Can bicarbonate replace phosphate to improve the sustainability of bioelectrochemical systems for H₂ production?

Abstract

Phosphate is generally used as an effective electrolyte buffer in bioelectrochemical systems, but it is not sustainable. Bicarbonate, if it can replace phosphate as an alternative buffer, may improve the overall economic feasibility of microbial electrolysis cells (MEC) for its application in wastewater treatment and H₂ production. In this study, the performance of single-chamber MECs with combo buffers with different PO₄³⁻ to HCO₃⁻ (P/C) ratios was investigated. The results demonstrate that large (about 80%) but not complete replacement of PO₄³⁻ with HCO₃⁻ is feasible, indicating phosphate is necessary to maintain the electric current and the stability of MECs. The current density of MEC with P/C at 20/80 (in mol%) was comparable to that with a full phosphate buffer, and can be kept stable for as long as 1800 h, under 0.5 to 0.9 V applied voltages. Analysis by using molecular approaches, including denaturing gradient gel electrophoresis and DNA sequencing, shows that P/C ratios affect the microbial community structure of the bioanode biofilm, especially on the population of Geobacter, which is a predominant and key exoelectrogenic bacteria to produce an electric current in MEC. The results of this study will broaden the knowledge of the buffer effect and promote the potential applicability of MECs for H₂ production.

---

Introduction

As a nascent technology for hydrogen production from organic waste, microbial electrolysis cells (MEC) have been rapidly developed in recent years. However, there are still some problems, such as high internal resistance, low hydrogen recovery rate, the inappropriate reactor architecture and difficulty to scale up, etc., which restrict the application of MEC.^1–3 The reduction of internal resistance, which is greatly influenced by the electrolyte, can improve the performance of MEC and allow it to produce H₂ rapidly.^4–6 As a neutral pH system, a phosphate buffer solution is usually used in MECs, because phosphate provides a suitable growth environment for microbes, increases the conductivity of the solution and reduces the solution resistance.^7–11 But the difficulty of post-processing for phosphate may bring the risk of eutrophication in receiving water. So looking for the alternative buffer system or reducing the phosphorus content is greatly significant to the application of MEC.

The phosphate buffers have two main functions in MEC: (1) stabilize the solution pH; (2) facilitate the proton transfer. The number of dissociable protons, carried by monobasic and dibasic phosphates, is much more than the free proton.¹² Compared to phosphate, bicarbonate has smaller molecular weight and exhibits higher conductivity and mass transfer coefficient. Thus, bicarbonate can also effectively increase the transfer rate of proton and lower the solution resistance. Moreover, bicarbonate is much cheaper than phosphate. Fan et al. reported that phosphate can be fully replaced by bicarbonate in a microbial fuel cell (MFC), and a 38.6% increase of output power was obtained.¹² Merrill and Logan¹³ have explored the effects of different electrolytes and found that high bicarbonate condition can improve the conductivity of the solution to optimize the performance of MEC. While the impact on the long-term performance and stability of MEC system under carbonate-rich electrolyte condition is still lack of knowledge.

In this study, aiming to increase the sustainability of MEC by less relying on phosphate, a combo buffer system, namely phosphate and bicarbonate in combination with different ratios, were applied to investigate the compatibility and the impact of buffer on hydrogen production in MEC. This research will reveal the insight of buffer effects and laid the foundation for improving the economy of MEC in its application.

Experimental

The setup and the operation of MECs

To investigate the effect of different type of buffers on MEC, four single-chamber membrane-free MECs were prepared in two steps, following the procedure of previous study.¹⁴ First, bioanodes (graphite felts; 25 mm × 30 mm; 5 mm thick) were enriched respectively in four air-cathode MFCs that adapting to different buffers, which were mixed by the phosphate (PBS: NH₄Cl 0.62 g L⁻¹, NaH₂PO₄·2H₂O 5.54 g L⁻¹, Na₂HPO₄·12H₂O 23.08 g L⁻¹, KCl 0.26 g L⁻¹) and the bicarbonate buffer (NH₄Cl 0.16 g L⁻¹, NaHCO₃ 8.401 g L⁻¹, KCl 0.26 g L⁻¹), according to the different molar ratio of PO₄³⁻/HCO₃⁻, namely pure carbonate (P/C@0%), trace amount of phosphate (0.8%, mol%) supplemented carbonate (P/C@0.8%), phosphate 20 mM/bicarbonate 80 mM (P/C@20%), and pure phosphate buffer (P/C@100%). Each MFC was inoculated with a 10 mL of activated sludge (2000 mg L⁻¹) and 15 mL fresh medium solution containing sodium acetate (NaAc) as the sole carbon source. The cathode was made of a square carbon film (30 mm × 22 mm) made of acetylene black and was coated with 0.5 mg cm⁻² of Pt. After 2 weeks of inoculation, the power output of the air-cathode MFC exceeded 0.4 V in 100 Ω external resistors, indicating the bioanodes were successfully acclimated to generate current. Second, the acclimated MFC bioanodes were transferred into single-chamber MECs with a Pt-based carbon cloth cathode (0.5 mg Pt cm⁻²). The MEC bioanodes were further cultivated with the same medium and buffers as in step 1, together with 1 g L⁻¹ NaAc as the carbon source. After 10 cycles of operation of MEC applied with a voltage of 0.5 V, a stable current was generated, indicating the exoelectrogenic bacteria in bioanode film were successfully acclimated in MECs. Later, through changing the applied voltages from 0.5 V to 1.0 V, the performance of MEC operated with each of the distinct buffers was compared to obtain the optimum condition. After that, six MECs with the optimum buffer were prepared and were operated at respective applied voltages, i.e. 0.5 V, 0.6 V, 0.7 V, 0.8 V, 0.9 V, 1.0 V, for over 1800 h to find out the long-term stability of MECs.

Electrochemical analysis

Cyclic voltammetry (CV) of MECs bioanode were performed using an electrochemical analyzer (CHI 660C, Chenhua, China) with a saturated calomel electrode (SCE) as the reference electrode placed close to the anode. The scanning was conducted for three cycles under anode potential ranging from −0.72 V to 0.1 V vs. SCE at a scanning rate of 10 mV s⁻¹. The medium solution used in the electrochemical analysis was kept the same as that used in the batch test of MECs, but was constantly purged with N₂.

Microbial community and phylogenetic analysis

The microbial community and phylogenic information of the bioanode operated under different buffer conditions and applied voltages were investigated by using 16S rRNA gene-based molecular techniques. The extraction of bioanode DNA was conducted with an UltraClean®Soil DNA Isolation Kit (MO BIO Laboratories, San Diego, US) following the manufacturer's protocol. Polymerase chain reaction (PCR) amplification and denaturing gradient gel electrophoresis (DGGE) of PCR products were conducted as previously described.¹⁵ The PCR primer set used for DGGE were 341FGC (5′-CGC CCG CCG CGC GCG GCG GGCGGG GCG GGG GCA CGG GGG GCC TAC GGG AGG CAG CAG-3′) and 518R (5′-ATT ACC GCG GCT GCT GG-3′). Every distinct band in DGGE gel was sliced with a sterile blade and purified by using the “crush-and-soak” method.¹⁶ The purified DNA was then amplified by PCR again with the same primer set 341F (without GC clamp) and 518R. The PCR products were sequenced with 341F primer and the sequencing results were searched using BLAST to analyze the phylogenetic information.

Results and discussion

The effect of buffers with different P/C ratios on MECs' performance

Electrolyte has a direct impact on the electrocatalytic capability of exoelectrogenic bacteria in bioanode and consequently affect the performance of MEC. In this study, four electrolyte buffers with different molar ratios of phosphate and bicarbonate (P/C@100%, P/C@20%, P/C@0.8%, P/C@0%, all in 0.1 M) were applied to investigate their influence on the current generation of MECs. Results show that the buffer with higher phosphate proportions is benefit for electrocatalytic performance of bioanodes (Fig. 1). Both current densities (Fig. 1a) and the linear sweep voltammogram (LSV) (Fig. 1b) of MECs with four different electrolyte buffers varied greatly.

Fig. 1 (a) The variation of current density with applied voltages on MECs and (b) CV of bioanodes with buffers at different P/C ratios (the scanning rate: 10 mV s⁻¹).

Generally, current density of MECs increases with applied voltages. However, our MECs did not behave in this way, especially for the MECs with buffers at low P/C conditions, where the current density of MECs increased positively with applied voltages initially, but began to drop when the applied voltage was over 0.8 V (Fig. 1a). LSV results also show that the bioanode of MECs with lower phosphate proportions was not as electrochemically active as those with higher amount of phosphate, indicating that the content of phosphate is a key factor for the better performance of MECs.

It is postulated that the type of buffer, rather than pH, have greater effect on MEC's performance. Usually, lower solution resistance is conducive to raise the current density and improve hydrogen production in MEC. However, it is quite uncommon that the MEC with pure bicarbonate (P/C@0%), which has the highest conductivity among the four MECs, exhibits the worst H₂ production performance (see ESI, Table S1†). Another interesting phenomenon is that buffer P/C@20% and P/C@0% are similar in pH, however, the MECs with these two buffers performed quite different. On the other hand, MECs with buffer P/C@100% and P/C@20% are considerably different in pH, but their H₂ production and current generation are in similar level. Ribot-Llobet et al.¹⁷ also found that the initial pH has no appreciable effect on H₂ production rates or total production. The above results indicate pH of electrolyte is not the key factor for the current generation of MEC and the slightly alkaline pH will not have adverse effect on the exoelectrogenic property of electro-active biofilm.

The microbial community analysis of bioanode in the MECs with different buffer systems

There is a great diversity of microbial population in the biocommunity of MFC's bioanode.^18–22 However, it is still not clear about the shift of bioanodic microbial community from MFC to MEC, which could be a result of a variety of factors, i.e. applied voltages, current densities, buffers and carbon sources that applied in MECs.

In this study, the dynamic structure of bacterial community of anode biofilms in the original MFCs and their transformation to MEC were investigated based on the assumption that the artificial augmentation of the circuit by external power could generally result in changes of the bacteria population.²¹ DGGE results demonstrate that there is a great phylogenetic diversity in the microbial communities of the anode biofilms of MFC and MEC (Fig. S1†). After purifying and PCR amplifying the distinct bands in DGGE gel, the 16S rDNA sequencing results show that Geobacter was the main bacteria (operational taxonomic units, OTUs) in MECs (Table S2†), which is consistent with the result reported by Kiely et al.²³ that microbial community of bioanodes in single-chamber MECs was dominated by clones with similarity close to Geobacter sp. Under phosphate-rich buffer conditions, Geobacter could reach over 60% of population in the bioanode microbial communities. Call et al.²⁴ also found that the sequences similar to G. sulfurreducens were predominant in the clone library obtained from a mixed-culture MEC (72% of clones), but with carbonate-rich buffer.

The population abundance of Geobacter in the bioanode microbial communities of our MECs was greatly influenced by the buffers, as shown in Fig. S1.† Results show that when P/C ratio of the buffer was extremely low, Geobacter was almost absent in the microbial communities, while Clostridium sp. and Soehngenia sp. became the main bacteria species, which are non-exoelectrogenic bacteria. The paradox is that carbonate buffer has been deemed important for generating significant Geobacter population.²³ Actually, the reported carbonate-buffered solution for the enrichment of Geobacter bacteria also contained a moderate amount of phosphate content (NaH₂PO₄ 0.6 g L⁻¹),²⁵ which indicates that phosphate, rather than carbonate, would be more necessary for enriching Geobacter. The reason for the absence of Geobacter under low phosphate condition is not clear so far. It is postulated that carbonate-rich solution will raise the pH to around 8.5, which is not the optimum pH for the growth of Geobacter. However, the scarcity of Geobacter population in the bioanode of MFCs under carbonate-rich buffers did not exhibit the inhibition for their power generation, most probably because the current generated in MFCs was much lower than that in MECs. As a consequence, the abundance of exoelectrogenic bacteria needed for electron generation in MFCs is not as much as those in MECs.

The long-term stability of MECs with buffer P/C@20%

Because of the better performance, the MECs with buffer P/C@20% were operated 1800 h to study the long-term stability of MEC, shown in Fig. 2. As usual, current densities of MECs were positively correlated to the increase of applied voltages. Results also demonstrate that after two months, all the current densities of MECs operated with respective applied voltages decreased more or less, probably because the degeneration of microbial activity happened. However, under 1.0 V of applied voltage, the current density of MEC decreased more significantly.

Fig. 2 The stability of the MECs operated under different applied voltages. (a–c) are different sampling points to study the microbial community shift in MEC bioanode. Buffer: P/C@20%, 0.1 M.

The CV test of bioanode sampled at 400 h and 1500 h of operation also shows that, the electrochemical activity and the double layer current of bioanode under 1.0 V applied voltage reduced significantly (Fig. 3), indicating not only the current generation capability of bacteria has reduced, but the electrochemically active bacteria population abundance decreased as well. Results also demonstrate that higher applied voltage, although brings about higher current, does not benefit for maintaining the exoelectrogenic bacteria population in microbial community of bioanode during the long-term operation of the MEC.

Fig. 3 The CV of bioanode sampled at 400 h and 1500 h of the MEC operated under applied voltage of 1.0 V. Buffer: P/C@20%, 0.1 M. The scanning rate: 10 mV s⁻¹.

Microbial community analysis of bioanode has shown that the main exoelectrogenic bacteria in our MECs is Geobacter sp., which is a strictly anaerobe existing in the interior of the anode biofilm close to the electrode. It has been proved that microorganisms in some strata of the biofilm, where nutrients have been locally depleted, may enter a nongrowing state in which they are less susceptible.²⁶ On the other hand, with the growing of bacteria, the biofilm will become thicker and more inert material left behind in the interior of the biofilm, which will also affect the contact of electroactive bacteria with the electrode. The higher of the current density, the more nutrients are needed to support the activity of exoelectrogenic microorganisms. Thus, higher applied voltages may bring the risk of the depletion of the substrate close to the interior of biofilm and cause of degeneration of exoelectrogenic bacteria in MEC, finally causes the decrease of the current and H₂ production of MECs.

The microbial population shift in the bioanode during long-term operation of MECs

Since current density decreased significantly for the MEC operated under 1.0 V applied voltage in two months' of operation, the microbial population in bioanode were studied by using the 16S rDNA based PCR-DGGE and cloning-sequencing analysis. DGGE results show that there was no significant difference for the microbial diversity in the anode biofilms (Fig. S2†). Among ten main OTUs, most of them belonged to Geobacter; except that OTU-S1, OTU-S2 and OTU-S3 were belong to Clostridium, Arcobacter and Pseudomonas, respectively. However, the microbial population abundance varied greatly, and the decrease in proportion of total amount Geobacter quite well with the decay of current in MEC, as shown in Fig. 4. This indicates the decrease of Geobacter contributes most to the decline of current density of MEC.

Fig. 4 The microbial population shift in the bioanode of the MEC applied with 1.0 V of voltage during two months of operation.

Conclusions

Buffer type has a direct impact on the electrocatalytic capability of bioanode in MECs. Largely replacing phosphate with bicarbonate in the electrolyte buffer is feasible to maintain the electroactivity of bioanode in MEC. With a buffer of P/C@20%, MEC can perform stably as long as 1800 h under the applied voltages from 0.5 to 0.9 V. Results also demonstrate that voltage higher than 1.0 V is not benefit for the current generation of MEC in long-term operation because higher voltage applied on MEC mainly causes the decrease of the population abundance of Geobacter in bioanode.

Acknowledgements

This work was financial supported by grants from the National Natural Science Foundation of China (Grant no. 21373022, 51108014, U1137602 and 51422301), Beijing Nova program (Z131109000413008), National Basic Research Program of China (973 Program) (no. 2011CB935700), Fundamental Research Funds for the Central Universities and Research Fund for the Doctoral Program of Higher Education of China (20111102120045).

Notes and references

1. S. Cheng and B. E. Logan, Proc. Natl. Acad. Sci. U. S. A., 2007, 104, 18871.
2. H. S. Lee, C. I. Torres, P. Parameswaran and B. E. Rittmann, Environ. Sci. Technol., 2009, 43, 7971.
3. B. E. Logan, D. Call, S. Cheng, H. V. M. Hamelers, T. H. J. A. Sleutels, A. W. Jeremiasse and R. A. Rozendal, Environ. Sci. Technol., 2008, 42, 8630.
4. H. Liu, H. Hu, J. Chignell and Y. Fan, Biofuels, 2010, 1, 129.
5. J. Y. Nam and B. E. Logan, Int. J. Hydrogen Energy, 2012, 37, 18622.
6. K. Rabaey, P. Girguis and L. K. Nielsen, Curr. Opin. Biotechnol., 2011, 22, 371.
7. D. S. Aaron, Transport in fuel cells: electrochemical impedance spectroscopy and neutron imaging studies, 2010.
8. J. C. Biffinger, R. Ray, B. Little and B. R. Ringeisen, Environ. Sci. Technol., 2007, 41, 1444.
9. S. Cheng, H. Liu and B. E. Logan, Environ. Sci. Technol., 2006, 40, 2426.
10. Z. He, Y. Huang, A. K. Manohar and F. Mansfeld, Bioelectrochemistry, 2008, 74, 78.
11. J. R. Kim, S. Cheng, S. E. Oh and B. E. Logan, Environ. Sci. Technol., 2007, 41, 1004.
12. Y. Fan, H. Hu and H. Liu, Environ. Sci. Technol., 2007, 41, 8154.
13. M. D. Merrill and B. E. Logan, J. Power Sources, 2009, 191, 203.
14. D. Liang, Y. Liu, S. Peng, F. Lan, S. Lu and Y. Xiang, Front. Environ. Sci. Eng., 2014, 8, 624.
15. D. Liang, T. Zhang and H. H. P. Fang, Water Res., 2007, 41, 2879.
16. A. M. Maxam and W. Gilbert, Proc. Natl. Acad. Sci. U. S. A., 1997, 74, 560.
17. E. Ribot-Llobet, J. Y. Namb, J. C. Tokash, A. Guisasola and B. E. Logan, Int. J. Hydrogen Energy, 2013, 38, 2951.
18. B. E. Logan, C. Murano, K. Scott, N. D. Grayc and I. M. Headc, Water Res., 2005, 39, 942.
19. B. H. Kim, H. S. Park, H. J. Kim, G. T. Kim, I. S. Chang, J. Lee and N. T. Phung, Appl. Microbiol. Biotechnol., 2004, 63, 672.
20. N. T. Phung, J. Lee, K. H. Kang, I. S. Chang, G. M. Gadd and B. H. Kim, FEMS Microbiol. Lett., 2004, 233, 77.
21. D. E. Holmes, D. R. Bond, R. A. O'Neil, C. E. Reimers, L. R. Tender and D. R. Lovley, Microbial Ecology, 2004, 48, 178.
22. K. Rabaey, N. Boon, S. D. Siciliano, M. Verhaege and W. Verstraete, Appl. Environ. Microbiol., 2004, 70, 5373.
23. P. D. Kiely, J. M. Regan and B. E. Logan, Curr. Opin. Biotechnol., 2011, 22, 378.
24. D. F. Call, R. C. Wagner and B. E. Logan, Appl. Environ. Microbiol., 2009, 75, 7579.
25. J. F. Caccavo, D. J. Lonergan, D. R. Lovley, M. Davis, J. F. Stolz and M. J. McInerney, Appl. Environ. Microbiol., 1994, 60, 3752.
26. P. S. Stewart, J. Bacteriol., 2003, 185, 1485.

---

Footnote

† Electronic supplementary information (ESI) available: pH and conductivity of different electrolyte buffers; the BLAST results of the distinct DGGE band in MECs; DGGE profile showing the microbial diversity and population in the bioanodes of MECs, etc. See DOI: 10.1039/c5ra00702j

---