- Email: [email protected]
Steady-state performance and chemical efficiency of Microbial Electrolysis Cells
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/he
Steady-state performance and chemical efficiency of Microbial Electrolysis Cells Tom H.J.A. Sleutels a,*, Annemiek Ter Heijne b, Cees J.N. Buisman a,b, Hubertus V.M. Hamelers a a
Wetsus, Centre of Excellence for Sustainable Water Technology, Agora 1, P.O. Box 1113, 8900 CC Leeuwarden, The Netherlands b Sub-Department of Environmental Technology, Wageningen University, Bornse Weilanden 9, P.O. Box 17, 6708 WG Wageningen, The Netherlands
article info
abstract
Article history:
The objective of this paper was to study MEC performance at steady-state conditions in
Received 25 January 2013
continuous mode and to analyse MEC performance in terms of chemical efficiency. At
Received in revised form
steady-state operation, a current density of 10.2 A m2 (applied voltage 1.0 V) for a set-up
8 April 2013
with an AEM was produced, compared to 7.2 A m2 for a set-up with a CEM. For all applied
Accepted 13 April 2013
voltages, total internal resistance for the AEM configuration was lower than or the CEM
Available online 6 May 2013
configuration. Therefore, energy input for the AEM configuration is lower than for the CEM configuration. In case a CEM is used, the conductivity in the cathode reaches high values:
Keywords:
>130 mS cm1. This conductivity is mainly caused by the presence of Naþ (7.8 g L1), Kþ
Bioelectrochemical system
(12.2 g L1) and OH (8.3 g L1). Furthermore, MECs perform better at high buffer and
Microbial Fuel Cell
electrolyte concentrations. However, as current density does not increase proportionally
Microbial Electrolysis Cell
with increase in chemicals, the effectiveness of chemical addition decreases when more
Internal resistance
chemicals are added. Therefore, addition of chemicals and buffer does not necessarily
Efficiency
enhance performance but increases operational costs. Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
1.
Introduction
Organic material in waste streams is a suitable source for the production of renewable energy and chemicals [1]. Bioelectrochemical systems (BES) are a promising technology to extract the energy from these waste streams to produce a variety of products like, for example, hydrogen gas [2,3]. The bioanode, at which microorganisms convert the chemical energy in organic material to electrical energy, forms the basis of most BESs. Depending on the cathode reaction, these systems are referred to as Microbial Fuel Cells (MFCs) when electricity is produced or Microbial Electrolysis Cells (MECs) when extra energy is added to the electrons to produce products at the cathode (e.g. hydrogen, methane) [4e8].
The challenge to bring BES research further towards application lies in the integration of the many different fields that are relevant for the topic: from electrochemistry, to biotechnology, to microbiology, to modelling, and process technology. This combination of fields makes BES research challenging, and the need for standardization of electrochemical measurements and experimental design has been recently acknowledged [9,10]. Essential in BES research is that results are obtained under conditions that are relevant for practical application of the technology, i.e. relevant electrolytes (low conductivity, neutral pH at the anode), continuous mode, and stable current generation. Moreover, addition of high concentrations of phosphate buffers should be avoided [9] as it is not feasible in practice [3].
* Corresponding author. Tel.: þ31 58 2843000; fax: þ31 58 2843001. E-mail address: [email protected] (T.H.J.A. Sleutels). 0360-3199/$ e see front matter Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijhydene.2013.04.067
7202
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
In contrast, many long term studies on BESs have been performed and their analysis has been done in a situation where only a steady current density has been reached [11,12]. A true steady-state situation however, is not only reflected in stable current generation, but also includes stable pH, conductivity, and ion concentrations in both electrolytes. In most cases the concentration of ions, and thus the conductivity, in anolyte and catholyte still changes even though the current has equilibrated, and therefore a shift within the internal resistance still takes place. Our objective in this study was twofold: (i) to study MEC performance at steady-state conditions in continuous mode, and (ii) to use a novel method to analyse MEC performance in terms of chemical efficiency. To study steady-state performance, our approach was to monitor all variables that are required to obtain a true steadystate. Once this steady-state was reached, we analysed the performance and partial internal resistances of systems operated with an AEM and a CEM. To analyse MEC performance in terms of chemical efficiency, we compared the input of chemicals in the form of buffer to the produced current.
2.
Materials and methods
2.1.
Microbial Electrolysis Cell
The MEC consisted of an anode and cathode compartment first separated by a CEM (Fumasep FKE, FuMa-Tech GmbH, Germany) and later by an AEM (AEM; AMX e Neosepta, Tokuyama Corp., Japan). The anode compartment contained a 1 mm thick carbon felt (Technical Fibre Products Ltd., Kendal, United Kingdom) which was separated from the membrane by spacer material (PETEX 07-4000/64, Sefar BV, Goor, The Netherlands) to enable a perpendicular forced flow through the anode [13]. The cathode compartment contained a platinum coated (50 g m2) titanium mesh (thickness 1 mm, specific surface area 1.7 m2 m2 e Magneto Special Anodes BV, Schiedam, The Netherlands). Both anode and cathode had a working volume of 280 mL and the anolyte and catholyte were circulated over the compartment at a flow rate of 340 mL min1. The hydraulic retention time of the anode compartment was 0.9 h. Anode, cathode and membrane all had a projected surface area of 250 cm2. Potentials of anode and cathode were measured using Ag/AgCl reference electrodes (þ0.200 V vs NHE. ProSense QiS, Oosterhout, The Netherlands) which were connected to the cell using capillaries filled with 3 M of KCl solution separated from the electrolyte by an agar salt bridge. The cell voltage was applied (0.6, 0.8 and 1.0 V) by an adjustable power supply (ES 03-5, Delta Electronica BV, Zierikzee, The Netherlands). All experiments were performed at 303 K.
2.2. Experimental procedures and steady-state conditions The MEC was inoculated with 100 mL of effluent from a working MEC [14]. The anode of the system was continuously fed with synthetic wastewater at a rate of 5 mL min1. The
synthetic wastewater contained 1.36 g L1 NaCH3COO$3H2O, 0.68 g L1 KH2PO4, 0.87 g L1 K2HPO4, 0.74 g L1 KCl, 0.58 g L1 NaCl, 0.28 g L1 NH4Cl, 0.1 g L1 MgSO4.7H2O, 0.1 g L1 CaCl2.2H2O and 0.1 mL L1 of a trace element mixture [15]. At the start of the experiment the catholyte consisted of 10 mM of phosphate buffer. The system with the CEM was first operated at an applied voltage of 0.6 V until a steady state in ion transport through the membrane was reached (approximately two weeks depending on the produced current). Steady state was determined by a constant current output, a stable conductivity and pH in anode and cathode and by a constant ion content of both anode and cathode compartment. Constant ion content of both anolyte and catholyte was confirmed by measurement of all ions present. After these steady-state conditions the experiment was repeated at an applied voltage of 0.8 and 1.0 V. In both cases, again it took approximately two weeks to reach steady state. Finally, this whole procedure was repeated with the AEM. Cation concentrations were determined with inductively coupled plasma-optical emission spectroscopy (ICP-OES, Perkin Elmer Optima 3000XL). Anion concentrations were determined using ion chromatography (Metrohm 761 Compact IC equipped with a conductivity detector and a Metrosep A Supp 5 6.1006.520 column). Acetate concentrations were measured using ion chromatograph (Metrohm 761 Compact IC) equipped with a conductivity detector and an anion column (Metrosep A Supp 5 6.1006.520). Bicarbonate concentrations were determined using a total organic carbon analyzer (Shimadzu TOCVCPH). Finally, ammonium was measured using a standardized test kit for ammonium (ammonium cuvette test LCK303, XION 500 spectrophotometer, Dr. Lange Nederland B.V., The Netherlands). Conductivity and pH of both electrolytes was measured externally by taking a sample from the compartments (WTW pH/cond 340i, Weilheim, Germany).
2.3.
Calculations
The energy input and output of an MEC are determined by the produced current (A) and the applied voltage (V). At a higher applied voltage more current will be produced but also the required energy to produce the product is higher. The applied voltage (Ecell) is required to overcome the different internal resistances of the system and to overcome the thermodynamically required energy input for hydrogen production compared to the thermodynamically determined energy gain from the oxidized substrate which is called the equilibrium voltage. The internal resistance includes the ionic resistance of the electrolytes, anode and cathode overpotentials and the membrane transport resistance [16,17]. Summarizing the applied voltage consists of: (i) equilibrium voltage (Eeq), (ii) anode overpotential (han), (iii) cathode overpotential (hcat), (iv) ionic resistance (Eionic) and (v) membrane transport resistance (ET). The equilibrium voltage was calculated using the measured conditions with Eeq ¼ Ecat Ean ¼
E0cat
RT pH2 ln 2 2F Hþ cat
!
E0an
RT ½CH3 COO ln 2 9 8F HCO3 Hþ an
!
(1)
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
where Ecat is the theoretical cathode potential (V), Ean is the theoretical anode potential (V), E0cat is the standard cathode potential (V), E0an is the standard anode potential (V), R is the universal gas constant (8.3145 J mol1 K1), T is the temperature (303 K) and F is the Faraday’s constant (96,485 C mol1). The anode and cathode overpotentials (V) were calculated with han ¼ Ean;
measured
Ean
hcat ¼ Ecat Ecat;
measured
(2)
where Ean, measured (V) is the measured anode potential and Ecat, measured (V) is the measured cathode potential. The anode and cathode overpotentials were calculated using the actual concentrations of acetate, bicarbonate and protons and include the mass and charge transfer overpotential and the energy used for maintenance and growth of the organisms. The Ohmic loss of the electrodes was not measured separately and is included in the overpotential of the anode and cathode [18]. The ionic loss of the electrolytes (V) is related to the electrolyte resistance of the anolyte and catholyte and was calculated with [19] Eionic ¼ Iions
1 1 dan dcat þ Ran þ Rcat ¼ Iions 2Asan 2Ascat 2 2
(3)
where Iions is the flow of ions through the electrolyte (A) which is equal to the current, dan is the distance between the anode and the membrane (0.004 m), dcat is the distance between the cathode and the membrane (0.001 m), A is the surface area (0.025 m2), san is the anode conductivity (S m1) and scat is the cathode conductivity (S m1). The factor ½ is added to estimate the location of the microorganisms and of the produced electrons within the anode compartment since the microorganisms are not all located at equal distance from the membrane. The membrane transport loss (ET) is the potential loss caused by transport of ions through the membrane and was calculated from all other potential losses using [16] Ecell ¼ Eeq han hcat Eionic ET
(4)
The partial internal resistance of all the different components was then calculated from these potential losses according to ohms law (R ¼ V/I ). The specific energy input (kW h m3 H2) in the system was calculated with Eelectricity ¼
nPF Ecell 1 RT 3:6$106 rcat
Cchem Q in F I
where Cchem is the chemical input (mol eq L1), I is the current density (A m2), Qin is the influent flow rate (L s1) and F is Faradays constant (C mol1 e).
3.
Results and discussion
3.1.
Steady-state performance
True steady-state situation in MECs is indicated not only by a constant produced current, but also by a constant cathodic pH and cathodic conductivity. Fig. 1 shows an example of MEC start-up towards steady-state performance in this case, equipped with a CEM and at an applied voltage of 1 V. For this specific experiment, the steady-state performance was reached after approximately 12 days, but for other experiments this time period varied depending on the applied voltage and produced current. The change in ion concentration is highly dependent on the current density produced by the system and steady-state is therefore reached faster at high current densities. Table 1 shows an overview of the steady-state performance of MECs equipped with a CEM and an AEM at different applied voltages. The first indication of the performance of an MEC is the produced current density. At low applied voltage (0.6 V) the current densities for both systems are similar. At higher applied voltages (1.0 V) the differences in performance become larger. This eventually led to a current density of 10.2 A m2 at an applied voltage of 1.0 V for the set-up with the AEM, compared to 7.2 A m2 for the set-up with the CEM. The main cause for the difference in performance between CEM and AEM was the difference in pH and conductivity of the catholyte. At high cathode pH, the thermodynamically required energy to form hydrogen increases. At the same time, low cathode conductivity increases the internal resistance of the system. Both processes therefore lead to decreased performance. At 0.6 V applied voltage, the pH and conductivity of the catholyte in the AEM configuration were lower than the pH and conductivity of the catholyte in the CEM configuration. At 0.8 V and 1.0 V applied voltages, the pH for the AEM
(5)
where P is the pressure (105 Pa) and T is the absolute temperature (303.15 K), 3.6$106 is a conversion factor for the conversion of Joules to kW h, and rcat is the cathodic hydrogen recovery [2]. The anode compartment was continuously fed and therefore changes in produced current had little influence on the pH and conductivity. The anode compartment had a constant pH of 6.6 and a conductivity of 5.6 mS cm1. These values were used in the calculations. The contribution of the relative chemical input (Crel) per mol of produced electrons and was calculated with Crel ¼
7203
(6)
Fig. 1 e Example of typical steady-state performance of an MEC equipped with a CEM at an applied voltage of 1 V. Steady-state conditions are indicated by a constant current density, cathodic conductivity and cathodic pH.
7204
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
Table 1 e Steady-state performance of MECs equipped with CEM and AEM at applied voltages of 0.6, 0.8 and 1.0 V. The energy input is calculated assuming a cathodic coulombic efficiency of 95%. Membrane CEM
AEM
Applied voltage (V)
Current density (A m2)
pH cathode (e)
Conductivity cathode (mS cm1)
Energy input (kW h m3 H2)
0.6 0.8 1.0 0.6 0.8 1.0
0.95 2.74 7.2 1.12 4.81 10.18
13.18 13.52 13.69 11.77 13.1 13.83
41 75 133 16 36 34
1.57 2.10 2.62 1.57 2.10 2.62
configuration increases to values comparable to those for the CEM configuration. The conductivity for the CEM configuration however, reached much higher values than the conductivity for the AEM configuration although this did not lead to higher current densities, since the contribution of the conductivity to the total internal resistance was relatively small. A high conductivity indicates a high ion concentration in the catholyte. In the case when a CEM is used these ions were mainly Naþ (7.8 g L1), Kþ (12.2 g L1) and OH (8.3 g L1) at an applied voltage of 1 V. Although the pH increase in the cathode for different types of membranes has been the topic of research before [6,20,21], their performance has only been studied in short term experiments.
3.2.
Partial internal resistance analysis
A more detailed analysis of the performance of the MEC can be provided by analysing the contribution of the internal resistance of the different components during steady-state performance of the MEC. Fig. 2 gives an overview of these internal resistances for both configurations at different applied voltages. Firstly, from this figure it becomes clear that the total internal resistance for the AEM configuration is lower than the internal resistance for the CEM configuration for all applied voltages. So, at all applied voltages the AEM configuration outperformed the CEM configuration, which is in agreement with previous studies [6,16]. Secondly, in the AEM configuration, increase in applied voltage leads to a more pronounced reduction in internal resistance compared to the CEM configuration.
When looking at the contribution of the different components of the configurations it can be seen that for both MECs, at all applied voltages, the equilibrium resistance for hydrogen production contributes for over 60% of the total internal resistance. The equilibrium voltage (resistance) represents the amount of Gibbs-energy that needs to be invested for the production of hydrogen. In principle, this energy can be regained in the form of chemical energy stored in the produced hydrogen. The actual required energy however, is higher than the theoretical value, since the pH in both the anode and cathode changes. At low applied voltages, the increase of the cathode pH is more pronounced in the CEM configuration than in the AEM configuration. Therefore, the required energy input for the AEM configuration is lower than for the CEM configuration. This is advantageous in terms of energy efficiency; although the hydrogen production rate will be lower, the required energy for hydrogen production is also lower. When designing a full scale system a balance has to be found between energy input and production rate. Both the anode and cathode resistance decrease with increasing current density for both membrane configurations. Generally, it is considered that the anode overpotential increases at higher applied voltages [9]. The anode resistance however, is the anode overpotential divided by the current density, and because the current density increases with higher applied voltages at higher rate than the overpotential, the anode resistance decreases at higher current densities. The resistance for transport of ions through the membrane was lower for the AEM configuration compared to the CEM configuration at all applied voltages, which is in agreement with previous findings [16]. Finally, the ionic resistance for this
Fig. 2 e Distribution of the internal resistances of MECs equipped with CEM and AEM at applied voltages of 0.6, 0.8 and 1.0 V. The ionic resistance is too small to be visible.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
specific reactor configuration was low (<10 mU) and hardly contributed to the total required energy input even at this low ionic strength (5 mS cm1). For both configurations, the pH increase in the cathode compartment causes the largest energy loss as shown by the equilibrium resistance. Therefore, the largest reduction of the required specific energy input of the system would be a decrease of this cathode pH. Use of an AEM decreases the cathode pH compared to the use of a CEM. Another proposed possibility to reduce the pH in the cathode is by addition of CO2 to the catholyte [22,23]. There have been studies that directly measured all partial internal resistances of MFCs using electrochemical impedance spectroscopy (EIS) [24,25]. Interpretation of EIS data however, highly depends on the model used and often several parameters are mixed in the measured spectra [26]. The internal resistance analysis proves a valuable alternative method to directly calculate the contribution of the partial resistances to the total internal resistance from measured data, which can give valuable information about the limitations of the system.
3.3.
Energy input and output
The energy efficiency of an MEC is determined by the energy output divided by the energy input. The energy input is determined as the electrical energy that is required to produce 1 m3 of hydrogen gas (kW h m3 H2), when the energy input from the waste product (in this case acetate) is not considered. The energy output is expressed as the chemical energy stored in this hydrogen gas. The energy input is highly dependent on the applied voltage and the cathodic efficiency. The applied voltage determines how much energy is added for every produced electron while the cathodic efficiency determines how many of these electrons end up as hydrogen gas. Therefore, at higher applied voltages the specific energy input, being the energy input per electron, is higher than at lower applied voltage at constant cathodic efficiency (Table 1). At higher applied voltages, generally, the rate at which electrons are produced is higher. For economical purposes, the goal is to produce a high current at low applied voltages (low internal resistance). Therefore, economically speaking, it is more interesting to use an AEM as physical separator between anode and cathode because of the lower internal resistance of an MEC equipped with an AEM.
3.4.
Chemical efficiency
In many lab-scale studies, considerable amounts of salts and buffer are added to improve the conductivity and to prevent biofilm acidification ([27]). Although these high concentrations of salts and buffer improve BES performance, their addition is undesirable from an economic and environmental point of view. Addition of extra chemicals not only increases the operational costs of the system, but also leads to an extra treatment step before the effluent of the BES can be discharged. For practical application of BESs, it is therefore interesting to study the effect of the addition of chemicals and buffer on the conversion rate. In general, a high conversion rate (of COD) at low chemical input is desirable. We therefore studied the
7205
contribution of electrolyte and buffer to the performance (current density) of the system; and for its evaluation we introduce a new concept: the relative chemical input, which indicates the moles chemical equivalents needed per mole of produced electrons. Three types of chemical inputs were considered: electrolyte, buffer, and substrate. First of all, the differences in MEC performance with an anion exchange membrane (AEM) and a cation exchange membrane (CEM) were considered. It has been shown in previous research and in this study that an MEC with an AEM outperforms an MEC with a CEM [6,16]. This also has a direct influence on the relative chemical input for the system. Fig. 3A shows the relative chemical input for both systems and shows that the total chemical input for the AEM is half of the chemical input of the CEM (data from [16]). This can be explained by the fact that both systems used influent with the same chemical composition while the AEM produced higher current density (5.3 A m2) than the CEM (2.3 A m2). Therefore, the relative contribution of these chemicals in the best performing system (AEM) was the lowest. This shows that when an AEM is applied, not only the performance is better, but because of this, the investment in the form of chemicals, expressed per electron, is lower compared to a CEM. Secondly, we studied the differences between systems with different buffer concentrations. Previous research showed that addition of buffer has a positive effect on the current production of MECs [13,27,28]. Also in this case, it is interesting to study the effect of chemical input on productivity. Fig. 3B shows the relative chemical input for identical systems using influent buffer concentrations of 10 mM and 50 mM (data from [13]). Although the system with the high buffer concentration (50 mM P) produced considerably higher current (10.2 A m2 vs. 7.8 A m2 at 10 mM), the chemical input for 10 mM buffer relative to the current produced was only one third of the chemical input for 50 mM buffer. Only when the current density would be 5 times as high as well, the chemical efficiency would stay the same. However, because the current density does not increase proportionally with the increase in buffer concentration, the contribution in mol equivalents of buffer per electron is much lower at high buffer concentrations. On the other hand, a lower buffer concentration results in a lower ionic strength and thus a higher ohmic resistance. Therefore, thirdly, we studied the differences between systems with different ionic strength. It is generally accepted that systems with low conductivity, which have high ohmic losses, perform worse than systems with high conductivity. Fig. 3C shows the relative chemical requirement for MECs with influent with 10 mM of phosphate buffer, 0 mM of phosphate buffer and with addition of electrolyte, not in the form of buffer (data from [29]). This figure shows that required input in the form of chemicals to produce a certain current (mole of electrons per time unit) decreases at low buffer concentrations. Again, the implication is that the system performs better at high buffer and electrolyte concentrations, but as the current density does not increase proportionally with the increase in chemicals, the effectiveness of chemical addition decreases when more chemicals are added. However, at low
7206
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
This study analysed the performance of MECs at steadystate conditions. Steady-state conditions are particularly interesting because these will be the actual operating conditions of future continuous BESs. It would therefore be most useful to see more steady-state studies in the future to make a realistic estimate of future applicability of BESs. Furthermore, it was shown that addition of large amounts of buffer and electrolyte is unnecessary from a chemical efficiency point of view since it only increases costs of the system.
4.
Fig. 3 e Relative chemical input for an MEC equipped with a CEM compared to a system equipped with an AEM (A); for an MEC operated with 10 and 50 mM of phosphate buffer (B) and for an MEC operated with 10 mM and 0 mM of phosphate buffer and without addition of extra salts in the anolyte (C).
Conclusions
Our objective in this study was twofold: (i) to study MEC performance at steady-state conditions in continuous mode, and (ii) to use a novel method to analyse MEC performance in terms of chemical efficiency. At steady-state operation, a current density of 10.2 A m2 at an applied voltage of 1.0 V for the set-up with the AEM was produced, compared to 7.2 A m2 for the set-up with the CEM. At steady-state conditions, the total internal resistance for the AEM configuration is lower than the internal resistance for the CEM configuration for all applied voltages. Also, in the AEM configuration, increase in applied voltage leads to a more pronounced reduction in internal resistance compared to the CEM configuration. Therefore, the required energy input for the AEM configuration is lower than for the CEM configuration. The pH gradient over the membrane causes the largest part of the internal resistance for both membrane configurations. At lower (<1.0 V) applied voltages, the relative contribution to the total internal resistance of the cathode pH is more pronounced in the CEM configuration than in the AEM configuration. In case a CEM is used, the conductivity in the cathode reaches very high values: (>130 mS cm1). This conductivity is mainly caused by the presence of Naþ (7.8 g L1), Kþ (12.2 g L1) and OH (8.3 g L1). Furthermore, MECs perform better at high buffer and electrolyte concentrations. However, as the current density does not increase proportionally with the increase in chemicals, the effectiveness of chemical addition decreases when more chemicals are added. We found that, when expressed per produced electron, addition of chemicals and buffer does not necessarily enhance performance but increases operational costs and is therefore not a cost-effective strategy towards practical application.
Acknowledgements ionic strength, it is important that good mass transfer is ensured. One way to improve electrode performance is the use of electrodes with a high specific surface like graphite felt, granules or brushes [13,30,31], which has also been shown to be very suitable to improve cathode performance [32]. However, inside these electrodes a sufficient mass transfer, e.g. via forced flow, is crucial [13,14]. So, to a certain extent, addition of (low concentrations of) chemicals and buffer is beneficial for the performance of BESs, but their addition is less effective at higher concentrations.
This work was performed in the TTIW-cooperation framework of Wetsus, centre of excellence for sustainable water technology (www.wetsus.nl). Wetsus is funded by the Dutch Ministry of Economic Affairs, the European Union European Regional Development Fund, the Province of Fryslan, the City of Leeuwarden and by the EZ-KOMPAS Program of the “Samenwerkingsverband Noord-Nederland”. The authors like to thank the participants of the research theme “Resource Recovery” for the fruitful discussions and their financial support.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
Nomenclature
Eeq Equilibrium voltage, V Cathode overpotential, V hcat Ionic resistance, V Eionic Anode overpotential, V han Membrane transport resistance, V Et Theoretical cathode potential, V Ecat Theoretical anode potential, V Ean standard cathode potential, V E0cat standard anode potential, V E0an R Universal gas constant, 8.3145 J mol1 K1 T Temperature, K F Faraday constant, 96,485 C mol1 Ean, measured Measured anode potential, V Ecat, measured Measured cathode potential, V Flow of ions through electrolyte, A Iions Distance between anode and membrane, 0.004 m dan Distance between cathode and membrane, 0.001 m dcat A Projected surface area, 0.025 m2 san Anode conductivity, S m1 scat Cathode conductivity, S m1 Eelectricity Specific energy input, kW h m3 H2 P Pressure, 105 Pa Cathodic hydrogen recovery, % rcat Relative chemical input, Mol eq (mol e1)1 Crel Cchem Chemical input, mol eq L1 I Current density, A m2 Qin Influent flow rate, L s1
references
[1] Rabaey K, Rozendal RA. Microbial electrosynthesis e revisiting the electrical route for microbial production. Nature Reviews Microbiology 2010;8:706e16. [2] Logan BE, Call D, Cheng S, Hamelers HVM, Sleutels THJA, Jeremiasse AW, et al. Microbial electrolysis cells (MECs) for high yield hydrogen gas production from organic matter. Environmental Science & Technology 2008;42:8630e40. [3] Sleutels THJA, Ter Heijne A, Buisman CJN, Hamelers HVM. Bioelectrochemical systems: an outlook for practical applications. ChemSusChem 2012;5:1012e9. [4] Cheng S, Xing D, Call DF, Logan BE. Direct biological conversion of electrical current into methane by electromethanogenesis. Environmental Science and Technology 2009;43:3953e8. [5] Hamelers HVM, Ter Heijne A, Sleutels THJA, Jeremiasse AW, Strik DPBTB, Buisman CJN. New applications and performance of bioelectrochemical systems. Applied Microbiology and Biotechnology 2010;85:1673e85. [6] Cheng S, Logan BE. Sustainable and efficient biohydrogen production via electrohydrogenesis. Proceedings of the National Academy of Sciences of the United States of America 2007;104:18871e3. [7] Rozendal RA, Hamelers HVM, Euverink GJW, Metz SJ, Buisman CJN. Principle and perspectives of hydrogen production through biocatalyzed electrolysis. International Journal of Hydrogen Energy 2006;31:1632e40.
7207
[8] Van Eerten-Jansen MCAA, Heijne AT, Buisman CJN, Hamelers HVM. Microbial electrolysis cells for production of methane from CO2: long-term performance and perspectives. International Journal of Energy Research 2012;36:809e19. [9] Jeremiasse AW, Hamelers HVM, Kleijn JM, Buisman CJN. Use of biocompatible buffers to reduce the concentration overpotential for hydrogen evolution. Environmental Science and Technology 2009;43:6882e7. [10] Jeremiasse AW, Hamelers HVM, Buisman CJN. Microbial electrolysis cell with a microbial biocathode. Bioelectrochemistry 2010;78:39e43. [11] Zhuang L, Yuan Y, Wang Y, Zhou S. Long-term evaluation of a 10-liter serpentine-type microbial fuel cell stack treating brewery wastewater. Bioresource Technology 2012;123:406e12. [12] Escapa A, Gil-Carrera L, Garcı´a V, Mora´n A. Performance of a continuous flow microbial electrolysis cell (MEC) fed with domestic wastewater. Bioresource Technology 2012;117:55e62. [13] Sleutels THJA, Lodder R, Hamelers HVM, Buisman CJN. Improved performance of porous bio-anodes in microbial electrolysis cells by enhancing mass and charge transport. International Journal of Hydrogen Energy 2009;34:9655e61. [14] Sleutels THJA, Hamelers HVM, Buisman CJN. Effect of mass transfer speed and direction in porous anodes on microbial electrolysis cell performance. Bioresource Technology 2011;102:399e403. [15] Zehnder AJB, Huser BA, Brock TD, Wuhrmann K. Characterization of an acetate-decarboxylating, nonhydrogen-oxidizing methane bacterium. Archives Of Microbiology 1980;124:1e11. [16] Sleutels THJA, Hamelers HVM, Rozendal RA, Buisman CJN. Ion transport resistance in microbial electrolysis cells with anion and cation exchange membranes. International Journal of Hydrogen Energy 2009;34:3612e20. [17] Fan Y, Sharbrough E, Liu H. Quantification of the internal resistance distribution of microbial fuel cells. Environmental Science and Technology 2008;42:8101e7. [18] Rozendal RA, Sleutels THJA, Hamelers HVM, Buisman CJN. Effect of the type of ion exchange membrane on performance, ion transport, and pH in biocatalyzed electrolysis of wastewater. Water Science and Technology 2008;57:1757e62. [19] Ter Heijne A, Hamelers HVM, De Wilde V, Rozendal RA, Buisman CJN. A bipolar membrane combined with ferric iron reduction as an efficient cathode system in microbial fuel cells. Environmental Science & Technology 2006;40:5200e5. [20] Rozendal RA, Hamelers HVM, Molenkamp RJ, Buisman CJN. Performance of single chamber biocatalyzed electrolysis with different types of ion exchange membranes. Water Research 2007;41:1984e94. [21] Kim JR, Cheng S, Oh SE, Logan BE. Power generation using different cation, anion, and ultrafiltration membranes in microbial fuel cells. Environmental Science & Technology 2007;41:1004e9. [22] Torres CI, Lee HS, Rittmann BE. Carbonate species as OH carriers for decreasing the pH gradient between cathode and anode in biological fuel cells. Environmental Science and Technology 2008;42:8773e7. [23] Fornero JJ, Rosenbaum M, Cotta MA, Angenent LT. Carbon dioxide addition to microbial fuel cell cathodes maintains sustainable catholyte pH and improves anolyte pH, alkalinity, and conductivity. Environmental Science and Technology 2010;44:2728e34. [24] Sleutels THJA, Darus L, Hamelers HVM, Buisman CJN. Effect of operational parameters on coulombic efficiency in
7208
[25]
[26]
[27]
[28]
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
bioelectrochemical systems. Bioresource Technology 2011;102:11172e6. Deeke A, Sleutels THJA, Hamelers HVM, Buisman CJN. Capacitive bioanodes enable renewable energy storage in microbial fuel cells. Environmental Science and Technology 2012;46:3554e60. Jeremiasse AW, Hamelers HV, Croese E, Buisman CJ. Acetate enhances startup of a H2-producing microbial biocathode. Biotechnology and Bioengineering 2012;109:657e64. Torres CI, Marcus AK, Rittmann BE. Proton transport inside the biofilm limits electrical current generation by anoderespiring bacteria. Biotechnology and Bioengineering 2008;100:872e81. Liu H, Cheng S, Logan BE. Power generation in fed-batch microbial fuel cells as a function of ionic strength,
[29]
[30]
[31]
[32]
temperature, and reactor configuration. Environmental Science and Technology 2005;39:5488e93. Sleutels THJA, Hamelers HVM, Buisman CJN. Reduction of pH buffer requirement in bioelectrochemical Systems. Environmental Science & Technology 2010;44:8259e63. Logan B, Cheng S, Watson V, Estadt G. Graphite fiber brush anodes for increased power production in air-cathode microbial fuel cells. Environmental Science & Technology 2007;41:3341e6. Rabaey K, Ossieur W, Verhaege M, Verstraete W. Continuous microbial fuel cells convert carbohydrates to electricity. Water Science and Technology 2005:515e23. Jeremiasse AW, Hamelers HVM, Saakes M, Buisman CJN. Ni foam cathode enables high volumetric H2 production in a microbial electrolysis cell. International Journal of Hydrogen Energy 2010;35:12716e23.
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/he
Steady-state performance and chemical efficiency of Microbial Electrolysis Cells Tom H.J.A. Sleutels a,*, Annemiek Ter Heijne b, Cees J.N. Buisman a,b, Hubertus V.M. Hamelers a a
Wetsus, Centre of Excellence for Sustainable Water Technology, Agora 1, P.O. Box 1113, 8900 CC Leeuwarden, The Netherlands b Sub-Department of Environmental Technology, Wageningen University, Bornse Weilanden 9, P.O. Box 17, 6708 WG Wageningen, The Netherlands
article info
abstract
Article history:
The objective of this paper was to study MEC performance at steady-state conditions in
Received 25 January 2013
continuous mode and to analyse MEC performance in terms of chemical efficiency. At
Received in revised form
steady-state operation, a current density of 10.2 A m2 (applied voltage 1.0 V) for a set-up
8 April 2013
with an AEM was produced, compared to 7.2 A m2 for a set-up with a CEM. For all applied
Accepted 13 April 2013
voltages, total internal resistance for the AEM configuration was lower than or the CEM
Available online 6 May 2013
configuration. Therefore, energy input for the AEM configuration is lower than for the CEM configuration. In case a CEM is used, the conductivity in the cathode reaches high values:
Keywords:
>130 mS cm1. This conductivity is mainly caused by the presence of Naþ (7.8 g L1), Kþ
Bioelectrochemical system
(12.2 g L1) and OH (8.3 g L1). Furthermore, MECs perform better at high buffer and
Microbial Fuel Cell
electrolyte concentrations. However, as current density does not increase proportionally
Microbial Electrolysis Cell
with increase in chemicals, the effectiveness of chemical addition decreases when more
Internal resistance
chemicals are added. Therefore, addition of chemicals and buffer does not necessarily
Efficiency
enhance performance but increases operational costs. Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
1.
Introduction
Organic material in waste streams is a suitable source for the production of renewable energy and chemicals [1]. Bioelectrochemical systems (BES) are a promising technology to extract the energy from these waste streams to produce a variety of products like, for example, hydrogen gas [2,3]. The bioanode, at which microorganisms convert the chemical energy in organic material to electrical energy, forms the basis of most BESs. Depending on the cathode reaction, these systems are referred to as Microbial Fuel Cells (MFCs) when electricity is produced or Microbial Electrolysis Cells (MECs) when extra energy is added to the electrons to produce products at the cathode (e.g. hydrogen, methane) [4e8].
The challenge to bring BES research further towards application lies in the integration of the many different fields that are relevant for the topic: from electrochemistry, to biotechnology, to microbiology, to modelling, and process technology. This combination of fields makes BES research challenging, and the need for standardization of electrochemical measurements and experimental design has been recently acknowledged [9,10]. Essential in BES research is that results are obtained under conditions that are relevant for practical application of the technology, i.e. relevant electrolytes (low conductivity, neutral pH at the anode), continuous mode, and stable current generation. Moreover, addition of high concentrations of phosphate buffers should be avoided [9] as it is not feasible in practice [3].
* Corresponding author. Tel.: þ31 58 2843000; fax: þ31 58 2843001. E-mail address: [email protected] (T.H.J.A. Sleutels). 0360-3199/$ e see front matter Copyright ª 2013, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijhydene.2013.04.067
7202
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
In contrast, many long term studies on BESs have been performed and their analysis has been done in a situation where only a steady current density has been reached [11,12]. A true steady-state situation however, is not only reflected in stable current generation, but also includes stable pH, conductivity, and ion concentrations in both electrolytes. In most cases the concentration of ions, and thus the conductivity, in anolyte and catholyte still changes even though the current has equilibrated, and therefore a shift within the internal resistance still takes place. Our objective in this study was twofold: (i) to study MEC performance at steady-state conditions in continuous mode, and (ii) to use a novel method to analyse MEC performance in terms of chemical efficiency. To study steady-state performance, our approach was to monitor all variables that are required to obtain a true steadystate. Once this steady-state was reached, we analysed the performance and partial internal resistances of systems operated with an AEM and a CEM. To analyse MEC performance in terms of chemical efficiency, we compared the input of chemicals in the form of buffer to the produced current.
2.
Materials and methods
2.1.
Microbial Electrolysis Cell
The MEC consisted of an anode and cathode compartment first separated by a CEM (Fumasep FKE, FuMa-Tech GmbH, Germany) and later by an AEM (AEM; AMX e Neosepta, Tokuyama Corp., Japan). The anode compartment contained a 1 mm thick carbon felt (Technical Fibre Products Ltd., Kendal, United Kingdom) which was separated from the membrane by spacer material (PETEX 07-4000/64, Sefar BV, Goor, The Netherlands) to enable a perpendicular forced flow through the anode [13]. The cathode compartment contained a platinum coated (50 g m2) titanium mesh (thickness 1 mm, specific surface area 1.7 m2 m2 e Magneto Special Anodes BV, Schiedam, The Netherlands). Both anode and cathode had a working volume of 280 mL and the anolyte and catholyte were circulated over the compartment at a flow rate of 340 mL min1. The hydraulic retention time of the anode compartment was 0.9 h. Anode, cathode and membrane all had a projected surface area of 250 cm2. Potentials of anode and cathode were measured using Ag/AgCl reference electrodes (þ0.200 V vs NHE. ProSense QiS, Oosterhout, The Netherlands) which were connected to the cell using capillaries filled with 3 M of KCl solution separated from the electrolyte by an agar salt bridge. The cell voltage was applied (0.6, 0.8 and 1.0 V) by an adjustable power supply (ES 03-5, Delta Electronica BV, Zierikzee, The Netherlands). All experiments were performed at 303 K.
2.2. Experimental procedures and steady-state conditions The MEC was inoculated with 100 mL of effluent from a working MEC [14]. The anode of the system was continuously fed with synthetic wastewater at a rate of 5 mL min1. The
synthetic wastewater contained 1.36 g L1 NaCH3COO$3H2O, 0.68 g L1 KH2PO4, 0.87 g L1 K2HPO4, 0.74 g L1 KCl, 0.58 g L1 NaCl, 0.28 g L1 NH4Cl, 0.1 g L1 MgSO4.7H2O, 0.1 g L1 CaCl2.2H2O and 0.1 mL L1 of a trace element mixture [15]. At the start of the experiment the catholyte consisted of 10 mM of phosphate buffer. The system with the CEM was first operated at an applied voltage of 0.6 V until a steady state in ion transport through the membrane was reached (approximately two weeks depending on the produced current). Steady state was determined by a constant current output, a stable conductivity and pH in anode and cathode and by a constant ion content of both anode and cathode compartment. Constant ion content of both anolyte and catholyte was confirmed by measurement of all ions present. After these steady-state conditions the experiment was repeated at an applied voltage of 0.8 and 1.0 V. In both cases, again it took approximately two weeks to reach steady state. Finally, this whole procedure was repeated with the AEM. Cation concentrations were determined with inductively coupled plasma-optical emission spectroscopy (ICP-OES, Perkin Elmer Optima 3000XL). Anion concentrations were determined using ion chromatography (Metrohm 761 Compact IC equipped with a conductivity detector and a Metrosep A Supp 5 6.1006.520 column). Acetate concentrations were measured using ion chromatograph (Metrohm 761 Compact IC) equipped with a conductivity detector and an anion column (Metrosep A Supp 5 6.1006.520). Bicarbonate concentrations were determined using a total organic carbon analyzer (Shimadzu TOCVCPH). Finally, ammonium was measured using a standardized test kit for ammonium (ammonium cuvette test LCK303, XION 500 spectrophotometer, Dr. Lange Nederland B.V., The Netherlands). Conductivity and pH of both electrolytes was measured externally by taking a sample from the compartments (WTW pH/cond 340i, Weilheim, Germany).
2.3.
Calculations
The energy input and output of an MEC are determined by the produced current (A) and the applied voltage (V). At a higher applied voltage more current will be produced but also the required energy to produce the product is higher. The applied voltage (Ecell) is required to overcome the different internal resistances of the system and to overcome the thermodynamically required energy input for hydrogen production compared to the thermodynamically determined energy gain from the oxidized substrate which is called the equilibrium voltage. The internal resistance includes the ionic resistance of the electrolytes, anode and cathode overpotentials and the membrane transport resistance [16,17]. Summarizing the applied voltage consists of: (i) equilibrium voltage (Eeq), (ii) anode overpotential (han), (iii) cathode overpotential (hcat), (iv) ionic resistance (Eionic) and (v) membrane transport resistance (ET). The equilibrium voltage was calculated using the measured conditions with Eeq ¼ Ecat Ean ¼
E0cat
RT pH2 ln 2 2F Hþ cat
!
E0an
RT ½CH3 COO ln 2 9 8F HCO3 Hþ an
!
(1)
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
where Ecat is the theoretical cathode potential (V), Ean is the theoretical anode potential (V), E0cat is the standard cathode potential (V), E0an is the standard anode potential (V), R is the universal gas constant (8.3145 J mol1 K1), T is the temperature (303 K) and F is the Faraday’s constant (96,485 C mol1). The anode and cathode overpotentials (V) were calculated with han ¼ Ean;
measured
Ean
hcat ¼ Ecat Ecat;
measured
(2)
where Ean, measured (V) is the measured anode potential and Ecat, measured (V) is the measured cathode potential. The anode and cathode overpotentials were calculated using the actual concentrations of acetate, bicarbonate and protons and include the mass and charge transfer overpotential and the energy used for maintenance and growth of the organisms. The Ohmic loss of the electrodes was not measured separately and is included in the overpotential of the anode and cathode [18]. The ionic loss of the electrolytes (V) is related to the electrolyte resistance of the anolyte and catholyte and was calculated with [19] Eionic ¼ Iions
1 1 dan dcat þ Ran þ Rcat ¼ Iions 2Asan 2Ascat 2 2
(3)
where Iions is the flow of ions through the electrolyte (A) which is equal to the current, dan is the distance between the anode and the membrane (0.004 m), dcat is the distance between the cathode and the membrane (0.001 m), A is the surface area (0.025 m2), san is the anode conductivity (S m1) and scat is the cathode conductivity (S m1). The factor ½ is added to estimate the location of the microorganisms and of the produced electrons within the anode compartment since the microorganisms are not all located at equal distance from the membrane. The membrane transport loss (ET) is the potential loss caused by transport of ions through the membrane and was calculated from all other potential losses using [16] Ecell ¼ Eeq han hcat Eionic ET
(4)
The partial internal resistance of all the different components was then calculated from these potential losses according to ohms law (R ¼ V/I ). The specific energy input (kW h m3 H2) in the system was calculated with Eelectricity ¼
nPF Ecell 1 RT 3:6$106 rcat
Cchem Q in F I
where Cchem is the chemical input (mol eq L1), I is the current density (A m2), Qin is the influent flow rate (L s1) and F is Faradays constant (C mol1 e).
3.
Results and discussion
3.1.
Steady-state performance
True steady-state situation in MECs is indicated not only by a constant produced current, but also by a constant cathodic pH and cathodic conductivity. Fig. 1 shows an example of MEC start-up towards steady-state performance in this case, equipped with a CEM and at an applied voltage of 1 V. For this specific experiment, the steady-state performance was reached after approximately 12 days, but for other experiments this time period varied depending on the applied voltage and produced current. The change in ion concentration is highly dependent on the current density produced by the system and steady-state is therefore reached faster at high current densities. Table 1 shows an overview of the steady-state performance of MECs equipped with a CEM and an AEM at different applied voltages. The first indication of the performance of an MEC is the produced current density. At low applied voltage (0.6 V) the current densities for both systems are similar. At higher applied voltages (1.0 V) the differences in performance become larger. This eventually led to a current density of 10.2 A m2 at an applied voltage of 1.0 V for the set-up with the AEM, compared to 7.2 A m2 for the set-up with the CEM. The main cause for the difference in performance between CEM and AEM was the difference in pH and conductivity of the catholyte. At high cathode pH, the thermodynamically required energy to form hydrogen increases. At the same time, low cathode conductivity increases the internal resistance of the system. Both processes therefore lead to decreased performance. At 0.6 V applied voltage, the pH and conductivity of the catholyte in the AEM configuration were lower than the pH and conductivity of the catholyte in the CEM configuration. At 0.8 V and 1.0 V applied voltages, the pH for the AEM
(5)
where P is the pressure (105 Pa) and T is the absolute temperature (303.15 K), 3.6$106 is a conversion factor for the conversion of Joules to kW h, and rcat is the cathodic hydrogen recovery [2]. The anode compartment was continuously fed and therefore changes in produced current had little influence on the pH and conductivity. The anode compartment had a constant pH of 6.6 and a conductivity of 5.6 mS cm1. These values were used in the calculations. The contribution of the relative chemical input (Crel) per mol of produced electrons and was calculated with Crel ¼
7203
(6)
Fig. 1 e Example of typical steady-state performance of an MEC equipped with a CEM at an applied voltage of 1 V. Steady-state conditions are indicated by a constant current density, cathodic conductivity and cathodic pH.
7204
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
Table 1 e Steady-state performance of MECs equipped with CEM and AEM at applied voltages of 0.6, 0.8 and 1.0 V. The energy input is calculated assuming a cathodic coulombic efficiency of 95%. Membrane CEM
AEM
Applied voltage (V)
Current density (A m2)
pH cathode (e)
Conductivity cathode (mS cm1)
Energy input (kW h m3 H2)
0.6 0.8 1.0 0.6 0.8 1.0
0.95 2.74 7.2 1.12 4.81 10.18
13.18 13.52 13.69 11.77 13.1 13.83
41 75 133 16 36 34
1.57 2.10 2.62 1.57 2.10 2.62
configuration increases to values comparable to those for the CEM configuration. The conductivity for the CEM configuration however, reached much higher values than the conductivity for the AEM configuration although this did not lead to higher current densities, since the contribution of the conductivity to the total internal resistance was relatively small. A high conductivity indicates a high ion concentration in the catholyte. In the case when a CEM is used these ions were mainly Naþ (7.8 g L1), Kþ (12.2 g L1) and OH (8.3 g L1) at an applied voltage of 1 V. Although the pH increase in the cathode for different types of membranes has been the topic of research before [6,20,21], their performance has only been studied in short term experiments.
3.2.
Partial internal resistance analysis
A more detailed analysis of the performance of the MEC can be provided by analysing the contribution of the internal resistance of the different components during steady-state performance of the MEC. Fig. 2 gives an overview of these internal resistances for both configurations at different applied voltages. Firstly, from this figure it becomes clear that the total internal resistance for the AEM configuration is lower than the internal resistance for the CEM configuration for all applied voltages. So, at all applied voltages the AEM configuration outperformed the CEM configuration, which is in agreement with previous studies [6,16]. Secondly, in the AEM configuration, increase in applied voltage leads to a more pronounced reduction in internal resistance compared to the CEM configuration.
When looking at the contribution of the different components of the configurations it can be seen that for both MECs, at all applied voltages, the equilibrium resistance for hydrogen production contributes for over 60% of the total internal resistance. The equilibrium voltage (resistance) represents the amount of Gibbs-energy that needs to be invested for the production of hydrogen. In principle, this energy can be regained in the form of chemical energy stored in the produced hydrogen. The actual required energy however, is higher than the theoretical value, since the pH in both the anode and cathode changes. At low applied voltages, the increase of the cathode pH is more pronounced in the CEM configuration than in the AEM configuration. Therefore, the required energy input for the AEM configuration is lower than for the CEM configuration. This is advantageous in terms of energy efficiency; although the hydrogen production rate will be lower, the required energy for hydrogen production is also lower. When designing a full scale system a balance has to be found between energy input and production rate. Both the anode and cathode resistance decrease with increasing current density for both membrane configurations. Generally, it is considered that the anode overpotential increases at higher applied voltages [9]. The anode resistance however, is the anode overpotential divided by the current density, and because the current density increases with higher applied voltages at higher rate than the overpotential, the anode resistance decreases at higher current densities. The resistance for transport of ions through the membrane was lower for the AEM configuration compared to the CEM configuration at all applied voltages, which is in agreement with previous findings [16]. Finally, the ionic resistance for this
Fig. 2 e Distribution of the internal resistances of MECs equipped with CEM and AEM at applied voltages of 0.6, 0.8 and 1.0 V. The ionic resistance is too small to be visible.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
specific reactor configuration was low (<10 mU) and hardly contributed to the total required energy input even at this low ionic strength (5 mS cm1). For both configurations, the pH increase in the cathode compartment causes the largest energy loss as shown by the equilibrium resistance. Therefore, the largest reduction of the required specific energy input of the system would be a decrease of this cathode pH. Use of an AEM decreases the cathode pH compared to the use of a CEM. Another proposed possibility to reduce the pH in the cathode is by addition of CO2 to the catholyte [22,23]. There have been studies that directly measured all partial internal resistances of MFCs using electrochemical impedance spectroscopy (EIS) [24,25]. Interpretation of EIS data however, highly depends on the model used and often several parameters are mixed in the measured spectra [26]. The internal resistance analysis proves a valuable alternative method to directly calculate the contribution of the partial resistances to the total internal resistance from measured data, which can give valuable information about the limitations of the system.
3.3.
Energy input and output
The energy efficiency of an MEC is determined by the energy output divided by the energy input. The energy input is determined as the electrical energy that is required to produce 1 m3 of hydrogen gas (kW h m3 H2), when the energy input from the waste product (in this case acetate) is not considered. The energy output is expressed as the chemical energy stored in this hydrogen gas. The energy input is highly dependent on the applied voltage and the cathodic efficiency. The applied voltage determines how much energy is added for every produced electron while the cathodic efficiency determines how many of these electrons end up as hydrogen gas. Therefore, at higher applied voltages the specific energy input, being the energy input per electron, is higher than at lower applied voltage at constant cathodic efficiency (Table 1). At higher applied voltages, generally, the rate at which electrons are produced is higher. For economical purposes, the goal is to produce a high current at low applied voltages (low internal resistance). Therefore, economically speaking, it is more interesting to use an AEM as physical separator between anode and cathode because of the lower internal resistance of an MEC equipped with an AEM.
3.4.
Chemical efficiency
In many lab-scale studies, considerable amounts of salts and buffer are added to improve the conductivity and to prevent biofilm acidification ([27]). Although these high concentrations of salts and buffer improve BES performance, their addition is undesirable from an economic and environmental point of view. Addition of extra chemicals not only increases the operational costs of the system, but also leads to an extra treatment step before the effluent of the BES can be discharged. For practical application of BESs, it is therefore interesting to study the effect of the addition of chemicals and buffer on the conversion rate. In general, a high conversion rate (of COD) at low chemical input is desirable. We therefore studied the
7205
contribution of electrolyte and buffer to the performance (current density) of the system; and for its evaluation we introduce a new concept: the relative chemical input, which indicates the moles chemical equivalents needed per mole of produced electrons. Three types of chemical inputs were considered: electrolyte, buffer, and substrate. First of all, the differences in MEC performance with an anion exchange membrane (AEM) and a cation exchange membrane (CEM) were considered. It has been shown in previous research and in this study that an MEC with an AEM outperforms an MEC with a CEM [6,16]. This also has a direct influence on the relative chemical input for the system. Fig. 3A shows the relative chemical input for both systems and shows that the total chemical input for the AEM is half of the chemical input of the CEM (data from [16]). This can be explained by the fact that both systems used influent with the same chemical composition while the AEM produced higher current density (5.3 A m2) than the CEM (2.3 A m2). Therefore, the relative contribution of these chemicals in the best performing system (AEM) was the lowest. This shows that when an AEM is applied, not only the performance is better, but because of this, the investment in the form of chemicals, expressed per electron, is lower compared to a CEM. Secondly, we studied the differences between systems with different buffer concentrations. Previous research showed that addition of buffer has a positive effect on the current production of MECs [13,27,28]. Also in this case, it is interesting to study the effect of chemical input on productivity. Fig. 3B shows the relative chemical input for identical systems using influent buffer concentrations of 10 mM and 50 mM (data from [13]). Although the system with the high buffer concentration (50 mM P) produced considerably higher current (10.2 A m2 vs. 7.8 A m2 at 10 mM), the chemical input for 10 mM buffer relative to the current produced was only one third of the chemical input for 50 mM buffer. Only when the current density would be 5 times as high as well, the chemical efficiency would stay the same. However, because the current density does not increase proportionally with the increase in buffer concentration, the contribution in mol equivalents of buffer per electron is much lower at high buffer concentrations. On the other hand, a lower buffer concentration results in a lower ionic strength and thus a higher ohmic resistance. Therefore, thirdly, we studied the differences between systems with different ionic strength. It is generally accepted that systems with low conductivity, which have high ohmic losses, perform worse than systems with high conductivity. Fig. 3C shows the relative chemical requirement for MECs with influent with 10 mM of phosphate buffer, 0 mM of phosphate buffer and with addition of electrolyte, not in the form of buffer (data from [29]). This figure shows that required input in the form of chemicals to produce a certain current (mole of electrons per time unit) decreases at low buffer concentrations. Again, the implication is that the system performs better at high buffer and electrolyte concentrations, but as the current density does not increase proportionally with the increase in chemicals, the effectiveness of chemical addition decreases when more chemicals are added. However, at low
7206
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
This study analysed the performance of MECs at steadystate conditions. Steady-state conditions are particularly interesting because these will be the actual operating conditions of future continuous BESs. It would therefore be most useful to see more steady-state studies in the future to make a realistic estimate of future applicability of BESs. Furthermore, it was shown that addition of large amounts of buffer and electrolyte is unnecessary from a chemical efficiency point of view since it only increases costs of the system.
4.
Fig. 3 e Relative chemical input for an MEC equipped with a CEM compared to a system equipped with an AEM (A); for an MEC operated with 10 and 50 mM of phosphate buffer (B) and for an MEC operated with 10 mM and 0 mM of phosphate buffer and without addition of extra salts in the anolyte (C).
Conclusions
Our objective in this study was twofold: (i) to study MEC performance at steady-state conditions in continuous mode, and (ii) to use a novel method to analyse MEC performance in terms of chemical efficiency. At steady-state operation, a current density of 10.2 A m2 at an applied voltage of 1.0 V for the set-up with the AEM was produced, compared to 7.2 A m2 for the set-up with the CEM. At steady-state conditions, the total internal resistance for the AEM configuration is lower than the internal resistance for the CEM configuration for all applied voltages. Also, in the AEM configuration, increase in applied voltage leads to a more pronounced reduction in internal resistance compared to the CEM configuration. Therefore, the required energy input for the AEM configuration is lower than for the CEM configuration. The pH gradient over the membrane causes the largest part of the internal resistance for both membrane configurations. At lower (<1.0 V) applied voltages, the relative contribution to the total internal resistance of the cathode pH is more pronounced in the CEM configuration than in the AEM configuration. In case a CEM is used, the conductivity in the cathode reaches very high values: (>130 mS cm1). This conductivity is mainly caused by the presence of Naþ (7.8 g L1), Kþ (12.2 g L1) and OH (8.3 g L1). Furthermore, MECs perform better at high buffer and electrolyte concentrations. However, as the current density does not increase proportionally with the increase in chemicals, the effectiveness of chemical addition decreases when more chemicals are added. We found that, when expressed per produced electron, addition of chemicals and buffer does not necessarily enhance performance but increases operational costs and is therefore not a cost-effective strategy towards practical application.
Acknowledgements ionic strength, it is important that good mass transfer is ensured. One way to improve electrode performance is the use of electrodes with a high specific surface like graphite felt, granules or brushes [13,30,31], which has also been shown to be very suitable to improve cathode performance [32]. However, inside these electrodes a sufficient mass transfer, e.g. via forced flow, is crucial [13,14]. So, to a certain extent, addition of (low concentrations of) chemicals and buffer is beneficial for the performance of BESs, but their addition is less effective at higher concentrations.
This work was performed in the TTIW-cooperation framework of Wetsus, centre of excellence for sustainable water technology (www.wetsus.nl). Wetsus is funded by the Dutch Ministry of Economic Affairs, the European Union European Regional Development Fund, the Province of Fryslan, the City of Leeuwarden and by the EZ-KOMPAS Program of the “Samenwerkingsverband Noord-Nederland”. The authors like to thank the participants of the research theme “Resource Recovery” for the fruitful discussions and their financial support.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
Nomenclature
Eeq Equilibrium voltage, V Cathode overpotential, V hcat Ionic resistance, V Eionic Anode overpotential, V han Membrane transport resistance, V Et Theoretical cathode potential, V Ecat Theoretical anode potential, V Ean standard cathode potential, V E0cat standard anode potential, V E0an R Universal gas constant, 8.3145 J mol1 K1 T Temperature, K F Faraday constant, 96,485 C mol1 Ean, measured Measured anode potential, V Ecat, measured Measured cathode potential, V Flow of ions through electrolyte, A Iions Distance between anode and membrane, 0.004 m dan Distance between cathode and membrane, 0.001 m dcat A Projected surface area, 0.025 m2 san Anode conductivity, S m1 scat Cathode conductivity, S m1 Eelectricity Specific energy input, kW h m3 H2 P Pressure, 105 Pa Cathodic hydrogen recovery, % rcat Relative chemical input, Mol eq (mol e1)1 Crel Cchem Chemical input, mol eq L1 I Current density, A m2 Qin Influent flow rate, L s1
references
[1] Rabaey K, Rozendal RA. Microbial electrosynthesis e revisiting the electrical route for microbial production. Nature Reviews Microbiology 2010;8:706e16. [2] Logan BE, Call D, Cheng S, Hamelers HVM, Sleutels THJA, Jeremiasse AW, et al. Microbial electrolysis cells (MECs) for high yield hydrogen gas production from organic matter. Environmental Science & Technology 2008;42:8630e40. [3] Sleutels THJA, Ter Heijne A, Buisman CJN, Hamelers HVM. Bioelectrochemical systems: an outlook for practical applications. ChemSusChem 2012;5:1012e9. [4] Cheng S, Xing D, Call DF, Logan BE. Direct biological conversion of electrical current into methane by electromethanogenesis. Environmental Science and Technology 2009;43:3953e8. [5] Hamelers HVM, Ter Heijne A, Sleutels THJA, Jeremiasse AW, Strik DPBTB, Buisman CJN. New applications and performance of bioelectrochemical systems. Applied Microbiology and Biotechnology 2010;85:1673e85. [6] Cheng S, Logan BE. Sustainable and efficient biohydrogen production via electrohydrogenesis. Proceedings of the National Academy of Sciences of the United States of America 2007;104:18871e3. [7] Rozendal RA, Hamelers HVM, Euverink GJW, Metz SJ, Buisman CJN. Principle and perspectives of hydrogen production through biocatalyzed electrolysis. International Journal of Hydrogen Energy 2006;31:1632e40.
7207
[8] Van Eerten-Jansen MCAA, Heijne AT, Buisman CJN, Hamelers HVM. Microbial electrolysis cells for production of methane from CO2: long-term performance and perspectives. International Journal of Energy Research 2012;36:809e19. [9] Jeremiasse AW, Hamelers HVM, Kleijn JM, Buisman CJN. Use of biocompatible buffers to reduce the concentration overpotential for hydrogen evolution. Environmental Science and Technology 2009;43:6882e7. [10] Jeremiasse AW, Hamelers HVM, Buisman CJN. Microbial electrolysis cell with a microbial biocathode. Bioelectrochemistry 2010;78:39e43. [11] Zhuang L, Yuan Y, Wang Y, Zhou S. Long-term evaluation of a 10-liter serpentine-type microbial fuel cell stack treating brewery wastewater. Bioresource Technology 2012;123:406e12. [12] Escapa A, Gil-Carrera L, Garcı´a V, Mora´n A. Performance of a continuous flow microbial electrolysis cell (MEC) fed with domestic wastewater. Bioresource Technology 2012;117:55e62. [13] Sleutels THJA, Lodder R, Hamelers HVM, Buisman CJN. Improved performance of porous bio-anodes in microbial electrolysis cells by enhancing mass and charge transport. International Journal of Hydrogen Energy 2009;34:9655e61. [14] Sleutels THJA, Hamelers HVM, Buisman CJN. Effect of mass transfer speed and direction in porous anodes on microbial electrolysis cell performance. Bioresource Technology 2011;102:399e403. [15] Zehnder AJB, Huser BA, Brock TD, Wuhrmann K. Characterization of an acetate-decarboxylating, nonhydrogen-oxidizing methane bacterium. Archives Of Microbiology 1980;124:1e11. [16] Sleutels THJA, Hamelers HVM, Rozendal RA, Buisman CJN. Ion transport resistance in microbial electrolysis cells with anion and cation exchange membranes. International Journal of Hydrogen Energy 2009;34:3612e20. [17] Fan Y, Sharbrough E, Liu H. Quantification of the internal resistance distribution of microbial fuel cells. Environmental Science and Technology 2008;42:8101e7. [18] Rozendal RA, Sleutels THJA, Hamelers HVM, Buisman CJN. Effect of the type of ion exchange membrane on performance, ion transport, and pH in biocatalyzed electrolysis of wastewater. Water Science and Technology 2008;57:1757e62. [19] Ter Heijne A, Hamelers HVM, De Wilde V, Rozendal RA, Buisman CJN. A bipolar membrane combined with ferric iron reduction as an efficient cathode system in microbial fuel cells. Environmental Science & Technology 2006;40:5200e5. [20] Rozendal RA, Hamelers HVM, Molenkamp RJ, Buisman CJN. Performance of single chamber biocatalyzed electrolysis with different types of ion exchange membranes. Water Research 2007;41:1984e94. [21] Kim JR, Cheng S, Oh SE, Logan BE. Power generation using different cation, anion, and ultrafiltration membranes in microbial fuel cells. Environmental Science & Technology 2007;41:1004e9. [22] Torres CI, Lee HS, Rittmann BE. Carbonate species as OH carriers for decreasing the pH gradient between cathode and anode in biological fuel cells. Environmental Science and Technology 2008;42:8773e7. [23] Fornero JJ, Rosenbaum M, Cotta MA, Angenent LT. Carbon dioxide addition to microbial fuel cell cathodes maintains sustainable catholyte pH and improves anolyte pH, alkalinity, and conductivity. Environmental Science and Technology 2010;44:2728e34. [24] Sleutels THJA, Darus L, Hamelers HVM, Buisman CJN. Effect of operational parameters on coulombic efficiency in
7208
[25]
[26]
[27]
[28]
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 7 2 0 1 e7 2 0 8
bioelectrochemical systems. Bioresource Technology 2011;102:11172e6. Deeke A, Sleutels THJA, Hamelers HVM, Buisman CJN. Capacitive bioanodes enable renewable energy storage in microbial fuel cells. Environmental Science and Technology 2012;46:3554e60. Jeremiasse AW, Hamelers HV, Croese E, Buisman CJ. Acetate enhances startup of a H2-producing microbial biocathode. Biotechnology and Bioengineering 2012;109:657e64. Torres CI, Marcus AK, Rittmann BE. Proton transport inside the biofilm limits electrical current generation by anoderespiring bacteria. Biotechnology and Bioengineering 2008;100:872e81. Liu H, Cheng S, Logan BE. Power generation in fed-batch microbial fuel cells as a function of ionic strength,
[29]
[30]
[31]
[32]
temperature, and reactor configuration. Environmental Science and Technology 2005;39:5488e93. Sleutels THJA, Hamelers HVM, Buisman CJN. Reduction of pH buffer requirement in bioelectrochemical Systems. Environmental Science & Technology 2010;44:8259e63. Logan B, Cheng S, Watson V, Estadt G. Graphite fiber brush anodes for increased power production in air-cathode microbial fuel cells. Environmental Science & Technology 2007;41:3341e6. Rabaey K, Ossieur W, Verhaege M, Verstraete W. Continuous microbial fuel cells convert carbohydrates to electricity. Water Science and Technology 2005:515e23. Jeremiasse AW, Hamelers HVM, Saakes M, Buisman CJN. Ni foam cathode enables high volumetric H2 production in a microbial electrolysis cell. International Journal of Hydrogen Energy 2010;35:12716e23.