Surface Modification for Enhanced Biofilm Formation and Electron Transport in Shewanella Anodes

aCenter for Micro-Engineering Materials (CMEM), Department of Chemical & Biological Engineering, The University of New Mexico, Albuquerque, New Mexico 87131, USA bCenter for Biochemical Engineering, Department of Chemical & Biological Engineering, University of New Mexico, Albuquerque, New Mexico 87131, USA cCenter for Emerging Energy Technology (CEET), Department of Civil Engineering, The University of New Mexico, Albuquerque, New Mexico 87131, USA

The possibility to directly convert the energy of chemical bonds into electricity has been recognized as an alternative and effective approach for energy transformation.This quest has been embodied in electrochemical devices including batteries and fuel cells.Commonly used batteries and fuel cells employ inorganic catalysts and harmful, costly electrolytes.In order to overcome the limited reserve of noble metals usually used and avoid the use of harmful compounds, a new avenue of electrochemical systems has been developed.These new systems rely on bio-catalytic ability of enzymes and microorganisms in fuel cells. 1 The latter has gained attention of researchers, government agencies and industry driven by, among other things, the potential for combining efficient wastewater purification with concomitant electricity production.This achievement will provide an opportunity for the development of portable, self-sustaining wastewater treatment units.
4][5][6][7][8] Among the general requirements, such as good conductivity, chemical stability, mechanical strength, high surface area and low cost, anode materials should posses several key characteristics that will determine the rate of bacteria-electrode interactions.] Based on these requirements, various surface modification techniques have been explored and recognized as a new avenue for MFCs optimization.][14] Both of these factors strongly depend on the electrode material properties, such as real surface area, conductivity, hydrophilicity and surface chemistry.
7][18] These materials can be modified to provide higher bacteria adhesion and increased biofilm formation and development. 29 In addition, surface chemical or electrochemical oxidation can also be an effective treatment method for enhanced bacteria-electrode interactions. 9Surface chemistry modification techniques such as benzene diazonium treatment 20 and ammonia gas treatment 20 provide a positively charged electrode surface, which as been speculated to be beneficial for bacterial attachment. 3,9,11,20In addition, simultaneous utilization of nitric acid and sulfuric acid, provide surface oxidation and introduction of oxygen containing groups.This modification approach increased the ratio of saturated/unsaturated carbon on the surface and consequently decreased the electrode resistance, which also had a positive effect on bacteria attachment and biofilm formation. 9Another example of electrochemical oxidation of the electrode material resulted in a significant increase in current produced from MFC graphite anodes over MFCs containing untreated anodes. 21These methods, however effective, are not optimal for large-scale applications due to cost, safety chemical handling and technical requirements.Therefore, we propose a simple, fast and effective and therefore low-cost surface modification method for enhanced biofilm formation, increased electron transfer rate, and thus higher current density generation.This method focuses in lowering the degree of hydrophobicity of the anode material and consists of partial oxidation of carbonaceous material by UV/O 3 treatment. 22n this process, a UV/O 3 irradiation chamber utilizes UV light to excite ambient O 2 molecules, thus lowering their activation energy and moving them into a more reactive state.The interaction between highly reactive oxygen and carbon atoms from the carbon felt structures leads to the formation of different functional groups at surface, including -OH, -COOH, =O, =CO and -CO.4][25] The procedure provides a type of surface treatment that can be carried out at atmospheric pressures thus, lowering treatment cost, and minimizes damaging effects inherent in other treatment techniques such as strong acid oxidation. 21The influence of the UV/O 3 time of exposure to the material surface chemistry and morphology was examined and correlated with the electrochemical performance of these electrodes when explored in Shewanella oneidensis MFC.
As shown in our previous studies, we have determined that the degree of hydrophobicity/hydrophilicity and surface charge of the material are very important parameters to be considered for enhanced bacterial anode attachment and thus, improved extracellular electron transfer (EET).The latter leads to higher MFC overall performance with both pure and mixed cultures. 11,19The current start up time in MFCs is a crucial parameter that is highly influenced by the EET rate and is therefore carefully taken into consideration in our electrochemical characterization methods.We have shown that the start up time of MFCs inoculated with wastewater was also decreased by electrochemical oxidation of the anode, introducing different surface chemistries. 9,11. oneidensis utilizes several biomolecules that take part in EET.It is know that metabolites such as flavins along with and even in association with outer membrane proteins (cytochromes) play an important part in this process. 12,26In particular, riboflavin molecules secreted by S. oneidensis present electrochemical activity at the biofilm-electrode interface 12 and more importantly play the role of a cofactor for the outer membrane cythochrome OmcA.In these studies, we verified the importance of riboflavin in the EET by showing a direct correlation between riboflavin-enriched anodes and anodic performance.

Materials and Methods
Bacterial culturing conditions.-Wildtype Shewanella oneidensis MR-1 were stored in as 40% glycerol frozen stocks at −80 • C. Cultures were maintained on Triptic Soy Broth (TSB; Becton Dickinson, Franklin Lakes, NJ) agar plates and incubated for 24 hours at 30 • C. Isolated colonies were subcultured into TSB media and incubated aerobically at 150 rpm and 30 • C for approximately 18 hours.The cultures were then washed three times with 50 mM phosphate buffer at a pH of 7.4 and resuspended at 7 × 10 7 cells mL −1 in the same buffer.Upon the last wash, 30 mM of sodium fumarate (Sigma) was added as an electron acceptor and 80 mM of sodium lactate (Sigma) was added as an electron donor.
Electrode surface modification.-Surfacemodification relied on partial oxidation of carbon material by UV/ozone treatment.Carbon felt disks with 1.2 cm in diameter were treated at different UV/ozone exposure times ranging from 0 to 60 minutes using a UV/O 3 cleaner 144AX (Jelight Company, Irvine, CA).As a benchmark, a carbon felt electrode was treated with Isopropyl alcohol (IPA) for 10 min and rinsed 3 times in sterile buffer.
Electrochemical testing.-Theelectrochemical experiments were conducted in a three-electrode electrochemical cell with a reference electrode (saturated Ag/AgCl), a platinum wire counter electrode and a carbon felt as the working electrode (1.2 cm diameter). 27Shewanella MR-1 suspended in 50 mM phosphate buffer was introduced in the cells prior to electrochemical experiments.All electrochemical tests were done using a Gamry Reference 600 potentiostat (Gamry Instruments, Warminster, PA).Chronoamperometry studies were performed for 24 hours at a constant potential of −0.30V vs. Ag/AgCl.Potentiostatic polarization curves were carried out based on several chronoamperometry steps starting from open circuit potential to 0.30 V vs. Ag/AgCl with a step of 0.05 V.The electron transfer mechanism was examined via cyclic voltammetry in the potential region of −0.45 V to 0.40 V vs. Ag/AgCl at a scan rate of 10 mV s −1 .The electrochemical cell was purged with nitrogen and sealed with Parafilm (Sigma) throughout the experiments to maintain micro-aerobic conditions. 27termination of riboflavin content.-Cyclicvoltammetry measurements of the S. oneidensis anodes after 24 hours of potentiostatic polarization were carried out at various scan rates.A couple of redox peaks associated with the riboflavin oxidation/reduction was observed, and a linear dependence of the peak currents from the scan rate was detected.The last indicates surface confined electrochemistry, which allows the utilization of the following equation for the calculation of the riboflavin surface coverage (mol cm −2 ): 28 Where i p is the peak current (A), n is the number of electrons necessary for riboflavin redox transformation, F is the Faradaic constant, R is the universal gas constant, T is the temperature in K, A is the surface area of the electrodes (cm 2 ) and υ is the scan rate (V s −1 ).The riboflavin surface coverage was calculated first based on the geometrical surface area and then normalized to the electrochemical accessible surface area (ECSA).
Scanning electron microscopy.-Scanningelectron microscopy (SEM) analysis was used for visual confirmation of biofilm formation on the carbon-felt electrodes after the electrochemical characterization.
Upon completion of these studies, bacterial cells were fixed to the electrode using a 2.5% glutaraldehyde solution for 24 hours and then exposed to increasing concentrations of ethanol (50, 60, 70, 80, 90, 95, and 100%) ending in 3 separate washes of 100% ethanol.The samples were then dried by immersion in 50% (v/v) hexamethyl-disilazane (HMDS) in ethanol for 15 min, followed by 15 min in 100% HMDS and finally air exposure overnight.The samples were then coated with a 15 nm layer of gold-palladium using an Emitech K950X sputter coater.A Hitachi S-5200 Nano SEM was used to capture the samples' images at an accelerating voltage of 10 kV.
X-ray photoelectron spectroscopy.-Surfacechemistry of the carbon felt materials modified through partial oxidation by different times of exposure to UV/O 3 were characterized using XPS analyses.XPS measurements were performed with a Kratos Axis Ultra DLD X-ray photoelectron spectrometer using a monochromatic Al Kα source operating at 225 W. Survey and C1s, O1s and N1s high resolution spectra were acquired at 80 and 20 eV pass energy, respectively.Charge compensation was not necessary.The data obtained are average of 3 different areas per sample.
Data analysis and quantification were performed in CasaXPS software.A linear background was used for quantification.Quantification utilized sensitivity factors that were provided by the manufacturer.High resolution C 1s spectra were curve-fitted using individual peaks with 70% Gaussian / 30% Lorentzian (GL (30)) peak shape and 1.0 +/− 0.2 eV full width at half maximums (FWHMs).

Evaluation of electrochemical accessible surface area (ECSA).-
The ECSA of the carbon felt electrodes after their modification was evaluated based on changes in materials capacitance. 29Cyclic voltammetry of the electrodes in 50 mM phosphate buffer was carried out to determine the capacitance of the modified carbon felts experimentally.The ECSA was determined using the following equation: where C sp is the specific capacitance, i is the difference in the anodic and cathodic current at a chosen potential, and υ is the scan rate in V s −1 .The specific capacitance for the carbonaceous material was taken to be 35 μF cm −2 only for electrodes comparison, and i under nitrogen conditions was taken at a scan rate of 10 mV s −1 to calculate the ECSA.

Results and Discussion
Shewanella oneidensis MR-1 is a model organism in MFCs due to its capability to transfer electrons across the outer cell membrane to reduce the number of alternative electron acceptors beside oxygen.1][32] Previous studies have shown that, after 24 h of polarization, when an S. oneidensis biofilm is covering the electrode surface, the number of cells attached to the surface was defined mostly by the ECSA and the expanded uncertainty of the anode's response decreased significantly when the current was normalized to that parameter. 27At the same time, our group has also demonstrated the importance of the hydrophilicity and surface morphology on bacteria attachment and the dynamics of biofilm formation determining the start up time of MFCs. 16Therefore in this study the important parameters related to the properties of the electrode material have been examined.These include electrochemical accessible surface area of the electrodes, their resistance and surface chemistry as well as S. oneidensis biofilm formation, riboflavin accumulation, and anode performance.
Surface chemistry and morphology.-X-rayphotoelectron spectroscopy (XPS) was used to study the changes in the surface composition of the materials as a result of the oxidative etching.Table I summarizes the evaluated chemistry.
According to the XPS data, the performed modification procedure led to an increase in the content.A linear dependence between the time of UV/O 3 exposure and the amount of carboxyl and hydroxyl groups was observed as shown in Figure 1a.At the same time, a decrease in the amount of sp 2 carbon was also seen (Figure 1b).The increase in oxygen containing functional groups should most likely cause an increase in the hydrophilicity of the electrodes and thus enhance the biofilm formation and development.It was also suggested that the decrease in C=C bonds leads to a decreased resistance as was observed before, 9 but no such observation can be made here.The resistance of the modified electrodes in this study, determined via electrochemical impedance spectroscopy, increased with the increased time of UV/O 3 exposure or with the decreased amount of C=C bonds.The resistance of the untreated carbon felt electrode was 24 increasing to 136 after 20 min of UV/O 3 exposure, reaching 285 and 366 after 45 and 60 min of UV/O 3 exposure, respectively.This observation, besides being controversy to previous studies, is expected since the electrons in carbonaceous materials are being transferred through the C=C bonds.
The oxidative etching of carbon materials creates defects and holes in their structure increasing their specific surface area.The electrochemical accessible surface area was calculated using electrode capacitance.It is important to clarify that the ECSA determined from this study can only be used for relative comparison of the studied materials.From the increase in the amount of oxygen with longer exposure time, we can assume that the ECSA will also increase.Figure 1c shows the dependence of the electrode ECSA on the time of UV/O 3 exposure.The increase of the ECSA with the treatment time confirmed our assumption and raised another hypothesis that increased surface area available for bacteria attachment and biofilm growth will improve the anode performance.It is well known that more porous materials typically produce higher current densities per geometrical surface area compared to their smoother counterparts.It has been also established that electrochemical accessible surface area of carbon felt is a limiting factor when S. oneidensis biofilms are formed. 27Therefore we expected increased anode current densities at higher UV/O 3 exposure time besides the increasing electrodes resistance.
Electrochemical performance and biofilm coverage.-Ithas been established that biofilm formation for Shewanella oneidensis MR-1 occurs preferentially on an electrode polarized with a constant poten-tial of −0.3 V vs. Ag/AgCl. 27Biofilms formed under these conditions yield behavior indicative of mediated electron transport (MET) at lower potentials and direct electron transport (DET) from outer membrane-bound multiheme cytochromes at potentials above 0.00 V vs. Ag/AgCl. 27ll electrodes were subjected to a constant potential of −0.3 V for 24 hours while current production was recorded (Figure 2a).Additional lactate was introduced into the electrolyte after 10 hours of operation to ensure fuel excess.The observed maximum current for 20 and 45 min treated electrodes reached roughly 32 μA after 24 hours.At the same time, the electrode treated by 60 minutes of UV/O 3 exposure exhibited maximum current similar to those of IPA-treated carbon felt, used as a benchmark, which was notably lower then the rest of the modified electrodes.No significant current was detected when using untreated carbon felt.The increase in generated current for the 45 min treated electrode after 7-8 hours of polarization was an indication of initial biofilm formation as it has been observed previously. 27he start up time for 20 min treated electrode was significantly higher than the 45 min electrode indicating slower biofilm development.After 24 hour of constant electrode polarization we expected that the 20 and 45 min treated electrodes would possess higher bacterial coverage than all electrodes studied.The polarization curves taken at the end of the chronoamperometry measurement showed a similar trend in anodes performance, where 45 min time point presented the highest performance among all samples (Figure 2b).The shape of the recorded polarization curves is in accordance with previous studies of S.oneidensis anodes, demonstrating the ability of these bacteria to exhibit both MET and DET at the proper potentials.It is clear that the electrode modification through UV/O 3 partial oxidation promotes DET and provides higher surface for direct contact between bacterial cells and electrode surface which may be due to both increased hydrophilicity 22,25 of carbon felt and increased ECSA (Figure 1c and Figure 3).SEM of the ozonized carbon felts demonstrated the highest bacterial coverage on samples treated with 45 min of UV/O 3 (Figure 3), whereas untreated electrodes supported the lowest amount of bacterial attachment.Although the electrode subjected to 60 min of UV/O 3 oxidation had the highest ECSA (Figure 1c), the bacteria coverage was lower than the 45 min treated anode (Figures 3 and 4).In contrary to the highest ECSA the 60 min treated electrode possessed the highest resistance (366 ), which was 1.3 times higher than the resistance of 45 min treated carbon felt and 2.7 times higher than 20 min treated electrode.The latter suggests that the increased electrode resistance is additional factor regulating bacteria attachment and biofilm growth  and The order in which the current densities generated by the electrodes increase follows the order in which the bacterial coverage changes, demonstrating an expected correlation between MFC performance and biofilm formation on the surface of the anode. 5,10,16tection of metabolites implicated in extracellular electron transfer.-Cyclicvoltammograms of the tested S. oneidensis anodes (Figure 5) demonstrate increasing capacitance with the increasing time of UV/O 3 exposure, which can be explained by the increasing ECSA of the treated electrode materials.One can also identify peaks that can be attributed to both MET and DET as it has been observed before. 27The first couple of redox peaks with formal redox potential (E • ' ) of approximately -0.45 V vs. Ag/AgCl can be assigned to the reversible electrochemical transformation of riboflavin, a metabolic product known to play a key role in the extracellular electron transfer in Shewanella spp. 12 The catalytic wave with half wave potential (E 1/2 ) at 0.02 V vs. Ag/AgCl is due to DET carried out through the outer surface cytochromes.And the irreversible oxidation peak at potentials at 0.40 V vs. Ag/AgCl was ascribed to cytochromes' oxidation at the electrode surface. 27,33n order to quantify the amount of riboflavin secreted by bacteria in contact with the anodes, cyclic voltammograms were generated at scan rates varying from 5 mV s −1 to 200 mV s −1 .Plotting the peaks current associated with the riboflavin redox reaction versus the scan rate, a linear trend line was found, which indicated surface confined reaction of riboflavin reduction.Therefore, we can claim that the peaks observed at potentials around −0.30 V vs. Ag/AgCl are due to riboflavin oxidation/reduction, when it is adsorbed on the electrode surface as previously demonstrated when electrodes were colonized with S. oneidensis biofilms. 12iboflavin molecules are implicated in DET pathway of Shewanella anodes, 26 with flavins enhancing the electron transfer through the outer membrane cytochromes as bound cofactors, not as freeform flavins.This concept for EET is also found in the model organism Geobacter sulfurreducens, which secrete flavins at concentrations comparable to those of S. oneidensis and exploits these for enhanced communication with an electrode. 13For S. oneidensis bioelectrodes, flavins adsorbed on the anode interact with MtrC or OmcA proteins found on the outermembranes and nanowires and enhance the enzymatic reaction. 33iboflavin secretion by Shewanella is a sensor and response to redox changes, and as such, it promotes cell attachment and biofilm formation. 33In this model, riboflavin is retained at the electrode surface and we hypothesize that by that it creates a redox gradient that can be sensed by bacteria and directs cell swimming towards the electrode surface and attachment.
The amount (nmol cm −2 ) of riboflavin adsorbed on the electrodes was determined based on the response of the peak current to riboflavin concentration. 28Since the geometrical surface area is not representative of the specific surface area of the materials, the amount of riboflavin adsorbed was calculated based on both the geometrical surface area and the ECSA (Figure 1c).During flavin-mediated electron transfer, one electron is transferred between the outermembrane cytochromes and flavin mononucleotide (FMN), resulting in the formation of semiquinone rather than fully reduced FMN; 26 therefore we based our model on such a 1 electron transfer.Figure 6a shows linear correlation between the riboflavin coverage and the UV/O 3 treatment time when the riboflavin coverage was normalized to the geometrical surface area of the electrode.In contrast, when the riboflavin coverage was normalized to the ECSA a polynomial dependence was observed with highest amount of riboflavin adsorbed on 45 min treated electrodes, followed by 20 min oxidized one (Figure 6b), which also showed the highest current densities (Figure 2b).This observation is in accordance with the hypothesis that flavins are involved in DET and that respiratory current from dissimilatory metal reducing bacteria is strongly coupled to the amount of electrochemically active flavins.
It has previously been shown that oxidized graphite can adsorb and retain higher amounts of mediator, most likely due to the increased specific surface area of the electrodes. 17The same observation can be made in this study.Increasing the time of UV/O 3 exposure provides higher electrochemical accessible surface area, increased amount of oxygen containing groups and thus increased amount of adsorbed riboflavin.
3] This is due to the ability of soluble flavin to be adsorbed at the surface of the electrode.Our reports corroborate these findings after performing further electrochemical studies.These consisted of non-treated UV/O 3 carbon felt electrode incubated in a solution of 1 mM riboflavin for 24 hours prior to electrode polarization in the MFC.The results were compared to measurements in the absence of additional riboflavin from non-treated carbon felt.
We found that the electrode incubated in riboflavin for 24 hours prior to electrochemical experiments demonstrated higher current generation during the polarization step, greater anode performance and higher CV peaks associated with riboflavin oxidation/reduction in comparison to the unmodified anode (Figure 7).We speculate that by introducing additional riboflavin molecules to the electrode surface, which are readily available to interact with the outer membrane cytochromes and leads to enhanced electrochemical behavior.

Conclusions
Electrochemical studies have been performed on S. oneidensis utilizing UV/O 3 treated carbon felt electrodes.The proposed treatment is a simple, low-cost, and effective method, which can be easily exploited for modification of high surface anode materials.Carbon electrodes exposed to 45 min of surface treatment provided the best electrochemical results and richer bacterial cell attachment.This was further confirmed via SEM imaging and analysis.In addition, we have shown that enrichment of the electrode surface with flavin correlates with the increased anodic performance.This observation is in accordance with the hypothesis that flavins are involved in DET.

Figure 1 .Figure 2 .
Figure 1.Dependence of a) C-OH, C=O, b) C=C bonds from the time of UV/O 3 exposure and c) Electrochemical accessible surface area of carbon felt electrodes treated at different UV/ozone times of exposure.

Figure 3 .
Figure 3. SEM images of carbon felt electrodes subjected to different times of UV/O 3 exposure taken after electrodes polarization and electrochemical characterization.

Figure 4 .
Figure 4. Bacterial surface area coverage on carbon felts treated at different UV/O 3 times.The area coverage was calculated after processing SEM images through Image J.

Figure 5 .
Figure 5. Cyclic voltammograms of carbon felt anodes treated through UV/O 3 for various time of exposure at scan rate of 10 mV s −1 .

Figure 6 .
Figure 6.Riboflavin coverage at the electrode surface, calculated based on: a) the electrode geometrical surface area; b) the ECSA.

Figure 7 .
Figure 7. Electrochemical measurements from non-treated carbon felt electrodes (0 min) incubated in riboflavin solution prior to Shewanella inclusion (red trace) and when no additional riboflavin is included (blue trace).a) Chronoamperometry of Shewanella; b) Polarization curves; c) Cyclic voltammograms.