An interesting field of science and technology is microbial fuel cells (MFC). MFC use bacteria as the catalysts to convert the chemical energy of fuel (such as organic compounds) to electrical energy [1] [2] . MFCs have limitations of power output. This affects the various uses of MFC (electricity generation, wastewater treatment, biosensor, and hydrogen production) [3] . Efficiencies of MFCs are affected by cell configuration, electrode materials, electrolyte compositions, microorganisms, nutritional substrates [4] , and the presence or absence of oxygen [5] [6] .

Different mechanisms for electron transfer from the microbe to the electrode surface have been proposed. Direct electron transfer can be via either direct contact between outer membrane cytochromes (MtrC and OmcA) with electrodes or through extracellular conductive connections called bacterial nanowires [7] [8] . Mediated electron transfer via soluble redox mediators (such as flavin and riboflavin) also carries charge to the electrode [9] [10] .

The effect of added ions to the solution is still not fully understood. Metal ions such as Cu²⁺, Cd²⁺, and Fe³⁺ have been studied to affect in enhancing MFC activity [11] . Wu et al. [12] showed that 10 mM Fe³⁺ addition to solution in S. oneidensis MR-1 increased the power output of MFCs by 1.1 times than control. Kim et al. [13] found that the addition of 100 mM Fe³⁺ in solution increased the power output of MFCs 80%. Fitzgerald et al. [14] reported that the addition of Ca²⁺ at the concentration of 1.4 mM increased the current density by more than 80% in S. oneidensis MR-1. Heavy metals, too, can increase power output. As an example, Xu et al. [11] found that the addition of 5-6 nM Cu²⁺ and Cd²⁺ increased the power output of MFCs by 1.3 - 1.6 times.

One possibility is that metal ions affect solution properties. It is known that metal ions increase the conductivity of electrolytes, decrease internal resistance, and can increase the power output of MFCs [15] . Electrolyte salt concentration correlated with ionic strength has been shown to affect power output [4] [16] [17] .

In our research, we utilize S. oneidensis MR-1. S. oneidensis MR-1 is a gram-negative cell that contains three layers: the cytoplasmic membrane (CM) or inner membrane, the peptidoglycan layer which is external to the cytoplasmic membrane, and the outer membrane (OM) which is lipopolysaccharide-rich. The outer membrane (OM) consists of the bilayer that composed lipid, protein, and lipopolysaccharide [18] . The extracellular polysaccharides (EPS) play a major role in mediating the bacterial colonization of surfaces and cell adhesion [19] . Metal ions bind to EPS and can affect the EPS structure, which affects adhesion. Ca²⁺ is important as an ionic cross-bridging molecule for negatively-charged polysaccharides [20] [21] . It has been suggested that negatively charged groups in the O-antigen are able to mediate Ca²⁺ dependent cell adhesion as well [22] . Ca²⁺ and Mg²⁺ can affect bacterial adhesion and biofilm formation by impacting electrostatic interactions and physiology-dependent attachment processes [23] .

The presence of metal ions in solution passively increases bacterial attachment to soil surfaces [24] . There it was suggested that salt non-specifically compresses of the double layer between negatively charged solid surfaces and negatively charged bacterial surfaces. The literature shows that salt affects bacterial adhesion to Teflon, glass, metal surface wettability, and metal oxide surface by increasing the ionic strength [25] . As an ion compresses the double layer, it might affect the microbes approach an electrode and attach to the surface [26] .

The enhancement of microbially driven electrochemistry enhancement by metal ions may be offset by toxic biological effects tend to bind to a sulfhydryl group inside the cell [27] . In plants, they can affect photosynthesis, gaseous exchange, and nutrient absorption of plants [28] .

The goal of this study is to study whether metal ions enhance the efficiency of bacterial extracellular electron transfer (EET) in S. oneidensis MR-1 by enhancing adhesion to the electrode for improving the current output of microbial fuel cells. We use Indium Tin Oxide (ITO) electrode surface to study microbial adhesion because it is a metal oxide [29] . Bacteria are known to attach on natural metal oxides [30] . Adhesion of bacteria and that subsequent effect on electrochemical behavior is studied by variation of metal ions such as Ca²⁺, Pb²⁺, Cd²⁺, and Mg²⁺ and their concentrations.

Bacteria samples were injected into the electrochemical cell for a two-hour anodic loading period at the potential of +0.2 V vs. Ag/AgCl. The net current density was monitored during the loading process. Figure 1(a), shows a typical loading current density vs. time with blanks (no bacteria) tryptic soy broth (TSB) only, TSB with metal added, control (TSB with bacteria), and Cd²⁺ ion addition. Addition of new food source, as anticipated, allowed proliferation of the bacteria, Figure 1(b).

Figure 1(a) shows measurable features related to attachment. A rapid decay of current (Figure 1(a), feature 1) is associated with surface charging. This occurs within the first five seconds. Feature 2 (lag phase of time (t_lag) with a nearly constant lag current (i_lag)) has been observed elsewhere. This start-up phase depends on substrate (i.e. glucose vs acetate), electron acceptor (fumarate, Fe³⁺, O₂), and ionic strength [34] . For example, to reduce the start-up time Zhang et al. [35] used a glass fiber separator in MFC, as it has a low oxygen transfer coefficient and a low charge transfer resistances. Zou et al. [36] reported a much faster start-up rate for polymerized riboflavin (RF) onto carbon cloth (CC) electrode. Feature 3 is the maximum anodic loading currents density that occurs after the lag period grows. Feature 3 is followed by a decrease in current production. The drop can be altered by the addition of food source, Figure 1(b).

Figure 2 shows the effect of addition of metal ions Mg², Cd²⁺, Pb²⁺, and Ca²⁺ on start-up time and maximum loading current density compared to the control. It can be observed that both the concentration of metal ion and the species of the metal ion are important factors in altering the electrochemical behavior. An increase in the metal ion concentration decreased the start-up time and increased maximum loading current density. The type of ion also had a significant impact on start-up time and maximum loading current density compared to the control.

The data show decreased start-up time consistent with literature experiments [34] [35] [36] . Here an increase in the metal ion concentration decreased the start-up time. This may be related to adhesion. Bacterial adhesion to the ITO electrode was observed by the scanning electron microscope (SEM) (Figure 3). Figure 3(a) the control (TSB + Bacteria) shows that few bacteria are attached to the electrode surface in the presence of growth media only. In contrast, the addition of Mg²⁺, Cd²⁺, Pb²⁺, and Ca²⁺ ions at 800 μM increased the bacterial attachment to the surface of the electrode Figures 3(b)-(e).

Confocal microscopy was used to measure percentage surface covered by S. oneidensis MR-1. ITO surface coverage was significantly enhanced by the addition of metal ions Mg²⁺, Cd²⁺, Pb²⁺, and Ca²⁺ compared to the control. The control resulted in a total of 10% surface area coverage. The surface area coverage increased over the control to approximately 51.6% for Mg²⁺, 64.7% for Cd²⁺, 70.2% for Pb²⁺, and 78.5% for Ca²⁺ at ion concentration of 800 μM.

There is a direct relationship between maximum loading current density and bacterial coverage at the electrode surface (Figure 1(a), Feature 3). Figure 4

shows the average of triplicate maximum loading current density vs % bacterial coverages. The relationship between maximum loading current density and % bacterial coverages increases linearly (R² = 0.99). Once attached, at the same % coverage, the same maximum loading current density is observed between all metal ions. The effect is not metal ion-specific. F test values are larger than 0.05 showing that this relationship. The metal ions, however, also show a specific effect on loading current. The maximum loading currents obtained (for coverages at 800 µM) for Mg²⁺, Cd²⁺, Pb²⁺, and Ca²⁺ were 2.25, 3.13, 3.58, and 4.09 times higher than that of the control. F tests show that these values were significantly different.

A second set of electrochemical data belongs to cyclic voltammetry obtained after electrode loading at +0.2 V vs Ag/AgCl for two hours. The first of 6 CVs are shown in Figure 5. The TSB only and metal ion blanks show no discernible peaks, as expected. It is observed that the reduction peak current is increased over the control when Mg²⁺, Cd²⁺, Pb²⁺, and Ca²⁺ are added. The shape of the CV suggests an adsorbed species [29] [37] [38] . The integrated peaks area under the reduction peak gives the amount of material that attached to the electrode surface.

Q ( C ) = n F N (1)

where (n) is the number of electrons transferred per molecule, (F) is Faraday’s constant (96,485 C∙mol⁻¹), and N is the amount of material electrolyzed in moles of either and or both the outer membrane cytochrome and the mediator flavin. Total Q should be related to the percent coverage.

Consistent with Figure 4, the increase in the measured cathodic peak charge vs. percentage of S. oneidensis MR-1 coverage was linear (R² = 0.99) for each metal (Figure 6). There is a slight difference in linearity as measured by the

slope on charge and the type metal ion. F ratio values are less than 0.05 indicating that there is a difference in the slopes between charge and % bacterial coverage. Above it was concluded when the cells are attached to the electrode surface, at a given total surface coverage, current output is non-specific between metal ions. Here it is noted that the ability to accept a charge (cathodic process) is slightly a metal ion-specific.

There are several hypotheses related to the increase in the metal ion concentration could increase bacterial attachment. Firstly, Shewanella surfaces are negatively charged at pH 7 [39] due to EPS structural anionic groups. Non-specific electrostatic forces may be applicable. Researchers have applied classical Derjaguin, Landau, Vervey, and Overbeek (DLVO) double layer theory to describe bacterial attachment to inorganic surfaces. The classical DLVO theory is described as the net interaction between a cell and a flat surface, which results from attractive van der Waals interactions and repulsive interactions between the electrical double layer of the cell and the substratum [40] . Choi et al. [26] studied the attachment and detachment of bacteria during various ionic strengths by using quartz sand surface. They found that increasing the ionic strength of the solution increased attachment. Similarly, Loosdrecht et al. [41] adhesion at polystyrene surfaces increased with increasing ionic strength of the electrolyte.

However, the theory does not adequately describe all of the observed behavior. Rijnaarts et al. [42] attempted to apply classical DLVO theory by controlling for the ionic strength and found only a moderate adherence to the theory. Rijnaarts et al. [42] study of bacterial adhesion on Teflon and glass surfaces hydrophobicity found different behavior at low ionic strength (0.001 M) and high ionic strength (0.1 M). At low ionic strength attachment was inhibited by DLVO-type electrostatic repulsion, but at high concentration attachment was controlled by steric interactions. We also found that classical DLVO theory failed to explain all the data in two ways. Ionic strength was held constant (0.09 M), which should have resulted in no alteration of adhesion. DLVO theory would suggest that at similar concentrations there would be no effect in changing the cation species. Thus, like Rijnaarts classical DLVO theory was found insufficient to explain the results.

Secondly, DLVO theory can be altered by considering the hydrophobicity. Extracellular polymeric substance (EPS) is an important component in controlling the hydrophobicity of the bacterial cell surface. Tsuneda et al. [43] reported that if the amount of EPS is small, cell adhesion onto solid surfaces is inhibited by electrostatic interaction, and if the amount of EPS is large, cell adhesion is promoted by polymeric interaction.

An explanation for the results relates to metal ion binding within the EPS. Metal ions are known to bind to EPS [44] [45] and, therefore, likely affect the properties of the EPS. EPS on the bacteria may extend through the barrier into a region where attraction can dominate. As an example, Redman et al. [46] suggested that when the bacteria have attached from the pili or flagellum that can occur across the barrier when the microbe lies within the secondary minimum. Different adhesion effects result when the configuration of EPS is either tight or loose.

In addition to the interaction of metal ions with EPS Wei et al. [47] have found carboxyl and phosphate groups were mostly responsible for Cd²⁺ adsorption on bacterial cells and affect the affinity between the bacterial cells and metal ions. Dobrowolski et al. [48] found vibration band alterations associated with carboxylic acids and alcohols when Cd²⁺ and Pb²⁺ were bound extracellular substances obtained from bacteria. These effects are, in the literature, also predicted to be non-specific and related to trends in hydrated ionic radii, hydrated radii, and hydration energies [49] [50] .

Thirdly, a supplemental explanation for the differences between metal ions is the impact of their water of hydration on binding to the EPS sites. As predicted, for a non-specific electrostatic mechanism, attachment increases with ionic radii, Figure 7. Mg²⁺, Cd²⁺, and Ca²⁺ follow the expected trend well. For these ions, bacterial coverage is linear with ionic radii (R² = 0.99). As expected, Ca²⁺ results in the highest current density. Pb²⁺ is out of sequence between Mg²⁺, Cd²⁺, and Ca²⁺. Inclusion of Pb²⁺ into the sequence resulted in a significantly poorer correlation between % bacterial coverage and ionic radii (R² = 0.79). This difference may be related to alterations in valence shell electrons. The structure of the hydrated Pb²⁺ ion is affected by the lone electron pair, which is giving complex hemi-directed in aqueous solution [51] .

Finally, metal ions may affect the metabolism of the bacteria. Cd²⁺ is known to be far more toxic than Pb²⁺. As the minimal inhibitory concentration (MIC) is 0.5 g/cm³ for Cd²⁺ compared to 5.0 g/cm³ for Pb²⁺. The effect is suggested to be a metabolic one in Cd²⁺ where Cd²⁺ interferes with both Ca²⁺ and Zn²⁺ related sulfhydryl group [27] . In order to check metabolism, several experiments were performed.

First, the optical density (O.D.) was measured from the extracted volume of growth media in the electrochemical cell (Figure 8). There is no observed difference in the optical density from the beginning to the end of the experiment between metal ions. This suggests that there is no toxic effect inhibiting microbial growth in solution. This is despite the difference in minimum inhibitory concentration (MIC) for the metal ions.

Second, HPLC was utilized to measure the glucose consumed during the electrochemical experiment (Figure 8). Glucose consumption was dependent on the type of metal ions in the cultures. Glucose concentrations found in solutions of 800 μM Mg²⁺, Cd²⁺, Pb²⁺, and Ca²⁺, are 0.45, 0.67, 0.85, and 1.53 g/l, respectively. These values are much higher than that consumed by the control 0.14 g/l.

Fitzgerald et al. [14] found that the lactate consumed was independent of the concentration of CaCl₂ in the medium.

Third, riboflavin production was monitored. Riboflavin has been identified as an indirect extracellular electron transfer (EET) mediator, acting as an electron shuttle for the EET of S. oneidensis MR-1 in MFCs [10] [52] . There is a difference between metal ions which may be metabolitic [14] [53] . Xu et al. [11] found that the riboflavin was increased with Cu²⁺ and Cd²⁺ (1.78 and 1.73 μM) compared to the control (1.46 μM). The concentration of riboflavin was measured by HPLC (Figure 8). Riboflavin concentrations found in solutions of 800 μM for Mg²⁺, Cd²⁺, Pb²⁺, and Ca²⁺ were 2.28, 3.08, 4.27, and 6.23 μM, respectively. These values are much higher than that produced by the control, 0.26 μM. As riboflavin is a mediator for electron transfer, increased riboflavin product should increase electrochemical activity of the bacteria. The trends were consistent with all of the electrochemical trends observed above. Both sets of experiments indicate that glucose (consumed) and riboflavin (produced) follow the same trends in the electrochemical behavior shown above.

The results implied that metal ions Mg²⁺, Cd²⁺, Pb²⁺, and Ca²⁺ enhanced bacteria attachment to the electrode surface and that might, in part, contribute to current output improvement. There are other factors related to the current output. For example, cellular metabolism and activity of electron transfer chain (e.g., c-type cytochrome). These factors might be influenced by metal ions Mg²⁺, Cd²⁺, Pb²⁺, and Ca²⁺ and might play a role in current output improvement.

This study shows that the current output generation from Shewanella. oneidensis MR-1 is affected by the addition of metal ions. Heavy metals ions are expected to have a toxic effect on bacterial growth. In this study, heavy metals had no effect in optical density of extracted solutions indicating the bacterial growth in the solution was unaffected. Mg²⁺, Cd²⁺, Pb²⁺, and Ca²⁺ significantly increased current output compared to that of the control. Further analyses indicated that metal ions increased anodic biofilm formation. Once attached, a linear relationship was found between the % bacterial coverage and electrochemical performance. This may be related to alterations in the EPS structure, as observed elsewhere. These results suggest avenues of research in the improvement of performance and operation of MFCs.

The author sincerely acknowledges scholarship from Al-Baha University for the doctoral program. The author would like to thank the Department of Chemistry at Northeastern Illinois University for providing the HPLC instrument for sample analysis. This research utilized a scanning electron microscope, which was funded by the NSF, Major Research Instrumentation (MRI) Program via Award No. 1726994.

The authors declare no conflicts of interest relevant to this article.

Alshahrani, A., Fitch, A. and Al-Bazi, J. (2019) Metal Ions Impact on Shewanella Oneidensis MR-1 Adhesion to ITO Electrode and the Enhancement of Current Output. American Journal of Analytical Chemistry, 10, 428-443. https://doi.org/10.4236/ajac.2019.109031