Research

Carbon and Steel Surfaces Modifiedby Leptothrix discophora SP-6:Characterization and ImplicationsT U A N A N H N G U Y E N , † Y U Z H U O L U , † , ‡

X I N G H O N G Y A N G , § A N DX I A N M I N G S H I * , †

Corrosion, Electrochemistry & Analysis Laboratory (CEAL),Western Transportation Institute, PO Box 174250, College ofEngineering, Montana State University, Bozeman,Montana 59717-4250, School of Materials Science andEngineering, Tianjin University, Tianjin 300072, P.R. China,and Veterinary Molecular Biology Department, College ofAgriculture, 960 Technology Boulevard, Montana StateUniversity, Bozeman, Montana 59718-4000

Received May 19, 2007. Revised manuscript received August20, 2007. Accepted September 19, 2007.

Leptothrix discophora SP-6, a type of manganese(Mn)-oxidizing bacteria, has been known to accumulate Mn oxidesfrom the aqueous environment and thus play a key role inmicrobiologically influenced corrosion by increasing theelectrochemical potential of steel and other metals. Similarly,this bacterium was found to modify the surface of glassy carbonin aqueous solution and increase its potential (i.e., ennoblement).In the latter case, biomineralized Mn oxides can be usedas cathodic reactants for a new generation of microbial fuelcells featuring a biocathode. In this preliminary study, factorsaffecting the biofilm formation and biomineralization processeswere examined. The inflow of air into the culture mediumwas found essential to sustain the ennoblement of substrateelectrodes. The OCP and FESEM/EDS data indicated that asmoother initial substrate surface generally led to betterennoblement. Polarizing the carbon electrode at +500 mVSCEfor 15 min was found to facilitate the ennoblement on carbonelectrodes, and so did coating it with a poly(L-lysine) layer.Independent of substrate material, initial surface roughness, andpretreatment, there were three parameters in the EIS equivalentcircuit that correlated well with the OCP indicating the levelof ennoblement by L. discophora SP-6, i.e., electrolyte resistance,double-layer capacitance, and low-frequencies capacitance.These fascinating findings merit further investigation as they mayshed light on the fundamental bacteria/substrate interactionsand help advance the knowledge base needed for the engineeringapplications.

IntroductionMicrobial fuel cells (MFCs) present an attractive pathway

to directly convert chemical energy into electrical energyand have been extensively investigated in the past severaldecades. Numerous studies have focused on microorganisms

to work as a biocatalyst for the anodic reaction of the MFC,generating electricity via degradation of organic matter. Untilrecently, biocathode MFCs have been a relatively uncharteredterritory.

The microbial colonization of passive metals has beenreported to increase their open circuit potential (OCP) tofinal values between +200 mVSCE (1) and +450 mVSCE (2, 3)through a series of reactions collectively termed ennoblement.Initial studies have explained ennoblement by a wide varietyof mechanisms (4): microbially generated protons near thesurface (5); microbially generated hydrogen peroxide, pos-sibly combined with low pH (6); microbially producedorganometallic catalysts of oxygen reduction (1, 3); specificenzymes (7); and passivating siderophores (8). A number ofstudies have attempted to establish a direct link betweenbiofilm formation and ennoblement. Ennobled potential hasbeen correlated with cell density (9) and with biologicalactivities in the biofilms by measuring the accumulation ofadenosin triphosphate (ATP) (10) or electron transport activityand lipopolysaccharide content (11). Nonetheless, the struc-tural heterogeneity of the biofilm may also affect its activitiesand mass transfer dynamics, and thus the OCP of its substrate(12). Recent studies have linked the ennoblement of passivemetals with the activity of manganese-oxidizing bacteria(MOB), such as the freshwater Gram-negative Leptothrixdiscophora (4, 13).

As an essential micronutrient for all living organisms,manganese (Mn) is Earth’s second most abundant transitionmetal, next to iron (14). In natural waters, soluble Mn(II) canreach up to millimolar concentrations, even in the presenceof oxygen (14), and biological catalysis has been wellestablished as the dominant mechanism of oxidizing Mn(II)to insoluble Mn(III, IV) oxides in circumneutral freshwater(15, 16). It was proposed that the microorganisms first oxidizeMn(II) to a soluble or enzyme-complexed Mn(III) intermedi-ate, then subsequently to Mn(IV) oxides primarily a phyl-lomanganate most similar to δ-MnO2 or acid birnessite (14).A reasonable reaction concomitant with the bacterial Mn(II)oxidation reactions is the reduction of oxygen to water asshown in eq 1.

O2 + 4H+ + 4e- f 2H2O (1)

Biomineralized manganese oxides can be used as cathodicreactants significantly superior to oxygen. First, they are someof the strongest oxidizing agents (next to oxygen) found innature. In many natural waters, microbial activities lead tothe formation of particulate Mn oxides with average oxidationstates exceeding 3.4 (14). Second, the reduction of manganeseoxides deposited by L. discophora increased the potential ofa MFC by approximately 300 mV and delivered a currentdensity up to 2 orders of magnitude higher, compared withthose reached with the reduction of oxygen (17). Moreimportantly, manganese oxides are in direct contact withthe electrode, where the cathodic reaction is thus not limitedby the mass transfer process of cathodic reactants. Finally,manganese oxides are renewable. When used as a cathodefor the MFC, manganese oxides are reduced to Mn(II), whichin return is microbially reoxidized. This sequence of eventsproduces renewable cathodic reactants, MnOOH and MnO2,in direct contact with the electrode (4).

Mn(II)-oxidizing microorganisms, primarily bacteria andfungi, are ubiquitous in nature and play an integral role inthe biogeochemical cycling of manganese, iron, nitrogen,carbon, sulfur, and trace metals (18). They accelerate the

* Corresponding author phone: 1-406-994-6486; fax: 1-406-994-1697; e-mail: Xianming_s@coe.montana.edu.

† College of Engineering, Montana State University.‡ Tianjin University.§ College of Agriculture, Montana State University.

10.1021/es071178p CCC: $37.00 2007 American Chemical Society VOL. 41, NO. 23, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7987

Published on Web 10/26/2007

rate of Mn biomineralization several orders of magnitudefaster than either abiotic catalysis on mineral surfaces orhomogeneous oxygenation in aqueous solution (18). Thiswas demonstrated by in situ immersion of 316L stainlesssteel into a freshwater creek, where the OCP of steel wasincreased by 350 mV in 30 days resulting from Mn biom-ineralization (19).

MOB are likely to survive cold temperatures or otherrelatively harsh conditions in the field environment. Forinstance, the “Wild-type” L. discophora are mostly the sheath-forming strain SP-6, which can be maintained indefinitelyin slow growing conditions at temperatures between 20 and25 °C. For culture management, it can be stored at 4 and-80°C (20).

Therefore, it is envisioned that the ennoblement phe-nomena can be utilized for a new generation of MFCsfeaturing a biocathode that involves L. discophora or otherMOB. Such MFCs that harness the native population ofbacteria abundant in natural waters could potentially bemade self-sufficient with simply the pretreated electrodesand nutrients continuously available in the environment. Assuch, they would be ideal for a host of autonomousapplications that demand a minimum amount of mainte-nance (e.g., monitoring water quality in lakes and streams),which is the focus of an ongoing research project funded bythe National Academies. Figure 1 illustrates the conceptualdesign of the novel MFC, consisting of a sacrificial anode ofaluminum alloy and a cathode of porous graphite coveredby manganese dioxide and L. discophora biofilms. Unpub-lished results in our laboratory demonstrated the feasibilityof this concept at the laboratory-bench scale.

Within this context, it is crucial to investigate how variousfactors would influence the ennoblement of surfaces by L.discophora and thus the performance of novel MFCs. Theultimate goal is to maximize the energy output and reliabilityof a biocathode utilizing the L. discophora biofilm. Forinstance, initial surface roughness is a factor that can becontrolled to optimize how well the bacteria initially attachto the substrate and form the biofilm and to minimize thebiofilm detachment as well. Cell density and dissolved oxygenconcentration are two factors that may evolve during theennoblement and affect the adhesion, growth, and biom-ineralization processes of L. discophora.

Biominerals on water-immersed metal surfaces also playa key role in microbiologically influenced corrosion (MIC)and the biomineralization by L. discophora and related metal/microbe interactions have been mostly examined in thiscontext (4, 19, 21, 22). Shi et al. studied the Mn biominer-alization by L. discophora SP-6 (4) and argued that thefollowing electrochemical half-reactions are responsible forthe ennoblement of the electrochemical potential of 316Lstainless steel and a low-iron titanium alloy (Ti–6Al–4V):

MnO2 +H++ e-fMnOOH E0 ) + 0.81 VSCE (2)

MnOOH+ 3H++ e-fMn2++ 2H2O E0 ) + 1.26 VSCE

(3)

With the overall reaction shown in eq 4, the half-cellequilibrium potential of a noble cathode (such as graphite)covered with biomineralized Mn oxides could reach +0.36VSCE:

MnO2 + 4H++ 2e-fMn2++ 2H2O E0 ) + 1.28VSCE,

i.e., E′pH)7.2 ) + 0.36 VSCE(4)

The reversal of eq 4, i.e., the Mn(II) oxidation to Mn(IV)oxide, can be balanced with the oxygen reduction shown ineq 1 and lead to the following reaction:

Mn2++ 1/2O2 +H2OfMnO2 + 2H+ (5)

In addition to the practical applications such as MICcontrol and MFCs, the ennoblement by L. discophora is alsoa phenomenon of intellectual importance. The mechanisticdetails of biofilm formation and biomineralization by MOBare still not fully unraveled. A recent review of three MOBspecies suggested that in all of them proteins containing thehighly conserved Cu-binding ligands characteristic of mul-ticopper oxidases seemed to participate in the oxidationprocess (23). L. discophora has been known to oxidize Mn(II)and Fe(II) during its stationary growth phase (24, 25). Bothc-type hemes and multicopper oxidase are believed to playa role in Mn(II) oxidation, whereas Mn(II)-oxidizing factorsare speculated to be secreted as part of a complex originatingfrom sheath remnants of L. discophora (23).

In this work, we examined the influence of electrochemicalpolarization on the ennoblement of carbon and steel surfacesby L. discophora SP-6. Biomineralization by L. discophorahas a physiological function, which may be supplying energyfor growth, scavenging of harmful oxygen species, or servingas terminal electron acceptor that supports anaerobic growth.The vital processes that bacteria use to capture, store, andmobilize energy involve electron transport and catalysis ofredox reactions and therefore, the hypotheses to be testedinclude the following. (1) Polarizing the substrate materialmay alter the ennoblement processes; and (2) Coating anelectroactive polymer on the substrate material may alterthe ennoblement processes.

Such alternations are expected to occur by changing themetabolic pathways of L. discophora; affecting its growthkinetics; affecting the attachment and detachment behaviorsof L. discophora; changing the pathways of Mn-oxidizationby L. discophora; and/or changing the products of Mn-oxidization by L. discophora. For instance, polarizing theinert substrate cathodically is expected to “pump” electronsto the electrode surface and thus promote the reactionsshown in eq 3 or 4 while hindering the Mn(II) oxidation. Itis also speculated to potentially inhibit the secretion of Mn(II)-oxidizing factors, slowing down the bacterial growth byconsuming available oxygen, accelerating the biofilm agingand detachment, or reducing the existing biogenic oxides.Polarizing the inert substrate anodically, on the other hand,may affect the ennoblement processes in an opposite manner.

FIGURE 1. Schematic of the novel MFC, consisting of a sac-rificial anode of aluminum alloy and a cathode of porousgraphite covered by manganese dioxide and L. discophorabiofilms.

7988 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 23, 2007

Modification of substrate surface by coating with an elec-troactive polymer is speculated to have combined effects ofreducing the initial surface roughness and electrochemicalpolarization of the substrate.

In this preliminary study, we used the electrochemical,microbiological, and surface analytical tools to achieve abetter understanding of the influential factors in the en-noblement of carbon and steel surfaces by L. discophoraSP-6. Such knowledge is essential for us to better control thecapabilities of the bacterium to form a healthy and reliableMn-oxidizing biofilm. The factors investigated include dis-solved oxygen concentration, substrate material (glassycarbon vs carbon steel), initial surface roughness, electro-chemical polarization, and substrate pretreatment by anelectroactive polymer. In addition, the evolution of celldensity during the ennoblement was examined.

Materials and MethodsMaterials and Reactor. Glassy carbon (GC-20SS grade,thickness 3 mm, #800 finish) and 304 stainless steel werepurchased from Tokai Carbon USA Inc. (Hillsboro, OR) andMetal Supermarkets (Boise, ID), respectively. To make testcoupons, both materials were cut from larger sheets intosmall size of about 1 × 1 cm. The unexposed side of eachcoupon was then electrically connected to a copper wire byusing Ag-conductive glue, followed by sealing with epoxyresin (MAS epoxies-FLAG). After the curing of the epoxysealer, the coupons were polished to provide a specific levelof uniform surface roughness. They were wet-sanded withtap water on the metallographic grinding disk covered withsilicon carbide papers of decreasing grit sizes: 200, 400, 600,800, and 1000. After sanding at each grit size, the holder andcoupon were rinsed with running tap water to remove anyremaining grit. The last step in cleaning the surface was tosonicate the coupons in distilled water. The electrodes werethen rinsed with ethanol and mounted in a silicone stopper.

The reactor was of a glass bowl type, 10 cm tall, and 30cm in top-diameter. Eight electrodes were mounted to theirholders, which were attached to the top of the reactor. Plastictubes were mounted in the reactor to aerate the culturemedium using an air pump (Elite 799; with 900 mL/min rate).Pall-Gelman sterile mini-capsule filters were attached to thesetubes to prevent contamination of the reactor. A magneticbar was placed at the bottom of the reactor to provide stirring.The reactor was autoclaved on the dry setting (via depres-surization method) at 123 °C and 1.2 atm for 30 min. Oncecooled to room temperature, the reactor was then sealedwith silicon gel. To study the effect of pumping on the celldensity in culture medium, two variable flow pumps (FisherScientific) were used to pump the culture medium in andout the reactor, respectively.

Prior to the experiment, all electrodes assembly andreference electrode were sterilized by 99% ethanol and then

connected to a computer via an 8-channel multiplexer to theOCP monitoring and electrochemical impendance spec-troscopy (EIS) measurements. The reactor and its accessorieswere then placed in the plastic custom-made glovebox (60cm × 60 cm × 120 cm) that had been sterilized by UV lightfor 24 h. Prior to inoculating the reactor with L. discophora,the coupons were immersed in the culture medium for 24 hto make sure that the OCPs were stabilized.

Biofilm Preparation Procedures. To grow the MOB weprepared 15 L of American type Culture Collection (ATCC)culture medium 1917 MSPV (similar to the one used in ref(4)). The nutrient solution was autoclaved on the liquid settingat 123 °C and 1.2 atm for 25 min. After cooling the solutionto room temperature, we aseptically added 15 mL of filter-sterilized vitamin solution, as required by the MSPV medium(17), 60 mL of filter-sterilized 50 mM manganese sulfatesolution, and 75 mL of filter-sterilized 20% sodium pyruvatesolution. All chemicals were from Fisher Scientific.

The bacterium, L. discophora SP-6 (ATCC 51168) waspurchased from ATCC and stored at -80 °C before use. Toinoculate the reactor, we poured 2 L of the MSPV mediumwith vitamins, sodium pyruvate, and manganese sulfate intoreactor. All the stock bacteria L. discophora SP-6 grown onone agar plate for over 12 h were then added in the aeratedreactor.

Cell Density Evaluation. The cell density in culturemedium was measured by using a spectrophotometerBioState 3 (Thermo Spetronic). The cell density was definedas number of bacteria per milliliter and characterized in termsof UV absorption at 600 nm, i.e., optical density (OD600).

Electrochemical Measurements. Electrochemical mea-surements were conducted using a three-electrode system.The OCP of 8 coupons (working electrodes) were monitoredsimultaneously by using a Gamry Electrochemical MultiplexerECM8, which was also used periodically to measure the EISof each coupon. The counter electrode used was a curvedplatinum wire placed in the middle of the reactor, whereasthe reference electrode was a saturated calomel electrode(SCE) connected to the reactor through a salt bridge. For EISmeasurement, the working electrode was polarized at ( 10mV around its OCP. Frequency was ranging between 10 KHzand 10 mHz, with 10 points per decade. We used the Gamryanalysis software for plotting and fitting the EIS data.

Surface Analyses. Before and after the ennoblementexperiment, surface analytical tools were used to examinethe surface of glassy carbon and stainless steel coupons. Inaddition to visual inspection under a digital microscope, fieldemission scanning electron microscopy/energy-dispersiveX-ray spectrometry (FESEM/EDS) investigation was con-ducted. To analyze the surface of the ennobled coupons, thecoupons were mechanically removed from the holders, air-dried, and kept in a vacuumed desiccator for up to 48 h prior

FIGURE 2. OCP evolution during ennoblement by L. discophora SP-6, as a function of initial surface roughness: (a) on glassy carbonelectrodes; and (b) on stainless steel electrodes.

VOL. 41, NO. 23, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7989

to FESEM/EDS analysis to ensure dryness. The biofilms onthese coupons were preserved by coating with a thin carbonlayer (<50 nm) and analyzed for surface morphology andchemistry. The instrument used for FESEM/EDS investigationwas a new model of Zeiss Supra 55VP PGT/HKL, which hadan ultrahigh resolution at low voltage (1.7 nm at 1 kV) andcould detect the small variations of trace element content.

Results and DiscussionEffect of Pumping of Air and Culture Medium on theEnnoblement. For both glassy carbon and stainless steelelectrodes immersed in the same bioreactor, OCP dataindicated that inflow of air into the culture medium wasessential to sustain the ennoblement of substrate by L.discophora SP-6. This is consistent with existing researchand confirms that the oxygen availability played an integralrole in the biomineralization process by this type of MOB,as indicated in eq 5. In addition, it is reasonable to speculate

that the growth of this aerobic bacterium depended on thedissolved oxygen concentration in the aqueous culturemedium. For both carbon and steel electrodes, OCP droppedsignificantly when the inflow of air halted and came backagain once the pumping resumed. Therefore we employedthe pumping of air (bubbling at the rate of 900 mL/min) intothe culture medium during the ennoblement experiment,except when it was necessary to avoid perturbation duringthe EIS measurements (duration of about 30 min each).

In addition, the pumping of culture medium into and outof the bioreactor had an interesting influence on the celldensity in the bioreactor. This is expected to affect theennoblement as well, since the ennobled potential has beencorrelated with cell density (9). When the bioreactor wasused as a static culture system (with no pumping), theevolution of cell density of L. discophora SP-6 was as follows.The cell density was 2.36 × 107 cells/mL at the second dayof inoculation, and it increased by almost 10 times to 2.28

FIGURE 3. SEM image of glassy carbon electrodes: (a), (b), (c) before ennoblement, with initial surface roughness of # 800 (asreceived), #600, and #200, respectively; and (a2), (b2), (c2) after ennoblement.

7990 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 23, 2007

× 108 cells/mL at the sixth and ninth day, and maintainedat 1.8 × 108 cells/mL after 12 days. This is somewhatcomparable to another study with the same bacterium (13),in which the cell density reached 108 cells/mL after 20 h in

inoculum. This order of cell density (e.g., 108) was also usedto maintain the ennoblement by another bacterium (Thio-bacillus ferroxidant) in ref (26). The increase in cell densityduring the ennoblement without culture medium replenished

FIGURE 4. OCP evolution during ennoblement by L. discophora SP-6, as a function of a 15-min electrochemical polarization of (a)and (c) glassy carbon electrodes; and (b) and (d) stainless steel electrodes.

FIGURE 5. SEM image of ennobled (a) glassy cabon electrode polished to #600, without polarization applied; (b) same as (a), with a15-min polarization at E ) +500mVSCE; (c) stainless steel electrode polished to #1000, without polarization applied; and (d) same as(c), with a 15-min polarization at E ) +100 mVSCE.

VOL. 41, NO. 23, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7991

may be explained by the hypothesis that L. discophora SP-6could grow with Mn(II) as an energy source. It has beenargued that the reaction shown in eq 5 has a standard freeenergy of -70 kJ mol-1 at pH 7 and thus theoretically cansustain bacterial growth, and pathways of electron transportwere proposed in a model to suggest the generation of energy(ATP) as the result of bacterial manganese oxidation (23).

On the other hand, when the bioreactor was used as aflow culture system (medium being pumped into and out ofthe reactor at same rate), the evolution of cell density of L.discophora SP-6 was very different. It was found that thecontinuous pumping reduced the cell density by 1/3 and 1/5on a daily basis, at the pumping rate of 0.5 mL min-1 and0.2 mL min-1, respectively. This indicates that in the systeminvestigated, the growth of L. discophora SP-6 was relativelyslow and the enhanced growth due to the inflow ofreplenishing nutrients was eclipsed by the dilution effect viathe outflow of culture medium. Therefore, we chose to usethe static culture system for the rest of the experiments.

The preliminary findings above have significant implica-tions when it comes to the biocathode MFC utilizing L.discophora or other MOB. Especially when placed in a fieldenvironment (e.g., a stream), the energy output of such aMFC will fluctuate over time, as it depends on the ennoble-ment by the bacteria and thus on the dissolved oxygenconcentration in and flow rate of water. The dissolved oxygenconcentration is a function of numerous factors in the field,such as the temperature and flow rate of water, presence ofcontaminants, and photosynthesis. The flow of water notonly will affect the availability of oxygen and nutrients, butalso may cause the detachment of biofilm or dilute the celldensity. These natural or man-made fluctuations underlinethe necessity of a power management mechanism for theMFC, which is part of our ongoing project.

Effect of Initial Surface Roughness on the Ennoblement.For both glassy carbon and stainless steel electrodes im-mersed in the same bioreactor, OCP data indicated that asmoother initial substrate surface generally led to betterennoblement by L. discophora SP-6. As shown in Figure 2a,in the early stage of ennoblement (20 h after innoculation,i.e., 72 Ks), the OCP of carbon electrode with a #800 finishgrade reached approximately +135 mVSCE, whereas that ofcarbon electrode polished to #600 and #200 was around +25and +15 mVSCE, respectively. For the one polished to #200,the dropping from +75 mV at 50 Ksec indicated a possibledetachment of biofilm from the electrode surface. The OCPof the electrodes exposed to the abiotic culture medium didnot increase over time as in the presence of L. discophora.The OCP of glass carbon electrodes generally increased with

FIGURE 6. OCP evolution during ennoblement by L. discophora SP-6, as a function of PLL coating on carbon electrodes: (a) at lowcell density; (b) at high cell density.

FIGURE 7. (a) OCP evolution of carbon electrode with and without PLL coating, at high cell density during the ennoblement; (b) EISNyquist diagrams of these electrodes after ennoblement.

FIGURE 8. Equivalent circuit used to characterize theinterfaces: Re associated with the electrolyte resistance; CDL,capacitance of double layer; RCT, resistance of charge transfer;W, Warburg impedance (diffusion line); RL and CL areresistance and capacitance at low frequencies (30).

7992 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 23, 2007

time after innoculation, except on the one polished to #200due to possible detachment of biofilm from its electrodesurface. Figure 2b shows that the ennoblement of stainlesssteel electrodes by L. discophora SP-6 was much slower,compared with that of glassy carbon electrodes. Nonetheless,a smoother initial substrate surface also led to betterennoblement. The OCP of stainless steel electrodes generallyincreased with time after innoculation, except on the onepolished to #1000 where the OCP fluctuated significantly inthe first day or so for unknown reasons.

In an open growth system, bacterial adhesion is knownto occur in 4 phases: “the transport of the bacterium to thesurface, the initial adhesion with a reversible and irreversiblestage, the attachment by specific interactions, and finallythe colonization in order to form a biofilm” (27). Contraryto the findings in this research, rough surface has beenreported to promote plaque formation and maturation, andhigh-energy surfaces are known to collect more plaque, tobind the plaque more strongly, and to select specific bacteria(27). Another study confirmed that the surface roughnessvalue and the amount of plaque adhesion decreased withthe increase in polishing level (28). A recent study indicatedthat a rougher surface promoted bacterial adhesion andcolonization; however, there was a certain threshold (below200 nm) below which there was no significant decrease inbacterial adhesion and colonization (29).

The FESEM/EDS data provided a possible explanationfor the effect of initial surface roughness on the ennoblementby L. discophora SP-6. The SEM images illustrated that thebiofilm grown on the smoother carbon electrode surface(Figure 3a2) seemed to be more compact and homogeneousthan those grown on the rougher surfaces (Figure 3b2 and3c2). The EDS data indicated that the biofilm grown on thesmoother carbon electrode surface (e.g., #800 finish grade)featured higher concentration of biominerals (Mn signalstrength of 0.72 wt%) than those grown on the roughersurfaces (e.g., no detectable Mn signal on the electrodepolished to #200 after ennoblement).

Effect of Electrochemical Polarization on the Ennoble-ment. As discussed in the Introduction it was hypothesizedthat polarizing the substrate material might alter the en-noblement processes as the external electric field servedeither as an electron sink or reservoir. To test this hypothesis,a short-duration electrochemical polarization was appliedto glass carbon and stainless steel electrodes in the earlystage of ennoblement by L. discophora SP-6. These electrodeswere polished to the #600 grit size prior to the ennoblement.During the first-day immersion in the bioreactor, a constant(anodic or cathodic) potential was maintained for 15 min onthe electrode, and its OCP was subsequently monitored alongwith that of nonpolarized electrodes.

Figure 4a illustrates the evolution of OCP of carbonelectrodes after the 15-min anodic polarization at+300 mVSCE

and +500 mVSCE, respectively. The OCP data indicated thatthe polarization at +500 mVSCE helped to enhance the OCPof carbon electrode by approximately 100 mV and thisenhancement was sustained till the end of the ennoblement,

as shown in Figure 4c. Such was not the case for theennoblement on Pt and Au electrodes by other bacteria:Thiobacillus ferrooxidans (23), Pseudomonas fluorescens (31),and marine bacteria (32). Interestingly, the polarization at+300 mVSCE caused a loss of OCP by approximately 100 mV.Similarly, a 15-min cathodic polarization at -300 mVSCE hadan irreversible damaging effect on the ennoblement of carbonelectrode by L. discophora SP-6, as shown in Figure 4c.

Figure 4b illustrates the evolution of OCP of stainless steelelectrodes after the 15-min anodic polarization at +300 and+500 mVSCE, respectively. The OCP data indicated that thepolarization at +300 mVSCE helped to enhance the OCP ofstainless steel electrode by approximately 50 mV whereasthe polarization at+500 mVSCE caused a loss of OCP that wasarguably due to the anodic polarization-induced-pittingcorrosion of the 304 stainless steel (22). Even at a lower anodicpotential (+100 mVSCE), a 15-min polarization had anirreversible damaging effect on the ennoblement of stainlesssteel electrode by L. discophora SP-6, as shown in Figure 4d.

The FESEM/EDS data provided more insight on the effectof electrochemical polarization on the ennoblement by L.discophora SP-6. The SEM images illustrated that the biofilmgrown on both carbon electrode polarized at +500 mVSCE

(Figure 5b) and stainless steel electrode polarized at +100mVSCE (Figure 5d) seemed to be more compact than thosegrown on the nonpolarized surfaces (Figure 5a and 5c). Forthe stainless steel electrode, while the polarization at +100mVSCE had a negative effect on the OCP ennoblement, it didimprove the growth or attachment of L. discophora SP-6biofilm. The EDS data indicated that the biofilm grown onthe polarized carbon electrode surface (+500 mVSCE) featuredhigher concentration of biominerals (7.11 wt. % Mn) thanthose grown on the nonpolarized surface (2.61 wt% Mn).

The preliminary findings above confirm our hypothesisand merit further investigation into the role of such polar-ization in the adhesion, growth, and biomineralizationprocesses of L. discophora SP-6. In addition, these fascinatingfindings have significant implications when it comes to thebiocathode MFC utilizing L. discophora or other MOB.Especially when placed in a field environment, the energyoutput of such a MFC will fluctuate over time, and sometimesthere may be a need to apply an external stimulus to promotethe adhesion, growth, and biomineralization processes ofthe bacteria. With a better understanding of how electro-chemical polarization may alter the ennoblement processes,researchers can design the algorithms accordingly to optimizethe performance of the MFC using its own energy, which ispart of our ongoing project. The idea has also beendemonstrated by other researchers on different bacteria; forinstance, the biochemical production of hydrogen fromacetate can be enhanced by applying an electrochemicalpotential to a microbial fuel cell (33).

Effect of PLL Coating on the Ennoblement. As discussedin the Introduction, it was hypothesized that coating anelectroactive polymer on the substrate material might alterthe ennoblement processes as it not only modified the initialsurface roughness of substrate but also served either as an

TABLE 1. Parameters of Equivalent Circuits Representing Various Electrode Surfaces Modified by L. discophora SP-6

substrateennobled OCP

(vs SCE) Re (Ω cm2) RCT (Ω cm2) RL (Ω cm2) CDL (F cm-2) CL (F cm-2) W (S s1/2 cm-2)

glassy carbon (#800) +135 mV 34 24 23 4.9 × 10–07 1.2 × 10-05 2.6 × 10-15

glassy carbon (#200) +60 mV 6 469 47 6.5 × 10–08 1.4 × 10-04 3.4 × 10-13

stainless steel (#1000) +80 mV 14 611 49 5.8 × 10–08 1.9 × 10-05 1.8 × 10-08

stainless steel (#200) -50 mV 2 367 34 6.0 × 10–08 3.4 × 10-05 5.9 × 10-10

glassy carbon (#800) +235 mV 35 593 18 2.4 × 10–05 1.1 × 10-03 1.4 × 10-13

glassy carbon (#800,coated with PLL) +239 mV 32 191 37 3.8 × 10–05 1.5 × 10-03 6.7 × 10-14

VOL. 41, NO. 23, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7993

electron sink or reservoir. To test this hypothesis, poly (L-lysine), or PLL, with a highly positively charged amino acidchain, was applied as a coating layer to glassy carbonelectrodes (as received, #800 finish grade). To this end, anaqueous solution of 0.01% wt. PLL was dropped on theelectrodes, which were then kept naturally drying. The coatingof the surface with PLL in the absence of L. discophora onlyincreased its OCP by approximately 20 mV. Once immersedin the bioreactor, the OCP of these electrodes were monitoredalong with that of noncoated electrodes.

Figure 6a and 6b illustrate the effect of the PLL coatingon the OCP evolution of carbon electrodes, at low cell density(∼107 cells/mL) and high cell density (∼108 cells/mL),respectively. The OCP data indicated that the carbonelectrodes coated with a PLL layer ennobled to higher OCPin a very short time. At high cell density, the OCP of PLLcoated electrode immediately reached +250 mVSCE, thenstabilized for long time. Interestingly, the noncoated carbonelectrodes reached similar OCP after a much longer time(Figure 7a). This implies that the pretreatment of electrodesurface by PLL coating enhanced the ennoblement by L.discophora SP-6 via mechanisms probably different fromthose of electrochemical polarization. At the similar level ofennoblement (OCP value), the EDS data indicated that thebiofilm grown on the PLL-coated carbon electrode surfacefeatured a lower concentration of biominerals (Mn signalstrength of 1.5 wt%) than those grown on the noncoated

surface (Mn signal strength of 4.6 wt%). It is possible that thepretreatment of electrode surface by PLL led to the formationof different types of Mn oxides, as indicated by the significantdecrease in the Mn/O concentration ratio (from 0.19 to 0.1).

PLL is known to bind to DNA, red membranes, and anynegatively charged protein, and has been used to promotecell adhesion in microbiological applications (34). In thisresearch, the PLL pretreatment of electrodes resulted in fasterand more reliable ennoblement, possibly via reducing theinitial surface roughness, promoting the bacterial adhesionand biofilm attachment to the substrate, in addition toanodically polarizing the substrate.

The preliminary findings confirm our original hypothesisrelated to the effect of PLL coating on the ennoblement andunderline the need for more fundamental and engineeringresearch on this topic. In the engineering domain, thisprovides another means for us to enhance the performanceof the biocathode MFC. The effect of pretreating the carbonelectrodes with other electroactive polymers (e.g., polya-nilines) is being investigated in our laboratory as well.

EIS as an Effective Means To Characterize the Ennoble-ment. In addition to OCP monitoring, we used EIS measure-ments throughout this research in order to gain more insightabout the electrolyte/biofilm and biofilm/substrate interfacesand to validate or supplement some of the assumptions usedto interpret the OCP and FESEM/EDS data. Figure 7b providesan example of EIS Nyquist diagrams, illustrating the dif-

FIGURE 9. Correlations between equivalent circuit parameters and ennobled OCP: (a) Re; (b) logarithm of CDL; and (c) logarithmof CL.

7994 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 23, 2007

ference between two ennobled carbon electrodes, onepretreated with PLL and the other without such pretreatment.In order to analyze the EIS data, we adopted an equivalentcircuit as shown in Figure 8 for all ennobled electrodes,whether they were glass carbon or stainless steel. Theequivalent circuit consisted of parameters characterizing theelectrolyte resistance (Re), double-layer capacitance (CDL),charge-transfer resistance (RCT), Warburg diffusion imped-ance (W), low-frequencies resistance (RL), and capacitance(CL) (30).

As shown in Table 1, the EIS data indicated a smootherinitial substrate surface led to significantly higher double-layer capacitance (CDL) on glass carbon electrodes, whichmay be due to the surface modification by less-conductivebiominerals. Both reducing the initial surface roughness ofcarbon electrode from #200 to #800 and coating it with a PLLlayer significantly reduced both its charge-transfer resistance(RCT) and its Warburg diffusion impedance (W), which maybe considered a result of better surface continuity andhomogenuity. Interestingly, as the ennoblement progressed,the charge-transfer resistance (RCT) increased, likely due tothe surface modification by less-conductive biominerals (e.g.,MnOOH and MnO2), as suggested by the EDS data. As such,the value of RCT for glass carbon electrodes ennobled to+135and +235 mVSCE was 24 and 593 Ω.cm2, respectively.

Independent of substrate material, initial surface rough-ness, and pretreatment by PLL, there were three parametersin the equivalent circuit that correlated well with the OCPindicating the level of ennoblement by L. discophora SP-6.First of all, among six ennobled electrodes, their electrolyteresistance (R′e, in Ω) increased with the OCP nonlinearly, asshown in Figure 9a. That may be explained by the increasinglythicker biofilm accumulated on the substrate surface thatserved as a physical barrier for the ionic species in theelectrolyte trying to access the substrate. Second, the double-layer capacitance (CDL, in F cm-2) correlated with the OCPeven better than Re (Re*S), as shown in Figure 9b. In the earlystage of the ennoblement by L. discophora SP-6, CDL changedlittle. But as the ennoblement progressed, CDL increasedexponentially. This may be explained by the surface modi-fication by less-conductive biominerals over the duration ofennoblement. Finally, there was also a strong, yet morecomplex correlation between the low-frequencies capaci-tance (CL, in F cm-2) and the OCP, as shown in Figure 9c. Inthe early stage of the ennoblement by L. discophora SP-6, CL

changed little or even decreased. But as the ennoblementprogressed, CDL increased nonlinearly. This suggests twocompeting mechanisms at work, one of which eventuallyoverwhelmed the other. More research is needed in order tounderstand the various factors influencing CL as well as otherequivalent circuit parameters (e.g., W indicating the diffusionof divalent Mn cations and its cycling within the biofilm (4)).

Independent of substrate material, initial surface rough-ness and pretreatment by PLL, the three parameters (Re, CDL,CL) in the EIS equivalent circuit correlated well with the OCPindicating the level of ennoblement by L. discophora SP-6.

These above fascinating findings merit further investiga-tion as they may shed light on the fundamental bacteria/substrate interactions and help advance the knowledge baseneeded for the engineering applications.

AcknowledgmentsThis work was supported by the Research and InnovativeTechnology Administration under the U.S. Department ofTransportation through the University Transportation Centerresearch grant awarded to the Western TransportationInstitute. We specially thank Dr. Recep Avci and the ICALfacility he directs at Montana State University–Bozeman forthe use of FESEM/EDS. We also extend our appreciation toProfessor Shizhe Song at School of Materials Science and

Engineering, Tianjin University, for her advising the Ph.D.dissertation of Y.L.

Literature Cited(1) Johnsen, R.; Bardal, E. Cathodic properties of different stainless

steels in natural seawater. Corrosion 1985, 41 (5), 296.(2) Mollica, A.; Trevis, A. Correlation Entre la Formation de la

Pellicule Primaire et la Modification de la Cathodique sur desAciers Inoxydables Experimentes en eau de Mer aux Vitessesde 0.3 a 2 m/s. In Proceedings of 4th International Congress onMarine Corrosion and Fouling, Antibes, France, 1976.

(3) Motoda, S.; Suzuki, Y.; Shinohara, T. The effect of marine foulingon the ennoblement of electrode potential for stainless steels.Corros. Sci. 1990, 31, 515.

(4) Shi, X.; Avci, R.; Lewandowski, Z. Electrochemistry of PassiveMetals Modified by Manganese Oxides Deposited by Leptothrixdiscophora: Two-step Model Verified by TOF-SIMS. Corros. Sci.2002, 44 (5), 1027.

(5) Dexter, S. C.; Gao, G. Y. Effect of Seawater Biofilms on CorrosionPotential and Oxygen Reduction of Stainless Steel. Corrosion1988, 44 (10), 717.

(6) Chandrasekaran, P. S.; Dexter, C. Proceedings of Corrosion 1993;Houston, TX, NACE,1993, Paper No. 493.

(7) Scotto, V.; DiCintio, R.; Marcenaro, G. The Influence of MarineAerobic Microbial Film on Stainless Steel Corrosion Behaviour.Corros. Sci. 1985, 25 (3), 185.

(8) Eashwar, M.; Maruthamuthu, S. Mechanism of BiologicallyProduced Ennoblement: Ecological Perspectives and a Hypo-thetical Model. Biofouling 1995, 8, 203.

(9) Dexter, S. C.; Zhang, H. J. Effect of biofilms on corrosion potentialof stainless steel alloys in estuarine waters. In Proceedings of the11th International Corrosion Congress, Florence, Italy, 1990, 333.

(10) Mollica, A.; Trevis, A.; Traverso, E.; Ventura, G.; DeCarolis, G.;Dellepiane, R. Cathodic Performance of Stainless Steels inNatural Seawater as a Function of Microorganism Settlementand Temperature. Corrosion 1989, 45, 48.

(11) Scotto, V.; Beggiato, M.; Marcenaro, G.; Dellepiane, R. Microbialand Biochemical Factors Affecting the Corrosion Behavior ofStainless Steels in Seawater. In A working Party Report on MarineCorrosion of Stainless Steels: Chlorination and Microbial Effects;European Federation of Corrosion Publication No. 10; TheInstitute of Materials: London, 1993; p 21.

(12) Yang, X.; Beyenal, H.; Harkin, G.; Lewandowski, Z. QuantifyingBiofilm structure using image analysis. J. Microbiol. Methods2000, 39, 109.

(13) Dickinson, W. H.; Caccavo, F., Jr.; Olesen, B.; Lewandowski, Z.Ennoblement of Stainless Steel by the Manganese-DepositingBacterium Leptothrix discophora. Appl. Environ. Microbiol. 1997,63 (7), 2502.

(14) Tebo, B. M.; Bargar, J. R.; Clement, B. G.; Dick, G. J.; Murray,K. J.; Parker, D.; Verity, R.; Webb, S. M. Biogenic ManganeseOxides: Properties and Mechanisms of Formation. Annu. Rev.Earth Planet Sci. 2004, 32, 287.

(15) Ghiorse, W. C.; Chapnick, S. D. Environ. Biogeochem., Ecol. Bull.(Stockhom) 1983, 35, 367.

(16) Nealson, K. H.; Tebo, B. M.; Rosson, R. A. Occurrence andMechanisms of Microbial Oxidation of Manganese. Adv. Appl.Microbiol. 1988, 33, 279.

(17) Rhoads, A.; Beyenal, H.; Lewandowski, Z. Microbial Fuel CellUsing Anaerobic Respiration as an Anodic Reaction andBiomineralized Manganese as a Cathodic Reactant. Environ.Sci. Technol. 2005, 39, 4666.

(18) Tebo, B. M.; Johnson, H. A.; McKarthy, J. K.; Templeton, A. S.Geomicrobiology of Manganese(II) Oxidation. Trends Microbiol.2005, 13 (9), 421.

(19) Dickinson, W.; Lewandowski, Z. Electrochemical and Micro-electrode Studies of Stainless Steel Ennoblement. In Proceedingsof Corrosion 1995; Houston, TX, NACE, 1995; paper no. 223.

(20) Emerson, D.; Ghiorse, W. C. Isolation, Cultural Maintenance,and Taxonomy of a Sheath-forming Strain of Leptothrix dis-cophora and Characterisation of Manganese-oxidizing ActivityAssociated with the Sheath. Appl. Environ. Microbiol. 1992, 58,4001.

(21) Shi, X.; Avci, R.; Lewandowski, Z. Biomineralized Manganeseand Iron Oxides on Passive Metals: Their Chemistry andConsequences for Material Performance. Corrosion 2002, 58(9), 728.

(22) Shi, X.; Avci, R.; Lewandowski, Z. Comparative Study inChemistry of Microbially and Electrochemically Initiated Pitsof Type 316L Stainless Steel. Corros. Sci. 2003, 45 (11), 2577.

VOL. 41, NO. 23, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7995

(23) Brouwers, G. J.; Vijgenboom, E.; Corstjens, P. L. A. M.; De Vrind,J. P. M.; De Vrind De Jong, E. W. Bacterial Mn2+ Oxidizing Systemsand Multicopper Oxidases: An Overview of Mechanisms andFunctions. Geomicrobiol. J. 2000, 17, 1.

(24) Adams, L. F.; Ghiorse, W. C. Characterization of ExtracellularMn2+-oxidizing Activity and Isolation of an Mn2+-oxidizingProtein from Leptothrix Discophora SS-1. J. Bacteriol. 1987, 169,1279.

(25) Boogerd, F. C.; De Vrind, J. P. M. Manganese Oxidation byLeptothrix discophora. J. Bacteriol. 1987, 169, 489.

(26) Natarajan, K. A. Effect of Applied Potential on the Activity andGrowth of Thiobacillus ferrooxidans. Biotechnol. Bioeng. 1992,39, 907.

(27) Quirynen, M.; Bollen, C. M. The Influence of Surface Roughnessand Surface-free Energy on Supra- and Subgingival PlaqueFormation in Man. A Review of the Literature. J. Clin. Periodontol.1995, 22 (1), 1.

(28) Kawai, K.; Urano, M.; Ebisu, S. Effect of Surface Roughness ofPorcelain on Adhesion of Bacteria and Their SynthesizingGlucans. J. Prosthet. Dent. 2000, 83 (6), 664.

(29) Tang, H.; Cao, T.; Wang, A.; Liang, X.; Salley, S. O.; McAllister,J. P., II; Simon Ng, K. Y. Effect of Surface Modification of Siliconeon Staphylococcus epidermidis Adhesion and Colonization.J. Biomed. Mater. Res. Part A 2007, 80A (4), 885.

(30) Ferlony, P.; Mastracgostino, M.; Meneghel, L. ImpedanceAnalysis of Electronically Conducting Polymers. Electrochim.Acta 1996, 41 (1), 27.

(31) Busalmen, J. P.; De Sanchez, S. R. Electrochemical Polarization-Induced Changes in the Growth of Individual Cells and Biofilmsof Pseudomonas fluorescens. Appl. Environ. Microbiol. 2005, 71(10), 6235.

(32) Gordon, A. S.; Gerchakov, S. M.; Udey, L. R. The Effect ofPolarization on the Attachment of Marine Bacteria to Copperand Platinum Surfaces. Can. J. Microbiol. 1981, 27 (7), 698.

(33) Liu, H.; Grot, S.; Logan, R. Electrochemically Assisted MicrobialProduction of Hydrogen from Acetate. Environ. Sci. Technol.2005, 39, 4317.

(34) Zhang, Y.; Li, J.; Shen, Y.; Wang, M.; Li, J. Poly-L-lysineFunctionalization of Single-Walled Carbon Nanotubes. J. Phys.Chem. B 2004, 108, 15343.

ES071178P

7996 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 23, 2007