A novel small scale Microbial Fuel Cell design for increased electricity generation and waste water treatment

A novel small scale Microbial Fuel Cell design for increased electricity generation and waste water treatment
of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  A novel small scale Microbial Fuel Cell design forincreased electricity generation and waste watertreatment  George Papaharalabos  a , John Greenman  b , Chris Melhuish  a ,Ioannis Ieropoulos  a, *  a Bristol BioEnergy Centre, Bristol Robotics Laboratory, University of the West of England, Bristol BS16 1QY, UK b Faculty of Applied Sciences, University of the West of England, Bristol, UK a r t i c l e i n f o Article history: Received 10 April 2014Received in revised form20 October 2014Accepted 19 January 2015Available online 21 February 2015 Keywords: Small scale microbial fuel cellsRapid prototype materialsUrineTwist n '  Play designCOD removal a b s t r a c t Microbial Fuel Cells (MFCs) are a sustainable energy technology with minimal carbonfootprint, which is promising for wastewater remediation and generation of usefulamounts of electricity. This study focuses on the architecture and rapid prototyping ma-terials used for building MFCs and their effect on overall performance. Three MFC variantsof the same design were constructed using ABS, PC-ISO and RC25 materials and werecompared with an established MFC design. MFCs were assessed in terms of power pro-duction and COD reduction both individually and when connected electrically in parallel.In all cases the new design showed a better power output and COD removal. The order of performance in terms of power production and COD reduction for individual MFCs was PC-ISO, RC25 and ABS. However when triplets of the same materials were joined electricallytogether, then the order was different with RC25 outperforming ABS and PC-ISO, whichwas dependent on the materials '  properties. It is concluded that the best performing in-dividual MFC may not necessarily result in the best performing stack.Copyright  ©  2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rightsreserved. Introduction MFC technology Microbial Fuel Cells (MFCs) are bio-electrochemical deviceswhose constituent electro-active bacteria harvest electronsand protons by oxidising organic matter. Electrons travelthrough the anode to the cathode electrode  via  an externalload, and cations diffuse through a cation exchangemembrane that separates the anode with the cathode. At-mospheric oxygen in the cathode reacts with the incoming electrons and protons to produce water [1]. Miniaturisation of  MFCs has been reported in the literature as a more efficientway of generating electricity [2] and can be utilised for pow-ering small devices [3]. To date, the highest volumetric power density of miniaturised MFCs reported, is 667  m W/cm 3 [4].However this is still 1000-fold lower than that of lithium-ionbatteries (7.2    10 7 e 2.16    10 8 W/m 3 , with a theoretical den-sity of 3000 kg/m 3 ) [5,6]. MFCs can nevertheless generate This article was srcinally submitted as part of the Special Section for the 5th edition of the  “ European Fuel Cell Technology  &  Ap-plications Piero Lunghi Conference  &  Exhibition ”  (EFC13), Rome, Italy, December 11 e 13, 2013 [Int J Hydrogen Energy Vol. 39, Issue 36]. *  Corresponding author . Tel.:  þ 44 (0) 1173286318,  þ 44 (0) 1173286322; fax:  þ 44 (0) 1173283960.E-mail address: (I. Ieropoulos).  Available online at ScienceDirect  journal homepage: international journal of hydrogen energy 40 (2015) 4263 e 4268  ©  2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.  power from waste and can be maintained continuously. Inaddition, during the last decade, the MFC technology hasimproved in terms of current density by over a 1000-fold [6,7]. Physical and biochemical advantages of small-sized MFCs Small scale MFCs benefit from lower activation losses, andhighersubstrateutilisation(masstransfer),duetoadecreaseddiffusion resistance, which lowers the overall internal resis-tance[8,9].Ineffect,thereisimprovedpercolationoffeedstock in the anode, which provides the anodophilic biofilm with ahigher supply rate of fuel. Furthermore, shorter molecularpaths allow better diffusion of protons in the biofilm, whichgreatly enhances pH buffering  [10]. These effects improve the reaction kinetics of the anodophiles, which lead to higherpower densities. Decreased paths also affect the distance thatelectrons need to travel from the microbial source (planktonicand biofilm) to the external circuit, which also decreases theinternal resistance [11]. Various types of MFC reactors have been developed from MFC research groups, including minia-ture, cylindrical, up flow, and stacked reactors [12]. For example, high power densities have been reported in small,single-chamber MFC reactors (6.25 mL) utilising catalyst-freecarbon fibre veil electrodes, generating 0.44 W/m 3 [13]. Manufacturing materials of small MFCs To date, many studies have focussed on the improvement of electrodematerialsorthereactor ' sarchitecture[2,13 e 16]. Themajority of MFC publications, involves MFC casings made outof glass (borosilicate) or Plexiglas (polymethyl-methacrylate,PMMA or Perspex) [12,17 e 19]. However, little is known aboutthe selection of manufacturing materials used to build MFCreactors and only a few studies report on MFC reactors madefrom various thermo-polymer plastics (polypropylene, poly-carbonate, nanocure and acrylonitrile-butadiene-styrene)[20 e 23] where their effect on performance is examined.Thus, the main reason for experimenting with plastic poly-mers is the structural properties of thermoplastics, whichcombinedwith3Dprinting,canproduceintricatedesignsforavariety of shapes and sizes.Thisstudy focusesonthestructuralmaterialsaswell asthereactor architecture of a novel MFC design (Twist n '  Play) andemphasises the importance of criteria such as overall biocom-patibility, selective electrical insulation and oxygen diffusion(low for anode; high for cathode) for high power output. Materials and methods Twist n '  Play  MFC-chamber design and fabrication Improvements compared to the EcoBot-III MFC casing In order to evaluate the performance standards of the newdesignandthematerialsinvolved,a directcomparison with analready proven small-sized MFC [23] of the same internal vol-ume was performed. For this reason three RC25 Nanocure typeEcoBot-IIIMFCswereusedascontrolssoastoexaminewhetherthe architecture of the new design MFC or the materialsinvolvedinthemaking,canimprovetheperformance.Thenew‘ Twist n '   Play ’  MFC casing was designed using SolidWorks edu-cationEdition2010SP5.0software(DassaultSystemes,US)andmaintained the same internal reactor volume (6.25 mL) as thefully tested model (EcoBot-III MFC) developed in 2008 andfinalised in 2010 [13,23]. The improvements on the new MFC design comprised the following features:I. Same internal volume with smaller external footprint(less building and support material).II. Simple assembly without fixtures such as screws, clipsor clamps.III. Minimised exposure of the anode chamber to atmo-spheric O 2 . 3D printing of the new design In-house rapid prototyping facilities were employed to fabri-cate the new MFC in three different thermoplastics: PC-ISO(medical-grade biocompatible Polycarbonate; Laserlines, UK),ABS (Acrylonitrile Butadiene Styrene; Laserlines, UK) andRC25 Nanocure (ceramic-filled photo curable resin; envi-sionTEC GmbH, Germany). MFCs made out of RC25 Nanocurewere fabricated with a Stereolithography 3D printer, whereasPC and ABS MFCs were produced using the method of FusedDeposition Modelling (FDM Titan/Dimension BST, Laserlines,UK). Due to the hygroscopic nature of ABS, parts were coatedwith a layer of methyl-ethyl-ketone (Sigma-Aldrich, UK) so asto render the units watertight.Theselectionoftheabovematerialswasbasedonpreviousstudies [2,3,13,20,22 e 26], in which they were considered to bethe most common thermoplastics used for rapid prototyping and also they could be fabricated in-house. MFC operation and monitoring Inoculation and fuel supply Triplicates of single-chamber air-breathing MFCs wereassembled for each material; both the anode and the cathodeelectrodes employed catalyst-free carbon fibre veil sheets30g/m 2 .Sheetsof11  14cm(155cm 2 ,totalsurfacearea)werefoldeddownfivetimessoastoforma1.8cm  2.9cm  1.0cmcuboid. Titanium wire was used as a currentcollector,piercedthrough the carbon veil cuboids. A cation exchange mem-brane (CMI-7000, Membranes International Inc., NJ, USA) witha surface area of 12 cm 2 was placed between two siliconrubbergaskets(Fig.1), separatingtheanodefromthecathode.The MFCs were initially inoculated in batch-mode with acti-vatedsewagesludge(WessexWater,Saltford,UK).AnolytepHwas 7.3 and replenished every 24 h with 1 mL of TYE (1%Tryptone, 0.5% Yeast extract, Fisher Scientific, UK) for the first14 daysoftheexperiment. A2.7 kOhmresistorwasconnectedto each MFC during this period for selecting an anodophilicbacterial consortium; this value was selected in order to bestmatch the Rint. Following 15 days from inoculation, thefeedstock was replaced with fresh non-treated human urineandallMFCswereconnectedtoa24-channelperistalticpump(Watson Marlow, UK) for continuous flow, at a rate of 1 mL/hcorresponding to a hydraulic retention time of 6.8 h. Sampleswere received on a daily basis at fixed time of the day from ahealthy individual. Measured pH on fresh urine samples international journal of hydrogen energy 40 (2015) 4263 e 4268 4264  ranged between 5.5 and 5.8, conductivity was 38 mS/cmaverage, and the mean COD value from a sampling period of 20 days was 15.5 g/L. All experiments were conducted underan ambient temperature of 22  C  ±  2. Power curves and data collection IndividualMFCandoverallstack(triplet)voltagewasrecordedusing an HP Agilent multiplex logging unit (34907A, HP). Theperformance ofindividualMFCsandtripletswasmeasuredbyapplying a range of 50 resistance values from 30 kOhms downto 3 Ohms every 3 min using an automated variable resistor[27]. Voltage was measured in volts (V), and current (uA) wascalculated according to Ohm ' s law, I  ¼  V/R. Power in micro-watts was subsequently calculated from P  ¼  V*I. Power den-sity(P D )wascalculatedbydividingtheabsolutepower( m W)bythe total electrode surfacearea ( a ¼ 155 cm 2 ) and expressed insquare-metres (m 2 ). Similarly, the current was divided by theelectrode ' s  a  so as to estimate the current density (C D ).Recorded data were processed into detailed graphs using GraphPad Prism ® version 5.01 software package (GraphPad,San Diego, California, U.S.A.). Chemical oxygen demand (COD) For measuring the Chemical Oxygen Demand of fresh urine,high range (0 e 20,000 mg/L) potassium dichromate oxidationmethod (CAMLAB, UK) was used and COD values were calcu-lated via colorimetric analysis (Photometer-System MD200,Lovibond). Fresh urine samples (200  m L) were filtered (0.20  m m,Minisart ® ) and COD content measured prior to entering andafter exiting the MFCs (24 e 48 h). Treated urine samples wereinitially filtered and then centrifuged to 500 g for 5 min. Results  &  discussion Performance from individual new design MFCs Thematerialsselectedinthisstudywerebasedon3Dprinting suitability and use in MFC work, as well as biocompatibilitywith respect to microbial toxicity. Specific toxicity analyseswere not performed however, implicit information could stillbe drawn from the MFC performance levels; possible toxiceffectsfromthestructuralmaterial,wouldhavedetrimentallyaffected the MFC performance.Results shown are from 20 days (D20) and 40 days (D40)after inoculation, and there is clear evidence in terms of improvement in performance due to maturity. The 20-daypoint was chosen as a point in the transitory periodfollowing urine addition as the fuel, and the 40-day point waschosen as an exemplar of steady state conditions (Fig. 2).Polarisation runs on individual units 20 days after inoculation(Fig. 3A) showed that the new design MFCs outperformed thecontrol units in all three different material cases by amaximum of 74% in terms of power. The control MFCs pro-duced a maximum power transfer (MPT) of 31  m W (2 mW/m 2 )at 121  m A (8.1 mA/m 2 ). MFCs made of PC-ISO showed thehighest power and current generation amongst the differentmaterials with values of 54  m W (3.6 mW/m 2 ) and 136  m A(9.1 mA/m 2 ) which was 74% and 12% higher than the controlvaluesrespectively.ThesecondbestperformingMFCmaterialwas the RC25 Nanocure, reaching 44  m W (2.6 mW/m 2 ) and136  m A (9.1 mA/m 2 ) which was 42% higher power and 12%higher current compared to the control. MFCs built with ABSshowed also 16% higher power generating 36  m W (2.4 mW/m 2 )and a 7% increase in current to 130  m A (8.7 mA/m 2 ). To assessperformance after a further period of maturity, polarisationexperiments were carried out for individual units 40 days  post inoculation, so as to examine establishment of the electro-active biofilm (Fig. 3B). With respect to the maturing be-tween the D20 and D40 period, the control MFCs powerimproved to 50  m W (3.3 mW/m 2 ) and the current output to210  m A (14 mA/m 2 ), which resulted in an increase of 61% and73%respectivelyattheendofthe40dayperiod.Asintheearlystages, a similar order in performance was displayed with PC-ISO MFCs increasing their MPT by 22% to 66  m W (4.4 mW/m 2 )and the average current by 160% to 354  m A (23.6 mA/m 2 ); RC25Nanocure MFCs increasedtheir power outputby 29% to 57  m W(3.8 mW/m 2 ) andthecurrentoutputto 329  m A(21.9mA/m 2 ) an Fig. 1  e  (Left)  Twist n '  Play  MFC design. (Right) EcoBot-IIIcontrol MFC design assembly made from RC25 Nanocure.Fig. 2 e Currentgeneration during20 days and 40 days  post  inoculation as an indication of biofilm maturity and theplateau phase near the end of this period. international journal of hydrogen energy 40 (2015) 4263 e 4268  4265  increase of 141%; ABS units produced a power of 50  m W(3.3 mW/m 2 ) and a current of 218  m A (14.5 mA/m 2 ) whichstandsforanincreaseof38%and67%respectively.Intermsof performance differences amongst the examined MFCscompared to the control 40 days after inoculation, the PC-ISO,theRC25Nanocureexhibitedanincreasein powerby 32%and14% respectively. The current was also higher by 68% (PC-ISO)and 56% (RC25). In the case of ABS, the power was similar tothe control and the current showed only an increase of 4%.Power sweeps performed in this extended period after theinoculation presented overshoots peaks in the graph for allthe examined materials and designs. This could suggest thatthe biofilm had yet to reach the maturity stage or the resis-tance value intervals rate were too large and fast for the bio-filmto establisha steadystate[28]. InbothmaturingstagesallMFCsdisplayedasimilarinternalresistanceof2kOhm,wheremaximum power transfer was achieved.These results are consistent with findings from a previousstudy by Ledezma et al. (2010) that compared ABS, RC25 andPC-ISO as structural materials for a larger size (25 mL) anddifferent architecture dual-chamber MFC [20]. Moreover, theincreased outputs from the new  Twist n '   Play  design whencompared to the control EcoBot-III MFC casing built in 2008,suggest that the improvements made on the design led to amore functional anode chamber, which allowed for betterbiofilm establishment conditions and electricity generation. MFCs fluidically connected Parallel electrical connection (  n  ¼  3  ) In this part of the experiments triplets made from the samematerial were connected electrically in parallel, with fluidicconnection between the units. Series connection was notattempted in this study, due to the inherent fluidic conduc-tance (from MFC to MFC), which would short-circuit the MFCsand would thus act adversely, under these conditions [21].Proper fluidic isolation is part of the immediate next steps,which will allow performing a fuller investigation, including series connection.Polarisation results showed a variance in performancebased on the output from individual MFCs during D40 period.The control MFC triplet reached 129  m W (2.8 mW/m 2 ) at acurrent of 0.84 mA, (18.3 mA/m 2 ) (Fig. 4). The RC25 Nanocuretriplet produced 203  m W (4.4 mW/m 2 ) and a current of 1.3 mA(29 mA/m 2 ). ABS triplet generated 152  m W (3.3 mW/m 2 ) and0.66 mA (14.3 mA/m 2 ). The PC-ISO MFC underperformedcompared to the other new design MFCs generating 133  m W(2.9 mW/m 2 ) and 0.64 mA (13.8 mA/m 2 ).With regard to P D  and C D  obtained from individual units,the control EcoBot-III design displayed a 15% decrease in P D but a 30% increase in C D . The parallel-connected MFCs madeof RC25 Nanocure showed a 16% and 38% increase respec-tively. MFCsmadeofABS,maintainedsimilarP D andC D levelsas the individual MFCs. Increases in C D  are to be expected dueto the parallel electrical configuration of the MFCs in thetriplets. On the contrary, PC-ISO new design parallel-connected MFCs decreased by 34% and 40% their P D  and C D .A possible explanation for this reason could be the material ' sintegrity, as it has the highest tensile strength of all materialsbut it possesses a very low endurance to fracture stress [29]especially after 40 days of operation under hydraulic pres-sure in the anode chambers. This led to the appearance of ductiletearingzones(Fig.5)followedbyductilebrittleness[30] on the external surface on all anode PC-ISO casings, thatallowed anolyte to leak, therefore allowing oxygen to pene-trate into the anodes, affecting the overall performance. Fig. 3 e Powercurvesfrom individual units,(A) 20days and(B) 40 days after inoculation. Data with error bars shownfor n ¼ 4. Fig. 4  e  Performance of MFCs when connected in parallel. international journal of hydrogen energy 40 (2015) 4263 e 4268 4266  The in parallel configuration pushed the system collec-tively to a more optimum equilibrium which resulted in adecreased internal resistance for all tested MFCs and showeda uniform value of 105  U  in all cases of 3 MFCs connected inparallel, which is equivalent to a 315 U for an individual MFC,whereas the internal resistance from individual units in thebeginning of the experiment was 10-times higher. ControlMFCs exhibited a 305 U internal resistance which stands for atheoretical of 915  U  in an individual unit which equals a 2.2-folddecreasecomparedtotheirresistanceatthestartoftheof this study.Again the newly designed MFC proved to be superior overthe control MFC architecture with the exception of the PC-ISOand an overall comparison between the  Twist n '   Play  and thecontrol RC25 Nanocure highlighted the improvement of thenew design. COD reductionIndividual units.  The new MFC design was tested in terms of chemical oxygen demand (COD) reduction, both as a stand-alone unit and when in stack operation configured electri-cally in parallel. For the sake of consistency the srcinal CODvalue of the urine sample used for all the individual units was16.8g/LonthedayoftheCODmeasurement.Thecontrolunitsreduced this value by 29% (11.9 g/L). MFCs made of PC-ISOachieved an average of 44% COD removal down to 9.4 g/L.RC25 Nanocure units reduced the organic content by 37% to10.5 g/L. The mean COD treatment from ABS units resulted in36%reductionoforganiccontentoftheoriginalurineto10.7g/L. In terms of the MFCs running as individual units, therecorded COD removal was proportional to the order of powerperformance. It could therefore be suggested that the MFCswith the best performance characteristics are expected toachieve the highest COD removal within a number of MFCsdisplaying various outputs. Stacked MFCs in parallel (n  ¼  3).  COD values were recordedwhen MFCs of the same material were connected in parallel.As it would be expected, COD remediation was increasedwhen more elements were connected fluidically and config-ured electrically in parallel [31]. In this case the initial CODcontent found in the supplied urine sample during the CODexperiment was 19.4 g/L. The control MFC as in the individualtest, showed the lowest range of removal, decreasing the CODvalue by 38% (7.4 g/L) to a level of 12 g/L. Again the PC-ISOtriplet removed a total of 53% (10.3 g/L) which was thehighest amongst the different materials. The COD value fromtheRC25Nanocurestackwasdecreasedby 46%leaving10.6g/L of COD in the treated sample. The ABS stack showed asimilar COD treatment with the RC25, reducing the organiccontent by 45% down to a value of 10.7 g/L. It is important topoint out, that the twist and play MFC was primarily designedto remain watertight and minimize oxygen presence in theanodic chamber when stacked in continuous flow mode,which was not a feature of the control (EcoBot) MFC design. Inall material cases besides the PC-ISO, the stacked MFCstreated the organic content of urine in line with the poweroutputs. Nonetheless, the high COD removal value from thePC-ISO was not in accordance with the power output, whichcould be related with the material ' s structural failure. It isexpected that the failure resulted in extensive leakages andreduction/dilution of the samples, which probably affectedthe end-treatment product from the stack. Based on thismaterial bottleneck, more experiments should take place toconfirm the COD removal efficiency. Conclusions In principle, an MFC ' s performance is affected by many vari-ables, and behaviour of one factor may be directly influencedby other parts of the MFC. A common feature in the processthat affects electricity production and COD removal is largelydependent on the materials used. It could be argued that thematerials and conditions optimised for one type of MFC arenot necessarily optimal for other MFC types and structures.Several challenges need to be resolved, including high arealresistivity, high oxygen leakage and non-compatibility withmicro-fabrication.This study showed how three versions of a novel MFCreactor can greatly affect their overall performance of MFCsunderfluidicandelectricalscenarios. Thenewdesignshowedan overall increased performance compared to the controlMFC reactor and depending on the electrical and the fluidicconnection the RC25 Nanocure and the ABS seemed toperform better in terms of power and COD treatment of up to15% increase whilst retaining better power and COD valuesthan the control.Even for rapid prototype materials, which are expected tobe of a finite lifetime, RC25 Nanocure proved to be the mostrobust.TheseresultsdemonstratedthattheRPtechnologyisauseful tool for examining various materials as structural Fig. 5  e  Characteristics of (A) the ABS with the porous surface, (B) the RC25 Nanocure with the ceramic integral ruffledsurface and (C) the PC-ISO with a cavity highlighted. Micrometre reference: 200  m m. international journal of hydrogen energy 40 (2015) 4263 e 4268  4267
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks