THE INFLUENCE OF EXTERNAL LOAD ON THE PERFORMANCE 1 OF MICROBIAL FUEL CELLS

22 In this work the effect of the external load on the current and power generation as well as 23 on the pollutant removal by a microbial fuel cell (MFC) has been studied by step-wise 24 modifying the external load. The modification was composed by a direct scan, in which the 25 external load was increased from 120 to 3300 Ω , and a subsequent reverse scan in which the 26 external load was decreased to 120 Ω . The obtained results indicated that the exerted 27 electrical current decreased when the external resistance increased. This reduction in the 28 current was even maintained in the reverse scan, when the external resistance was step-29 wise decreased. Regarding to the power exerted, when the external resistance was 30 increased, below the value of the internal resistance, an enhancement of the power exerted 31 was observed. However, when operating nearby the values of the internal resistance a 32 stable power exerted was reached, about 1.6·10 -3 mW. In the reverse scan, the power 33 exerted decreased in all the cases. The current and power behavior was explained by the 34 change in the population distribution which shift to a more fermentative than electrogenic 35 culture. 36


INTRODUCTION
The increasing world population and the rising standard of living are the main reasons for the increase in energy demand.The increase of the energy consumption is accompanied by the deterioration of the natural environment.With the aim to protect the environment, a number of steps have been taken to minimize the negative effects of the energy production.
During the last years, attention was drawn to alternative energy sources that are abundant and environment-friendly (Beegle & Borole, 2018).The solution to this problem is related to the consumption of Renewable Energy Sources.Unfortunately, the energy production of some of the renewable energy sources depend on atmospheric conditions, which makes it very difficult to use these resources during peak demand.Because of that, the need for continuity of energy supply could not be met without energy storage and/or the use of alternative fuels available at any time.
About the use of alternative technologies, the use of fuel cells stand up as a promising option.The main advantage over other conventional renewable energy sources, such as sunlight or wind energy, are the possibility of uninterrupted operation and the energy production independent of weather fluctuations.In addition, the superiority of fuel cells over other conventional energy sources are amongst other the zero emission of harmful toxic gases (SOx, NOx) and of solid particles (O'Hayre et al., 2006;Slate et al., 2019).The advantages also include the simple structure of the cell and the mechanism of its operation.
Due to the direct conversion of chemical energy into electricity, fuel cells owe higher gross efficiency to traditional generators or power units (O'Hayre et al., 2006).
A version of the fuel cells are the Microbial Fuel Cell (MFC).In the MFC, microbial cultures can be involved in the oxidation/reduction of the electron donor/acceptor.Usually, the MFC consist of anode and cathode chambers, physically separated by a Proton Exchange Membrane (PEM).In most of the cases, an active microbial culture located at the anode oxidizes organic substrates releasing electrons to the anode and protons to the liquid bulk.
The electrons flow through an external electrical circuit, generating an electrical current, and the protons are transported to the cathode through the membrane (Gonzalez del Campo et al., 2014b;Jung et al., 2007).At the cathode, the electrons and the protons are used to reduce and oxidant, usually oxygen (Mateo et al., 2015a).The technology of producing electricity from bacteria was already known in 1911 by M. C. Potter (Potter, 1911) but only at the turn of the decades began to be seriously considered, being very relevant during the last twenty years (Shabani et al., 2020).
The major bottleneck of MFC application is its relatively low power density, the present power output density is not high enough for industrial applications (Mateo et al., 2018a).In this way, MFC power density can be increased by optimizing operating conditions such as COD, pH, flow rate, temperature or configuration options such as electrodic area, external resistance, etc. (Aelterman et al., 2008;Gonzalez del Campo et al., 2013b;Ieropoulos et al., 2010;Mateo et al., 2016;Rozendal et al., 2008).
In the MFC configuration, the external resistance is used to dissipate the electrical energy when MFCs are operated independently of an electrical device, and as an integrated part of an electrical grid that controls the characteristic outputs of fuel cells (Rismani-Yazdi et al., 2011).The external resistance controls the ratio between the current generation and the cell voltage (Logan et al., 2006).A high external resistance results in a high cell voltage and low current; the opposite is observed when operating with low external resistance.According to the literature, a way to minimize losses is to operate the MFC at the optimal conditions for power production which is related the external resistance of the MFC (Clauwaert et al., 2008).According to the Jacobi's law, the maximum external power will be obtained once the external resistance equal the internal resistance of the MFC.
One of the main problems of the MFC technology is to maximize the power output and the development of future power management system, the study of the external resistance is very relevant for engineers (Katuri et al., 2011;Lyon et al., 2010).
MFC operated under different external loads indicates that this is one of the most important factors conditioning the short and long-term performance of the MFC.In the literature it has been described the effect of external load operating in the batch mode (Koók et al., 2020).
However, for full-scale implementation, as well as for process automatization, it is required to study the continuous operation of MFC.Because of that, the aim of this paper is to study the effect of long-term modification of the external resistance on the performance of MFCs.
Attention was paid to the electricity production, fuel consumption and microorganism's population of the anodic compartment in order to evaluate the effect of the external resistance.The main contribution of the present work is the approach the relation between biotic and electric features of MFC under continuous operation.

MATERIALS AND METHODS
In the order to study the influence of external load on the performance of an MFC operating continuously, different external loads were connected to the MFC.These experiments were carried out at constant influent flow rate of 7.2 L / d.The cycle of the entire experiment was 92 days.The initial external load, 120 Ω, was step-wise increased to 560, 1000, 1500, 2200, 2700 and 3300 Ω in the direct scan.After that, the value of the external load was step-wise decreased back to the initial external load, 120 Ω.

Experimental set-up
In this work, a two chambered MFC was used.The anodic and cathodic chambers were built on graphite plate, the volume of the anodic and the cathodic chambers were 0.95 cm 3 and 0.5 cm 3 respectively.The compartments of the MFC were separated by a, 180 μm thick, Sterion® proton exchange membrane (PEM) with high ion exchange capacity (0.9-0.02 meq g -1 ), high ionic conductivity (8•10 -2 S/cm) and low electronic conductivity (<10 -10 S cm -1 ).As anodic and cathodic electrodes Toray carbon papersTGPH-120 (E-TEK, USA) were used.The projected superficial area of the anodic electrode was 4.65 cm 2 and that of the cathodic electrode was 2.85 cm 2 .The anodic and cathodic electrodes contained a 20% and a 10% of Teflon respectively in order to improve its mechanical properties causing a very small drop in the electrochemical performance.At the cathode, a catalytic layer with 0.5 mg Pt/cm 2 loading was deposited with the aim to enhance the electrode performance (Gonzalez del Campo et al., 2014a).The electrodes and the membranes were coupled as a Membraneelectrode Assembly (MEA), the MEA was performed according to literature (Mateo et al., 2016).The cathodic and the anodic electrodes were connected by means of an external load (Rext) of 120 Ω (initial conditions); The load of the external resistance was selected in order to reduce as much as possible the activation losses and to make easier the electron transfer during the startup stage (Yi et al., 2019).
A scheme of the set-up is presented in Figure 1.The anodic compartment was inoculated as previously described in the literature (Mateo et al., 2018b).After acclimatization, the MFC was operated in continuous mode.To do that, the anodic chamber was fed with a medium containing a synthetic wastewater at 0.5 ml/min.
The composition of the synthetic wastewater used in the experiments is presented in Table 1.In order to avoid the degradation of the wastewater during its storage, it was sterilized for 30 min at 105°C (Gonzalez del Campo et al., 2014b).An air breathing cathode was used because of its improved performance according to the literature (Mateo et al., 2015b;Wang et al., 2017).

Characterization techniques
The MFC operation was continuously monitored by means of a digital multimeter connected to the external load (Rext).The potential (V) was directly related to the electrical current (I) between the edges of the external load by the Ohms law (Equation 1).
Power was calculated using Equation 2, where, I is current and V is voltage.
In this set-up, the electrical current was generated due to the oxidation of the organic matter contained in the wastewater.This oxidation process was monitored by measuring the COD removal percentage (% COD) and the COD removal rate (rCOD).These parameters can be calculated according to Equations 3 and 4, respectively, where Q is the flow rate to the anodic chamber, ∆COD is the COD removed from the wastewater and CODo is the COD initial of the wastewater (343 mg/l).
The COD concentration in the anodic effluent was measured by spectrophotometric With regard to the biomass characterization, a MALDI-TOF AXIMA Assurance by Shimadzu was used.The matrix solutions were prepared by saturating α-cyano-4-hydroxycinnamic acid in a 1:48:2 acetonitrile:water:trifluoroacetic acid matrix solution.Then, the microorganisms were sterilized using ethanol at 75%, and the solution was centrifuged at 1000 rpm for ten minutes.After centrifugation, the supernatant was removed, and the biomass recovered from the precipitate using 20 µL of an acetonitrile/formic acid/water (50:35:15) solution according to the procedure described in the literature (Mateo et al., 2016).
Additionally, the concentration of acetic, propionic and butyric acids in the effluents were determined by gas chromatography (Perkin Elmer) with flame ionization detector (FID) using a Crossbond Carbowax Column (15 m x 0.32 mmID, 0.25 mm df) (Fernandez-Morales et al., 2010).The oven temperature was 140 °C for 1.5 min, followed by a ramp of 25°C/min until the temperatures reached 190°C and it was kept for 2 min.The temperature of the injector and detector were 200 and 230°C, respectively.Nitrogen was used as a carrier gas.Also, lactic and formic acid concentrations were determined by HPLC (Agilent) equipped with an ultraviolet diode array detection (UV-DAD) and a Zorbax SBAq column (4.6 x 150 mm 5 µm).
Mobile phase was a buffer pH 2 (0.02 M phosphate) composed by 99 vol.% of water and 1 vol.% of acetonitrile.

RESULTS AND DISCUSSION
In this study, the influence of the external load on the electricity production, fuel consumption, wastewater treatment capacity and microbial culture population was evaluated by varying the external load coupled to the electrical circuit of the MFC.With this aim, the external load was step-wise increased from 120 Ω to 560, 1000, 1500, 2200, 2700 and 3300 Ω.The external resistance of 3300 Ω was selected in order to overcome the internal resistance of the MFC used in this study, which was about 2200 Ω.After that, the value of the external resistance was step-wise decreased back to 120 Ω, in order to analyze the effect of the reduction of the external load of the MFC.
In order to ensure the reproducibility of the tests, each external resistance value was kept in the MFC until the exerted voltage reached a steady state response.Once the response reached the steady state, the electrical performance of the MFC as well as the pH, COD and microorganisms concentration of the effluent were determined.

External load effect on electricity production
The effect of the external load on electricity generation was studied taking into account the current and power exerted.

Current generated by the cell
In the Figure 2, it is presented the electrical current exerted by the cell when changing the external load.In order to ensure that the steady state was reached, the external load was kept constant during at least 7 days.
Figure 2. Current exerted by the MFC at every external load.
In the Figure 2, it can be observed that the current decreased when the external load increased.The greatest changes were obtained when the resistance increased from 560 to 1500 Ω.However, from 2200 Ω the current remained almost constant, at around 0.02 mA, until the end of the direct scan when the external load reached 3300 Ω.This trend can be explained because the external load controls the ratio between the current generation and the cell voltage (Aelterman et al., 2008).In the literature it has been described that high external loads result in high cell voltages and low currents (Katuri et al., 2011).
Once the direct scan had been finished, a reverse scan was performed with the aim to compare the intensity obtained when the external resistance was step-wise increased and then step-wise decreased.During the reverse scan, contrary to the expectations, the intensity exerted by the MFC did not increase.The current was kept almost constant around 0.02 mA independently of the external resistance applied.The current generated by MFC for each external load was higher in the direct than in the reverse scan.In other words, after the operation with high external loads, the intensity exerted decreased.This behavior indicated that the system had irreversibly changed during the experiment.This change only could be explained by a change in the population distribution.In the literature, it has been suggested that external load changes influence the growth and population distribution in the biotic compartments of the MFC (Pinto et al., 2010).In order to ratify that, the microbial population was analyzed by means of maldi-tof observing the apparition of clostridium when the external load increased.This behavior has been also identified in other recent research works, (Kook et al., 2020).These fermenter microorganisms present a higher growth rate, about 0.1 h -1 (Sjöblom et al., 2015), than the electrogenic ones, about 0.05 h -1 (Zhang et al., 2014), which could explain why the population did not return to the initial conditions when the external load was decreased.

Power generated by the cell
The Jacobi theorem states that when the external load is equal to the internal resistance, the maximum power is obtained.Because of that, a polarization curve was carried out to determine the internal resistance of the MFC.From this analysis, the internal resistance was determined being its value 2210 Ω.In order to evaluate Jacobi theorem, the Figure 3 presents the power generated at every external resistance.As can be seen in Figure 3, when increasing the external resistance from 120 to 560 Ω a great increment in power exerted was observed.The power exerted increased from 3.8•10 -4 to 1.6•10 -3 mW.Then, the subsequent step-wise increments in the external resistance slightly increased the power exerted, reaching a maximum at about 2200 Ω, exerting 1,6•10 -3 mW.When the external resistance was step-wise decreased, it was observed a significant reduction in the power exerted from 1.6•10 -3 to 4.8•10 -5 mW.It is important to highlight that this behavior generates a hysteresis loop.Because of that, the power exerted for each external resistance was always higher in the direct scan, when the external resistance was step-wise increased, compared to operating in reverse scan, when the external resistance was step-wise decreased.In the literature, it has been proposed that these changes could be explained by changes in the distribution of the microbial population of the culture when the external load changed (Lyon et al., 2010).It could be the reason for the reduction in the power exerted after the operation under high external loads.In the literature it has been described that differences observed in the anodic potentials when the MFC is operated at different external loads acts as a selection force of the different electrogenic microorganisms (Schröder, 2007).These changes in the microbial distribution could modify the activation losses and the internal resistance of the anode, which are function of electrochemical activity of the anodic electrogenic microorganisms (Rismani-Yazdi et al., 2011).In this work, as described above, the microbial population distribution changed after the experiments, showing a significant growth of fermenters microorganisms mainly from the Clostridium genus.

External load effect on fuel consumption
In the MFC, the microorganisms of the anodic chamber oxidize the fuel to carry out its anabolic and catabolic reactions.As the changes in the external resistance influences the performance of the microbial culture of the MFC, it was considered interesting to evaluate its influence over the fuel consumption.It must be stated that in this case the fuel was wastewater, being the fuel the biodegradable organic substrates contained in the wastewater.In order to evaluate the fuel consumption, the Chemical Oxygen Demand (COD) of the wastewater used as fuel was determined before and after its reaction in the MFC.
Figure 4, presents the COD removal as a function of the external resistance applied.As can be seen, the fuel consumption slightly increased when the internal resistance was increased below the value of the internal resistance of the MFC 2200 Ω.However, when the external resistance was higher than this value, the fuel consumption started to decrease.This decrement was maintained even when the external resistance values were decreased during the reverse scan.Taking into account that the electricity production also decreases, it was considered interesting to determine the Coulombic Efficiency (CE).The CE is a measurement that indicates the percentage of electron equivalents converted into electricity when oxidizing a fuel in the MFC.From the values of the CE obtained it was observed that the CE decreased along the experiments.This behavior could be explained by changes in the population's distribution of the microbial culture previously described.In this case, the reduction in the CE could be explained because the growth of non-electrogenic microorganisms outcompete the growth of the electrogenic ones.In the Figure 5, it can be seen that the effluent microorganisms concentration remained unaltered when the external resistance was modified.However, taking into account that the current generated at high external resistances was lower, it can be concluded that the population shift to a less electrogenic culture.
Regarding to the electrolyte pH, it could play an important role on the performance of MFCs.
It is known that pH is an important parameter for microorganisms metabolisms because they can only carry out their vital functions within a limited pH range (Infantes et al., 2012).For this reason, most MFCs operated at neutral pH to optimize bacterial growth conditions (Gonzalez del Campo et al., 2013a).The pH can influence oxidative metabolisms and in turn affect the electron transference to the electrode and the proton generation mechanisms (Behera & Ghangrekar, 2009).Values outside the 6.5-8.5 range are known to strongly affect microorganism metabolism.However, the low concentration of protons at about neutral pH makes the internal resistance of the cell relatively high, especially if compared to that of conventional chemical PEM fuel cells that use acidic electrolytes.
In the Figure 5, it can be seen that pH was maintained constant, around 6.7, when external resistance was lower than 2000 Ω, this pH is an optimal pH for the growth of the electrogenic microorganisms.Then, when external resistance was increased over the internal one, the pH decreased until 5.8.These results indicates an enhancement of the acidogenic metabolisms when the external resistance overcome the value of the internal one.Then, in the reverse scan, when external resistance was decreased from 3300 until 120 Ω, the pH was maintained at acidic values, out of the optimum range of the electrogenic microorganisms.This behavior could be explained by the development of acidogenic microorganisms, which release VFA as result of the acidogenic fermentation of the substrate, mainly when operating at external loads with resistance higher than the internal one.These results indicate an irreversible change in the population distribution during the direct scan that was maintained even during the reverse scan.This hypothesis was ratified by the MALDI-TOF analysis previously performed that identified the appearance of clostridium microorganisms when the internal resistance increases over the internal resistance.
In the literature, Pinto and al. predicted that independent of the microbial composition of the inoculum, the development of the electrogenic culture could only be achieved when the external resistance values are the same or lower than the MFC internal resistance.In this way, the lower values of external loads facilitate the electron transfer process, facilitating therefore the development of the electrogenic culture (Pinto et al., 2011), whereas high values of external resistance facilitate the development of non-electrogenic cultures.
In order to verify this hypothesis, the VFA released to the liquid bulk were analyzed.
Finally, in order to determine if the drop of pH was caused by the VFA production, its concentration in the effluent was analyzed.From the analysis, acetic and propionic acid were identified as the main VFA contained in the anodic effluent.In the Figure 6, the VFA concentrations for each external load are presented.The acetic acid concentration was very similar, about 0.8 mM, for all the external loads studied.However, the propionic acid concentration significantly increases from 0 to 0.6 mM when the external load was increased.Then, when the external load was decreased a slight reduction of the propionic concentration was observed.As result, an increasing and then slightly decreasing VFA concentration was observed and presented in   , 2011).These observations also agree qualitatively with microbial population modelling and experimental works previously published by Picioreanu et al., which showed that an increased external load favors the growth of the anaerobic microorganisms (Picioreanu et al., 2007;Picioreanu et al., 2008).The increased production of VFA indicates a shift from an electrogenic culture to a fermentative one when the system operates under higher external resistances.This can be explained by the fact that the higher the external load, the more difficult the electron transference to the anode and therefore the lower the electron transfer rate.This difficulty causes a lower growth rate of the electrogenic microorganisms which was reflected in lower current and the electrical energy generation (Katuri et al., 2011).
In order to identify the contributions to the fuel consumption in the MFC, a fuel mass balance based on COD measurements was carried out.From the influent COD, the effluent COD, and taking into account the VFA COD contribution to the effluent, the net fuel consumption in the electrogenic and fermentative metabolisms were determined.In the Figure 6, the different COD contributions are presented as a function of external load.
As explained above, effluent COD decreased when operating at external loads.On the one hand, in the Figure 6 is demonstrated that the VFA contribution to the effluent COD increases when external load of the system is increased.Because of the fermentative metabolisms, the raw substrate COD in the effluent decreases from 200 to 100 mgO2/l when external load was increased from 120 to 3300 Ω.
Thus, it can be concluded that the increment of external load causes an increment of the VFA production and a decrease in the effluent substrate concentration due to the contribution of the non-electrogenic fermentative culture.
In order to evaluate the net fuel consumption, the substrate removal rate was calculated ranging their values from 3 to 4.5 mg/h when the external load was increased from 120 to 3300 Ω.This increase can be explained by the non-electrogenic metabolisms observed at higher external resistances.Hence, the maximum pollutant removal from the wastewater of the MFC was obtained when operating at the highest external load.Although these operational conditions do not allow the system to reach the highest electricity generation.

Conclusions
From the results obtained in the present study, the following conclusions can be extracted.
The external load modified the population distribution in the MFC.The microbial population evolve from a more electrogenic to a more fermenting microbial culture.This modification in the population distribution affect all the operational variables, being mainly reflected on the current density exerted which decreased from mA to mA.With regard to the power exerted, the system behaves as predicted by the Jacobi Theorem although the peak in the power production was obtained in a wide range, from 1000 to 3000 Ω, instead of being located at the point were the external and internal resistance meet.Finally, with regard to the fuel consumption, its value was almost the same in all the cases but a slight decreased was observed after the operation at high external loads.Taking into account that the electricity production significantly decreases but the fuel consumption slightly decreased it can be concluded that the system yields a lower electrical efficiency after the operation at high external loads.
methods.Volatile suspended solids were measured according to the literature (American Public Health Association, 2005).Conductivity and pH were measured with a Jenway 470 conductivity meter and a PCE-228 pH meter, respectively.

Figure 3 .
Figure 3. Power exerted by the MFC at every external load.

Figure 4 .
Figure 4. Fuel consumption in steady state at every external resistance.

Figure 5 .
Figure 5. Volatile suspended solids concentration and pH of effluent at every external

Fig 6 .
The increasing VFA concentration a its dissociation explained the pH drop observed(Fernandez-Morales et al., 2010).

Figure 6 .
Figure 6.Acids COD, substrate COD and total COD of effluent at every external resistance.

Table 1 .
Composition of the synthetic wastewater.