Bioresource Technology 100 (2009) 2189–2197

lable at ScienceDirect

Contents lists avai

Bioresource Technology

journal homepage: www.elsevier .com/locate /b ior tech

Electro-Fenton, hydrogenotrophic and Fe2+ ions mediated TOC and nitrateremoval from aquaculture system: Different experimental strategies

Jurate Virkutyte a,b,1, Veeriah Jegatheesan b,*

a University of Kuopio, Department of Environmental Sciences, Waste Management Group, Yliopistonranta 1E, Kuopio 70211, Finlandb James Cook University, School of Engineering, Townsville, 4811 QLD, Australia

a r t i c l e i n f o

Article history:Received 18 August 2008Received in revised form 21 October 2008Accepted 27 October 2008Available online 12 December 2008

Keywords:Electro-FentonAquacultureNitratesHydrogenotrophic denitrificationFe2+

0960-8524/$ - see front matter � 2008 Elsevier Ltd. Adoi:10.1016/j.biortech.2008.10.050

* Corresponding author. Tel.: +61 7 4781 4871; faxE-mail addresses: jurate.virkutyte@uku.fi (J. V

jcu.edu.au (V. Jegatheesan).1 Fax: +358 153556363.

a b s t r a c t

A number of methods for denitrification were studied including Electro-Fenton method, hydrogenotroph-ic as well as innovative Fe2+ mediated denitrification and their technical feasibility in terms of changes inTOC and nitrate concentrations, effect of different Fenton’s reagent dosage, current and the effect of thepH was investigated. This study was carried out using tailor made electrodialytic reactor. It was foundthat the highest TOC removal was achieved at pH 2.2 and 2.4 (77.1% and 97.8%, respectively) at the anodeand the lowest accumulation of 33% at pH of 6.2 at the cathode. The highest TOC removal in terms ofusing different H2O2 concentrations was achieved at 40 mM reaching as high as 97.3%. Regardless exper-imental strategy, initially nitrates migrated towards the cathode due to the strong hydraulic gradientunder the applied electric current. During the course of experiments, nitrates were transported towardsthe anode where their concentration decreased. The highest nitrate removal was achieved at0.12 mA cm�2 electric current density (94.8%) at the anode and a complete removal at the cathode.Hydrogenotrophic denitrification was the highest reaching 92.5%, however, when Fe2+ ions as electrondonor was used for the destruction of nitrates, only 66.6% removal was achieved. Denitrification usingonly Fe2+ ions was a factor 1.4 less than using electrically generated hydrogen or a Fenton’s reagent.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Due to the increased demand for the fish production throughoutthe world, including Australia, aquaculture has become a very fast-growing and important industry in the past several decades. Aqua-culture systems have advantages of minimal water input andwastewater discharge while allowing full control of the cultureenvironment (Qin et al., 2005). Water quality is an important factorin the production process of intensive aquaculture systems (Singhet al., 1999). According to Jegatheesan et al. (2007), one of the ma-jor environmental issues facing aquaculture is the management ofdischarge of nutrient-rich, i.e. ammonia, nitrite and nitrate watersfrom land based aquaculture systems into the coastal waters.Ammonia is the main contaminant produced by the fish. Short-term exposure of fish to a high concentration of ammonia causesincreased gill ventilation, loss of equilibrium, convulsion and death(Thurston et al., 1981). Nitrite is an intermediate product of deni-trification, which may cause oxidation of blood hemoglobin ironto ferric ion, forming methemoglobin that may result in hypoxia

ll rights reserved.

: +61 7 4781 6678.irkutyte), jega.jegatheesan@

and death of fish (Qin et al., 2005). The nitrate ion is relativelynon-toxic to fishes, however, in excess concentrations may causeeutrophication of water bodies therefore impairing the healthyenvironment for fish.

The removal of NO�3 from water and wastewater has beenextensively researched over the last few decades. Ion exchangeand reverse osmosis processes have been used over the past 15years to treat high nitrate water. However, both processes merelyconcentrate the nitrate into an effluent stream requiring disposal(Kapoor and Viraraghavan, 1997). Moreover, both methods haveseveral disadvantages such as (bio) fouling of membranes and hightechnological (energy) and maintenance costs. A number of alter-native methods for denitrification have been investigated includ-ing electrodialytic, hydrogenotrophic and Fenton oxidationamong a few (Banasiak et al., 2007; Oldani et al., 1992). Electro-chemical methods have shown potential in the transport as wellas destruction of numerous organic and inorganic contaminants.Moreover, electrolytic disinfection of water may also be beneficialin fish culture, increasing survival and the number of fish whichcan be maintained in given volumes of water (Jorguera et al., 2002).

Electrodialytic treatment is based on the application of an elec-tric field to a matrix, where electromigration of charged ions andspecies in the applied electric field as well as uncharged speciesand water due to electroosmosis occurs (Virkutyte et al., 2004).

2190 J. Virkutyte, V. Jegatheesan / Bioresource Technology 100 (2009) 2189–2197

Positively charged species move towards the cathode, negativelycharged species towards the anode and water with unchargedspecies migrate towards the cathode. The electric field will mainlypass the contaminated material in the micropores due to the lowelectric resistance at those sites, and thus, the electric current actswhere the contaminants are mainly found (Rulkens et al., 1998).Moreover, electrodialysis does not require the addition of hazard-ous chemicals and is a simple-to-operate process, which can be eas-ily and fully automated. It adapts to non-continuous feed suppliesand different loads immediately, and temperature changes have anegligible effect on this process. All these characteristics togetherwith its high recovery rate of more than 97% make electrodialysisespecially an attractive method, both ecologically and economi-cally, for the elimination of nitrates from water (Kesore et al., 1997).

Hydrogen peroxide is a strong oxidant with a standard potentialof 1.80 and 0.87 V at pH 0 and 14, respectively and its applicationto the treatment of various inorganic and organic pollutants iswidely documented (Neyens and Baeyens, 2003). The main advan-tage of the process is the complete destruction of contaminants toharmless compounds such as CO2, H2O and inorganic salts.

In closed, such as recirculating aquaculture systems, nitrate re-moval can also be accomplished by enhanced biological denitrifica-tion in which nitrate is reduced to gaseous nitrogen products thatare released into the atmosphere (Grommen et al., 2006). However,the process requires organic donor such as methanol, which canlead to the deterioration of water quality due to the formation oftoxic hydrogen sulfides (Lee et al., 2000). To overcome this prob-lem, hydrogen and Fe2+ ions as electron donors may be used. Bio-logical denitrification coupled to hydrogen-induced reduction hasbeen efficiently utilized to remove nitrates from potable watersupplies (Liessens et al., 1992). A select group of denitrifying bac-teria can autotrophically reduce nitrate, using hydrogen insteadof organic compounds or sulfur as an electron donor (Smithet al., 2001). According to Haugen et al. (2002), hydrogen gas is asafe alternative to organic electron donors, required for the effi-cient denitrification, is not toxic and does not produce undesiredby-products. Furthermore, the reaction products – nitrogen gasand water are innocuous and compatible with human consump-tion and the reaction has a small biomass yield (Smith et al.,2005). Also, upon depletion of unwanted nitrates, excess hydrogencan be easily removed by air stripping or exchange with atmo-sphere upon depletion of nitrates (Smith et al., 2001). The hydro-genotrophic denitrification generates 50% less microbial biomassthan heterotrophic denitrification with methanol as electron donor

Fig. 1. Schematic representation of electrodialytic reactor for t

(Rutten and Schnoor, 1992). Moreover, there is an insignificantthreat of biological regrowth in distribution systems and disinfec-tion by-product formation (Haugen et al., 2002).

The objective of the study was to determine the technical fea-sibility of combined electrical and chemical treatment methodssuch as Electro-Fenton (which utilizes low level electric currentand a Fenton’s reagent), hydrogenotrophic (electrically generatedhydrogen gas on the cathode) and innovative Fe2+ mediated TOCand nitrate removal from a real wastewater – aquaculture sys-tem taking into consideration changes in TOC and nitrate concen-trations, effect of different Fenton’s reagent dosage, current andthe effect of pH.

2. Methods

2.1. Electrodialytic reactor

Electrodialytic set up consisted of a DC power supply (Hewlett–Packard 613 Altai, Germany), which was used to constantly main-tain 0.02–0.16 mA cm�2 DC, glass container (width 20 cm � length25 cm � height 15 cm), connection cables, electrodes and peristal-tic pumps to re-circulate the solution (Fig. 1). The total volume ofthe reactor was 3 L. Platinized titanium rod electrodes (Titaniumcomponents, India) were used in experiments as they have highconductivity, are inert and therefore suitable for long-term appli-cations. The voltage fluctuations were monitored with a Fluke112 multimeter (Fluke, Eindhoven, The Netherlands). Cathodiccompartment was separated by the cation-exchange and anodiccompartment by the anion-exchange membranes (Ionics Inc.,Watertown, MA, USA). The electrodialytic reactor operated be-tween 10 and 80 mA at a constant 30 V for 48 h for each experi-ment. The initial pH was in the range of 7.5–7.7. The initialnitrate concentration was 120–195.5 mg L�1 and TOC was 9.2–11.2 mg L�1. Detailed experimental parameters are presented inTable 1.

At the end of the treatment, the power supply was turned off,the electrode wires were disconnected and the electrodialytic reac-tor was disassembled. After each experiment, anion and cation-ex-change membranes were immersed into the diluted acidic solution(0.5 M) to extract nitrates. The doses of Fe2+ and H2O2 were ad-justed by 1:15 ratio of FeSO4 � 7H2O (T.J. Baker, Deventer, TheNetherlands) and fresh H2O2 (Sigma–Aldrich LaborchemikalienGmbH, Seelze, Germany) aqueous solutions of the correspondingconcentrations.

he TOC and nitrate removal from aquaculture wastewater.

Table 1Experimental conditions.

Parameter Value

pH 7.5–7.7Temperature (�C) 24 ± 1Duration of experiments (h) 48Initial nitrate concentration (mg L�1) 120–195.5Initial TOC concentration (mg L�1) 9.2–11.2Current densities (mA cm�2) 0.02; 0.06; 0.12; 0.16Voltage (V) 30Current (mA) 10; 30; 60; 80H2O2 concentration (mM) 5; 20; 40; 80Fe2+ ion concentration (mM) 10; 30; 100; 200

J. Virkutyte, V. Jegatheesan / Bioresource Technology 100 (2009) 2189–2197 2191

2.2. Calculation

The amount of hydrogen gas produced during experiments inthe electrodialytic cell was calculated according to Grommenet al. (2006):

H2ðgÞ ¼ I � t � ðe� AÞ�1 ðiÞ

where I is current (A), t is time (s), e is the elemental charge of theelectron (1.602 � 10�19 C) and A is Avogadro number (6.02 � 1023).The nitrate and TOC removal is determined by

R ¼ ð1� C1=C0Þ � 100% ðiiÞ

where C1 is the concentration of NO�3 –N or TOC (mg L�1) after thetreatment and C0 is the initial concentration of NO�3 –N or TOC(mg L�1).

2.3. Chemical analyses

Collected samples were stored in clean glass bottles at 4 �C for24 h prior analysis. Nitrates were analyzed colorimetrically usingHach DR980 analyzer using modified cadmium reduction tech-nique (Smith et al., 2001). Cadmium metal is used to reduce ni-trates (NO�3 ) to nitrites (NO�2 ). Next, the nitrite ions react in anacidic medium with sulfanilic acid to form an intermediate diazo-nium salt, which, when coupled with gentisic acid, forms an am-ber-colored compound. Color intensity of the compound is indirect proportion to the nitrate concentration of the water sample(APHA, 1995).

TOC (Shimadzu TOC-5000A, Japan) analyzer was used to mea-sure the dissolved organic concentration in wastewater samplesusing combustion methods at temperature 680 �C, measuringrange <20 mg L�1 and sample injection volume of 50–200 lLaccording to Standard Method 5310C (Haugen et al., 2002). Theaccuracy of measured values for TOC was estimated around 5%.Samples were tested in triplicates. The pH was measured using

0

20

40

60

80

100

120

140

160

6.8 5.7 4.8 4.2 3.2 2.2 7.

Var

iati

on

in T

OC

(%

)

3

Fig. 2. Evolution of TOC with change in pH dur

radiometer electrode. Temperature was kept at ambient(24 ± 1 �C) levels.

2.4. Quality assurance

The analysis of samples followed the standard quality assuranceand control (QA/QC) procedures (Segura et al., 2004). To assure thereproducibility of testing procedure, following precautions weretaken into consideration: (1) dilution was performed using highquality deionized water (0.055 lS, 18 mX, T = 22 ± 1 �C and pH7.1); (2) new anion and cation-exchange membranes were usedfor each experiment; (3) after each test, platinized titanium rodelectrodes were immersed into 0.5 M HCl solution for 24 h, rinsedwith deionized water to avoid contamination; (4) after each treat-ment, glass container was immersed into 0.05 M HCl solution for24 h, rinsed with tap water and then with deionized water; (5)all chemical analyses were performed in triplicates; (6) the TOCcalibration was checked after 15 samples and (7) standard devia-tion was calculated for each test.

3. Results and discussion

3.1. Electro-Fenton induced TOC removal and denitrification ofaquaculture

3.1.1. Effect of pH on TOC removalLaboratory scale experiments were carried out at acidic (the

anodic compartment) and alkaline (the cathodic compartment)pH to evaluate the effect of pH on TOC removal. The pH of the solu-tion is an important factor for effective oxidation by the Fenton’sreagent (Catalkaya and Kargi, 2007):

2Fe2þ þH2O2 þ 2Hþ ! 2Fe3þ þ OH� þ 2H2O ð1Þ

Thus, for the efficient Fenton process, the presence of H+ (acidicenvironment) is required for the decomposition of H2O2 (Smithet al., 2005) and the production of maximum amount of hydroxylradicals (Gogate and Pandit, 2004). Fig. 2 shows the variation inthe TOC removal from aquaculture wastewater at the anode andcathode with changes in the pH of the solution. It is interestingto note that there was no TOC removal observed, in fact, TOC evenaccumulated at the anode when pH was in the range of 6.8–7.7.Such TOC behavior was also indicated by Kobya and Delipinar(2008) who researched TOC and COD removal from wastewaterusing electrocoagulation method. TOC removal at pH 2.4 was thehighest reaching 97.8%, followed by 77.1% at pH 2.2 and 72.8% atpH 2.5. However, when pH at the anode was 5–6, the TOC removalwas only in the range of 35–48% (Fig. 2). In addition, the highestTOC accumulation at the cathode was at pH 12.5 reaching as highas 153.2%. The lowest accumulation at the cathode was achieved atpH 6.2 (33%) at the beginning of the treatment (Fig. 2).

7 6.2 11.1 11.5 12.2 12.3 12.4 12.4 12.5

pH

ing the electrodialytic treatment (30 mA).

2192 J. Virkutyte, V. Jegatheesan / Bioresource Technology 100 (2009) 2189–2197

Fig. 3 presents the relationship between the pH and TOC con-centration at the anode and cathode in the selected experiment.The results were similar in all the experiments and therefore thisfigure represents all the data obtained in the study. Thus, the high-est TOC degradation was observed at pH between 2.2 and 3.5. Inanother study, optimal pH range for Fenton processes was alsofound between 2.0 and 4.5 for the treatment of landfill leachatewith combined UASB and Fenton coagulation (Wang et al., 2000).When pH fell below 2 (data not shown), the TOC removal wasdiminished. This may be attributed to that less OH� radicals areproduced because of the formation of [Fe(H2O)]2+ that is less reac-tive with H2O2 (Gallard et al., 1998). Furthermore, at low pH, thereaction between OH� and H+ becomes predominant and the regen-eration of Fe2+ by reaction of Fe3+ with H2O2 is inhibited (Deng andEnglehardt, 2006).

During the electrochemical treatment of aquaculture wastewa-ter, cathodic processes result in an increase of pH with time (Abuz-aid et al., 1999). When pH medium becomes slightly alkaline oralkaline, TOC removal becomes hindered (Fig. 3). Deng and Engle-hardt (2006) proposed five mechanisms that are responsible forthe decreased treatment efficiency and which are also relevant forthe current study: (1) the absence of H+ when pH is extremely alka-line (at the cathode in this study), may inhibit the decomposition ofH2O2 that reduces the production of OH� radicals and therefore nosignificant degradation of contaminants is observed; (2) when pHis above 5, H2O2 rapidly decomposes to water and oxygen and thusno radicals are formed that are responsible for the destruction ofcontaminants; (3) at high pH, the ferrous ion precipitates asFe(OH3), hindering the reaction between Fe3+ and H2O2, and there-fore the regeneration of Fe2+ ions that are necessary for the efficientFenton process is diminished; (4) at high pH, not only pH itself butalso aqueous carbonate species (CO2�

3 and HCO�3 ) that scavenge OH�

radicals impair the remediation process, and (5) oxidation potentialsignificantly decreases with increasing pH from E0 = 2.8 V at pH 0 toE14 = 1.95 V at pH 14. In addition, Bautista et al. (2007) observedthat formed Fe(OH)3 may also catalyze the decomposition of H2O2

to O2 and H2O, thus decreasing the production of OH� radicals andyet again impairing the remediation process.

3.1.2. Effect of H2O2 dosage on TOC removalThe amount of H2O2 highly influences the operating costs and

the efficiency of the process. It is reported by Konstantinou andAlbanis (2004) that removal of organics increases with the increasein H2O2 concentration. However, the excess of peroxide may resultin iron sludge formation due to O2 off-gassing caused by the auto-decomposition of H2O2 (Kim et al., 2001). Moreover, the excess ofH2O2 remaining in the solution after the treatment entails a toxic-ity which may be above the corresponding limit. Evidently, it is dif-ficult to determine the optimal Fenton’s reagent dosage as it highly

0

5

10

15

20

25

2.5 2.4 3.1 3.5 4.2 5.2 5.5 6.9 7.8

TO

C c

on

cen

trat

ion

(m

g/L

)

p

Fig. 3. The relationship between pH and TOC conc

depends on the wastewater or effluent quality. A typical range forFenton reactants is about one part of iron per 5–25 parts (%wt/wt)of hydrogen peroxide (Zazo et al., 2005).

H2O2 concentration in the experiments varied between 5 and80 mM, while Fe(II) concentration was kept according to 15:1 ratio.The TOC removal increased with an increase in added ferrous ionsconcentration. However, the extent sometimes became marginalabove 40 mM or was similar to experiments where 5 mM solutionswere used (Fig. 4). Moreover, excess concentrations of ferrous ionsmay lead to a formation of elevated amounts of iron salts that willsignificantly contribute to an increase in the total dissolved solids(TDS) content of the effluent stream (Gogate and Pandit, 2004).

In all the experiments there was some TOC fluctuation in both,the anodic and cathodic compartments reaching as high as43.3 mg L�1 of TOC after 40 h when just the electric current with-out Fenton‘s reagent, and 25.1 mg L�1 after 16 h when no Fenton‘sreagent or electricity was applied (Fig. 4). When aquaculturewastewater was subjected to electricity (without the addition ofFenton’s reagent), the removal of TOC at the anode was insignifi-cant reaching only 19% at the end of 48 h experiment (Fig. 4a).However, the application of 30 mA direct current for 48 h de-creased the TOC from the initial 9.2 to 4.5 mg L�1, which corre-sponds to 49% removal. When different H2O2 concentrationswere added to investigate the TOC removal at the anode, the high-est TOC decrease (0.3 mg L�1 from the initial 11.2 mg L�1) was ob-tained at 40 mM, corresponding to 97.3% removal, followed by20 mM (1.9 mg L�1 from the initial 9.2 mg L�1), which yielded79.3% and 5 mM (2.1 mg L�1 from the initial 9.2 mg L�1) with77% removal. Moreover, when the highest H2O2 concentration(80 mM) was applied, 3.3 mg L�1 of TOC from the initial 9.2 mg L�1

was found after the 48 h of treatment, corresponding to 64.1% re-moval. The reason for the reduced TOC removal when elevatedconcentrations of H2O2 were used may be that an excess of hydro-gen peroxide may also act as OH� scavenger (auto-scavenging reac-tions), hence an increase in its concentration leads to a decrease inthe concentration of free hydroxyl radicals, which causes the de-crease in the TOC removal (Zelmanov and Semiat, 2008).

Interestingly, when 20 and 5 mM of H2O2 was added to thereactor, there was insignificant reduction in TOC after first 16 h(4% and 11%, respectively). However, after 24 h of the experiment,TOC increased to 15.2 and 15.1 mg L�1, corresponding to 65.2% and64.1% accumulation, respectively (Fig. 4a). As depicted from Fig. 4b,there was no removal just accumulation of TOC at the cathode inall the conducted experiments. The lowest TOC accumulation(63%) was found at 20 mM of H2O2, followed by 5 mM H2O2 con-centration (76%) and 82% when no electricity or H2O2 was applied.Evidently, the addition of 40 and 80 mM of H2O2 did not reduceTOC concentration at the cathode, corresponding to 151% and120% accumulation, respectively.

6.1 7.9 11.5 12.2 11.8 12.2 12.1 12.4 12.1

H

entration at the anode and cathode at 30 mA.

Fig. 4. Distribution of TOC in electrodialytic reactor during 48 h of experiment with different Fenton’s reagent concentrations: (a) anodic and (b) cathodic compartments.(AQ – Aquaculture wastewater and AQ, EK – Aquaculture wastewater treated with electricity).

J. Virkutyte, V. Jegatheesan / Bioresource Technology 100 (2009) 2189–2197 2193

3.1.3. Nitrate transport in electrodialytic reactor at different currentdensities

Fig. 5 shows nitrate distribution in the electrodialytic reactor. Itis clear that after 8 h of experiments regardless the experimentalstrategy, considerable amounts of nitrate was found by the cath-ode. It may be explained that when the direct current was applied,strong hydraulic gradient (electroosmotic flow), which is alwaysdirected towards the cathode prevailed and transported nitrates,which were still complexed with the organic matter and were un-charged, along the electrodialytic reactor (Manokararajah and Ran-

jan, 2005). However, after 16 h the amount of nitrates found at theanode increased. This may be due to that nitrates became chargedand electromigration was already the predominant phenomenonin the system. Therefore, negatively charged NO�3 were transportedtowards the positively charged anode. Observations made by Jiaand co-authors (2005) indicate similar behavior of nitrates inover-saturated soils or slurries.

According to Fig. 5, the highest nitrate removal was recorded atcurrent density of 0.12 mA cm�2 (60 mA) and a complete removalof nitrate was attained after 48 h at the cathode and 94.8% at the

Fig. 5. Distribution of nitrates in electrodialytic reactor during 48 h of experiment with different electric currents: (a) anodic and (b) cathodic compartments(AQ – Aquaculture wastewater and AQ, EK – Aquaculture wastewater treated with electricity).

2194 J. Virkutyte, V. Jegatheesan / Bioresource Technology 100 (2009) 2189–2197

anode. It is in agreement with Abuzaid and co-authors (1999) whostudied electrochemical nitrite and nitrate removal from aquacul-ture wastewater. Moreover, nitrate migration towards the anodewas higher with an increase in current, which may be attributedto that higher current densities yield higher nitrate movement to-wards the anode (Eid et al., 2000).

There was a significant concentration of nitrates detected inthe anion-exchange membrane after each treatment (45.2–60.2 mg L�1). This may be due to the possible oxidation ofaquaculture ammonium directly in the semi-permeable mem-

brane. Indeed, Smith et al. (2005) observed that ammonium, usu-ally passes through the system unaltered and may besubsequently oxidized to nitrate in the filter or a membrane. Itwas found that the change in nitrate concentration directly re-lated to changes in TOC, e.g. the decrease in nitrate concentrationwas directly proportional to an increase in TOC and vice versa.There was a gradual decrease in nitrate concentration from 120to 0.3 mg L�1 during 48 h of the experiments. At the same time,the gradual increase in TOC from 1.3 to 23.3 mg L�1 was alsodetected.

J. Virkutyte, V. Jegatheesan / Bioresource Technology 100 (2009) 2189–2197 2195

3.2. Hydrogenotrophic and Fe2+ mediated denitrification ofaquaculture

3.2.1. Hydrogenotrophic denitrificationWhen the low level direct electric current is applied, main

cathodic reaction of nitrate reduction to dinitrogen may be ex-pressed according to Szpyrkowicz et al. (2006)

5H2 þ 2NO�3 ! 4H2Oþ N2 þ 2OH� ð2Þ

The theoretical consumption of hydrogen is 0.35 mg H2 per1 mg NO�3 –N reduced to dinitrogen (Mateju et al., 1992). When10 mA current was applied, the amount of formed hydrogen wasthe lowest, followed by a threefold increase when the applied cur-

Fig. 6. Distribution of nitrates in electrodialytic reactor during 48 h of treatment with Fewastewater and AQ, EK – Aquaculture wastewater treated with electricity).

rent was 30 mA, almost sixfold increase at 60 mA current and morethat eightfold increase at 80 mA. Moreover, the amount of hydro-gen significantly increased with time, suggesting that the durationof the process as well as the applied electric current may be thelimiting factors for the hydrogen production. Also, the decreasein nitrate concentration, i.e. denitrification was the most pro-nounced at pH 2.6–5 and 7.4–8.8 and reached 92.5% and 89.6%,respectively. Indeed, when researching nitrates removal fromaquaria by electrochemically generated hydrogen, Grommen andco-authors (2006) determined that denitrification is the most effi-cient at the optimum pH range of 7.7–8.6 and may reach above90%. Interesting to note that there was a significant decrease in ni-trate concentration at the anode at even pH 2.4–4.8. This may be

2+ as electron donor: (a) anodic and (b) cathodic compartments (AQ – Aquaculture

2196 J. Virkutyte, V. Jegatheesan / Bioresource Technology 100 (2009) 2189–2197

attributed to that charged NO�3 ions were transported by electro-migration towards the anode, where H+ from the anodic reactionsreduced NO�3 either to N2 at low current densities or to NHþ4 at highcurrent densities (Peel et al., 2003).

There was no significant denitrification observed when pH was9 and above. Therefore, there is some controversy regarding theinfluence of alkaline pH on hydrogen-induced denitrification.According to Glass and Silverstein (1998), when pH exceeds 8,there is an increased reduction rate of nitrate to nitrite, which isnot reduced any further to dinitrogen or ammonium. However,Thomsen et al. (1994) argued that for example Paracoccus denitrif-icans (these species can denitrify using hydrogen even when an or-ganic carbon source is not present), can efficiently denitrify even atpH 9.5.

The TOC rose from the initial 9.2 mg L�1 to 10.3 mg L�1 after 8 hand to 43.3 mg L�1 after 40 h, which may be attributed to the in-creased H2 concentrations at the cathode due to the electrolysisof water. In addition, Haugen et al. (2002) determined that apartfrom the elevated H2 and CO2 concentrations in the system, TOC in-crease may also be attributed to the increased number of auto-trophs in the reactor. However, at the end of the 48 h ofexperiment, TOC at the cathode was slightly reduced to17.8 mg L�1, which is in contrast to findings of Rittmann et al.(1994) who argued that over the period of time, TOC is not ex-pected to increase due to the production of soluble microbial prod-ucts and transported biomass and thus TOC fluctuation mustremain minimal.

In order to avoid TOC increase and thus to ensure low TOC at theend of the experiments, it is assumed that usage of some materialthat effectively adsorbs organic matter and biomass would signif-icantly improve the electrodialytic system. For example, Haugenand co-authors (2002) proposed rock media or a real aquifer mate-rial that supports biological growth as well as filtration capacityand thus controls the TOC concentration within the system.

3.2.2. Fe2+ mediated denitrificationDenitrification using Fe2+ ions as electron donors may be ex-

pressed as follows:

NO�3 þ 5Fe2þ ! 0:5N2 þ 5FeOOHþ 9Hþ ð3Þ

To investigate the applicability of Fe2+ as an electron donor toreduce nitrate concentration in aquaculture wastewater, differentconcentrations (10, 30, 100 and 200 mM) of Fe2+ (as FeSO4 � 7H2O)were used coupled with low level (30 mA) electric current (Fig. 6).When no electricity was applied and no Fe2+ ions were added tothe reactor, there was no significant change in the nitrate concen-tration at the anode, however, there was some reduction of nitrateat the cathodic side (Fig. 6). This may be attributed to that therewas organic matter available for the microorganisms that alsoacted as electron donor and thus reduction took place. Indeed, asobserved by Klas et al. (2006), the reduction of nitrates is possibleat some extent using the electron donor supplied from within thesystem itself – intrinsic energy source – available organic matter inthe aquaculture system. Also, Boley et al. (2000) determined thatwhen high concentration of dissolved organic matter coupled withelectron donor is used, significant denitrification rates areachieved.

When Fe2+ concentration was 10 mM, there was a significantdecrease in the nitrate concentration at the cathode after 16 and24 h (100 and 80 mg L�1, respectively). However, nitrate concen-tration at the anode did not change significantly, decreasing onlyto 123.5 mg L�1 at the end of the experiment corresponding to37% removal. When the concentration of Fe2+ was increased, thehighest reduction (66.6%) in nitrate concentration was achievedwith 100 mM Fe2+ solution (65.2 mg L�1) at the cathode, followed

by 30 mM (78.6 mg L�1, 59.8% removal) (Fig. 6b). On the contrary,the nitrate concentration at the anode did not follow the samepathway as at the cathode, i.e. the concentration decreased twofold(49% removal) when 100 mM was used and 1.8-fold (43% removal)when 200 mM of Fe2+ was used (Fig. 6b). The reduction of nitratesat the cathodic side might be attributed to that charged Fe2+ ionswere transported via electromigration towards the cathode wherethe reduction of nitrates took place.

4. Conclusion

The technical feasibility to remove TOC and nitrates from aqua-culture wastewater using a combined electrical and chemicaltreatment methods such as Electro-Fenton, hydrogenotrophic andFe2+ mediated denitrification was investigated on a laboratoryscale. It was found that the most efficient nitrate removal followedthe order Electro-Fenton > hydrogenotrophic > Fe2+ mediateddenitrification.

Nitrate removal using Fe2+ was a factor 1.4 less than usinghydrogen (92.5%) or a Fenton’s reagent (94.8%).

The highest TOC removal was achieved at pH 2.2 and 2.4 (77.1%and 97.8%, respectively) at the anode. The lowest accumulation of33% at pH 6.2 at the cathode was obtained.

The highest TOC removal in terms of using various H2O2 con-centrations was achieved at 40 mM reaching as high as 97.3%.

Regardless the experimental strategy, initially nitrates migratedtowards the cathode under the applied electric current. During thecourse of experiments, nitrates were transported towards the an-ode where their concentration was reduced. The highest nitrate re-moval was achieved at 0.12 mA cm�2 electric current density(94.8%) at the anode and a complete removal at the cathode.

The highest efficiency of hydrogenotrophic reduction of nitrateswas 92.5%. Fe2+ ions mediated process achieved 66.6% nitrateremoval.

Acknowledgements

The Academy of Finland is acknowledged for the research fund-ing 2005–2008 (Decision No. 212649); partial fund provided by‘‘JCU Uninet Pty Ltd.” for the first author to conduct research atthe School of Engineering of James Cook University from April toJuly 2006 is also acknowledged.

References

Abuzaid, N.S., Al-Hamouz, Z., Bukhari, A.A., Essa, M.E., 1999. Electrochemicaltreatment of nitrite using stainless steel electrodes. Water Air Soil Poll. 109,429–442.

APHA, 1995. Standard Methods for the Examination of Water and Wastewater, 18thed. American Public Health Association, Washington, DC.

Banasiak, L.J., Kruttschnitt, T.W., Schäfer, A.I., 2007. Desalination usingelectrodialysis as a function of voltage and salt concentration. Desalination205, 38–46.

Bautista, P., Mohedano, A.F., Gilarranz, M.A., Casas, J.A., Rodriguez, J.J., 2007.Application of Fenton oxidation to cosmetic wastewaters treatment. J. Haz.Mater. 143, 128–134.

Boley, A., Müller, W.R., Haider, G., 2000. Biodegradable polymers as solid substrateand biofilm carrier for denitrification in recirculated aquaculture systems.Aquacult. Eng. 22, 75–85.

Catalkaya, E.C., Kargi, F., 2007. Color, TOC and AOX removals from pulp mill effluentby advanced oxidation processes: a comparative study. J. Haz. Mater. 139, 244–253.

Deng, Y., Englehardt, J.D., 2006. Treatment of landfill leachate by the Fenton process.Water Res. 40, 3683–3694.

Eid, N., El Shorbagy, W., Larson, D.L., Slack, D.C., 2000. Prediction of nitrateelectromigration in sandy soil. J. Haz. Mater. 79, 133–149.

Gallard, H., De Laat, J., Legube, B., 1998. Effect of pH on the oxidation rate of organiccompounds by Fe-II/H2O2, mechanisms and simulation. New J. Chem. 3, 263–268.

Glass, C., Silverstein, J., 1998. Denitrification kinetics of high nitrate concentrationwater: pH effect on inhibition and nitrite accumulation. Water Res. 32, 831–839.

J. Virkutyte, V. Jegatheesan / Bioresource Technology 100 (2009) 2189–2197 2197

Gogate, P.R., Pandit, A.B., 2004. A review of imperative technologies for wastewatertreatment I: oxidation technologies at ambient conditions. Adv. Environ. Res. 8,501–551.

Grommen, R., Verhaege, M., Verstraete, W., 2006. Removal of nitrate in aquaria bymeans of electrochemically generated hydrogen gas as electron donor forbiological denitrification. Aquacult. Eng. 34, 33–39.

Haugen, K.S., Semmens, M.J., Novak, P.J., 2002. A novel in situ technology forthe treatment of nitrate contaminated groundwater. Water Res. 36, 3497–3506.

Jegatheesan, V., Zeng, C., Shu, L., Manicom, C., Steicke, C., 2007. Technologicaladvances in aquaculture farms for minimal effluent discharge to oceans. J.Clean. Prod. 15, 1535–1544.

Jia, X., Larson, D., Slack, D., Walworth, J., 2005. Electrokinetic control of nitratemovement in soil. Eng. Geol. 77, 273–283.

Jorguera, M.A., Valencia, G., Eguchi, M., Katayose, M., Riquelme, C., 2002.Disinfection of seawater for hatchery aquaculture systems using electrolyticwater treatment. Aquaculture 207, 213–224.

Kapoor, A., Viraraghavan, T., 1997. Nitrate removal from drinking water – review. J.Environ. Eng. 123, 371–380.

Kesore, K., Janowski, F., Shaposhnik, V.A., 1997. Highly effective electrodialysis forselective elimination of nitrates from drinking water. J. Membrane Sci. 127, 17–24.

Kim, J.S., Kim, H.Y., Won, C.H., Kim, J.G., 2001. Treatment of leachate produced instabilized landfills by coagulation and Fenton oxidation process. J. Chin. Inst.Chem. Eng. 32, 425–429.

Klas, S., Mozes, N., Lahav, O., 2006. Development of a single-sludge denitrificationmethod for nitrate removal from RAS effluents: lab-scale results vs modelprediction. Aquaculture 259, 342–353.

Kobya, M., Delipinar, S., 2008. Treatment of the baker’s yeast wastewater byelectrocoagulation. J. Hazard. Mat. 154, 1133–1140.

Konstantinou, I.K., Albanis, T.A., 2004. TiO2-assisted photocatalytic degradation ofazo dyes in aqueous solution: kinetic and mechanistic investigations. A review.Appl. Cat. B: Environ. 49, 1–14.

Lee, P.G., Lea, R.N., Dohmann, E., Prebilsky, W., Turk, P.E., Ying, H., Whitson, J.L.,2000. Denitrification in aquaculture systems: an example for a fuzzy logiccontrol problem. Aquacult. Eng. 23, 37–59.

Liessens, J., Vanbrabant, J., DeVos, P., Kersters, K., Verstraete, W., 1992. Mixedculture hydrogenotrophic nitrate reduction in drinking water. Microb. Ecol. 24,271–290.

Manokararajah, K., Ranjan, R.S., 2005. Electrokinetic retention, migration andremediation of nitrates in silty loam soil under hydraulic gradients. Eng. Geol.77, 263–272.

Mateju, V., Cizinská, S., Krejcí, J., Janoch, T., 1992. Biological water denitrification – areview. Enzyme Microb. Technol. 14, 170–183.

Neyens, E., Baeyens, J., 2003. A review of classic Fenton’s peroxidation as anadvanced oxidation technique. J. Haz. Mater. B98, 33–50.

Oldani, M., Killer, E., Miguel, A., Schock, G., 1992. On the nitrate and monovalentcation selectivity of ion exchange membranes used in drinking waterpurification. J. Membrane Sci. 75, 265–275.

Peel, J.W., Reddy, K.J., Sullivan, B.P., Bowen, J.M., 2003. Electrocatalytic reduction ofnitrate in water. Water Res. 37, 2512–2519.

Qin, G., Liu, C.C.K., Richman, N.H., Moncur, J.E.T., 2005. Aquaculture wastewatertreatment and reuse by wind-driven reverse osmosis membrane technology: apilot study on coconut island. Hawaii. Aquacult. Eng. 32, 365–378.

Rittmann, B.E., Regan, J.M., Stahl, D.A., 1994. Nitrification as a source of solubleorganic substrate in biological treatment. Water Sci. Technol. 30, 1–8.

Rulkens, W.H., Tichy, R., Grotenhuis, J.T.C., 1998. Remediation of polluted soil andsediment: perspectives and failures. Water Sci. Technol. 37, 27–35.

Rutten, P., Schnoor, G., 1992. Five year’s experience of nitrate removal from drinkingwater. Water Supply 10, 183–190.

Segura, M., Cámara, C., Madrid, Y., Rebollo, C., Azcárate, J., Kramer, G.N., Gawlik,B.M., Lamberty, A., Quevauville, Ph., 2004. Certified reference materials (CRMs)for quality control of trace-element determinations in wastewater. TrAC Trendsin Anal. Chem. 23, 194–202.

Singh, S., Ebeliong, J., Wheaton, F., 1999. Water quality trials in four recirculatingaquacultural system configurations. Aquacult. Eng. 20, 75–84.

Smith, R.L., Miller, D., Brooks, M., Widowson, M., Killingstad, M.W., 2001. In situstimulation of groundwater denitrification with formate to remediate nitratecontamination. Environ. Sci. Technol. 35, 196–203.

Smith, R.L., Buckwalter, S.P., Repert, D.A., Miller, D.N., 2005. Small-scale hydrogen-oxidizing-denitrifying bioreactor for treatment of nitrate-contaminateddrinking water. Water Res. 39, 2014–2023.

Szpyrkowicz, L., Daniele, L., Radaelli, M., Specchia, S., 2006. Removal of NO�3 fromwater by electrochemical reduction in different reactor configuration. Appl.Catal. B: Environ. 66, 40–50.

Thomsen, J.K., Geest, T., Cox, R.P., 1994. Mass spectrometric studies of the effect ofpH on the accumulation of intermediates in denitrification by Paracoccusdenitrificans. Appl. Environ. Microbiol. 60, 536–541.

Thurston, R.V., Russo, R.C., Vinogradov, G.A., 1981. Ammonia toxicity to fishes:effect of pH on the toxicity of the un-ionized ammonia species. Environ. Sci.Technol. 15, 837–840.

Virkutyte, J., Sillanpää, M., Latostenmaa, P., Martisius, J., 2004. Electrokinetic copperremoval from lake sand: the process design and kinetics. Int. J. Surf. Min.Reclam. Environ. 18, 220–231.

Wang, P.I., Lau, W.C., Fang, H.H.P., Zhou, D., 2000. Landfill leachate treatment withcombined UASB and Fenton coagulation. J. Environ. Sci. Health A: Toxic/Hazardous Substances Environ. Eng. 35, 1981–1988.

Zazo, J.A., Casas, J.A., Mohedano, A.F., Gilarranz, M.A., Rogriguez, J.J., 2005. Chemicalpathway and kinetics of phenol oxidation by Fenton’s reagent. Environ. Sci.Technol. 39, 9295–9302.

Zelmanov, G., Semiat, R., 2008. Iron(3) oxide-based nanoparticles as catalysts inadvanced organic aqueous oxidation. Water Res. 42, 492–498.