Electricity generation by Z. mobilis ZM4
During ethanol fermentation by Z. mobilis ZM4, the ORP value of broth kept decreasing in the initial 36 h, followed by a slightly recovery until 48 h (Figure 1A). It has been reported that the glucose consumption rate was the main cause for the altered ORP[20]. Before 36 h, rapid glucose consumption, attributed to active metabolism, tended to release and accumulate the reducing power from the substrate, which consequently pull down the ORP value. Afterward, cell lysis started the oxidized compounds release, which led to a little bit restoration of ORP. Therefore, Z. mobilis has potential to build up an reducing environment and form the ORP difference to produce electricity in MFC.
To further identify the potential capacity of electricity generation for Z. mobilis ZM4, OCV and WV were showed in the Figure 2A. Compared to medium-loaded MFC with stable voltage, ZM4-inoculated MFC exhibited a significant voltage output, which increased rapidly and peaked at 30 h, and then kept at high level. The maximal WV was about three-fold higher than that of the medium-loaded MFC, which meant that ZM4 was able to yield more electrons to enhance electricity generation. However, due to inefficient electron generation and transfer, the voltage output of ZM4 was relative lower than other well-known exoelectrogens[3, 4].
Cyclic voltammetry was conducted to reveal the redox reaction that occurred in the equipment. As shown in Figure 2B, a prominent redox peak appeared around -0.18 V in medium-loaded MFC while the redox peak moved to around -0.29 V in ZM4-inoculated MFC, which was caused by the direct involvement of conductive c-type cytochromes. These changed peaks indicated that the mechanism of extracellular electron transfer was different from medium mediated ones. To further investigate electrochemical performance, the polar curve was obtained (Figure 2C). Compared with the power generated in uninoculated MFC, ZM4-inoculated MFC exhibited a 2-fold higher maximum power density of approximately 2.0 mW/m2.
Ethanol production was monitored in an open circuit MFC, a closed circuit MFC, and flasks. Because of no significant difference among these conditions, electricity generation showed no competition with ethanol production for ZM4(Figure 1B). An energy balance was calculated to clarify the energy flow during power generation and ethanol production. Approximately 86% of the total energy was captured by ethanol (Figure 1C). By establishing the MFC equipment, approximately 3% of the energy was derived from the internal energy in the form of electricity. Compared with Shewanella oneidensis, ZM4’s recovered electricity energy was 80-fold lower[21]. But considering the energy captured by ethanol, the total generated energy of ZM4 was still significant.
The mechanism of electricity generation
According to Figure 3, The specific WV varied with the concentration of glucose. As an electron donor of ZM4-inoculated MFC, a higher glucose concentration would yield more electrons. Hence, MFC with 150 g/L glucose exhibited the highest voltage output 72.3 mV per OD600, while as glucose concentration decreased to 20 g/L, there was only 23.6 mV voltage output.
Additionally, the specific WV also varied over time. From 12 to 60 h, the specific WV increased and peaked at 30 h, followed by a slight regression. Based on the results in section 3.1, the ORP difference was formed by reducing compounds accumulation in the broth via consuming the substrate by Z. mobilis. Therefore, the glucose consumption rate can be considered as a main pointer for electricity generation. Interestingly, another voltage peak appeared between 0 and 12 h, when glucose was not used at all by microorganism, indicating that electricity generation was independent of glucose utilization before 12 h. Accordingly, the entire electricity generation could be divided into two types: a glucose-independent process and a glucose-dependent process.
For better understanding of two-type electricity generation, data scaling was adopted to normalize the variable range of WV at different glucose concentrations. In the left-bottom area of Figure 4A, the WV increased sharply when the glucose consumption rate remained unchanged, signifying that a glucose-independent electricity generation process occurred in the lag growth phase. In the right part of Figure 4A, representing the exponential growth phase, the WV obeyed positive linearity with the glucose consumption rate. In the left top area of Figure 4A, ZM4 entered a stationary growth phase, when two types of electricity generation simultaneously emerged. Furthermore, scatter diagrams of other initial glucose concentrations also showed the similar trends to 100 g/L, confirming a two-type process was a common phenomenon for ZM4-inoculated MFC, and that the glucose consumption rate was the direct driver of electricity generation progress.
Since the increased voltage during 0–12 h was independent of glucose uptake, the supernatant should contain some chemicals that contribute to the electricity generation. To verify this hypothesis, supernatants of broth at 6 h and 30 h were added in MFC (Figure 4B). The electricity performance of the supernatant at 6 h was restored immediately, but the supernatant at 30 h maintained basal level. Hence, the glucose consumption rate was the only contributor to electricity generation at 30 h. While the supernatant was the key driving force for MFC at 6 h, as the removal of cells did not strongly influence electricity performance. Although Z. mobilis cells cease growth and glucose usage during 0-12 h, they continue to release reducing substances, leading to the ORP difference and voltage output. Meanwhile, those reducing substances could be depleted because of low voltage output in the supernatant without cells at 30 h.
In addition, according to Figure 4C, once oxygen was sparged into the supernatant at 6 h, extra voltage generated by reducing substances were vanished, which confirmed the fact that reducing substances contributing to electricity generation could be eliminated by oxygen. In contrast, sparging oxygen at 30 h had no influence on the electrical properties, indicating the depletion of these reducing substances. Overall, the cooperation between the reducing substances in supernatant and glucose consumption by cells contribute to the total electricity generation.
Improvement of MFC by biofilm removal
Previous research showed that biofilm attached on the electrode has significant influence on electricity production[22, 23]. Concerning ability of ZM4 to form a biofilm on hydrophobic surfaces[24], biofilm formation on carbon cloth and evaluation of its role in electrogenesis were undertaken.
According to Figure 5A, the biofilm proliferated between 6-30 h and remained stable after it reached the maximum. Meanwhile, the charge transfer resistance (Rct) of the electrode calculated by EIS decreased from 0-30 h. As soon as the biofilm ceased growing, Rct was restored to its original level. To elucidate the exact relationship between Rct and the biofilm, the bacterial viability, defined as the ratio of live cells in biofilm, was quantified at different time points. Live cells dominated in the biofilm before 30 h, but as the biofilm continued growing, the number of dead cells began to increase and exceeded live cells at 60 h. Rct rose as the viability of cells dropped. The dead cells were not only capable of electricity generation, but also hampered the electron transfer process on the electrode surface[25, 26]. Therefore, the rise of charge transfer resistance at the later phase results from the continuously decreasing viability.
To prevent increased resistance, the biofilm attached to the electrode was cleared at 0, 6, 30, and 60 h (Figure 5A). There was no significant difference in Rct caused by the biofilm between the attached and detached electrode until 60 h. The Rct of the detached electrode was almost half of the attached electrode at 60 h, indicating that the removal of the biofilm dominated by dead cells helps enhance electron transfer to the electrode.
Due to positive effect on decreased resistance, biofilm removal was performed in three repeated batches to investigate long-term influence of biofilms on electrical performance. In each batch, the electrode was rinsed to detach biofilm when glucose was exhausted. According to Figure 5B, in the first two batches, the electricity generation of the detachment and attachment groups were maintained at the same level. But the WV of the detachment group was 10% higher than that in the attachment group in the third batch.
Given the biofilm and charge transfer resistance, although biofilm in the attachment group grew continuously due to the replenishment of fresh medium during the first two batches, the limited quantity of biofilm remained incapable of significantly influencing resistance and electricity performance. In the third batch, the entire surface of electrode had been covered by accumulated biofilm, which interrupted the electron transfer process resulting in a sharp rise in resistance. Hence, biofilm removal increased the WV in the third batch. However, this change was subtle due to the insufficient electricity generation ability of ZM4. Therefore, other strategies should be implemented to strengthen the EET.
Improvement of MFC by EET enhancement
Frequently, the EET pathway is a bottleneck for further improving the voltage output of exoelectrogens[27]. It consists of an indirect electron shuttle-meditated pathway[28] and the direct pathways, including appendages[29], nanowires[30] and c-type cytochrome[31]. Hence, adding electron shuttles and c-type cytochrome was implemented to enhance the EET pathway. To exclude the influence of the medium, voltage improvement ratios of ZM4 (Ros and Rws) and reagents (Ror and Rwr) were adopted to evaluate the contribution of the bacterial and the reagents to voltage respectively.
Figure 6A shows that methylene blue, neutral red, and TEMPOL significantly influenced the electricity performance: there was a noticeable increase on Rwr of methylene blue and Ror of neutral red, and a prominent decline on Ror of TEMPOL. However, without inoculation of ZM4, these changes could still be achieved by adding these electron shuttles only. In term of the voltage improvement ratio of ZM4 (Ros and Rws) (Figure 6B), the cells did not indicate an enhanced ability to generate electricity with the addition of electron shuttles that showed significant changes in Figure 6A. However, c-type cytochrome promoted ZM4 to output much more voltage (Ros and Rws were both improved). Meanwhile, c-type cytochrome has less influence on MFC (Ror and Rwr remained nearly unchanged). Therefore, instead of an indirect electron shuttle-mediated pathway, enhancing direct electron transfer could significantly benefit the electricity performance, which supported the CV measurement results in reported in section 3.1 that the main EET mechanism was direct involvement of conductive c-type cytochrome.
In addition, improving cell membrane permeability was also tried to enhance the transport of electron shuttles across cell membranes and achieve more efficient EET[32]. Tween 80 as a surfactant with concentration of 0, 5, 20, and 80 mg/L were chosen to increase permeability of ZM4[33]. Figure 7 shows that low addition (5 and 20 mg/L) of Tween promoted the OCV after 24 h, but the highest concentration (80 mg/L) inhibited OCV. The results reflected that appropriate Tween 80 enhanced the transport of electron shuttles. On the other hand, with the addition of Tween 80, WV reached peaks earlier, but the entire . Because the total electricity power is determined by the available electron pool, which is not closely related to the surfactant. Hence, improving the total quantity of electrons is the key factor for enhancing the whole electricity generation(E-supplementary data for this work can be found in e-version of this paper online)
Improvement of MFC by intracellular redox homegeneis
Commonly, electrons were stored in cells in form of NADH[34]. In theory, increasing ratio of NADH to NAD+ could help elevate electron pool to enhance voltage output[35]. Adding NAD+ precursor nicotinic acid (NA) and nicotinamide (NM), or overexpressing cofactor related genes (ZMO0899, ZMO1116, ZMO1885) were adopted. Figure 8 revealed that these two methods are successful to alter ratio of NADH/NAD+: precursors addition and overexpressing ZMO1116 (Glutamate synthase: reduce NAD+ to NADH and increase the ratio of NADH/NAD+) decreased the ratio while overexpressing ZMO0899 (NAD+ synthase: increase the pool of NAD(H/+) and ZMO1885 (NADH oxidase: oxidize NADH to NAD+ and decrease the ratio of NADH/NAD+) improved it. Because these genes functions were results of homologous blast, they may perform opposed functions in ZM4. Unfortunately, the maximum WV decreased with addition of nicotinic acid or nicotinamide and overexpressing redox-related genes. Owing to Z. mobilis sensitive to the intracellular redox change, altered ratio of NADH/NAD+ possibly inhibited electricity performance. Therefore, maintaining redox homogeneis is the precondition for improving voltage output of ZM4. Besides, this fact emphasized that if gene manipulation was adopted to enhance electricity performance, keeping redox balance should be focused, because the voltage improvement resulted from gene manipulation may be counteracted by the negative effect of redox imbalance during this modification process.
Improvement of MFC by the strategies combination
To further upgrade voltage output, cooperation of several effective methods (biofilm removal, addition of c-type cytochrome and Tween 80) were also tested. Figure 9 revealed that with addition of c-type cytochrome, electricity power increased by 30% (from 2.0 to 2.6 mW/m2) and Tween 80 helped reduce the time needed to reach the WV climax for MFC by about 6 h. Any treatment involving these two methods obtained the positive effect on electricity generation. However, biofilm removal did not affect MFC obviously. As previous discussion in section 3.3, only repeated batch fermentation was able to influence WV in Z. mobilis-inoculated MFC. Enough accumulation and long-term growth leads to more dead cells trapped in the biofilm, thus refreshing the biofilm with short fermentation time can’t exhibit significant improvement of electricity production. Therefore, combination of c-type cytochrome and Tween 80 improved power density and accelerated the entire electricity generation simultaneously showing that advantages of these two methods could be superposed to benefit electricity performance more.