Effect of temperature on a miniaturized microbial fuel cell (MFC)

Abstract

A microbial fuel cell (MFC) is a bioinspired energy converter which directly converts biomass into electricity through the catalytic activity of a specific species of bacteria. The effect of temperature on a miniaturized microbial fuel cell with Geobacter sulfurreducens dominated mixed inoculum is investigated in this paper for the first time. The miniaturized MFC warrants investigation due to its small thermal mass, and a customized setup is built for the temperature effect characterization. The experiment demonstrates that the optimal temperature for the miniaturized MFC is 322–326 K (49–53 °C). When the temperature is increased from 294 to 322 K, a remarkable current density improvement of 282% is observed, from 2.2 to 6.2 Am⁻². Furthermore, we perform in depth analysis on the effect of temperature on the miniaturized MFC, and found that the activation energy for the current limiting mechanism of the MFC is approximately between 0.132 and 0.146 eV, and the result suggest that the electron transfer between cytochrome c is the limiting process for the miniaturized MFC.

Background

A microbial fuel cell electrochemically converts biomass into electricity through the catalytic activity of specific species of bacteria, named exoelectrogen or anode-respiring bacteria, which are capable of transfer electrons outside their outer membrane [1,2,3]. During the past two decades, significant improvement on MFC has been demonstrated; for instance, the power density of microbial fuel cells have improved by more than four orders of magnitude, from 0.1 mWm⁻² to 7.72 Wm⁻² [4,5,6]. In order to improve the power density, a number of research have been performed, such as implementing the anode with high surface area to volume ratio [5, 7,8,9], investigating the performance of different species of exoelectrogen [10, 11], investigating different MFC configurations, and investigating different operation conditions, including temperature, pH, flow rate, etc. [12,13,14,15]. The two most investigated exoelectrogen are Shewanella and Geobacter, which generally generate higher current and power density than other types of exoelectrogen. The effect of temperature on microbial fuel cells with Geobacter as exoelectrogen has been investigated by many researchers, and it is reported that the optimal temperature to be 298–303 K (25–30 °C) for macro/mesoscale MFCs [13, 15].

Miniaturized MFCs are emerging technologies that utilize microfabrication for MFCs, which results in the characteristics of small footprint, low expense, batch fabrication, improved diffusion, high surface area to volume to ratio, etc. [1, 16–18]. Particularly, the small footprint makes miniaturized MFCs with a very small thermal mass, which is suitable for the temperature to be controlled within a short period of time [19]. We perform first-order calculation and found that the miniaturized MFC can be heated to the target temperature within several minutes due to its low heat capacity, which is suitable for characterizing the effect of temperature on MFCs, while macroscale MFCs takes much longer time (supporting materials). We have demonstrated miniaturized MFCs and microbial supercapacitors with record power densities of 1.12 × 10⁴ and 3.95 × 10⁵ Wm⁻³; yet the effect of temperature on miniaturized MFCs haven’t been studied to date [5, 7, 20]. In this paper, we report the first study of temperature effect on a miniaturized MFC with Geobacter dominated mixed species. Our results show that the miniaturized MFC has an optimal temperature at 322–326 K.

Methods

A customized setup including a miniaturized microbial fuel cell (MFC) with an anode chamber of 50 µL and a close-loop temperature controlled oven was built, as illustrated in Fig. 1. The process flow for the miniaturized MFC fabrication is illustrated in Fig. 1a. The miniaturized MFC had a proton exchange membrane (PEM) (Nafion 117, Sigma Aldrich), to permit cation transport and to avoid electrical short-circuiting and electrolyte cross mixing. Two silicone gaskets with thicknesses of 500 µm were sandwiched between two glass slides. Ti/Pt (20/100 nm) was deposited by electron beam evaporation for both glass slides in order to form anode and cathode. The volume of each chamber was 50 μL, and the size of electrode was 1 cm². Two nanoports (10–32 Coned assembly, IDEX Health and Science) were used to provide microfluidic pathways into and out of each chamber. The MFC was assembled with four screw bolts and nuts to minimize oxygen/electrolyte leakage. An optical image of the miniaturized MFC after assembly is depicted in Fig. 1b.

Fig. 1

A miniaturized microbial fuel cell (MFC) for the characterization of temperature effect: a fabrication process flow for the miniaturized MFC; b Optical image of the miniaturized MFC after assembly; c illustration of the miniaturized MFC mounted inside a gravity convection oven with close-loop temperature control. The miniaturized MFC has a chamber volume of merely 50 µL, allowing the temperature inside the MFC to be controlled by the oven due to its small thermal mass

The inoculum for the MFC reactor was obtained from an acetate-fed microbial electrolysis cell (MEC) that had Geobacter-enriched bacterial community originally from anaerobic-digestion sludge. Clone libraries of the 16S-rRNA gene showed that the inoculum was a mixed bacterial culture dominated by Geobacter sulfurreducens. The anolyte was 25-mM sodium acetate medium with 1680 mg KH₂PO₄, 12,400 mg Na₂HPO₄, 1600 mg NaCl, 380 mg NH₄Cl, 5 mg EDTA, 30 mg MgSO₄·7H₂O, 5 mg MnSO₄·H₂O, 10 mg NaCl, 1 mg Co(NO₃)₂, 1 mg CaCl₂, 0.001 mg ZnSO₄·7H₂O, 0.001 mg ZnSO₄·7H₂O, 0.1 mg CuSO₄·5H₂O, 0.1 mg AlK(SO₄)₂, 0.1 mg H₃BO₃, 0.1 mg Na₂MoO₄·2H₂O, 0.1 mg Na₂SeO₃, 0.1 mg Na₂WO₄·2H₂O, 0.2 mg NiCl₂·6H₂O, and 1 mg FeSO₄·7H₂O (per liter of deionized water) (pH 7.8 ± 0.2). For the start-up process, inoculum and anolyte were mixed with a volume ratio of 1:1. The catholyte was 100-mM potassium ferricyanide in 100-mM phosphate buffer solution (pH 7.4). A gravity convection oven was utilized to control the temperature of the MFC (Lindberg Blue M), as illustrated in Fig. 1c. The temperature of the oven was then controlled digitally with a resolution of 0.1 °C and was calibrated with a lab red spirit-filled precision thermometer. The current was monitored by recording the voltage drop across an external resistor connected between the anode and the cathode using a data acquisition system (DAQ/68, National Instrument). At least three replicates were conducted for all experiments.

Due to its small thermal mass, the miniaturized MFC allowed the temperature inside anode chamber of the MFC to match well with the ambient temperature in the oven. The start-up processes of the miniaturized MFCs in elevated temperature of 314 K and at a room temperature of 294 K and elevated temperature of 314 K are shown in Additional file 1: Figures S1 and S2, respectively.

Results

First, we characterized the impact of the temperature on the performance of the miniaturized MFC when a thick biofilm is formed on top, and the biofilm is grown at a temperature of 314 K. The temperature was changed periodically from 294 to 332 K with six intervals. For each temperature, we waited three to 6 h for the current became stable. As shown in Fig. 2, the current density first increased with the increase of temperature, and then decreased. At a temperature of 294 K, a power density of mere 2.2 Am⁻² was obtained while at a temperature of 322 K, the highest current density of 6.2 Am⁻² was reached. The reason for the drop of current density was believed to be that the high temperature made the Geobacter less active.

Fig. 2

Current density versus operation temperature of a miniaturized MFC with a fully grown biofilm grown at 314 K. As the operation temperature increases, the current density first increase and then decrease

Analyzing Fig. 2, the increase of the current comes from two parts: (1) at high temperature, the output current for exoelectrogen increases; (2) at high temperature, the number of exoelectrogen which become active also increases. A second experiment was performed on the miniaturized MFC during its start-up process under a temperature of 294 K, and the temperature was controlled by the oven ranging from 294 to 325 K (the maximum temperature is kept at 325 K so that the device will not fail due to high temperature). For each temperature, a period of 5–7 min was kept until increasing the temperature. The short time prevent the exoelectrogen from proliferating since there is not enough time. Therefore, we can assume that the amount of exoelectrogen is kept constant during the temperature adjustment phase. The result of the current density versus temperature is shown in Fig. 3. The current increases with the temperature, which is in similar trend with Fig. 2. As a result, it also shows that the temperature has a direct impact on the current output of the MFC.

Fig. 3

Output current versus temperature for a miniaturized MFC during biofilm growth (grown at 294 K). For a single temperature increase adjustment, only 5–7 min are kept, so the biofilm does not have enough time to grow and the result represent the effect of temperature on the same biofilm

Discussion

We try to perform more in depth analysis on the temperature effect on the miniaturized MFC result. Firstly, we try to derive equation about the relationship between the output current versus the temperature.

According to the Marcus theory, the rate constant for electron transfer is expressed as [21]:

$$ \begin{aligned} k_{{et}} & = \frac{{2\pi }}{\hbar }H_{{AB}}^{2} \frac{1}{{\sqrt {4\pi k_{b} T} }}\exp \left( { - \frac{{(\lambda + \Delta G^{ \circ } )^{2} }}{{4\lambda k_{b} T}}} \right) \\ & = \frac{{2\pi }}{\hbar }H_{{AB}}^{2} \frac{1}{{\sqrt {4\pi k_{b} T} }}\exp \left( { - \frac{{\Delta G^{\dag } }}{{k_{b} T}}} \right) \\ \end{aligned} $$

(1)

where k _et is the rate constant of the rate-limiting step in catalytic activity of G. sulfurreducens, H _AB is the electron coupling between initial and final state, λ is the reorganization energy, ħ is the reduced Planck constant, ∆G° is the total Gibbs energy change during the electron transfer, k _b is the Boltzmann constant, T is the absolute temperature, and ∆G ^† is the activation energy associated with a rate-limiting step in catalytic activity of G. sulfurreducens, where

$$ \Delta G^{\dag } = \frac{{(\lambda + \Delta G^{ \circ } )^{2} }}{4\lambda } $$

As a result, assuming the rate-limiting step is a first order reaction, the current in the catalytic activity of G. sulfurreducens can be written as

$$ I = C \cdot k_{et} \cdot M = \frac{2\pi }{\hbar }C \cdot H_{AB}^{2} \frac{1}{{\sqrt {4\pi k_{b} T} }}\exp \left( { - \frac{{\Delta G^{\dag } }}{{k_{b} T}}} \right) \cdot M $$

(2)

Here C is a constant for the reaction. Rearranging Eq. 2 results in

$$ \ln \left( {I \cdot \sqrt T } \right) = \ln \left( {\frac{2\pi }{\hbar }C \cdot H_{AB}^{2} \frac{1}{{\sqrt {4\pi k_{b} } }}M} \right) - \frac{{\Delta G^{\dag } }}{{k_{b} T}} $$

(3)

Equation 3 shows a linear relationship between the ln(I∙T ^0.5) and T ⁻¹ and the slope is −∆G ^† /k _b . As a result, by linearly fitting the curve of ln(I∙T ^0.5) versus T ⁻¹, the activation energy of the rate-limiting step in EET is obtained.

Ln(I∙T ^0.5) and T ⁻¹ curve for biofilm during the start-up process with different steady-state current at 294 K, ranging from 0.04 to 0.84 Am⁻² is plotted based on Fig. 3, as shown in Fig. 4, where the 0.04 Am⁻² represent very initial biofilm, while the 0.84 Am⁻² represent fully grown biofilm. According to Fig. 4a, a linear relationship between ln(I∙T ^0.5) and T ⁻¹ is found for all four curves, and linearly fitting the curves in Fig. 4a results in the corresponding activation energies. As the activation energy shows, in the very initial biofilm and mid-stage (beginning) biofilm, the activation energy ranges from 0.189 to 0.202 eV. For mid-stage (end) biofilm to fully grown biofilm, the activation energy drops to 0.132–0.146 eV.

Fig. 4

a ln(I∙T^0.5) and T⁻¹ curve for biofilm at biofilm formation stages with different maximum steady state current densities during the start-up process, linear relationships are observed for all curves, and activation energy are obtained through Eq. 3, as shown in b

We also plot the ln(I∙T ^0.5) versus T ⁻¹ curve based on the data for Fig. 2, as shown in Fig. 5.

Fig. 5

a ln(I∙T^0.5) and T⁻¹ curve for thick fully-grown biofilm grown at 314 K; b logarithmic amount of redox cofactors contributing to the steady-state current generation as temperature changes during fully grown biofilm by assuming the activation energy is 0.132 eV based on Fig. 4b; as temperature increases, a larger amount of redox cofactors are responsible for the catalytic activity of Geobacter sulfurreducens, suggesting that the catalytic activity of Geobacter sulfurreducens is temperature dependent and is favored by higher temperature

A linear relationship between ln(I∙T ^0.5) and T ⁻¹ is also found for the fully grown biofilm; however, the slope in Fig. 5a is much larger than Fig. 4a. The reason for this phenomenon is that as temperature increases, not only does the rate constant, k _et increase, but also the amount of redox cofactors, M _red increases exponentially based on in Eq. 3. Thus because we use the same miniaturized MFC setup and the same inoculum to obtain Figs. 4 and 5, we can assume the k _et for fully grown biofilm is the same as Fig. 4 for fully grown biofilm, 0.132 eV. Therefore, we can calculate the change in amount of redox cofactor contributing to the EET as temperature changes, as shown in Fig. 5b, which shows a linear relationship between ln(M _red ) and T ⁻¹. Linearly fitting the curve also results in an activation energy of 0.155 eV.

The activation energy in the rate-limiting step of catalytic activity of G. sulfurreducens is compared with activation energies of individual catalytic activity of G. sulfurreducens steps, as shown in Table 1. It is found that the activation energy of this work fits very well to the electron transfer between cytochrome c, while being very far from activation energies of other steps associated with EET. This suggests that the rate-limiting step is the cytochrome c.

Table 1 the activation energy of the rate-limiting step of this work is compared with the activation energy reported by prior studies of the reactions related to EET

Moreover, the activation energy for the amount of M _red versus 1/T, 0.155 eV, falls in the activation energy range of cytochrome c, which also supports that for thick biofilm, the electron transfer between cytochrome c is critical for electron transfer from G. sulfurreducens to the anode, i.e. EET [29].

These findings suggest that the rate-limiting step for catalytic activity of G. sulfurreducens is the EET, which has long been a focus for studies in microbial electrochemical technology (MET). The result is consistent with prior evidences. For instance, Liu and Bond [29] found that all cytochrome c proteins were oxidized at the initial biofilm formation stage, whereas reduced cytochrome c accumulated in thick biofilms, suggesting a part of the cytochrome c located far from the anode cannot transport their electrons to the anode. It suggests that the long range electron transfer is the limiting mechanism for EET. Strycharz-Graven et al. [30] observed cyclic voltammetry peaks at a slow scan rate and reported that EET driven by redox reaction is limited either by electron transfer from inside the exoelectrogen to the outside of the exoelectrogen or from outside the exoelectrogen to the anode. Bonanni et al. [31] performed numerical modeling on the EET found that electron transfer limited the respiration rate of the cells for biofilm far from the anode. Our result not only concurs existing findings, but also is the first findings to experimentally pinpoint that electron transport among cytochrome c is the rate-limiting mechanism for EET.

Furthermore, the activation energy for the amount of redox cofactors versus 1/T, 0.155 eV, supports that electron transfer between cytochrome c is a critical step for EET. For thick biofilm, as temperature increases, the amount of cytochrome c contributing to EET also increases, suggesting that electron transfer between cytochrome c is the main long-range EET mechanism. This finding is in agreement with prior studies by Snider et al. [32] that long-range electron transport in G. sulfurreducens biofilms is driven by cytochrome c redox gradient. In the meantime, it is not contradictory with the findings that nano-pili secreted by G. sulfurreducens are conductive [31,32,35], as the nano-pili may not be the major electron transfer mechanism.

In addition, from our finding, miniaturized MFC performs better at an elevated temperature, which concurs with the trend of macro-sized MFCs [36]. This finding may help our future MFC design to seek for optimal temperature to be established.

Conclusion

A miniaturized MFC is built to characterize its current output capability as a function of the temperature. The result shows that the output current increases as temperature increases from 294 to 322 K, and an optimal temperature for the MFC is 322–326 K (49–53 °C), which indicates the thermal activation characteristic of the miniaturized MFC. Further in depth analysis reveals that the activation energy for the current limiting mechanism of the MFC is 0.132–0.146 eV, suggesting that the electron transfer between cytochrome c is the limiting process for the miniaturized MFC.