Evaluation of solar photo-Fenton and ozone based processes as citrus wastewater pre-treatments

Accepted Manuscript Evaluation of Solar Photo-Fenton and Ozone Based Processes as Citrus Wastewater Pre-treatments José Guzmán, Rosa Mosteo, Judith Sarasa, José A. Alba, José L. Ovelleiro PII: DOI: Reference:

S1383-5866(16)30130-7 http://dx.doi.org/10.1016/j.seppur.2016.03.025 SEPPUR 12912

To appear in:

Separation and Purification Technology

Received Date: Accepted Date:

12 February 2016 16 March 2016

Please cite this article as: J. Guzmán, R. Mosteo, J. Sarasa, J.A. Alba, J.L. Ovelleiro, Evaluation of Solar PhotoFenton and Ozone Based Processes as Citrus Wastewater Pre-treatments, Separation and Purification Technology (2016), doi: http://dx.doi.org/10.1016/j.seppur.2016.03.025

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Evaluation of Solar Photo-Fenton and Ozone Based Processes as Citrus Wastewater Pretreatments. José Guzmán*, a, b, Rosa Mosteoa, Judith Sarasaa, José A. Albaa, José L. Ovelleiroa a

Department of Chemical Engineering and Environmental Technologies, Environmental Sciences

Institute (IUCA),University of Zaragoza, C/ María de Luna 3, 50018 Zaragoza, Spain b

Institute on Tropical Fruit Research, Ave 7mano. 3005 e/30 and 32 Miramar, La Habana, Cuba

* Corresponding author. Tel.: + 34 876555039. Fax: + 34 976762142. Email: [email protected].

Nomenclature: Code

Full name

AOPs

advanced oxidation processes

BOD5

biochemical oxygen demand in 5 days

COD

chemical oxygen demand

CWW

citrus wastewater

DOC

dissolved organic carbon

PF

solar photo-Fenton process

ROS

reactive oxygen species

SHE

standard hydrogen electrode

TSS

total suspended solids

1

ABSTRACT In the present study, different ozone based processes (O3, O3/OH-, O3/UV, O3/H2O2 and O3/UV/H2O2) and solar photo-Fenton treatment were evaluated for the pre-treatment of synthetic samples of citrus wastewater. The highest removal of organic matter in ozone based processes was achieved at high pH values, using H2O2 and UV radiation at 254 nm. After the application of ozone for 150 min. at 1.9 gO3/L, 1017 mg/L of H2O2, UV radiation and pH~ 7, the removal efficiency of chemical oxygen demand and dissolved organic carbon was 15.7% and 10.9%, respectively. Apparent color, real color and turbidity were reduced significantly, mainly at the initial stages of the reaction. The solar photoFenton treatment (pH of citrus wastewater, [H2O2] = 15937 mg/L, 510 mg/L of Fe3+ and reaction time of 30 min) produced a removal of 77% of chemical oxygen demand and 53% of dissolved organic carbon. This photocatalytic process is the most appropriate technique for the pretreatment of citrus effluent. An estimation of the total production costs of the photo-Fenton process shows that when solar light is used the cost of the treatment is 13.8€/m3.

KEYWORDS: Citrus wastewater; Pre-treatments; Solar Photo-Fenton; Ozone; Cost estimation.

2

1. Introduction Citrus cultivation (mainly oranges) is a major industry and a significant economic sector in the United States, Brazil, Mexico, China, India, Iran, and most Mediterranean countries, including Spain and Greece [1]. According to the Statistical Database of the Food and Agriculture Organization of the United Nations, world orange production in 2012 was estimated to be 68.2 million tons. A high percentage of this produce (~84%) is used to manufacture products such as juice, marmalade, etc., which generates large quantities of waste. Citrus wastewaters are generated mainly during the production of citrus juice, the extraction of essential oils as byproduct, and during the cleaning operations of equipment and industrial installations. These wastewaters contain biodegradable organic matter and they can be treated by aerobic or anaerobic biological processes [1–7]. However, these effluents are characterized by very high organic loads (COD: 1000–10000 mgO2/L), high variability of pH (normally acid), the presence of low concentrations

of

nutrients

(especially

nitrogen

and

phosphorus)

and

flavonoids

and

heteropolysaccharides, as for instance hesperidin and pectin, in colloidal form [8]. Another peculiarity of the citrus effluent is the presence of essential oils in high concentrations that can produce serious problems in the aerobic biological treatment plants, commonly used for the treatment of this industrial wastewater [9]. Recently, citrus wastewater evaluated using toxicity bioassays on aquatic macroinvertebrates and using biochemical biomarkers was classified as toxic [10]. The pre-treatment of this kind of effluent by chemical oxidation, especially with advanced oxidation processes (AOPs), is able to oxidise biorrefractory pollutants into a more easily biodegradable form and to reduce the high concentration of organic matter. The AOPs are defined as oxidation processes in which hydroxyl radicals are the main oxidants involved. This type of radical is a very powerful oxidant (E0 2.80 V vs. SHE) which leads to a very effective oxidation process. In addition to their efficiency, these technologies have another important advantage: no dangerous or persistent by-products are formed as a consequence of the reduction of the oxidizing agent. 3

The hydroxyl radical is produced from single oxidants such as ozone (O3), or from a combination of strong oxidants such as O3 and hydroxide (OH-), O3 and hydrogen peroxide (H2O2), or ferrous or ferric ions (Fe2+/Fe3+) with H2O2 and combining these processes with UV radiation [11]. The combination of Fe2+and H2O2 is called the Fenton reaction. Ozone in water can follow two pathways: direct oxidation of compounds by molecular ozone (E 0 2.07 V vs. SHE) and indirect oxidation through hydroxyl free radicals produced during the decomposition of ozone and from reactions between ozone and some organic and inorganic species in water [12]. Hoigné and Bader [13] found that under acidic conditions, direct oxidation with molecular ozone is of primary importance. Under conditions favoring hydroxyl free radical production, such as high pH (O3/OH-), exposure to UV or addition of hydrogen peroxide (O3/H2O2), the hydroxyl oxidation starts to dominate [14]. As well as hydroxyl radicals, other radicals are generated such as superoxide, ozonide and hydroperoxide radicals. Ozonation and ozone related AOPs have been used for the removal of organic compounds from fruit and vegetable processing industry wastewaters [14-16]. The Fenton oxidation process is based on the use of a combination of hydrogen peroxide and ferrous or ferric ion to form active hydroxyl radicals in acidic solutions in a very simple and cost-effective way [17]. In the photo-Fenton process (PF), additional reactions occur in the presence of light that produce hydroxyl radicals or increase the production rate of hydroxyl radicals [18], thus increasing the efficiency of the process. This technique has been studied for the treatment of winery wastewater [19– 21] and olive mill wastewater [22], among other agroindustrial effluents. The aim of this research is thus to investigate the efficiency and feasibility of natural solar photoFenton treatment and different ozone-based advanced oxidation processes (O3, O3/OH-, O3/UV, O3/H2O2 and O3/UV/H2O2) for the pre-treatment of citrus wastewater. This research is focused to reduce the organic load and inhibitory substances present in the industrial effluent for subsequent

4

aerobic biological treatment. Furthermore, an estimation of the total costs involved in the best treatment (in terms of organic matter removal) has been made.

2. Experimental 2.1. Analytical methods and reagents Different physicochemical parameters were analyzed in the samples. The concentration of organic matter was determined by the chemical oxygen demand (COD) and the dissolved organic carbon (DOC). The COD was determined by the colorimetric method of closed reflux, according to EPA Method 410.4 [23]. The DOC analysis was performed on the Shimadzu TOC-VCSH analyzer following SM 5310-B of the “Standard Methods” [24]. Other parameters as pH (SM 4500-H+ B), total suspended solids (TSS) (SM 2540-D), biochemical oxygen demand (BOD5) (SM 5210-B) and color (SM 2120-C) were done according to “Standard Methods” [24]. Conductivity was performed following ISO Norm UNE 27888:1994, turbidity according to ISO 7027:1999 and volatile acidity following UNE 34229:1981. The total acidity was determined by titration with sodium hydroxide 0.1N and is expressed as percentage of citric acid. The d-limonene was determined as essential oils recovered by extraction-titration by Scott method [25]. All analysis were performed in triplicate. Pure oxygen (high purity oxygen 100%) from Air Products Group was used in the ozone based treatments for ozone generation. The ozone generated and ozone not consumed was calculated by the iodometric method [26]. Residual ozone, dissolved in the sample, was measured by a color-meter test (Ozone Photometric Tester, Merck) together with a UV/Vis spectrophotometer (Thermospectronic, Helios α). The reagents used for the solar photo-Fenton experiments were ferric chloride solution (FeCl3·6H2O, Probus®) and commercial hydrogen peroxide (H2O2 (30% (v/v), Carlo Erba®). The concentration of total and dissolved iron was analyzed with the phenantroline method 3500-B of 5

“Standard methods” [24] using a Multiparameter Hanna HI 83099 photometer in order to check iron precipitation. The concentration of hydrogen peroxide was monitored using Merck Merckoquant Peroxide Test strips (0–25 mgH2O2/L, 0–100 mgH2O2/L and 0–1000 mgH2O2/L) and with the spectrophotometric method of metavanadate [27].The pH parameter was adjusted using sodium hydroxide solution (1N NaOH) from Panreac®. Sodium thiosulphate (0.0551N Na2S2O3) from Panreac® was used for neutralizing the residuals of ozone and H2O2 at the end of the experiments. 2.2 Wastewater Owing to the seasonal character of citrus juice production, which generates wastewaters for about 6 months of the year, it was necessary to reproduce citrus effluents by means of synthetic samples (CWW). These samples were prepared by diluting natural orange juice in distilled water, taking into account the characteristic of real effluent generated in a citrus industry from Cuba [28]. In general, real wastewaters show a low pH, high biodegradability, and high concentrations of organic matter, suspended solids and turbidity (Table 1). After dilution of orange juice, the samples were filtered using pressure filters nonwoven fabric (Filter-Lab® NW25L) to remove solid particles and suspended solids. The objective was to avoid possible interference in the oxidation process and transmission of the radiation in the photocatalytic treatments. The physicochemical characterization of the CWW and its comparison with real wastewaters is presented in Table 1. The processes based on ozone were performed with a dilution of CWW (CODinitial of 6000 mgO2/L). The photo-Fenton process was carried out without dilution of CWW for all assays (CODinitial of 10000 mgO2/L). 2.2. Ozone based processes procedure Ozone based treatments were performed in a spherical reactor (2 L) made of pyrex glass. Ozone was produced using an ozone generator (Fischer Model 500). The ozone and oxygen mixture was continuously introduced into the reactor through a bubble diffuser contactor placed at the bottom (inlet 6

ozone partial pressure in the gas mixture 0.82 kPa) and 1.5 L of CWW was treated for 150 min. The gas flow rate was 50 LO2/h, resulting in an input ozone concentration of 1.9 gO3/L (O3generated). Ozone not transferred into the process water during contact was released from the reactor as not consumed O3 (O3not consumed). This not consumed O3 was routed to two ozone destruct units containing 250 mL of a 2% KI solution. When the ozone reacted with this KI, it was reduced to oxygen and released to the atmosphere as an innocuous element. The ozone residual dissolved in the sample (O3residual dissolved) was lower than 0.1 mg/L for all assays in accordance with other authors [29]. This value can be considered insignificant for the calculation of the ozone consumed (O3consumed) by the sample, which was obtained as the difference between O3generated and O3not consumed.The ozonation efficiency was calculated according to the following expression: ozonation efficiency (%) = (O3consumed/O3generated) *100 [16]. In the O3/H2O2 process (Peroxone), the selected H2O2 concentrations were 1017 and 2000 mg/L, taking into account two criteria: the molar ratio [O3/H2O2] = 0.75 [30] and the weight ratio [COD/H2O2] = 3 [14]. Furthermore, the influence on the methodology of the addition of H2O2 was studied: (i) one addition at the initial time and (ii) stepwise addition. In the photo-oxidative treatments (O3/UV and O3/UV/H2O2), the reactor was equipped with a lowpressure mercury lamp (light intensity of 660 W/m2, Helios Italquartz) emitting nearly monochromatic light at 254 nm. In the O3/UV/H2O2 process a concentration of 1017 mg/L of H2O2 was used. Ozone based treatments were carried out at different pHs: real pH of CWW (pH~4), 9 and 10. In the O3/H2O2 and O3/UV/H2O2 processes, some experiments with control of the pH (pH~7) during the reaction time were carried out. According to Staehelin and Hoigné [31], the lower limit for the effectiveness of the H2O2/O3 process is in a pH range of 5–7. In each experiment, samples were periodically taken at regular time intervals from the reaction vessel and immediately analyzed to avoid further reaction. All experiments were carried out in duplicate at room temperature. 2.3. Solar Photo-Fenton treatment procedure 7

The photo-Fenton experiments were assisted with natural solar radiation and carried out in batch systems in quartz flasks of 100 ml. The assays were performed during sunny days in September at the University of Zaragoza (Spain) located at 247 m above sea level, 41°39’N, 1°00’29E, using natural solar light between 11:00–12:00 hours. The mean global irradiation was close to 2424 kJ/m2, as values reported by the Spanish Climate Agency [32]. The conditions were selected taking into account the previous results of the photo-Fenton process in solar chamber that demonstrated its effectiveness in the treatment of effluents from the citrus industry [33]. The experimental runs were reproduced using natural solar irradiation in the combination of factors that maximize the yield of the removal of organic matter (measured as DOC): concentration Fe3+: 510 mg/L and hydrogen peroxide dosage in an interval of 50-125% (established as a percentage of the stoichiometric ratio in weight between the H2O2 and the COD, R = H2O2/COD = 2.125 [34]). The CWW was stirred before the addition of hydrogen peroxide and during the treatment at 45 rpm. All experiments were conducted at the natural pH of the synthetic samples during 30 min, which ensured total consumption of the H2O2 concentration employed. In order to detect possible interferences, a control sample without the addition of Fenton reagents was used in each solar photoFenton experiment to quantify losses by volatilization and other causes. Each experiment was carried out in triplicate.

3. Results and discussion 3.1. Ozonation Table 2 shows the results of ozonation experiments carried out at different pHs. As can be observed, the dissolved organic matter present in CWW is not removed by ozonation since the maximum reduction was 3%. According to the literature, this can be explained in terms of the generation of high concentrations of intermediates that cannot be oxidized such as carboxylic acids which hardly react 8

with ozone [12]. Consequently, they are accumulated in the system and contribute to the concentration of organic matter in the treated effluent. In addition, a decrease in pH was observed in all the tests, which is attributed to the generation of dicarboxylic acids and the formation of several compounds with a low molecular weight. Acids, hydroxyl aldehydes, carbonyl compounds and epoxides are the most frequently cited in the literature as oxidation products of β-carotene [35]. At acidic pH (real pH of CWW), only a direct ozone attack can be expected to degrade the organic matter, while in the ozonation process at alkaline pH several oxidation mechanisms coexist. At alkaline pH, ozone self-decomposition to radicals occurs as a result of the initiation reaction (Eq. (1)) [36]. Similar results of the organic matter degradation have been obtained in the pH range studied in the present experiments, which indicate the low oxidizability of the pollutants contained in the CWW by ozonation. O3 +OH −→ HO2− + O2

(k= 40 M−1 s−1, pH = 10)

(1)

However, the color was very effectively reduced, reaching values close to 100%, and the turbidity achieved reductions higher than 70% in all cases. The color of citrus effluents is due to the presence of compounds such as carotenoids which contribute to the yellow, orange or red color in orange juice [37]. The basic chemical structure of the carotenoids consists of tetraterpenoids connected by opposite units at the center of the molecule with a polyenic chain ranging from 3 to 15 conjugated double bonds [38]. All these double bonds are potential sites for the occurrence of reactions with ozone, promoting the color removal. Fig. 1a shows the evolution of color removal during the reaction time in the ozonation experiments. As a result of the treatments, absorbance at 420 nm decreased significantly. CWW total decolorization (real color) occurred mainly at the initial stage of the reaction (50 min).The apparent color showed a marked reduction especially in the first 60 min of reaction, although it is not totally removed.

9

The ozonation efficiency was also determined in the experiments carried out in the present study. The efficiency values are summarized in Table 2. In all the experiments, efficiencies after 60 min of reaction (time required for total real color removal) reached values of around 40% and rose up to 46% when the pH was increased up to 10. These efficiencies indicate that a high ozone dosage left the reactor without being consumed by degradation reactions. 3.2. Ozone combined with hydrogen peroxide (O3/H2O2) process The results of ozone combined with hydrogen peroxide (Peroxone process) are shown in Table 3. It can be observed that the O3/H2O2 system was more effective in the experiments with initial conditions of pH>7.However, the results of COD and DOC removal were not significant. The production of hydroxyl radicals by the reaction of ozone with H2O2 (Eq. (2)) occurs at a slower rate. These results confirmed that the oxidation of the organic matter by ozone and hydroxyl radicals generated in the aqueous solution may only lead to partial oxidation of the pollutants present in CWW. O3+ H2O2 →HO2• + •OH + O2

(k= 6.5×10−2M−1 s−1)

(2) At the real pH of the CWW, and in conditions in which the pH is not controlled, ozone does not decompose to more reactive radical species since the reaction is developed mainly at acidic pH values. Furthermore, in these conditions the "scavenger" action of H2O2 is accentuated because the HO2concentration is negligible (Eq. (3)) and H2O2 is found in excess, thus decreasing the total production of reactive oxygen species (ROS). H2O2 + H2O

HO2- + H+

(pKa = 11.6)

(3)

An increase in the H2O2 concentration from 1017 to 2000 mg/L at pH 9 improves the yield of organic matter removal (Table 3). However, the results do not improve the removal of apparent color removal, turbidity, DOC and COD achieved in the experiments carried out with 1017 mg/L and pH

10

control (pH~7) during the treatment. In fact, when the applied H2O2 concentration is above the optimum value, H2O2 acts as a radical scavenger [39]. The stepwise addition of hydrogen peroxide during the Peroxone process is more efficient than the use of one addition at the initial time, increasing the removal percentage of organic matter (Table 3). This trend agrees with the results of Kosaka et al. [40]. This result shows that the composition of CWW varies throughout the treatment, and therefore the requirement of H2O2 is a function of the time and the use of such doses. This is reflected in the residual hydrogen peroxide analysis, since the H2O2 concentration was lower in the experiments with stepwise addition than in the assay with the addition at the initial time. The evolution of the color removal for all the conditions is also presented in Fig. 1b. There was a significant decrease in decolorization during the first minutes of reaction (30 min), similar to the ozonation processes. The apparent color shows an important reduction in 60 min of treatment, except for the experiments at the real pH of the sample. It can be assumed that at the beginning of the reaction, the easily oxidisable substances are removed. As the oxidation continues, the organic compounds that can be easily oxidised become less available and some generated intermediate compounds become increasingly important scavengers of hydroxyl radicals. The ozonation efficiency after 60 minutes of treatment (Table 3) shows values from 43% up to 66% in the experiments carried out with 1017 mg/L of hydrogen peroxide. In some cases, the presence of H2O2 increases the ozone efficiency, although this is probably not due to reactions with the organic matter present in the CWW but rather to the direct reaction of ozone with hydrogen peroxide. The greater efficiency observed when the pH is increased is also probably due to the increase in the reaction rates (Eqs.(1 and 2)). 3.3. UV-assisted ozonation (O3/UV) and combined with hydrogen peroxide (O3/UV/H2O2) processes

11

Tables 4 and 5 show the results of UV photo-assisted ozonation (O3/UV) and combined with hydrogen peroxide (O3/UV/H2O2), respectively. In these assays, the highest COD and DOC removal was observed at alkaline pH as a result of the fast reaction of the organic matter with the radical species (i.e., •OH, HO2•, O2•− and O3•−). At alkaline pH, ozone self-decomposition to radicals occurs (Eq.(1)). In the presence of UV radiation, the photodecomposition of ozone leads to the formation of H2O2 and hydroxyl radicals, as is shown in Eq.(4). Furthermore, the reaction of ozone with H2O2 (Eq. (2)) also produces hydroxyl radicals, although at a slower rate. O3 + H2O + hʋ → H2O2 + O2

hʋ< 310nm

(4)

In addition, the H2O2 is deprotonated preferentially to water to give the hydroperoxide anion HO2− (Eq. (3)). The photolysis of H2O2 by UV radiation yields two hydroxyl radicals from each molecule of hydrogen peroxide (Eq. (5)). This also explains the partially higher yields of organic matter degradation observed with the O3/UV and O3/UV/H2O2 processes in comparison to the other ozone-based processes. In the O3/UV/H2O2 process, the production of hydroxyl radicals by Eq. (5) becomes highly significant due to the supplementary H2O2 added to the solution, as well as Eq. (6) under an excess of peroxide. These results confirm that the oxidation of this effluent was mainly due to hydroxyl radicals. H2O2 + hʋ → 2•OH •

(φOH. = 1)

OH + H2O2 → O2•− + H2O + H+

(5) (k= 2.7×107M−1 s−1)

(6)

The pH in both treatments followed the same evolution as that in the ozone-based experiments carried out in the absence of radiation. This effect is counteracted in the O3/UV/H2O2 process where the pH is controlled to remain neutral, resulting in higher yields of COD and DOC removal, 15.4 and 10.9 mg/L, respectively. The use of O3/UV and O3/UV/H2O2 processes enhanced the oxidation rate of the CWW compared to the ozone-alone or the Peroxone process. The results show a fast decrease in the removal of the real color, up to values close to 100% in the first 15–20 min of treatment (Figs. 1c and 1d). Similar to 12

previous treatments, the apparent color shows a marked reduction at the beginning of the process (60 min). It is interesting to observe that for the real pH of CWW, the apparent color removal is faster in the O3/UV/H2O2 process compared to the other treatments. The turbidity shows a reduction greater than 90% in the experiments under the initial conditions of pH~7, except for the treatment O3/UV at real pH of the sample. Furthermore, the ozone efficiency after 60 minutes of reaction (Tables 4 and 5) increases with the pH in both treatments, from 44% (acidic pH) to 57–70% (alkaline pH). At the real pH of the CWW, there was no difference between the different ozone-based treatments, whilst at pH 9 or 10 there was about a 10–20% difference between the ozone efficiencies obtained during the ozone-alone treatment and about a 25–30% difference with respect to the O3/H2O2 process. In the O3/H2O2/UV process higher ozone consumption was observed suggesting a more effective use of the ozone supplied to the system. 3.4. Solar Photo-Fenton treatment In the solar photo-Fenton treatment neither H2O2 nor Fe3+ should be overdosed, to ensure that the maximum amount of •OH radicals is available for the oxidation of organic compounds [41].Generally, the degradation rate for organic compounds increases as the H2O2 concentration increases until a critical H2O2 concentration is achieved [42]. It is clearly shown in Table 6 that increasing the H2O2 concentration leads to increases in the removal efficiency of up to 71% at a dose of 125% of stoichiometric value (26562 mg/L). This is explained by the effect of the higher production of •OH radicals as a result of the addition of H2O2. Nevertheless, the small difference between the DOC removal attained with 1.6 and 2.7 of the ratio [H2O2/COD] (w/w) (an increase of 66% of the stoichiometric dosage) indicates that improvements in terms of degradation may not be worth the large loads of oxidant used. This result might be attributed to the scavenging effect of peroxide on the hydroxyl radicals, which became stronger as the ratio [H2O2/Fe3+] increased.

13

With a dosage of 75% of the stoichiometric ratio in weight (weight ratio of [H2O2/Fe3+] < 35.3), Fe3+ = 510 mg/L and 15937 mg/L of H2O2), the oxidation efficiency decreased, probably because there was not enough hydrogen peroxide to form the hydroxyl radicals and/or because of the scavenging effect of Fe2+ on hydroxyl radicals (Eqs. (7 and 8)), thereby inducing low reactivity of the reaction between the radicals and the contaminant to be oxidized [43]. Fe2+ + HO• → Fe3+ + HO-

(7)

Fe2+ + HO2•/O2- • → Fe3+ + H2O2

(8)

In general, it can be observed that the efficiency of organic matter reduction (measured as DOC) after solar photo-Fenton treatment is similar to that obtained in a previous work with artificial radiation in solar chamber [33], although the hydrogen peroxide consumption during the reaction is slower for the solar assisted treatment. It is remarkable the advantage of this treatment as the same results are obtained under variable climate conditions and different light intensity and spectral composition. The greatest contribution to the DOC degradation in these systems is given by preferential attack of the polyphenolic compounds regarding the presence of other contaminants in these effluents. Considering the initial high organic content (COD: 10000 mg/L) and that the polyphenolic compounds are only 1/30 of the organic compounds in citrus wastewater, a large group of compounds still has a negative impact on the degradation of DOC. Thus, the optimum H2O2 dosage was 75% of the stoichiometric value (15937 mg/L). In these operational conditions a good level of degradation of organic matter was achieved after 30 min of treatment. Table 7 shows the characterization of the effluents of CWW after using the optimum solar photoFenton process. Organic matter removal measured as COD was 76.9% and 53.3% measured as DOC. The dissolved total iron after treatment was 287.4 mg/L, significantly less than the initial concentration used. This is mainly due to an increase in the formation of stable complexes between the organic

14

ligands and Fe (III). Consequently, it was observed an increase of the color and conductivity in the treated samples. The detection of volatile acids in the samples after demonstrates the good development of the process. After treatment were determined values of 2.45 g/L of volatile acidity expressed as grams of acetic acid (Table 7). The presence of a significant fraction of easily biodegradable compounds in the effluent in the form of organic acids of low molecular weight accelerates the efficiency of the aerobic metabolism of biomass in a biological treatment. Compounds such as R-(+)limonene, the most important terpene present in essential oil from citrus (90% of the total), were not detected in the samples after completion of treatment. This aspect is of great importance since this compound has shown an inhibitory effect of microbiological activity, which makes the solar photo-Fenton system an adequate pretreatment before biological treatment. From the data reported in the previous sections, it can be inferred that ozone based processes are less efficient than the solar photo-Fenton process for the pre-treatment of citrus effluents, in terms of organic matter removal efficiency. In the photo-Fenton process the rate of HO• regeneration is significantly increased by the interaction of light and complexes in the solution. The efficient use of light quanta in the photo-reduction of ferric ion and the photolysis of Fe3+ organic intermediate chelates are the major reasons for the high efficiency of this method [22]. The application of photo-Fenton processes assisted by sunlight might lower the cost of this treatment, providing a great advantage for application in the debugging of citrus effluents to scale industry. Considering the reduction of the concentration of organic matter and inhibitory substances by the solar photo-Fenton treatment, it can be affirmed that this process is appropriate as a pre-treatment of citrus effluents

15

3.5. Cost estimation of the solar photo-Fenton process In this section, an estimation of the total costs is calculated relating to the operating costs of the photo-Fenton process for the selected conditions described in section referred to solar photo-Fenton treatment, since it is the best alternative of pre-treatment of citrus effluent in terms of organic matter removal. The photo-Fenton processes are often costly in terms of initial outlay of plant equipment, energy requirements, reagents and sludge disposal, although the use of solar energy as a source of radiation reduces operating costs. In this sense, in the cost estimation of the process was considered the treatment assisted with natural solar radiation. Average yearly radiation, 20.8 W/m2, was estimated as the mean UV (typical for the geographic region of the Caribbean) and the mean yearly illuminated hours per day, 12 h [44],were taken as working hours (tw).The operation costs are mainly comprised of the personnel, maintenance, energy and reagents. The analysis includes amortization costs. One of the main factors that increase costs of the photo-Fenton process is the type of photoreactor. In this cost estimation, a flat reactor has been proposed in order to reduce installation costs [45]. The flat reactors are cheap and technically simple (fabrication and maintenance). Although, more accurate studies are needed in order to the extrapolation of results from one experimental device to another, and the use of these results in the scaling-up. The amortization costs (ACf) in €/m3, were obtained taking into account the costs of installation and acquisition of the basic equipment of the treatment system (If) by means of Eq. (9),where Vt is the capacity of plant (m3/d), D is the number of days a year the plant is expected to work (240 days considered the average duration of the season of citrus processing industry in Cuba) and L is the life cycle of the plant (20 years). ACf = If/ (Vt*D*L)

(9)

The operational cost of the photo-Fenton process in € per m3 was estimated considering the amount of Fenton reagents and their commercial prices (0.38 €/kg of FeCl3.6H2O and 0.21€/kg of 30% v/v 16

H2O2), the energy cost (CE) estimated by calculating the power to blower proposed of the aerationagitation system (Eq.(10)), the amount of alkaline agent (calcium hydroxide (0.15 €/kg)) used to remove the iron remaining in the solution by precipitation and to condition the pH to 7–8 at the end of the photo-Fenton process, necessary for the subsequent biological treatment, and the sludge disposal cost (0.12 €/kg). The dose of alkaline agent and volume of sludge produced were determined from assays secondary of the photo-Fenton treatment. Maintenance costs are considered as 2% of amortization costs [46] and the cost of personnel is the cost estimated for a wastewater treatment plant, 0.0712 €/m3[47]. CE = (PE* W* tw)/ Vt

(10)

Where W is the power required, estimated through an energy balance; PE is the power cost, 0.1429 €/kWh. Table 8 shows all the estimated operating costs and the total cost for the PF process in order to treat 3685 m3/day of citrus effluents with an initial COD concentration of 10 g/L (data for a citrus processing plant located in Cuba) [28]. The proposed level of organic matter degradation is 50–60%. The total process cost is estimated by adding the amortization costs and the operating costs. The photo-Fenton treatment assisted with natural solar radiation is a very promising technique for industrial application since the treatment cost would be about 13.8 €/m3. Chatzisymeon et al. [48] estimated a total cost of 273 €/m3 for the treatment of olive-oil mill wastewaters (initial COD of 5100 mgO2/L) by photocatalysis (TiO2/UV), of which 94% were energy consumption costs. These authors achieved conversions of 18% of COD, 63% of total phenols and 66% of discoloration. Muñoz et al. [49] calculated the total cost of a hypothetical industrial solar photo-Fenton treatment plant at 14.1 €/m3, indicating that solar radiation does not result in such high operating costs. Similar results were determined by Machado et al. [50] in the economic estimation of solar water treatment for chip-board production, specifying an estimated total cost of 9 €/m3. 17

The proposed treatment has a higher cost than the one which is currently carried out by the wastewater treatment plant of the studied industry. This plant consists of a first screening stage for solid separation, followed by a neutralization stage and nutrients addition to treat the wastewater by an activated sludge biological system. The operation costs of this treatment are around 0.6 €/m 3 and are associated to the biological treatment. However, this purification system does not reach the required organic matter levels that are established by Cuban legislation [51]. 4. Conclusions According to the results, it can be concluded that ozonation processes are only effective for removing color and turbidity of CWW since the highest organic matter removal was 15.7% and 10.9%, measured by COD and DOC, respectively. The results indicate that, compared with ozone-based processes, the photo-Fenton process produces higher COD and DOC removal, 76.9% and 53.3% respectively. An economic analysis of the total production costs of the photo-Fenton process showed that the cost for photocatalytic treatments when solar light is used, is 13.8 €/m 3. The photo-Fenton treatment assisted with natural solar radiation is a very promising technique for industrial application.

Acknowledgements We acknowledge financial support for this work from the DGA-FSE Research Team T33. We thank the University of Zaragoza and Santander Bank for a Ph.D. grant awarded to José Guzmán Hidalgo. “Proyecto financiado por el Ministerio de Educación en el marco del Programa Campus de Excelencia Internacional”, Campus Iberus (CEF11–0017).

18

References [1] M.A. Martin, J.A. Siles, A.F. Chica, A. Martin, Biomethanization of orange peel waste, Bioresour. Technol. 101 (2010) 8993–8999. [2] F. Osorio, J.C. Torres, E. Hontoria, Study of biological aerated filters for the treatment of effluents from the citrus industry, J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng. 41 (2006) 2683–2697. [3] J.A. Siles, M.D.L. Martin, A. Martin, F. Rapaso, R. Borja, Anaerobic digestion of wastewater derived from the pressing of orange peel generated in orange juice production, J. Agric. Food Chem. 55 (2007) 1905–1914. [4] M.D.L.M. Santos, J.A. Siles, A.F.C. Perez, A.M. Martin, Modelling the aerobic digestion of wastewater derived from the pressing of orange peel produced in juice manufacturing, Bioresour. Technol. 101 (2010) 3909–3916. [5] M. Elnekave, S.O. Celik, M. Tatlier, N. Tufekci, Artificial Neural Network Predictions of UpFlow Anaerobic Sludge Blanket (UASB) Reactor Performance in Treatment of Citrus Juice Wastewater, Pol. J. Environ. Studies 21 (2012) 49–56. [6] D.A. Zema, S. Andiloro, G. Bombino, V. Tamburino, R. Sidari, A. Caridi, Depuration in aereated ponds of citrus processing wastewater with a high concentration of essential oils, Environ. Technol. 33 (2012) 1255-1260. 19

[7] A. Koppar, P. Pullammanappallil, Anaerobic digestion of peel waste and wastewater for on site energy generation in citrus processing facility, Energy 60 (2013) 62–68. [8] M. Saverini, I. Catanzaro, G. Sciandrello, G. Avellone, S. Indelicato, G. Marci, L. Palmisano, Genotoxicity of citrus wastewater in prokaryotic and eukaryotic cells and efficiency of heterogeneous photocatalysis by TiO2, J. Photochem. Photobiol. B. 108 (2012) 8–15. [9] J. Guzmán, Sustainability of the process of obtaining citrus essential oil (Sostenibilidad del proceso de obtención de aceite esencial cítrico), Academic Publishing GmbH & Co. KG., Alemania, 2012, ISBN: 978-3-8484-5530-0. [10] I. Karaouzas, E. Cotou, T.A. Albanis, A. Kamarianos, N.T. Skoulikidis, U. Giannakou, Bioassays and biochemical biomarkers for assessing olive mill and citrus processing wastewater toxicity, Environ. Toxicol. 26 (2011) 669–676. [11] S. Renou, J.G. Givaudan, S. Poulain, F. Dirassouyan, P. Moulin, Landfill leachate treatment: review and opportunity, J. Hazard. Mater. 150 (2008) 468–493. [12] J. Hoigné, H. Bader, Rate constants of reactions of ozone with organic and inorganic compounds in water– I. Non-dissociating organic compounds, Water Res. 17 (1983) 173–183. [13] J. Hoigné, H. Bader, The Role of Hydroxyl Radical Reactions in Ozonation Processes in Aqueous Solutions, Water Res. 10 (1997) 377–386. [14] M.S. Lucas, J.A. Peres, G. Li Puma, Treatment of winery wastewater by ozone based advanced oxidation processes (O3, O3/UV and O3/UV/H2O2) in a pilot-scale bubble column reactor and process economics, Sep. Purif. Technol. 72 (2010) 235–241. [15] W.K. Lafi, B. Shannak, M. Al-Shannag, Z. Al-Anber, M. Al-Hasan, Treatment of olive mill wastewater by combined advanced oxidation and biodegradation, Sep. Purif. Technol. 70 (2009) 141–146.

20

[16] F.J. Beltrán, J.M. Encinar, J.F. González, Industrial wastewater advanced oxidation. Part 2. Ozone combined with hydrogen peroxide or UV radiation, Water Res. 31 (1997) 2415–2428. [17] Y. Deng, J.D. Englehardt, Treatment of landfill leachate by the Fenton process, Water Res. 40 (2006) 3683–3694. [18] J.J. Pignatello, D. Liu, P. Huston, Evidence for an additional oxidant in the photo assisted Fenton reaction, Environ. Sci. Technol. 33 (1999) 1832–1839. [19] R. Mosteo, M.P. Ormad, E. Mozas, J. Sarasa, J.L. Ovelleiro, Factorial experimental design of winery wastewaters treatment by heterogeneous Photo-Fenton process, Water Res. 40 (2006) 1561–1568. [20] R. Mosteo, M.P. Ormad, J.L. Ovelleiro, Photo-Fenton processes assisted by solar light used as previous step to biological treatment applied to winery wastewaters, Water Sci. Technol. 56 (2007) 89–94. [21] R. Mosteo, J. Sarasa, M.P. Ormad, J.L. Ovelleiro, Sequential Solar Photo-Fenton–Biological System for the Treatment of Winery Wastewaters, J. Agric. Food Chem. 56 (2008) 7333–7338. [22] L. Rizzo, G. Lofrano, M. Grassi, V. Belgiorno, Pre-treatment of olive mill wastewater by chitosan coagulation and advanced oxidation processes, Sep. Purif. Technol. 63 (2008) 648– 653. [23] U.S. Environmental Protection Agency (USEPA), Method 410.4. The determination of chemical oxygen demand, 1993. [24] A.D. Eaton, L.S. Clesceri, E.W. Rice, A.E. Greenberg, M.A.H. Franson, Standard Methods for the Examination of Water and Wastewater, 21st ed., APA-AWWA-WEF, 2005. [25] W. Horwitz. Official methods of analysis of AOAC International, 18th Ed., Gaitherburg, Md: AOAC International, 2005. [26] I.M. Kolthoff, R. Belcher, Volumetric Analysis III, New York, Interscience, 1957.

21

[27] R.F.P. Nogueira, M.C. Oliveira, W.C. Parterlini, Simple and fast spectrophotometric determination of H2O2 in photo-Fenton reactions using metavanadate. Talanta 66 (2005) 86-91. [28] National Company for Analysis and Technical Services UEB Matanzas (Empresa Nacional de Análisis y Servicios Técnicos UEB Matanzas), Monitoring report and characterization of wastewater, Citrus Industrial Company "Héroes de Girón" Jagüey Grande (Informe de monitoreo y caracterización de residuales, Empresa Industrial de Cítricos “Héroes de Girón” de Jagüey Grande), Matanzas, Cuba, 2009. [29] R. García, S. Cortes, J. Sarasa, M.P. Ormad, J.L. Ovelleiro, TiO2-Catalysed ozonation of raw Ebro river water, Water Res. 34 (2000) 1525–1532. [30] M.P. Ormad, Effluents of manufacturing of pesticides derived from DDT and trichlorobenzene. Characterization, control and ozone oxidation (Vertidos de la fabricación de plaguicidas derivados del DDT y triclorobenceno. Caracterización, control y oxidación con ozono), Doctoral Thesis, University of Zaragoza, Spain, 1996. [31] J. Staehelin, J. Hoigné, Decomposition of ozone in water: rate of initiation by hydroxide ions and hydrogen peroxide, Environ. Sci. Technol. 16 (1982) 676–681. [32] www.aemet.es. Official website of Spanish Climate Agency (Agencia Estatal de Meteorología). Accessed March 2015. [33] J. Guzmán, R. Mosteo, M.P. Ormad, J.L. Ovelleiro, Combined Photo-Fenton–SBR Processes for the Treatment of Wastewater from the Citrus Processing Industry, J. Agric. Food Chem. 63 (2015) 391–397 [34] M.S. Lucas, J.A. Peres, Removal of COD from olive mill wastewater by Fenton’s reagent: Kinetic study, J. Hazard. Mater. 168 (2009) 1253–1259.

22

[35] C.M.J. Benevides, M.C.C. Veloso, P.A.P. Pereira, J.B.A de Andrade, Chemical study of βcarotene oxidation by ozone in an organic model system and the identification of the resulting products, Food Chem. 126 (2011) 927–934. [36] H. Tomiyasu, H. Fukutomi, G. Gordon, Kinetics and mechanism of ozone decomposition in basic aqueous solution, Inorg. Chem. 24 (1985) 2962–2966. [37] C. Cortés, M.J. Esteve, D. Rodrigo, F. Torregrosa, A. Frígola, Changes of colour and carotenoids contents during high intensity pulsed electric field treatment in orange juices, Food Chem. Toxicol. 44 (2006) 1932–1939. [38] A.J. Melendez-Martınez, I.M. Vicario, F.J. Heredia, Review: analysis of carotenoids in orange juice, J. Food Compos. Anal. 20 (2007) 638–649. [39] A.Y. Lin, C. Lin, J. Chiou, P.A. Hong, O3 and O3/H2O2 treatment of sulfonamide and macrolide antibiotics in wastewater, J. Hazard. Mater. 171 (2009) 452–458. [40] K. Kosaka, H. Yamada, K. Shishida, S. Echigo, R.A. Minear, H. Tsuno, S. Matsui, Evaluation of the treatment performance of a multistage ozone/hydrogen peroxide process by decomposition by-products, Water Res. 35 (2001) 3587–3594. [41] S. Cortez, P. Teixeira, R. Oliveira, M. Mota, Evaluation of Fenton and ozone-based advanced oxidation processes as mature landfill leachate pre-treatments, J. Environ. Monag. 92 (2011) 749–755. [42] C.L. Hsueh, Y.H. Huang, C.C. Wang, C.Y. Chen, Degradation of azo dyes using low iron concentration of Fenton and Fenton-like system, Chemosphere 58 (2005) 1409–1414. [43] F. Ay, F. Kargi, Advanced oxidation of amoxicillin by Fenton’s reagent treatment, J. Hazard. Mater. 179 (2010) 622–627.

23

[44] M. Álvarez-Guerra, M. Ramos, F. Menéndez, Solar radiation atlas of the Republic of Cuba (Atlas de radiación solar de la República de Cuba), Ed. Academia de Ciencias, La Habana, Cuba, 1995. [45] J. Giménez, D. Curcó, M.A. Queral, Photocatalytic treatment of phenol and 2,4-dichlorophenol in a solar plant in the way to scaling-up, Catal. Today 54 (1999) 229–243. [46] J.A. Sánchez, I.M. Román, I. Carra, A. Cabrera, J.L. Casas, S. Malato, Economic evaluation of a combined photo-Fenton/MBR process using pesticides as model pollutant. Factors affecting costs, J. Hazard. Mater. 244-245 (2013) 195–203. [47] M. Molinos-Senante, F. Hernandez-Sancho, R. Sala-Garrido, Economic feasibility study for wastewater treatment: a cost-benefit analysis, Sci. Total Environ. 408 (2010) 4396–4402. [48] E. Chatzisymeon, E. Diamadopoulos, D. Mantzavinos, Comparison and predesign cost estimation of advanced oxidation processes for olive mil wastewater treatment, 2nd International Conference on Hazardous and Industrial Waste Management, Crete, Greece, 5-8 October, 2010. [49] I. Muñoz, S. Malato, A. Rodriguez, X. Domenech, Integration of environmental and economic performance of processes. Case study on advanced oxidation processes for wastewater treatment, J. Adv. Oxid. Technol. 11 (2008) 270–275. [50] A.E.H. Machado, T.P. Xavier, D.R. De Souza, J.A. De Miranda, E.T.F.M. Duarte, R. Ruggiero, L. De Oliveira, C. Sattler, Solar photo-Fenton treatment of chip board production wastewater, Sol. Energy 77 (2004) 583–589. [51] National Bureau of Standards (NC). Cuba, 1999. NC 27: 1999. Wastewater discharge to terrestrial water and sewerage (Vertimiento de Aguas Residuales a las Aguas Terrestres y al alcantarillado). Specifications (Especificaciones).

24

Figure caption Fig. 1.Effect of pH on color removal in ozone based treatments. (a) Ozonation; (b)O3/H2O2; (c) O3/UV; (d) O3/UV/H2O2.

25

Tables. Table 1. Physicochemical characterization of citrus wastewater (CWW).

Parameter pH Conductivity(S/cm 20 oC) TSS (mg/L) Turbidity (NTU) Real color (CPU) COD (mg/L) BOD5 (mg/L) DOC (mg/L) Biodegradability (BOD5/COD)

Real citrus effluents (mean values) [28] 3.8 998 777 669 10019 6619 0.66

CWW 3.6–4.5 450–550 240–280 130–160 450–100 10000 4246–5252 4218–4260 0.43–0.53

Table 2. Effect of initial pH on COD, DOC, color and turbidity removal efficiencies in ozonation process. Parameter COD removal (%) DOC removal (%) Apparent color removal (%) Real color removal (%) Turbidity removal (%) Final pH O3 efficienciesa (%) a

Initial pH 4 9 2.5 2.7 1.0 2.7 75.9 90.2 100 100 71.5 73 3.0 6.2 40.4 40.4

10 3.0 3.0 93.3 100 73.9 6.4 46.3

60 minutes of reaction time.

26

Table 3. Effect of initial pH and H2O2 concentration on COD, DOC, color and turbidity removal efficiencies in O3/H2O2 process. [H2O2] = 1017 mg/L [H2O2] = 2000 mg/L [H2O2] = 1017 mg/L Initial pH pH control Initial pH pH control Parameter 4 9 10 pH~7 9 pH~7 stepwise addition of H2O2 COD removal (%) 3.5 2.0 4.0 7.3 6.9 13.9 DOC removal (%) 2.7 2.3 3.6 6.1 5.5 10.6 Apparent color removal (%) 54.9 98.5 97.6 97.0 94.0 80.7 Real color removal (%) 100 100 100 100 100 100 Turbidity removal (%) 68.9 98.9 94.4 76.4 92.2 78.9 Final pH 3.4 4.9 5.3 7.8 4.7 7.6 O3 efficienciesa (%) 43.8 52.3 65.7 63.4 ND ND a

60 minutes of reaction time. ND: no data available. Table 4. Effect of initial pH on COD, DOC, color and turbidity removal efficiencies in O3/UV process. Parameter COD removal (%) DOC removal (%) Apparent color removal (%) Real color removal (%) Turbidity removal (%) Final pH O3 efficienciesa (%) a

Initial pH 4 9 1.6 7.0 3.1 13.0 86.4 97.4 100 100 78.7 93.5 2.9 5.3 43.8 57.6

10 9.5 9.2 98.9 100 97.1 5.3 65.9

60 minutes of reaction time.

Table 5. Effect of initial pH on COD, DOC, color and turbidity removal efficiencies in O3/UV/H2O2 process. Parameter COD removal (%) DOC removal (%) Apparent color removal (%) Real color removal (%) Turbidity removal (%) Final pH O3 efficienciesa (%) a

Initial pH 4 9 4.5 11.0 3.0 4.8 95.4 97.4 100 100 91.8 96.4 3.0 4.5 44.4 57.6

10 15.4 8.8 98.7 100 95.9 4.6 70.2

pH control pH~7 15.7 10.9 98.7 100 95.4 7.9 68.4

60 minutes of reaction time. 27

Table 6. Effect of the initial hydrogen peroxide concentration on DOC removal efficiencies in solar photo-Fenton treatment (Fe3+ = 510 mg/L). H2O2 (% H2O2/COD Parameter stoichiometric) (w/w) DOC removal (%) Final pH Residual H2O2 50 1.1 37.3 1.7 <30 75 1.6 53.3 1.7 30-100 83 1.8 59.1 1.6 30-100 125 2.7 71.5 1.5 800-1000

Table 7. Physicochemical parameters of the treated effluent by the optimum solar photo-Fenton process (Fe3+ = 510 mg/L; H2O2 = 15937 mg/L). Parameter pH Conductivity (mS/cm 20 oC) TSS (mg/L) Turbidity (NTU) Real color (CPU) COD (mg/L) BOD5 (mg/L) DOC (mg/L) Biodegradability (BOD5/ COD) Total Fe (mg/L) Volatile acidity (g/L) Fats and oil (mg/L) DL: detection limit

Value 1.5 7.2 77.5 42 267 2315 1198 1980 0.52 287.4 2.45
Removal efficiency (%) 70.2 71.1 76.9 79.4 53.3 -

28

Table 8. Estimated total cost of the process. Expense Operation costs .

Requirements

3

FeCl3 6H2O (kg/m ) 30% (v/v) H2O2(kg/m3) Ca(OH)2 (kg/m3) Subtotal 3

Energy (Kwh/m ) Sludge disposal (kg/m3) Subtotal Maintenance Staff Total costs of operation Amortization Total Cost

Reagents 2.5 53.1 1.5 Facilities 0.3547 10.25 -

Unitary costs (€/m3)

0.94 11.2 0.22 12.36 0.0479 1.23 1.28 0.0026 0.0712 13.7 0.07 13.8

29

Figure 1

30

Graphical abstract

31

Highlights Ozonation and solar photo-Fenton were evaluated for the pre-treatment of CWW The ozone-based processes are only effective to remove color and turbidity The solar photo-Fenton is the most adequate process to remove organic matter An economic study on the best pre-treatment alternative is presented

32