You are on page 1of 12

Environmental Pollution 259 (2020) 113796

Contents lists available at ScienceDirect

Environmental Pollution
journal homepage: www.elsevier.com/locate/envpol

Treatment of biodigested coffee processing wastewater using Fenton’s


oxidation and coagulation/flocculation*
Valciney Gomes de Barros a, b, Carmen S.D. Rodrigues c, *, Wilmar Alirio Botello-Sua rez a, b,
Rose Maria Duda a, d, Roberto Alves de Oliveira a, Eliana S. da Silva e, Joaquim L. Faria e,
Rui A.R. Boaventura e, Luis M. Madeira c
a rio de Saneamento Ambiental, Departamento de Engenharia Rural, Faculdade de Ci^
Laborato encias Agrarias e Veterina
rias, Universidade Estadual Paulista,
UNESP, Ca^mpus de Jaboticabal, Av. Prof. Paulo Donato Castallene, Km 5, 14884-900, Jaboticabal, SP, Brazil
b
Programa de Po s-Graduaça~o em Microbiologia Agropecua ria, Faculdade de Ci^ rias e Veterina
encias Agra rias, Universidade Estadual Paulista (UNESP),
Campus de Jaboticabal, Jaboticabal, SP, Brazil
c
LEPABE - Laboratory for Process Engineering, Environment, Biotechnology and Energy, Faculty of Engineering, University of Porto, Rua Dr. Roberto Frias,
4200-465, Porto, Portugal
d
Faculdade de Tecnologia de Jaboticabal, “Nilo St
efani”, Av. Eduardo Zambianchi, 31, 14883-130, Vila Industrial, Jaboticabal, Sa~o Paulo, Brazil
e
Laboratory of Separation and Reaction Engineering e Laboratory of Catalysis and Materials (LSRE-LCM), Faculdade de Engenharia, Universidade do Porto,
Rua Dr. Roberto Frias, 4200-465 Porto, Portugal

a r t i c l e i n f o a b s t r a c t

Article history: Biodigested coffee processing wastewater (CPW) presents a high organic load and does not meet the
Received 4 July 2019 limits imposed by legislation (namely in Brazil) for discharge into water bodies. Anaerobic digestion
Received in revised form generally cannot provide a satisfactory organic matter reduction in CPW as a significant fraction of
22 October 2019
recalcitrant compounds still persists in the treated effluent. So, this study aims to find alternative ways to
Accepted 10 December 2019
remove refractory organic compounds from this wastewater in order to improve the biodegradability and
Available online 13 December 2019
reduce the toxicity, which will allow its recirculation back into the anaerobic digester. Three treatment
approaches (Fenton’s oxidation - Approach 1, Coagulation/flocculation (C/F) - Approach 2, and the
Keywords:
Coffee processing wastewater
combination of C/F with Fenton’s process - Approach 3) were selected to be applied to the biodigested
Fenton oxidation CPW in order to achieve that objective.
Coagulation-flocculation The application of the Fenton process under the optimal operating conditions (initial pH ¼ 5.0;
Biodegradability T ¼ 55  C, [Fe3þ] ¼ 1.8 g L1 and [H2O2] ¼ 9.0 g L1) increased the biodegradability (the BOD5:COD ratio
Toxicity raised from 0.34 ± 0.02 in biodigested CPW to 0.44 ± 0.01 after treatment) and eliminated the toxicity
(0.0% of Vibrio fischeri inhibition) along with moderate removals of organic matter (51.3%, 55.7% and
39.7% for total organic carbon e TOC, chemical oxygen demand e COD and biochemical oxygen demand
- BOD5, respectively). The implementation of a coagulation/flocculation process upstream from Fenton’s
oxidation, under the best operating conditions (pH 10e11 and [Fe3þ] ¼ 250 mg L1), also allowed to
slightly increase the biodegradability (from 0.34 to 0.47) and reduce the toxicity, whereas providing a
higher removal of organic matter (TOC ¼ 76.2%, COD ¼ 76.5 and BOD5 ¼ 66.3% for both processes
together). Approach 1 and Approach 3 showed to be the best ones, implying similar operating costs
(~74 R$ m3/~17 V m3) and constitute an attractive option for managing biodigested CPW.
© 2019 Elsevier Ltd. All rights reserved.

1. Introduction residues of the coffee industry, which is generated in large amounts


during post-harvest procedures, mainly by the application of the
Coffee processing wastewater (CPW) represents one of the main wet processing technology (Alves et al., 2017). The uncontrolled
disposal of this effluent is of great concern in coffee-producing
countries, since it shows high concentration of suspended or-
*
ganics like sugars, pectins, proteins and polyphenols, and generates
This paper has been recommended for acceptance by Charles Wong.
* Corresponding author.
adverse effects on the receiving water bodies (Dadi et al., 2018).
E-mail address: csdr@fe.up.pt (C.S.D. Rodrigues). This scenario led to the development of economically attractive and

https://doi.org/10.1016/j.envpol.2019.113796
0269-7491/© 2019 Elsevier Ltd. All rights reserved.
2 V. Gomes de Barros et al. / Environmental Pollution 259 (2020) 113796

easily adaptable treatment processes aiming at an efficient reduc- combination with biological degradation (Rodrigues et al., 2017a),
tion of the polluting potential of CPW. as mentioned below.
Anaerobic digestion (AD) is the most widely used technology for The coagulation/flocculation (C/F) process can be applied alone
treating CPW. This treatment process provides several benefits or coupled to Fenton’s oxidation. This process is relatively mature,
including generation of methane (renewable energy) along with cost-effective and user-friendly (GilPavas et al., 2017; Guerreiro
low sludge production (Beyene et al., 2014). However, even with the et al., 2016), and consists in the destabilization of the colloidal
application of high-rate anaerobic reactors such as upflow anaer- particles, which is achieved by using a coagulant (Sharp et al.,
obic sludge blanket (UASB) reactors, anaerobic digestion generally 2006). The destabilization of the colloidal particles can occur by
cannot provide a satisfactory organic matter reduction as a signif- four different mechanisms: i) inclusion in a precipitate, when a
icant fraction of recalcitrant compounds still persists in the treated high concentration of metal is added to an alkaline medium
effluent (Bruno and Oliveira, 2013; Campos et al., 2013; Villa- forming an hydroxide that traps the colloidal particles, ii) adsorp-
Montoya et al., 2017), namely lignin, tannins and humic acid that tion and neutralization of charge that occurs when using positively
are resistant to biological degradation (Zayas et al., 2007). charged coagulants that have a high affinity to colloid surfaces, iii)
The presence of recalcitrant compounds in the biodigested double layer compression resulting from the addition of salts to the
effluent is very undesirable when part of the treated CPW is solution, which increases the ionic strength that leads to a double
intended to be recirculated back into the anaerobic reactor (Sua rez ion bed compression of the colloids, thereby reducing the repulsion
et al., 2018). Actually, this strategy is adopted to benefit from the potential allowing them to aggregate, and iv) adsorption and
alkalinity of biodigested CPW, then reducing the consumption of bridging between particles by the addition of polyelectrolytes to
chemicals to raise the pH of CPW fed to the anaerobic digester and/ the solution, which can adsorb the colloids and form bridges with
or to maintain an appropriate organic loading rate in the system, each other (Duan and Gregory, 2003). Usually, in wastewater
especially for wastewaters with high concentration of organic treatment by coagulation, the destabilization of colloidal particles
matter (Barros et al., 2017; Sua rez et al., 2018). In order to improve occurs by adsorption and neutralization of charge and/or inclusion
the quality of the biodigested CPW, and therefore allowing its in a metal precipitate.
recirculation back to the digester, subsequent post-treatment The microflocs formed are aggregated by the addition of a
strategies must be considered. flocculant, generating particles with large size that can be removed
Previous studies have been successfully focused on advanced by sedimentation, flotation or filtration. In this process, aluminum
oxidation processes, such as Fenton oxidation, for the treatment of and iron salts are the most widely used as coagulants. However,
different effluents containing biologically recalcitrant compounds, aiming at reducing costs with chemicals, the application of iron
providing biodegradability improvement at low treatment cost salts is advantageous when dissolved iron resulting from C/F acts as
(Lucas and Peres, 2009; Nieto et al., 2011; Yalfani et al., 2009). This a catalyst in a subsequent Fenton process (Rodrigues et al., 2017a).
oxidation process involves different species of iron (namely ferrous The main goal of this work was to assess some possibilities of
ion) reacting with H2O2 to generate other reactive oxygen species, biodigested CPW treatment in order to reduce the concentration of
especially the hydroxyl radical (HO) (Eq. (1)) (Gil Pavas et al., 2019; recalcitrant organic compounds, thus improving the biodegrad-
Kantar et al., 2019; Pignatello et al., 2006), which oxidizes organics ability and reducing the toxicity, making possible the recirculation
contained in the effluent and generates intermediate compounds of the treated wastewater to the anaerobic reactors (this way
(Eq. (2)) that are further converted, hopefully to carbon dioxide and increasing the methane production). For that, three strategies were
water (complete mineralization e Eq. (3)) (Guerreiro et al., 2016). adopted, namely Fenton’s oxidation (Approach 1), coagulation/
flocculation (C/F) - (Approach 2), and combination of C/F plus
Fe2þ þ H2O2 / Fe3þ þ HOþ OH (1) Fenton’s oxidation (Approach 3). Up to the author’s knowledge, in
the scientific literature there are no reports on biodigested CPW
HOþ organics /H2O þ intermediates (2) treatment by Fenton’s oxidation and C/F, as well as by the combi-
nation of both processes. Moreover, the treatment costs, which is a
HOþ intermediates /H2O þ CO2 (3) crucial issue for industrial application, were never taken into ac-
count when selecting the best treatment approach.
During the oxidation process, Fe2þ regeneration occurs through
the reaction of Fe3þ with H2O2 (Eq. (4)) to give continuity to the 2. Materials and methods
previous reactions. In addition, other competitive and undesirable
reactions between the hydroxyl radical and excess of oxidant and/ 2.1. Biodigested coffee processing wastewater
or catalyst can occur in parallel (Eqs. (5) and (6)) (Kantar et al.,
2019; Walling, 1975), whereby a less amount of hydroxyl radicals The raw coffee processing wastewater (CPW) used in this study
is present in solution to oxidize the organic compounds (Rodrigues was collected in a coffee-producing farm in the Sa ~o Paulo State,
et al., 2017a). Pedregulho (Brazil), during the harvest period (AprileJuly) of 2017
and the respective characteristics are shown in Table 1. After pH
Fe3þ þ H2O2 / Fe2þ þ HO2þ Hþ (4) adjustment to ~7.0, the influent (with COD ¼ 28500 mg O2 L1) was
treated in a 20 L capacity continuously operated up-flow anaerobic
HOþ H2O2 / H2O þ HOO (5) sludge blanket (UASB) reactor with the dimensions presented
elsewhere (Villa-Montoya et al., 2017) e see UASB1), fed at an
HOþ Fe2þ / Fe3þþ OH (6) organic loading rate of 18.2 g COD (L d)1, operating at a hydraulic
retention time of 30 h and up-flow velocity of 0.038 m h1. Oper-
The Fenton process is considered to be environmentally friendly ating this way, stable conditions were achieved in the anaerobic
because non-toxic reagents are employed; in addition, it operates digestion process and the methane production rate was 2.2 L CH4
at ambient temperature and atmospheric pressure, is easy to d1. The volatile solids content in the sludge was 40 g L1, with a
operate and implies low capital cost. In addition, it is effective in sludge loading rate of 0.2 g COD (g VS d)1.
removing the organic pollutants or reducing their toxicity and The discharge effluent of this reactor (biodigested CPW) was
increasing the biodegradability of effluents allowing the treated using three different approaches: i) Fenton’s oxidation, ii)
V. Gomes de Barros et al. / Environmental Pollution 259 (2020) 113796 3

Table 1
Characteristics of raw and biodigested CPW, after Fenton’s oxidation (Approach 1), C/F process (Approach 2) and C/F followed by Fenton’s process (Approach 3) - removal
efficiencies are shown between brackets e and global removals for the combined processes (Approach 3) with standard errors for a confidence level of 95% and, the respective
operating costs.

Raw CPW Biodigested After Fenton (Approach Approach 3 Discharge


Parameter CPW 1) Limits
After C/F (Approach 2) After Fenton Global Removal
(%)

pH 4.3 ± 0.3 6.9 ± 0.1 7.0 ± 0.2a 7.0 ± 0.1a 7.0 ± 0.2a n.d. 6.0e9.0 b

COD (mgO2 L1) 28500 ± 1780 4263 ± 170 1889 ± 85 (55.7 ± 0.5%) 2613 ± 100 1003 ± 80 76.5 ± 0.3 180 b
(38.7 ± 0.2%) (62.2 ± 0.3%)
BOD5 (mgO2 L1) 13000 ± 700 1387 ± 31 816 ± 25 (39.7 ± 1.8%) 846 ± 24 (39.0 ± 1.0%) 467 ± 10 (44.9 ± 0.6%) 66.3 ± 0.8 10 c

TOC (mgC L1) n.d. 1447 ± 5 706 ± 6 (51.3 ± 1.2%) 907 ± 8 (37.3 ± 0.6%) 345.2 ± 7 76.2 ± 0.7
(61.8% ± 0.8)
BOD5:COD 0.46 ± 0.01 0.34 ± 0.02 0.44 ± 0.01 0.32 ± 0.02 0.47 ± 0.01 n.d.
b
Turbidity (NTU) n.d. 506 ± 6.0 n.d. 21.4 ± 2.0 n.d. n.d. 100
V. fischeri inhibition (%), 5 min n.d. 11 ± 1.0 0.0 ± 0.5 0.0 ± 0.3 0.0 ± 0.5 n.d.
V. fischeri inhibition (%), n.d. 14 ± 0.5 0.0 ± 0.6 1.6 ± 0.4 0.0 ± 0.3 n.d.
15 min
V. fischeri inhibition (%), n.d. 20 ± 1 0.0 ± 0.4 3.3 ± 0.5 0.0 ± 0.6 n.d.
30 min
Operating costd (R$ m3/V - e 74.1/16.5 3.4/0.76 70.8/15.8 74.2/16.6
m3)
a
- after neutralization, n.d. - Not determined.
b
CONAMA No. 357/2005 (Art. 16) (Brazil, 2005).
c
Commission for Environmental Policy of Minas Gerais, (COPAM, 2008).
d
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 BAUMINAS Química and Pero  xidos do Brasil (Brazil): H2O2 (50.0% (w/v), density at 25  C ¼ 1.2 g mL1) e 2.30 R$ per kg; FeCl3 (38% (w/w)) e 1.78 R$ per kg. For
converting from Brazilian Real to Euro, it was used the exchange rate of 4.48 R$/V - European Central Bank and Portugal Bank, August 20, 2018 (www.bportugal.pt).

coagulation/flocculation (C/F) and iii) combination of both pro- oxidation experiments, only replacing the biodigested CPW by
cesses in series. distilled water. The residual concentration of oxidant (H2O2) and
the presence of hydroxyl radicals were evaluated along the reaction
time.
2.2. Fenton’s oxidation procedure

The Fenton’s reactions were performed in batch mode using a 2.3. Chemical coagulation/flocculation procedure
250 mL-capacity glass batch reactor. The reactor was equipped with
a water recirculation jacket connected to a water thermostatic bath Coagulation/flocculation was carried out in a jar test apparatus
(Huber, polystat cc1) to maintain the temperature constant. A (VELP Scientifica, model JLT6). In each beaker, 200 mL of bio-
volume of 150 mL of biodigested CPW was put into the reactor and, digested CPW was placed and the pH adjusted to the desired value
after reaching the required temperature, the pH was adjusted to the with 10 M NaOH or 1 M H2SO4; then, FeCl3 (Labchem®) solution
desired value with 1 M H2SO4 or 1 M NaOH. Subsequently, the (12 g L1) was added and the pH immediately readjusted (if
required amount of iron (as ferric chloride - FeCl3.6H2O, from necessary). All runs were performed at room temperature (~28  C).
Labchem®) was added, and when the pH became steady, H2O2 (30% The coagulation step comprised high stirring speed (150 rpm)
w/v, from Carlo Erba®) was put into the reactor, this corresponding during 3 min and, for the flocculation stage, the stirring was slowed
to the initial instant of the oxidation runs (t ¼ 0 in the figures down to 15 rpm for a period of 20 min. These operating conditions
below), which were extended for 3 h. As the effluent from the were established according to data reported in the literature
Fenton reaction was intended to be subjected to anaerobic diges- (Eckenfelder , 2000; Poland and Pagano, 2010; Rodrigues et al.,
tion, the reaction was catalyzed with FeCl3 instead of FeSO4 (nor- 2014a). In the flocculation step no flocculant was added because
mally used). On the one hand, the sulfates are reduced to hydrogen in previous studies it was verified that it did not increase the
sulfide by action of anaerobic microorganisms, which present s organic load reduction (Karthik et al., 2008; Nandy et al., 2003;
toxic effects and generate odors and, on the other hand, it generates Rodrigues et al., 2013).
sulfuric acid by the action of oxygen, which can induce corrosion. In order to separate the flocs from the clarified effluent, sedi-
The homogenization into the reactor was achieved by stirring at mentation was carried out for 20 h and then a portion of super-
200 rpm, using a bar and a magnetic plate (Falc®). natant was taken to determine turbidity, BOD5, COD, TOC, toxicity
Along time, the temperature and pH values were recorded, and and biodegradability (expressed as BOD5:COD ratio).
samples were collected to measure the total organic carbon (TOC),
after adding an excess of sodium sulfite to eliminate the residual 2.4. Analytical methods
hydrogen peroxide and stop the reaction in the sampling vial. In the
last samples, collected after 180 min, BOD5, COD and toxicity The chemical oxygen demand was determined according to
(evaluated by the inhibition of Vibrio fischeri) were measured in the Method 5220D (APHA, 2005). Firstly, the samples were digested at
effluent, after increasing the pH until ~12 (this way precipitating 150  C for 2 h in closed reflux (termoreactor Merck, model 420) and
the iron and eliminating the residual oxidant present) with 10 M then the amount of potassium dichromate not consumed was
NaOH and further neutralization to pH ~7.0 with concentrated determined, measuring the absorbance at 605 nm (photometer
H2SO4. The samples for toxicity assessment were neutralized using Merck, model Nova 60 Spectroquant). The dissolved COD (CODdiss.)
1 N HCl, as proposed by the analytical methodology was determined in the same way as described for COD, after
(Standardization, 2005). filtration on a glass microfiber filter GF/D from Whatman.
The blank runs were performed the same way as the Fenton’s BOD5 corresponds to oxygen consumption in 5 days by
4 V. Gomes de Barros et al. / Environmental Pollution 259 (2020) 113796

microorganisms in the organic matter biodegradation process The 3D-EEM spectrum provides information about several dis-
(Method 5210 D (APHA, 2005)). The samples were placed in a BOD solved organic compounds that might be present, such as aromatic
analysis apparatus (Velp Scientifica, model BOD EVO) and incu- proteins (EM < 380 nm and EX < 250 nm), fulvic-like compounds
bated (Velp Scientifica, model FOC 215 E incubator) at 20  C for (EM and EX of >380 nm and <250 nm, respectively), microbial
assessing the dissolved oxygen consumption. metabolic by-products (EM < 380 nm and EX > 250 nm) and humic
The total organic carbon was measured in a TOC analyzer (Shi- acid-like compounds (EM > 380 nm and EX > 250 nm) (Chen et al.,
madzu, model TOC- L) equipped with an autosampler (Shidmazu, 2003; Kang et al., 2015). The results showed that the presence of
model ASI-L). In the analytic procedure the samples are subjected fluorescence in CPW corresponds to humic acid-like compounds,
to catalytic combustion at 680  C and the concentration of the once only one fluorescence center was observed in the region of
formed CO2 is measured in an infrared detector (Method 5310 D such substance (see Fig. 1a).
(APHA, 2005)). The biodigested CPW is not suitable for direct discharge into
The turbidity was determined using a HI88703 turbidimeter water bodies because, except for pH, all parameters are above the
from Hanna Instruments, following the Method 2130 B (APHA, legislated values (CONAMA Resolution No. 357/2005 (Brazil, 2005)
2005). or Normative COPAM No.01/2008 (COPAM, 2008) - cf. Table 1. On
The inhibition of Vibrio fischeri (V. fischeri) was the selected the other hand, the recirculation to anaerobic reactors (which is a
method to evaluate the acute toxicity of the wastewater. The common approach to increase the alkalinity of the raw effluent fed
respective determination was carried out according to ISO 11348e3 to the reactors) or the treatment by another biological process
(Standardization, 2005), comparing the luminescence intensity should be not feasible, considering the low biodegradability (and
emitted by the bacteria before and after the V. fischeri exposure to some toxicity) of the biodigested CPW. So, this effluent needs a
the samples for 5, 15 and 30 min at 15  C. For this purpose, a subsequent treatment for degradation of the recalcitrant com-
Microtox analyzer (ModerWater, model 500) was used. pounds, biodegradability improvement and toxicity reduction, thus
The concentration of hydrogen peroxide was measured ac- allowing its recirculation to the anaerobic digester. In this study,
cording to the method developed by Sellers (1980), which is based three strategies were considered for that purpose: Fenton process
on the measurement at 400 nm of the absorbance of the yellow (Approach 1), chemical coagulation/flocculation (C/F) (Approach 2)
color developed by the reaction of hydrogen peroxide with tita- and combination of C/F and Fenton reaction (Approach 3).
nium oxalate (Thermo Unicam spectrophotometer, model Helios g).
The presence of hydroxyl radicals was evaluated by measuring 3.2. Optimization of Fenton process (approach 1)
the absorbance at 563 nm of the brown color of 1,5-
diphenylcarbazone formed by oxidation of 1,5-diphenylcarbazide The Fenton process is influenced by several operating condi-
by the hydroxyl radicals (Wang et al., 2011). tions, namely: catalyst dose (Fe3þ), initial pH, initial oxidant (H2O2)
Three-dimensional fluorescence excitation-emission matrix (3D concentration and temperature. A parametric study was carried out
EEM) spectra were obtained using a JASCO spectrofluorometer (FP- to evaluate the effect of each variable on organic compounds
8300). Samples were diluted with Milli-Q water in order to avoid removal, improvement of the biodegradability (assessed in terms of
inner filter effect and analyzed in a 1 cm quartz cuvette. Excitation the BOD5:COD ratio) and reduction of the toxicity. The obtained
wavelength range varied from 200 to 500 nm with 2 nm steps results are presented and discussed in the following sections.
increment while the emission spectrum was recorded between 250
and 750 nm with 1 nm steps. The scan speed was fixed at 3.2.1. Effect of the catalyst dose
1000 nm min1 and the slit width was fixed at 5 nm and 10 nm in The effect of iron dose was evaluated in the range
excitation and emission, respectively. Milli-Q water blanks were 0.90e2.70 g L1 (corresponding to Fe3þ:H2O2 ratios (w/w) of
analyzed in the same conditions as the samples and the respective 0.1e0.3, close to the typical ratios reported for the Fenton process -
spectrum was subtracted to samples spectra in order to minimize between 0.05 and 0.2 (USP Technologies, 2019)), and the remaining
Raman scattering. The 1st and 2nd order Rayleigh scattering lines variables, H2O2 concentration, initial pH and temperature, respec-
were removed using internal filter. tively, were fixed at 9.0 g L1, 3.0 and 30  C. The TOC reduction is
All analytical determinations were performed in duplicate and fast in the first 60e90 min of reaction, then the removal rate decays
the coefficients of variation were less than 7.0% for pH, 9.0% for up to 150 min and TOC remains approximately constant for higher
BOD5, 6.0% for COD, 2.0% for TOC and hydrogen peroxide concen- reaction times e see Fig. 2a. A similar trend was observed for pH
tration, 3.0% for absorbance at 530 nm (indirect measurement of evolution (Fig. 2b), which decreases from the beginning of the re-
the hydroxyl radicals), 8.0% for turbidity and 10.0% for V. fischeri action (t ¼ 0) up to 60 min, being the variation attenuated after-
inhibition. ward. The decrease of pH for t < 0 is caused by the addition of the
Fenton’s reagents (particularly the iron salt) and during the reac-
3. Results and discussion tion (t > 0) is commonly attributed to the carboxylic acids (inter-
mediate compounds) formed, as reported in other studies (Lima
3.1. Characteristics of biodigested CPW et al., 2018; Rodrigues et al., 2018; Rodrigues et al., 2017b; Soares
et al., 2019).
The physicochemical characteristics of the biodigested CPW The maximum TOC removal (43.1% after 180 min) was obtained
used in this work are reported in Table 1. In general, the effluent is for the catalyst dose of 1.8 g L1. The greatest overall COD (49.4%)
characterized by the presence of a high amount of organic com- and BOD5 (30.4%) removals were also observed for this catalyst
pounds, since it presents high concentrations of BOD5, TOC and dose, along with the maximum BOD5:COD ratio (0.45), as shown in
COD (the concentration of dissolved organic matter, expressed as Table 2. Higher Fe3þ concentrations caused a decrease in the pro-
dissolved COD, represents 95% of the total COD) and low biode- cess efficiency, probably due to the fact that the iron in excess reacts
gradability (BOD5:COD <0.4), which suggests difficult stabilization with HO (Eq. (6)) (Walling, 1975), thereby decreasing the hydroxyl
by biological degradation. As expected from the anaerobic diges- radical concentration in solution. This fact is corroborated by the
tion of CPW, the pH is neutral and high turbidity levels were found. decay of HO formation when [Fe3þ] increases from 1.8 to 2.7 g L1
The biodigested CPW has low toxicity as indicated by V. fischeri in the blank tests (Fig. 3a) whereas a rapid hydrogen peroxide
inhibitions of 11.0e20.0%. consumption is observed (Fig. 3b).
V. Gomes de Barros et al. / Environmental Pollution 259 (2020) 113796 5

Fig. 1. 3D EEM fluorescence spectra of biodigested CPW (a), after Fenton’s oxidation alone (b), after coagulation/flocculation (c) and after C/F followed by Fenton’s process (d).

Regarding the acute toxicity, only the Fe3þ dose of 0.9 g L1 the one hand, at very high pH occurs the decomposition of
originated some toxicity, but the inhibition of V. fischeri (2.4e6.9% - hydrogen peroxide into water and oxygen (Szpyrkowicz et al.,
Table 2) was less than that obtained for the initial biodigested CPW 2001) and the formation of iron hydroxide (Pignatello, 1992),
(11.0e20.0% - Table 1). which is therefore not solubilized and available to catalyze the
reaction. On the other hand, at very acidic pH (<2.0) H2O2 reacts
with the Hþ in solution and forms H3Oþ 2 ; moreover, the Fe

con-
3.2.2. Effect of the initial pH centration, responsible for ensuring the continuity of the reaction,
The effect of pH was assessed in the range of 2e7, keeping decreases because in such conditions the Fe(OH)2þ and Fe(OH)þ 2
constant the temperature at 30  C, the H2O2 concentration at species prevail (Pignatello, 1992). Thus, both the amount of oxidant
9.0 g L1 and the Fe3þ at the previously obtained optimal dose and the catalyst available to generate the hydroxyl radicals
(1.8 g L1). This range of pH values was selected to encompass the decrease, which leads to a reduced efficiency of the process.
commonly reported values (2e5) as optimal for the Fenton process
(Pignatello, 1992; Rivas et al., 2001; Zhang et al., 2019). The highest
mineralization was reached at initial pH of 5.0, corresponding to 3.2.3. Influence of initial hydrogen peroxide concentration
46.5% TOC reduction (see Fig. 2c). In addition, this pH value led to Four runs were carried out to determine the effect of the initial
the highest removals of BOD5 (33.6%) and COD (51.7%) and biode- H2O2 dose by varying its concentration in the range of 4.5 g L1 to
gradability improvement (BOD5:COD ratio increased from 0.34 to 18.0 g L1 (corresponding to 0.5e2.0 times the stoichiometric dose
0.45) whereas generating a non-toxic effluent as shown in Table 2. based on the COD value); usually, higher amounts are required for
The optimal pH for treating the biodigested CPW was not very the complete removal of COD since part of the H2O2 decomposes
acidic as typically reported for other treatments by the Fenton into oxygen and water (Southworth and Voelker, 2003) and it is
process (pH 2.5e3.0). This fact was associated to the decay of pH also consumed in scavenging reactions. The concentration of Fe3þ
from 5.0 to near 2.5 as shown in Fig. 2d, due to the addition of FeCl3 (1.80 g L1), initial pH (5.0) and temperature (30  C) were kept
with the simultaneous formation of the above-mentioned carbox- constant. In Fig. 2e one can see that, for all H2O2 concentrations
ylic acids. This aspect is advantageous for the industrial application tested, the TOC removal increases rapidly in the first 30 min of
of the Fenton process in the treatment of this type of effluent since reaction, which coincides with the higher pH decay during the
it allows reducing the costs associated with the consumption of reaction (see Fig. 2f), and still increases but at a lower rate up to
acid for the preliminary acidification. 150 min and afterward is approximately constant. The maximum
The existence of an optimum pH is justified by the fact that, on TOC removal was achieved for [H2O2] ¼ 9.0 g L1, corresponding to
6 V. Gomes de Barros et al. / Environmental Pollution 259 (2020) 113796

Fig. 2. Effect of Fe3þ dose (a,b), initial pH (c,d), H2O2 concentration (e,f) and temperature (g,h) in TOC removal and evolution of pH, with standard errors for a confidence level of
95%, during the Fenton process alone (experimental conditions were as follows: a,b) pHinitial ¼ 3.0, T ¼ 30  C and [H2O2] ¼ 9.0 g L1; c,d) T ¼ 30  C, [Fe3þ] ¼ 1.8 g L1 and
[H2O2] ¼ 9.0 g L1; e,f) pHinitial ¼ 5.0, T ¼ 30  C, [Fe3þ] ¼ 1.8 g L1; g,h) pHinitial ¼ 5.0, [Fe3þ] ¼ 1.8 g L1 and [H2O2] ¼ 9.0 g L1).
V. Gomes de Barros et al. / Environmental Pollution 259 (2020) 113796 7

Table 2
Operating conditions used in the Fenton’s oxidation runs (Approach 1), removal efficiencies, BOD5:COD ratio and toxicity reached after 180 min of reaction with standard errors
for a confidence level of 95%. Bold conditions refer to the run providing better efficiencies when each variable was changed in the parametric study.

Parameter [H2O2] (g L1) [Fe3þ] (g L1) Initial pH T (oC) Removal efficiency (%) BOD5:COD Inhibition V. fischeri at 5/15/30 min (%)

TOC COD BOD5

Fe3þ 9 0.9 3.0 30 26.8 ± 0.5 26.8 ± 0.8 18.1 ± 0.8 0.36 ± 0.08 2.4 ± 0.2/3.8 ± 0.4/6.9 ± 0.6
1.8 43.1 ± 0.8 49.4 ± 0.2 30.4 ± 0.3 0.45 ± 0.02 0.0 ± 0.1/0.0 ± 0.3/0.0 ± 0.4
2.25 35.4 ± 0.6 36.9 ± 0.2 25.5 ± 0.2 0.38 ± 0.03 0.0 ± 0.3/0.0 ± 0.4/0.0 ± 0.3
2.7 32.3 ± 1.0 33 ± 0.6 22.9 ± 1.0 0.37 ± 0.01 0.0 ± 0.1/0.0 ± 0.2/0.0 ± 0.4
pH 9 1.8 2 30 18.1 ± 0.5 21.2 ± 0.4 14.6 ± 3.0 0.27 ± 0.05 3.2 ± 0.2/6.2 ± 0.3/10.2 ± 0.5
3 43.1 ± 0.8 49.4 ± 0.2 30.4 ± 0.3 0.45 ± 0.02 0.0 ± 0.1/0.0 ± 0.3/0.0 ± 0.4
5 46.5 ± 0.9 51.7 ± 0.4 33.6 ± 1.2 0.45 ± 0.06 0.0 ± 0.1/0.0 ± 0.2/0.0 ± 0.1
6 41.5 ± 1.2 39.2 ± 0.8 24.5 ± 1.7 0.01 ± 0.02 0.0 ± 0.2/0.0 ± 0.5/0.0 ± 0.6
7 24.8 ± 0.4 28.7 ± 0.05 17.3 ± 0.7 0.01 ± 0.01 2.6 ± 0.3/3.3 ± 0.2/6.8 ± 0.5
H2O2 4.5 1.8 5 30 26.3 ± 0.8 26.5 ± 0.4 20.8 ± 1.5 0.35 ± 0.03 2.9 ± 0.5/4.0 ± 0.4/6.7 ± 0.1
9 46.5 ± 0.9 51.7 ± 0.4 33.6 ± 1.2 0.45 ± 0.06 0.0 ± 0.1/0.0 ± 0.2/0.0 ± 0.1
13.5 38.3 ± 1.3 42.6 ± 2.6 27.6 ± 1.6 0.41 ± 0.03 0.0 ± 0.2/0.0 ± 0.3/0.0 ± 0.1
18 30.9 ± 1.8 35.9 ± 0.1 23.4 ± 1.6 0.39 ± 0.05 0.0 ± 0.4/0.0 ± 0.8/0.0 ± 0.2
T 9 1.8 5 20 34.3 ± 0.5 39 ± 1.2 29.7 ± 2.0 0.37 ± 0.03 0.0 ± 0.4/0.0 ± 0.3/0.0 ± 0.7
30 46.5 ± 0.9 51.7 ± 0.4 33.6 ± 1.2 0.45 ± 0.06 0.0 ± 0.1/0.0 ± 0.2/0.0 ± 0.1
55 51.3 ± 1.2 55.7 ± 0.5 39.7 ± 1.8 0.44 ± 0.01 0.0 ± 0.5/0.0 ± 0.6/0.0 ± 0.4

Fig. 3. Effect of Fe3þ dose on the formation of hydroxyl radicals (by measuring the absorbance at 563 nm in the presence of 1,5-diphenyl carbazide) (a) and in the consumption of
hydrogen peroxide (b), with standard errors for a confidence level of 95%, in blank tests.

46.5% after 180 min of reaction. The same dose also led to the best of the effluent discharged from the anaerobic reactors (55  C), and
COD (51.7%) and BOD5 (33.6%) removal efficiencies, as well as to a therefore there are no additional costs associated with heating the
higher BOD5:COD ratio (0.45), as can be seen in Table 2. The exis- effluent. Furthermore, it is known that for temperatures above
tence of an optimal H2O2 concentration has been reported by other 60e70  C the thermal decomposition of hydrogen peroxide in
researchers that treated industrial wastewater by Fenton’s oxida- water and oxygen can occur.
tion (GilPavas et al., 2017; Guerreiro et al., 2016) and can be As shown in Fig. 2g and Table 2, overall TOC removals raised
explained by the scavenging of hydroxyl radicals by excess of H2O2, significantly when increasing the temperature from 20 to 30  C,
generating HOO species with a lower oxidation potential (Eq. (5)) however, a smaller increase was achieved between 30 and 55  C.
(Walling, 1975). Again, a pH decay was observed during the oxidative process,
As regards acute toxicity (evaluated by inhibition of Vibrio which was most noticeable for the assay at 55  C as shown in
fischeri), only the sample corresponding to the lowest H2O2 con- Fig. 2h. The positive effect of temperature on the process efficiency
centration (4.5 g L1) inhibits the bacterial growth, but the inhi- is associated with the increase of the kinetic constants of the re-
bition values (2.9e6.7% e Table 2) were smaller than those actions, as expressed by the Arrhenius law. For the temperature
obtained for the initial biodigested CPW (11e20% - Table 1). For leading to the best efficiency (55  C), removals of 51.3, 55.7 and
higher oxidant doses, no toxicity was detected in the final effluent. 39.7% for TOC, BOD5 and COD, respectively, were obtained (see
Table 2). Moreover, the biodegradability was improved and the
3.2.4. Effect of temperature toxicity of wastewater was totally eliminated (Table 2).
The influence of the temperature on the process efficiency was The effect of temperature on the process efficiency found in the
evaluated by carrying out three runs at 20  C, 30  C and 55  C, present work has been early reported by other authors when
respectively. These values are practically inside the optimal range treating biodigested sugarcane vinasse (Guerreiro et al., 2016), real
(20e50  C) reported in the literature for the best performance of paper wastewater (Torrades et al., 2003) and textile dyeing efflu-
Fenton’s oxidation (Parmar, 2014). Still, the minimum temperature ents (Rodrigues et al., 2014b; Rodrigues et al., 2013), who also
has been established considering the weather conditions in S~ ao observed a positive effect of temperature on the Fenton’s oxidation
Paulo State (Brazil) during the coffee harvest period (AprileJuly) efficiency.
and when CPW is produced, at an ambient temperature never The biodigested CPW treated by Fenton’s oxidation, using the
below 20  C; the maximum value corresponds to the temperature optimal conditions previously found (9 g L1 of H2O2, initial pH of 5,
8 V. Gomes de Barros et al. / Environmental Pollution 259 (2020) 113796

1.8 g L1 of Fe3þ and temperature of 55  C), shows only humic acid- [Fe3þ]  250 mg L1, a residual inhibition around 1.6% and 3.4% was
like compounds as fluorescent dissolved organic matter as observed for contact times with the bacteria of 10 and 15 min,
observed for biodigested CPW; however, the fluorescence center respectively.
decreased when compared to that obtained for the wastewater As regards biodegradability, Fig. 4b shows that the BOD5:COD
before Fenton’s oxidation (see Fig. 1). ratio remained approximately constant (0.32e0.34) whatever the
The treated biodigested CPW shows a high organic content iron dose employed and very similar to that obtained for the raw
(COD ¼ 1889 mg O2 L1; TOC ¼ 706 mg C L1 and BOD5 ¼ 816 mg O2 biodigested CPW.
L1 e see Table 1). The concentrations of COD and BOD5 are clearly The existence of a plateau when changing the coagulant dose
above the discharge limits established in Brazilian legislation can be explained by the fact that the colloidal particles get desta-
(COD ¼ 180 mg O2 L1 and BOD5 ¼ 10 mg O2 L1), so the effluent bilized when the coagulant added to the biodigested CPW interacts
cannot be discharged into water bodies. Considering that a signif- with negative colloids and neutralizes their charges. However,
icant increase in biodegradability (BOD5:COD raised from 0.34 to when coagulant is in excess, it can be adsorbed by the colloids, so
0.44) and decrease in toxicity were observed (see Table 1), it is these become positively charged and electrostatic repulsion occurs,
anticipated that the proposed treatment by the Fenton process can which leads to re-stabilization of the colloids, thus impairing the
allow to recirculate the effluent (which still contains a considerable coagulation (Verma et al., 2010).
content of organic matter) back to the anaerobic reactor. The coagulation/flocculation process improves the effluent
The operating cost of this treatment process is considerable quality at a much lower operating cost (3.4 R$ m3 z 0.76/V m3 -
(74.1 R$ m3/16.5 V m3 cf. Table 1). Therefore, another strategy Table 1) as compared to the Fenton process, only associated with
was sought, starting with the more conventional stage of coagu- the iron chloride consumption because the costs of acid and base
lation/flocculation, aiming at a better efficiency of the organic for pH adjustment are negligible. The 3D EEM spectrum points out
matter removal and at a reduction of costs. again to the presence of humic acid-like compounds (Fig. 1 c), but
the fluorescence center is below that observed for biodigested CPW.
3.3. Chemical coagulation/flocculation (approach 2) The effluent still presents a high organic load (COD ¼ 2613 mg O2
L1; TOC ¼ 907 mg C L1 and BOD5 ¼ 846 mg O2 L1), which means
3.3.1. Effect of pH that it does not meet the discharge limits (see Table 1). Thus, a
The influence of pH on TOC, COD and BOD5 removals, as well as subsequent treatment is required; however it cannot be a biological
on turbidity and toxicity reduction during the coagulation/floccu- degradation because the effluent generated after C/F presents some
lation process, was assessed by performing a number of experi- toxicity and low biodegradability (not reaching the BOD5:COD
ments varying this parameter in the range 2e11. The selection of threshold ratio of 0.4 that is the value reported in the literature for
this range took in account the pH values reported in the literature easily biodegradable wastewaters (Pulgarin et al., 1999; Zaghdoudi
for the best performance of coagulation using ferric salts (Eck- et al., 2017)). Therefore, its recirculation to the anaerobic digestor is
enfelder Jr., 2000). The highest removal efficiencies were obtained not an option either.
at high pH values (10e11), i.e., removals of 19e22% (Fig. 4a) were An alternative approach for biodigested CPW treatment, as re-
achieved. Coagulation/flocculation allows suspended, colloidal and ported by Guerreiro et al. (2016), is coagulation/flocculation as a
dissolved material to be removed through the formation of flocs pretreatment step upstream from the Fenton process (which con-
(Ghernaout and Ghernaout, 2012), but a significant proportion of stitutes the Approach 3 in this study). This strategy aims to reduce
organic matter can still remain in the effluent after this process. the costs associated with chemicals consumption in the second
Coagulation/flocculation was not satisfactory to enhance the process. On the one hand, a less amount of hydrogen peroxide is
effluent biodegradability (assessed in terms of BOD5:COD ratio), consumed because the organic content is reduced in the coagula-
which remained invariable at ca. 0.33 (Fig. 4a). However, the pro- tion/flocculation step and, on the other hand, the iron salt still
cess led to a significant reduction of the effluent turbidity, up to 92% remaining in solution after coagulation/flocculation will be used as
(Fig. 4c). In addition, a reduction of the inhibition of Vibrio fischeri catalyst in the chemical oxidation step.
also occurred (Fig. 4c). Taking into account the results obtained, a
pH of 10 was selected as the best one for operating the coagulation/ 3.4. Chemical coagulation/flocculation plus Fenton’s oxidation
flocculation process. At this pH value, the anionic forms of iron (approach 3)
(Fe(OH)-4) and Fe(OH)3 become dominant, consequently the coag-
ulation/flocculation process occurs probably by the mechanism of In order to increase the mineralization and/or biodegradability
inclusion in a precipitate (Duan and Gregory, 2003; Ghernaout and and to reduce the toxicity of CPW at the lowest treatment cost, the
Ghernaout, 2012), i.e. the colloidal organic matter is trapped during strategy of combining coagulation/flocculation (C/F) with Fenton’s
the formation of Fe(OH)3. oxidation was adopted. Thus, the effluent is first treated by C/F
under the best operating conditions reached in Approach 2
3.3.2. Effect of Fe3þ dosage (pH ¼ 10.0 and [Fe3þ] ¼ 250 mg L1) and, afterward submitted to
The concentration of Fe3þ was assessed within the range of Fenton’s oxidation. In the second treatment step the optimal con-
62.5e1000 mg L1 at pH 10. It was found that the organic matter ditions of pH and temperature previously found when testing
removal increased with the coagulant dose until 250 mg L1; Approach 1 (pH ¼ 5.0, T ¼ 55  C) were selected, only evaluating the
higher doses did not provide any efficiency improvement (Fig. 4b). effect of the initial concentrations of catalyst and oxidant, as pre-
For the optimal dose of coagulant (250 mg Fe3þL1) a significative sented below. In Fenton’s oxidation the optimal iron and hydrogen
reduction in the organic matter content was achieved peroxide concentrations found in Approach 1 were not used
(COD ¼ 38.7%, BOD5 ¼ 39.0%, TOC ¼ 37.3%) (see Table 1 and Fig. 4b). because the organic load is lower since part of it was removed in
The turbidity of the effluent significantly decreased up to the first treatment step (i.e. in the coagulation/flocculation process)
[Fe3þ] ¼ 125 mg L1, but a lower decrease occurred when the dose
was increased to 250 mg L1 and finally remained at an approxi- 3.4.1. Influence of Fe3þ concentration
mately constant value (27 NTU) for the higher doses (Fig. 4d). The dose of Fe3þ was varied in the range 0.55e2.2 g L1, main-
Although the acute toxicity for a contact time of 5 min has been taining [H2O2] ¼ 5.6 g L1 (corresponding to the stoichiometric
initially reduced to 1.6%, it disappeared for [Fe3þ] > 125 mg L1; for amount of COD of the effluent after C/F - 2613 mg O2 L1, as shown
V. Gomes de Barros et al. / Environmental Pollution 259 (2020) 113796 9

Fig. 4. Effect of pH and Fe3þ concentration on the removals of TOC, COD and BOD5 and on biodegradability expressed as BOD5:COD ratio (a,b) and turbidity and acute toxicity of the
effluent (c,d), with standard errors for a confidence level of 95%, after the coagulation/flocculation process (vcoagulation ¼ 150 rpm, tcoagulation ¼ 3 min, [Fe3þ]a,c ¼ 62.5 mg L1,
pHb,d ¼ 10, T ¼ ~28  C, vflocculation ¼ 20 rpm, tflocculation ¼ 15 min).

in Table 1). As can be seen in Fig. 5a and Table 3, the removals of once they allow higher organic matter removals as compared with
TOC, BOD5 and COD increased with the catalyst dose up to 1.8 g L1 the C/F process alone (see Table 1). Despite the reduction of the
and decreased for higher concentrations. This iron concentration organic content, COD and BOD5 remain above the discharge limits
also leads to one of the largest pH decreases during the Fenton imposed by the Brazilian legislation (see Table 1). Thus, the treated
reaction (Fig. 5b). The optimal Fe3þ dose reached was the same biodigested CPW cannot be discharged into water bodies, but
observed when treating biodigested CPW only by Fenton’s oxida- considering the elimination of the toxicity and the strong biode-
tion. The existence of an optimal catalyst dose results from the fact gradability improvement, the wastewater can be recirculated to the
that the excess of ferric iron in solution reacts with the hydroxyl anaerobic reactor; this will be the aim of future work, in order to
radical (Eq. (6)), as mentioned above. assess the impact on bio-methane production.
The biodegradability increased from 0.34 for the raw bio-
digested CPW to 0.43 after C/F plus Fenton’s oxidation and a non- 3.5. Comparison with literature data
toxic effluent was reached (see Table 3).
Some reports on the treatment of raw coffee wastewaters are
3.4.2. Effect of initial H2O2 concentration available in the open literature. Anaerobic digestion has shown be
The influence of the initial H2O2 concentration was evaluated in efficient in the removal of organic load (COD removals between 60
the range 5.6e11.2 g L1, setting the catalysts dose at the optimal and 86%) with high methane production (Battista et al., 2016;
value previously determined (1.8 g L1). As shown in Fig. 5c and Beyene et al., 2014; Cruz-Salomo  n et al., 2018; Fia et al., 2012; Villa-
Table 3 the mineralization increased with the oxidant concentra- Montoya et al., 2017). In a system composed by two anaerobic UASB
tion up to 8.4 g L1, corresponding to TOC, COD and BOD5 removals reactors in series treating CPW and operating at high organic
of 61.8, 62.2 and 44.9% after 180 min of reaction, respectively. loading rate (18.2 g COD (L d)1), the organic matter removal
Moreover, the pH decreased during the reaction, more pro- (expressed as COD) attained 84%, which is within the range above
nouncedly for such oxidant dose (see Fig. 5d), the biodegradability mentioned, along with high methane production (2.2 L CH4 d1
of the effluent was further improved, and its toxicity was null for all (Su
arez et al., 2018)).
H2O2 concentrations tested. Besides the application of anaerobic biological processes, there
The combination of the two processes (coagulation/flocculation are some studies reported in the literature where physical/chemical
followed by Fenton’s reaction) reached global removals of 76.2, 76.5 processes, such as coagulation/flocculation, photo-Fenton oxida-
and 66.3% for TOC, COD and BOD5, respectively - Table 1 - and, tion, electrochemical advanced oxidation and adsorption, were
again, only the presence of humic acid-like fluorescent compounds adopted to treat raw CPW. COD removal efficiencies of 67% and 63%
was observed but with a decrease in the fluorescence center were obtained for coagulation/flocculation and photo-Fenton
compared to the effluent before treatment (see Fig. 1). The Fenton’s oxidation of CPW, respectively (Zayas et al., 2007). The treat-
oxidation alone (Approach 1) or combined with C/F (Approach 3) ability of coffee processing wastewater was also evaluated by
implies similar operating costs (~74.0 R$ m3/~17.0 V m3). Both anodic oxidation with electrogenerated H2O2 and TOC and color
treatment strategies are the best ones to treat the biodigested CPW, removals of 73e84% and 89e93%, respectively, were achieved
10 V. Gomes de Barros et al. / Environmental Pollution 259 (2020) 113796

Fig. 5. TOC removal and evolution of pH along reaction time for different Fe3þ doses (a,b) and H2O2 concentrations (c,d), with standard errors for a confidence level of 95%, in CPW
previously treated by coagulation/flocculation (experimental conditions were as follows: a,b) pHinitial ¼ 5.0, T ¼ 55  C and [H2O2] ¼ 5.6 g L1; c,d) pHinitial ¼ 5.0, T ¼ 55  C and
[Fe3þ] ¼ 1.8 g L1).

Table 3
Operating conditions used in Fenton’s oxidation of biodigested CPW after treatment by C/F, respective removal efficiencies, along with BOD5:COD ratio and toxicity achieved
after 180 min of reaction with standard errors for a confidence level of 95%. Bold conditions refer to the run providing better efficiencies when each variable was changed in the
parametric study.

Parameter [H2O2] (g L1) [Fe3þ ] (g L1) Initial pH T (oC) Removal efficiency (%) BOD5:COD Inhibition V. fischeri at 5/15/30 min (%)

TOC COD BOD5

Fe3þ 5.6 0.55 5.0 55 43.4 ± 0.6 40.3 ± 0.2 21.4 ± 2.0 0.42 ± 0.07 0.0 ± 0.9 / 0.0 ± 0.4 / 0.0 ± 0.6
1.12 49.3 ± 0.9 47.7 ± 0.5 30.3 ± 1.4 0.42 ± 0.04 0.0 ± 0.2 / 0.0 ± 0.4 / 0.0 ± 0.3
1.8 52.5 ± 0.5 54.9 ± 0.9 38.5 ± 0.5 0.43 ± 0.02 0.0 ± 0.6 / 0.0 ± 0.5 / 0.0 ± 0.3
2.2 50.6 ± 1.0 48.6 ± 0.2 34.4 ± 2.4 0.41 ± 0.03 0.0 ± 0.2 / 0.0 ± 0.4 / 0.0 ± 0.3
H2O2 5.6 1.8 5 55 52.5 ± 0.5 54.9 ± 0.9 38.5 ± 0.5 0.43 ± 0.02 0.0 ± 0.6 / 0.0 ± 0.5 / 0.0 ± 0.3
8.4 61.8 ± 0.8 62.2 ± 0.3 44.9 ± 0.6 0.47 ± 0.01 0.0 ± 0.5 / 0.0 ± 0.3 / 0.0 ± 0.6
9.6 57.9 ± 0.7 57.5 ± 0.2 40.7 ± 0.7 0.44 ± 0.04 0.0 ± 0.5 / 0.0 ± 0.7 / 0.0 ± 0.2
11.2 55.4 ± 1.2 55.9 ± 0.7 38.8 ± 1.0 0.44 ± 0.08 0.0 ± 0.8 / 0.0 ± 0.4 / 0.0 ± 0.6

(Villanueva-Rodríguez et al., 2014). High removals of COD and BOD PILC as catalyst to the anaerobically digested effluent and obtained
(>98%) were also reached by using commercial activated carbon, removals of 50 and 66% for TOC and COD, respectively. The overall
avocado peel carbon and avocado seed carbon as adsorbents (Devi, COD removal efficiency, corresponding to anaerobic digestion plus
2009; Devi et al., 2008). wet peroxidation, was 97%.
Few investigations were performed on the treatment of anaer- In the present study, Fenton oxidation per se or combined with
obic digested CPW, which is the goal of the present study. coagulation/flocculation proved to be promising strategies for
Selvamurungan et al. (2010) studied the integration of anaerobic biodigested CPW treatment, allowing COD removals above 93e96%
biodigestion in a UASB with aerobic biodegradation to treat bio- when considering the previous anaerobic digestion. These effi-
digested CPW. The authors achieved overall removals of 68.6 and ciencies are higher than those obtained by Selvamurungan et al.
74.6% for COD and BOD, respectively. A similar study was carried (2010) and Rossmann et al. (2013) using combined anaerobic and
out Rossmann et al. (2013) and similar removals (87 and 84% for aerobic degradation. However, they are similar to that reached by
BOD and COD) were achieved with and without the aeration step. A Sanabria et al. (2012) when employing anaerobic digestion fol-
different work was performed by Sanabria et al. (2012) that applied lowed by wet peroxidation.
a post-treatment consisting of wet peroxidation using AleCeeFe-
V. Gomes de Barros et al. / Environmental Pollution 259 (2020) 113796 11

4. Conclusions 2017. Improved methane production from sugarcane vinasse with filter cake in
thermophilic UASB reactors, with predominance of Methanothermobacter and
Methanosarcina archaea and Thermotogae bacteria. Bioresour. Technol. 244,
In this study it was evaluated the treatment of biodigested CPW 371e381.
by Fenton and coagulation/flocculation processes, alone and com- Battista, F., Fino, D., Mancini, G., 2016. Optimization of biogas production from
bined, to obtain an effluent with characteristics allowing its recir- coffee production waste. Bioresour. Technol. 200, 884e890.
Beyene, A., Yemane, D., Addis, T., Assayie, A.A., L.T., 2014. Experimental evaluation of
culation back to anaerobic digesters. It was found that: anaerobic digestion for coffee wastewater treatment and its biomethane re-
covery potential. Int. J. Environ. Sci. Technol. 11, 1881e1886.
Brazil, 2005. Ministe rio Do Meio Ambiente. Consellho Nacional Do Meio Ambiente -
i) The Fenton’s oxidation per se (Approach 1) represents an
Conama. Resoluç~ ao nº 357 de 17 de março de 2005. http://www.mma.gov.br/
efficient strategy for improving the biodegradability (from port/conama/res/res05/res35705.pdf. accessed 07.30.18.
0.34 to 0.44) and eliminating the toxicity of the biodigested Bruno, M., Oliveira, R.A.d., 2013. Performance of uasb reactors in two stages fol-
CPW while obtaining a considerable organic matter removal lowed by post- treatment with activated sludge in wastewater batch of wet-
processed coffee. Engenharia Agrícola, Jaboticabal 33, 808e819.
(55.7%, 39.7% and 51.3% for COD, BOD5 and TOC removals, Campos, C.M.M., Prado, M.A.C., Pereira, E.L., 2013. Anaerobic digestion of waste-
respectively). water from coffee and chemical analysis of biogas produced using gas chro-
ii) The Coagulation/flocculation alone (Approach 2) showed matography: quantification of methane, and potential energy gas exchanger.
Biosci. J. 29, 570e581.
lower levels of organic matter removal (~37%e~39%), did not Chen, W., Westerhoff, P., Leenheer, J.A., Booksh, K., 2003. Fluorescence excitation-
increase the biodegradability and generated an effluent that emission matrix regional integration to quantify spectra for dissolved organic
maintained some toxicity, although showing lower inhibi- matter. Environ. Sci. Technol. 37, 5701e5710.
COPAM, 2008. Deliberaça ~o Normativa Conjunta COPAM/CERH-MG N.o 1, de 05 de
tion of V. fischeri than the biodigested CPW.
maio de 2008. In: http://www.mma.gov.br/port/conama/processos/EFABF603/
iii) The combination of the two processes (C/F plus Fenton’s DeliberaNormativaConjuntaCOPAM-CERHno01-2008.pdf. accessed 07.20.18.
oxidation) allowed to make the wastewater more biode- Cruz-Salomo n, A., Ríos-Valdovinos, E., Pola-Albores, F., Lagunas-Rivera, S., Meza-
gradable and non-toxic, like the Fenton process alone, and to Gordillo, R., Ruíz-Valdiviezo, V.M., 2018. Evaluation of hydraulic retention time
on treatment of coffee processing wastewater (CPWW) in EGSB bioreactor.
obtain the highest global removals of TOC (76.2%), COD Sustainability 10, 83.
(76.5%) and BOD5 (66.3%). Dadi, D., Mengistie, E., Terefe, G., Getahun, T., Haddis, A., Birke, W., Beyene, A.,
Luis, P., Van der Bruggen, B., 2018. Assessment of the effluent quality of wet
coffee processing wastewater and its influence on downstream water quality.
Taking in account that it is intended to recirculate the bio- Ecohydrol. Hydrobiol. 18, 201e211.
digested CPW back to the anaerobic reactor, the strategies sug- Devi, R., 2009. Innovative technology of COD and BOD reduction from coffee pro-
gested are the Fenton process (Approach 1) or the combination of C/ cessing wastewater using avocado seed carbon (ASC). Water Air Soil Pollut. 207,
299e306.
F plus Fenton’s oxidation (Approach 3), with similar operating costs Devi, R., Singh, V., Kumar, A., 2008. COD and BOD reduction from coffee processing
(~74 R$ m3/~17 V m3), once both approaches reduce the content wastewater using Avacado peel carbon. Bioresour. Technol. 99, 1853e1860.
of refractory organic compounds making the effluent more biode- Duan, J., Gregory, J., 2003. Coagulation by hydrolysing metal salts. Adv. Colloid
Interface Sci. 100e102, 475e502.
gradable and non-toxic. However, the effect of the recirculation on Eckenfelder Jr., W.W., 2000. Industrial Water Pollution Control, third ed. ed
biogas production has to be evaluated. Still, for the application of (Singapore).
such treatment industrially, the study will have to proceed for Fia, F.R.L., Matos, A.T., Borges, A.C., Fia, R., Cecon, P.R., 2012. Treatment of wastewater
from coffee bean processing in anaerobic fixed bed reactors with different
larger laboratory and/or pilot facilities (scale-up).
support materials: performance and kinetic modeling. J. Environ. Manag. 108,
14e21.
Declaration of competing interest Ghernaout, D., Ghernaout, B., 2012. Sweep flocculation as a second form of charge
neutralisationda review. Desalin. Water Treat. 44, 15e28.
Gil Pavas, E., Correa-Sa nchez, S., Acosta, D.A., 2019. Using scrap zero valent iron to
The authors declare that there is no conflict of interest regarding replace dissolved iron in the Fenton process for textile wastewater treatment:
the publication of this article. optimization and assessment of toxicity and biodegradability. Environ. Pollut.
252, 1709e1718.
GilPavas, E., Dobrosz-Go  mez, I., Go
mez-García, M.A.,  2017. Coagulation-flocculation
Acknowledgments sequential with Fenton or Photo-Fenton processes as an alternative for the
industrial textile wastewater treatment. J. Environ. Manag. 191, 189e197.
Guerreiro, L.F., Rodrigues, C.S.D., Duda, R.M., de Oliveira, R.A., Boaventura, R.A.R.,
This work was financially supported by Projects UID/EQU/ Madeira, L.M., 2016. Treatment of sugarcane vinasse by combination of coag-
00511/2019 - Laboratory for Process Engineering, Environment, ulation/flocculation and Fenton’s oxidation. J. Environ. Manag. 181, 237e248.
Kang, J., Ma, T., Zhou, Q., Gao, X., Chen, Y., Wu, J., Chen, J., Xiang, Y., 2015. New
Biotechnology and Energy - LEPABE - and by Associate Laboratory insight into DOC and DON in a drinking water biological aerated filter (BAF) by
LSRE-LCMdUID/EQU/50020/2019 funded by national funds ultimethod and correlation analysis of 3D-EEM. Anal. Methods 7, 9885e9893.
through FCT/MCTES (PIDDAC). We would also like to thank the Kantar, C., Oral, O., Urken, O., Oz, N.A., Keskin, S., 2019. Oxidative degradation of
chlorophenolic compounds with pyrite-Fenton process. Environ. Pollut. 247,
scientific collaboration under project “AIProcMat@N2020 -
349e361.
Advanced Industrial Processes and Materials for a Sustainable Karthik, M., Dafale, N., Pathe, P., Nandy, T., 2008. Biodegradability enhancement of
Northern Region of Portugal 2020” (NORTE-01-0145-FEDER- purified terephthalic acid wastewater by coagulationeflocculation process as
000006, NORTE 2020, Portugal 2020 Partnership Agreement, pretreatment. J. Hazard Mater. 154, 721e730.
Lima, V.N., Rdrigues, C.S.D., Madeira, L.M., 2018. Application of the Fenton’s process
through FEDER). Valciney Gomes de Barros is grateful to the Bra- in a bubble column reactor for hydroquinone degradation. Environ. Sci. Pollut.
zilian Coordination for the Improvement of Higher Education Control Ser. 25, 34851e34862.
Personnel (CAPES) for his Sandwich Doctorate scholarship (Process Lucas, M.S., Peres, J.A., 2009. Treatment of olive mill wastewater by a combined
process: Fenton’s reagent and chemical coagulation. J. Environ. Sci. and Health,
88881.132416/2016e01). The authors are thankful to the WWTP of Part A 44, 198e205.
Parada for providing the aerobic biological sludge for BOD analyses. Nandy, T., Shastry, S., Pathe, P.P., Kaul, S.N., 2003. Pre-treatment of currency printing
ink wastewater through coagulation-flocculation process. Water Air Soil Pollut.
148, 15e30.
References Nieto, L.M., Hodaifa, G., Rodríguez, S., Gime nez, J.A., Ochando, J., 2011. Degradation
of organic matter in olive-oil mill wastewater through homogeneous Fenton-
Alves, R.C., Rodrigues, F., Antonia Nunes, M., Vinha, A.F., Oliveira, M.B.P.P., 2017. like reaction. Chem. Eng. J. 173, 503e510.
Chapter 1 - State of the art in coffee processing by-products. In: Galanakis, C.M. Parmar, A., 2014. Fenton process: a case study for treatment of industrial waste
(Ed.), Handbook of Coffee Processing By-Products. Academic Press, pp. 1e26. water. Int. J. Innovative and Emerging Res. in Eng. 1, 23e30.
APHA, AWWA, WEF, 2005. Standard Methods for the Examination of Water and Pignatello, J.J., 1992. Dark and photoassisted iron(3þ)-catalyzed degradation of
Wastewater, twenty-first ed. ed. American Public Health Association, American chlorophenoxy herbicides by hydrogen peroxide. Environ. Sci. Technol. 26,
Water Works Association, Water Pollution Control Federation, Washington, DC. 944e951.
Barros, V.G.d., Duda, R.M., Vantini, J.d.S., Omori, W.P., Ferro, M.I.T., Oliveira, R.A.d., Pignatello, J.J., Oliveros, E., MacKay, A., 2006. Advanced oxidation processes for
12 V. Gomes de Barros et al. / Environmental Pollution 259 (2020) 113796

organic contaminant destruction based on the fenton reaction and related fenton reaction in the presence of fulvic acid. Environ. Sci. Technol. 37,
chemistry. Crit. Rev. Environ. Sci. Technol. 36, 1e84. 1130e1136.
Poland, J., Pagano, T., 2010. Jar Testing. Engineering Department, Virginia Poly- Standardization, I.O.f., 2005. Water Quality - Determination of the Inhibitory Effect
technic Institute and State University, USA. of Water Samples on the Light Emission of Vibrio Fischeri (Luminescent Bacteria
ringer, P., 1999. Strategy
Pulgarin, C., Invernizzi, M., Parra, S., Sarria, V., Polania, R., Pe Test) - Part 3: Method Using Freeze-Dried Bacteria.
for the coupling of photochemical and biological flow reactors useful in Suarez, W.A.B., Vantini, J.d.S., Duda, R.M., Giachetto, P.F., Cintra, L.C., Ferro, M.I.T.,
mineralization of biorecalcitrant industrial pollutants. Catal. Today 54, Oliveira, R.A.d., 2018. Predominance of syntrophic bacteria, Methanosaeta and
341e352. Methanoculleus in a two-stage up-flow anaerobic sludge blanket reactor
n, F.J., Frades, J., Buxeda, P., 2001. Oxidation of p-hydroxybenzoic
Rivas, F.J., Beltra treating coffee processing wastewater at high organic loading rate. Bioresour.
acid by Fenton’s reagent. Water Res. 35, 387e396. Technol. 268, 158e168.
Rodrigues, C.S.D., Boaventura, R.A.R., Madeira, L.M., 2014a. Technical and economic Szpyrkowicz, L., Juzzolino, C., Kaul, S.N., 2001. A Comparative study on oxidation of
feasibility of polyester dyeing wastewater treatment by coagulation/floccula- disperse dyes by electrochemical process, ozone, hypochlorite and fenton re-
tion and Fenton’s oxidation. Environ. Technol. 35, 1307e1319. agent. Water Res. 35, 2129e2136.
Rodrigues, C.S.D., Boaventura, R.A.R., Madeira, L.M., 2014b. A new strategy for Torrades, F., Perez, M., Mansilla, H.D., Peral, J., 2003. Experimental design of Fenton
treating a cotton dyeing wastewater - integration of physical-chemical and and photo-Fenton reactions for the treatment of cellulose bleaching effluents.
advanced oxidation processes. Int. J. Environ. Sci. Technol. 14, 232e255. Chemosphere 53, 1211e1220.
Rodrigues, C.S.D., Borges, R.A.C., Lima, V.N., Madeira, L.M., 2018. p-Nitrophenol USP Technologies, 2019. Fenton’s reagent general chemistry using H2O2. acessed
degradation by Fenton’s oxidation in a bubble column reactor. J. Environ. 09.06.19. http://www.h2o2.com/industrial/fentons-reagent.aspx?
Manag. 206, 774e785. pid¼143&name¼General-Chemistry-of-Fenton-s-Reagent.
Rodrigues, C.S.D., Madeira, L.M., Boaventura, R.A.R., 2013. Treatment of textile dye Verma, S., Prasad, B., Mishra, I.M., 2010. Pretreatment of petrochemical wastewater
wastewaters using ferrous sulphate in a chemical coagulation/flocculation by coagulation and flocculation and the sludge characteristics. J. Hazard Mater.
process. Environ. Technol. 34, 719e729. 178, 1055e1064.
Rodrigues, C.S.D., Neto, A.R., Duda, R.M., Oliveira, R.A.d., Boaventura, R.A.R., Villa-Montoya, A.C., Ferro, M.I.T., de Oliveira, R.A., 2017. Removal of phenols and
Madeira, L.M., 2017a. Combination of chemical coagulation, photo-Fenton methane production with coffee processing wastewater supplemented with
oxidation and biodegradation for the treatment of vinasse from sugar cane phosphorous. Int. J. Environ. Sci. Technol. 14, 61e74.
ethanol distillery. J. Clean. Prod. 142, 3634e3644. Villanueva-Rodríguez, M., Bello-Mendoza, R., Wareham, D.G., Ruiz-Ruiz, E.J., Maya-
Rodrigues, C.S.D., Soares, O.S.G.P., Pinho, M.T., Pereira, M.F.R., Madeira, L.M., 2017b. Trevin~ o, M.L., 2014. Discoloration and organic matter removal from coffee
p-Nitrophenol degradation by heterogeneous Fenton’s oxidation over activated wastewater by electrochemical advanced oxidation processes. Water, Air, Soil
carbon-based catalysts. Appl. Catal. B Environ. 219, 109e122. Pollut. 225, 1e11.
Rossmann, M., Matos, A.T., Abreu, E.C., Silva, F.F., Borges, A.C., 2013. Effect of influent Walling, C., 1975. Fenton’s reagent revisited. Acc. Chem. Res. 8, 125e131.
aeration on removal of organic matter from coffee processing wastewater in Wang, J., Guo, Y., Gao, J., Jin, X., Wang, Z., Wang, B., Li, K., Li, Y., 2011. Detection and
constructed wetlands. J. Environ. Manag. 128, 912e919. comparison of reactive oxygen species (ROS) generated by chlorophyllin metal
Sanabria, N.R., Peralta, Y.M., Montan ~ ez, M.K., Rodríguez-Valencia, N., Molina, R., (Fe, Mg and Cu) complexes under ultrasonic and visible-light irradiation.
Moreno, S., 2012. Catalytic oxidation with Al-Ce-Fe-PILC as a post-treatment Ultrason. Sonochem. 18, 1028e1034.
system for coffee wet processing wastewater. Water Sci. Technol. 66, Yalfani, M.S., Contreras, S., Medina, F., Sueiras, J., 2009. Phenol degradation by
1663e1668. Fenton’s process using catalytic in situ generated hydrogen peroxide. Appl.
Sellers, R.M., 1980. Spectrophotometric determination of hydrogen peroxide using Catal. B Environ. 89, 519e526.
potassium titanium(IV) oxalate. Analyst 105, 950e954. Zaghdoudi, M., Fourcade, F., Soutrel, I., Floner, D., Amrane, A., Maghraoui-
Selvamurugan, M., Doraisamy, P., Maheswari, M., 2010. An integrated treatment Meherzi, H., Geneste, F., 2017. Direct and indirect electrochemical reduction
system for coffee processing wastewater using anaerobic and aerobic process. prior to a biological treatment for dimetridazole removal. J. Hazard Mater. 335,
Ecol. Eng. 36, 1686e1690. 10e17.
Sharp, E.L., Parsons, S.A., Jefferson, B., 2006. The impact of seasonal variations in Zayas, T.P., Geissler, G., Hernandez, F., 2007. Chemical oxygen demand reduction in
DOC arising from a moorland peat catchment on coagulation with iron and coffee wastewater through chemical flocculation and advanced oxidation pro-
aluminium salts. Environ. Pollut. 140, 436e443. cesses. J. Environ. Sci. 19, 300e305.
Soares, O.S.G.P., Rodrigues, C.S.D., Madeira, L.M., Pereira, M.F.R., 2019. Heteroge- Zhang, M.-h., Dong, H., Zhao, L., Ri, D.-x., Meng, D., 2019. A review on Fenton process
neous fenton-like degradation of p-nitrophenol over tailored carbon-based for organic wastewater treatment based on optimization perspective. Sci. Total
materials. Catalysts 9, 1e18. Environ. 670, 110e121.
Southworth, B.A., Voelker, B.M., 2003. Hydroxyl radical production via the photo-

You might also like