Natural eggshell membrane as separator for improved coulombic efficiency in air-cathode microbial fuel cells

Abstract

Natural eggshell membrane (ESM) is proposed as a separator to enhance electricity recovery in air-cathode microbial fuel cells (MFCs). The unique fibrous structure and surface chemistry of ESM can mitigate substrate and oxygen crossover, and biofouling of the cathode, achieving a remarkable coulombic efficiency of 67.14–95.03%, depending on the current density. This study demonstrates an effective, simple and green route for recovering electricity from organic matter in MFCs.

---

Microbial fuel cells (MFCs) are a new, promising technology for sustainable recovery of electricity through microbial oxidation of organic substances based on the metabolism of electrochemically active microorganisms.¹ MFCs are capable of producing electricity using a broad range of organic materials including carbohydrates, volatile fatty acids, alcohols, and proteins.² Being similar to chemical fuel cells, the microbially produced electrons are sinked at the cathode where air (21% oxygen) is reduced to form water.³ It was, and is common practice to apply proton exchange membranes (PEMs) for separating the anode and cathode, with the target of conducting protons, avoiding substrate crossover, and preventing oxygen diffusion to the anode.⁴ In most cases, a PEM is virtually unnecessary in air-cathode MFCs because the aqueous solution is conductive to cationic ions for charge neutralization. Many studies addressed that the power density could be increased greatly when the PEM is removed because of the decease in ohmic resistance of the MFC.⁵ On the other hand, in the absence of PEM, the oxygen can be passively transferred to the anode to react directly with the substrate (electron loss).⁶ For example, a number of studies reported a quite low coulombic efficiency (CE), i.e. lower than 20% at external resistance of 1000 Ω in membrane-free MFC.⁷ Obviously, this constitutes an unwanted side reaction that causes substantial decline in coulombic efficiency (CE) and oxygen concentration available for electrochemical reduction, as well as duration problems associated with biofouling and catalyst poisoning.⁸

To mitigate above problems, a variety of separator materials are placed over the cathode to inhibit substrate crossover and oxygen reverse diffusion.⁹ However, the extent to which CE can be increased appears to be very limited, most often, accounting for the magnitude of increase not exceeding 20–40% (e.g. external resistance of 1000 Ω).¹⁰ This should be the consequence of several reasons, such as pH changes resulting from selective ion transport,¹¹ increased internal resistance caused by ionic impedance of separators,¹² and non-selective permeability toward proton, oxygen and organic substrates.^13,14 Additionally, the high cost of ion exchange membrane also impedes the larger-scale application of MFC. Within this context, it will be highly desirable to develop efficient, cost-effective and sustainable separator materials for improved MFC performances.

In this communication, eggshell membrane (ESM), a natural permselective semipermeable membrane with ion transportation, hydrophilicity, mechanical-resistant, low cost and sustainability is first used as separator to improve power generation in air-cathode MFC.¹⁵ To further identify the role of ESM, the structure and constitutes of ESM are characterized. ESM has a biopolymeric coalescing fibrous network situated between eggshell (ES) and eggwhite (EW) (Fig. 1A), which is crucial for protecting EW from infecting bacteria through the aperture of ES, and prevention of water evaporation from inside EW and yolk.¹⁶ Thereinto, the outermost part of ESM is outer membrane (OM) located under ES with the fibre size of 0.5–3 μm and mesh size of 0.7–8 μm (Fig. 1B). Some parts extend into the mammillary knobs of ES, making it difficult be separated for further utilization.¹⁷ The inner membrane (IM) is interlaced with OM while is separated from it in the air-filled cavity (Fig. 1A).¹⁸ Besides, the thickness of OM (50–70 μm) is higher than that of IM (15–26 μm) with a smaller fibrous size and more compact structure from the outer side to the inner side.¹⁹ Depending on the degree of humidity, both two layers exhibit much lower oxygen transfer coefficient (k₀ = 1.78 × 10⁻⁶ cm s⁻¹ for OM and k₀ = 0.11 × 10⁻⁶ to 1.56 × 10⁻⁶ cm s⁻¹ for IM) than the separators (k₀ = 0.42 × 10⁻⁴ to 1.3 × 10⁻⁴ cm s⁻¹) currently used in MFC.²⁰ Moreover, the innermost part of ESM is limiting membrane (LM), an extremely thin nodular structure surrounding the EW without apparent macropores (Fig. 1C), presenting a further handicap for permeability of O₂. Thus, the unique structure and properties of ESM makes it possible to serve as separator for enhanced electricity recovery and power output due to the limited oxygen permeability, excellent mass transfer (H⁺, OH⁻) capability, and impermeability of macromolecule nutrients (glucose). In a word, the electron–oxygen–hydrion interface reactions are prospected to be greatly enhanced by preventing the oxygen and nutrients crossover the natural ESM to achieve four-electron high-efficiency long-time power output in MFC (Fig. 1D).

Fig. 1 (A) Artistic rendition of the structure of an egg, and SEM images of (B) the outermost structure and (C) the innermost part of the eggshell membrane. (D) Schematic illustration of the structures of air-driven cathode equipped with ESM separator and the cathode reactions.

The performance of ESM serving as separator was evaluated in single-chamber glucose-feeding (1 g L⁻¹) MFC. As is revealed in Fig. 2A, the cell voltage of approximately 0.572 ± 0.03 V was obtained at external resistance of 1000 Ω in the presence of ESM separator over Pt/C cathode (Pt/C-ESM), and the voltage was increased to 0.598 ± 0.05 V when the EMS is removed from the cathode. The observation of higher voltage lacking ESM (0.026 V) originated from decrease in ohmic resistance, which was well consistent with previous studies.²¹ Notably, such minor difference did not dismiss superior coulombic recovery for the Pt/C-ESM MFC, indicated by extraordinary long-time voltage duration of about 70 h (Fig. 2A). This value is almost as five times as membrane-free Pt/C cathode (14 h), corresponding to a much longer single-cycle time (120 h) for Pt/C-ESM cathode than that of 50 h for Pt/C one, and many other results reported previously.²²

Fig. 2 (A) Time course of cell voltage produced in MFC, (B) power density, (C) electrode potential and (D) CE as a function of current density. Red circles: with ESM; grey squares: without ESM. The data ± S.D. bars indicate the experiments in triplicate.

To further investigate the role of ESM in MFC, the polarization tests were conducted and the resultant CE values were also calculated. Fig. 2B shows that the MFC equipped with ESM separator produced a maximum power density of 1441 ± 15 mW m⁻² at 3.47 ± 0.05 A m⁻², a value just 1.84% higher than that without (1415 ± 10 mW m⁻², 3.46 ± 0.08 A m⁻²). Such slight difference should be due to the enhanced anode potentials and unaffected cathode potentials. As shown in Fig. 2C, when reaching the maximum power density, the use of ESM gave much lower anode potential of −0.509 ± 0.01 V vs. SCE, which was 13.62% lower than ESM-free one (−0.448 ± 0.02 V), while the cathode potential remained almost the same. These results further confirmed the superiority of ESM as separator for enhancing persistent power production without the loss of power density, by contrast to prior separators that presented considerable decrease in power density in spite of increasing CE.⁹ Moreover, the CE of the two cathodes were further calculated and compared. With increased current density by stepwise change of external resistance, much higher CE values of 67.14–95.03% were obtained in the presence of ESM than that of 22.11–38.21% in the absence of ESM (Fig. 2D). It was worth emphasizing that the maximum CE reached a unprecedentedly high level as great as 95.03% in air-cathode MFC, and this was in line with the extremely low oxygen mass transfer coefficient (2.3 × 10⁻⁷ cm s⁻¹, maximal humidity, 71–106 μm).²³ The CE values were also quite competitive with other reported separators (Fig. 3) such as glass and nylon fibres having different pore sizes (37–76%, 0.7–4.8 A m⁻², k₀ = 0.19 × 10⁻⁴ to 0.5 × 10⁻⁴ cm s⁻¹ with thickness of 45–700 μm),^7,9,22 ultrafiltration membrane (38–49%, 0.7–3.1 A m⁻², k₀ = 0.19 × 10⁻⁴ to 0.42 × 10⁻⁴ cm s⁻¹, 265 μm),⁴ J-cloth with different layers (20–38%, 0.7–3.6 A m⁻², 2.9 × 10⁻³ cm s⁻¹, 300 μm),^9,24 ion exchange membranes (35–72%, 0.7–2.8 A m⁻², 0.94 × 10⁻⁴ to 1.3 × 10⁻⁴ cm s⁻¹, 190–460 μm),^4,9 as well as polyvinyl alcohol (PVA) polymeric membranes (42–80%, 0.8–4.9 A m⁻², 2.26 × 10⁻⁴ cm s⁻¹, 120 μm).²¹

Fig. 3 CE at the current density of maximum power as a function of maximum power density for different separators. Data for ESM and ESM-free MFCs are provided in this study. Data for nylon filters, glass fibres (GFF), and in absence of a separator (NS) are from Zhang et al.⁷ Data for glass fibre DC1.0 (GF) and J-cloth (JC) are from Zhang et al.²² Data for ultrafiltration membranes with molecular weight cut offs of 1k daltons (UF), PEM (Nafion), AEM, and CEM are from Kim et al.⁴ Data for poly (vinyl alcohol) separators without porogen (PVA) is from Chen et al.²¹

In previous works, there was a trade-off existing between power density and CE. Despite higher CE by blocking oxygen diffusion, the mass transfer process is also hindered and vice versa.¹³ Herein, the MFC equipped with ESM separator exhibit both high power density and CE compared to the data from other studies (Fig. 3).^4,9,21 Besides, the stability-based biofouling and biodegradation issues have been explored for ESM long-term operation. As shown in Fig. S1,† the ESM cathode exhibits a relative stable voltage output of about 800 h (6 cycles, ∼135 h per cycle, 28 mL reactor). After that, the voltage decreases from 0.6 to 0.5 V but can still maintain 400 h, then the voltage shows a continuous declination with no obvious degradation after 11 cycles' (1500 h) running (Fig. S1†). In order to further study the change of ESM after operation, the FT-IR and SEM images have been listed in Fig. S2 and S3.† As shown, there is no obvious change of the FT-IR characteristic peak of ESM before and after operation, demonstrating the bacteria attached on ESM haven't biodegraded or changed the nature of these proteins. This result is in accordance with voltage output (Fig. S1†), optical and SEM images (Fig. S3A–D†) of the ESM. Besides, more scattered rod-shaped bacteria are formed on the side facing the nutrients of ESM as operation continues (Fig. S3B–D†). And no obvious biodegradation shows up after 1500 hours' operation, only gradually serious biofouling. This will significantly block the mass transfer (H⁺, OH⁻) channels then decrease the performance of MFC although they are not enriched to the naked eye compared with pure Pt/C (Fig. S3A and S4†). Even so, the ESM cathode can still obtain excellent CE, power output, and operation time than the pure Pt/C cathode. So considering its cost-effective, high-efficiency, and sustainability compared with other separators, the natural wasted ESM is proposed to be used as separator in MFC for its further development.

To examine whether the ESM has a negative impact to the performance of cathodic oxygen reduction, the linear sweep voltammetry (LSV) and electrochemical impedance spectroscopy (EIS) were performed. The results show a slightly lower current density for ESM-cathode (34.91 ± 0.25 A m⁻²) than that for bare cathode (38.14 ± 0.32 A m⁻²) due to the increased transfer resistance (Fig. 4A).²¹ Fig. 4B illustrates similar ohmic resistance and charge-transfer resistance of MFC, indicating the negligible impact of ESM to electrochemically active microorganisms and electrolyte properties. Nevertheless, the diffusion resistance of ESM was slightly higher than that of without, due to the resistance of ESM to mass transfer, but this did not necessarily affect power production. However, a thick macroscopic biofilm (∼2.3 mm) inevitably formed on the surface of bare cathode during long-term operation owing to direct contact with microorganisms and nutrients in anolyte (Fig. S4†). The growth of dense biofilm can significantly increase both charge transfer and diffusion resistance by consuming the oxygen and hydrions involved in ORR procedure and inhibiting proton transfer (Fig. 4B).¹⁴ On the contrary, the MFC with ESM displayed a relatively lower resistance due to the superior ion transmission capacity toward H⁺ and OH⁻. Besides, the ESM can also mitigate the substrate transfer effectively to prevent the biofilm formation on the cathodic active interface.¹³ For this reason, there was no visible biofilm on the cathode surface in the presence of ESM following long-term operation (Fig. S4†). That is to say, the unique properties of ESM make it capable of improving the reactions at electron–oxygen–hydrion interface by inhibiting reverse migration of oxygen and substrate. The two processes have been known to be responsible for coulombic loss and ion transfer resistance. On the other hand, the excellent permeability of H⁺ and OH⁻ can enable four-electron oxygen reduction at the cathodic interface (Fig. 1D).

Fig. 4 (A) LSV curves and (B) EIS plots for Pt/C and Pt/C-ESM. (C) Schematic illustration of ESM properties and its application in MFC.

On the basis of the chemical composition and microstructure, the mechanisms of ESM for improved power production in air-cathode MFC were analysed and discussed. The fibres of ESM are mainly composed of various types of proteins (80–85%), of which 10% are collagens (type I, V, and X), and 70–75% are other proteins and glycoproteins.²⁵ The number of protein types in ES matrix (>500) is much larger than that in EW (148),²⁶ vitelline membrane (137),²⁷ and egg yolk (316).²⁸ The chemical compositions and morphologies mean a complicated selective permeability, especially an excellent permeability toward some small-size ions (e.g. H⁺, OH⁻),²⁹ while impermeability toward some macromolecules (glucose, lactose) and bacteria,^30,31 and limited permeability toward oxygen molecules (O₂) (Fig. 4C) under humidified condition.³² Besides, the ESM contains many C, O, N-based groups, which is beneficial for hydrophilicity and ion-binding capacity.³³ Considering all these unique properties of ESM, it is prospected to be a high-efficiency separator material applied in MFCs for several reasons below.

First, the charge-neutral structure and hydrophilicity can mitigate OH⁻ accumulation by avoiding selective ion transport over the cathode interface. The excellent permeability of H⁺ and OH⁻ via mesh pore channels, protein-driven diffusion, and mediator transmission can provide sufficient H⁺ for oxygen reduction and alleviate OH⁻ accumulation effect.²⁹ Second, the oxygen mass transfer coefficient of ESM is two orders of magnitude lower than that of other separators used in MFC,²⁰ and thus resulting in adequate electron recovery (CE of 95%) because O₂ reverse diffusion is impaired significantly. Third, the ESM exhibits impermeability toward biologically metabolic substrate like glucose,³⁴ such that the crossover phenomena and biofilm formation are inhibited due to the lack of carbon source. Last, the ESM also prevent the volatilization losses of substrates effectively, maximizing the electron recovery via electrochemical route.³⁵

In conclusion, the eggshell is a common bio-waste extensively available from daily foods. Herein, the ESM is proposed to be in situ used as separator for air-cathode MFC, achieving a remarkable enhancement of power density and CE. The unique chemical composition and micro structure allow for selective permeability toward different species, making the ESM more advantageous than most of commercially available separator materials.

Acknowledgements

This research was financially supported by the National Natural Science Foundation of China (No. 51378143), State Key Laboratory of Urban Water Resource and Environment (Grant No. 2015TS01), the Fundamental Research Funds for the Central Universities (Grant No. HIT.BRETIII.201419).

Notes and references

1. B. E. Logan, B. Hamelers, R. Rozendal, U. Schröder, J. Keller, S. Freguia, P. Aelterman, W. Verstraete and K. Rabaey, Environ. Sci. Technol., 2006, 40, 5181–5192.
2. D. Pant, G. V. Bogaert, L. Diels and K. Vanbroekhoven, Bioresour. Technol., 2010, 101, 1533–1543.
3. F. Harnisch and U. Schröder, Chem. Soc. Rev., 2010, 39, 4433–4448.
4. J. R. Kim, S. Cheng, S.-E. Oh and B. E. Logan, Environ. Sci. Technol., 2007, 41, 1004–1009.
5. H. Liu and B. E. Logan, Environ. Sci. Technol., 2004, 38, 4040–4046.
6. S. Cheng, H. Liu and B. E. Logan, Environ. Sci. Technol., 2006, 40, 364–369.
7. X. Zhang, S. Cheng, X. Wang, X. Huang and B. E. Logan, Environ. Sci. Technol., 2009, 43, 8456–8461.
8. S. Hays, F. Zhang and B. E. Logan, J. Power Sources, 2011, 196, 8293–8300.
9. W.-W. Li, G.-P. Sheng, X.-W. Liu and H.-Q. Yu, Bioresour. Technol., 2011, 102, 244–252.
10. G. Chen, F. Zhang, B. E. Logan and M. A. Hickner, Electrochem. Commun., 2013, 34, 150–152.
11. E. A. Hernándes-Pagán, N. M. Vargas-Barbosa, T. Wang, Y. Zhao, E. S. Smotkin and T. E. Mallouk, Energy Environ. Sci., 2012, 5, 7582.
12. F. Harnish and U. Schröder, ChemSusChem, 2009, 2, 921–926.
13. Y. Yuan, S. Zhou and J. Tang, Environ. Sci. Technol., 2013, 47, 4911–4917.
14. J. An, H. Jeon, J. Lee and I. S. Chang, Environ. Sci. Technol., 2011, 45, 5441–5446.
15. M. Baláz, Acta Biomater., 2014, 10, 3827–3843.
16. I. Gantois, R. Ducatelle, F. Pasman, F. Haesebrouck, R. Gast, T. J. Humphrey and F. V. Immerseel, FEMS Microbiol. Rev., 2009, 33, 718–738.
17. J. W. W. Liong, J. F. Frank and S. Bailey, J. Food Prot., 1997, 60, 1022–1028.
18. R. Bellairs and A. Boyde, Z. Zellforsch. Mikrosk. Anat., 1969, 96, 234–249.
19. J. Zhou, S. Wang, F. Nie, L. Feng, G. Zhu and L. Jiang, Nano Res., 2010, 4, 171–179.
20. S. R. Kayar, G. K. Snyder, G. F. Birchard and C. P. Black, Respir. Physiol., 1981, 46, 209–221.
21. G. Chen, B. Wei, Y. Luo, B. E. Logan and M. A. Hickner, ACS Appl. Mater. Interfaces, 2012, 4, 6454–6457.
22. X. Zhang, P. Liang, J. Shi, J. Wei and X. Huang, Bioresour. Technol., 2013, 130, 529–535.
23. R. A. Ackerman and H. Rahn, Respir. Physiol., 1980, 45, 1–8.
24. Y. Fan, H. Hu and H. Liu, J. Power Sources, 2007, 171, 348–354.
25. R. M. Leach, Poult. Sci., 1982, 61, 2040–2047.
26. K. Mann, J. V. Olsen, B. Macek, F. Gnad and M. Mann, Proteomics, 2007, 7, 106–115.
27. K. Mann, Proteomics, 2008, 8, 2322–2332.
28. K. Mann and M. Mann, Proteomics, 2008, 8, 178–191.
29. Y. W. L. Beanent, Bull. Entomol. Res., 1949, 39, 467–488.
30. M. Russel, Y. Qian, J. Yao and M. M. F. Choi, Int. J. Environ. Anal. Chem., 2014, 94, 1317–1329.
31. K. D. Reu, K. Grijspeerdt, W. Messens, M. Heyndrickx, M. Uyttendaele, J. Debevere and L. Herman, Int. J. Food Microbiol., 2006, 112, 253–260.
32. O. D. Wangensteen, D. Wilson and H. Rahn, Phys. Rev. Lett., 1971, 11, 16–30.
33. Y. Shen, P. O. Saboe, I. T. Sines, M. Erbakan and M. Kumar, J. Membr. Sci., 2014, 454, 359–381.
34. R. E. Wood, F. P. Wirth and H. E. Morgan, Biochim. Biophys. Acta, 1968, 163, 171–178.
35. K. Kaweewong, W. Garnjanagoonchorn, W. Jirapakkul and S. Roytrakul, Protein J., 2013, 32, 297–308.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details, voltage output, FT-IR, SEM, and optical images of cathodes after operations. See DOI: 10.1039/c6ra13052f

---