 October 2010 Available online 31 October 2010 Keywords: Pharmaceuticals Wastewater Advanced oxidation processes Hydrolysis LC-TOF-MS Degradation pathway * Corresponding author. Tel.: þ55 16 3301 96 E-mail address: nogueira@iq.unesp.br (R. 1 Present address: UFU e Universidade Fed Uberlândia, MG, Brazil. 0043-1354/$ e see front matter ª 2010 Elsev doi:10.1016/j.watres.2010.10.029 a b s t r a c t The influence of iron species on amoxicillin (AMX) degradation, intermediate products generated and toxicity during the photo-Fenton process using a solar simulator were evaluated in this work. The AMX degradation was favored in the presence of the potassium ferrioxalate complex (FeOx) when compared to FeSO4. Total oxidation of AMX in the presence of FeOx was obtained after 5 min, while 15 min were necessary using FeSO4. The results obtained with Daphnia magna biossays showed that the toxicity decreased from 65 to 5% after 90 min of irradiation in the presence of FeSO4. However, it increased again to a maximum of 100% after 150 min, what indicates the generation of more toxic interme- diates than AMX, reaching 45% after 240 min. However, using FeOx, the inhibition of mobility varied between 100 and 70% during treatment, probably due to the presence of oxalate, which is toxic to the neonates. After 240 min, between 73 and 81% TOC removal was observed. Different pathways of AMX degradation were suggested including the opening of the four-membered b-lactamic ring and further oxidations of the methyl group to aldehyde and/or hydroxylation of the benzoic ring, generating other intermediates after bound cleavage between different atoms and further oxidation to carboxylates such acetate, oxalate and propionate, besides the generation of nitrate and ammonium. ª 2010 Elsevier Ltd. All rights reserved. 1. Introduction 2004; Foti et al., 2009). The antibiotic AMX, a broad spectrum The presence of pharmaceuticals in the environment has been reported as an emerging risk to the biotic environment. Among the pharmaceuticals, special attention is focused on antibiotics, since bacterial resistance and toxic effects on several organisms such as algae and crustaceous have been found not only at high concentrations, but also at low concentrations resulting in chronic effects (Andreozzi et al., 06; fax: þ55 16 3301 6692. F. Pupo Nogueira). eral de Uberlândia, Instit ier Ltd. All rights reserve aminopenicillin antibiotic, widely used in human and veteri- nary medicine, was tested for toxicity on microalgal species (growth inhibition) and found to be nontoxic to Pseudo- kirkneriella subcapitata e Closterium ehrenbergii (EC50 100mg L�1), but showed marked toxicity to the Synechococcus leopolensis (EC50 2 mg L�1) (Andreozzi et al., 2004). This compoundhas been identified in municipal sewage treatment plant effluent at concentration of 13 ng L�1 in Italy (Castiglioni et al., 2006). So, uto de Quı́mica, Av. João Naves de Ávila, 2121, CP 593, 38400-902, d. mailto:nogueira@iq.unesp.br http://www.sciencedirect.com http://www.elsevier.com/locate/watres http://dx.doi.org/10.1016/j.watres.2010.10.029 http://dx.doi.org/10.1016/j.watres.2010.10.029 http://dx.doi.org/10.1016/j.watres.2010.10.029 0 2 4 6 8 10 4.0x106 6.0x106 8.0x106 1.0x107 AMX C4 C3 C2 in te ns ity ( cp s) time (min) C1 Fig. 1 e LC/TOF-MS chromatogram obtained from hydrolysis of AMX solution (10 mg LL1) in water after 90 min at pH 2.5 in the dark. wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 1 3 9 4e1 4 0 2 1395 it is necessary to evaluate new alternatives to prevent water contamination, considering the risks that residual pharma- ceuticals can present to human health and to the environment. The use of solar advanced oxidation processes (AOPs), such as photo-Fenton for the treatment of non-biodegradable and/ or toxic compounds can be an alternative to the conventional processes. Previous works has shown the efficiency of this process on the degradation of different therapeutic classes of pharmaceuticals such as antibiotics, anti-inflammatory and analgesic drugs (Pérez-Estrada et al., 2005, 2007; Shemer et al., 2006; Bautitz and Nogueira, 2007; Trovó et al., 2008, 2009). Some authors have reported fast and effective degradation of AMX under different conditions, such O3, O3/H2O2, H2O2/ UV, TiO2, Fenton and photo-Fenton (Arslan-Alaton and Dogruel, 2004; Arslan-Alaton and Caglayan, 2005; Andreozzi et al., 2005; Elmolla and Chaudhuri, 2009; Mavronikola et al., 2009; Martins et al., 2009). Arslan-Alaton and Dogruel (2004) applied a variety of advanced oxidation processes, O3/OH-, H2O2/UV, Fe2/H2O2, Fe 3þ/H2O2, Fe 2/H2O2/UV, Fe3þ/H2O2/UV, to penicillin formulation effluent, concluding that best results regarding complete removal of active substance AMX were obtained using the photo-Fenton process and alkaline ozon- ation. Although the literature reports the efficiency of AMX degradation by different treatments, no studies have been published on its treatment by solar photo-Fenton, including intermediates generated and evolution of toxicity. AMX – Mw = 365 g mol-1 HO H N NH2 O N S O C H3C CH3 O OH C1 and C2 - Mw HO NH NH2 O HO H Fig. 2 e Products of AMX The aim of the present work was to study the use of the photo-Fenton process for the degradation of AMX using a solar simulator. The study includes the influence of the iron species used, the identification of the intermediate products generated during the process by the use of liquid chroma- tography coupled to time-of-flight mass spectrometry (LC- TOF-MS) and the toxicity assessment. 2. Experimental 2.1. Chemicals All theAMXsolutionswereprepared indistilledwater.AMXwas purchased fromSigmaeAldrichandusedas received.Hydrogen peroxide (30% w/v) (POCH, SA), FeSO4.7H2O and sulphuric acid (POCH, SA), NaOH and bovine liver catalase (SigmaeAldrich) were also used as received. Potassium ferrioxalate (K3Fe (C2O4)3$3H2O), named as FeOx, was prepared and purified as described previously (Hatchard and Parker, 1956) using iron nitrate and potassium oxalate (Mallinckrodt). Ammonium metavanadate (SigmaeAldrich) solution was prepared at a concentration of 0.060 M in 0.36 MH2SO4. All reagents were of analytical grade. HPLC-grade acetonitrile andmethanol (Merck) and formic acid (Fluka) were used for HPLC analyses. 2.2. Hydrolysis, photolysis and photo-Fenton experiments The solutions for hydrolysis and photolysis experiments were prepared by dissolving AMX in distilled water at an initial concentration of 10 mg L�1 (TOC5.3 mg C L�1, natural pH6.2). Hydrolysis experiments were performed in 250 mL beakers at two different pH (2.5 and 6.2). The beakerswere kept in the dark at room temperature for 330 min. The photolysis and photo- Fenton experiments were conducted in a solar simulator (Suntest CPSþ from Heraeus, Germany) equipped with a 1100W xenon arc lamp and special filters restricting trans- mission of light below 290 nm. The lamp was set to minimum intensity (250 W m�2), since under high intensity the inter- mediates generated could be quickly degradedmaking difficult their detection. Pyrex glass vessels (Schott Durand, Germany) providedwith an internal recirculatingwater systemwere used to maintain the internal temperature at 25 � 2 �C. The initial AMX concentration for the photo-Fenton experiments was 50 mg L�1 (TOC ¼ 26.3 mg C L�1). Although this concentration is higher than that found in aqueous = 383 g mol-1 C3 and C4 – Mw = 339 g mol-1 H N S C C CH3 H3C O O OH HO NH NH2 NHS O CCH3 H3C O OH hydrolysis in water. http://dx.doi.org/10.1016/j.watres.2010.10.029 http://dx.doi.org/10.1016/j.watres.2010.10.029 Table 1 e Accurate mass measurements found by LC-ESI-TOF-MS spectra of protonated AMX degradation products and fragmentation ions by hydrolysis in water. Compound Retention time (min) Formula Calculated mass (m/z) Expected Mass (m/z) Error (ppm) *DBE C1 5.563 C16H22N3O6S 384.1222 384.1223 �0,48 7.5 C16H21N3O6NaS 406.1007 406.1043 �8,9 7.5 C16H19N2O6S 367.0952 367.0958 �1,7 8.5 C15H22N3O4S 340.1317 340.1325 2,5 6.5 C15H19N2O4S 323.1057 323.106 �0,94 7.5 C7H13N2O2S 189.0687 189.0692 �2.8 2.5 C6H10NO2S 160.0420 160.0426 �4.2 2.5 C2 6.613 C16H22N3O6S 384.1229 384.1223 1.3 7.5 C16H19N2O6S 367.0964 367.0958 1.5 8.5 C15H22N3O4S 340.1323 340.1325 �0.75 6.5 C15H19N2O4S 323.1063 323.106 0.91 7.5 C7H13N2O2S 189.0694 189.0692 0.92 2.5 C6H10NO2S 160.0435 160.0426 5.1 2.5 C3 7.572 C15H22N3O4S 340.1325 340.1325 �0.16 6.5 C15H21N3O4NaS 362.1158 362.1444 3.6 6.5 C15H19N2O4S 323.1057 323.106 �0.94 7.5 C14H19N2O3S 295.1105 295.111 �2.0 6.5 C9H13N2O3S 229.0642 229.0641 0.26 4.5 C7H13N2O2S 189.0687 189.0692 �2.8 2.5 C9H11N2O 163.0861 163.0865 �3.0 5.5 C7H7O 107.0491 107.0491 �0.39 4.5 AMX 7.897 C16H20N3O5S 366.1114 366.1118 �1.1 8.5 C16H19N3O5NaS 388.0919 388.0937 �4.8 8.5 C16H17N2O5S 349.0849 349.0852 �1.1 9.5 C15H17N2O4S 321.0894 321.0903 3.0 8.5 C10H10NO2S 208.0420 208.0426 �3.2 6.5 C6H10NO2S 160.0433 160.0426 3.9 2.5 C4H4NOS 114.0006 114.0008 1.8 3.5 C4 8.675 C15H22N3O4S 340.1331 340.1325 1.6 6.5 C15H21N3O4NaS 362.1149 362.1144 1.1 6.5 C15H19N2O4S 323.1065 323.106 1.5 7.5 C14H19N2O3S 295.1112 295.111 0.37 6.5 C9H13N2O3S 229.0623 229.0641 �8.0 4.5 C7H13N2O2S 189.0700 189.0692 4.1 2.5 C9H11N2O 163.0869 163.0865 1.9 5.5 C7H7O 107.0502 107.0491 9.9 4.5 * DBE ¼ double-bond equivalency. wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 1 3 9 4e1 4 0 21396 environment, it was chosen to permit an adequate detection of intermediates. The initial concentration of FeSO4.7H2O or FeOx was 0.05 mM, which is below the maximum concentra- tion of iron allowed in wastewater according to Brazilian legislation. Based on previous work (Trovó et al., 2009), the initial H2O2 concentration used was 120 mg L�1 and new additions of H2O2 were made during the experiments according to the consumption of H2O2. The initial pH was adjusted to between 2.5e2.8, the optimum pH range for the photo-Fenton process. 2.3. Chemical analysis Before LC-MS analysis, the solution pH was adjusted to between 6 and 8, and 0.5 mL catalase solution (0.1 g L�1) was added to25mLof sample toquench the reactionandguarantee the absence of hydrogen peroxide before the bioassays. The samples were then filtered through 0.45 mm membranes. The AMX concentrations and intermediates were analysed by liquid chromatography electrospray time-of-flight mass spectrometry (LC-ESI-TOF-MS), in positive ionisation mode, usinganHPLC (AgilentSeries1100) equippedwitha3�250mm reverse-phaseC18 analytical column, 5 mmparticle size (Zorbax SB-C18, Agilent Technologies). A and B mobile phases were acetonitrile and water with 0.1% formic acid, respectively, at a flow rate of 0.4 mLmin�1. The injection volume was 20 mL. A linear gradient progressed from 10% A (initial conditions) to 100% A in 50 min, and was maintained at 100% A for 3 min. A 15 min post-run time back to the initial mobile-phase compo- sitionwas allowedafter each analysis. Under these conditions, AMX retention time was between 7.9 and 8.2 min, with adetection limitof 5mgL�1. ThisHPLCsystemwasconnected to an Agilent MSD time-of-flight mass spectrometer with an electrospray interface operating under the following condi- tions: capillary, 4000V;nebulizer, 40psi;dryinggas, 7.0Lmin�1; gas temperature, 300 �C; skimmer voltage, 60 V; octapole dc1, 36.5 V; octapole rf, 250 V; fragmentor, 190 V. The mineralization of AMX during the experiments was evaluated by measuring the decay of the total organic carbon (TOC)usingaTOCanalyser (ShimadzuTOC5050A)equippedwith http://dx.doi.org/10.1016/j.watres.2010.10.029 http://dx.doi.org/10.1016/j.watres.2010.10.029 0 50 100 150 200 250 0.0 0.2 0.4 0.6 0.8 1.0 time (min) T O C /T O C 0 FeSO 4 FeOx A 0 1 2 3 4 5 6 N (m g L -1) 0 50 100 150 200 250 0 20 40 60 80 100 FeSO 4 - Daphnia m. FeOx - Daphnia m. FeSO 4 - Vibrio f. FeOx - Vibrio f. In hi bi ti on ( % ) time (min) B Fig. 3 e Influence of iron species on the (A) removal of TOC (solid symbols) and release of nitrogen (open symbols), (B) evolution of toxicity during AMX photo-Fenton treatment using Daphnia magna (48 h) and Vibrio fischeri (30 min) bioassays. Initial concentrations: [AMX] [ 50 mg LL1 (26.3 mg LL1 TOC); [Fe2D] [ 0.05 mM; [H2O2] [ 120 mg LL1 (further additions); pH [ 2.5. wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 1 3 9 4e1 4 0 2 1397 an ASI5000 autosampler. Ammonium, carboxylates and NO3 - concentrations were measured with a Dionex DX-600 ion chro- matograph using a Dionex Ionpac AS 11-HC 4 mm � 250 mm column.Thegradientprogramwaspre-runfor5minwith20mM Table 2 e Accurate mass measurements found by LC-ESI-TOF- fragmentation ions by photo-Fenton process in water with sol Compound Retention time (min) Formula Calculat C5 4.187 C16H20N3O6S 3 C16H17N2O6S 3 C15H17N2O5S 3 C12H11N2O4 2 C6H10NO2S 1 C6 4.513 C16H22N3O7S 4 C16H19N2O7S 3 C16H17N2O6S 3 C15H19N2O5S 3 C8H15N2O4S 2 C6H10NO2S 1 C7H8NO 1 C1 5.549 C16H22N3O6S 3 C16H21N3O6NaS 4 C16H19N2O6S 3 C15H22N3O4S 3 C15H19N2O4S 3 C7H13N2O2S 1 C6H10NO2S 1 C7 5.906 C16H20N3O6S 3 C16H17N2O6S 3 C10H10NO3S 2 C7H13N2O2S 1 C6H10NO2S 1 C4H4NOS 1 NaOH, 8minwith 20mMNaOHand 7minwith 35mMNaOH, at a flow rate of 1.5 mLmin�1. The hydrogen peroxide was measured during the experi- ments using the spectrophotometric method employing MS spectra of protonated AMX degradation products and ar simulator. ed mass (m/z) Expected mass (m/z) Error (ppm) *DBE 82.1069 382.1067 0.43 8.5 65.0808 365.0801 1.7 9.5 37.0857 337.0852 1.3 8.5 47.0715 247.0713 0.67 8.5 60.0432 160.0432 �0.16 2.5 00.1170 400.1172 �0.75 7.5 83.0916 383.0907 2.2 8.5 65.0792 365.0801 �2.7 9.5 39.1013 339.1009 1.1 7.5 35.0754 235.0747 2.9 2.5 60.0431 160.0426 2.6 2.5 22.0598 122.0600 �2.0 4.5 84.1222 384.1223 �0,48 7.5 06.1041 406.1043 �0.56 7.5 67.0955 367.0958 �0.91 8.5 40.1309 340.1325 �4.9 6.5 23.1069 323.1060 2.8 7.5 89.0694 189.0692 0.92 2.5 60.0430 160.0426 2.0 2.5 82.1066 382.1067 �0.35 8.5 65.0794 365.0801 �2.1 9.5 24.0370 224.0375 �2.6 6.5 89.0686 189.0692 �3.3 2.5 60.0422 160.0426 �3.0 2.5 14.0011 114.0008 2.5 3.5 (continued on next page) http://dx.doi.org/10.1016/j.watres.2010.10.029 http://dx.doi.org/10.1016/j.watres.2010.10.029 Table 2 (continued). Compound Retention time (min) Formula Calculated mass (m/z) Expected mass (m/z) Error (ppm) *DBE C2 7.010 C16H22N3O6S 384.1223 384.1223 �0.22 7.5 C16H19N2O6S 367.0958 367.0958 �0.095 8.5 C15H22N3O4S 340.1317 340.1325 �2.5 6.5 C15H19N2O4S 323.1063 323.1060 0.91 7.5 C7H13N2O2S 189.0682 189.0692 �5.4 2.5 C6H10NO2S 160.0425 160.0426 �1.1 2.5 C5H8NS 114.0369 114.0371 �2.6 2.5 C4H10NO 88.0755 88.0756 �22 0.5 C3 7.969 C15H22N3O4S 340.1321 340.1325 �1.3 6.5 C15H19N2O4S 323.1049 323.1060 �3.4 7.5 C7H13N2O2S 189.0693 189.0692 0.39 2.5 C9H11N2O 163.0866 163.0865 0.064 5.5 C4 8.909 C15H22N3O4S 340.1324 340.1325 �0.45 6.5 C15H19N2O4S 323.1063 323.1060 0.91 7.5 C14H19N2O3S 295.1100 295.1110 �3.7 6.5 C10H9N2O2S 189.0667 189.0658 4.5 7.5 C9H11N2OS 163.0880 163.0865 8.6 5.5 C8 9.941 C15H20N3O5S 354.1132 354.1118 3.9 7.5 C15H17N2O5S 337.0855 337.0852 0.68 8.5 C15H15N2O4S 319.0744 319.0747 �0.96 9.5 C14H17N2O3S 293.0957 293.0954 0.88 7.5 C9H9N2O3 193.0611 193.0607 1.7 6.5 C6H11N2OS 159.0586 159.0586 �0.38 2.5 C5H8NOS 130.0322 130.0321 0.68 2.5 C7H7O 107.0493 107.0491 1.5 4.5 C9 10.2 C14H20N3O3S 310.1225 310.1219 1.6 6.5 C14H17N2O3S 293.0953 293.0954 �0.48 7.5 C10 10.701 C8H11N2O4S 231.0435 231.0434 0.41 4.5 C8H10N2O4NaS 253.0262 253.0253 3.4 4.5 C6H10NO2S 160.0429 160.0426 1.4 2.5 C5H12NO 102.0909 102.0913 �4.3 0.5 C11 11.89 C16H18N3O9S 428.0754 428.0758 �1.0 9.5 C16H17N3O9NaS 450.0581 450.0577 0.73 9.5 C16H16N3O8S 410.0660 410.0652 1.8 10.5 C15H18N3O7S 384.0852 384.0859 �2.1 8.5 C6H10NO2S 160.0429 160.0426 1.4 2.5 C4H4N3 94.0401 94.0399 1.3 4.5 C12 12.238 C14H20N3O2S 294.1273 294.127 0.76 6.5 C14H17N2O2S 277.1008 277.1005 0.99 7.5 C10H9N2O2 189.0663 189.0658 2.4 7.5 C10H7N2O 171.0549 171.0552 �2.3 8.5 C13 12.71 C6H10NO3S 176.0374 176.0375 �1.1 2.5 C6H8NO2S 158.0266 158.027 �2.7 3.5 C5H8NOS 130.0320 130.0321 �0.86 2.5 C4H8NS 102.0375 102.0371 3.0 1.5 C14 13.704 C16H20N3O5S 366.1116 366.1118 0.60 8.5 C10H11N2O3 207.0760 207.0764 �2.0 6.5 C6H10NO2S 160.0426 160.0426 �0.48 2.5 C5H8NS 114.0369 114.0371 �2.6 2.5 C15 15.86 C16H18N3O8S 412.0811 412.0809 0.45 9.5 C16H17N3O8NaS 434.0630 434.0628 3.3 9.5 C10H9N2O2 189.0670 189.0658 6.1 7.5 C6H10NO2S 160.0426 160.0426 �0.48 2.5 C16 16.382 C16H19N2O7S 383.0911 383.0907 0.91 8.5 C16H17N2O6S 365.0795 365.0801 �1.9 9.5 C15H19N2O5S 339.1011 339.1009 0.53 7.5 C6H10NO2S 160.0431 160.0426 2.6 2.5 * DBE ¼ double-bond equivalency. wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 1 3 9 4e1 4 0 21398 http://dx.doi.org/10.1016/j.watres.2010.10.029 http://dx.doi.org/10.1016/j.watres.2010.10.029 AMX C5 C16 C7 C13 C1 and C2 C11 C3 and C4 C8 C9 C12 C14 C15 C6 C10 HO NH NH2 H N SO C C CH3 H3C HO O O OH H HO N NH2 O N S O C H3C CH3 O OH OH HO NH NH2 HN S O CH3 C O H HO H N NH2 O N S O C H3C CH3 O OH HO HO H N O N S O C H3C CH3 O OH HO OH HO H N NH2 O N S O C H3C CH3 O OH HO NH NH2 NHS O CCH3 H3C O OH HO NH NH2 NHS O CCH3 C O OH O H HO NH NH2 NHS O CH3 H3C HO N NH O N S O C H3C CH3 O OH OH HO HO OH N S O C CH3 C O H O HO H2N HN S C CH3 C O H O HO HO N NH O N S O C H3C CH3 O OH OH HO OH HO H N S C CH3 CH3 O HO HN NH O O HO NH NH2 HN S O C C CH3 CH3 HO O O HO H HO Fig. 4 e Proposed photo-Fenton degradation pathway of AMX in distilled water using different iron species. wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 1 3 9 4e1 4 0 2 1399 ammonium metavanadate (Unicam-II spectrophotometer) as described by Nogueira et al. (2005a). 2.4. Toxicity evaluation The same sample treatment described for the analysis by liquid chromatography was made prior to the toxicity evalu- ation. The toxicity tests were made with commercial bioas- says, Biofix�Lumi-10, based on inhibition of the luminescence emitted by the marine bacteria Vibrio fischeri, and Daphnia magna immobilization (Daphtoxkit F� magna, Creasel, Belgium) as described previously by Trovó et al. (2009). S NH H3C CH3 O OH N S H3C O H2 N HO CH3 m/z: 160 (C6H10NO2S) m/z: 189 (C7H13N2O3S) A B Fig. 5 e Proposed structures for fragments at m/z 160 and m/z 189. 3. Results and discussion 3.1. Hydrolysis and photolysis Several blank experiments were performed at the initial AMX concentration of 10 mg L�1, to assure that the results found during the photocatalytic tests were consistent and not due to hydrolysis and/or photolysis. No decay in AMX concentration during irradiation at natural pH (6.2) after 6 h irradiation was observed. However, 64% of AMX was hydrolyzed after 1.5 h reaching 74% after 5.5 h at pH 2.5 in the dark, while no hydrolysis at pH 6.2 was observed. However, no mineraliza- tion occurred after the same time, which indicates the generation of intermediates formed during hydrolysis at acid medium. Four intermediates were detected using LC/TOF-MS analysis (Fig. 1), not so interesting for the behavior of AMX in the environment but important for interpretation of photo- Fenton results. Them/z of two intermediates detectedwas 340 and of other two 384. The hydrolysis of AMX starts with the opening of the four-membered b-lactam ring and yields the product AMX penicilloic acid withm/z 384 (C1 and C2), which contains a free carboxylic acid group. Then decarboxylation of the free carboxylic acid occurs, generating two intermediates withm/z http://dx.doi.org/10.1016/j.watres.2010.10.029 http://dx.doi.org/10.1016/j.watres.2010.10.029 0 5 10 15 20 25 30 0 1x107 2x107 3x107 4x107 5x107 A re a time (min) C2 C5 C6 C7 C11 C15 FeSO 4 0 10 20 30 40 50 60 0.0 5.0x105 1.0x106 1.5x106 2.0x106 A re a time (min) C3 C4 C8 C9 C10 C12 C13 C14 C16 C17 FeSO 4 0 10 20 30 40 50 0.0 2.0x10 5 4.0x10 5 6.0x10 5 8.0x10 5 1.0x10 6 A re a time (min) C1 C2 C5 C6 C7 C8 C12 C13 C14 C15 C16 FeOx Fig. 6 e Evolution of the intermediates of AMX degradation in water by photo-Fenton process using different iron species. wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 1 3 9 4e1 4 0 21400 340 (C3 and C4) (Fig. 2). These results are in accordance with the work of Nagele and Moritz (2005). However, the products formed are stereoisomer compounds, what is demonstrated by the same fragmenta- tion pattern of the ions for the two different retention times (Table 1). The literature reports a chromatographicmethod for the detection and determination of these stereoisomers (Valvo et al., 1998). The formation of these stereoisomers occurs with opening of the four-membered b-lactam ring, since the carbon between nitrogen and sulphur in the thia- zolidine ring is chiral. In this way, during hydrolysis the formation of the respective stereoisomer occurs generating the products of m/z 340 [C3: amoxilloic acid (5S) and C4: amoxilloic acid (5R)] and the products of m/z 384 [C1: amox- icilloic acid (5S,6R) and C2: amoxicilloic acid (5R,6R)]. 3.2. Photo-Fenton experiments The efficiency of degradation to different classes of organic compounds can be influenced by the iron species (Nogueira et al., 2005b). Therefore, the influence of different iron species (FeOx and FeSO4) on AMX degradation by photo-Fen- ton process was evaluated. A 50 mg L�1 AMX solution in the presence of 0.05 mM of iron and 120 mg L�1 H2O2 was irradi- ated. New additions of H2O2 were made when necessary. The AMX oxidation andmineralizationwas favouredwhen FeOx was used as iron species. Total removal of AMX was observed after 5 min irradiation, while 15 min were necessary using FeSO4 (data not shown). A similar behaviour was observed for the removal of TOC. However total mineraliza- tionwas not observed. Under these conditions, a removal of 73 and 81% of TOC was achieved after 240 min in the presence of FeSO4 and FeOx, respectively (Fig. 3A). The difference was more relevant at earlier stages, with a removal of 50% of TOC after 75 and 25 min in the presence of FeSO4 and FeOx, respectively. Better results of AMX degradation were also observed when FeOx was used in comparison to Fe(NO3)3 (Trovó et al., 2008). The nitrogen in the AMX molecule was released mainly as ammonium, expressed as total nitrogen in Fig. 3A. The total nitrogen concentration released after 240 min corresponded to 100% of the theoretic amount, and this shows that the residual TOC after 240 min did not contain nitrogen (Fig. 3A). No difference on ammonium release rate was observed when different iron species were used. The main carboxylates indentified during irradiation were: acetate, oxalate and propionate. AMX (50 mg L�1) caused 30% inhibition of Vibrio fischeri bioluminescence in bioassay after 30 min exposure. During the photo-Fenton degradation using the different iron species, no significant change in the toxicity was observed (Fig. 3B). However, using Daphnia magna bioassays, 65% inhibition of the neonates mobility (after 48 h of exposure) was observed in the initial AMX solution. After 90 min, the inhibition of mobility of the neonates decreased to 5% using FeSO4. However, it increased again to a maximum of 100% after 150 min achieving 45% at the end of the experiment (240 min) (Fig. 3B). However, using FeOx, the inhibition of mobility varied between 95 and 70% during the treatment, probably due to the presence of oxalate, which is toxic to the neonates. A control experiment in the presence of only FeOx showed 100% of mobility inhibition. The increase of toxicity after 120 min was observed with both FeSO4 and FeOx (Fig. 3B), what coincides with the plateau observed for TOC removal, suggesting that the toxicity is due to carboxylic acids gener- ated during degradation process (Pintar et al., 2004). 3.3. Identification of intermediates and AMX degradation pathway in water by photo-Fenton process in a solar simulator The identification of themain degradation products generated during the photo-Fenton treatment was carried out in order to propose a degradation pathway and assess the intermediates that possibly contribute to the toxicity. Analysis by LC-TOF- MS allowed the detection and identification of sixteen inter- mediates during AMX degradation by photo-Fenton process (Table 2). Themain fragments observed in themass spectra of each intermediate are indicated on Fig. 4. Among these intermediates, fifteen were detected when FeSO4 was used http://dx.doi.org/10.1016/j.watres.2010.10.029 http://dx.doi.org/10.1016/j.watres.2010.10.029 wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 1 3 9 4e1 4 0 2 1401 (C2-C16), while only eleven were detected using FeOx (C1; C2; C5-C8; C12-C16). Starting with AMX, there are three possible pathways for further degradation. The first one is hydroxylation, which can occur in different positions, yielding the intermediates C5, C7, C11, C15 and C16. Fromone (C5 and C7) to four hydroxyl groups (C11) are added to the AMX molecule. The assignation of the exact position of these groups is not always possible by the use of the proposed technique, but the observation of the frag- mentation pattern can help in this task. Thus, all these inter- mediates presented the fragment at m/z 160.0431, which corresponds to the thiazolidine ring (C6H10NO2S; DBE: 2.5) (Fig. 5A). The presence of the same fragment in the mass spectrum of AMX indicates that this moiety remains unchanged and thus the attack of the hydroxyl radicals is expected to occur at the positions in themolecule of AMXmore susceptible to an electrophilic attack, i.e the benzoic ring or the nitrogen atom, with electrons lone pair. The differentiation of the C5 and C7 isomers was possible mainly by the presence of the fragment at m/z 189 in C7 (Fig. 5B). This fragment, which corresponded to the C7H13N2O2S formula (DBE: 2.5) confirmed that the OH attack had no place in the benzoic ring but on the nitrogen atom. This fragment was also present in C15. DBE increase by one unit, observed in C11 and C15 with respect to AMX, was due to oxidation of the amino group to an imino group. This explains the absence of NH3 loss, characteristic of theAMXandtherestofderivativescontaining theaminogroup. Another degradation pathway corresponds to the opening of the four-membered b-lactam ring and yields the stereo- isomers of penicilloic acid, C1 and C2, and a series of deriva- tives (C3, C4, C6, C8, C9 and C12), which do not presented the b-lactam ring. C1 and C2 have been already reported as hydrolysis products. A further decarboxylation reaction yiel- ded the intermediates C3 and C4, which were also detected as hydrolysis products. The oxidation of themethyl groups in the thiazolidine ring was evidenced by the identification of the intermediates C8 and C9. C8 showed an increase in the DBE of a unit with respect to C3/C4, while containing an additional oxygen atom. The presence of fragments at m/z 293 (C14H17N2O3S), m/z 159 (C6H11N2OS) and m/z 130 (C5H8NOS), coming from the decarboxylation and subsequent cleavage of the molecule, confirm this proposal. C9 has been proposed as originated by decarboxylation of C8, however the scarce fragmentation observed in themass spectrum cannot confirm this structure. The bond cleavage between nitrogen of the amino group and the carbonyl group is evidenced by compound C10 (m/z 231.0435; C8H11N2O4S). The bond cleavage reduced the double- bond equivalency (DBE) from 8.5 to 4.5, due to the loss of benzoic ring and the double bond of the carbonyl group. The next intermediate generated (C13), presents a difference of 2 DBE in relation to C10, due to the loss of the CO group and of the four-membered b-lactam ring. This compound could be also generated by the cleavage of C8 molecule. Finally, a third degradation route was through the forma- tion of C14 (amoxicillin diketopiperazine 20,50), an isobaric derivative of AMX generated by the opening of the b-lactam ring with further arrangement of the molecule. Recently, Lammet al., (2009) have reported by the first time the presence of this derivative in wastewater. The different iron species influenced the AMX degradation and the intermediates generated. A higher number of inter- mediates from AMX degradation was observed using FeSO4 when compared to FeOx. Besides, in the presence of FeOx, the intermediates were degraded after 10 min, except the inter- mediate C13 (45 min), while in the presence of FeSO4 between 15 and 60 min were necessary. However, with both iron species, total degradation of the intermediates determined was observed (Fig. 6). The differences detected between intermediates behavior could be rationalized only from a kinetics (and not mechanistically) point of view. The inter- mediates not determined in FeOx were those less abundant (C3, C4, C9, C10 and C11) or quickly appearing and degrading since FeOx was a quicker process than FeSO4 (more evident during the first stages of the treatment). The more persistent products were detected during both treatments. 4. Conclusions Although AMX was more rapidly degraded in the presence of ferrioxalate, the sample presented significant toxicity during all the treatment due to the presence of oxalate. Using FeSO4, significant decrease of toxicity to Daphnia magnawas observed after 90min irradiation (from 65% to 5%), when 53% of the TOC was removed. The residual TOC did not contain nitrogen since 100% of the theoretic amount of the total nitrogen was released after 240 min, mainly as ammonium. Sixteen inter- mediates were identified using HPLC-ESI-TOF, generated after hydroxylation of benzoic ring, opening of b-lactam ring and several other hydroxylation steps. The identification of these intermediates was based on the measurement of the accurate masses of the intermediates and their main fragments. These intermediates were responsible for the toxicity during the treatment, which was then reduced to 45% after 240 min irradiation, when they were converted to short chain carboxylic acids, allowing further conventional treatment. Acknowledgments FAPESP, CAPES, CNPq, Spanish Ministry of Education (Contract CONSOLIDER-INGENIO 2010 CSD2006-00044) and Andalusia Regional Government (Project no. P06-TEP-02329). r e f e r e n c e s Andreozzi, R., Canterino, M., Marotta, R., Paxeus, N., 2005. Antibiotic removal from wastewaters: the ozonation of amoxicillin. Journal of Hazardous Materials 122 (3), 243e250. Andreozzi, R., Caprio, V., Ciniglia, C., Champdoré, M., Giudice, R., Marotta, R., Zuccato, E., 2004. Antibiotics in the environment: occurrence in Italian STPs, fate, and preliminary assessment on algal toxicity of amoxicillin. Environmental Science and Technology 38 (24), 6832e6838. Arslan-Alaton, I., Caglayan, A.E., 2005. Ozonation of procaine penicillin G formulation effluent part I: process optimization and kinetcs. Chemosphere 59 (1), 31e39. http://dx.doi.org/10.1016/j.watres.2010.10.029 http://dx.doi.org/10.1016/j.watres.2010.10.029 wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 1 3 9 4e1 4 0 21402 Arslan-Alaton, I., Dogruel, S., 2004. Pre-treatment of penicillin formulation effluent by advanced oxidation processes. Journal of Hazardous Materials 112 (1e2), 105e113. Bautitz, I.R., Nogueira, R.F.P., 2007. Degradation of tetracycline by photo-Fenton process e Solar irradiation and matrix effects. Journal of Photochemistry and Photobiology A: Chemistry 187 (1), 33e39. Castiglioni, S., Bagnati, R., Fanelli, R., Pomati, F., Calamari, D., Zuccato, E., 2006. Removal of pharmaceuticals in sewage treatment plants in Italy. Environmental Science and Technology 40 (1), 357e363. Elmolla, E.E., Chaudhuri, M., 2009. Degradation of the antibiotics amoxicillin, ampicillin and cloxacillin in aqueous solution by the photo-Fenton process. Journal of Hazardous Materials 172 (2e3), 1476e1481. Foti, M., Giacopello, C., Bottari, T., Fisichella, V., Rinaldo, D., Mammina, C., 2009. Antibiotic resistance of Gram negatives isolates from loggerhead sea turtles (Caretta caretta) in the central Mediterranean Sea. Marine Pollution Bulletin 58 (9), 1363e1366. Hatchard, C.G., Parker, C.A., 1956. A new sensitive chemical actinometer II. Potassium ferrioxalate as a standard chemical actinometer. Proceedings of the Royal Society of London 235 (1203), 518e536. Lamm, A., Gozlan, I., Rotstein, A., Avisar, D., 2009. Detection of amoxicillin-diketopiperazine-2’, 5’ in wastewater samples. Journal of Environmental Science and Health. Part A, Toxic/ Hazardous Substances & Environmental Engineering 44 (14), 1512e1517. Martins, A.F., Mayer, F., Confortin, E.C., Frank, C.S., 2009. A study of photocatalytic processes involving the degradation of the organic load and amoxicillin in hospital wastewater. Clean e Soil, Air, Water 37 (4e5), 365e371. Mavronikola, C., Demetriou, M., Hapeshi, E., Partassides, D., Michael, C., Mantzavinos, D., Kassinos, D., 2009. Mineralisation of the antibiotic amoxicillin in pure and surface waters by articial UVA e and sunlight-induced Fenton oxidation. Journal of Chemical Technology and Biotechnology 84 (8), 1211e1217. Nagele, E., Moritz, R., 2005. Structure elucidation of degradation products of the antibiotic amoxicillin with ion trpa MSn and accurate mass determination by ESI TOF. Journal of the American Society for Mass Spectrometry 16 (10), 1670e1676. Nogueira, R.F.P., Oliveira, M.C., Paterlini, W.C., 2005a. Simple and fast spectrophotometric determination of H2O2 in photo- Fenton reactions using metavanadate. Talanta 66 (1), 86e91. Nogueira, R.F.P., Silva, M.R.A., Trovó, A.G., 2005b. Influence of the iron source on the solar photo-Fenton degradation of different classes of organic compounds. Solar Energy 79 (4), 384e392. Pérez-Estrada, L.A., Malato, S., Agüera, A., Férnandez-Alba, A. R., 2007. Degradation of dipyrone and its main intermediates by solar AOPs. Identification of intermediate products and toxicity assessment. Catalysis Today 129 (1e2), 207e214. Pérez-Estrada, L.A., Maldonado, M.I., Gernjak, W., Agüera, A., Fernández-Alba, A.R., Ballesteros, M.M., Malato, S., 2005. Decomposition of diclofenac by solar driven photocatalysis at pilot plant scale. Catalysis Today 101 (3e4), 219e226. Pintar, A., Besson, M., Gallezot, P., Gibert, J., Martin, D., 2004. Toxicity to Daphnia magna and Vibrio fischeri of Kraft bleach plant effluents treated by catalytic wet-air oxidation. Water Research 38 (2), 289e300. Shemer, H., Kunukcu, Y.K., Linden, K.G., 2006. Degradation of the pharmaceutical metronidazole via UV, Fenton and photo- Fenton processes. Chemosphere 63 (2), 269e276. Trovó, A.G., Melo, S.A., Nogueira, R.F.P., 2008. Photodegradation of the pharmaceuticals amoxicillin, bezafibrate and paracetamol by the photo-Fenton process e application to sewage treatment plant effluent. Journal of Photochemistry and Photobiology A 198 (2e3), 215e220. Trovó, A.G., Nogueira, R.F.P., Aguera, A., Fernandez-Alba, A.R., Sirtori, C., Sixto, M., 2009. Degradation of sulfamethoxazole in water by solar photo-Fenton. Chemical and toxicological evaluation. Water Research 43 (16), 3922e3931. Valvo, L., Ciranni, E., Alimenti, R., Alimonti, S., Draisci, R., Giannetti, L., Lucentini, L., 1998. Development of a simple liquid chromatographic method with UV and mass spectrometric detection for the separation of substances related to amoxicillin sodium. Journal of Chromatography A 797 (1e2), 311e316. http://dx.doi.org/10.1016/j.watres.2010.10.029 http://dx.doi.org/10.1016/j.watres.2010.10.029 Degradation of the antibiotic amoxicillin by photo-Fenton process – Chemical and toxicological assessment Introduction Experimental Chemicals Hydrolysis, photolysis and photo-Fenton experiments Chemical analysis Toxicity evaluation Results and discussion Hydrolysis and photolysis Photo-Fenton experiments Identification of intermediates and AMX degradation pathway in water by photo-Fenton process in a solar simulator Conclusions Acknowledgments References