Author's personal copy

Degradation of pentachlorophenol with the presence of fermentableand non-fermentable co-substrates in a microbial fuel cell

Liping Huang a, Linlin Gan a, Qingliang Zhao b,⇑, Bruce E. Logan c, Hong Lu a, Guohua Chen a,d

a Key Laboratory of Industrial Ecology and Environmental Engineering, Ministry of Education (MOE), School of Environmental Science and Technology, Dalian University ofTechnology, Dalian 116024, Chinab State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology, Harbin 150090, Chinac Department of Civil and Environmental Engineering, The Pennsylvania State University, University Park, PA 16802, USAd Department of Chemical and Biomolecular Engineering, Kowloon, Hong Kong University of Science and Technology, Hong Kong, China

a r t i c l e i n f o

Article history:Received 29 May 2011Received in revised form 17 July 2011Accepted 19 July 2011Available online 27 July 2011

Keywords:Microbial fuel cellPentachlorophenolCo-metabolismAcetateGlucose

a b s t r a c t

Pentachlorophenol (PCP) was more rapidly degraded in acetate and glucose-fed microbial fuel cells(MFCs) than in open circuit controls, with removal rates of 0.12 ± 0.01 mg/L h (14.8 ± 1.0 mg/g-VSS-h)in acetate-fed, and 0.08 ± 0.01 mg/L h (6.9 ± 0.8 mg/g-VSS-h) in glucose-fed MFCs, at an initial PCP con-centration of 15 mg/L. A PCP of 15 mg/L had no effect on power generation from acetate but power pro-duction was decreased with glucose. Coulombic balances indicate the predominant product waselectricity (16.1 ± 0.3%) in PCP-acetate MFCs, and lactate (19.8 ± 3.3%) in PCP-glucose MFCs. Current gen-eration accelerated the removal of PCP and co-substrates, as well as the degradation products in bothPCP-acetate and PCP-glucose reactors. While 2,3,4,5-tetrachlorophenol was present in both reactors, tet-rachlorohydroquinone was only found in PCP-acetate MFCs. These results demonstrate PCP degradationand power production were affected by current generation and the type of electron donor provided.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Pentachlorophenol (PCP) has been extensively used as an herbi-cide, insecticide, fungicide, wood preservative, resin, and lubricant.It is present in surface soils from drying areas near wood treatmentplants, industrial wastewater effluents, and treatment lagoons(Field and Sierra-Alvarez, 2008). PCP can be degraded aerobicallyby certain bacteria as the sole carbon and energy source, but itsco-metabolic degradation under anaerobic conditions is regardedas an effective strategy for it removal (Field and Sierra-Alvarez,2008; Karn et al., 2010, 2011).

Anaerobic co-metabolism of PCP is mainly based on reductivedechlorination, in which PCP serves as an electron acceptor andthe co-substrate (an easily degradable substrate) is used to reducethe toxicity and growth inhibition of PCP on microorganisms. Theco-substrate can also act as an inducing agent for biodegradativeenzymes as well as an electron donor for bacterial growth. Severalco-substrates have been used, including glucose, sucrose, formate,methanol, and propionate (Field and Sierra-Alvarez, 2008;Majumder and Gupta, 2008; Damianovic et al., 2009). One chal-lenge to this degradation approach in practice, however, is lowdegradation rates, excess sludge generation, as well as high operat-

ing costs (Field and Sierra-Alvarez, 2008; Damianovic et al., 2009).New processes are needed that can achieve more rapid PCP degra-dation rates and improve the existing process limitations.

One possibility for improved PCP degradation is to use a micro-bial fuel cell (MFC). An MFC is a device that uses microbes to convertthe chemical energy stored in organic and inorganic compoundsinto electricity, providing a low-cost and low-maintenance reactoras well as a process that produces very little sludge (Logan, 2009).While many previous studies showed a wide range of organic sub-strates can be degraded in an MFC, ranging from easily degradableorganics such as acetate to complex wastewaters (Pant et al., 2010),there is now great interest in using the process for bioremediationof aquatic sediments and groundwater pollutants (Lovley andNevin, 2011; Pham et al., 2009; Huang et al., 2011a,b,c). Aromatichydrocarbons (Zhang et al., 2010), alkanes (Morris et al., 2009),chloroethane (Pham et al., 2009), pyridine (Zhang et al., 2009), phe-nol (Luo et al., 2009) and indole (Luo et al., 2010) have been shownto be degraded in MFCs, suggesting that MFCs might be useful forPCP degradation (Dabo et al., 2000; Huang et al., 2011a,c). However,PCP degradation has not yet been examined in MFCs.

The type of co-substrate has been shown to be an important fac-tor for PCP degradation in conventional biological processes(Damianovic et al., 2009; Field and Sierra-Alvarez, 2008; Majumderand Gupta, 2008; Shen et al., 2005). In MFCs it has also been shownthat system performance (power densities, coulomic efficiencies(CEs), biomass production, and other factors) varies with substrate

0960-8524/$ - see front matter � 2011 Elsevier Ltd. All rights reserved.doi:10.1016/j.biortech.2011.07.063

⇑ Corresponding author.E-mail addresses: qlzhao@hit.edu.cn, zhql1962@yahoo.com.cn (Q. Zhao).

Bioresource Technology 102 (2011) 8762–8768

Contents lists available at ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Author's personal copy

type, particularly for fermentable substrates such as glucose com-pared to non-fermentable substrates such as acetate (Huang et al.,2011a; Kiely et al., 2010; Lee et al., 2008; Yang et al., 2010; Zhanget al., 2011). In this study, the effect of acetate and glucose on PCPdegradation in MFCs was explored in order to more thoroughlyexamine the effect of PCP degradation in MFCs in the presence ofdifferent co-substrates. Performance was evaluated in terms ofPCP and co-substrate degradation rates, power production, CEs,and biocatalytic activities including identification and quantifica-tion of PCP metabolites. Deeper insight into these aspects of PCPdegradation in MFCs could enable future applications of MFCs forPCP bioremediation.

2. Methods

2.1. Fuel cell assembly

Two-chamber MFCs (duplicates) were used in all experiments,with the electrodes separated by a cation exchange membrane(CEM) (CMI-7000 Membranes International, Glen Rock, NJ)(2.5 cm in diameter), as described by Wang et al. (2008). Graphitefelt (Sanye Co., Beijing, China) was packed in each compartmentand served as the anode and the cathode. The net working volumeof each chamber was 100 mL. A reference electrode (Ag/AgCl elec-trode, 195 mV versus standard hydrogen electrode, SHE) was usedto obtain cathode and anode potentials, with all voltages reportedhere versus SHE. Two controls (duplicate reactors) were also oper-ated: one was used for abiotic processes and therefore it was notinoculated (abiotic control); the other was run in the open circuitmode to examine changes in PCP and co-substrates in the absenceof current generation. All of the reactors were wrapped with alumi-num foil to exclude light.

2.2. Inoculation and operation

Domestic wastewater collected from primary sedimentationtank of Lingshui Wastewater Treatment Plant in Dalian, China,was used to inoculate the anode. Prior to use, wastewater wassparged with N2 for 15 min. For the initial acclimation, 50 mL ofwastewater was inoculated into 50 mL of a nutrient solution whichcontained (per liter) (NH4)2SO4 0.386 g, K2SO4 0.149 g, NaH2-

PO4�2H2O (3.31 g), Na2HPO4�12H2O (10.31 g), vitamins (12.5 mL/L) and minerals (12.5 mL/L) (Huang and Logan, 2008). Acetate orglucose was added at a final concentration of 780 mg/L (COD ba-sis). After the formation of stable and repeatable power peaks, ana-lytical grade PCP (Sigma, 99.8%) dissolved in 0.2 M NaOH, togetherwith acetate or glucose in a nutrient solution, was refilled as indi-cated. The replacement of anodic solution was done at the end ofeach fed-batch cycle (defined as a voltage of <20 mV). Unlessotherwise stated, the external resistance was set at 500 X. Forthe cathode chamber, the same NaH2PO4–Na2HPO4 buffer wasused along with hexacyanoferrate (50 mM). All reactors were oper-ated at a room temperature of 22 ± 3 �C.

2.3. Analyses

Chemical oxygen demand (COD) and glucose were measuredusing standard methods (State Environmental Protection Adminis-tration, 2002). Biomass in the anodic chamber was assessed as pre-viously reported (Lee et al., 2008). The voltage across an externalresistor was recorded (30 min intervals) using a date acquisitionboard (PISO813, Taiwan). Power density was calculated from thevoltage, external resistance, and normalized by anodic net workingvolume. CE was calculated as total Coulombs collected and dividedby the Coulombs available based on the COD removed. Values of

polarization and power density curve were obtained by MFCs pro-ducing a maximum stable voltage by changing the external resis-tance of the reactors every 30 min from 20 kX down to 50 X.

Samples (0.5 mL) were periodically withdrawn from the reac-tors and filtered through 0.22 lm pore diameter membrane filtersto remove bacteria. The analysis of PCP and volatile fatty acids (for-mate, acetate, lactate and propionate) was performed using a highperformance liquid chromatograph (HPLC Agilent 1100), equippedwith a C18 capillary column (4.6 mm in diameter and 250 mm inlength, ODS-2 Hypersil, Thermo). For PCP determination, the ultra-violet detector was set at 254 nm. The mobile phase was preparedby dissolving trifluoroacetic acid with ultrapure water (pH 2.8) andthe ratio of this solution and methanol was 20:80 (V/V). In the caseof formate, acetate, lactate and propionate determination, theultraviolet detector was set at 215 nm while the mobile phasewas prepared by dissolving phosphoric acid with ultrapure water(2:1000 (V/V)) and the ratio of this solution and methanol was50:50 (V/V). The intermediates of ethanol, butyrate, isobutyrate,and methane and hydrogen in the headspace were analyzed usinga gas chromatograph (Agilent 6820). Standards were prepared fortetrachlorophenols (TeCPs) such as 2,3,5,6-tetrachlorophenol(2,3,5,6-TeCP) (Supelco, 99.8%), 2,3,4,5-tetrachlorophenol (2,3,4,5-TeCP) (Supelco, 99.8%), 2,3,4,6-tetrachlorophenol (2,3,4,6-TeCP)(Riedel-de Haen, 99.9%), as well as tetrachlorohydroquinone(TeCHQ) (Chem Service, 98.7%) due to their presences in conven-tional anaerobic PCP degradation pathways (Damianovic et al.,2009). PCP metabolites were quantitatively determined by HPLCafter a qualitative analysis with an APCI (�) ion trap mass spec-trometer coupled with the LC (Agilent HPLC–MS/MS 6410). Elec-tro-spray ionization (ESI) was operated in a negative mode, withthe scan mass range set from 70.0 to 350.0.

A complete Coulombic analysis was performed during the deg-radation of PCP and co-substrates (acetate or glucose) in order tofollow the fate of the electrons in the system as previouslydescribed (Huang and Logan, 2008). Briefly, the initial total Cou-lombs (CT) are Cx0 = nxbxF, where nx is the number of moles of initialco-substrate, bx is the moles of electrons per mole of co-substrate,and F is Faraday’s constant. As a co-substrate is degraded, the elec-trons can be distributed to electricity (CE), intermediates (CI), bio-mass (CB), PCP dechlorination products (CP), remaining substrate(CS) and other processes (CL) such as aerobic respiration and inter-mediates that were not measured or unknown extracellular poly-meric substances. Therefore,

CT ¼ CS þ CE þ CI þ CB þ Cp þ CL

where CT is the total Coulombs, Cs the co-substrate remaining inanodic chamber, CE the Coulombs recovered as electricity, CI theCoulombs calculated from measured intermediates (Huang andLogan, 2008), CB the Coulombs calculated from biomass accordingto the reported (Lee et al., 2008), CP the Coulombs consumed byPCP dechlorination process, and CL the Coulombs lost to other pro-cesses such as aerobic respiration and intermediates that were notmeasured or unknown extracellular polymeric substances. Since acomplete de-chlorination of 15 mg/L PCP in 100 mL anodic solutiontheoretically only yields 5.4 Coulombs, which is much less than thetheoretical initial 941.3 Coulombs provided by the co-substrate(780 mg/L as COD), Cp in the above equation was neglected andthe equation was simplified to:

CT ¼ CS þ CE þ CI þ CB þ CL

Coulombic balances for intermediates, electrons, biomass, theremaining co-substrate, the lost and the total were then calculatedas the ratio of the corresponding total Coulombic recoveries relativeto the initial Coulombic content (Cx0).

The bioelectrochemical behavior of anodic biofilms was exam-ined using cyclic voltammetry (CV) and a three electrode configu-

L. Huang et al. / Bioresource Technology 102 (2011) 8762–8768 8763

Author's personal copy

ration with a potentiostat (CHI 650A, Chenhua, Shanghai). Thescanned potential between �0.6 and +0.6 V (versus SHE) was per-formed at a scan rate of 1.0 mV/s under quiescent conditions.

3. Results and discussion

3.1. Reactor acclimation

Following inoculation, both acetate-fed and glucose-fed MFCsneeded ca 130 h to reach the peak in power although the peak val-ues were different (Fig. 1A and B). After three-cycles, these reactorsexhibited stable and repeatable peaks in power production. When5–15 mg/L of PCP was added to acetate-fed MFCs, power was notadversely affected and reached ca 2.0 W/m3 for a period of 10–20 h (Fig. 1A). However, adding 20 mg/L of PCP decreased powerto 1.3 W/m3, showing an inhibition of the activity of electrochem-ically active bacteria at this PCP concentration (Fig. 1A).

Power decreased from 1.7 to 1.3 W/m3 when PCP was added at aconcentration of 5 mg/L to glucose-fed reactors, showing that therewas a much greater sensitivity of glucose-acclimated reactors toPCP than acetate-fed reactors (Fig. 1B). Higher concentrations ofPCP (10–20 mg/L) did not further decrease power although 20 mg/L PCP prolonged the time to reach peak power (Fig. 1B). No powerwas generated using PCP (5–20 mg/L) as a sole carbon and energy,indicating the need for co-substrates (data not shown). However,when acetate or glucose was added to the reactors after experi-ments where only PCP was added, power generation was graduallyrecovered and reached the original levels after three cycles usingacetate and five cycles using glucose as a co-substrate. These resultsindicate that co-substrates were needed for PCP degradation.

Polarization curves also showed different effects of PCP additionon the performance of acetate (Fig. 2A and C) and glucose (Fig. 2B andD) fed MFCs. There was no appreciable effect on the maximum

power density when PCP was added to acetate-fed reactors, with amaximum power production of 2.0 ± 0.2 W/m3 at 6.6 A/m3

(Fig. 2A). However, when glucose was added with 15 mg/L of PCP,the maximum power was only 1.3 ± 0.1 W/m3, compared to1.7 ± 0.1 W/m3 when only glucose was added (Fig. 2B and D). Anodepotentials also changed much more over the current density rangewith acetate than glucose (Fig. 2C and D). The anode potential inthe PCP-acetate MFC changed from �0.21 to 0.025 V (Fig. 2C), com-pared to�0.12 to�0.095 V for the PCP-glucose MFC (Fig. 2D). Therewere only slight differences in anode potentials in the absence of PCPwith the different substrates. The development of more positive an-ode potentials did not appreciably affect MFC performance in termsof power densities. The more rapid change in anode potential usingacetate compared to glucose suggests that the use of a fermentablesubstrate such as glucose stabilized anode microbial activities atcurrent densities as high as 9 A/m3. This result suggests that the sub-strate may have had effect on the composition of the biofilm com-munities, a situation which will be examined in future studies.

3.2. Relationship between PCP degradation rate, CE and initial PCPconcentration

Overall, acetate-fed MFCs achieved more PCP removal thanglucose-fed reactors. In both cases, the PCP degradation rates in-creased slightly with initial PCP concentrations ranging from 5 to15 mg/L (Fig. 3A). Acetate-fed MFCs produced a PCP degradationrate of 0.12 ± 0.01 mg/L h (14.8 ± 1.0 mg/g-VSS-h) at an initial PCPof 15 mg/L. This rate was higher than that produced in glucose-fed reactors at the same PCP concentration (0.08 ± 0.01 mg/L h;6.9 ± 0.8 mg/g-VSS-h) (Fig. 3A). Both of these PCP degradation rateswere much higher than those obtained in conventional anaerobicprocesses under fed batch conditions of 0.038 mg/L h (0.0082 mg/g-VSS-h) at an initial PCP of 10 mg/L (Szewczyk and Długonski,2009), 0.051 mg/L h (0.0036 mg/g-VSS-h) at a PCP of 5.0 mg/L(Mun et al., 2008), 0.29 mg/g-VSS-h at a PCP of 3.5 mg/L (Ayudeet al., 2009), and in continuous flow reactors of 0.039 mg/L h at aPCP of 13 mg/L (Li et al., 2010).

A higher PCP concentration of 20 mg/L resulted in much lowerdegradation rates of 0.051 ± 0.01 mg/L h (acetate) and 0.062 ±0.01 mg/L h (glucose). This suggests that at higher PCP concentra-tions the removal of PCP in MFCs was more similar to that of con-ventional anaerobic reactors (Fig. 3A). In the absence of currentgeneration in the open-circuit controls (15 mg/L PCP) PCP degrada-tion rates were much low, with 0.049 ± 0.008 mg/L h (acetate) and0.034 ± 0.007 mg/L h (glucose). Thus, current generation wasshown to improve degradation rates of PCP at these tested concen-tration levels. Our observations were in agreement with previousreports with other chemicals, where the degradation rates of chlo-roethane and diesel were improved by current generation (Morriset al., 2009; Pham et al., 2009). These results may also reflect theimportance of environmental conditions inside the MFC reactorsto the establishment of the PCP degradation process.

The overall CEs decreased with an increase in initial PCP con-centrations in both acetate and glucose MFCs, exhibiting a similarand gradual decrease in CE with PCP concentrations ranging from 5to 15 mg/L (Fig. 3B). The CE of the control (acetate) without PCPwas 37 ± 1%, much higher than that obtained in the MFCs with only5 mg/L PCP (acetate) of CE = 28 ± 0.3%. There was a relatively smal-ler change in CEs in glucose-fed MFCs, with a decrease from31 ± 1% (no PCP) to 27 ± 1% (15 mg/L PCP) (Fig. 3B).

3.3. Power generation and concentrations of co-substrates, PCP, andintermediates over time

The length of a complete fed-batch cycle was 70 h for acetate,compared to 96 h for glucose for PCP-amended MFCs (Fig. 4A). The

Fig. 1. Power production from MFCs in the presence of: (A) acetate, and acetate andPCP; (B) glucose, and glucose and PCP. (Arrows indicate when reactors were fedwith fresh medium; numbers indicate concentration of PCP (mg/L); 4–6 repeatedcycles were performed at each PCP concentration, although only two cycles wereshown here.)

8764 L. Huang et al. / Bioresource Technology 102 (2011) 8762–8768

Author's personal copy

concentrations of co-substrates (Fig. 4B) and PCP (Fig. 4C) were con-tinuously decreased over time, with the rates of PCP degradationenhanced by current generation. The final PCP concentration in ace-tate-fed MFCs was 3.5 ± 0.7 mg/L with current generation, com-pared to 10.3 ± 0.7 mg/L in open circuit controls (Fig. 4C). Inglucose-fed reactors, the final PCP concentration was 7.3 ± 0.8 mg/L with current production, compared to 11.7 ± 0.5 mg/L in open

circuit controls (Fig. 4C). These results show that the degradationrates of both PCP and the co-substrates were improved by electricitygeneration. Acetate-fed reactors performed better than glucose-fedMFCs (Fig. 4A and C), either through increasing activity of bacteriathat could degrade PCP or as a result of acetate being a superiorco-substrate than glucose for PCP degradation. Current generationin an MFC results in the growth of electrochemically active microor-ganisms on the surface of the electrodes (Zhang et al., 2010). Thiscontinuous electron discharge leads to continuous consumption ofelectrons and protons at the cathode, and results in the consumptionof more electrons from substrate metabolism than in an open circuitcontrol (Srikanth et al., 2010; Huang et al., 2011a). This metabolismis apparently beneficial to PCP dechlorination.

Formate, which produced lower power densities than otherfermentation end products (Phuc et al., 2008; Kiely et al., 2010),was found in acetate-fed reactors amended with PCP under bothclosed- and open-circuit conditions (Fig. 4D). Formate can be pro-duced by acetogenic bacteria using hydrogen and carbon dioxideas sole substrates (Chapelle and Bradley, 1996) or formed from car-bon dioxide through microbial electrosynthesis as recently demon-strated on a cathode (Nevin et al., 2011). Alternatively, methanogenic archaea growing on carbon monoxide and water can alsosynthesize formate (Oelgeschlager and Rother, 2008). Thus, thepossibility that formate was produced via carbon dioxide andhydrogen, or by carbon monoxide and water in acetate-fed andPCP-amended MFCs, cannot be excluded.

In glucose MFCs amended with PCP, only lactate was presentunder closed circuit conditions (Fig. 4D) in comparison with ace-tate (data not shown) and lactate in open circuit conditions. Theabsence of acetate in the PCP-glucose closed circuit reactors illus-trates a pathway of glucose conversion to lactate. Other minororganics like propionate, ethanol, isobutyrate and butyrate werenot detected.

Ortho-dechlorination of PCP was observed in all reactorsamended with co-substrates, as evidenced by the formation of2,3,4,5-TeCP (shown as e/z 230.7 in LC–MS), reaching the highestconcentrations of 8.2 ± 0.7 mg/L (acetate) and 5.1 ± 0.8 mg/L (glu-cose) at 96 h (Fig. 4E). There were higher concentrations of2,3,4,5-TeCP produced in current generating MFCs than in open cir-cuit controls, illustrating the enhanced rate of PCP breakdown inthese systems. TeCHQ, a metabolite of PCP degradation in conven-tional biological processes (Field and Sierra-Alvarez, 2008), wasgradually increased to 0.5 ± 0.08 mg/L at 96 h in PCP-acetate fed

Fig. 2. Comparison of voltage output and power density: (A) acetate and PCP (solid symbol), and acetate only (open symbol); (B) glucose and PCP (solid symbol), and glucoseonly (open symbol)-fed MFCs, and anode and cathode potentials; (C) acetate and PCP (solid symbol), and acetate only (open symbol); (D) glucose and PCP (solid symbol), andglucose only (open symbol)-fed MFCs as a function of current density (PCP concentration: 15 mg/L).

Fig. 3. (A) PCP degradation rate and (B) CEs at various initial PCP concentrationswith the presence of acetate ( ) or glucose ( ). ( : PCP-acetate in open circuitcontrols, : PCP-glucose in open circuit controls.)

L. Huang et al. / Bioresource Technology 102 (2011) 8762–8768 8765

Author's personal copy

MFCs (Fig. 4F), suggesting more diverse PCP degradation pathwaysin PCP-acetate fed MFCs than in PCP-glucose fed reactors. 2,3,4,6-TeCP previously found in conventional anaerobic processes (Fieldand Sierra-Alvarez, 2008) was not detected in the MFCs. Therewas also no detectable 2,3,5,6-TeCP in the PCP-acetate fed systems,suggesting a lack of further transformation from TeCHQ. Any tri-chlorophenol and dichlorophenol were un-detectable in bothPCP-acetate and PCP-glucose fed MFCs, further confirming the pri-mary de-chlorination from PCP to TeCP.

Anaerobic degradation of PCP is a step by step process ofremoving consecutive chlorine atoms, in which de-chlorination ofortho-chlorines occurs at the fastest rate and dechlorination ofpara-chlorines occurs at the slowest rate. Chlorophenols hydroxyl-ation in para position is also quite a common reaction in the path-ways for this compound degradation. Thus, the prevalence ofortho-dechlorination and the formation of TeCHQ in the MFCs wasin agreement with data from conventional anaerobic processes(Field and Sierra-Alvarez, 2008; Szewczyk and Długonski, 2009).The presence of TeCP and TeCHQ in PCP-acetate reactors, as wellas TeCP in PCP-glucose reactors reflected the difference of PCP deg-radation pathways due to the different types of co-substrate. Addi-tionally, the accumulation of TeCP and TeCHQ indicated that thesechemicals were the rate limiting step for further degradation, con-sistent with findings using conventional anaerobic processes (Fieldand Sierra-Alvarez, 2008; Szewczyk and Długonski, 2009). Whileenvironmental conditions like oxygen concentration and redoxpotential can affect PCP degradation rates and pathways in conven-tional anaerobic, aerobic, and anoxic reactors (Field and Sierra-Alva-rez, 2008; Damianovic et al., 2009; Karn et al., 2010, 2011), the MFCsused here may have produced appropriate anaerobic–aerobic condi-tions and resulted in more rapid de-chlorination than conventionalbiological processes.

3.4. Coulombic analysis

As shown in Fig. 5A, in the first 12 h of an MFC batch cycle,57.9% of acetate was removed (42.1 ± 3% remaining), with14.5 ± 0.5% recovered as formate, 5.8 ± 0.8% as biomass, and2.4 ± 0.5% as electricity. Therefore, 35.2 ± 2% of the total Coulombsoriginally present in the acetate were rapidly lost to other pro-cesses. Methane was not detectable in the headspace in any exper-iments (data not shown), suggesting that either methanogenesisdid not occur or more likely that methane was produced and lostthrough the cathode. At 72 h, 9.3 ± 0.8% of the Coulombs remainedin the acetate, with 13.1 ± 0.9% in formate, 11.7 ± 0.4% in biomass,and 16.1 ± 0.3% as electricity, with 49.8% not accounted for in theseproducts (Fig. 5A). In the open circuit controls (Fig. 5B), at 72 h,57 ± 4% remained in acetate, 5.6 ± 1% in formate, 23.3 ± 4% in bio-mass, with 14.2% lost to other processes (Fig. 5B).

The Coulombic balance for PCP-glucose fed MFCs at a batch cy-cle time of 72 h (Fig. 5C) resulted in 19.8 ± 3.3% of total Coulombsin lactate, 10.4 ± 0.3% as electricity and 16.3 ± 1.2% as biomass,with 53.5% lost to other processes. In the open circuit controls(Fig. 5D), a majority of Coulombs (36.0 ± 4.0%) were recovered asbiomass with 17.6% in lactate and 8.0% in acetate, resulting in a38.4% lost to unknown processes. In terms of metabolism of elec-trochemically active bacteria, donor and acceptor variations gaverise to differences in bacterial metabolic flux (Yang et al., 2010).Thus, anodic biofilms formed under PCP-acetate or PCP-glucosefed and closed circuit conditions may have changed their metabolicfluxes under open circuit conditions and exhibited different Cou-lombic balances.

These results show that acetate, a non-fermentable substrate,produced higher electron and lower biomass recoveries than glu-cose, a fermentable substrate, in the presence of PCP. These present

Fig. 4. Time course of (A) power generation, (B) acetate and glucose degradation, (C) PCP degradation, (D) minor organics of formate (circle) and lactate (triangle), (E) TeCPand (F) TeCHQ accumulation under closed circuit (solid symbol) and open circuit (open symbol) conditions for acetate (circle) and glucose (triangle).

8766 L. Huang et al. / Bioresource Technology 102 (2011) 8762–8768

Author's personal copy

results relative to electron recoveries and biomass production withthese substrates are consistent with previous observations in theabsence of PCP (Lee et al., 2008; Chae et al., 2009).

3.5. Interactions of microorganisms with electrodes

CVs were used to understand the biocatalytic activities of theanodic biofilms. Three oxidation–reduction peaks were observedfor PCP-acetate biofilms (Fig. 6A), two in the PCP-glucose biofilm(Fig. 6B), and no peaks for the abiotic controls (Fig. 6A and B). Inthe absence of PCP, one set of oxidation–reduction peaks appeared

in both acetate and glucose MFCs, but at different potentials. Thepresence of PCP shifted the peak values from 1.0 mA at 0.16 Vand �1.0 mA at �0.28 V in acetate fed MFCs to 0.75 mA at0.36 V,�0.22 mA at 0.16 V and�0.54 mA at�0.24 V in PCP-acetatereactors (Fig. 6A). PCP-glucose MFCs had shifts in peak values thatwere different from that observed with PCP-acetate reactors(Fig. 6B). A similar oxidative peak of 1.50 mA at 0.36 V and a weak-er reductive peak of �0.31 mA at �0.02 V than that in the absenceof PCP appeared in PCP-glucose MFCs. These results suggest thatthere were changes in the electron transfer potentials used bythe bacteria on the anode as a result of the used of either glucoseor acetate and PCP. The sizes of oxidation–reduction peaks of allthe biofilm anodes became smaller than those where PCP wasnot added, illustrating weaker electrochemical activities in thepresence of PCP.

The composition of the microbial consortia can change with thesubstrate used in MFCs (Logan, 2009; Lovley and Nevin, 2011).Even in the absence of PCP the communities that evolve in MFCsfed glucose are different from those fed acetate (Zhang et al.,2011). Enrichment procedures that use additional compoundscan also affect the community depending on how they are added.For example, glucose and Congo red added sequentially produceddifferent communities than that obtained when the two chemicalswere added simultaneously (Hou et al., 2011). It is therefore rea-sonable to expect that addition of PCP to the MFCs altered the com-position of microbial consortia. Further investigations of the effectsof PCP on microbial communities in MFCs are therefore warranted.

4. Conclusions

Current generation accelerated the removal of PCP with rates of0.12 ± 0.01 mg/L h (14.8 ± 1.0 mg/g-VSS-h) in acetate-fed, and0.08 ± 0.01 mg/L h (6.9 ± 0.8 mg/g-VSS-h) in glucose-fed MFCs, atan initial PCP concentration of 15 mg/L (780 mg/L COD applied).Power production, Coulombic balances and PCP metabolites in ace-tate-fed MFCs were also different from those in glucose-fed MFCs.These results illustrate that PCP degradation rates and power den-sities were affected by current generation and the type of electrondonor provided for current generation.

Fig. 5. Coulombic balance over time from all substrates in (A) PCP-acetate fed MFCs: , total Coulombic balance; , biomass; , electrons; , formate; , acetate; (B) acetateopen circuit controls: , total Coulombic balance; , biomass; , formate; , acetate; (C) PCP-glucose fed MFCs: , total Coulombic balance; , biomass; , electrons; , lactate;

, glucose; and (D) glucose open circuit controls: , total Coulombic balance; , biomass; , lactate; , acetate; , glucose.

Fig. 6. Comparison of CVs in (A) acetate-PCP and (B) glucose-PCP fed MFCs ( : co-substrate and PCP; : co-substrate only; : co-substrate and PCP with abioticcontrol).

L. Huang et al. / Bioresource Technology 102 (2011) 8762–8768 8767

Author's personal copy

Acknowledgements

The authors would like to acknowledge the financial support forthis work provided by the Natural Science Foundation of China (No.21077017) and Projects of State Key Laboratory of Urban WaterResource and Environment, Harbin Institute of Technology (Nos.HC201021 and 2010DX17).

References

Ayude, M.A., Okada, E., González, J.F., Haure, P.M., Murialdo, S.E., 2009. Bacillussubtilis as a bioindicator for estimating pentachlorophenol toxicity andconcentration. J. Ind. Microbiol. Biotechnol. 36, 765–768.

Chae, K.J., Choi, M.J., Lee, J.W., Kim, K.Y., Kim, I.S., 2009. Effect of different substrateson the performance, bacterial diversity, and bacterial viability in microbial fuelcells. Bioresour. Technol. 100, 3518–3525.

Chapelle, F.H., Bradley, P.M., 1996. Microbial acetogenesis as a source of organicacids in ancient Atlantic coastal plain sediments. Geology 24, 925–928.

Dabo, P., Cyr, A., Laplante, F., Jean, F., Menard, H., Lessard, J., 2000. Electrocatalyticdehydrochlorination of pentachlorophenol to phenol or cyclohexanol. Environ.Sci. Technol. 34, 1265–1268.

Damianovic, M.H.R.Z., Moraes, E.M., Zaiat, M., Foresti, E., 2009. Pentachlorophenol(PCP) dechlorination in horizontal-flow anaerobic immobilized biomass (HAIB)reactors. Bioresour. Technol. 100, 4361–4367.

Field, J.A., Sierra-Alvarez, R., 2008. Microbial degradation of chlorinated phenols.Rev. Environ. Sci. Biotechnol. 7, 211–241.

Huang, L.P., Logan, B.E., 2008. Electricity production from xylose in fed-batch andcontinuous-flow microbial fuel cells. Appl. Microbiol. Biotechnol. 80, 655–664.

Huang, L.P., Cheng, S.A., Chen, G.H., 2011a. Bioelectrochemical systems for efficientrecalcitrant wastes treatment. J. Chem. Technol. Biotechnol. 86, 481–491.

Huang, L.P., Chai, X.L., Cheng, S.A., Chen, G.H., 2011b. Evaluation of carbon-basedmaterials in tubular biocathode microbial fuel cells in terms of hexavalentchromium reduction and electricity generation. Chem. Eng. J. 166, 652–661.

Huang, L.P., Regan, J.M., Quan, X., 2011c. Electron transfer mechanisms, newapplications, and performance of biocathode microbial fuel cells. Bioresour.Technol. 102, 316–323.

Hou, B., Sun, J., Hu, Y., 2011. Effect of enrichment procedures on performance andmicrobial diversity of microbial fuel cell for Congo red decolorization andelectricity generation. Appl. Microbiol. Biotechnol. 90, 1563–1572.

Karn, S.K., Chakrabarty, S.K., Reddy, M.S., 2010. Characterization ofpentachlorophenol degrading Bacillus strains from secondary pulp-and-paper-industry sludge. Int. Biodeter. Biodegr. 64, 609–613.

Karn, S.K., Chakrabarti, S.K., Reddy, M.S., 2011. Degradation of pentachlorophenol byKocuria sp. CL2 isolated from secondary sludge of pulp and paper mill.Biodegradation 22, 63–69.

Kiely, P.D., Rader, G., Regan, J.M., Logan, B.E., 2010. Long-term cathode performanceand the microbial communities that develop microbial fuel cells fed differentfermentation endproducts. Bioresour. Technol. 102, 361–366.

Lee, H.S., Parameswaran, P., Kato-Marcus, A., Torres, C.I., Rittmann, B.E., 2008.Evaluation of energy-conversion efficiencies in microbial fuel cells (MFCs)utilizing fermentable and non-fermentable substrates. Wat. Res. 42, 1501–1510.

Li, Z., Yang, S., Inoue, Y., Yoshida, N., Katayama, A., 2010. Complete anaerobicmineralization of pentachlorophenol (PCP) under continuous flow conditions by

sequential combination of PCP-dechlorinating and phenol-degrading consortia.Biotechnol. Bioeng. 107, 775–785.

Logan, B.E., 2009. Exoelectrogenic bacteria that power microbial fuel cells. Nat. Rev.Microbiol. 7, 375–381.

Lovley, D.R., Nevin, K.P., 2011. A shift in the current: new applications and conceptsfor microbe-electrode electron exchange. Curr. Opin. Biotechnol. doi:10.1016/j.copbio.2011.01.009.

Luo, H., Liu, G., Zhang, R., Jin, S., 2009. Phenol degradation in microbial fuel cells.Chem. Eng. J. 147, 259–264.

Luo, Y., Zhang, R., Liu, G., Li, J., Li, M., Zhang, C., 2010. Electricity generation fromindole and microbial community analysis in the microbial fuel cell. J. Hazard.Mater. 176, 759–764.

Majumder, P.S., Gupta, S.K., 2008. Effect of carbon sources and shock loading on theremoval of chlorophenols in sequential anaerobic–aerobic reactors. Bioresour.Technol. 99, 2930–2937.

Morris, J.M., Jin, S., Pruden, A., 2009. Microbial fuel cell in enhancing anaerobicbiodegradation of diesel. Chem. Eng. J. 146, 161–167.

Mun, C.H., He, J., Ng, W.J., 2008. Pentachlorophenol dechlorination by an acidogenicsludge. Wat. Res. 42, 3789–3798.

Nevin, K.P., Hensley, S.A., Franks, A.E., Summers, Z.M., Ou, J., Woodard, T.L.,Snoeyenbos-West, O.L., Lovley, D.R., 2011. Electrosynthesis of organiccompounds from carbon dioxide is catalyzed by a diversity of acetogenicmicroorganisms. Appl. Environ. Microbiol. 77, 2882–2886.

Oelgeschlager, E., Rother, M., 2008. Carbon monoxide-dependent energymetabolism in anaerobic bacteria and archaea. Arch. Microbiol. 190, 257–269.

Pant, D., Bogaert, G.V., Diels, L., Vanbroekhoven, K., 2010. A review of the substratesused in microbial fuel cells (MFCs) for sustainable energy production. Bioresour.Technol. 101, 1533–1543.

Pham, H., Boon, N., Marzorati, M., Verstraete, W., 2009. Enhanced removal of 1,2-dichloroethane by anodophilic microbial consortia. Wat. Res. 43, 2936–2946.

Phuc, T.H., Beomsoek, T., Chang, I.S., 2008. Performance and bacterial consortium ofmicrobial fuel cell fed with formate. Energy Fuels 22, 164–168.

Shen, D.S., Liu, X.W., Feng, H.J., 2005. Effect of easily degradable substrate onanaerobic degradation of pentachlorophenol in an upflow anaerobic sludgeblanket (UASB) reactor. J. Hazard. Mater. B119, 239–243.

Srikanth, S., Mohan, S.V., Sarma, P.N., 2010. Positive anodic poised potentialregulates microbial fuel cell performance with the function of open and closedcircuitry. Bioresour. Technol. 101, 5337–5344.

State Environmental Protection Administration, 2002. The water and wastewatermonitoring methods, 4th ed. China Environmental Science Press, Beijing.

Szewczyk, R., Długonski, J., 2009. Pentachlorophenol and spent engine oildegradation by Mucor ramosissimus. Int. Biodeter. Biodegr. 63, 123–129.

Wang, G., Huang, L., Zhang, Y., 2008. Cathodic reduction of hexavalent chromium[Cr (VI)] coupled with electricity generation in microbial fuel cells. Biotechnol.Lett. 30, 1959–1966.

Yang, T.H., Coppi, M.V., Lovley, D.R., Sun, J., 2010. Metabolic response of Geobactersulfurreducens towards electron donor/acceptor variation. Microbial. CellFactories 9, 90.

Zhang, C., Li, M., Liu, G., Luo, H., Zhang, R., 2009. Pyridine degradation in themicrobial fuel cells. J. Hazard. Mater. 172, 465–471.

Zhang, T., Gannon, S.M., Nevin, K.P., Franks, A.E., Lovley, D.R., 2010. Stimulating theanaerobic degradation of aromatic hydrocarbons in contaminated sediments byproviding an electrode as the electron acceptor. Environ. Microbiol. 12, 1011–1020.

Zhang, Y.F., Min, B., Huang, L.P., Angelidaki, I., 2011. Electricity generation andmicrobial community response to substrate changes in microbial fuel cell.Bioresour. Technol. 102, 1166–1173.

8768 L. Huang et al. / Bioresource Technology 102 (2011) 8762–8768