Application of cathode modified by reduced graphene oxide/polypyrrole to enhance anammox activity

Abstract

In this paper, a modified carbon felt (serving as the cathode) prepared by coating reduced graphene oxide (RGO) with polypyrrole (PPy) was applied in an electrode-anammox reactor. Compared with an anammox reactor with RGO addition directly, the anammox reactor with the RGO/PPy electrode applied had a higher nitrogen removal rate. Under the optimal voltage of 1.0 V, the nitrogen removal rate of reactor 3 (R3, RGO/PPy modified cathode applied) reached 1524 g-N m⁻³ d⁻¹ on day 198, which was about 37.5% and 13.7% higher than those of reactor 1 (R1, the control) and reactor 2 (R2, RGO addition). Moreover, transmission electron microscope observation indicated that RGO could cross into the anammox cells and lead to a morphological change of anammox biomass cells. In addition, the high affinity of the RGO electrode could make a biofilm of anammox biomass gather onto the surface of the RGO/PPy electrode as observed by scanning electron microscope images. Besides, it was demonstrated that RGO/PPy cathode application increased the crude enzyme activities and the cell quantities of the anammox reactor.

---

1. Introduction

The anaerobic ammonium oxidation (anammox) process is a novel and promising bioprocess for nitrogen removal from wastewater, especially for treating high ammonium concentration and low chemical oxygen demand content.¹ Compared to conventional nitrification and denitrification processes, the anammox process has some unique advantages such as high nitrogen removal, low energy consumption, no demand for an external carbon source and low sludge yield.² Up to now, although a number of studies on a laboratory scale, even full-scale anammox processes, have been conducted, the rapid enrichment of anammox bacteria still faces some challenges owing to the relatively slow growth rate with doubling time ranging from 11 to 20 days.^3,4 Consequently, a promising alternative to overcome the disadvantage of long start-up time is to promote the metabolism of anammox biomass via enhancing the biomass activity. In this way, the growth rate of anammox bacteria could be accelerated and further achieve a fast start-up of the anammox process.

Reduced graphene oxide (RGO) and related nanomaterials have received considerable attention due to their high surface area, unique physico-chemical properties and low cytotoxicity.⁵ Ruiz et al. demonstrated that RGO coated on filters can induce a faster growth of Escherichia coli.⁶ Another study demonstrated that the activities of human fibroblast cells are enhanced by increasing the dose of RGO.⁷ Yin et al. proved that 1.5- to 2-fold enhancement of enzyme activities of anammox biomass was realized with RGO addition owing to its electron transfer rate being much faster than that of coenzyme Q.⁸ But Wang et al. reported that RGO addition could enhance the nitrogen removal performance of anammox biomass by only 10%.⁹ As mentioned above, the efficiency of the combination of RGO and anammox biomass was unsatisfactory compared to the excellent electron transfer ability. Thus, it would make sense to search for approaches to improve the promoting efficiency of RGO in the anammox process.

Also, it cannot be ignored that RGO is one kind of very good electrical conductor. The excellent electrical conductivity of RGO makes it possible for it to be applied to bio-electrochemical reactors (BER). Actually, BER combined with microorganisms for nitrogen removal from wastewater treatment had attracted more attention due to its high efficiency and flexibility.^10,11 One of the most important factors of microbial electrochemical systems is the desirability to capture energy from organic waste or convert waste into commodity chemicals. Low electron transfer efficiency continues to limit the practical application of microbial electrochemical systems.¹² Biofilms can substantially facilitate extracellular electron transfer between bacteria and electrodes superior to that of planktonic bacteria, owing to their higher local cell density, potentially higher local electron shuttle concentration, shorter electron transfer distance, and the involvement of direct electron transfer pathways through c-type cytochromes.¹³ Significantly, bacterial cells could be captured by RGO, where the graphene oxide (GO) acted as a net to catch the bacteria.^14,15 This enabled the incorporation of a large amount of biomass into the biofilm matrix, and formed multiple conductive pathways. Additionally, Wang et al. demonstrated that RGO could facilitate the secretion of extracellular polymeric substance and anammox bacteria could attach onto the surface of the RGO.⁷ Therefore, it could be inferred that the high affinity for microorganisms could help anammox biomass gather onto the surface of an electrode and form a biofilm if the electrode was modified by RGO. Additionally, although the detailed mechanism of the electron transfer between enzymes and electrodes remains to be elucidated, enzymes showed great catalytic performance with electrochemical technology. A universal belief was that the directional movement under weak electric charge of an electrode was convenient for electron transfer of enzymatic reactions.¹⁶ It could be inferred that the electrode reaction could enhance the enzymatic reaction rate involving RGO. Therefore, RGO applied in a bio-electrode reactor could be expected to enhance the activity of anammox biomass.

In this study, a RGO/polypyrrole (PPy) modified cathode was applied in an electrode-anammox reactor to enhance the activity of anammox biomass. Furthermore, the effect factors, such as the variation of enzyme activities, growth rate of anammox biomass and cell structure, were explored.

2. Materials and methods

2.1 Microorganisms and feed media

The anammox biomass used for continuous experiments was taken from a laboratory-scale anammox upflow column reactor in our lab. Anammox bacteria of KSU-1 strain (AB057453.1) accounted for about 70–75% of the total biomass in the seed biomass as determined by FISH observation. The media used in the experiments mainly consisted of ammonium and nitrite in the form of (NH₄)₂SO₄ and NaNO₂. The composition of the trace mineral medium was as described by van der Graaf et al.¹⁷

2.2 Continuous experiments

Three identical upflow column reactors, R1 (the control reactor), R2 (anammox reactor with RGO addition), R3 (electrode-anammox reactor), were applied for continuous experiments. 50 mg GO was uniformly mixed with seed sludge and then added into R2. The cathode of R3 used a carbon fiber felt which was modified by RGO/PPy as an electrode around the inner reactor. The anode electrode was a graphite rod. The working volumes were about 0.5 L with an inner diameter of 5 cm and height of 25 cm. All the reactors contained 50 g (wet weight) anammox biomass resulting in an initial mixed liquor volatile suspended solids (MLVSS) concentration of 4920 mg L⁻¹ for each reactor. The three reactors were continuously fed with the same media, and the influent was purged with 99.5% N₂ to maintain dissolved oxygen (DO) below 0.5 mg L⁻¹. Influent pH was adjusted to 7.0 ± 0.2 by dosing with 2 M HCl and the temperature was maintained at 35 ± 1 °C using a water bath. Fig. 1 shows a schematic diagram of the continuous experiments.

Fig. 1 Schematic diagram of three identical anammox reactors. R1, control reactor; R2, anammox reactor with 100 mg L⁻¹ RGO addition; R3, electrode-anammox reactor with RGO/PPy electrode application.

2.3 Cathode membrane modification by RGO

The carbon fiber felt membrane modifying process was described by Liu et al.¹⁸ Firstly, blank fiber felt was immersed in APS solution (20 mg mL⁻¹, 50 mL) for 10 min and drip-dried in air for 30 min. Then, 1 mL pyrrole dissolved in ethanol and water solutions were ultrasonically mixed with certain concentration of RGO solutions to get the RGO/PPy composites. RGO was obtained from reducing GO by using hydrazine monohydrate, a strong reducing agent. Lastly, the fiber felt was dipped into the composite solution. A large quantity of RGO/PPy composite was formed by polymerization on the surface of blank fiber felt in a short time. After 5 min, the membrane was taken out, suspended in air and rinsed with water.

2.4 Analytical methods

Concentrations of nitrite and nitrate were determined by using an ion-exchange chromatograph (ICS-1100, DIONEX, AR, USA) with an IonPac AS18 anion column after filtration with 0.22 μm pore size membranes. NH₄⁺–N and MLVSS concentrations were measured according to standard methods.¹⁹ pH measurement was done using a digital pH meter (PHS-25, Leici Company, China), while DO was measured using a digital DO meter (YSI, Model 55, USA). Transmission electron microscopy (TEM) and scanning electron microscopy (SEM) observation were carried out according to the methods described by Duan et al.²⁰

2.5 Preparation of biomass extracts and determination of enzyme activity

Firstly, the anammox biomass samples taken from each reactor were centrifuged at 8000 rpm at 4 °C for 20 min. Secondly, 2 g cells (wet weight) were weighed after removing the supernatant followed by washing twice with sodium phosphate buffer solution (20 mM, pH 7.0). Then the washed pellets were re-suspended in 20 mL of the same buffer and lysed by freezing and thawing followed by sonication (225 W, at 4 °C for 30 min, ultrasonic processor CPX 750, USA). Cell debris was separated by centrifugation (22 000 rpm), at 4 °C for 30 min. The supernatant was stored at 4 °C and used as cell extract in the determination of protein and enzyme activities. Protein concentration was measured according to the Bradford procedure,²¹ using bovine serum albumin (BSA) as a standard. Hydrazine dehydrogenase (HDH) activity was determined according to the methods described by Shimamura et al.,²² and the reactions were evaluated as an increase in the absorbance of cytochrome c at 550 nm in the standard mixture using a spectrophotometer (V-560 UV/Vis spectrophotometer, Jasco, Japan). The HDH activity was expressed as μmol of cytochrome c reduced per mg protein per min. Nitrate reductase (NAR) activity was assayed in accordance with the methods reported by Meincke et al.²³ by measuring the consumption of nitrite. Nitrite reductase (NIR) activity was assayed on the basis of the methods described by Hira et al.²⁴ The concentrations of nitrite were detected after 10 min of reaction under anaerobic conditions by a spectrophotometer. Both NAR and NIR activities were expressed as μmol of nitrite reduced per mg protein per min.

2.6 Quantitative PCR assay

Primer pairs Amx 809-F and Amx 1066-R (20) were used for real-time PCR quantification of the anammox bacteria. The reaction volume of 25 μL contained 12.5 μL SYBR Premix Ex Taq (TaKaRa, Dalian, China), 0.4 mg mL⁻¹ BSA, 200 nM final concentration of each primer, 0.5 μL Rox reference dye and 2 μL extracted DNA as a template. Three replicates were analyzed for each sample. The PCR program was as follows: denaturation for 2 min at 95 °C, followed by 40 cycles of 5 s at 95 °C, annealing for 30 s at 62 °C and prolongation for 30 s at 72 °C. Melting curve analysis showed only one peak at T_m = 87.0 °C. No detectable peaks that were associated with primer-dimer artifacts and no other nonspecific PCR amplification products were observed. The plasmid DNA concentration was determined with a Nanodrop ND-1000 UV-Vis spectrophotometer (NanoDrop Technologies, USA). And the anammox bacterial 16S rRNA gene copy number was calculated directly from the concentration of extracted plasmid DNA. Sixfold serial dilutions of a known copy number of the plasmid DNA were subjected to q-PCR assay in triplicate to generate an external standard curve.

3. Results and discussion

3.1 The morphology of the conductive membrane surface

Microscale morphology of the cathode was revealed with SEM observation. For the modified RGO/PPy electrode (Fig. 2B), the surface of the coated felt had some tiny particles compared with the blank carbon felt (Fig. 2A), which demonstrated that RGO/PPy was successful in modifying the surface of the carbon felt. Moreover, amounts of zoogloea granules grew onto the surface of the cathode at the end of the experiment on day 198. The carbon felt was wrapped in a layer of thick biofilms, which can be clearly seen in the SEM image (Fig. 2C) and physical photo (Fig. 2D).

Fig. 2 SEM observation and physical photo of cathode electrode applied in R3. (A) Blank carbon felt; (B) RGO/PPy modified carbon felt; (C) sample taken from cathode in R3 on day 198; (D) physical photo of cathode on day 198.

These images proved that the modified RGO/PPy carbon felt had a strong capacity for capturing anammox bacteria. The high total biomass loading of naturally formed biofilms onto cathode electrode surfaces promoted the nitrogen removal performance of the bio-electrode reactor.²⁵ Besides, it is necessary to form macroflocs for enhancing the cell density because the anammox bacteria were not active until the cell concentration was higher than 10¹⁰ to 10¹¹ cells per mL.²⁶ Thus, the formation of biofilms onto the surface of the cathode might be beneficial for the nitrogen removal performance in the anammox reactor.

3.2 Continuous experiments

Firstly, 100 mg L⁻¹ GO was added into R2 and different voltages at six levels (0.2 V, 0.4 V, 0.6 V, 0.8 V, 1.0 V, 1.2 V) were applied to the electrodes in R3. Fig. 3 shows that both R2 and R3 exhibited better nitrogen removal performance compared with R1 at the beginning of the tests. From day 0 to day 34, the average NRRs of R2 and R3 reached about 698 g-TN m⁻³ d⁻¹, which was 11.9% higher than that of the control. Obviously, RGO enhanced the activity of the anammox biomass, which further improved the nitrogen removal performance of the anammox reactor. During this period, comparing the nitrogen removal performances of R2 with R3, results showed the same tendency, although the voltage of R3 was increased from 0.2 to 0.4 V. However, when the voltage was raised to 0.6 V, the NRR of R3 gradually reached 814 g-TN m⁻³ d⁻¹, which was not only 20.6% higher than that of R1 (675 g-TN m⁻³ d⁻¹) but also 8.24% higher than that of R2 (752 g-TN m⁻³ d⁻¹) on day 61. It was demonstrated that the modified RGO/PPy electrode applied in R3 under appropriate voltage could be more efficient for promoting the activity of anammox biomass compared with only RGO addition of R2. Then, with the voltage continuing to rise, the NRR of R3 showed a continuously increasing trend. A peak NRR of 906 g-TN m⁻³ d⁻¹ (R3) was obtained with a voltage of 1.0 V, which was 1.33 times as high as that of control on day 91. But at the same period, the NRR of R2 with the RGO addition was 822 g-TN m⁻³ d⁻¹, only 13.7% higher than that of the control. And yet, as the voltage reached 1.2 V, the NRR of R2 and R3 returned to the same level on day 106. Considering the higher growth rate of anammox biomass of R3 in previous incubation days (discussion later), the excessive voltage of 1.2 V might exert an extremely negative influence on the anammox biomass. These results demonstrated that the RGO/PPy electrode could enhance the activity of anammox biomass under a certain voltage range from 0.6 V to 1.0 V in the electrode-anammox reactor. Nevertheless, on increasing the voltage over 1.0 V, the nitrogen removal performance of the electrode-anammox process would decrease.

Fig. 3 Comparison of the nitrogen removal performance of the three reactors under different voltages. R1, control reactor; R2, anammox reactor with 100 mg L⁻¹ RGO addition; R3, electrode-anammox reactor with RGO/PPy electrode application. (A) NH₄⁺–N; (B) NO₂⁻–N; (C), NO₃⁻–N; (D) NLR and NRR.

Then, the nitrogen loading rates (NLRs) of the three reactors were increased by shortening HRT with constant influent substrate concentrations to investigate the stability of the RGO/PPy electrode applied and the long-term effects of optimal voltage on the anammox process. As shown in Fig. 4, the NRRs of the three reactors presented very large differences. R3 with the RGO/PPy electrode applied always exhibited a more understanding nitrogen removal performance compared with R1 and R2. At the end of this experiment, the NRR of R3 reached 1524 g-N m⁻³ d⁻¹, which was about 37.5% higher than that of R1 (1108 g-N m⁻³ d⁻¹) and 13.7% higher than that of R2 (1341 g-N m⁻³ d⁻¹) when the NLRs of the three reactors increased to 2166 g-N m⁻³ d⁻¹ on day 198. Although the RGO addition also enhanced the activity of anammox and further increased the nitrogen removal performance of the anammox process, the electrode-anammox reactor with the modified RGO/PPy electrode as cathode showed a better capacity for enhanced nitrogen removal compared with RGO as additive into the anammox reactor.

Fig. 4 Comparison of the nitrogen removal performance of the three reactors under the optimal voltage. R1, control reactor; R2, anammox reactor with 100 mg L⁻¹ RGO addition; R3, electrode-anammox reactor with RGO/PPy electrode application. (A) NH₄⁺–N; (B) NO₂⁻–N; (C) NO₃⁻–N; (D) NLR and NRR.

3.3 TEM observations

Samples for TEM observation were taken from R1, R2 and R3 in order to investigate the variation of cellular structure of anammox bacteria with RGO addition and electrode application on day 198. Evidently, abundant cells in R2 and R3 showed irregular shape, which was different from the normal anammox cells in R1. Additionally, compared with anammox cells of R1, there were some dark particles distributed in the inner cytomembrane of anammox biomass taken from R2 (Fig. 5B). In this study, R1 and R2 were operated at the same conditions except RGO addition or not. It could be speculated that the particles might be RGO. To the best of our knowledge, nearly all the key enzymes including HDH, NIR and NAR are located in anammoxosome.⁴ Because of the efficient electron transfer ability, RGO could get inside anammox cells even cross into anammoxosome to accelerate the enzymatic reaction rate. Moreover, as shown in Fig. 5C and F, amounts of flocculent aggregates were observed located in the edge of the cell membrane of anammox biomass (R3) but one could not see the RGO particles that were seen for R2. It could be speculated that the cytoplasmic ground substances of anammox cells could be tightly adhered to the RGO/PPy electrode. Actually, there have been uncomfirmed reports on the biocompatibility and absorbability of RGO.²⁷ As shown in Fig. 5G–I, with RGO addition and RGO/PPy modified electrode application, the anammox zoogloea of R2 and R3 could become dark red rather than the normal red-brown. The variation of color might result from the biofilm matrix formation fixed with RGO. Yin et al.⁸ demonstrated that RGO could act as an electron shuttle instead of coenzyme Q to participate in enzymatic reaction. Undoubtedly, the RGO wrapped by cytoplasmic ground substances dramatically shortened the electron transfer distance of electron shuttles. Besides, the microbial electron transfer process in the bio-electrode reactor was associated with the bacterial inward transfer of extracellular electrons to intracellular electrons. The observation of anammox biomass in R2 proved that RGO could enter the interior of anammox cells. Due to the strong electroconductivity, RGO could accelerate the directional migration of electrons driven by electric current, which indicated a small internal resistance and would speed up the bacterial inward transfer.²⁸ Based on these results, application of the RGO/PPy modified electrode could provide much better performance compared with only RGO addition in the anammox process.

Fig. 5 TEM observations and physical photos of sludge samples from the anammox reactors. (A), (D) and (G) samples taken from R1 on day 198; (B), (E) and (H) samples taken from R2 on day 198; (C), (F) and (I) samples taken from R3 on day 198.

3.4 Comparison of key enzyme activities

Fig. 6A depicts the crude HDH activities to investigate the variation of the enzyme activities of the three reactors during the incubation time, which were measured on days 0, 45, 90, 135 and 180. On day 0, the HDH activity of seed anammox sludge was calculated as 0.57 μmol cyto. c per min per mg protein. Then, the HDH activities of anammox biomass in three reactors all exhibited an increasing tendency within the cultivation period. 45 days later, the values for the three reactors increased to 0.62, 0.73 and 0.78 μmol cyto. c per min per mg protein, respectively. Compared with the values of the control reactor, the HDH activities of anammox biomass in R2 and R3 were enhanced by 25.8% and 17.7%. As shown in Fig. 6A, a peak value of 1.67 μmol cyto. c per min per mg protein of anammox biomass was achieved in R3 with the RGO/PPy electrode on day 180, which was 1.42 and 1.11 times higher than that of R1 and R2. In addition, the crude NIR and NAR activities exhibited a similar changing trend to that of HDH activities during the incubation time. After 180 days cultivation, both the peak NIR activity of 38.7 μM NO₂⁻–N per mg protein per min and NAR activity of 2.74 μM NO₂⁻–N per mg protein per min were observed from the anammox biomass taken from R3.

Fig. 6 Comparison of the variation of enzyme activities in the whole incubation period. R1, control reactor; R2, anammox reactor with 100 mg L⁻¹ RGO addition; R3, electrode-anammox reactor with RGO/PPy electrode application. (A) HDH; (B) NIR; (C) NAR.

The highest key enzyme activities of R3 were ascribed to the following reasons. Firstly, the RGO as the electron shuttle possessed efficient electron transfer capacity, which might accelerate the electron transfer process of enzymatic reactions between substrates and cytochrome c.⁸ Secondly, electrochemical technology also could enhance the enzyme activities because of the acceleration of electron transfer due to the directional movement under weak electric charge on the electrode.¹⁶ The RGO-modified carbon felt used as the cathode combined the advantages of RGO and bio-electrode reactor. Moreover, the RGO could be wrapped with cytoplasmic ground substances including enzymes and substrates to participate in the enzymatic reaction, which could shorten the electron transfer distance. Thus, it was reasonable for the peak enzyme activities of anammox biomass to occur in the electrode-anammox reactor with the RGO/PPy modified electrode applied.

3.5 qPCR results

qPCR experiments can indirectly reflect the levels of anammox bacteria in reactors. Therefore, the 16S rRNA anammox bacterial copy numbers of the three reactors were also investigated during the running period. At the beginning of this experiment, a seed anammox sludge containing the copy numbers of 4.84 × 10⁹ copies per g biomass was passed into the three reactors. The results are shown in Fig. 7. Clearly, the 16S rRNA copy numbers of all three reactors demonstrated an increasing trend during the running time. After 45 days, the copy numbers of R1, R2 and R3 increased to 5.43, 5.99 and 6.16 × 10⁹ copies per g biomass, which were calculated as 12.1%, 23.8% and 27.3% higher than that of the seed sludge. However, there were distinct differences in the growth rate for the three reactors. On day 180, the peak values of 8.38 × 10⁹ copies per g biomass in R2 and 9.97 × 10⁹ copies per g biomass in R3 were 1.18 and 1.40 times as high as those of the control.

Fig. 7 Comparison of the qPCR results in the whole incubation period. R1, control reactor; R2, anammox reactor with 100 mg L⁻¹ RGO addition; R3, electrode-anammox reactor with RGO/PPy electrode application.

From all the above results, there was a similar changing tendency between the qPCR results and the enzyme activities. The RGO addition and the electrode reaction of R3 with RGO/PPy applied could improve the catalysis of key enzymes for transformation of substrates, which further brought about the consumption of more ammonium and nitrite. Furthermore, the RGO/PPy electrode application was beneficial for the formation of macroflocs, which also could bring about better anammox activities and more efficient nitrogen treatment.²⁹ In fact, the anammox biomass yield was proportional to the nitrogen consumption (0.066 ± 0.01 mol mol⁻¹ ammonium).²⁸ In the cultivation period, R3 always had higher nitrogen removal efficiency during the incubation period corresponding to the long-term acceleration of growth rate of anammox biomass. Thus, the amount of normal flora of R3 with RGO/PPy application was more than that of R1 and R2 after long-term cultivation.

Electrochemical application and RGO addition have been proven to be great ways to enhance the activity of anammox biomass. However, what cannot be ignored is the low electron transfer efficiency of electrochemical systems. Also, regarding RGO insolubility in water as additive, not being mixed with anammox sludge, this method was also inefficient. The electrode-anammox reactor with RGO/PPy modified cathode applied to increase the nitrogen removal rate of the anammox process solved the two inefficiencies. Thus, it could be inferred that RGO/PPy modified cathode application would be a potential benefit to the anammox process for mass cultivation of anammox sludge in continuous experiments.

4. Conclusion

The electrode-anammox reactor with RGO/PPy modified cathode application increased the nitrogen removal rate by 37.5%. Moreover, compared with being directly added into the anammox reactor, RGO coated on the electrode could play much better role in enhancing the activity of anammox biomass. Besides, the RGO/PPy modified cathode applied in the anammox reactor under optimal voltage could promote the crude enzyme activities and accelerate the growth rate of anammox biomass. Furthermore, TEM observation indicated that RGO could enter into anammox cells leading to a variation of anammox cell structure.

Acknowledgements

This work was supported by the Natural Science Foundation of China (no. 21377014), Poyang Lake Water Resources and Ecology Center Open Foundation of Ministry of Water Resources (no. ZXKT201506) and China Postdoctoral Science Foundation Funded Project (no. 2016M592110).

References

1. S. E. Vlaeminck, A. Terada, B. F. Smets, D. Van der Linden, N. Boon, W. Verstraete and M. Carballa, Environ. Sci. Technol., 2009, 43, 5035–5041.
2. H. J. M. Op den Camp, B. Kartal, D. Guven, L. van Niftrik, S. C. M. Haaijer, W. R. van der Star, K. T. vad de Pas-Schoonen, A. Cabezas, Z. Ying, M. C. Schmid, M. M. M. Kuypers, J. van de Vossenberg, H. R. Harhangi, C. Picioreanu, M. C. M. van Loosdrecht, J. G. Kuene, M. Strous and M. S. M. Jetten, Biochem. Soc. Trans., 2006, 34, 174–178.
3. B. Kartal, L. van Niftrik, J. T. Keltjens, H. J. Op den Camp and M. S. M. Jetten, Adv. Microb. Physiol., 2012, 60, 211–262.
4. S. Lackner, E. M. Gilbert, S. E. Vlaeminck, A. Joss, H. Horn and M. C. M. van Loosdrecht, Water Res., 2014, 55, 292–303.
5. K. A. Geim and K. S. Novoselov, Nat. Mater., 2011, 6, 183–191.
6. O. N. Ruiz, K. A. S. Fernando, B. Wang, N. A. Brown, P. G. Luo, N. D. Mcnamara, M. Vangsness, Y. P. Sun and C. E. Bunker, ACS Nano, 2011, 5, 8100–8107.
7. A. Wang, K. Pu, B. Dong, Y. Liu, L. Zhang, Z. Zhang, W. Duan and Y. Zhu, J. Appl. Toxicol., 2013, 33, 1156–1164.
8. X. Yin, S. Qiao, C. Yu, T. Tian and J. T. Zhou, Chem. Eng. J., 2015, 276, 106–112.
9. D. Wang, G. W. Wang, G. Q. Zhang, X. C. Xu and F. L. Yang, Bioresour. Technol., 2013, 131, 527–530.
10. S. Kondaveeti, S. H. Lee, H. D. Park and B. Min, Water Res., 2014, 512014, 25–36.
11. B. Virdis, K. Rabaey, R. A. Rozendal, Z. H. Yuan and J. Keller, Water Res., 2010, 44, 2970–2980.
12. B. E. Logan and K. Rabaey, Science, 2012, 337, 686–690.
13. C. J. Chen, X. X. Huang, C. X. Lei, W. J. Zhu, Y. X. Chen and W. X. Wu, Chemosphere, 2012, 89, 1224–1229.
14. W. F. Chen, S. R. Li, C. H. Chen and L. F. Yan, Adv. Mater., 2011, 23, 5679–5683.
15. H. B. Heersche, P. J. Herrero, J. B. Oostinga, L. M. K. Vandersypen and A. F. Morpurgo, Nature, 2007, 446, 56–59.
16. Y. Sugimoto, Y. Kitazumi, S. Tsujimura, O. Shirai, M. Yamamoto and K. Kano, Biosens. Bioelectron., 2015, 63, 138–144.
17. A. A. van der Graaf, P. de Bruijn, L. A. Robertson, M. S. M. Jetten and J. G. Kuenen, Microbiology, 1996, 142, 2187–2196.
18. L. F. Liu, F. Zhao, J. D. Liu and F. L. Yang, J. Membr. Sci., 2013, 437, 99–107.
19. APHA, 19th. New York, 1995.
20. X. M. Duan, J. T. Zhou, S. Qiao, X. Yin, T. Tian and F. D. Xu, J. Environ. Sci., 2012, 24, 1083–1090.
21. M. M. Bradford, Anal. Biochem., 1976, 1976, 248–254.
22. M. Shimamura, T. Nishiyama, H. Shigetomo, T. Toyomoto, Y. Kawahara, K. Furukawa and T. Fujii, Appl. Environ. Microbiol., 2007, 73, 1065–1072.
23. M. Meincke, E. Bock, D. Kastrau and P. M. H. Kroneck, Arch. Microbiol., 1992, 158, 127–131.
24. D. Hira, H. Toh, C. Migita, H. Okubo, T. Nishiyama, M. Hattori, K. Furukawa and T. Fujii, FEBS Lett., 2012, 586, 1658–1663.
25. W. C. Lee, C. H. Y. X. Lim, H. Shi, L. A. L. Tang, Y. Wang, C. T. Lim and K. P. Loh, Environ. Sci. Technol., 2009, 43, 7571–7577.
26. W. C. Lee, C. H. Y. X. Lim, H. Shi, L. A. L. Tang, Y. Wang, C. T. Lim and K. P. Loh, ACS Nano, 2011, 5, 7334–7341.
27. M. Strous, J. A. Fuerst, E. H. M. Kramer, S. Logemann, G. Muyzer, K. T. Van De Pas-Schoonen, R. Webb, J. G. Kuenen and M. S. M. Jetten, Nature, 1999, 400, 446–449.
28. K. Rabaey and R. A. Rozendal, Annu. Rev. Microbiol., 2010, 8, 706–716.
29. M. Strous, J. J. Heijnen, J. G. Kuenen and M. S. M. Jetten, Appl. Microbiol. Biotechnol., 1998, 50, 589–596.

---