A high-performance photo-microbial desalination cell

Accepted Manuscript Title: A high-performance photo-microbial desalination cell Author: Yuxiang Liang Huajun Feng Dongsheng Shen Na Li Yuyang Long Yuyang Zhou Yuan Gu Xianbin Ying Qizhou Dai PII: DOI: Reference:

S0013-4686(16)30760-5 http://dx.doi.org/doi:10.1016/j.electacta.2016.03.177 EA 27007

To appear in:

Electrochimica Acta

Received date: Revised date: Accepted date:

1-2-2016 16-3-2016 29-3-2016

Please cite this article as: Yuxiang Liang, Huajun Feng, Dongsheng Shen, Na Li, Yuyang Long, Yuyang Zhou, Yuan Gu, Xianbin Ying, Qizhou Dai, A high-performance photo-microbial desalination cell, Electrochimica Acta http://dx.doi.org/10.1016/j.electacta.2016.03.177 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

A high-performance photo-microbial desalination cell

Yuxiang Liang1,2, Huajun Feng1,2, Dongsheng Shen1,2, Na Li1,2, Yuyang Long1,2, Yuyang Zhou 1,2

,Yanfeng Wang1,2, Yuan Gu1 , Xianbin Ying1, Qizhou Dai3

School of Environmental Science and Engineering, Zhejiang Gongshang University, Hangzhou 310012, China; 1 Zhejiang Provincial Key Laboratory of Solid Waste Treatment and Recycling, Hangzhou 310012, China; 2 College of Environment, Zhejiang University of Technology, Hangzhou 310032, China; 3

Abstract: This paper introduces a high-performance photo-microbial desalination cell (PMDC) based on photo-electrochemical interactions. The anode of the cell was successfully modified with nanostructured α-Fe2O3. The maximum current density of the PMDC during operation was 8.8 A m−2 at an initial salt concentration of 20 g L−1, which was twice that of the unmodified microbial desalination cell. The results of electrochemical impedance spectroscopy and cyclic voltammetry indicated the current increase of PMDC was mainly contributed by the high electron transfer rate at electrode/biofilm interface. The salt concentration of the effluent from the middle chamber was below ca. 1.4 mg L−1 and the salt removal performance of the PMDC was always higher than 96%. The calculated number of harvested electrons agreed well with NaCl removal. In conclusion, the present PMDC effectively offers simultaneous electricity generation and desalination.



Corresponding author. Address: School of Environmental Science and Engineering, Zhejiang Gongshang University, Hangzhou 310012, China. Tel.: +86 571 87397126; fax: +86 571 87397126.

E-mail address: [email protected] 1

Keywords: bio-photocatalyst anode, current generation, desalination, nanostructured α-Fe2O3

1. Introduction Freshwater scarcity [1] and energy shortages [2] are the main dilemmas for most small-sized islands. In the last few decades, enormous efforts have been focused on the supply of both freshwater and energy from renewable sources. At present, numerous typical desalination technologies have been commercialized such as reverse osmosis filtration system, electroosmosis and multiple-effect evaporation [3, 4]. However, such desalination technologies running at the expense of massive energy consumption are not sustainable. A new desalination method named microbial desalination cell (MDC) was invented in 2009 [5], which was an emerging desalination technology that offers simultaneous wastewater treatment, electricity generation, and desalination [6]. Compared with traditional distillation, MDC provided a more energy-efficient and eco-friendly option for desalination [7] through using microbes as catalysts to drive oxidation reactions in order to convert organic waste into electricity [8]. However, MDC technology today appears to be constrained by low desalination efficiency, mainly resulting from the low current density. For example, Mehanna et al. [9] reported a microbial electrodialysis cell for water desalination and hydrogen gas production, while the removal ratio was just 37% due to the low current (2 A m-2). Jacobson et al.[10] used a continuously operated upflow microbial desalination cell with an air cathode, but this system only produced a low current density of 0.7 A m-2 which was difficult to achieve standard desalting effect (< 20%). In addition, there have been a lot of similar reports with low desalination efficiency (<80%). Therefore, the improvement of current densities to achieve better desalination performance is an extremely urgent research topic. The electricity generating mechanism of MDC is similar to that of microbial fuel cell 2

(MFC) which the current is generated by using microbes on anode as catalysts to drive oxidation reactions. At present, a large amount of anodic surface modifications such as heat treatment, electrochemical oxidation, carbon nanotube coating has been successful in increasing the current densities in MFC [11-14]. Therefore, anodic surface modification is available for increasing the current densities of MDC likewise. Recently bio-photoelectrochemical cells have attracted great interest as a new way to increase the power output of microbial fuel cells by exploiting sunlight as a driver without increasing operational cost. For example, hematite nanowire photoanodes for high-performance microbial fuel cells [15], nanowire-bacteria hybrids for unassisted solar carbon dioxide fixation [16], and nanowire photocathodes for solar-driven microbial photoelectrochemical cells [17] have been reported. Therefore, it would be attractive to combine an MDC with a bio-photoelectrochemical cell to obtain a high-performance photo-microbial desalination cell (PMDC). Herein, we aimed to design a new composite bio-photocatalyst anode (PBA) and PMDC. Hematite (α-Fe2O3), as a promising photocatalyst, has attracted considerable attention owing to its small band gap (2.2 eV), which is favorable for the efficient absorption of energy in the ultraviolet and visible light regions [18]. Thus, nanostructured α-Fe2O3 was chosen as the photocatalyst for the PMDC. Graphite was chosen as the anode carrier because of its excellent biocompatibility [19]. The current generation and desalination efficiency of the PBDC were used to evaluate its capacity.

2. Experimental 2.1 Synthesis of bio-photocatalyst anode A circular graphite flake (diameter 40 mm, thickness 5 mm, projected surface area 12.5 cm2) was used as the support for the nanostructured α-Fe2O3. Before use, the flake was ultrasonically cleaned for 30 min using acetone, ethanol, and deionized (DI) water in turn. The α-Fe2O3 film anode was prepared by the anodization method as the figure S1 [20]. First, the graphite flake was immersed into 0.5M FeSO4 solution 3

(ethylene glycol/H2O = 1:8) and an anodization potential of 1.2 V vs Ag/AgCl was applied for 15 min using an electrochemical station (Biologic VSP, Claix, France). All potentials in this work are quoted relative to the Ag/AgCl reference electrode (3.5 M KCl). The obtained electrodes were sintered in a furnace at 500 °C for 2 h and then allowed to cool naturally to room temperature. Last, the unmodified side of the anode was immersed in light liquid paraffin for 10 min (the photo-modified side did not touch the paraffin), and then cleaned to remove the excess paraffin using surfactant for 10 min. 2.2 Reactor construction and operation The reactor was a sealed rectangular three-chamber (anode, middle desalination, and cathode) reactor (Fig. 1) which was separated with an anion exchange membrane (AMI-7001, Membranes International, USA) and cation exchange membranes (CMI-7000, Membranes International). The membranes were immersed in 5% NaCl solution for 24 h before use. The active volume of the anode, middle desalination, and cathode chamber were 40, 10, and 40 mL, respectively. Two reactors were constructed for the experiments: (a) PMDC: equipped with the BPA. The photo side of the BPA was pointed outwards to receive light from a xenon lamp (12V-35W, Shenlei, China), while the other side was pointed inwards for immobilization of the biofilm (projected surface area 12.5 cm2); (b) MDC: equipped with an untreated anode in the same way as PMDC. The cathodes of both reactors were graphite felt (4 cm diameter and 0.4 cm thickness). The anolyte consisted of CH3COONa (1 g L−1), M9 solution [21] (NH4Cl, 0.1 g L−1; NaCl, 0.5 g L−1; KH2PO4, 4.4 g L−1; K2HPO4, 3.4 g L−1; MgSO4, 0.1 g L−1; NaHCO3, 2 g L−1), and trace elements (FeSO4·7H2O, 1.0 mg L−1; CuSO4·5H2O, 0.02 mg L−1; H3BO3, 0.014 mg L−1; MnSO4·4H2O, 0.10 mg L−1; ZnSO4·7H2O, 0.10 mg L−1; Na2MoO4·2H2O, 0.02 mg L−1; CoCl2·6H2O, 0.02 mg L-1). The anolyte was flushed with high-purity nitrogen gas for 10 min before use. The catholyte was 10g L−1 K3[Fe(CN)6]. Both anolyte and catholyte were continuously recirculated through the anode and cathode chambers at a rate of 2 mL h−1 using a peristaltic pump. The initial 4

microorganisms for the reactor were collected from the fresh anodic effluent (10 mL) of an anaerobic granule sludge blanket reactor in Hangzhou, China. All reactors were cultivated at a stable ambient temperature (30 ± 2 °C). 2.3. Analytical methods Photoelectrochemical activities were investigated in the custom-built reactor with the α-Fe2O3 film modified electrode as the working electrode and a clean graphite electrode as the counter electrode. The conductive side of the modified electrode was placed facing the simulated light. A xenon lamp was used as the solar simulator with an intensity of 10 mW cm-2 which measured by light intensity detector (Radiometer, FA-A, China) (25 W). The modified electrode was immersed into the mixed solution with 0.1 M Na2SO4 and 0.1 M Na2SO3 to measure the I-t properties when the potential of the working electrode was set at 0.01 V. Scanning electron microscope (SEM) (G(pro), Phenom, Holland) images of the electrode surface were obtained at an acceleration voltage of 5 kV. X-ray Photoelectron Spectroscopy (XPS) was used to measure the elemental composition of electrodes surface. The O1s (525-535 eV) and Fe2p peaks (704-716 eV) were collected using an EscaLab250Xi spectrometer with a monochromated Al Kα source (Thermo,English) and analyzed by XPSPEAK41 software. The samples for XPS and SEM tests just need to dry at room temperature and cut into small pieces (2mm* 2mm). The primary particle size of the Fe2O3 NPs was evaluated by transmission electron microscope (TEM) (JEOL JEM-1230, Japan) operated at 80 kV. The sample was scraped lightly from α-Fe2O3 film on anode. Then the powder was added to 1 mL of anhydrous ethanol and sonicated in an ultrasonic bath continuously for 0.5 h. Finally, the dispersion dripped on copper screen (20nm) and waited for drying. Salinity was measured using a conductivity meter (Seven Multi, Mettle Toledo, Switzerland). The following in situ electrochemical tests were running during operation with an initial salt concentration of 40 g L−1 and the anode chamber was continuously fed media at a flow rate of 30 mL h-1 to keep the anolyte conductivity 5

stable. Electrochemical impedance spectroscopy (EIS) was carried out at open circuit potential using an electrochemical station (Biologic VSP, Claix, France). The amplitude was 10 mV and the frequency ranged from 10 kHz to 0.005 Hz. The cyclic voltammetry (CV) response of the biofilm on anode was conducted within a potential window of −0.7 V to 0 V at a scan rate of 5 mVs-1. The light On/Off experiment was carried out to test the response of the anodic biofilms to light. The On/Off illumination cycles were controlled by a piece of black paper which could cover the outside of the PBA. The external resistance was 1 Ω and current generation data were collected using a data acquisition instrument (34970A, Agilent, USA). Biofilm samples were subjected to the LIVE/DEAD BacLight bacterial viability test (LIVE/DEAD® BacLight™ Bacterial Viability Kit, Molecular Probes, America) followedthe manufacturer’s instructions. Labeled cells were visualized and z-stacks were captured using a Zeiss LSM 780 confocal laser scanning microscope. The three-dimensional (3D) biofilm images were reconstructed using ZEN 2010 software.

3 Results and discussion Figure 1 and S2 shows the desalination mechanism of the PMDC. The band-edge position of α-Fe2O3 straddles a range of 2.2 eV from −0.3 eV (the edge of the conduction band) to 1.9 eV (the edge of the valence band, VB) [15, 20], while the potential of the biofilm (Geobacter) on the anode surface is ca. −0.28 V [8]. After the photogenerated electrons flowed to the cathode, the real potential on the surface of anode became more positive under the influence of valence band (1.9V). The surface potential of PMDC was still higher than that of MDC (-0.2V) in spite of bending downwards for the surface potential band,. So the electron charge transfer rate will increase with the enlargement of the potential difference between the surface of anode and bacteria. During the process of electricity generation, the number of protons in the anode chamber increases continually because the bacteria oxidize the substrate and transfer electrons to the anode [5]. Therefore, the salt anion (Cl−) will pass through the 6

anion exchange membrane and transfer to the anode. Similarly, the salt cation (Na+) will pass through the cation exchange membrane and transfer to the cathode to maintain the potential balance [5]. After synthesis of α-Fe2O3 film on the electrode, visible red material could be observed on the surface of the photo-bio-anode. The SEM results (Fig. S3) showed that almost 100% of the surface area of graphite anode was covered by a continuous network of rhombic shapes (length ca.100 nm, thickness ca.20 nm), which was same with that in previously report [20]. In addition, the XPS results in the Fe2p and O1s regions of the modified electrode surface showed that both Fe and O peaks can split into three peaks (Fig. S4). The 710.8(Fe2p) and 529.5(O1s) peaks proved that the substance successfully grafted onto the graphite electrode was α-Fe2O3. As shown in Fig. S5, the transient photocurrent of the sample was measured during a repeated ON– OFF illumination cycle. The photocurrent increased to 0.73 A m-2 once the light illuminated. After 1-week acclimation, both the PMDC and MDC were started up and reached a stable maximum current. Figure 2a and S6 shows the current generation measured over time at the untreated and α-Fe2O3 film-modified electrodes. The anode was continuously fed media at a flow rate of 30 mL h-1 avoiding substrate limitations for bacteria on the anode or changes in pH. The maximum current density produced by the PMDC and MDC during operation with an initial salt concentration of 20 g L−1 was 8.8 A m−2, and 4.2 A m−2, respectively. The current densities of both the PMDC and MDC were only about half of the maximum current density when the initial salt concentration of desalination chambers was 5 g L−1. This was because of the high resistance of the desalination chamber (the salt concentration of M9 was about 10 g L−1). So instead, when the salt concentration was higher than that of M9 solution, the main factor limiting the current produced was the salt concentration of M9. Therefore, the maximum current density of PMDC and MDC during operation with initial salt concentration of 10 g L−1, 20 g L−1 and 40 g L−1 was very close. Once the light was turned off, the current density of PMDC decreased to around 5.8 A 7

m−2 (Fig. 3a). Obviously, the photo-electrochemical interaction could enhance the performance of current generation. The light responds of MDC was caused by infrared radiation of the xenon lamp.

We further explored the effect of light

illumination on PBA. CV analysis showed that the acetate oxidation current started around −0.41 V (Fig. 3b). The redox peaks in a CV originate from the biological redox components responsible for electron transfer between the bacteria and the anode which agrees with previous reports [11]. The PMDC current rapidly achieved a current density of 6.9 A m−2 at -0.2 V, which is 2.3 times that of the MDC (3 A m−2) at the same potential. The results were consistent with the current generation in the figure 3a. In addition, A rapid decrease of current density (down to ca. 4.0 A m-2 at -0.2 V) and a rapid increase of current density (up to ca. 6.0 A m-2 at -0.2 V) were recorded for the OFF and ON periods of illumination, respectively, indicating an excellent performance of PBA for increasing electrons between bacteria and the electrode surface. To further prove this point, the EIS was measured to research the charge-transfer resistance between biofilm and anode (Fig. 3c). The charge-transfer resistance was a fairly good indicator of the electron transfer process between bacteria and electrode. The results of EIS were fitted by “R1+R2/Q2+R3/Q3” equation [22], where the “R1”, “R2” and “R3” were denoted as ohmic resistance, charge-transfer resistance, and diffusion resistance of the anode, respectively. Before used, the charge-transfer resistance of PBA (399.3Ω) was similar with that of untreated anode (375.4) (Fig. S8). After 1-week acclimation, The R2 of PMDC was 2.5 Ω which was far lower than that of MDC (54.8 Ω). After switching off the light, the R2 of PMDC increased to 6.2 Ω and then decreased to 3.1 Ω while switching on the light again. The results showed that the current increase of PMDC was mainly contributed by the high electron transfer rate at electrode/biofilm interface under light illumination. However, it is curious that the current density of PMDC in dark was still higher than that of MDC both in CV (Fig. 3b) and current generation (Fig. 3a). What’s more, the R2 of PMDC in dark was also far lower than that of MDC. The reason may be from 8

the difference of the biofilms. A thicker biofilm was acquired on the PBA compared with control based on the CLSM 3D images (Fig. S9). In addition, the protein concentration of PMDC (42.5μg cm-2) was just about 1.5 times than that of MDC (28.1μg cm-2) which was consistent with the descend range of current for the OFF of the illumination. Those results implied that PBA benefits biofilm formation because of its high potential difference. Then the higher concentrations of electroactive bacteria may further reduce the mass transport. In this work, we replaced the salt solution of the desalination chamber of PMDC when the current density dropped below about 0.5 A/m2. The salt solution in desalination chamber of MDC was replaced in the same time even if the current did not damp out. Figure 2b shows that the water in the middle chamber of the PMDC was efficiently desalinated at all four initial salt concentrations. The final salt concentration of the desalinated water was about 296 ± 53 mg L−1 (5 g L−1), 199 ± 82 mg L−1 (10g L−1), 530 ± 82 mg L−1 (20 g L−1), and 1350 ± 1229 mg L−1 (40 g L−1). All of the salt removals were higher than 96%. Compared with that of the traditional MDC proposed by Huang et al. (88–93%) [5], the present PMDC had the better desalination performance. In each cycle (the salt solution of desalination chamber was replaced), the current density remained stable at most time but dropped rapidly at the last moment (when the salt concentration of salt solution was lower than that of M9 solution). To elucidate this process, the resistance was measured using EIS. The ohmic resistance measured for the different initial salt concentrations was found to increase from about 39 Ω at the beginning to 120, 135, 148, and 163 Ω, respectively (Fig. 4a), at the end of the cycle. The difference of ohmic resistance was caused by the various salt concentration of the desalinated water. A finding observed that higher electrolyte concentration could decrease the membrane solution interface resistance due to the compression of electrical double layer [3]. Those results indicated that under low electrolyte conditions, the ion exchange resistance of the middle chamber must have increased. The primary impetus driving the Cl– and Na+ to transfer from the middle chamber to 9

the anode and cathode chambers was the inflow current. It is strange that under open circuit control (CK), small amounts of salt were found to be removed. This is likely caused by the lager concentration gradient [23]. Figure 4b shows the coulombic efficiency between the electrons harvested and NaCl removed. There was good agreement between the amount of electrons harvested and the amount of NaCl removed for initial salt concentration of 5 g L−1 and 10 g L−1. However, the coulombic efficiency became lower with increasing initial salt concentration. This may be caused by the resistance generated by the reverse osmosis [24] of the NaCl from the anode and cathode chambers to the middle chamber later.

4. Conclusions A PBA was successfully fabricated by modifying an anode with a layer of nanostructured α-Fe2O3. The PBA significantly enhanced the electrogenesis capacity of the MDC. The maximum current density of the PMDC during operation was 8.8 A m−2 at an initial salt concentration of 20 g L−1, which was twice that of the MDC. In all experiments, the salt removal of the PMDC was higher than 96%, and the amount of electrons harvested agreed well with the observed NaCl removal. Owing to the aforementioned advantages, PMDC is an effective way to offer simultaneous electricity generation and desalination.

Acknowledgments This work was supported by the National Natural Science Foundation of China (51478431).

References [1] K.C. Zuo, J.X. Cai, S. Liang, S.J Wu, C.Y. Zhang, P. Liang, X. Huang, A ten liter stacked microbial desalination cell packed with mixed ion-exchange resins for secondary fffluent desalination. Environ. Sci. Technol. 48 (2014) 9917-9924. 10

[2] H.P. Luo, P. Jenkins, Z.Y. Ren, Concurrent desalination and hydrogen generation using microbial electrolysis and desalination cells. Environ. Sci. Technol. 45 (2011) 340-344. [3] A. Rommerskirchen, Y. Gendel, M.Wessling, Single module flow-electrode capacitive deionization for continuous water desalination. Electrochem. Comm. 60 (2015) 34-37. [4] Y. Gendel, A.K.E. Rommerskirchen, O. Davidb, M. Wessling, Batch mode and continuous desalination of water using flowing carbon deionization (FCDI) technology. Electrochem. Comm. 45 (2014) 152-156. [5] X.X. Cao, X. Huang, P. Liang, K. Xiao, Y.J. Zhou, X.Y. Zhang, B.E. Logan, A new method for water desalination using microbial desalination cells. Environ. Sci. Technol. 43 (2009) 7148-7152. [6] Y. Kim, B.E. Logan, Microbial desalination cells for energy production and desalination. Desalination 308 (2013) 122-130. [7] M. Elimelech, W. Phillip, The future of seawater desalination: energy, technology, and the environment. Science 333 (2011) 712–717. [8] B.E. Logan, K. Rabaey, Conversion of wastes into bioelectricity and chemicals by using microbial electrochemical technologies. Science 337(2012), 686-690. [9] M. Mehanna, P.D. Kiely, D.F. Call, B.E. Logan, Microbial electrodialysis cell for simultaneous water desalination and hydrogen gas production, Environ. Sci. Technol. 44 (2010) 9578–9583. [10] K.S. Jacobson, D.M. Drew, Z. He, Efficient salt removal in a continuously operated upflow microbial desalination cell with an air cathode, Bioresour. Technol. 102 (2011) 376–380. [11] K. Guo, S. Freguia, P.G. Dennis, X. Chen, B.C. Donose, J.J. Keller, J. Gooding, K. Rabaey, Effects of surface charge and hydrophobicity on anodic biofilm formation, community composition, and current generation in bioelectrochemical systems. Environ. Sci. Technol. 47 (2013) 7563-7570. [12] X. Wang, S.A. Cheng, Y.J. Feng, M.D. Merrill, T. Saito, B.E. Logan, Use of carbon mesh anodes and the effect of different pretreatment methods on power 11

production in microbial fuel cells. Environ. Sci. Technol. 43 (2009) 6870−6874. [13] X.H. Tang, K. Guo, H.R. Li, Z.W. Du, J.L. Tian, Electrochemical treatment of graphite to enhance electron transfer from bacteria to electrodes. Bioresour. Technol. 102 (2011) 3558−3560. [14] S. Nambiar, C.A. Togo, J.L. Limson, Application of multiwalled carbon nanotubes to enhance anodic performance of an Enterobacter cloacae-based fuel cell. Afr. J. Biotechnol. 8 (2009) 6927−6932. [15] F. Qian, H.Y. Wang, Y.C. Ling, G.M. Wang, M.P. Thelen, Y. Li, Photoenhanced electrochemical interaction between shewanella and a hematite nanowire photoanode. Nano Lett. 14 (2014) 3688-3693. [16] C. Liu, J.J. Gallagher, K.K. Sakimoto, E.M. Nichols, C.J. Chang, Y.M.C. Chang, P.D. Yang, Nanowire−bacteria hybrids for unassisted solar carbon dioxide fixation to value-added chemicals. Nano Lett. 15 (2015) 3634-3639. [17] G.L. Zang, G.P. Sheng, C. Shi, Y.K. Wang, W.W. Li, H.Q. Yu, A bio-photoelectrochemical

cell

with

a

MoS3-modified

silicon

nanowire

photocathode for hydrogen and electricity production. Energy. Environ. Sci. 7 (2014) 3033-3039. [18] R. Morrish, M. Rahman, J.M.D. MacElroy, C.A. Wolden, Activation of hematite nanorod arrays for photoelectrochemical water splitting. ChemSusChem 4 (2011) 474-479. [19] X.H. Tang, K. Guo, H.R. Li, Z.W. Du, J.L. Tian, Electrochemical treatment of graphite to enhance electron transfer from bacteria to electrodes. Bioresource Technol. 102 (2011) 3558-3560. [20] L. Fu, H.M. Yu, Y.K. Li, C.K. Zhang, X.Y. Wang, Z.G. Shao, B.L. Yi, Ethylene glycol adjusted nanorod hematite film for active photoelectrochemical water splitting. Phys. Chem. Chem. Phys. 16 (2014) 4284-4290. [21] X.Q. Zhang, H.J. Feng, Y.X. Liang, Z.Q. Zhao, Y.Y. Long, Y. Fang, M.Z. Wang, J. Yin, D.S. Shen, Electrical stimulation improves microbial salinity resistance and organofluorine removal in bioelectrochemical systems. Appl. Microbiol. Biotechnol. 11 (2015) 3737-3744. 12

[22] Y. Du, Y.J. Feng, Y.Q. Qu, J. Liu, N.Q Ren, H. Liu, Electricity generation and pollutant degradation using a novel biocathode coupled photoelectrochemical cell. Environ. Sci. Technol. 48 (2014) 7634−7641. [23] H. Zarrin, S. Farhad, F. Hamdullahpur, V. Chabot, A. Yua, M. Fowlera, Z.W. Chen, Effects of diffusive charge transfer and salt concentration gradient in electrolyte on Li-ion battery energy and power densities. Electrochim. Acta 125 (2014) 117-123. [24] J.M. Ochando-Pulido, V. Verardo, A. Segura-Carretero, A. Martinez-Ferez, Analysis of the concentration polarization and fouling dynamic resistances under reverse osmosis membrane treatment of olive mill wastewater. J. Ind. Eng. Chem. 31 (2015) 132-144.

13

Figure captions Fig. 1 Desalination mechanism of PMDC. First, the oxidation of organics by the microorganisms occurs spontaneously in the anode chamber. Second, the electrons are continuously attracted to the anode by the impetus from the electron holes and are then transferred to cathode. The bacterial extracellular electron transfer capacity is efficiency intensive. Finally, the water in the middle chamber is efficiently desalinated under the impetus of the inflow current.

14

Fig. 2 (a) Current density and (b) change in solution concentration for three-chamber MDCs with different initial salt concentration in the middle chamber (n=3).

-2

-1

-1

20 g L

40 g L

PMDC MDC PMDC without bacteria

Salt solution replacement

10 8 6 4 2 0

b 40 -1

12

Current density(A m )

-1

Concentration (g L )

-1

5g L 10g L

a

30

start end (PMDC) end (MDC) end (open circuit) end (PMDC without bacterial)

20

10

0

0

50

100

150

200

Time(h) −

15

5gL

-1

10 g L

-1

20 g L

-1

40 g L

-1

Fig. 3 (a) The transient current responses to the ON–OFF illumination cycle. (b) Cyclic voltammetric (5 mV/s) response of the anode in fresh culture. The CVs of PMDC were obtained starting from the second cycle. The third cycle was obtained for switching off of the light and the fourth cycle was obtained in light again.

(c)

Electrochemical impedance spectroscopy of the anode. The EIS of PMDC was measured for three times. First test measured in the light, the next immediate second test was measured in the dark and the last test measured in light again.

a -2

Current density(A m )

8 6 PMDC MDC

4 2 0

0

100

200 300 Time (min)

7.5

MDC PMDC in light PMDC in dark PMDC in light again

6.0

-2

Current density(A m )

b

400

4.5 3.0 1.5 0.0

ca. open circuit -1.5 -0.6

-0.4

-0.2

0.0

Ewe (V vs Ag/AgCl)

c 150

PMDC PMDC in dark PMDC in light again MDC

-Z"(Ω)

120 90 60 30 0

0

30

60

90

120

150

-Z'(Ω)

16

Fig. 4 (a) EIS at the start and end of a batch cycle with different initial salt concentration. (b) NaCl removal compared with total electrons harvested during one batch cycle for different initial salt concentrations. 120

100

start

end

-1

5gL -1 10 g L -1 20 g L -1 40 g L

-Z"(Ω)

80

60

40

20

0

40

60

120

160

180

200

-Z'(Ω)

4000

NaCl removal expressed as coulomb(C)

140

PMDC MDC

3500 3000 2500 2000 1500 1000 500 0

0

500

1000

1500

2000

2500

3000

3500

Electron harvested(C)

17

4000