This chapter describes biobattery start-up and results of real-time power output optimization during biobattery exposure to changes in temperature, salinity and carbon source availability. Also, changes in the biobattery electrochemical parameters are discussed. Finally, results obtained during the long-term biobattery operation are described.
3.1. Biobattery Start-Up
The biobattery operation was initiated in open circuit mode. The open circuit voltage (OCV) of each electrode pair was observed to rapidly increase reaching 450–520 mV after first 9 days of biobattery operation. The mixed microbial community of the solid anolyte composed of humus and saw dust contained a sufficient quantity of anodophilic microorganisms, which colonized the carbon felt electrodes forming an anodophilic biofilm [
22]. Biofilm formation at the anode of a microbial fuel cell has been extensively studied [
23,
24,
25] showing progressive development of the anodophilic biofilm. Upon connecting 1000 Ohm resistors to each electrode pair on day 9, an average power output of 0.25 mW was obtained almost instantly (
Figure 3). The short startup time can be attributed to a rapid development of the anodophilic biofilm and the associated increase of internal capacitance [
15].
All electrode pairs were connected to a single electrical load on day 24 (
Figure 3) and the P/O algorithm for maximizing power output by dynamically adjusting the
Vhigh (MFC open circuit voltage at which the electrical load is connected) was started. Details of the P/O algorithm are provided in the Materials and Methods section. Essentially, this optimization algorithm matched the external resistance (electrical load) to the internal resistance of the biobattery. It should be mentioned that in order to increase the biobattery current all electrode pairs were connected in parallel to the electrical load, as opposed to connection in series, which can be used to obtain a higher output voltage. Even minor differences in carbon source distribution and availability within the solid anolyte could affect electrochemical characteristics (OCV, internal resistance and capacitance) of each electrode pair. If electrodes are connected in series, these differences could lead to inferior MFC performance due to the internal resistance mismatch, as demonstrated in several studies [
26,
27,
28]. In addition, four circuits for on/off connection of the electrical load would be required. Consequently, parallel, rather than in series electrode connection was chosen for the test.
The 3D biobattery power output observed during the first two days after parallel connection of the electrodes and startup of the P/O algorithm for power output maximization varied between 1.3 mW and 1.5 mW (
Figure 3). This is higher than a summation of the power outputs from each electrode pair connected to a fixed 1000 Ohm resistor. This performance improvement was achieved due to real-time
Vhigh optimization. Also, a decrease of the internal resistance due to growth of anodophilic microorganisms under optimal operating conditions can be hypothesized [
29]. Over time, power output decreases until it reaches 0.51 mW (1.05 mW L
−1, where power density is calculated using an anode volume of 476 mL), about 44 days from start-up (
Figure 3). This volumetric power density is comparable to an earlier study using the same solid anolyte in an MFC with a 70 mL anode compartment volume. High power output immediately after connection of the electrical load was followed by a decrease in energy production attributable to carbon source–limited conditions due to fast consumption of the available dissolved organic matter and slow hydrolysis of solid organic materials.
This power generation profile follows the same trend as in our earlier study with the same solid anolyte [
9] as well as another study that evaluated performance of a SA-MFC on agar [
11].
3.2. The Impact of Environmental Conditions on Power Production
Practical applications of 3D biobattery for powering environmental sensors could lead to biobattery exposure to various environmental conditions that might affect electricity generation, such as changes in temperature, salt content (salinity) of water in the anode compartment, and organic matter availability. Accordingly, performance of the biobattery equipped with real-time power-output optimizing algorithm was evaluated in a broad range of these environmental conditions. In particular, conductivity and organic matter availability was changed by adding either NaCl or carbon source solutions to the water holding reservoir of the 3D biobattery. Also, temperature was varied by placing the biobattery in a temperature-controlled chamber or freezing a smaller (2D) biobattery, as described in the Materials and Methods section.
Higher conductivity could enhance MFC performance due to reduced ohmic resistance. At the same time, an increase in salt concentration can negatively affect Coulombic efficiency due to the sensitivity of freshwater anodophilic microorganisms to NaCl [
30]. Organic matter availability is another important limitation in SA-MFCs, therefore the presence of even relatively low dissolved organic carbon in water is expected to boost power production therefore affecting the optimal operating point. Performance testing under differing environmental conditions was initiated with an increase in anolyte conductivity by adding NaCl to water reservoir of the 3D biobattery. Once NaCl was added, the anolyte conductivity started to increase due to a combination of water percolation and evaporation through cathode walls. Anolyte liquid conductivity measurements obtained by withdrawing liquid sample from the middle section of the biobattery showed an increase from 0.03 mS cm
−1 to 0.8 mS cm
−1. The increase in conductivity was accompanied by initial increase of power output, followed by a power output decrease to 0.36 mW (
Figure 4A). The initial increase in the power output from 0.46 mW to 0.61 mW can be attributed to a decrease in ohmic resistance of the amended anolyte solution. Indeed, laboratory MFCs are typically operated at a conductivity much greater than 1 mS cm
−1 [
31], therefore at a conductivity of 0.03 mS cm
−1, high ohmic losses are expected. As mentioned above, addition of NaCl increases anolyte conductivity, but also could inhibit several microbial trophic groups responsible for organic matter hydrolysis [
30]. NaCl toxicity towards hydrolyzing microorganisms could explain the observed decrease in power production in spite of the increased anolyte conductivity during biobattery operation at the highest NaCl concentration. Overall, it is concluded that the biobattery performance can be improved by adding salts and nutrients to dilution water, but the composition of nutrients should be optimized to avoid NaCl-related toxicity.
Temperature is another important factor affecting biobattery performance by influencing microbial growth and metabolic activity as well as the oxygen reduction reaction (ORR) rate of the cathode. Biological activity is often positively correlated with temperature as long as the temperature does not exceed the optimal point [
32]. In the experiment aimed at studying the sensitivity of the 3D biobattery to temperature variations, the temperature was decreased from 25 °C to 4 °C in two steps. Accordingly, the biobattery power output decreased from 0.3 mW to 0.01 mW (
Figure 4B). Interestingly, a spike in power production (0.59 mW) was noticeable upon return to biobattery operation at 25 °C. Apparently, carbon source hydrolysis continued at low temperatures, while the consumption rate was reduced, thus leading to a power output surge once the temperature was increased. The surplus carbon source was quickly consumed, and power returned to approximately the same value of 0.37 mW.
A biobattery can be deployed in a region with below zero temperatures during winter months. Although biofilms can be adapted to low temperatures, biological activity stops once the biofilm is frozen. The resumption of bioelectrochemical activity once temperature increases (i.e., in the spring) and the effects of solid anolyte freezing on biobattery performance were investigated using a smaller 2D (single electrode pair) biobattery placed in a −30 °C chamber for 5 days and then removed and thawed at room temperature for 24 h. Polarization curves acquired before the test and after thawing were used to evaluate the impact of the freeze-thaw cycle on the biobattery performance.
Figure 4C compares polarization curves before freezing and after thawing. The OCV remained nearly unchanged at approximately 352 mV and the maximum power densities were also similar at 1.1 mW L
−1. The estimated internal resistance of 294 Ω before freezing increased slightly to 315 Ω after thawing. Overall, the ability of the biobattery (SA-MFC) to resume performance within short time after freezing shows its resilience and suitability for field applications in temperate regions.
Another important factor in determining MFC performance is organic matter availability. Power production in an MFC was observed to follow a Monod-type dependence on carbon source concentration [
33]. In a solid anolyte MFC dissolved organic matter is produced through hydrolysis of solids, which is a rate-limiting step. A substantial increase in power when there is an influx of organic matter is therefore expected. To test the impact of carbon source availability on the biobattery performance, the water reservoir was filled with diluted brewery wastewater. As the carbon source percolation from the reservoir (with subsequent concentrations of 0.2 g L
−1, 1 g L
−1 and 5 g L
−1 of brewery wastewater) began to increase organic matter availability in the biobattery, the power output increased from an average of 0.23 mW to 0.90 mW. (
Figure 4D). This test proved carbon source availability as one of the most limiting factors determining MFC power output.
Overall, the tests showed that power production in the 3D biobattery is strongly influenced by the environmental conditions and is expected to significantly vary over time. To maximize power output, the biobattery was operated with periodic connection of the electrical load with real time optimization of Vhigh (MFC OCV voltage at which the electrical load is reconnected). This approach proved to provide stable biobattery performance at all tested conditions, including biobattery exposure to low temperatures and carbon source availability fluctuations.
3.3. Real Time Biobattery Performance Diagnostics
In addition to enabling optimal performance of the biobattery under changing environmental conditions, periodic connection of the external resistor at an optimal value of
Vhigh (on/off mode of operation) also facilitated real time estimation of key internal parameters, such as open circuit voltage (OCV), resistance, and capacitance. Such real-time internal parameters monitoring allowed for constant performance diagnostics. The approach was implemented using the electrical equivalent circuit (EEC) model [
15]. Throughout all tests, the parameter estimation procedure described in Materials and Methods was carried out at six-hour intervals.
Overall, real-time monitoring of internal parameters provided timely detection of all external disturbances. For example, the change in anolyte conductivity due to the addition of NaCl was observed as a reduction of ohmic resistance (
R1) from 8.68 to 6.65 Ohm, as can be seen from the on-line monitoring results shown in
Figure 5A. In addition, the
R2 component of the internal resistance, which can be related to activation and concentration losses, decreased from 36.5 to 28.2 Ohm, suggesting enhanced electron transfer to carbon felt anodes. (
Figure 5B). However, the OCV and the internal capacitance, which could be related to the activity of the anodophilic microorganisms [
15] were reduced (
Figure 5C). Average OCV decreased from 396 mV to 369 mV, while internal capacitance estimation decreased from 5.1 F to 3.0 F, indicating that the anodophilic microorganisms were inhibited at the increased concentration of NaCl. Although anodophilic bacteria, such as
Geobacter spp., may withstand NaCl content up to 10 g L
−1 under carbon source replete conditions [
30], soil microorganisms can be stressed by increased salinity [
34]. Since these soil microorganisms are likely responsible for organic matter hydrolysis in the solid anolyte MFC, a higher NaCl content (salinity) could lead to decreased power production. Importantly, the salt content of water in the anode compartment of a biobattery could vary significantly during field applications. The real-time performance diagnostics enables timely detection of increasing salt content before the onset of microbial activity inhibition.
Temperature variation is another example of external disturbance significantly affecting biobattery performance. The effect of 3D biobattery exposure to progressively lower temperatures (from 25 °C to 4 °C) on the estimated internal parameters can be seen in
Figure 5D. The temperature decrease is reflected by the initial increase of
R1 from 7.0 to 105.0 Ohm, apparently due to a lower microbial activity. On the other hand, average
R2 reduced from 28.2 to 9.8 Ohm. As discussed previously, it can be hypothesized that organic matter degradation continued at low temperatures, while the consumption rate decreased, thus somewhat increasing the availability as the temperature dropped (
Figure 5E). Nevertheless, a general increase in the total internal resistance (
Rint =
R1 +
R2) of the biobattery due to the dominance of ohmic losses led to a significantly lower power output. It should be mentioned that a prolonged exposure of mixed microbial populations of the biobattery to low temperatures could result in microbial community adaptation to psychrophilic conditions [
25], and improved power output with time. By using real-time parameter monitoring, such adaptation to low temperatures could be clearly observed. In fact, the observed increase in the biobattery capacitance from 3 F to 20.6 F over the duration of this experiment could be attributed to increased biofilm thickness. However, the increased capacitance did not lead to higher power production, perhaps due to high ohmic losses. Also, low temperature led to a decline in the OCV from 367 mV to 145 mV (
Figure 5F). Increasing the temperature from 4 °C to 25 °C decreased the internal resistance from 114.92 to 76 Ohm, reduced capacitance to 9.2 F and increased OCV to 285 mV. When the temperature was returned to 25 °C,
R2 resistance was only slightly higher than before the low temperature test.
Brewery wastewater is rich in organic matter and high in conductivity, factors which can improve MFC performance [
3]. As expected, signs of improved performance included reduction of
R1 from 12.7 to 8.2 Ohm (
Figure 5G,H). Improved organic matter availability is expected to increase anodophilic microbial populations and, by extension, power production. Indeed, internal capacitance increased (from 17.1 to 32.1 F) as did OCV (from 235 mV to 343 mV,
Figure 5I). It should be noted that increased growth of microbial populations (increased biofilm thickness) may cause an accumulation of extracellular polymeric substances (EPS) within the solid anolyte matrix of the biobattery [
35], thereby reducing transport of water and ions in the solid anolyte. The reduced species transport could also affect the rate of hydrolysis, and low pH zones might be created [
36]. This could explain why the average
R2, which can be associated with diffusion limitations, increased from 60.0 to 82.7 Ohm towards the end of the experiment and power production declined.
Importantly, the real-time biobattery performance diagnostics successfully detected all external disturbances. When initial operating conditions were restored (e.g., in the temperature test), internal biobattery parameters also returned to initially estimated values. Analysis of the estimated internal parameters confirmed the importance of real-time monitoring, which enabled timely diagnostics of key performance issues, such as increasing salinity, changes in carbon source composition, and the related changes in internal resistance and capacitance. For example, a sharp drop in internal capacitance is indicative of biofilm deterioration, while an increase in internal resistance is indicative of carbon source limitation (solid anolyte deterioration) [
37]. However, care must be taken to consider multiple factors, such as simultaneous temperature and salinity fluctuations in the macro environment of the biobattery. In addition, this information can be used to schedule biobattery maintenance (e.g., cleaning or replacing the electrodes and solid anolyte).
3.4. Long-Term Biobattery Operation
After the completion of experiments aimed at studying the 3D biobattery response to various external perturbations, the biobattery operation was continued for a total of 388 days (run 1) as shown in
Figure 6. Data points corresponding to biobattery operation at different temperatures, NaCl concentration, and carbon source addition were excluded in this graph. As can be seen from this plot, the power output decreased with time, and was around 0.1 mW after 388 days of operation. Thereafter, the solid anolyte was replaced and the biobattery was operated for another 123 days (run 2). The power output profiles in both runs were similar, reflecting progressive depletion of the solid carbon source (fuel) in the biobattery.
As suggested by Westrich et al. [
38], solid anolyte organic matter can be considered to include a high-reactive fraction and a less-reactive fraction. Power production can thus be expected to be high in the first few months of operation due to a relatively fast hydrolysis of the high-reactive fraction. Once the less-reactive fraction is dominant, power production is expected to decrease. It can be suggested that it is the slow hydrolysis of the less-reactive fraction of organic matter that enables the biobattery to generate power for an extended period of time. Eventually, the less-reactive fraction will also be depleted, at which point the biobattery needs to be “recharged” by replenishing the solid anolyte.
Analytical determination of the reactive fractions of the solid anolyte in the first run showed that it contained 0.37 g g
−1 of high-reactive fraction and 0.25 g g
−1 of the less-reactive fraction. During the long-term operation of the biobattery shown in
Figure 6, power production exponentially decreases in the first 1–2 months of operation and it can be hypothesized that the high-reactive fraction of the solid anolyte was degraded in this period. The initial high power output (up to 1.6 mW) can thus be attributed to readily degradable dissolved organic matter, such as volatile fatty acids, from the high-reactive fraction. Thereafter, power production decreased at a slower rate, apparently corresponding to the much slower degradation of the less-reactive fraction. Once the biobattery was “recharged” and Run #2 started, the power output profile was reproduced (
Figure 6). Notably, the slightly higher power during Run #2 was due to the addition of 20 g of gellan gum (a polysaccharide) to the solid anolyte mix.