Available online at www.sciencedirect.

com

Bioresource Technology 99 (2008) 39653974

Review

Bio-electrochemical removal of nitrate from water and wastewaterA review

Department of Chemical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia

Abstract Nitrates in dierent water and wastewater streams raised concerns due to severe impacts on human and animal health. Diverse methods are reported to remove nitrate from water streams which almost fail to entirely treat nitrate, except biological denitrication which is capable of reducing inorganic nitrate compounds to harmless nitrogen gas. Review of numerous studies in biological denitrication of nitrate containing water resources, aquaculture wastewaters and industrial wastewater conrmed the potential of this method and its exibility towards the remediation of dierent concentrations of nitrate. The denitriers could be fed with organic and inorganic substrates which have dierent performances and subsequent advantages or disadvantages. Review of heterotrophic and autotrophic denitrications with dierent food and energy sources concluded that autotrophic denitriers are more eective in denitrication. Autotrophs utilize carbon dioxide and hydrogen as the source of carbon substrate and electron donors, respectively. The application of this method in bio-electro reactors (BERs) has many advantages and is promising. However, this method is not so well established and documented. BERs provide proper environment for simultaneous hydrogen production on cathodes and appropriate consumption by immobilized autotrophs on these cathodes. This survey covers various designs and aspects of BERs and their performances. 2007 Elsevier Ltd. All rights reserved.
Keywords: Water; Wastewater; Bio-electrochemical; Biological denitrication; BER

1. Introduction Water is of fundamental importance for life since the mechanism of metabolism and synthesis are in close relation with the specic characteristics of water. Transport of nutrients inside cells and interactions with the environment without water are impossible. On the other hand, water resources are limited and only 2.66% of the total global water resources comprising groundwater, lakes and rivers, polar ice and glaciers are fresh water. Furthermore, only a small fraction of the fresh water, about 0.6%, is usable as drinking water. For this reason, water resources must be necessarily treated properly and wastewater treatment must be done eciently (Shrimali and Singh, 2001). The fast movement of humans towards urbanization,
*

Corresponding author. Tel.: +603 79675313; fax: +603 79675319. E-mail address: [email protected] (M.K. Aroua).

industrialization and agricultural activities introduced various contaminants into the environment. Nitrogen-containing compounds are examples of those contaminants which can create serious problems when released into the environment, such as eutrophication of rivers (Sumino et al., 2006), deterioration of water quality and potential hazard to human or animal health. Nitrate is one of these compounds, though it does not pose a human or animal health threat by itself. It is potentially converted to nitrite in the gastrointestinal tract or to nitrous nitrogen compounds through reduction (Foglar et al., 2005). Up to the present time, researches are carried out towards nitrate removal from water resources, aquaculture ponds or aquaria, and industrial wastewaters. In this review, rstly, nitrate as a water contaminant and problems arising from its presence are studied. Secondly, various studies and methods applied for nitrate removal are discussed. Finally, aiming to reveal an eective method

0960-8524/$ - see front matter 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2007.05.026

3966

S. Ghafari et al. / Bioresource Technology 99 (2008) 39653974

applicable for dierent types of nitrate-containing water and wastewater, emphasis is on bio-electrochemical denitrication. 2. Nitrate-containing water and wastewaters 2.1. Nitrate in water resources Nitrate is identied as one of the hazardous contaminants in potable water that may reduce to nitrosamines in the stomach which is suspected to cause gastric cancer (Glass and Silverstein, 1999; Galvez et al., 2003). In addition, nitrate exposes infants and pregnant women to danger due to the potential reduction of nitrate to nitrite inside the stomach of foetus after digesting nitrate by the mother. The nitrite reacts with the hemoglobin in blood and converts the hemoglobin into methaemoglobin, which does not carry oxygen to cell tissues. This phenomenon results in a bluish color of infants skin so called methaemoglobinemia or the blue baby syndrome (Shrimali and Singh, 2001). Haemoglobin Fe2
NO 2

Methaemoglobin Cannot combine with O2

Can combine with O2

Water resources (ground and surface water) are contaminated by nitrate in several ways. Most of the researchers reported that contamination of groundwater by nitrate is primarily attributed to non-point agricultural sources (Feleke and Sakakibara, 2002). Nitrate originated from agriculture is increasingly growing all over the world due to the extreme use of fertilizers. Nitrate salts reach the groundwater as they percolate through the soil. Some other sources of nitrate in ground and surface water are from uncontrolled land discharges of treated or raw wastewater from domestic and industrial wastes, landlls (Islam and Suidan, 1998) and animal wastes particularly from animal farms (Terada et al., 2003). Nitrate in groundwater is increasingly an important problem, which prohibited the direct use of the groundwater resources for human consumption in some parts of the world including India, Japan, China, Saudi Arabia, USA, UK and several parts of Europe (Gayle et al., 1989). With the aim to protect the consumers from adverse eects of high nitrate intake, the United States, Canada, and the World Health Organization (WHO) have set standards to regulate the nitrate concentration in drinking water to 50 mg NO /l (Sayre, 1988). 3 Also, the United States Environmental Protection Agency (EPA) and WHO have set the maximum contaminant level (MCL) of 10 mg NO N=l in drinking water (Cast and 3 Flora, 1998). The same concern in Europe put MCL of 12 mg NO N=l in drinking water (Glass and Silverstein, 3 1999). 2.2. Nitrate in mariculture and aquaculture systems Cultivation of river and marine organisms in their natural or synthetic habitat is always along with the production

of ammonia (referring to NH3 or NH which is consid4 ered as a serious problem. Ammonia produces a bad avor and taste as well as harmful eects on aquatic animals. EPA proposed that ammonia concentrations in sh culture tanks be maintained at levels lower than 0.02 mg/l as the criterion for non-ionic ammonia (EPA, 1976). Therefore, indispensable need for ammonia removal complied with applying dierent pathways such as nitrication, sludge removal by sedimentation or mechanical ltration, and water exchange (Rijn et al., 1996). Nitrication is reported as one of the most common and eective methods applied by researchers (Hagopian and Riley, 1998; Singh et al., 1999; Zhu and Chen, 1999) in both freshwater and seawater. In this method ammonia is oxidized into nitrite and nally nitrate as the end-product. Reported maximum nitrate levels dier among various systems and values of 400500 mg NO N=l are reported (Otte and Rosenthal, 3 1979). Contrary to ammonia and nitrite, nitrate is relatively non-toxic to aquatic organisms. However, it should not be left to accumulate, because eventually leads to some undesirable results such as phytoplankton blooms, inhibition of nitrication and toxicity problems at certain concentrations of nitrate (Park et al., 2001; Rijn et al., 2006). As a result, daily replacement of a fraction of the system volume (510%) with new water may prevent accumulation of nitrate (Masser et al., 1999), though this solution does not sound environmentally. Increased eorts are now directed towards nitrate control in aquaculture systems, not only concerning toxic eect on sh, but also for complying to environmental regulations associated with permissible nitrate level in euent discharge which is as low as 11.3 mg NO N=l (European Council Directive, 1998). 3 2.3. Nitrate in industrial wastewaters Concentration of nitrogen compounds in some industrial wastewaters is tremendously higher than what is in groundwater and surface water. Ammonia and nitrate are the most problematic nitrogen compounds in this sort of wastewaters. Ammonia in industrial wastewater is normally eliminated by nitrication which is achieved by the complete oxidation of ammonia to nitrate (Sumino et al., 2006). Thus, nitrate removal from these types of industrial wastewater is an inevitable step in treatment. Dierent industrial wastewaters are reported to contain more than 200 mg NO N=l (Almeida et al., 1995; Zayed and Winter, 3 1998; Peyton et al., 2001) and some contain higher nitrate concentration. For instance, euent from some industries producing explosives, fertilizer (Watanabe et al., 2001), pectin, cellophane, and metals nishing industries, contain greater than 1000 mg NO N=l (Glass and Silverstein, 3 1999). Furthermore, nuclear industry produces nitrateloaded wastes in extremely higher concentration at many points during the nuclear fuel cycle. The processing of radioactive metal products at nuclear weapons production plants and research labs reported production of wastewater containing nitrate in excess of 50,000 mg NO N=l primar3

S. Ghafari et al. / Bioresource Technology 99 (2008) 39653974

3967

ily derived from nitric acid used for metal cleaning (Francis and Hatcher, 1980). 3. Denitrication methods 3.1. Abiotic nitrogen removal methods The conventional processes of water treatment method including coagulation, ltration, and disinfection which are applied for water potability are not ecient for the elimination of nitrate ion from the water. Therefore, to remove nitrate ion, a supplementary method is necessary. Nitrate removal from water or wastewater can be achieved using two main groups of treatment processes; physicochemical and biological treatment methods. The most common conventional treatment processes used to remove nitrate are reverse osmosis (RO), ion exchange (IE), electro-dialysis (ED) and activated carbon adsorption in conjunction with pH adjustment (Islam and Suidan, 1998; Ergas and Reuss, 2001; Shrimali and Singh, 2001; Feleke and Sakakibara, 2002; Prosnansky et al., 2002). Also, a wide interest towards some new methods such as metallic iron-aided abiotic nitrate reduction is recently reported by several researchers (Huang et al., 1998; Choe et al., 2000, 2004; Devlin et al., 2000; Hu et al., 2001; Huang and Zhang, 2002, 2004; Huang et al., 2003; Chi et al., 2004; Ginner et al., 2004). There are some disadvantages with these conventional methods that limit their utility, due to their expensive operation and subsequent disposal problem of the generated nitrate waste brine or by-products. For instance, IE process removes nitrate and sulfate simultaneously from the water being treated. Subsequent produced wastewater from the resin regeneration process rich in these two anions requires a complementary method before its discharge (Shrimali and Singh, 2001). RO, although is able to separate and concentrate the compounds contained in water without any change in their molecular structures, this method is limited due to its high cost and producing concentrated waste brine that may pose a disposal problem for further treatment (Ergas and Reuss, 2001; Shrimali and Singh, 2001). Even for the new method of metallic iron-aided abiotic nitrate reduction, although complete or partially nitrate removal are reported by the researchers, formation of ammonia along with nitrogen gas as end-products is an undesired consequence of this method (Biswas and Bose, 2005). 3.2. Biological denitrication methods In contrast to abiotic methods, which failed to treat nitrate and were merely able to separate or remove it from the stream and resulted in some problematic by-products, biological methods provided a promising chance for real treatment of nitrate and new novel techniques are reported accordingly. For removing nitrate, although aerobic denitrication has been reported (Joo et al., 2005; Robertson and Kuenen, 1984; Zart and Eberhard, 1998; Chiu et al.,

2007), most of the reported denitrication investigations are done by facultative anaerobes in the absence of oxygen (Rijn et al., 2006). Biological denitrication is a mechanism by which denitrifying bacteria use nitrate as terminal electron acceptor in their respiratory process in the absence of oxygen. Denitrifying bacteria, in this method, reduce inorganic nitrogen compounds such as nitrite and nitrate into harmless elemental nitrogen gas (Prosnansky et al., 2002) so that no further treatment is imposed. Also, unlike some contaminants which are in need for a certain microbe to be treated, denitrifying bacteria are ubiquitous in nature (Gamble et al., 1977; Szekeres et al., 2001) and numerous researchers cultivated them using mix cultures taken from wastewater treatment plants as seeds. Moreover, microbial removal of nitrate may be the most economical strategy for the reclamation of nitrate polluted waters and wastewaters (Soares, 2000). As a result, there has been a huge interest towards microbial removal of nitrate as the most environmentally friendly and cost-eective method, although biological denitrication is usually slow and lasts long time particularly for industrial wastewaters containing high concentrations of nitrate (Foglar et al., 2005). Later studies tried to speed up biological denitrication by applying new mechanisms through which better connection provided between the nitrate and the microorganisms in the water stream. The Inorganic Contaminants Research Committee of USA (Islam and Suidan, 1998) reported that different denitrication strategies, such as the use of granular activated carbon (GAC), packed beds, rotating biological contactors and soil-aquifer denitrication systems have been undertaken by various researchers. Also, eorts are still ongoing and some novelties in combination of biological and other abiotic methods, such as membrane biolm reactors (MBR), were recently manifested (Terada et al., 2003). 4. Heterotrophic versus autotrophic denitrication Biological denitrication is normally conducted by facultative anaerobes which are in essential need for some food and energy sources which are organic or inorganic (Cast and Flora, 1998; Rijn et al., 2006). This fact classies denitriers into two main groups of heterotrophs and autotrophs. Heterotrophs are microorganisms that require organic substrates to get their carbon for growth and development. They also obtain energy from carbohydrates and other organic material. In contrast, autotrophs are microorganisms that use inorganic substances as energy source and carbon dioxide as a carbon source. Heterotrophic denitrication has been applied in recirculating systems and traditional wastewater treatment plants, whereas autotrophic denitrication is lately notied within the last two decades. These substrates do not always exist in the water or wastewater which is to be treated and should be supplemented. For instance, biological removal of nitrogen from wastewaters which contain high levels of nitrate can be problematic due to the low concentration

3968

S. Ghafari et al. / Bioresource Technology 99 (2008) 39653974

of carbon source required for the denitrifying bacteria (Kim et al., 2004). Hence, a number of researchers have found that denitrication activity in organic carbon-limited waters or wastewaters can successfully be enhanced by adding some carbon sources such as methanol, ethanol, acetic acid, glucose, or starch (Killingstad et al., 2002; Kim et al., 2002). The choice of substrate depends on a number of considerations such as cost, capacity and conguration of reactor, and post-treatment of the denitried water (Szekeres et al., 2002). In case of heterotrophic denitrication, both liquid and solid forms of organic carbon sources are conventionally used though the aqueous type is more common for treatment of water and wastewater. Among liquid carbon sources, the most common ones are methanol, ethanol (Gomez et al., 2003) and acetic acid which have been used for heterotrophic denitrifying processes in full-scale plants of drinking water treatment (Feleke and Sakakibara, 2002). However, there is some concern regarding these external organic carbon sources in the heterotrophic denitrication. Some organic carbon sources are toxic and their excess will remain in water. In addition, by-products built-up from organic carbon sources seems to be unavoidable which dictates some further treatment methods for their removal as well as disinfection in case of water treatment (Shrimali and Singh, 2001; Feleke and Sakakibara, 2002). Besides, denitrication rate is strongly susceptible to type of carbon source, concentration of carbon source (Galvez et al., 2003; Gomez et al., 2003), and C/N ratio (Rijn et al., 2006). This would vary for dierent microorganisms, water streams, and environmental conditions (Chiu and Chung, 2003). Hence, some researchers have investigated the inuence of dierent types and magnitude of carbon sources to achieve the optimum removal (Kim et al., 2002; Chiu and Chung, 2003; Joo et al., 2005). A poor C/N ratio leads to improper denitrication while a high C/N ratio may cause accumulation of nitrite or extra production of nitrous other than nitrogen gas (Kim et al., 2002; Chiu and Chung, 2003). In contrast, autotrophic denitriers utilize inorganic carbon substrates (carbon dioxide or bicarbonate) as the food source and rely on electron donors like hydrogen or reduced sulfur compounds for energy (Biswas and Bose, 2005). Autotrophic denitrication using carbon dioxide gas as an inorganic carbon source and some electron donors like hydrogen (Kurt et al., 1987; Feleke et al., 1998; Islam and Suidan, 1998; Ergas and Reuss, 2001; Sakakibara and Nakayama, 2001; Feleke and Sakakibara, 2002; Prosnansky et al., 2002; Park et al., 2005), thiosulfate (Claus and Kutzner, 1985a,b), sulde (Kleerebezem and ` Mendeza, 2002), and sulfur stone (Flere and Zhang, 1999) have been reported in literature. Some advantages of autotrophic over heterotrophic denitrication are; evasion of poisoning eect of some organic carbon, low biomass build-up and less sludge production which result in reduction of reactor clogging and easier post-treatment (Shrimali and Singh, 2001; Rijn et al., 2006).

5. Hydrogenotrophic denitrication Hydrogenotrophic denitriers refer to the denitrifying microorganisms utilize hydrogen as energy source which are ubiquitous in nature (Till et al., 1998). In addition to the general advantages of autrotrophic denitrication, using hydrogen as electron donor brings more advantages into hydrogenothropic denitrication method compared to other autotrophic methods using other electron donors. This is attributed to the properties of hydrogen gas such as being harmless, naturally clean, and low solubility (1.6 mg/ l at 20 C) in water. Thus, in contrast to other substrates and electron donors, no further steps are needed to remove the residues. Furthermore, higher eciency is obtained with hydrogen rather than with organic electron donors; for example, Gentzar (1995) reported an 8.5-fold greater eciency with hydrogen when compared to organic electron donors. Besides, hydrogen is not as expensive as some other organics. As a result, a number of laboratory and eld systems have been reported in which hydrogen was sparged into the water to be treated hydrogenotrophically (Gros and Treutler, 1986; Gros et al., 1986; Kurt et al., 1987; Dries et al., 1988). Nevertheless, there are some drawbacks with sparging hydrogen. For instance, low solubility of hydrogen in water may hamper the rate of autotrophic denitrication by dissolution or poor mass transfer rate of hydrogen from gas phase to liquid phase. To overcome this obstacle, bioreactors with large volumes to increase the contact time, may be required to achieve the complete denitrication (Prosnansky et al., 2002). In addition, the hazardous (explosive) nature of hydrogen during use, transportation and storage may limit the use of hydrogen gas in denitrication reactors. Furthermore, the same as traditional biological processes, control of pH by introducing an acid equivalent is still required (Feleke and Sakakibara, 2002; Prosnansky et al., 2002). These disadvantages would result in higher initial costs as well as higher operating costs of hydrogenotrophic denitrication in comparison with heterotrophic denitrication. An appropriate method has to be found to overcome the above mentioned problems. As such, some researchers tried to produce hydrogen in situ to remove the risk in transfer, storage and operation in using hydrogen. Additionally, bacteria can utilize hydrogen at the time of generation more eectively since there is no need to dissolve it prior to feeding into the reactor. Consequently, the internal production of hydrogen has been recently studied extensively through hydrogen production from either chemical reaction with water or electrolysis of water (Prosnansky et al., 2002). The production of hydrogen within the reactor using chemical reactions has some drawbacks. Metallic ironassisted hydrogenotrophic denitrication process is an example of denitrication applying chemically produced hydrogen from corrosion of iron in water according to Eqs. (1) and (2) (Biswas and Bose, 2005).

S. Ghafari et al. / Bioresource Technology 99 (2008) 39653974

3969

Fe 2H2 O ! H2 Fe2 2OH anaerobic iron corrosion 2NO 5H2 ! N2 4H2 O 2OH 3 biological denitrification

1 2

The relatively slow rate of anaerobic in situ corrosion process (Eq. (1)) with consequent low hydrogen production rate slows denitrication process (Eq. (2)). The rate also depends on the type of metallic iron as well as on the surface area (Biswas and Bose, 2005). Another drawback of in situ chemical production of hydrogen is the presence of probable by-products and their undesirable reactions with other materials in the treated water or wastewater. On the contrary, in situ hydrogen production through electrolysis of water revealed some advantages due to the benecial characteristics of electricity. The most signicant advantages of electricity are that it is dirt free without any residual by-product as well as easy to control which means easy handling of hydrogen production. Hence, autotrophic denitrication using internal production of hydrogen by means of electricity inspired a promising process which entailed generation of electro-assisted bioreactors or bioelectro reactors (BERs). 6. Bio-electrochemical denitrication in BERs Biological denitrication assisted by electricity would theoretically be performed through in situ hydrogen production on the cathode surface (not on the anode surface). Hydrogen and a lower oxidation reduction potential (ORP) environment produced through the cathodic reactions would be utilized by hydrogenotrophs to reduce nitrate nitrogen into nitrogen gas (Zhang et al., 2005). A proper contact between microorganisms and hydrogen improves biological denitrication. BERs involving

direct immobilization of denitriers on the surface of the cathode, where hydrogen is being produced, facilitates appropriate access of hydrogen to the microorganisms (Park et al., 2005). BERs provide immobilized microorganisms with a compulsory supply, rather than accidental contact, of electron donors which results in faster denitrication. This is in addition to the advantages of the hydrogenotrophic method and in situ hydrogen production by electrolysis of water. BERs may be fed by organic or inorganic carbon sources which serve heterotrophic and autotrophic hydrogenothrops, respectively. Table 1 lists the reactor type, type of microorganisms, treated inuent and the range of nitrate for bio-electrochemical method of denitrication found in the literature. In general, utilizing autotrophs in a BER ideally serves all the advantages attributed to autotrophic denitrication, hydrogenotrophic electron donation, and electrolytic in situ hydrogen production in a single method. Consequently, researchers focused on denitrication by autotrophic microorganisms in a BER as a prospective pathway for nitrate removal from any nitrate-containing water or wastewater. 6.1. Theory of denitrication in BERs The presumed pathway of complete reduction of nitrate NO to nitrogen gas (N2) which involves four 3 consecutive steps is illustrated below. The oxidation state of the nitrogen atoms in each substance is presented in parentheses (Feleke and Sakakibara, 2002; Killingstad et al., 2002): NO ! NO ! NO ! N2 O ! N2 3 2
5 3 2 1 0

Decomposition of nitrate in BERs can be illustrated through seven equations including electrolysis of water which are shown below.

Table 1 List of some previous studies applying bio-electrochemical method to denitrify water or wastewater containing various doses of nitrate Reactor Microorganism Treated inuent Nitrate concentration (mg NO N =l 3 13.820.3 1540 20 20 20 22.6492 24 50300 100 200 Ref.

Multi-cathode BER Multi-cathode BER combined with microltration (MF) BER BER Separate reactors for electrolysis and denitrication BER BER combined with adsorption BER BER BER
a b

Autotrophic Autotrophic Autotrophic Autotrophic Heterotrophic Autotrophic Autotrophic Heterotrophic Heterotrophic Heterotrophic

Synthetic contaminated water Synthetic contaminated groundwater Synthetic contaminated groundwater Synthetic contaminated water Aquaria water Synthetic Synthetic water Synthetic Synthetic Synthetic w/wa IPTb contaminated w/w contaminated water w/w

Sakakibara and Nakayama (2001) Prosnansky et al. (2002) Feleke et al. (1998) Islam and Suidan (1998) Grommen et al. (2006) Park et al. (2005) Feleke and Sakakibara (2002) Beschkov et al. (2004) Cast and Flora (1998) Watanabe et al. (2001)

w/w: wastewater. IPT: isoprothiolan.

3970

S. Ghafari et al. / Bioresource Technology 99 (2008) 39653974

Electrolysis of water: On the anode: 5H2 O ! 2:5O2 10H 10e On the cathode: 10H2 O 10e ! 5H2 10OH 5 4

Stepwise autotrophic denitrication of nitrate to nitrogen gas is illustrated below. The main role of hydrogen gas as an electron donor is shown in this decomposition: 2NO 2H2 ! 2NO 2H2 O 3 2 2NO 2H2 ! N2 O H2 O 2OH 2 N2 O H2 ! N2 H2 O 6 7 8

Net reaction of denitrication on cathode obtained from Eqs. (5)(8): 2NO 6H2 O 10e ! N2 12OH 3 9

The overall reaction in bio-electrochemical reactor is a combination of Eqs. (4) and (9): 2NO H2 O ! N2 2:5O2 2OH 3 6.2. Design and specications of BERs Specications of electrodes and design of reactors play a signicant role in denitrication rate. The main parameters in constructing the reactor are materials, shape, number of electrodes and their arrangement as well as conguration 10

of BER. Table 2 lists some of these specications in previous studies. In previous studies, various types of metals and carbons were used as cathode and anode materials. Carbon in the form of GAC and graphite as well as metals such as stainless steel, titanium, nickel and copper has been studied. Some researchers tried to study eciency of dierent electrode materials; for example, Cast and Flora (1998) investigated metal cathode performance in comparison with graphite as the most common electrode because use of graphite electrodes in large-scale processes may be limited by its brittleness and bulk. The shape of electrodes and their arrangements are other important conditions in the construction of the reactor. In the case of the bio-electrochemical method, electrode rods, plane and cylindrical shapes are applied. A plane parallel conguration of electrodes is the most common electrode system, which grants many advantages such as availability of materials and uniform current. However, this conguration is only eective for very high concentrations of contaminants such as in the case of electrochemical reductions (or depositions) of metal ions (Sakakibara and Nakayama, 2001). In contrast, porous material such as packed beds, mesh or reticulated planes, screen metals, and expanded metals have been investigated in both plane and cylindrical shapes in BERs. Using numerous cathodes in BERs resulted in higher eciency due to providing extra surface area for immobilized denitriers and hydrogen production as well as expanded contact area with contaminants. In this case, arrangement of cathodes is signicantly

Table 2 Dierent materials and numbers of cathodes used in various shapes and congurations in bio-electrochemical reactors Electrode shape Plane Plane Number and material of used electrodes Cathode Stainless steel Five cathodes made by contacting GAC with stainless steel expanded metal One graphite felt cathode Anode One anode made of platinum-coated titanium One mesh-type dimensionally stable anode One amorphous carbon anode One graphite anode rod Two expanded metal Pt-coated anodes One carbon anode rod One graphite anode rod One stainless steel anode Surface area of cathode (cm2) 750.0 (Each cathode: 150) 105.0 Conguration Ref.

Cathodes set in parallel at upstream of anode Vertical anode and horizontal cathode Anode between cathodes in parallel Anode at the center surrounded by cylindrical cathode Anodes at the center surrounded by cylindrical cathodes concentrically Anode at the center surrounded by cathodes in parallel (attached to the cap) Anode at the center surrounded by cathodes in parallel (attached to the cap) In parallel and attached to the cap

Beschkov et al. (2004) Prosnansky et al. (2002) Park et al. (2005)

Plane

Plane

Two stainless steel cathodes + porous matrix One cylindrical graphite cathode

251.0

Cylindrical cathode and anode rod Cylindrical

Feleke et al. (1998), Feleke and Sakakibara (2002) Islam and Suidan (1998) Sakakibara and Nakayama (2001) Watanabe et al. (2001)

Rod

Eight expanded metal cathodes made of titanium + porous matrix Twelve Carbon cathode rods

1800.0

Rod

Four stainless steel cathode rods/ Four graphite cathode rods One stainless steel mesh cathode

Cast and Flora (1998)

Tubing electrodes

7.6

Skadberg et al. (1999)

S. Ghafari et al. / Bioresource Technology 99 (2008) 39653974

3971

considered and some pathways such as concentrically arrangement of cylindrical cathodes (Sakakibara and Nakayama, 2001) and successive parallel installment of plane cathodes (Prosnansky et al., 2002) have been examined. 6.3. Denitrication performances in BERs Cast and Flora (1998) applied heterotrophic denitriers to compare denitrication eciency of two cathode materials in water treatment. They examined stainless steel rods wrapped with stainless steel mesh (SS/SS) and graphite rod wrapped with polypropylene mesh (Gr/PP). The impact of copper on water denitrication was investigated. They utilized a reactor consisting of a graphite anode rod surrounded by four cathode rods. This bio-electrochemical reactor switched to continuous ow operation after 5 days of batch operation aiming to culturing biolm on the cathodes surface (for over 1 month). Initial concentration of 100 mg NO N=l was tested applying electric current of 3 1 mA with ow rate of 640 ml/d (correspond to a hydraulic residence (HRT) time of 5.5 days). Nitrite was produced by denitrication and accumulated in the electrochemical cell during all experiments and complete denitrication was not achieved in any of the tests. The dierence between reduction rates for two types of cathode materials was insignicant. There was no mention of the exact removal rate using each cathode material. They also found out that the denitrifying process would not be possible when a wastewater stream contained heavy metals due to deposition of heavy metals on the surface of cathodes, which inhibited biolm activity. In 2001, Watanabe and co-workers studied denitrication of high acidic wastewater of copper metal pickling using heterotrophic denitriers. They applied one carbon rod along the axial center of the cylindrical reactor as anode, which was surrounded by 12 carbon cathode rods wrapped with carbon ber felt. The bio-electrochemical reactor was continuously operated with a synthetic wastewater similar to real copper metal pickling wastewater in major compositions, adding acetate to support heterotrophic denitrifying bacteria. Dierent ranges of hydraulic retention time (HRT = 536 h), electric current intensity (1523 mA), and C/N ratio (0.71.2) were examined. Nitrate and copper removal at dierent concentrations of copper (135 mg/l) was also investigated. The results demonstrated that copper ion removal, denitrication and neutralization could be achieved simultaneously by using a single bio-electrochemical reactor. Inuence of copper ions on the performance of denitrifying bacteria was evaluated by means of the denitrication rate and the duration of lag time before denitrication started. Denitrication was eective and almost complete as long as the concentration of copper was less than 30 mg/l. Higher copper concentration led to the accumulation of nitrate while copper was almost removed implying that electric current contributed to copper removal by chemical precipitation instead of

hydrogen production. Nitrite was not produced and pH remained around neutral at all concentrations of copper. Islam and Suidan (1998) used a recycle biolm-electrode reactor consisting of a cylindrical graphite cathode located along the reactor wall and a graphite anode rod located in the center of the reactor. This electrode conguration was used to achieve a uniform current distribution throughout the reactor. Groundwater treatment was examined using a synthetic wastewater containing 20 mg NO N=l and 3 autotrophic denitrifying microorganisms immobilized on the inner surface of cylindrical cathode. Phosphate was used as the buer primarily, but was replaced with carbonate to simulate real groundwater in the absence of phosphate. The experiments were conducted in two phases. During the forward phase, electric current was increased from 0 to 100 mA, while during the reverse phase, current was decreased from 100 to 0 mA. Flow rate and recycle rate in both phases were 2.65 l/d and 1.14 l/min, respectively. Nitrate removal increased as electric current intensity increased to 20 mA. However, there was a decline in nitrate removal when the current was increased from 25 to 100 mA, which is attributed to hydrogen inhibition. Excess hydrogen along with methane started to appear when the current was beyond 60 mA. A stable nitrate removal eciency of 8287% was achieved for over 2 years. Besides, the highest 98% nitrate removal eciency was at current 20 mA using phosphate as a buer. Prosnansky et al. (2002) studied a combined BER and membrane ltration (MF) performance. Autotrophic denitrication using CO2 in a BER consisted of ve porous electrodes acting as multiple cathodes and an inert anode was applied. Installing a membrane after a BER can avoid microorganism escaping from the BER. Cathodes made by contacting GAC with stainless steel expanded metals and were connected in series with the platinum anode coated by titanium. Experimental results revealed that multi-cathode BER was capable of operating at high denitrication rates with low hydraulic retention time of 20 min (HRT = 0.33 h). They reported an enhancement denitrication rate of 16.4 mg NO N=h, which is 360 times 3 higher than previous studies. This implied that they overcame the slow operation of biological denitrication, the most signicant drawback of this method. They attributed this superior performance to the large eective surface area and low ORP zone formation in the multi-cathode region. A novel multi-electrode system was proposed by Sakakibara and Nakayama (2001) for biological water treatments by mean of autotrophic denitrifying biolm located on the surface of cathodes. A long term (over 500 days) continuous experiment was conducted to enhance treatment of dilute solutions such as groundwater or surface water containing nitrate, about 20 mg NO N=l. Eight cathodes and 3 two anodes in cylindrical shapes made by expanded metal were set concentrically in that BER. Polyurethane foam was used on the surface of the cathodes for supporting denitrifying biolms. They applied individual electrode potential or electric current to each part of the multi-electrode in

3972

S. Ghafari et al. / Bioresource Technology 99 (2008) 39653974

order to obtain uniform cathodic or anodic reactions without the limitation caused by low electrolytic conductance. The euent nitrite concentration was almost kept below 0.5 mg NO N=l. Sakakibara and Nakayama (2001) used 3 HRT range of 26 h which was two times smaller than those in previous BERs using monopolar electrode systems. They believed that the complete and enhanced denitrication in this long term experiment was attributed to using the multi-electrode system. The system provided a large eective surface area, the charge transfer mechanism by dissociative electrolytes HCO and CO2 , and the for3 3 mation of low ORP zone in the multi-electrode system. Park et al. (2005) denitried a synthetic wastewater by autotrophic microorganisms which gained electrons directly from the cathode. In contrast to the other researchers, they treated high nitrate laden wastewater which was not able to be treated if hydrogen was to be the electron donor. They acclimatized the denitrifying biolm present with 200 mA current for 60 days and achieved 98% nitrate reduction in applying the same current during experiments. With less or higher electric current, lower eciencies were obtained. At higher than 200 mA, electricity was consumed for hydrolysis of water and there was production of hydrogen on cathode. Inuent nitrate range of 20492 mg NO N=l was examined at 200 mA. In the case of 492 mg 3 NO N=l inuent nitrate, there was 434.78 mg NO N=l 3 3 removal. Hydrogen produced was 100-fold lower than the required amount for such reduction, implying that microorganisms had taken electron directly from the cathode. Therefore, unlike other studies, an extremely high inuent nitrate rate was converted to nitrogen gas because denitriers activities were not limited by the amount of produced hydrogen as electron donor. This is meaningful because nitrate levels are usually high in wastewater, up to as much as 5001000 mg NO N=l. Park and co-workers (2005) 3 made a comparison with other researchers results in terms of mg NO N=cm2 (biolm surface area) day). Their max3 imum denitrication rate was 0.17 mg NO N=cm2 d, 3 which was higher than the maximum rate achieved by Kurt et al. (1987) (0.031 mg NO N=cm2 d, Macdonald 3 2 (1990) (0.083 mg NO N=cm d, Liessens et al. (1993) 3 (0.023 mg NO N=cm2 d, and Sakakibara and Kuroda 3 (1993) (0.038 mg NO N=cm2 d. 3 7. Conclusion The available water resources are increasingly diminishing and we must preserve them through treating wastes before discharging. Nitrate removal in drinking water supply or treatment of wastewaters, cannot be carried out through conventional abiotic methods and this review showed the capability of bio-electrochemical method in denitrication. Some baseline information about this method and signicant principles, which inuence BERs performances are illustrated in this survey. There are many advantages of this method compared to other existing

methods in nitrate removal. Some results from previous researches illustrated the direct eect of electric current and reactor design on the nitrate removal. This can provide some new ways for researchers in the application of this method for improving denitrication. Since this method is relatively new and there is neither sucient documentation nor widely implementation in full scale, further studies are needed. Future research directions in this important eld should focus on studying the eects of electrode material and shape, surface area of cathode, and design of the BERs. Development of correlations and mathematical models for the simulation and optimization, and the economic evaluation of the process should be also considered in future research work. Acknowledgements This study was carried out with the aid of a research Grant of Kuok Foundation. References
Almeida, J.S., Reis, M.A.M., Carrondo, M.J.T., 1995. Competition between nitrate and nitrite reduction in denitrication by Pseudomonas uorescens. Biotechnol. Bioeng. 46, 476484. Beschkov, V., Velizarov, S., Agathos, S.N., Lukova, V., 2004. Bacterial denitrication of waste water stimulated by constant electric eld. Biochem. Eng. J. 17, 141145. Biswas, S., Bose, P., 2005. Zero-valent iron-assisted autotrophic denitrication. J. Environ. Eng. 131 (8), 12121220. Cast, K.L., Flora, J.R.V., 1998. An evolution of two cathode materials and the impact of copper on bio-electrochemical denitrication. Water Res. 32 (1), 6370. Chi, I., Zhang, S.T., Lu, X., Dong, L.H., Yao, S.L., 2004. Chemical reduction of nitrate by metallic iron. J. Water Supply: Res. Technol. AQUA 53 (1), 3741. Chiu, Y.C., Chung, M.S., 2003. Determination of optimal COD/nitrate ratio for biological denitrication. Int. Biodeterior. Biodegrad. 51, 43 49. Chiu, Y.C., Lee, L.L., Chang, C.N., Chao, A.C., 2007. Control of carbon and ammonium ratio for simultaneous nitrication and denitrication in a sequencing batch bioreactor. Int. Biodeterior. Biodegrad. 59, 17. Choe, S., Chang, Y.Y., Hwang, K.Y., Khim, J., 2000. Kinetics of reductive denitrication by nanoscale zero-valent iron. Chemosphere 41 (8), 13071311. Choe, S.H., Ljestrand, H.M., Khim, J., 2004. Nitrate reduction by zerovalent iron under dierent pH regimes. Appl. Geochem. 19 (3), 335 342. Claus, G., Kutzner, H.J., 1985a. Autotrophic denitrication by Thiobacillus denitricans. Appl. Microbiol. Biotechnol. 22 (2), 289296. Claus, G., Kutzner, H.J., 1985b. Physiology and kinetics of autotrophic denitrication by Thiobacillus denitricans. Appl. Microbiol. Biotechnol. 22 (2), 283288. Devlin, J.F., Eedy, R., Butler, B.J., 2000. The eects of electron donor and granular iron on nitrate transformation rates in sediments from a municipal water supply aquifer. J. Contam. Hydrol. 46 (1/2), 8197. Dries, D., Liessens, J., Verstrate, W., Stevens, P., Vos, P., Ley, J., 1988. Nitrate removal from drinking water by means of hydrogenotrophic denitriers in a polyurethane carrier reactor. Water Supply 6, 181192. Environmental Protection Agency, 1976. Quality criteria for water. Environmental Protection Agency, Washington, DC. Ergas, S.J., Reuss, A.F., 2001. Hydrogenotrophic denitrication of drinking water using a hollow bre membrane bioreactor. J. Water Supply: Res. Technol.AQUA 50 (3), 161171.

S. Ghafari et al. / Bioresource Technology 99 (2008) 39653974 European Council Directive, 1998. Directive no. 98/83/EC on the quality of water intented for human consumption. Adopted by the Council, on 3 November 1998. Feleke, Z., Sakakibara, Y., 2002. A bio-electrochemical reactor coupled with adsorber for the removal of nitrate and inhibitory pesticide. Water Res. 36, 30923102. Feleke, Z., Araki, K., Sakakibara, Y., Watanabe, T., Kuroda, M., 1998. Selective reduction of Nitrate to nitrogen gas in a biolm-electrode reactor. Water Res. 32 (9), 27282734. Flere, J.M., Zhang, T.C., 1999. Nitrate removal with sulfurlimestone autotrophic denitrication processes. J. Environ. Eng. 125 (8), 721 729. Foglar, L., Briski, F., Sipos, L., Vukovic, M., 2005. High nitrate removal from synthetic wastewater with the mixed bacterial culture. Biores. Technol. 96, 879888. Francis, C.W., Hatcher, C.W., 1980. Biological denitrication of highnitrate wastes generated in the nuclear industry. In: Cooper, P.F., Atkinson, B. (Eds.), Biological Fluidized Bed Treatment of Water and Wastewater. Ellis Horwood Ltd., Chichester. Galvez, J.M., Gomez, M.A., Hontoria, E., Gonzalez-Lopez, J., 2003. Inuence of hydraulic loading and air owrate on urban wastewater nitrogen removal with a submerged xed-lm reactor. J. Hazard. Mater. 101, 219229. Gamble, T.N., Betlach, M.R., Tiedje, J.M., 1977. Numerically dominant denitrifying bacteria from world soils. Appl. Environ. Microbiol. 33, 926939. Gayle, B.P., Boordman, G.D., Serrard, J.H., Benait, R.E., 1989. Biological denitrication of water. J. Environ. Eng. Div. 115, 930935. Gentzar, C.J., 1995. Membrane dissolution of hydrogen for biological nitrate removal. The 1995 Water Environ. Fed. Conference, pp. 4060. Ginner, J.L., Alvarez, P.J.J., Smith, S.L., Scherer, M.M., 2004. Nitrate and nitrite reduction by FeO: Inuence of mass transport, temperature, and denitrifying microbes. Environ. Eng. Sci. 21 (2), 219229. Glass, C., Silverstein, J., 1999. Denitrication of high-nitrate, high-salinity wastewater. Water Res. 33 (1), 223229. Gomez, M.A., Galvez, J.M., Hontoria, E., Gonzalez-Lopez, J., 2003. Inuence of concentration on biolm bacterial composition from a denitrifying submerged lter used for contaminated groundwater. J. Biosci. Bioeng. 95 (3), 245251. Grommen, R., Verhaege, M., Verstraete, W., 2006. Removal of nitrate in aquaria by means of electrochemically generated hydrogen gas as electron donor biological denitrication. Aquacul. Eng. 34, 3339. Gros, H., Treutler, K., 1986. Biological denitrication process with hydrogen-oxidizing bacteria for drinking water treatment. Aquaculture 5, 288290. Gros, H., Schnoor, G., Rutten, P., 1986. Nitrate removal from groundwater by autotrophic microorganisms. Water Sup. 4, 1121. Hagopian, D.S., Riley, J.G., 1998. A closer look at the bacteriology of nitrication. Aquacul. Eng. 18, 223244. Hu, H.Y., Goto, N., Fujie, K., 2001. Eect of pH on the reduction of nitrite in water by metallic iron. Water Res. 35 (11), 27892793. Huang, Y.H., Zhang, T.C., 2002. Kinetics of nitrate reduction by iron at near neutral pH. J. Environ. Eng. 128 (7), 604611. Huang, Y.H., Zhang, T.C., 2004. Eects of low pH on nitrate reduction by iron powder. Water Res. 38 (11), 26312642. Huang, C.P., Wang, H.W., Chiu, P.C., 1998. Nitrate reduction by metallic iron. Water Res. 32 (8), 22572264. Huang, Y.H., Zhang, T.C., Shea, P.J., Comfort, S.D., 2003. Eects of oxide coating and selected cations on nitrate reduction by iron metal. J. Environ. Qual. 32 (4), 13061315. Islam, S., Suidan, M.T., 1998. Electrolytic denitrication: long term performance and eect of current intensity. Water Res. 32 (2), 528536. Joo, H.Z., Hirai, M., Shoda, M., 2005. Characteristics of ammonium removal by heterotrophic nitricationaerobic denitrication by alcaligenes faecalis no. 4. J. Biosci. Bioeng. 100 (2), 184191. Killingstad, M.W., Widdowson, M.A., Smith, R.L., 2002. Modeling enhanced in situ denitrication in groundwater. J. Environ. Eng. 128 (6), 491504.

3973

Kim, Y.S., Nakano, K., Lee, T.J., Kanchanatawee, S., Matsumura, M., 2002. On-site nitrate removal of groundwater by an immobilized psychrophilic denitrier using soluble starch as a carbon source. J. Biosci. Bioeng. 93 (3), 303308. Kim, S., Jung, H., Kim, K.S., Kim, I.S., 2004. Treatment of high nitratecontaining wastewaters by sequential heterotrophic and autotrophic denitrication. J. Environ. Eng. 130 (12), 14751480. ` Kleerebezem, R., Mendeza, R., 2002. Autotrophic dentrication for combined hydrogen sulde removal from biogas and postdentrication. Water Sci. Technol. 45 (10), 349356. Kurt, M., Dunn, J., Bourne, J.R., 1987. Biological denitrication of drinking water using autotrophic organisms with H2 in a uidized-bed biolm reactor. Biotechnol. Bioeng. 29, 493501. Liessens, J., Germonpre, R., Beernaert, S., Verstraete, W., 1993. Removing nitrate with a methylotrophic uidized bed: technology and operating performance. J. AWWA 85, 144152. Macdonald, D.V., 1990. Denitrication by an expanded biolm reactor. J. WPCF 62, 796803. Masser, M.P., Rackocy, J., Losordo, T.M., 1999. Recirculating aquaculture tank production systems: management of recirculating systems. Southern Regional Aquaculture Center, Publication no. 452, 12pp. Otte, G., Rosenthal, H., 1979. Management of closed brackish-water system for high density sh culture by biological and chemical water treatment. Aquaculture 18, 169181. Park, E.J., Seo, J.K., Kim, M.R., Jung, I.H., Kim, J.K., Kim, S.K., 2001. Salinity acclimation of immobilized freshwater denitrier. Aquacul. Eng. 24, 169180. Park, H.I., Kim, D.K., Choi, Y., Pak, D., 2005. Nitrate reduction using an electrode as direct electron donor in a biolm-electrode reactor. Proc. Biochem. 40, 33833388. Peyton, B.M., Mormile, M.R., Petersen, J.N., 2001. Nitrate reduction with Halomonas Campisalis: kinetics of denitrication at pH9 and 12.5% NaCl. Water Res. 35, 42374242. Prosnansky, M., Sakakibarab, Y., Kuroda, M., 2002. High-rate denitrication and SS rejection by biolm-electrode reactor (BER) combined with microltration. Water Res. 36, 48014810. Rijn, V.J., Tal, Y., Barak, Y., 1996. Inuence of volatile fatty acids on nitrite accumulation by a Pseudomonas stutzeri strain isolated from a denitrifying uidized bed reactor. Appl. Environ. Microbiol. 62, 2615 2620. Rijn, J.V., Tal, Y., Schreier, H.J., 2006. Denitrication in recirculating systems: theory and applications. Aquacul. Eng. 34, 364376. Robertson, L.A., Kuenen, J.G., 1984. Aerobic denitrication: a controversy revived. Arch. Microbial. 139, 351354. Sakakibara, Y., Kuroda, M., 1993. Electric prompting and control of denitrication. Biotechnol. Bioeng. 42, 535537. Sakakibara, Y.M., Nakayama, T., 2001. A novel multi-electrode system for electrolytic and biological water treatments: electric charge transfer and application to denitrication. Water Res. 35 (3), 768778. Sayre, I.M., 1988. International standards for drinking water. Am. Water Works Assoc. 80, 53. Shrimali, M., Singh, K.P., 2001. New methods of nitrate removal from water. Environ. Pollut. 112, 351359. Singh, S., Ebeling, J., Wheaton, F., 1999. Water quality trials in four recirculating aquacultural system congurations. Aquacul. Eng. 20, 7584. Skadberg, B., Geoly-Horn, S.L., Sangamalli, V., Flora, J.R.V., 1999. Inuence of pH, current and copper on the biological dechlorination of 2,6-dichlorophenol in an electrochemical cell. Water Res. 33 (9), 19972010. Soares, M.I.M., 2000. Biological denitrication of groundwater. Water, Air, Soil Pollut. 123, 183193. Sumino, T., Isaka, K., Ikuta, H., Saiki, Y., Yokot, T., 2006. Nitrogen removal from wastewater using simultaneous nitrate reduction and anaerobic ammonium oxidation in single reactor. J. Biosci. Bioeng. 102 (4), 346351.

3974

S. Ghafari et al. / Bioresource Technology 99 (2008) 39653974 containing copper ion and high-strength nitrate to a bio-electrochemical reactor process. Water Res. 35 (17), 41024110. Zart, D., Eberhard, B., 1998. High rate of aerobic nitrication and denitrication by Nitrosomonas eutropha grown in a fermentor with complete biomass retention in the presence of gaseous NO2 and NO. Arch. Microbiol. 169, 282286. Zayed, G., Winter, J., 1998. Removal of organic pollutants and of nitrate from wastewater from dairy industry by denitrication. Appl. Microbiol. Biotechnol. 49, 469474. Zhang, L.H., Jia, J.P., Ying, D.W., Zhu, N.W., Zhu, Y.C., 2005. Electrochemical eect on denitrication in dierent microenvironments around anodes and cathodes. Res. Microbiol. 156, 8892. Zhu, S., Chen, S., 1999. An experimental study on nitrication biolm performances using a series reactor system. Aquacul. Eng. 20, 245259.

Szekeres, S., Kiss, I., Bejerano, T.T., Soares, M.I.M., 2001. Hydrogendependent denitrication in a two-reactor bio-electrochemical system. Water Res. 35 (3), 715719. Szekeres, S., Kiss, I., Kalman, M., Soares, M.I.M., 2002. Microbial population in a hydrogen-dependent denitrication reactor. Water Res. 36, 40884094. Terada, A., Hibiya, K., Nagai, J., Tsuneda, S., Hirata, A., 2003. Nitrogen removal characteristics and biolm analysis of a membrane-aerated biolm reactor applicable to high-strength nitrogenous wastewater treatment. J. Biosci. Bioeng. 95 (2), 170178. Till, B.A., Weathers, L.J., Alvarez, P.J., 1998. Fe(O)-supported autotrophic denitrication. Environ. Sci. Technol. 32 (5), 634639. Watanabe, T., Motoyama, H., Kuroda, M., 2001. Denitrication and neutralization treatment by direct feeding of an acidic wastewater