lable at ScienceDirect Journal of Cleaner Production 142 (2017) 3634e3644 Contents lists avai Journal of Cleaner Production journal homepage: www.elsevier .com/locate/ jc lepro Combination of chemical coagulation, photo-Fenton oxidation and biodegradation for the treatment of vinasse from sugar cane ethanol distillery Carmen S.D. Rodrigues a, Ana R. Neto a, Rose M. Duda b, Roberto A. de Oliveira c, Rui A.R. Boaventura d, Luis M. Madeira a, * a LEPABE e Laborat�orio de Engenharia de Processos, Ambiente, Biotecnologia e Energia, Departamento de Engenharia Química, Faculdade de Engenharia, Universidade do Porto, R. Dr. Roberto Frias, 4200-465, Porto, Portugal b Faculdade de Tecnologia de Jaboticabal, “Nilo St�efani”, Av. Eduardo Zambianchi, 31, 14883-130 Vila Industrial, Jaboticabal, S~ao Paulo, Brazil c Laborat�orio de Saneamento Ambiental, Departamento de Engenharia Rural, Faculdade de Ciências Agr�arias e Veterin�arias, Universidade Estadual Paulista, UNESP, Câmpus de Jaboticabal, Av. Prof. Paulo Donato Castallene, Km 5, 14884-900, Jaboticabal, SP, Brazil d LSRE e Laborat�orio de Processos de Separaç~ao e Reaç~ao, Laborat�orio Associado LSRE-LCM, Departamento de Engenharia Química, Faculdade de Engenharia, Universidade do Porto, R. Dr. Roberto Frias, 4200-465 Porto, Portugal a r t i c l e i n f o Article history: Received 17 June 2016 Received in revised form 29 September 2016 Accepted 20 October 2016 Available online 21 October 2016 Keywords: Sugarcane vinasse Biodegradability Toxicity Coagulation/flocculation Photo-Fenton oxidation Processes combination * Corresponding author. E-mail address: mmadeira@fe.up.pt (L.M. Madeira http://dx.doi.org/10.1016/j.jclepro.2016.10.104 0959-6526/© 2016 Elsevier Ltd. All rights reserved. a b s t r a c t The post-treatment of biodigested sugarcane vinasse was envisaged in the present study following three different approaches: i) coagulation/flocculation e Approach 1; ii) combination of coagulation/floccu- lation with photo-Fenton reaction e Approach 2; iii) coagulation/flocculation followed by biological degradation and photo-Fenton reaction e Approach 3. The coagulation/flocculation per se provided a slight increase in effluent's biodegradability, a decrease in toxicity and turbidity, and provided moderate removals of total organic carbon (TOC), biochemical oxygen demand (BOD5) and chemical oxygen demand (COD). The operating costs associated with chemicals consumption are low (5.7 R$/m3 z 1.4 V/m3). The second and third treatment approaches allowed to reach a nontoxic effluent, improved the biodegradability and led to similar high global organics removal efficiencies. The combined treatment consisting of coagulation/flocculation plus photo-Fenton oxidation (Approach 2) revealed to be a promising solution for treating this wastewater, in order to be recirculated into the high rate anaerobic reactor. Additionally, this solution provides smaller costs (84.9 R$/m3 z 20.2 V/m3) associated with chemicals and energy consumption, which are however very high. Nonetheless, this cost can be reduced by about 40% if solar radiation is used in the photo-Fenton process. © 2016 Elsevier Ltd. All rights reserved. 1. Introduction With the increasing shortage of fossil fuels, strategies have been implemented worldwide stimulating the production of biofuels, namely ethanol. In this scenario, since 1975 Brazil has encouraged the production of ethanol from sugarcane under the Brazilian Alcohol Program “Pro�alcool” (Moraes et al., 2015; Christofoletti et al., 2013), being the country with the highest production worldwide, approximately 26 billion litres in 2012/2013 (Júnior et al., 2014). ). The distillation of sugarcane juice and molasses to produce ethanol generates large quantities of vinasse: about 10e14 L per litre of ethanol produced (Júnior et al., 2015; Orteg�on et al., 2016). This effluent is characterized by strong brown colour, acidic pH, high turbidity, high content of organic compounds (acids, alcohols, aldehydes, ketones, esters and sugars) that contribute for its chemical oxygen demand (COD) and biochemical oxygen demand (BOD5), and high concentration of ions (Kþ, Ca2þ, Mg2þ, Fe3þ, Cl� and SO4 2�) (Barrera et al., 2013; Christofoletti et al., 2013; Marinho et al., 2014; Orteg�on et al., 2016). As vinasse presents high organicmatter and nutrient contents, it has been common practice to use it as fertilizer in sugarcane fields (Júnior et al., 2014; Moraes et al., 2015). The benefits should not be mailto:mmadeira@fe.up.pt http://crossmark.crossref.org/dialog/?doi=10.1016/j.jclepro.2016.10.104&domain=pdf www.sciencedirect.com/science/journal/09596526 http://www.elsevier.com/locate/jclepro http://dx.doi.org/10.1016/j.jclepro.2016.10.104 http://dx.doi.org/10.1016/j.jclepro.2016.10.104 http://dx.doi.org/10.1016/j.jclepro.2016.10.104 C.S.D. Rodrigues et al. / Journal of Cleaner Production 142 (2017) 3634e3644 3635 used as the sole justification for the indiscriminate disposal of vinasse into agricultural fields, so that this practice might be a palliative alternative providing a false impression of solving effi- ciently the management of vinasse (Fuess and Garcia, 2014). In fact, when used in large quantities, the soils become saturated and there is a risk of contamination of surface water and groundwater nearby the fields (Christofoletti et al., 2013; Marinho et al., 2014), causing serious environmental problems. Due to its strong organic matter content (high values of BOD5 and COD), vinasse causes depletion of dissolved oxygen in surface water, lowers the pH, causes the death of aquatic fauna and flora, and introduces contaminants that hinder the use of water for human routines; further, it provides unpleasant odours (Christofoletti et al., 2013). In order to minimize the envi- ronmental impacts of vinasse it is necessary to apply a treatment process prior to discharge into a water body or disposal on land. In general, anaerobic digestion is the process usually adopted (Syaichurrozi et al., 2013; Júnior et al., 2014, 2015; Barrera et al., 2015; Moraes et al., 2015) because this effluent presents high car- bon:nitrogen ratio, leading to methane production, which repre- sents an economic advantage (Wilkie et al., 2000). Anaerobic digestion presents high efficiencies with regards to organic com- pounds removal, however the treated wastewater still contains recalcitrant compounds that are not biodegradable, and are even inhibitory to biological activity. This is a very critical issue because part of the biodigested vinasse is recirculated back into the anaer- obic reactor. The advantage of recirculating the treated vinasse is to reduce the consumption of chemicals for neutralization of the stream fed to the anaerobic digester, taking advantage of the alkalinity of the biodigested vinasse (Barros et al., 2016; Guerreiro et al., 2016). In this context, it is crucial to partially oxidize the recalcitrant compounds present in biodigested vinasse using an efficient pro- cess able to improve its biodegradability and generate a nontoxic effluent. In this perspective, the advanced oxidation processes, and particular the photo-Fenton oxidation, seems to be a good option for that purpose. This process is based on principles that are com- mon to the dark Fenton one, namely the generation of extremely reactive hydroxyl radicals (that will unselectively attack the organic compounds), making use of hydrogen peroxide and a catalyst like ferrous ion in acidic medium (Eq. (1)) (Walling, 1975), but with simultaneous use of ultra-violet/visible radiation that allows the decomposition of hydrogen peroxide into hydroxyl radicals (Eq. (2)) (Sun and Pignatello, 1993). Additionally, there is the production of further HO� (and others) radicals upon Fe2þ regeneration either by Eq. (3), by photolysis of iron (III) hydroxide (Eq. (4)), or by photol- ysis of complexes formed between the organic compounds or their intermediates with Fe3þ (Eq. (5)) (Sun and Pignatello, 1993; Huang et al., 2008). Fe2þ þ H2O2/ Fe3þ þ HO� þ OH� (1) H2O2 þ ln / 2HO� (2) Fe3þ þ H2O2 þ ln /Fe2þ þ HO�þ Hþ (3) Fe(OH)2þ þ ln /Fe2þ þ HO� (4) [Fe(RCO2)]2þ þ ln /Fe2þ þ CO2 þ R� (R ¼ organic compound) (5) As vinasses are characterized by high turbidity that will filter the radiation, a pre-treatment to reduce turbidity is required. One possibility is to use coagulation/flocculation, because this process is easily applied and requires low capital and operating costs, being usually employed as a pre-treatment stage (Zayas et al., 2007). Coagulation/flocculation promotes the destabilization of the colloids by adding coagulants (usually iron or aluminium salts) and their further aggregation with addition of flocculants (polymers). The removal of the particles formed is achieved through filtration or sedimentation (Pang and Abdullah, 2013). Use of iron salt as coagulant might be particularly advantageous as the remaining dissolved metal might be employed as catalyst in a downstream Fenton/photo-Fenton process, thus reducing costs with chemicals. Following the strategy just described, the organic compounds not removed by coagulation/flocculation are then subjected to an advanced oxidation process to be partially oxidized. The subse- quent biological treatment, if required, can be carried out for instance in a sequential batch reactor (SBR) operating in 5 stages per cycle (feeding, reaction, sedimentation, discharge of the clari- fied effluent, sludge purge and idle) (Metcalf and Eddy, 2003). Such configuration presents some advantages comparatively to the conventional biological treatment, namely flexibility and simplicity, low cost and increased resistance to fluctuations in the feeding (Suresh et al., 2011). Additionally, equalization, reaction and clarification occur in the same reactor (EPA, 1999). In the open scientific literature, only one study was found dealing with the treatment of sugarcane vinasse by Fenton or photo-Fenton processes (Hadavifar et al., 2009) and another one about the use of chemical coagulation/flocculation (Zayas et al., 2007) per se. In both cases the processes were tested as stand- alone technologies and no costs data are provided. More recently it was shown that integration of dark Fenton with coagulation/floc- culation could be a promising solution, but costs are high (Guerreiro et al., 2016). The photo-Fenton process allows reaching similar or even better performances, with lower consumptions of chemicals as compared with the dark one. Moreover, if solar radi- ation is used, sustainability (economic and environmental) can be further enhanced. So, the main goal of this work is to reach the best treatment strategy considering different possibilities: coagulation/ flocculation alone e Approach 1, coagulation/flocculation plus photo-Fenton oxidation e Approach 2, and coagulation/floccula- tion combined with biological degradation and photo-Fenton oxidation e Approach 3, in order to improve the biodegradability and decrease the toxicity of the biodigested sugarcane vinasse for its recirculation into high rate anaerobic reactors, associated with the maximum mineralization of organic compounds, at the lowest treatment cost. Other combinations of these technologies have been discarded as discussed along the manuscript. 2. Materials and methods 2.1. Wastewater The vinasse used in this study was collected in a sugar cane ethanol distillery located in Ribeir~ao Preto, S~ao Paulo, Brazil. Firstly, this wastewater is fed to two upflow anaerobic sludge blanket (UASB) reactors in series operating at thermophilic temperature (55 �C) for organic matter degradation and methane production. The UASB reactors have a total volume of 12.1 L and of 5.6 L for first and second, respectively, and operated with hydraulic detention time (HDT) of 24.0 h and 11.1 h, respectively. The resulting effluent, used in this study, was characterized as described in Section 2.3. 2.2. Experimental procedure 2.2.1. Coagulation/flocculation The coagulation/flocculation runs were performed in a Jar Test apparatus (Velp Scientifica, model JLT6) at room temperature (24e28 �C). In each beaker, the coagulant (ferric chloride hexahy- drate - FeCl3.6H2O from LABCHEM) was added to 600 mL of vinasse and then the pHwas adjusted to the desired value using 10MNaOH C.S.D. Rodrigues et al. / Journal of Cleaner Production 142 (2017) 3634e36443636 and/or 1 M H2SO4. Subsequently, the mixture was subjected to rapid stirring (150 rpm) for 3 min and finally to slow stirring (20 rpm) for 15 min; the operating conditions concerning the stirring speed and time were fixed in accordance with studies re- ported in the literature (Eckenfelder, 2000; Satterfield, 2004; Bose, 2010; Poland and Pagano, 2010; Rodrigues et al., 2013a). No floc- culant was added because it was observed in a preliminary study that it did not improve the organic matter removal during the slow stirring stage (Rodrigues et al., 2013a). This way the operating costs are reduced. After coagulation/flocculation the effluent was decanted for 24 h for sedimentation of the flocs formed and separation of the liquid phase. The supernatant was collected and several parameters were measured, namely turbidity, BOD5, COD, TOC, toxicity and dissolved iron (when necessary) as detailed below. 2.2.2. Photo-Fenton's oxidation The photo-Fenton oxidation runs were performed in a 500-mL capacity glass cylindrical batch reactor equipped with a UV/ visible lamp (150 W high mercury vapour pressure lamp from Heraeus, model TQ 150), emitting 500 W/m2 at wavelengths from 200 to 600 nm. The lamp was located axially within an immersion quartz tube connected to a water thermostatic bath (Huber, poly- stat cc1) for temperature control. For each run, 250 mL of biodigested vinasse was used. After reaching the pre-defined temperature (in the range 20e55 �C), the pH was adjusted using 1 M H2SO4 or 1 M NaOH. Then, the required amount of iron (as FeCl3.6H2O from Labchem), when necessary, and H2O2 (30% w/v, from Chem-Lab®), was added. Finally, after the ra- diation source was switched on, the reaction was left to extend up to 3 h. In the case of Approach 2 there was no need to add FeCl3.6H2O because the dissolved iron resulting from the coagula- tion/flocculation stage (258.3 mg/L) was used as catalyst. Normally the catalyst for Fenton's process is Fe(II), added as ferrous sulphate; however, in this study ferric chloride was used to avoid odour and toxicity problems resulting from sulphate reduction to hydrogen sulphide by anaerobic bacteria present in the vinasse. The oxidation reaction proceeded under constant stirring, using a magnetic bar and a magnetic stirrer (Falc). Periodically, samples were taken from the reactor for measuring the total organic carbon (TOC); the reaction in the sampling vials was stopped by adding excess of sodium sulphite (that reacts instantaneously with remaining hydrogen peroxide). As regards to other parameters measured at the end of reaction (3 h), namely COD, BOD5 and toxicity (through inhibition of Vibrio fischeri), the reaction was stopped by eliminating the residual H2O2 and precipitating the iron in solution. For that, the pH was initially raised to 12.0 by the addition of 10 M NaOH, and then the sample was neutralized to pH ~7.0 using conc. H2SO4. The samples for toxicity assessment were neutralized with HCl 1 M as proposed by the analytical method. 2.2.3. Biological degradation The SBR consists of a cylindrical reactor with 2.0 L of capacity, operating for 19 cycles (i.e. up to a pseudo-steady state was reached), each with a duration of 24 h divided as follows: 1 h for feeding, 14 h for reaction, 8 h for sedimentation, 0.8 h and 0.2 h for discharge and idle, respectively. For the first cycle, 0.5 L of activated sludge from the aeration tank of Parada Wastewater Treatment Plant (WWTP) - located in �Aguas Santas, Portugal - with volatile suspended solids (VSS) of ~5 g/L was fed to the reactor. Then, the reactor was fed with 0.5 L of pre-treated effluent after addition of the amount of nitrogen and phosphorous usually required for biological treatment (BOD5:N:P of 100:5:1). In subsequent cycles, the reactor was fedwith 0.4 L of vinasse pre-treated by coagulation/ flocculation. During the reaction stage, the mixed liquor was stirred at 300 rpm using a magnetic bar and a stirrer (Falc). The dissolved oxygen (OD) was maintained at ~3.0 mg O2/L by using a diffused air aeration system coupled to an air pump (Aquamedic Mistral 2000). The biological reactor operated at ambient temperature (average value ~27 �C). COATI timers were used to turn on/off the peristaltic pumps (Gilson M312), the magnetic stirrer and the air pump. The duration of each cycle and each cycle stage as well as the concentration of dissolved oxygen were established in accordance with studies reported in the literature (Sirianuntapiboon and Sansak, 2008; Al-Amrani et al., 2014; Sathian et al., 2014; Rodri- gues et al., 2014a, 2014b; Esteves et al., 2015). 2.3. Analytical methods The inhibition of Vibrio fischeri test was performed according to standard DIN/EN/ISO 11348-3 (International Organization for Standardization, 2005). The bacteria were put in contact with the sample for incubation at 15 �C. After contact times of 5, 15 and 30 min the bioluminescence was measured in a Microtox Analyzer (model 500). The hydrogen peroxide was measured following the method developed by Sellers (1980). It quantifies the yellow-orange colour of the complex formed from the reaction of hydrogenperoxidewith titanium oxalate. For that, the absorbance at 400 nmwas measured in a Thermo Electron Corporation model Helios g spectrophotometer. The dissolved iron was measured by flame atomic absorption spectrometry (Method 3111 B (APHA, 1998)) using an AAS UNICAM spectrophotometer, model 939/959. The samples were firstly filtered through 0.45 mm pore size cellulose nitrate membranes (Filter-Lab). Other parameters were measured according to methods described in Standard Methods (APHA, 1998): the turbidity was evaluated by method 2130 B using a turbidimeter HI88703 from Hanna Instruments; the biological oxygen demand after 5 days (BOD5) was measured according to method D 5210, using an OxiTOP (Velp Scientifica) apparatus; the chemical oxygen demand (COD) was assessed by the closed reflux method (Method 5220 D); the TOC was determined by catalytic combustion at about 680 �C - method 5310 D, using a TC/TOC analyzer (Shimadzu 5000 A apparatus); the total phosphorus was quantified by color- imetry by the ascorbic acid method (Method 4500P e E), after digestion with ammonium persulfate; the total suspended solids (TSS) and volatile suspended solids (VSS) were quantified by gravimetric analysis e Method 2540 B and Method 2540 E, respectively; the total nitrogen was determined by colorimetry (Method D992-71 of ASTM Standards (ASTM, 1973)) after previous digestion (Method 4500 - N C). All analytical determinations were performed in duplicate, and the coefficients of variation were less than 2% for TOC, 6% for BOD5, 4% for COD and inhibition of Vibrio fischeri, and 3% for the other parameters. 3. Results and discussion The biodigested vinasse used in this study shows high levels of organic matter (quantified in terms of COD, TOC and BOD5), turbidity and TSS, low biodegradability (BOD5/COD ratio < 0.4) and some toxicity to Vibrio fischeri (see Table 1), reinforcing the importance of an appropriate post-treatment, for recirculation into anaerobic reactors and/or discharge into water bodies/land disposal. Table 1 Characteristics of the raw biodigested sugarcane vinasse effluent, after coagulation/flocculation (Approach 1) and after photo-Fenton oxidation previously treated by coag- ulation/flocculation (Approach 2), and respective global treatment efficiency (partial removal efficiencies are shown between brackets), for optimized operatory conditions, along with respective operating costs. Parameter Biodigested vinasse After coagulation/flocculation (Approach 1) After Photo-Fenton oxidation (Approach 2) Global treatment efficiency (%) pH 6.2 7.1b 7.0b e COD (mg O2/L) 12,550 7079.3 (43.6) 1153.2 (83.7) 91.0 BOD5 (mg O2/L) 4114 3074.0 (25.3) 795 (74.1) 81.0 TOC (mg C/L) 4202 2460.7 (41.4) 229.4 (90.4) 95.0 BOD5:COD ratio 0.33 0.43 (�) 0.69 (�) e Pt (mg P/L) 15.4 13.7 (11.0) n.d. (�) e Nt (mg N/L) 4.0 2.0 (50.0) n.d. (�) e Turbidity (NTU) 900 6.5 (�) n.d. (�) e TSS (mg/L) 1500 10 (�) n.d. (�) e Vibrio fischeri Inhibition 5 min (%) 45.1 5.6 (�) 0.0 (�) e Vibrio fischeri Inhibition 15 min (%) 46.7 6.3 (�) 0.0 (�) e Vibrio fischeri Inhibition 30 min (%) 47.1 7.7 (�) 0.0 (�) e Operating cost of each step (R$/m3/V/m3) a e 5.7/1.4 79.2/18.9 e nd. e not determined. a Operating costs are related to the consumption of chemicals (the costs of treatment/processing of the sludge generated were not considered). The costs of reagents were those given by OCC Química and Solvay - Per�oxidos do Brasil (Brazil): H2O2 (50.0% (w/v), density at 25 �C ¼ 1.2 g/cm3) e 2.0 R$/kg; FeCl3 (38% (w/w)) e 1.5 R$/kg. Energy cost for Brazilian companies ¼ 0.4 R$/kWh. For converting from Brazilian Real to Euro, it was used the exchange rate of 4.2 R$/V - European Central Bank and Portugal Bank, 07 April 2016 (www.bportugal.pt). b Upon neutralization of the effluent. C.S.D. Rodrigues et al. / Journal of Cleaner Production 142 (2017) 3634e3644 3637 3.1. Coagulation/flocculation (Approach 1) 3.1.1. Effect of pH For optimization during coagulation/flocculation, the pH was adjusted in the range 2.0e6.2 (the highest value corresponding to the pH of the biodigested vinasse). Fig. 1 shows that the maximum TOC (41.4%), BOD5 (25.3%) and COD (43.6%) removals occurred at pH 3, originating an effluent with low turbidity (6.5 NTU) e see Fig. 1b - and increased biodegradability, i.e., BOD5:COD ratio increased from 0.33 in biodigested vinasse to 0.43 after coagula- tion/flocculation. The toxicity decreased at pH 3 but the effluent still presents 5.6e7.7% inhibition of Vibrio fischeri (see Fig. 1b). These results allow establishing the pH of 3.0 as the best one for operating the coagulation/flocculation process. This pH value was the same reached by Guerreiro et al. (2016) for biodigested vinasse with a much lower amount of organic compounds (COD ¼ 6836 mg/L, TOC ¼ 1790 mg/L and BOD5 ¼ 2096 mg/L) comparatively to the one used in this study. At pH 3.0, the iron species present in solution are Fe3þ, Fe(OH)2þand Fe(OH)2þ, so the coagulation/flocculation occurs probably by the charge neutrali- zation mechanism (Duan and Gregory, 2003). 3.1.2. Effect of coagulant dose Once the optimum pH (3.0) was established, the effect of the concentration of Fe3þ (between 125 mg/L and 1000 mg/L) was assessed. It was found that removals of TOC, COD, BOD5 and turbidity increased with the coagulant dose until 500 mg Fe3þ/L, and higher doses did not provide any efficiency improvement (see Fig. 2). The appearance of an optimum coagulant dose can be explained by the charge neutralization theory.When coagulant was added to vinasse, Fe3þ, Fe(OH)2þand Fe(OH)2þ interacted with negative colloids and neutralized their charges, which promoted the colloids destabilization. However, when coagulant was in excess, colloids can absorb the cations and become positively charged, so, the colloids were stable again as a result of electrical repulsion [Verma et al., 2010]. For [Fe3þ] ¼ 500 mg/L the removals of TOC, COD and BOD5 were 41.4, 43.6 and 25.3%, respectively, the BOD5:COD ratio reached 0.43, turbidity was 6.5 NTU and the percent inhibition of Vibrio fischeri decreased to 5.6e7.7%. The coagulant dose (corresponding to 1.45 g FeCl3/L) was previously reported by Guerreiro et al. (2016) as the optimal dose for the treatment of vinasse from sugar cane ethanol distillery after treatment by anaerobic digestion with less quantity of organic matter, and also by Günes (2014) for the treatment of a landfill leachate with high organic matter content, similar to the bio- digested vinasse used in this study. However, Zayas et al. (2007) reported an optimum dose of 20 g FeCl3/L for treating a bio- digested vinasse (although with a slightly smaller COD of 8525 mg/ L). Operating at pH ¼ 3 and [Fe3þ] ¼ 500 mg/L, coagulation/floc- culation reduced the turbidity and the toxicity of effluent and led to a considerable organic matter reduction, but the biodegradability improvement (BOD5:COD ratio increased from 0.33 to 0.43) was very low. For this reason, it becomes important to subject the vinasse to a subsequent treatment. The coagulation is an important preliminary process, with low operating cost (5.7 R$/m3 z 1.4 V/m3) e see Table 1 e mainly associated with iron chloride con- sumption, because the costs of acid and base for pH adjustment are negligible. 3.2. Coagulation/flocculation followed by photo-Fenton reaction (Approach 2) Firstly, the biodigested vinasse was treated by photo-Fenton per se, at temperatures in the range 20e55 �C. Temperatures below 20 �C were not tested because this is the average ambient tem- perature in the region of S~ao Paulo e Brazil at the time of vinasse production (April till November). On the other hand, temperatures higher than 55 �C were also not tested because this is the tem- perature of the vinasse at the outlet of the anaerobic reactor. In addition, the use of higher temperatures would mean a higher treatment cost. Moreover, for temperatures above 60e70 �C the thermal decomposition of H2O2 is intensified, impairing the for- mation of hydroxyl radicals, thereby affecting the oxidation of the organic matter. The experimental results (not shown for brevity reasons) evi- denced that the photo-Fenton process by itself is inefficient, providing very low removals of TOC (9.1%), COD (2.5%) and BOD5 (1.3%) after 3 h. Such low efficiency of the process is associated with the high turbidity of vinasse that impairs the penetration of http://www.bportugal.pt 2 3 4 5 0 20 40 60 2.0 3.0 4.0 6.2 a COD BOD5 R em ov al (% ) pH TOC 0.0 0.1 0.2 0.3 0.4 0.5 BOD5:COD B O D 5:C O D 2 3 4 5 0 20 40 2.0 3.0 4.0 6.2 b 15 min 30 min In hi bi tio n of ir ehcsi f oirbiV (% ) pH 5 min 0 100 200 300 400 500 Turbidity Tu rb id ity (N TU ) Fig. 1. Removals of TOC, COD and BOD5 and biodegradability (a), turbidity and toxicity of the effluent (b) for different pH values in the coagulation/flocculation process - Approach 1 (vcoagulation ¼ 150 rpm, tcoagulation ¼ 3 min, [Fe3þ] ¼ 500 mg/L, vflocculation ¼ 20 rpm, tflocculation ¼ 15 min). 0 250 500 750 1000 0 20 40 60 125 250 500 750 1000 aCOD BOD5 R em ov al (% ) [Fe3+] (mg/L) TOC 0.0 0.1 0.2 0.3 0.4 0.5 BOD5:COD B O D 5:C O D 0 250 500 750 1000 0 20 40 125 250 500 750 1000 b 15 min 30 min In hi bi tio n of ir ehcsif oirbiV (% ) [Fe3+] (mg/L) 5 min 0 50 100 150 200 250 300 350 Turbidity Tu rb id ity (N TU ) Fig. 2. Effect of Fe3þ dose in the removal of TOC COD, BOD5 and BOD5:COD ratio (a), turbidity and toxicity of the effluent (b) in the coagulation/flocculation process - Approach 1 (vcoagulation ¼ 150 rpm, tcoagulation ¼ 3 min, pH ¼ 3.0, vflocculation ¼ 20 rpm, tflocculation ¼ 15 min). C.S.D. Rodrigues et al. / Journal of Cleaner Production 142 (2017) 3634e36443638 radiation in solution. Thus, the coagulation/flocculation as pre- treatment step has an important role in the treatment, because this process reduces the turbidity of the effluent (as shown in Section 3.1). Therefore, the photo-Fenton process was optimized for effluent previously treated by coagulation/flocculation under the best operating conditions found (pH¼ 3 and [Fe3þ]¼ 500mg/L). To improve the biodegradability and generate a non-toxic effluent the photo-Fenton reaction seems to be a good alternative, employing as catalyst the dissolved iron resulting from the previous coagulation/ flocculation stage (258.3 mg Fe/L). Therefore, a parametric study was carried out aiming to find the best dose of H2O2, temperature, pH and irradiance required to maximize the organic compounds removal, reduce the toxicity and improve the effluent biodegrad- ability. This will be dealt with in the following sections. 3.2.1. Influence of temperature The influence of temperature on the process efficiency was evaluated by performing some runs where this parameter was varied in the range 20e55 �C. Fig. 3a shows the histories of mineralization at different temperatures, with best results being reached at the highest temperature tested of 55 �C. The final COD and BOD5 removals, the BOD5:COD ratio, as well as the efficiency of hydrogen peroxide use (determined as the ratio between the TOC conversion and H2O2 consumption e XTOC:XH2O2) follow the same tendency of TOC reduction; the best results were achieved at 55 �C (see Fig. 3b and Table 2) with 88, 80, and 70% for TOC, COD and BOD5 removals, respectively, 0.64 for the BOD5:COD ratio and 0.88 for H2O2 efficiency. Regarding the toxicity of the effluent, quantified by the inhibi- tion of Vibrio fischeri, it should be noted that at 20 �C there was an increased toxicity compared to the effluent after coagulation/floc- culation (from 5.6-7.7% to 12.5e19.5% - see Table 2), which can be due to the formation of some oxidation by-products, more toxic than those coming from coagulation/flocculation. For the other 0 20 40 60 80 100 120 140 160 180 0 10 20 30 40 50 60 70 80 90 100 a ) %(lavo me R C OT t (min) T = 20 ºC T = 30 ºC T = 40 ºC T = 55 ºC 20 30 40 50 0 20 40 60 80 100 20 30 40 50 b COD BOD5 R em ov al (% ) T (ºC) TOC 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 BOD5:COD B O D 5:C O D Fig. 3. Influence of temperature on TOC removal along the reaction (a) and removal of TOC, COD and BOD5 and BOD5:COD ratio after 180 min of photo-Fenton reaction (b) after the coagulation/flocculation step - Approach 2 ([H2O2] ¼ 14.5 g/L, pHinitial ¼ 3.0, [Fe3þ] ¼ 258.3 mg/L, I ¼ 500 W/m2). C.S.D. Rodrigues et al. / Journal of Cleaner Production 142 (2017) 3634e3644 3639 temperatures tested, the effluent showed no toxicity e Table 2. The optimum value of temperature (55 �C) is the same as reported by Guerreiro et al. (2016) when treating a similar effluent by dark Fenton oxidation. The total operating costs related to the consumption of chem- icals (hydrogen peroxide and iron chloride used in the coagulation/ flocculation step) and energy (associated to the lamp used) are presented in Table 2. The energy costs associated with the effluent heating were considered to be null (because the natural effluent temperature ranges from 20 to 55 �C), whereby the total cost is the same for all temperatures tested as the other operating parameters remain constant. It must be emphasized that the energy cost considerably contributes (36.0 R$/m3 z 8.6 V/m3) to the overall process cost. The H2O2 consumption is the second variable that most influences the treatment cost. Using 14.5 g/L of H2O2, the residual iron concentration from the coagulation step, and an irradiance of 500W/m2, the total cost amounts to 105.6 R$/m3 (25.1 V/m3). 3.2.2. Effect of pH To evaluate the effect of pH four experiments were performed, varying this parameter in the range 1.5e4.0. All the other variables were fixed: [H2O2]¼ 14.5 g/L, T¼ 55 �C and irradiance¼ 500W/m2 (corresponding to a lamp power of 150 W). Fig. 4a presents the results of the TOC removal along the reaction time. It is possible to observe, for all pH values tested, that the TOC reduction is fast in the first 60 min of reaction, then the oxidation rate slightly decreases till 120min of reaction and finally TOC remains nearly constant. The maximum mineralization (and also COD and BOD5 removals), as well as BOD5:COD and XTOC: XH2O2 ratios, were reached at pH 3 with 70, 80 and 88% removals for BOD5, COD and TOC, respectively, 0.64 for BOD5:COD ratio and 0.88 for XTOC:XH2O2 ratio e see Fig. 4b and Table 2. Thus, pH 3 can be considered as the optimal value; the optimum pH for a photo-Fenton system usually lies in the range 2e3, regardless of the type of effluent to be treated (Wu et al., 1999). As a matter of fact, the pH affects the amount of Fe3þ pre- sent in solution. On one hand, for pH values higher than 4 the formation of insoluble iron oxides occurs reducing the fraction of Fe3þ dissolved (Kwon et al., 1999). On the other hand, in very acidic pH values the complex Fe(OH)2þ is present in a reduced amount, and the two more photo reactive compounds (Fe(OH)(H2O)52þ and Fe(OH)2(H2O)4þ) are also present in smaller quantities to generate the radicals and regenerate the catalyst. However, the Fe(H2O)63þ species is present in a greater amount but it has lower absorptivity, which limits the absorption of radiation (Safarzadeh-Amiri et al., 1996). Moreover, pH < 2.5 allows to occur the scavenging reac- tion between the hydroxyl radical and Hþ (HO� þ Hþ þ e�/H2O) (Spinks and Woods, 1990). Regarding toxicity data, evaluated by the percentage of Vibrio fischeri inhibition, these are shown in Table 2. Therewas an increase from 5.6 to 7.7% (biodigested vinasse toxicity recorded after coag- ulation/flocculation) to 12.0e21.3% after photo-Fenton reaction at pH 1.5 and 6.6e9.5% at pH 2 (see Table 2). For the other pH values tested - 3 and 4 - the effluent showed no toxicity. The increase of toxicity for pH 1.5 and 2 may be again due to the formation of by- products, more toxic than those present in the biodigested vinasse after coagulation/flocculation. 3.2.3. Effect of the initial H2O2 dose To evaluate the influence of the initial dose of hydrogen peroxide, five runs were carried out using the following H2O2 concentrations: 0, 4.5, 7.0, 9.0 and 14.5 g/L. In the first experiment the photo-Fenton process was applied to biodigested vinasse after coagulation/flocculation without addition of H2O2 to highlight the importance of its application to improve process efficiency. The effect of the amount of oxidant in TOC removal along the reaction time and in the BOD5:COD ratio, TOC, COD and BOD5 re- movals at the end of the run are presented in Fig. 5 a and b, respectively. It is possible to observe that the reaction without addition of oxidant ([H2O2] ¼ 0 g/L) achieved minor removal effi- ciencies (9, 3 and 1% for TOC, COD and BOD5, respectively) and a BOD5:COD ratio of 0.44. The resulting effluent showed toxicity, quantified in terms of percent inhibition of Vibrio fischeri (4.6% after 5 min of contact with the bacteria; 6.7% after 15 min and 8.1% after 30 min e see Table 2). These values are similar to those achieved in biodigested vinasse after coagulation/flocculation (5.6e7.7% - see Table 1). The presence of H2O2 led to an improvement of the process efficiency up to a H2O2 dose of 9 g/L. Such concentration allowed obtaining final removals of 74, 84 and 90% for BOD5, COD and TOC, Table 2 Conditions employed in the photo-Fenton oxidation runs (Approach 2), hydrogen peroxide consumption and efficiency of use, toxicity of the final effluent and operating costs for the photo-Fenton step. In bold are highlighted optimum conditions in each set of the parametric study. Run T (�C) pH Fe 3þ (mg/L) H2O2 (g/L) I (W/m2) H2O2consumed (g/L) X H2O2 (%) XTOC:XH2O2 Inhibition of Vibrio fischeri @5min/15 min/30 min (%) Cost (R$/m3/V/m3) T (�C) 20 3.0 258.3 14.5 500 14.5 100 0.60 12.5/17.3/19.5 105.6/25.1 30 14.5 100 0.77 0.0/0.0/0.0 40 14.5 100 0.84 0.0/0.0/0.0 55 14.5 100 0.88 0.0/0.0/0.0 pH 55 1.5 258.3 14.5 500 14.5 100 0.60 12.0/16.6/21.3 105.6/25.1 2.0 14.5 100 0.81 6.6/8.5/9.5 3.0 14.5 100 0.88 0.0/0.0 4.0 14.5 100 0.76 0.0/0.0 [H2O2] (g/L) 55 3.0 258.3 0 500 e e e 4.6/6.7/8.1 36.0/8.6 4.5 4.5 14.5 0.40 0.0/0.0 57.6/13.7 7.0 7.0 14.5 0.69 0.0/0.0 69.6/16.6 9.0 9.0 14.5 0.90 0.0/0.0 79.2/18.9 14.5 14.5 100 0.88 0.0/0.0 105.6/25.1 I (W/m2) 55 3.0 258.3 9.0 27.3 8.8 98.7 0.41 0.0/0.0 45.2/10.8 220 8.9 99.8 0.66 0.0/0.0 59.0/14.1 500 9.0 100 0.90 0.0/0.0 79.2/18.9 C.S.D. Rodrigues et al. / Journal of Cleaner Production 142 (2017) 3634e36443640 respectively, and improving considerably the biodegradability (BOD5:COD ratio increased from 0.43, after coagulation/floccula- tion, to 0.69, after photo-Fenton). The optimal oxidant dose (9 g/L) also proved to be more efficient as regards the oxidant use, i.e. led to a higher value of XTOC:XH2O2 (0.90) e Table 2. Such optimal dose is similar to what was reached by Primo et al. (2008) (10 g/L) for treating a landfill leachate by the photo-Fenton process with less organic matter (TOC ¼ 2120.3 mg/L, COD ¼ 3828.8 mg O2/L and BOD5 ¼ 680 mgO2/L). The existence of an optimal concentration of hydrogen peroxide has been reported in other photo-Fenton studies (Primo et al., 2008; Barnd~ok et al., 2016; Rodríguez-Chueca et al., 2016), and is explained by parallel reactions between H2O2 in excess and the hydroxyl species (HO� þ H2O2/ H2O þ HO2� (Walling, 1975)), whereby a less amount of radicals is present in solution to oxidize the organic compounds. Regarding toxicity (% inhibition of Vibrio fischeri), for all oxidant concentrations tested it was null e see Table 2. The costs associated with this process increased with the oxidant dose (see Table 2), as expected. The cost for the optimal concentration of H2O2 was 79.2 R$/m3 (corresponding to z18.9 V/m3). 3.2.4. Influence of irradiance For evaluating the effect of irradiance three runs were per- formed varying this parameter in the range 27.3e500 W/m2. This range was established taking into account the incidence of solar radiation (13.1e204 W/m2) in the State of S~ao Paulo and the average value in Brazil (277 W/m2). It was also decided to test the irradiance of 500 W/m2, which is the maximum value given by the mercury lamp used, in order to extrapolate the results to other sites exposed to higher solar radiation, where wastewater with organic loads similar to the biodigested vinasse used in this study needs an adequate treatment. The other parameters were fixed at: [H2O2] ¼ 9.0 g/L, pH ¼ 3.0 and T ¼ 55 �C. The experimental results shown in Fig. 6 demonstrated that the organic matter removal, biodegradability (BOD5:COD ratio) and efficiency of oxidant use (XTOC:XH2O2) e in Table 2 e increased with irradiance. This fact is explained by favouring the radicals forma- tion and regeneration of the ferrous ion when increasing the in- tensity of the incident radiation (Rodrigues et al., 2013b; Vaishnave et al., 2014). The toxicity of the final effluent was null for all irra- diances tested (Table 2). The global cost of the photo-Fenton process increases with the irradiance. For the maximumvalue of irradiance used - 500W/m2 - the total cost amounts to 79.2 R$/m3 (z18.9 V/m3), as reported in the previous section. 3.3. Coagulation/flocculation plus SBR and photo-Fenton process (Approach 3) In order to reduce the effluent treatment cost, after coagulation (conducted at the best operating conditions found in Approach 1) the biodigested vinasse was subjected to aerobic biological degra- dation, since coagulation increased the biodegradability from 0.33 to 0.43 and reduced the toxicity from 45.1-47.1% to 5.6e7.7% (see Table 1). Moreover, by decreasing the organic matter content by biodegradation less amount of oxidant is needed to the photo- Fenton process (last step in this approach). 3.3.1. Biological degradation The biological treatment was carried out in a SBR operated in accordance with the description provided in Section 2.2.3. The aerobic biological degradation was not applied to treat the vinasse, because it had been subjected to an anaerobic treatment, and then all biodegradable organic compounds have been removed. During the treatment of vinasse by the SBR (after coagulation/ flocculation), the pH, temperature and DO were measured in each cycle, and average values of 8.8, 27.2 �C and 3.0 mgO2/L, respec- tively, were obtained. During SBR operation, it was found that the removal of organics and nutrients increased along the first 9 cycles (see Fig. 7a) due to the biomass acclimatization to the effluent that is being fed to the reactor, and then remained almost constant. The average removal values, after reaching the pseudo-steady state (19 cycles), were as follows: 37% - COD, 40% - BOD5, 46% - TOC, 34% - total nitrogen and 18% - total phosphorus. The TSS content in the effluent discharged from the biological reactor declined along the first seven cycles (see Fig. 7b) and then remained practically con- stant with an average value of 1021 mg/L, which means that after 7 cycles a better sedimentation of the sludge occurred. The VSS content, inside the reactor (which gives idea of the growth/death biomass), had a very accentuated decay in the first four cycles, continuing to decaymore slowly in the four subsequent cycles, thus indicating that there is lysis and mineralization of biomass. After 8 cycles the VSS content remained nearly constant (average value of 3633 mg/L), indicating that biomass death is compensated by biomass growth. The operating cost of the SBR was considered to be only 0 20 40 60 80 100 120 140 160 180 0 10 20 30 40 50 60 70 80 90 100 a ) %(lavo me R C OT t (min) pH = 1.5 pH = 2.0 pH = 3.0 pH = 4.0 1 2 3 4 0 20 40 60 80 100 1.5 2.0 3.0 4.0 b COD BOD5 R em ov al (% ) pH TOC 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 BOD5:COD B O D 5:C O D Fig. 4. Effect of pH on TOC removal along the reaction (a) and removal of TOC, COD and BOD5 and BOD5:COD ratio after 180 min of photo-Fenton reaction (b) after the coag- ulation/flocculation step - Approach 2 ([H2O2] ¼ 14.5 g/L, T ¼ 55 �C, [Fe3þ] ¼ 258.3 mg/ L, I ¼ 500 W/m2). Fig. 5. Effect of oxidant concentration on TOC removal along the reaction (a) and removal of TOC, COD and BOD5 and BOD5:COD ratio after 180 min of reaction (b) after the coagulation/flocculation step - Approach 2 (T ¼ 55 �C, pHinitial ¼ 3.0, [Fe3þ] ¼ 258.3 mg/L, I ¼ 500 W/m2). C.S.D. Rodrigues et al. / Journal of Cleaner Production 142 (2017) 3634e3644 3641 associated with energy consumption for agitation (power required ¼ 0.26 W) and air insufflation (power required ¼ 4.5 W). Assuming an average energy cost of 0.4 R$/kWh (z0.10 V/kWh), the cost associated with the biological reactor operation was 10.7 R$/m3 (z2.5 V/m3) e see Table 3. 3.3.2. Photo-Fenton oxidation In viewof the high amount of organic compounds still present in the effluent from the biological degradation (SBR), it was further treated by photo-Fenton oxidation. Parameters like iron and hydrogen peroxide concentration and irradiance were optimized. The temperature and pH were fixed at optimum values (55 �C and 3, respectively) found for this process in Approach 2. 3.3.2.1. Influence of hydrogen peroxide concentration. The hydrogen peroxide concentration to be used in the photo-Fenton process was optimized by varying it in the range 2.5e9.0 g/L. In Figure S1a is shown the TOC removal history data, evidencing that performance increased up to the oxidant dose of 7.0 g/L. The same trend is observed for overall COD and BOD5 removal and final BOD5:COD ratio (see Figure S1b and Table 4). As explained in Approach 1, this optimal concentration of oxidant exists due to the fact that the excess of hydrogen peroxide reacts with hydroxyl radicals (scav- enging effect). The hydrogen peroxide dose of 7.0 g/L provided an effluent with BOD5:COD ¼ 0.69 and XTOC:XH2O2 ¼ 0.78; the toxicity of the effluent is null for all concentrations tested e see Table 4. Again, as expected, the operating costs for this process increased with the hydrogen peroxide dose (see Table 4). For the optimal dose (7.0 g/L) the cost amounts 72.6 R$/m3 (z17.3 V/m3). 3.3.2.2. Effect of iron dose. Given that the effluent is neutralized before being fed to the SBR, and consequently precipitation of iron occurs, it was needed to add and optimize the amount of catalyst in Fig. 6. Influence of irradiance on TOC removal along the reaction (a) and removal of TOC, COD and BOD5 and BOD5:COD ratio after 180 min of photo-Fenton reaction (b) after the coagulation/flocculation step - Approach 2 (T ¼ 55 �C, pHinitial ¼ 3.0, [H2O2] ¼ 9.0 g/L, [Fe3þ] ¼ 258.3 mg/L). Fig. 7. Variation of TOC, COD, BOD5, total nitrogen and phosphorus removal (a) and TSS and VSS (b) during 19 cycles of SBR operation, after effluent treatment by coagulation/ flocculation - Approach 3 (T ~ 27 �C, D.O. ~3.0 mg O2/L and vagitation ¼ 300 rpm). C.S.D. Rodrigues et al. / Journal of Cleaner Production 142 (2017) 3634e36443642 the photo-Fenton process. So, its effect was evaluated by carrying out four experiments where this parameter was varied in the range 129e517 mg/L. The results presented in Figure S2 and Table 4 show that TOC, COD and BOD5 removals, as well as biodegradability and efficiency of the oxidant use, increased with the iron concentration up to 258.3 mg/L. For higher catalyst doses the process efficiency decreased. The existence of an optimal iron dose is explained by the occurrence of radical scavenging reactions with the excess iron (Fe2þ þ HO� / Fe3þ þ OH�). The toxicity of the treated vinasse was null for all iron concentrations tested, as presented in Table 4. As regards operating costs, there is only a slight increase derived from the iron used in the process (see Table 4). For the iron concentration of 258.3 mg/L the cost was 72.6 R$/m3 (z17.3 V/m3). 3.3.2.3. Influence of irradiation. As in Approach 1, the effect of irradiance was evaluated from the results obtained in three runs carried out under irradiances varying in the range 27.3e500W/m2. The organic matter reduction, biodegradability and efficiency of H2O2 use increased with irradiance (see Figure S3 and Table 4) as expected, since the formation of hydroxyl radicals and catalyst regeneration are favoured by the increase of the irradiance. Regarding the inhibition of Vibrio fischeri, it was 0.0% for all tested irradiations. For the maximum irradiance (500 W/m2) removals of 78.0% for TOC and COD, and 64.0% for BOD5, BOD5:COD ¼ 0.69 and XTOC:XH2O2 ¼ 0.78 were achieved. The cost associated with energy consumption increased with irradiance (see Table 4), being very high as compared to hydrogen peroxide and particularly iron costs. For the maximum irradiance (500 W/m2) the total operating cost was 72.6 R$/m3 (z17.3 V/m3). The global removal efficiency of the treatment process (coagu- lation followed by aerobic biological degradation and photo-Fenton oxidation) was 92.0, 98.0 and 84.0% for COD, TOC and BOD5, respectively (see Table 3) with an associated operating cost of 89.0 R$/m3 (z21.2 V/m3). However, the treatment cost can be reduced to 52.9 R$/m3 (z12.6 V/m3), i.e. by about 40%, if solar radiation is used in the photo-Fenton process. 4. Conclusions Given the high organic load and the presence of refractory and toxic compounds in the sugar cane vinasse after anaerobic diges- tion, a treatment is needed before recirculation to the anaerobic reactors. Thus, this study evaluated the treatability of biodigested Table 3 Characteristics of the biodigested sugarcane vinasse effluent after each treatment step of Approach 3 and respective global efficiency (partial removal efficiencies are shown between brackets), for optimized operatory conditions, along with respective operating costs. Parameter After coagulation/flocculation After SBR After Photo-Fenton Global efficiency (%) pH 7.1b 8.8 7.1b e COD (mg O2/L) 7079.3 (43.6) 4468.9 (37.4) 983.1 (78.0) 92.0 BOD5 (mg O2/L) 3074.0 (25.3) 1879.8 (40.0) 678.0 (64.0) 84.0 TOC (mg C/L) 2460.7 (41.4) 1345.1 (45.7) 296.6 (78.0) 98.0 BOD5:COD ratio 0.43 (�) 0.42 (�) 0.69 (�) e Pt (mg P/L) 13.7 (11.0) 105.4 (33.6a) n.d. (�) e Nt (mg N/L) 2.0 (50.0) 25.8 (17.7a) n.d. (�) e Turbidity (NTU) 6.5 (�) n.d. (�) n.d. (�) e TSS (mg/L) 10 (�) 1021 (�) n.d. (�) e Vibrio fischeri Inhibition 5 min (%) 5.6 (�) n.d. 0.0 (�) Vibrio fischeri Inhibition 15 min (%) 6.3 (�) n.d. 0.0 (�) e Vibrio fischeri Inhibition 30 min (%) 7.7 (�) n.d. 0.0 (�) e Operating cost of each step (R$/m3/V/m3) c 5.7/1.4 10.7/2.5 72.6/17.3 e nd. e not determined. a Calculated from total phosphorus and nitrogen in effluent after coagulation/flocculation and after adding phosphate buffer and urea (17.6 mg P/L and 154.7 mgN/L). b Upon neutralization of the effluent. c Operating costs are related with consumption of chemicals (it was not considered the costs of treatment/processing of sludge generated). The costs of reagents were those given by OCC Química and Solvay - Per�oxidos do Brasil (Brazil): H2O2 (50.0% (w/v), density at 25 �C ¼ 1.2 g/cm3) e 2.0 R$/kg; FeCl3 (38% (w/w)) e 1.5 R$/kg. Energy cost for Brazilian companies ¼ 0.4 R$/kWh. For converting from Brazilian Real to Euro, it was used the exchange rate of 4.2 R$/V - European Central Bank and Portugal Bank, 07 April 2016 (www.bportugal.pt). Table 4 Conditions employed in the photo-Fenton oxidation runs (Approach 3), hydrogen peroxide consumption and efficiency of use, toxicity of the final effluent and total operating costs. In bold are highlighted optimum conditions in each set of the parametric study. Run T (�C) pH Fe 3þ (mg/L) H2O2 (g/L) I (W/m2) H2O2consumed (g/L) X H2O2 (%) XTOC:XH2O2 Inhibition of Vibrio fischeri @ 5min/15 min/30 min (%) Cost (R$/m3/V/m3) [H2O2] (g/L) 55 3.0 258.3 2.5 500 2.5 100 0.47 0.0/0.0/0.0 51.0/12.1 4.5 4.5 100 0.60 0.0/0.0/0.0 60.6/14.4 7.0 7.0 100 0.78 0.0/0.0/0.0 72.6/17.3 9.0 9.0 100 0.63 0.0/0.0/0.0 82.2/19.6 [Fe3þ] (mg/L) 55 3.0 129.0 7.0 500 7.0 100 0.44 0.0/0.0/0.0 71.1/16.9 193.0 7.0 100 0.63 0.0/0.0/0.0 71.8/17.1 258.3 7.0 100 0.78 0.0/0.0/0.0 72.6/17.3 517.0 7.0 100 0.65 0.0/0.0/0.0 75.5/18.0 I (W/m2) 55 3.0 258.3 7.0 27.5 7.0 100 0.44 0.0/0.0/0.0 38.5/9.2 220 7.0 100 0.62 0.0/0.0/0.0 52.4/12.5 500 7.0 100 0.78 0.0/0.0/0.0 72.6/17.3 C.S.D. Rodrigues et al. / Journal of Cleaner Production 142 (2017) 3634e3644 3643 sugar cane vinasse by three different strategies: i) chemical coag- ulation/flocculation per se (Approach 1), ii) coagulation/floccula- tion followed by photo-Fenton reaction (Approach 2) and iii) coagulation/flocculation, aerobic biological degradation in a SBR and photo-Fenton reaction (Approach 3). Approach 1 led to low organic matter removal efficiency; however the efficiency increased when vinasse was treated by combined processes (Ap- proaches 2 and 3). Removals higher than 80% for BOD5 and higher than 90% for the COD and TOC, a nontoxic effluent with increased biodegradability (BOD5:COD ratio ¼ 0.69) were achieved in Approach 2 and 3. However, the operating cost of treatment in Approach 3 (89.0 R$/m3 e z21.2 V/m3) is slightly higher than in Approach 2 (84.9 R$/m3 e z20.2 V/m3). Both costs are high, but can reduced by a factor of around 40% if solar photo-Fenton oxidation is employed. The combination of coagulation/flocculation with photo-Fenton oxidation (Approach 2) or biological degradation and photo-Fenton oxidation (Approach 3) increased the biodegradability and elimi- nated the toxicity of the wastewater, which allows to recirculate it into the anaerobic reactor, but the effect of the recirculation on the biogas production has to be evaluated. Acknowledgments This work was financially supported by: Project POCI-01-0145- FEDER-006939 (Laboratory for Process Engineering, Environment, Biotechnology and Energy e LEPABE) funded by FEDER funds through COMPETE2020 - Programa Operacional Competitividade e Internacionalizaç~ao (POCI) e and by national funds through FCT - Fundaçaeo para a Ciência e a Tecnologia. Carmen Rodrigues is grateful to LEPABE for financial support through the Postdoctoral grant. RoseMaria Duda is grateful to Ibero American Youth Scholarship Program teachers and researchers SANTANDER University. Roberto Alves de Oliveira is grateful to the Ministry of Education e Coor- denaç~ao de Aperfeicoamento de Pessoal de Nivel Superior (CAPES - Process PNPD-3137/2010) and the Ministry of Science and Tech- nology e Conselho Nacional de Desenvolvimento Cientifico e Tec- nologico (CNPq - Process 483118/2011-7 and 456426/2014-0) for the financial support to build, operate and monitor anaerobic re- actors for treating sugarcane vinasse. The authors are thankful to the WWTP of Parada for providing the aerobic biological sludge for BOD5 analyses and SBR inocculation. http://www.bportugal.pt C.S.D. Rodrigues et al. / Journal of Cleaner Production 142 (2017) 3634e36443644 Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jclepro.2016.10.104. References Al-Amrani, W.A., Lim, P.-E., Seng, C.-E., Ngah, W.S.W., 2014. Factors affecting bio- decolorization of azo dyes and COD removal in anoxiceaerobic REACT oper- ated sequencing batch reactor. J. Taiwan Inst. Chem. Eng. 45, 609e616. APHA, AWWA, WEF, 1998. Standard Methods for the Examination of Water and Wastewater, twentieth ed. American Public Health Association, American Water Works Association, Water Pollution Control Federation, Washington DC. ASTM, 1973. Annual Book of ASTM Standards - Part 23-Water; Atmospheric Anal- ysis. American Society of Testing and Materials, Philadelphia. Barnd~ok, H., Blanco, L., Hermosilla, D., Blanco, A., 2016. Heterogeneous photo- Fenton processes using zero valent iron microspheres for the treatment of wastewaters contaminated with 1,4-dioxane. Chem. Eng. J. 284, 112e121. Barrera, E.L., Spanjers, H., Dewulf, J., Romero, O., Rosa, E., 2013. The sulfur chain in biogas production from sulfate-rich liquid substrates: a review on dynamic modeling with vinasse as model substrate. J. Chem. Technol. Biotechnol. 88, 1405e1420. Barrera, E.L., Spanjers, H., Solon, K., Amerlinck, Y., Nopens, I., Dewulf, J., 2015. Modeling the anaerobic digestion of cane molasses vinasse: extension of the Anaerobic Digestion Model No. 1 (ADM1) with sulfate reduction for a very high strength and sulfate rich wastewater. Water Res. 71, 42e54. Barros, V.G., Duda, R.M., De Oliveira, R.A., 2016. Biomethane production from vinasse in UASB reactors inoculated with granular sludge. Braz. J. Microbiol. 47, 628e639. Bose, P., 2010. Water and Wastewater Engineering. Indian Institute of Technology, Kanpur, India. Christofoletti, C.A., Escher, J.P., Correia, J.E., Marinho, J.F.U., Fontanetti, C.S., 2013. Sugarcane vinasse: environmental implications of its use. Waste Manag. 33, 2752e2761. Duan, J., Gregory, J., 2003. Coagulation by hydrolysing metals salts. Adv. Colloid Interface Sci. 100e102, 475e502. Eckenfelder Jr., W.W., 2000. Industrial Water Pollution Control, thrid ed. McGraw- Hill International Editions, Singapore. EPA, 1999. U.S. Wastewater, Technology Fact Sheet: Sequencing Batch Reactors. U.S. Environmental Prottection Agency, Office of Water, EPA, Washington D.C. 932- F-99e073. Esteves, B.M., Rodrigues, C.S.D., Boaventura, R.A.R., Maldonado-H�odar, F.J., Madeira, L.M., 2015. Coupling of acrylic dyeing wastewater treatment by het- erogeneous Fenton oxidation in a CSTR with biological degradation in a SBR. J. Environ. Manag. 166, 193e203. Fuess, L.T., Garcia, M.L., 2014. Implications of stillage land disposal: a critical review on the impacts of fertigation. J. Environ. Manag. 145, 210e229. Guerreiro, L.F., Rodrigues, C.S.D., Duda, R.M., Oliveira, R.A., de Boaventura, R.A.R., Madeira, L.M., 2016. Treatment of sugarcane vinasse by combination of coag- ulation/flocculation and Fenton's oxidation. J. Environ. Manag. 181, 237e248. Günes, E., 2014. Seasonal characterization of landfill leachate and effect of seasonal variations on treatment processes of coagulation/flocculation and adsorption. Pol. J. Environ. Stud. 23, 1155e1163. Hadavifar, M., Zinatizadeh, A.A., Younesi, H., Galehdar, M., 2009. Fenton and photo- Fenton treatment of distillery effluent and optimization of treatment conditions with response surface methodology. Asia Pac. J. Chem. Eng. 5, 454e464. Huang, Y.-H., Huang, Y.-F., Chang, P.-S., Chen, C.-Y., 2008. Comparative study of oxidation of dye-Reactive Black B by different advanced oxidation processes: Fenton, electro-Fenton and photo-Fenton. J. Hazard. Mater. 154, 655e662. International Organization for Standardization, 2005. Water Quality - Determina- tion of the Inhibitory Effect of Water Samples on the Light Emission of Vibrio Fischeri (Luminescent Bacteria Test) - Part 3: Method Using Freeze-dried Bacteria. Júnior, A.D.N.F., Wenzel, J., Etchebehere, C., Zaiat, M., 2014. Effect of organic loading rate on hydrogen production from sugarcane vinasse in thermophilic acido- genic packed bed reactors. Int. J. Hydrogen Energy 39, 16852e16862. Júnior, A.D.N.F., Etchebehere, C., Zaiat, M., 2015. Mesophilic hydrogen production in acidogenic packed-bed reactors (APBR) using raw sugarcane vinasse as sub- strate: influence of support materials. Anaerobe 34, 94e105. Kwon, G., Lee, D.S., Kand, N., Yoon, J., 1999. Characteristics of p-chlorophenol oxidation by Fenton's reagent. Water Res. 33, 2110e2118. Marinho, J.F.U., Correia, J.E., Marcato, A.C., Pedro-Escher, J., Fontanetti, C.S., 2014. Sugar cane vinasse in water bodies: impact assessed by liver histopathology in tilapia. Ecotox. Environ. Safe 110, 239e245. Metcalf, Eddy, 2003. Wastewater Engineering: Treatment, Disposal and Reuse, fourth ed. McGraw e Hill, New York. Moraes, B.S., Zaiat, M., Bonomi, A., 2015. Anaerobic digestion of vinasse from sug- arcane ethanol production in Brazil: challenges and perspectives. Renew. Sust. Energy Rev. 44, 888e903. Orteg�on, G.P., Arboleda, F.M., Candela, L., Tamoh, K., Valdes-Abellan, J., 2016. Vinasse application to sugar cane fields. Effect on the unsaturated zone and ground- water at Valle del Cauca (Colombia). Sci. Total Environ. 539, 410e419. Pang, Y.L., Abdullah, A.Z., 2013. Current status of textile industry wastewater management and research progress in Malaysia: a review. Clean. e Soil, Air, Water 41, 751e763. Primo, O., Rivero, M., Ortiz, I., 2008. Photo-Fenton process as an efficient alternative to the treatment of landfill leachates. J. Hazard. Mater 153, 834e842. Poland, J., Pagano, T., 2010. Jar Testing. Civil Engineering Department, Virginia Polytechnic Institute and State University, USA. Rodrigues, C.S.D., Madeira, L.M., Boaventura, R.A.R., 2013a. Treatment of textile dye wastewaters using ferrous sulphate in a chemical coagulation/flocculation process. Environ. Technol. 34, 719e729. Rodrigues, C.S.D., Madeira, L.M., Boaventura, R.A.R., 2013b. Optimization and eco- nomic analysis of textile wastewater treatment by photo-Fenton process under artificial and simulated solar radiation. Ind. Eng. Chem. Res. 52, 13313e13324. Rodrigues, C.S.D., Madeira, L.M., Boaventura, R.A.R., 2014a. A new strategy in opti- mization of integration of advanced and biological oxidation processes for treat textile dyeing wastewaters. Ind. Eng. Chem. Res. 53, 2412e2421. Rodrigues, C.S.D., Madeira, L.M., Boaventura, R.A.R., 2014b. Synthetic textile dyeing wastewater treatment by integration of advanced oxidation and biological processes ePerformance analysis with costs reduction. J. Environ. Chem. Eng. 2, 1027e1039. Rodríguez-Chueca, J., Amor, C., Fernandes, J.R., Tavares, P.B., Lucas, M.S., Peres, J.A., 2016. Treatment of crystallized-fruit wastewater by UV-A LED photo-Fenton and coagulation/flocculation. Chemosphere 145, 351e359. Safarzadeh-Amiri, A., Bolton, J.R., Carter, S.R., 1996. The use of iron in advanced oxidation processes. J. Adv. Oxid. Technol. 1, 18e26. Sathian, S., Rajasimman, M., Radha, G., Shanmugapriya, V., Karthikeyan, C., 2014. Performance of SBR for the treatment of textile dye wastewater: optimization and kinetic studies. Alex. Eng. J. 53, 417e426. Satterfield, Z., 2004. Jar Testing. NESC, West Virginia University, USA. Sellers, R.M., 1980. Spectrophotometric determination of hydrogen peroxide using potassium titanium (IV) oxalate. Analyst 105, 950e954. Sirianuntapiboon, S., Sansak, J., 2008. Treatability studies with granular activated carbon (GAC) and sequencing batch reactor (SBR) system for textile wastewater containing direct dyes. J. Hazard. Mater 159, 404e411. Spinks, J.W.T., Woods, R.J., 1990. An Introduction to Radiation Chemistry, third ed. John Wiley & Sons, New York. Sun, Y., Pignatello, J.J., 1993. Photochemical reactions involved in the total miner- alization of 2-4-D by Fe3þ/H2O2/UV. Technol. 27, 304e310. Suresh, S., Tripathi, R.K., Rana, M.N.G., 2011. Review on treatment of industrial wastewater using sequential batch reactor. Intern. J. Sci. Technol. Manag. 2, 64e84. Syaichurrozi, I., Budiyono, Sumardiono, S., 2013. Predicting kinetic model of biogas production and biodegradability organic materials: biogas production from vinasse at variation of COD/N ratio. Bioresour. Technol. 149, 390e397. Vaishnave, P., Kumar, A., Ameta, R., Punjabi, P.B., Ameta, S.C., 2014. Photo oxidative degradation of azure-B by sono-photo-Fenton and photo-Fenton reagents. Arab. J. Chem. 7, 981e985. Verma, S., Prasad, B., Mishra, I.M., 2010. Pretreatment of petrochemical wastewater by coagulation and flocculation and the sludge characteristics. J. Hazard. Mater 178, 1055e1064. Walling, C., 1975. Fenton's reagent revisited. Acc. Chem. Res. 8, 125e131. Wilkie, A.C., Riedesel, K.J., Owens, J.M., 2000. Stillage characterization and anaerobic treatment of ethanol stillage from conventional and cellulosic feedstocks. Biomass Bioenerg. 19, 63e102. Wu, K., Xie, Y., Zhao, J., Hidaka, H., 1999. Photo-Fenton degradation of a dye under visible light irradiation. J. Mol. Catal. A Chem. 144, 77e84. Zayas, T., Romero, V., Salgado, L., Meraz, M., Morales, U., 2007. Applicability of coagulation/flocculation and electrochemical processes to the purification of biologically vinasse effluent. Sep. Purif. Technol. 57, 270e276. http://dx.doi.org/10.1016/j.jclepro.2016.10.104 http://dx.doi.org/10.1016/j.jclepro.2016.10.104 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref1 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref1 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref1 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref1 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref1 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref2 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref2 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref2 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref3 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref3 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref4 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref4 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref4 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref4 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref4 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref5 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref5 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref5 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref5 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref5 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref6 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref6 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref6 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref6 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref6 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref7 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref7 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref7 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref7 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref8 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref8 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref9 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref9 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref9 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref9 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref10 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref10 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref10 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref10 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref11 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref11 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref12 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref12 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref12 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref12 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref13 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref13 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref13 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref13 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref13 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref13 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref14 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref14 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref14 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref15 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref15 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref15 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref15 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref16 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref16 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref16 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref16 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref17 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref17 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref17 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref17 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref18 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref18 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref18 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref18 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref19 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref19 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref19 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref19 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref20 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref20 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref20 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref20 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref21 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref21 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref21 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref21 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref22 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref22 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref22 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref23 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref23 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref23 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref23 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref24 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref24 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref24 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref25 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref25 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref25 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref25 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref26 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref26 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref26 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref26 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref26 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref27 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref27 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref27 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref27 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref27 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref28 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref28 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref28 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref29 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref29 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref30 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref30 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref30 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref30 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref31 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref31 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref31 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref31 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref32 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref32 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref32 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref32 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref33 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref33 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref33 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref33 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref33 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref33 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref34 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref34 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref34 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref34 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref35 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref35 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref35 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref36 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref36 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref36 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref36 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref37 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref38 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref38 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref38 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref39 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref39 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref39 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref39 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref40 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref40 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref40 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref41 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref41 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref41 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref41 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref41 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref41 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref42 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref42 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref42 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref42 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref43 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref43 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref43 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref43 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref44 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref44 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref44 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref44 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref45 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref45 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref45 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref45 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref46 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref46 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref47 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref47 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref47 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref47 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref48 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref48 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref48 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref49 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref49 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref49 http://refhub.elsevier.com/S0959-6526(16)31725-5/sref49 Combination of chemical coagulation, photo-Fenton oxidation and biodegradation for the treatment of vinasse from sugar cane ... 1. Introduction 2. Materials and methods 2.1. Wastewater 2.2. Experimental procedure 2.2.1. Coagulation/flocculation 2.2.2. Photo-Fenton's oxidation 2.2.3. Biological degradation 2.3. Analytical methods 3. Results and discussion 3.1. Coagulation/flocculation (Approach 1) 3.1.1. Effect of pH 3.1.2. Effect of coagulant dose 3.2. Coagulation/flocculation followed by photo-Fenton reaction (Approach 2) 3.2.1. Influence of temperature 3.2.2. Effect of pH 3.2.3. Effect of the initial H2O2 dose 3.2.4. Influence of irradiance 3.3. Coagulation/flocculation plus SBR and photo-Fenton process (Approach 3) 3.3.1. Biological degradation 3.3.2. Photo-Fenton oxidation 3.3.2.1. Influence of hydrogen peroxide concentration 3.3.2.2. Effect of iron dose 3.3.2.3. Influence of irradiation 4. Conclusions Acknowledgments Appendix A. Supplementary data References