Effects of ferric iron on the anaerobic treatment and microbial biodiversity in a coupled microbial electrolysis cell (MEC) – Anaerobic reactor

Effects of ferric iron on the anaerobic treatment and microbial biodiversity in a coupled microbial electrolysis cell (MEC) – Anaerobic reactor

w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 5 7 1 9 e5 7 2 8 Available online at www.sciencedirect.com journal homepage: www.elsevier.com/locate/watre...

1MB Sizes 4 Downloads 52 Views

w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 5 7 1 9 e5 7 2 8

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/watres

Effects of ferric iron on the anaerobic treatment and microbial biodiversity in a coupled microbial electrolysis cell (MEC) e Anaerobic reactor Jingxin Zhang, Yaobin Zhang*, Xie Quan, Shuo Chen Key Laboratory of Industrial Ecology and Environmental Engineering (Dalian University of Technology), Ministry of Education, School of Environmental Science and Technology, Dalian University of Technology, Dalian 116024, China

article info

abstract

Article history:

Adding Fe(III) into a MEC e anaerobic reactor enhanced the degradation of organic matters.

Received 15 February 2013

To clarify the respective effects of combining Fe(III) dosage and a MEC and Fe(III) dosage

Received in revised form

only on strengthening anaerobic digestion, three anaerobic reactors were operated in

25 June 2013

parallel: a MEC e anaerobic reactor with dosing Fe(OH)3 (R1), an anaerobic reactor with

Accepted 27 June 2013

dosing Fe(OH)3 (R2) and a common anaerobic reactor (R3). With increasing influent COD

Available online 6 July 2013

from 1500 to 4000 mg/L, the COD removal in R1 was maintained at 88.3% under a voltage of 0.8 V, which was higher than that in reactor R2 and R3. When the power was cut off, the

Keywords:

COD removal in R1 decreased by 5.9%. The addition of Fe(OH)3 enhanced both anaerobic

Anaerobic

digestion and anodic oxidation, resulting in the effective mineralization of volatile fatty

Fe(III)

acids (VFAs). The reduced Fe(II) combined with electric field resulted more extracellular

Microbial electrolysis cell

polymeric substances (EPS) production. Quantitative real e time PCR showed a higher

Pyrosequencing

abundance of bacteria in the anodic biofilm and R1. Pyrosequencing and denaturing gradient gel electrophoresis (DGGE) analysis revealed that the dominant bacteria and archaea communities were richer and more abundant in the anode biofilm and R1. ª 2013 Elsevier Ltd. All rights reserved.

1.

Introduction

Anaerobic digestion is a desirable technology for highstrength organic wastewater treatment, simultaneously with the generation of bioenergy resources. Anaerobic degradation of organic wastes consists of three steps: hydrolysis/fermentation, acetogenesis and methanogenesis, during which methanogenesis is considered to be a main rate e limiting step. Both the slow metabolism of methanogens and its sensitive characteristic to environmental perturbation are liable to cause the unbalance between acidogenesis and methanogenesis, thereby resulting in the accumulation of volatile fatty acids (VFAs), inhibition of methanogenesis and even failure of

anaerobic process (Connaughton et al., 2006; Zhang et al., 2009). Therefore, how to improve the degradation of VFAs becomes a key problem for anaerobic system to maintain stable operation. Recently, bio-electrochemical reactors (BERs) are attracting attention because of their excellent performance for the digestion of organics and methane/hydrogen production (Sasaki et al., 2011a; Tartakovsky et al., 2011; Liu et al., 2005). The addition of external electrochemical system in an anaerobic reactor can be used to enhance microbial metabolism and control the electric flow (Thrash and Coates, 2008; Sasaki et al., 2010). The reactor system has been conducted by other researchers for enhancing anaerobic digestion (Tartakovsky

* Corresponding author. Tel.: þ86 411 84706460; fax: þ86 411 84706263. E-mail addresses: [email protected], [email protected] (Y. Zhang). 0043-1354/$ e see front matter ª 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.watres.2013.06.056

5720

w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 5 7 1 9 e5 7 2 8

et al., 2011). In our previous works, a pair of electrodes was inserted into an upflow anaerobic sludge blanket (UASB) reactor for enhancement of high salinity and azo dye wastewater treatment (Zhang et al., 2012a,b). This electric-anaerobic reactor could be considered as a coupling of UASB and Microbial electrolysis cell (MEC), and the COD removal was ascribed to the anaerobic digestion and anodic oxidation. To further strengthen the treatment performance of the MEC e anaerobic reactor, the anaerobic digestion efficiency and/or anode oxidation rate should be improved. During the anaerobic digestion, electrons from organics are transferred to various acceptors (Stams et al., 2003). As the electron acceptor, the addition of Fe(III) can significantly enhance the effect of microbial Fe(III) reduction, which further accelerated the degradation of VFA and methane production during methane fermentation (Coates et al., 2005). In addition, the anode oxidation rate of MEC/MFC was mainly attributed to the members of exoelectrogenic bacteria on anode biofilm. It is accepted that most of exoelectrogenic bacteria on anode such as Geobacter belong to iron reducing bacteria (IRB) that can be enriched successfully by the addition of Fe(III) oxides (Bond and Lovley, 2003; Gralnick and Newman, 2007; Ringeisen et al., 2004). Thus, it was speculated that the addition of Fe(III) in the MEC e anaerobic reactor may accelerate the degradation of VFAs. Theoretically, Fe(OH)3 could directly mineralize VFAs via microbial Fe(III) reduction process during anaerobic digestion. Further, it assumed that the addition of Fe(OH)3 might be favorable for improving anode oxidation process through enriching exoelectrogenic bacteria. However, the effect of Fe(OH)3 on the degradation of organics in the MEC e anaerobic reactor was not clarified until now. The response relationship among dosing Fe(OH)3, electrochemical system and microbial communities (bacteria and archaea) was not established to date. To clarify these issues, Fe(OH)3 was dosed into a MEC e anaerobic reactor for enhancement of high organic wastewater treatment in this study. The effect of Fe(OH)3 on anaerobic digestion and anode oxidation of organics was investigated respectively. The changes of the bacterial community structure and archaeal community structure in the bio-electrochemical system were studied firstly by the 454 GS e FLX pyrosequencing technology and the denaturing gradient gel electrophoresis (DGGE). The absolute quantity of bacteria was determined using real-time PCR.

2.

Materials and methods

2.1.

Experimental setup

A couple of carbon felt electrodes (60 mm width  60 mm length) were placed into an acrylic plastic up-flow anaerobic blanket reactor (UASB) (280 mm length  100 mm width  100 mm height) accompanied with the addition of Fe(OH)3 powder (30 g, analytical reagent). The anode felt was located in the anaerobic sludge phase and placing the cathode in the surface of settling section to form a MEC combined UASB reactor (hereafter referred to as R1). The electrodes were connected with a regulated DC power source through an electric wire. The working volume of the reactor was 2 L. The control

experiment was conducted in a common UASB reactor that was the same as R1 but without the electrodes (hereafter referred to as R2). The other control experiment were conducted in a common UASB reactor that was the same as R1 but without electrodes and Fe(OH)3 dosing (hereafter referred to as R3). After being seeded, these three reactors were operated with a hydraulic retention time (HRT) of 24 h at 35  1  C under a continuous mode. During the operation, the three reactors were conducted in parallel with increasing influent COD gradually from 1500 to 4000 mg/L. For the first 71 days, the reactor R1 was operated at a fixed voltage of 0.8 V with the increasing addition of COD. To clarify the effect of electric field, the voltage (0.8 V) was cut off from day 72 to day 91. The working voltage is significantly lower than the theoretical value of electrolysis of water (1.23 V).

2.2.

Inoculum and synthetic wastewater

Seed sludge was obtained from a laboratory-scale UASB reactor in our laboratory. The ratio of volatile suspended sludge to total suspended sludge (VSS/TSS) was 0.75 with initial TSS of 14.4 g/L. These three reactors were fed with synthetic wastewater. Sucrose, NH4Cl and KH2PO4 were added in the Synthetic wastewater as the carbon, nitrogen, and phosphorus sources, respectively, to give a COD:N:P ratio of 200:5:1. The trace elements were added according to the following composition: 1 mL/L of a trace element solution containing Zn at 0.37 mmol/L, Mn at 2.5 mmol/L, Cu at 0.14 mmol/L, Co at 8.4 mmol/L, Ni at 0.25 mmol/L, H3BO3 at 0.8 mmol/L and EDTA at 3.4 mmol/L. The pH of the influent wastewater was adjusted to 7.5 using NaHCO3 solution.

2.3.

Analysis

COD was determined according to the standard methods (APHA, 2005). After sampling from the effluent of reactors, a few drops of dilute hydrochloric was added into the effluent samples to prevent the oxidation of Fe(II). Then, these samples were centrifuged at 8000 g for 10 min to remove impurity. The supernatant was used to determine the concentration of Fe (II) ions using ortho phenanthroline spectrophotometer at an absorbance of 510 nm (Techcomp, UV-2301, Shanghai, China). Sludge sample was taken from the bottom of the reactors, which was added a few drops of dilute hydrochloric to prevent the oxidation of Fe(II) and then centrifuged at 8000 g for 10 min. After removing supernatant, the bottom sludge was washed three times by dilute hydrochloric acid (pH2) to dissolve remnant ferric hydroxide and then granular sludge was selected by a filter screen (bore diameter 0.5 mm). In this way, pure granular sludge was almost separated from remnant ferric hydroxide and most of Fe(II) ion in the sludge was reserved. Finally, these granular sludge were destructed with agua regia (mixture of 2.5 ml 65% HNO3 and 7.5 ml 37% HCl) for the roughly determination of Fe element i.e. Fe(II) in the sludge phase. The pH was recorded using a pH analyzer (Sartorius PB-20, Germany). MLSS and MLVSS were determined based on the weighing method after being dried at 103e105  C and burnt to ash at 550  C. Total VFAs including

5721

w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 5 7 1 9 e5 7 2 8

acetate, propionate and butyrate was determined using a gas chromatograph (Shimadzu, GC-2010/FID, Japan) according to the reported method (Yu et al., 2001). The CH4 production was determined using a gas chromatograph (GC-14C, Shimadzu, Japan) equipped with a thermal conductivity detector. The morphology of the sludge was observed by a transmission electron microscopy (TEM; JEM-2000EX, Japan). The extracellular polymeric substances (EPS) were extracted using a cation exchange resin (Frolund et al., 1996). Polysaccharides in the EPS were evaluated by the sulfuric acid-anthrone method (Wu et al., 2009) and protein in the EPS was analyzed according to the method described by Lowry et al. (Lowry et al., 1951). The abundance of bacteria was quantified by real-time PCR. 16S rDNA primers 1055F and 1392R was used to target almost all bacteria (Harms et al., 2003). The detailed steps of real-time PCR for the quantification of bacteria were conducted according to the methods described by Harms et al. (2003). To investigate the archaeal community structure, the denaturing gradient gel electrophoresis (DGGE) was conducted. The common archaeal primers were GC-787F, which contained a 40-bp GC clamp (forward primer) and 1059R (Yu et al., 2005). The specific steps of DGGE analysis for archaea were done according to the methods described by Hwang et al. (2010). DGGE profiles were analyzed using the Quantity One software.

3.

Results and discussion

3.1. Effects of ferric iron on COD removal and pH changes To investigate the effect of Fe(OH)3 on the electricity assisted anaerobic treatment performance, a MEC e anaerobic reactor (R1) and its two control reactors (R2 and R3) were operated in parallel. As shown in Fig. 1, a voltage of 0.8 V was applied to R1 in the first 71 days, and Fe(OH)3 was dosed into the reactor R1 and R2 once a time on day 5. During the initial 17 days, the COD removal in the three reactors were maintained between 85.4% and 90.7% at an influent COD of 1500 mg/L, and the pH

High e throughput 16S rDNA gene pyrosequencing

COD removal rate (%)

(A) 100

Without voltage

0.8 V voltage

5000

90 4000 80

Fe(OH)3 dosing

70

3000

60 2000

Eff in R1 Eff in R2 Eff in R3 Inf

50 40

InfluentCOD (mg/L)

1000

0

20

40

60

80

Time (d)

(B)

5000

Without voltage

0.8 V voltage

7.8

Fe(OH)3 dosing

4000

7.2

6.6

3000 Eff in R1 Eff in R2 Eff in R3 Inf

6.0

2000

Influent COD (mg/L)

After 91 days of operation, anode carbon felt of R1 was cut and crushed using scissors. Anaerobic sludge was sampled from the bottom of the reactor R1 and R3. The sludge samples were washed with phosphate-buffered saline (pH 7.4), after which the genomic DNA of the sample was extracted using an extraction kit (Bioteke Corporation, Beijing, China) according to the manufacturer’s instructions. The quality of the extracted DNA was checked by determining its absorbance at 260 nm and 280 nm. A set of bacterial primers 8F (50 e AGAGTTTGATCCTGGCTCAG e 30 ) and 533R (50 e TTACCGCGGCTGCTGGCAC e 30 ) was used to amplify the hypervariable V1 e V3 region of bacterial 16S rRNA gene (Lu et al., 2012). In order to sort multiple samples during pyrosequencing, 10-base barcodes were incorporated between 454 adapter and the forward primer. After being purified and quantified, the PCR products of V1eV3 region of 16S rRNA gene was determined by pyrosequencing using the Roche 454 FLX Titanium sequencer (Roche 454 Life Sciences, Branford, CT, USA) according to the methodology described by Margulies et al. (2005). To obtain the effective sequencing data, raw pyrosequencing results was processed as follows: 1) check the completeness of the barcodes and the adapter; 2) remove sequences containing ambiguities (“Ns”); 3) remove sequences shorter than 200 bps. 4) remove low e quality sequence i.e. a sequencing quality value lower than 20. Subsequently, effective sequences were clustered into operation taxonomic unit (OTUs) by a 3% or 5% level using the MOTHUR program. Rarefaction curves and Beta diversity index were conducted by MOTHUR to identify the species diversity for each sample. The OTUs defined by a 3% distance level were classified using the RDP-II classifier at a 50% confidence threshold. Hierarchical cluster analysis was performed by g plots package of R in Linux.

Effluent pH

2.4.

Principal component analysis (PCA) was completed using MOTHUR. Phylogenetic relationships of sequences were conducted according to the method described by Ye et al. (2011). Briefly, the effective sequences obtained from pyrosequencing were compared with Greengenes 16S rRNA gene database using NCBI’s BLASTN tool and the default parameters except for the maximum hit number of 100. And then these sequences were assigned to NCBI taxonomies with MEGAN software by using the lowest common ancestor (LCA) algorithm and the default parameters, namely absolute cut off: BLAST bitscore 35, and relative cut off: 10% of the top hits.

5.4 1000 0

20

40

60

80

Time (d) Fig. 1 e (A) COD removal and (B) pH changes in the three reactors (R1, R2 and R3) under different influent COD concentrations.

5722

w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 5 7 1 9 e5 7 2 8

(A)

Without Voltage

0.8 V voltage

1000

4500

TVFA (COD mg/L)

600

3500 3000 2500

400 Fe(OH)3 dosing

2000

200

1500

0

Influent COD (mg/L)

4000 Eff in R1 Eff in R2 Eff in R3 Inf

800

1000 0

20

40

60

80

Time (d)

(B)

Fe(OH)3 dosing

7

Current (mA)

4000

Current in R1 Inf COD

5

3000

4 3

2000

2 1

Influent COD (mg/L)

6

1000 0-4

6-17

19-42

44-71

Time (d) Fig. 2 e (A) Effluent TVFA concentrations in the three reactors (R1, R2 and R3) and (B) current value in R1 under different influent COD concentrations.

value ranged from 6.8 to 7.2. As influent COD increased gradually to 4000 mg/L, the COD removal in R1 was maintained at 88.3% on day 71, and the effluent pH showed no significant changes. Comparatively, the COD removal in reactor R2 and R3 decreased to 79.5% and 67.4% respectively, and the pH value also dropped to 6.4 and 5.5. Among the three reactors, R1 showed the highest COD removal on day 71. As compared with the reactor R2, reactor R1 presented a higher COD removal of 8.8%. The improved COD removal might be ascribed to the effect of anode oxidation. It is well known that microbes on the anode biofilm can oxidize small molecular VFAs, which could relieve the accumulation of VFAs and driving the COD removal. Besides, the

addition of Fe(OH)3 in R1 also contributed to the improved COD removal, which can be evaluated by the difference of the COD removal rate between reactor R2 and R3. Compared to reactor R3, the enhanced Fe(III) reduction process via the addition of Fe(OH)3 in R2 improved about 12.1% COD removal during anaerobic digestion, which was in agreement with the report of Coates et al. (2005) who demonstrated that stimulated microbial Fe(III) reduction in hog manure can rapidly remove the organic compounds and enhance methane production. Microbial reduction of Fe(III) is a process depended on IRB, which uses Fe(III) as an electron acceptor to oxidize organic matters. Nevertheless, the process of microbial Fe(III) reduction in itself might not be the main contribution for COD removal. As shown in Fig. S1, the mean amounts of Fe(II) reduced in R2 was maintained at 18.7 mg/(L d) in the first 71 days. According to the calculation based on the equation of microbial Fe(III) reduction i.e. 4Fe3þsolid þ CH2O þ H2O ¼ 4Fe2þ þ CO2 þ 4Hþ (1 g CH2O corresponding to 1.067 g COD) (Coates et al. (2005), the contribution of microbial Fe(III) reduction for COD removal was only 0.067%, which was much lower than the improved COD removal rate of 12.1%. At an influent COD concentration of 4000 mg/L from day 48 to day 71, the CH4 yields in the three reactors were 1.99  0.12 L/d (R1), 1.46  0.15 L/d (R2) and 0.51  0.21 L/d (R3) at the standard state (STP), respectively. This result indicated that the addition of Fe(III) in R2 increased methane production by 2 times compared with R3. It is reported that the addition of Fe(III) played a positive effect on the methanogenic fermentation of the organic wastewater through improving the metabolism of methanogens (Ivanov et al., 2002), which was a possible reason for the improved COD removal in R2. Moreover, the methane production increased by about 1/4 due to the electrodes. Proper electric stimulation can accelerate the growth of microbes via promoting microbial metabolism (Thrash and Coates, 2008), which was likely to enhance anaerobic methane production. This speculation was confirmed by the following real-time PCR and DGGE analysis. By calculations based on the reaction CO2 þ 8Hþ þ 8e- ¼ CH4 þ 2H2O (Sasaki et al., 2011b), the electrochemically produced methane (current 6.5 mA) accounted for lower than 3.5% of the increased methane production due to the addition of electrode. According to the calculation based on the theoretical methane production of 0.35 mL/mg CODremoved at standard state (STP) (Toprak, 1995), the methane production efficiency (methane COD/removed COD) were 80.6%, 65.7% and 27.1% in the reactors R1, R2 and R3, respectively. It indicated that the R3 efficiency was very low, probably due to the acidic pH less than 5.5 that made the methanogenic activity of methanogens weak (Bryers, 1985). According to the stoichiometry of the following reactions (Fang and Liu, 2002):

Table 1 e EPS contents in the sludge of R1, R2 and R3 on day 91. Sample

Sludge of R1 Sludge of R2 Sludge of R3

EPS Polypeptides (mg protein/g VSS)

Polysaccharides (mg glucose/g VSS)

Sum (mg EPS/g VSS)

44.6  1.37 41.2  1.71 39.6  1.12

111.2  1.44 69.1  0.73 37.1  0.58

155.8  2.81 110.3  2.44 76.7  1.7

w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 5 7 1 9 e5 7 2 8

4.0

Bacteria 16S rRNA gene 5 ( 10 copies/ng-DNA )

3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

R1

R2

R3

Anode biofilm

Fig. 3 e Real-time PCR quantification of total bacteria in the sludge samples of R1, R2, R3 and anode biofilm.

5723

converting 1 mol of glucose (0.5 mol sucrose C12H22O11) would produce 2 mol of carbon dioxide along with the production of VFAs. Therefore, higher VFAs accumulation in R3 will make more removed COD flow to carbon dioxide that was one of reasons for the low methane production efficiency. To further clarify the effect of anode oxidation, the voltage in R1 was cut off from day 72 to day 91. Subsequently, the COD removal in R1 decreased by 5.9%, which further verified the positive functions of anode oxidation on the degradation of organics. However, the COD removal efficiency in R1 was still maintained at 82.4%, which was slightly higher than that in R2 (78.9%) and R3 (67.1%) suggesting that the biodegradation ability of microbes could be strengthened after domestication under dissimilatory Fe(III) e reducing conditions.

3.2. Effects of ferric iron on current value and VFA production C12 H22 O11 þ H2 O/2C6 H12 O6

(1)

C6 H12 O6 þ 2H2 O/2CH3 COOH þ 4H2 þ 2CO2

(2)

C6 H12 O6 /CH3 ðCH2 Þ2 COOH þ 2H2 þ 2CO2

(3)

VFA is widely considered as process indicator during anaerobic process, because it is the main pre-methanogenic intermediate (Molina et al., 2009). The increase of organic loading easily resulted in the accumulation of VFA which can indicate

Fig. 4 e Comparisons of microbial community structure among R1, R2 and anode biofilm. (A) DGGE profiles of archaeal communities in R1, R3 and anode biofilm, accompanying with cluster analysis based on the UPGMA method. (B) Hierarchical cluster analysis of bacterial communities in R1, R3 and anode biofilm. The y-axis is the clustering of the 50 most abundant OUTs in the effective reads at 3% distance. The OTUs were instructed at phylum level. Sample communities were clustered according to the method of complete linkage. The color intensity of scale indicates relative abundance of each OTU read. Relative abundance was defined as the numbers of sequences affiliated with that OTU divided by the total number of sequences per sample. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

5724

w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 5 7 1 9 e5 7 2 8

the process imbalance (Ahring et al., 1995). As shown in Fig. 2, the VFA in the three reactors tended to be accumulated gradually with the increase of influent COD concentration from 1500 to 4000 mg/L. During the initial 17 days, the mean VFA concentrations in the three reactors ranged from 51.6 to 80.6 mg/L at an influent COD concentration of 1500 mg/L. As the influent COD concentration increased from 1500 to 4000 mg/L, the VFA concentration in R1 increased to 293.1 mg/L on day 71, which was significantly lower than that of 482.6 mg/L and 756.4 mg/L in R2 and R3 respectively. The relative low level of VFA in R2 was attributed to the enhanced microbial reduction of Fe(III) via the addition of Fe(OH)3. The introduction of anode in R1 further accelerated the mineralization of VFAs, which led to the lowest level of VFAs in the effluent. When the voltage was cut off in R1, effluent VFAs increased gradually to 359.5 mg/L which further confirmed the positive effect of electric field for the decomposition of VFAs. This result was in agreement with ours’ (Zhang et al., 2012b) and other reports (Lalaurette et al., 2009). At the first 4 days, Fe(OH)3 has not been dosed into the reactor R1 and R2. At this time, the current between the electrodes of R1 was maintained at 2.15  0.17 mA with the influent COD of 1500 mg/L. Subsequently, the Fe(OH)3 was dosed into the reactor R1 and R2 once a time on day 5. In the following 12 days after the addition of Fe(OH)3, the current of R1 increased rapidly to 4.95  0.21 mA with the same influent COD of 1500 mg/L. According to the calculations based on the anodic reaction of MEC C2H4O2 þ 2H2O þ 8e- ¼ 2CO2 þ 8Hþ (Wrana et al., 2010), the direct contribution of the increased current (about 2.8 mA) in R1 to the COD removal efficiency was lower than 0.5% that was negligible. Nevertheless, the electricity enhanced process on improving the performance of anaerobic fermentation might be slow due to the low metabolism of anaerobic bacteria especially methanogens. Even though there were no obvious and immediate effects for COD removal and/or less residual VFAs, it might be function well after a period of acclimation under the coupling effect of electricity and Fe(III). As the influent COD increased gradually from 1500 to 4000 mg/L, the current increased from 4.95  0.21 to 6.6  0.23 mA. The increased influent COD can lead to more VFAs production via anaerobic fermentation process and direct anode oxidation process that can be further used by exoelectrogens on anode biofilm. More VFAs production could provide more substance for the growth of exoelectrogens, which was likely to explain the increase of current.

attributed to the positive effect of electricity. However, based on the roughly estimation, the increased Fe2þ concentration of 5.6 mg/l was corresponded to a current of around 0.22 mA generated from Fe(III) reduction that was unbalance with the increased current value. It is reported that the outer membrane c-type cytochromes of IRB have a high binding affinity to Fe(III) oxide that can be utilized as electron shuttle to reduce distant terminal acceptors (Xiong et al., 2006; Kato et al., 2010). Therefore, it might be one of the reasons for the increased current after dosing Fe(III) in R1 that was also in agreement with the report of Ji et al. (2011) who used Fe2O3 -modified electrode as an anode of bio-electrochemical reactor that significantly improved the electricity production comparing with the bare anode. To further investigate the effect of environmental changes on microbial cell morphology, the sludge samples were observed by TEM as shown in Fig. S2. TEM images showed that a great amount of flocculus EPS was adhered to the surface of species in R1. Yet, the surface of species in R3 was relative smooth. The microbial EPS was mainly consisted of proteins and polysaccharides. As shown in Table 1, the level of EPS including proteins and polysaccharides extracted from the sludge of reactor R1 was significantly higher than the levels in reactor R2 and R3. EPS preferred to bind with divalent metals such as Fe2þ to form a stable structure, thereby helping the enrichment of microbes. To further confirm the relationship between Fe2þ and EPS production, the Fe2þ content in the bottom sludge of these three reactors was determined. The results showed that the Fe2þ content in the bottom sludge of R1

3.3. Effects of ferric iron on the enrichment of bacteria and sludge characteristics The addition of Fe(OH)3 in R1 enhanced the growth of IRB, resulting in microbial reduction of Fe(III) together with the degradation of organics (Tugel et al., 1986). As a product of microbial Fe(III) reduction, the amount of Fe2þ reflect the intensity of Fe(III) reduction reaction. As shown in Fig. S1, the amounts of Fe(II) reduced in R2 was maintained between 17.5 mg/L and 19.5 mg/L during the operation. As compared, effluent Fe2þ concentration in reactor R1 ranged from 21.2 mg/L and 25.2 mg/L. The increased Fe2þ concentration was

Fig. 5 e Taxonomic classification of the dominant phylogenetic groups from anode biofilm, R1 and R3 at the phylum level. Relative abundance is defined as the number of sequences affiliated with that taxon divided by the total number of sequences per sample (%). The relative abundance of phyla less than 1% of total composition in the three libraries was defined as “other”.

w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 5 7 1 9 e5 7 2 8

were 68.3 mg Fe2þ/g TSS, which was higher than that of 46.1 mg Fe2þ/g TSS in R2. The higher Fe2þ content in the bottom sludge and effluent of R1 was in line with its high EPS content, indicating that Fe(II) production assisted electric field helped to accelerate the formation of EPS. Analysis of real-time PCR quantification of bacteria showed that the numbers of copies of the Bacteria 16S rRNA gene were 2.13  0.41  105 copies/ngDNA in the R1, which was significantly higher than that in the bottom sludge of R2 and R3. In addition, the abundance of bacteria in the anode biofilm also presented 1.77-folds higher than that in the bottom sludge of R1 even though the both samples was in the same reactor. This result indicated that the addition of electric field might be favorable for the enrichment of bacteria. It is believed that proper electric stimulation can promote microbial metabolism (Thrash and Coates, 2008). Therefore, I speculated that electric field might accelerate the metabolism of IRB, resulting in more reduced Fe2þ in R1. Further, high levels of EPS was favorable for cell aggregation

5725

that was also a major reason for the high abundance of bacteria in the R1 and anode biofilm.

3.4.

Analysis of the microbial community

3.4.1.

Biodiversity of bacteria phylotypes

Through pyrosequencing, the effective sequence tags of samples from R1, R3 and anode surface were 8545, 6967 and 8363 respectively. As shown in the rarefaction curves of Fig. S3, at a 3% distance, the observed number of operational taxonomic units (OTUs) were 932 (R1), 964 (anode biofilm) and 607 (R3) respectively, suggesting that the bacterial community in the anodic biofilm and R1 was more richness (Fig. 3). High biodiversity could increase the ecological stability (Tilman et al., 2006; Wrighton et al., 2008), which means that the coupled anaerobic system of R1 with high biodiversity preferred to has higher capacity to resist environmental stress. The difference of microbial community between the

Fig. 6 e Phylogenetic relationships of sequences from the dominant bacterial communities of reactor R1, R3 and anode biofilm. These sequences were assigned into NCBI taxonomies with BLAST and MEGAN. Pie charts indicate the relative abundance of each phyla, classes and family. The relative abundance of the corresponding phyla, classes and family in R1, R3 and anode biofilm is defined as the ratio of each corresponding color area to pie area. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

5726

w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 5 7 1 9 e5 7 2 8

anode biofilm and bottom sludge of R1 was apparent, but reactor R2 had the similar microbial consortia to the bottom sludge of R1 including effective sequence tags (8503), operational taxonomic units (OTUs) (901) and microbial community structure at phylum level (shown in Fig. S5). Therefore, further analysis was only conducted among R1, anode biofilm and R3.

3.5.

Microbial community comparisons

The comparisons of bacterial and archaeal communities were done as shown in Fig. 4. From the hierarchical cluster analysis in Fig. 4B, the bacterial community in anode biofilm and R1 was classed to one cluster, while this cluster was disparted from R3, suggesting that the addition of Fe(III) in R1 play an important role in the formation of anode biofilm. From Table. S1, the beta diversity index between anode and R1 was 0.33. By comparison, the beta diversity index between anode and R3 showed a higher value of 0.638, which means that the effect of electric field changed the bacterial community structure significantly besides Fe(III) after long-time domestication. Principal component analysis (PCA) (Fig. S4) showed that principal components 1 and 2 explained 67.14% and 32.86% of the total community variations, respectively. For most anaerobic reactors, methanogens can represent Archaea (Sekiguchi et al., 1999). As shown in Fig. 4A, the dominant archaeal species from the numbered bands were difference among R1, R3 and anode biofilm. The intensity and numbers of the dominant bands in anode biofilm was significantly higher than that in R1 and R3, suggesting that more archaea were enriched in the anode biofilm.

3.6.

Through pyrosequencing and DGGE analysis, we demonstrated that microbial communities in R1 was more richness and diverse than that in R3 after being enriched. Special bacteria and archaea were selectively enriched on the surface of anode. This result was in agreement with the higher abundance of bacteria on the anodic biofilm. Syntrophic interactions between bacterial groups and archaeal groups in the anodic biofilm and R1 helped to consume more VFAs, which was favorable for the accumulation of VFAs and methane production.

4.

Conclusions

Fe(III) supplemented into a MEC e anaerobic reactor (R1) significantly improved the treatment performance for high concentration organic wastewater with respect to Fe(III) dosage only (R2). The lumped system had a higher rate for organics degradation through enhancing anaerobic degradation ability of VFAs under dissimilatory Fe(III) e reducing conditions. The reduced Fe(II) from Fe(III) reducing process together with the electric field led to more EPS production, which was favorable for the enrichment of bacteria in the anode biofilm and R1. Also, pyrosequencing and DGGE analysis confirmed that both bacterial and archaeal communities in the anode biofilm and R1 became more diverse and abundance than that in R3. Therefore, it is feasible and useful to improve the wastewater treatment performance of bio-electrochemical reactors via dosing Fe(III).

Bacterial taxonomy analyses

Acknowledgments As shown in Fig. 5, the relative bacterial community abundances were identified on the phylum level. Elusimicrobia, Spirochaetes, Armatimonadetes, Actinobacteria, Chloroflexi, Proteobacteria, Bacteroidetes and Firmicutes and Unclassified bacteria were dominated in these three reactors. The main difference among the three bacterial communities was attributed to the different distribution of phylum Proteobacteria, Actinobacteria, Chloroflexi, and Firmicutes. The relative abundance of the phylum Proteobacteria in the anode biofilm and R1 were 25.1% and 19.1% respectively, which was significantly higher than that in R3 (10.5%). To further compare the difference of microbial community in deep, the sequences from the dominant bacterial communities were analyzed at the family level by MEGAN software. As shown in Fig. 6, most of the families were found simultaneously in the three reactors. Yet, there still exist the difference among the three reactors. Some families were only found in the anode biofilm (such as Anaerophaga, Fibrobacteres/Acidobacteria group, Rhodocyclaceae, Syntrophobacterales, epsilon proteobacteria and Gammaproteobacteria), indicating that electrical stimulation helped to enrich some specific species selectively. Also, many other some families, including Clostridium, Clostridiales, Clostridial firmicutes, Bacteroidales, Bacilli etc. were marked by orange color, exist only in R3. Most of these families belong to acidogenic bacteria, which was in agreement with its pH declining.

The authors acknowledge the financial support from the National Basic Research Program of China (21177015), New Century Excellent Talent Program of the Ministry of Education of China (NCET-10-028) and Geping GreenAid ProjectEnvironmental Scientific Research “123 Project” of Liaoning, China.

Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.watres.2013.06.056.

references

Ahring, B.K., Sandberg, M., Angelidaki, I., 1995. Volatile fatty acids as indicators of process imbalance in anaerobic digestors. Appl. Microbiol. Biotechnol. 43 (3), 559e565. APHA, 2005. Standard Methods for the Examination on of Water and Wastewater, eighteenth ed. American Public Health Association, Washington D.C., USA. Bond, D.R., Lovley, D.R., 2003. Electricity production by Geobacter sulfurreducens attached to electrodes. App. Environ. Microbiol. 69 (3), 1548e1555.

w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 5 7 1 9 e5 7 2 8

Bryers, J.D., 1985. Structures modeling of the anaerobic digestion of biomass particulate. Biotechnol. Bioeng. 27, 638e649. Coates, J.D., Cole, K.A., Michaelidou, U., Patrick, J., Mcinerney, M.J., Achenbach, L.A., 2005. Biological control of hog waste odor through stimulated microbial Fe(III) reduction. Appl. Environ. Microbiol. 71 (8), 4728e4735. Connaughton, S., Collins, G., O’Flaherty, V., 2006. Development of microbial community structure and activity in a highrate anaerobic bioreactor at 18  C. Water Res. 40 (5), 1009e1017. Fang, H., Liu, H., 2002. Effect of pH on hydrogen production from glucose by a mixed culture. Bioresour. Technol. 82 (1), 87e93. Frolund, B., Palmgren, R., Keiding, K., Nielsen, P.H., 1996. Extraction of extracellular polymers from activated sludge using a cation exchange resin. Water Res. 30 (8), 1749e1758. Gralnick, J.A., Newman, D.K., 2007. MicroReview: extracellular respiration. Mol. Microbiol. 65 (1), 1e11. Harms, G., Layton, A.C., Dionisi, H.M., Garrett, V.M., Hawkins, S.A., Robinson, K.G., Sayler, G.S., 2003. Real-time PCR quantification of nitrifying bacteria in a municipal wastewater treatment plant. Environ. Sci. Technol. 37 (2), 343e351. Hwang, K., Song, M., Kim, W., Kim, N., Hwang, S., 2010. Effects of prolonged starvation on methanogenic population dynamics in anaerobic digestion of swine wastewater. Bioresour. Technol. 101 (1), S2eS6. Ivanov, V.N., Stabnikova, E.V., Stabnikov, V.P., Kim, I.S., Zubair, A., 2002. Effects of iron compounds on the treatment of fat-containing wastewaters. Appl. Biochem. Microbiol. 38 (3), 295e299. Ji, J.Y., Jia, Y.J., Wu, W.G., Bai, L.L., Ge, L.Q., Gu, Z.Z., 2011. A layer-by-layer self-assembled Fe2O3 nanorod-based composite multilayer film on ITO anode in microbial fuel cell. Colloids Surf. A: Physicochem. Eng. Aspects 390, 56e61. Kato, S., Nakamura, R., Kai, F., Watanabe, K., Hashimoto, K., 2010. Respiratory interactions of soil bacteria with (semi) conductive iron-oxide minerals. Environ. Microbiol. 12 (12), 3114e3123. Lalaurette, E., Thammannagowda, S., Mohagheghi, A., Maness, P.C., Logan, B.E., 2009. Hydrogen production from cellulose in a twostage process combining fermentation and electrohydrogenesis. Int. J. of Hydrogen Energy 34 (15), 6201e6210. Liu, H., Grot, S., Logan, B.E., 2005. Electrochemically assisted microbial production of hydrogen from acetate. Environ. Sci. Technol. 39 (11), 4317e4320. Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement with the folin phenol reagent. J. Biol. Chem. 193 (1), 265e275. Lu, L., Xing, D.F., Ren, N.Q., 2012. Pyrosequencing reveals highly diverse microbial communities in microbial electrolysis cells involved in enhanced H2 production from waste activated sludge. Water Res. 46 (7), 2425e2434. Margulies, M., Egholm, M., Altman, W.E., Attiya, S., Bader, J.S., Bemben, L.A., Berka, J., Braverman, M.S., Chen, Y.J., Chen, Z.T., Dewell, S.B., Du, L., Fierro, J.M., Gomes, X.V., Godwin, B.C., He, W., Helgesen, S., Ho, C.H., Irzyk, G.P., Jando, S.C., Alenquer, M.L.I., Jarvie, T.P., Jirage, K.B., Kim, J.B., Knight, J.R., Lanza, J.R., Leamon, J.H., Lefkowitz, S.M., Lei, M., Li, J., Lohman, K.L., Lu, H., Makhijani, V.B., McDade, K.E., McKenna, M.P., Myers, E.W., Nickerson, E., Nobile, J.R., Plant, R., Puc, B.P., Ronan, M.T., Roth, G.T., Sarkis, G.J., Simons, J.F., Simpson, J.W., Srinivasan, M., Tartaro, K.R., Tomasz, A., Vogt, K.A., Volkmer, G.A., Wang, S.H., Wang, Y., Weiner, M.P., Yu, P.G., Begley, R.F., Rothberg, J.M., 2005. Genome sequencing in micro fabricated high-density picolitre reactors. Nature 437 (7057), 376e380. Molina, F., Garcia, C., Roca, E., Lema, J.M., Molina, F., Castellano, M., 2009. Selection of variables for on-line

5727

monitoring, diagnosis, and control of anaerobic digestion processes. Water Sci. Technol. 60 (3), 615e622. Ringeisen, B.R., Henderson, E., Wu, P.K., Pietron, J., Ray, R., Little, B., Biffinger, J.C., Stabnikov, V.P., Tay, S.T.L., Tay, D.K., Ivanov, V.N., 2004. Effect of iron hydroxide on phosphate removal during anaerobic digestion of activated sludge. Appl. Biochem. Microbiol. 40 (4), 376e380. Sasaki, K., Sasaki, D., Morita, M., Hirano, S., Matsumoto, N., Ohmura, N., Igarashi, Y., 2010. Bioelectrochemical system stabilizes methane fermentation from garbage slurry. Bioresour. Technol. 101 (10), 3415e3422. Sasaki, K., Morita, M., Sasaki, D., Hirano, S., Matsumoto, N., Watanabe, A., Ohmura, N., Igarashi, Y., 2011a. A bioelectrochemical reactor containing carbon fiber texiles enables efficient methane fermentation from garbage slurry. Bioresour. Technol. 102 (13), 6837e6842. Sasaki, K., Morita, M., Sasaki, D., Hirano, S., Matsumoto, N., Watanabe, A., Ohmura, N., Igarashi, Y., 2011b. A bioelectrochemical reactor containing carbon fiber textiles enables efficient methane fermentation from garbage slurry. Bioresour. Technol. 102 (13), 6837e6842. Sekiguchi, Y., Kamagata, Y., Nakamura, K., Ohashi, A., Harada, H., 1999. Fluorescene in situ hydridization using 16S rRNAtargeted oligonucleotides reveals localization of methanogens and selected uncultured bacteria in mesophilic and thermophilic sludge granules. Appl. Environ. Microbiol. 65 (3), 1280e1288. Stams, A.J.M., Oude Elferink, S.J.W.H., Westermann, P., 2003. Metabolic interactions between methanogenic consortia and anaerobic respiring bacteria. Adv. Biochem. Eng. Biotechnol. 81, 31e56. Tartakovsky, B., Mehta, P., Bourque, J.S., Guiot, S.R., 2011. Electrolysis-enhanced anaerobic digestion of wastewater. Bioresour. Technol. 102 (10), 5685e5691. Thrash, J.C., Coates, J.D., 2008. Review: direct and indirect electrical stimulation of microbial metabolism. Environ. Sci. Technol. 42 (11), 3921e3931. Tilman, D., Reich, P.B., Knops, J.M.H., 2006. Biodiversity and ecosystem stability in a decade-long grassland experiment. Nature 441 (7093), 629e632. Toprak, H., 1995. Temperature and organic loading dependency of methane and carbon dioxide emission rates of a full-scale anaerobic waste stabilization pond. Water Res. 29 (4), 1111e1119. Tugel, J.B., Hines, M.E., Jones, G.E., 1986. Microbial iron reduction by enrichment cultures isolated from estuarine sediments. Appl. Environ. Microbiol. 52 (5), 1167e1172. Wrana, N., Sparling, R., Cicek, N., Levin, D., 2010. Hydrogen gas production in a microbial electrolysis cell by electrohydrogenesis. J. Clean. Prod. 18 (1), S105eS111. Wrighton, K.C., Agbo, P., Warnecke, F., Weber, K.A., Brodie, E.L., DeSantis, T.Z., Hugenholtz, P., Andersen, G.L., Coates, J.D., 2008. A novel ecological role of the Firmicutes identified in thermophilic microbial fuel cells. ISME J 2 (11), 1146e1156. Wu, J., Zhou, H.M., Li, H.Z., Zhang, P.C., 2009. Impacts of hydrodynamic shear force on nucleation of flocculent sludge in anaerobic reactor. Water Res. 43 (12), 3029e3036. Xiong, Y.J., Shi, L., Chen, B.W., Mayer, M.U., Lower, B.H., Londer, Y., Bose, S., Hochella, M.F., Fredrickson, J.K., Squier, T.C., 2006. High-affinity binding and direct electron transfer to solid metals by the Shewanella oneidensis MR-1 outer membrane c-type cytochrome OmcA. J. Am. Chem. Soc. 128, 13978e13979. Ye, L., Shao, M.F., Zhang, T., Tong, A.H.Y., Lok, S., 2011. Analysis of the bacterial community in a laboratory-scale nitrification reactor and a wastewater treatment plant by 454-pyrosequencing. Water Res. 45 (15), 4390e4398.

5728

w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 5 7 1 9 e5 7 2 8

Yu, H.Q., Tay, J.H., Fang, H.H.P., 2001. The roles of calcium in sludge granulation during UASB reactor start-up. Water Res. 35 (4), 1052e1060. Yu, Y., Lee, C., Kim, J., Hwang, S., 2005. Group-specific primer and probe sets to detect methanogenic communities using quantitative real-time polymerase chain reaction. Biotechnol. Bioeng. 89 (6), 670e679. Zhang, P., Chen, Y.G., Zhou, Q., 2009. Waste activated sludge hydrolysis and short e chain fatty acids accumulation under

mesophilic and thermophilic conditions: effect of pH. Water Res. 43 (15), 3735e3742. Zhang, J.X., Zhang, Y.B., Quan, X., 2012a. Electricity assisted anaerobic treatment of salinity wastewater and its effects on microbial communities. Water Res. 46 (11), 3535e3543. Zhang, J.X., Zhang, Y.B., Quan, X., Li, Y., Chen, S., Zhao, H.M., Wang, D., 2012b. An anaerobic reactor packed with a pair of Fe-graphite plate electrodes for bioaugmentation of azo dye wastewater treatment. Biochem. Eng. J. 63 (15), 31e37.