17-04-2012, 04:24 PM
Characterization of energy losses in an upflow single-chamber microbial electrolysis cell
[attachment=20251]
a b s t r a c t
We characterized electrode energy losses and ohmic energy loss in an upflow, singlechamber
microbial electrolysis cell (MEC) with no metal catalyst on the cathode. The MEC
produced 0.57m3-H2/m3-d at an applied voltage of w1 V and achieved a cathodic conversion
efficiency of 98% and a H2 yield of 2.4 mol H2/mol acetate. Eliminating the membrane
lowered the ohmic energy loss to 0.005 V, and the pH energy loss became as small as
0.072 V. The lack of metal catalyst on the cathode led to a significant cathode energy loss of
0.56 V. The anode energy loss also was relatively large at 0.395 V, but this was artificial, due
to the high positive anode potential, poised at þ0.07 V (vs. the standard hydrogen electrode).
The energy-conversion efficiency (ECE) was 75% in the single-chamber MEC when
the energy input and outputs were compared directly as electrical energy. To achieve an
energy benefit out of an MEC (i.e., an ECE >100%), the applied voltage must be less than
0.6 V with a cathodic conversion efficiency over 80%. An ECE of 180% could be achieved if
the anode and cathode energy losses were reduced to 0.2 V each.
ª 2009 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.
1. Introduction
Although H2 gas has many merits as an energy-carrier, it will
not become a viable renewable option until it is produced from
a non-fossil-fuel source and the cost of its production and
delivery decreases substantially [1]. A microbial electrolysis
cell (MEC) is a biomass-based approach that has the potential
to meet these future energy requirements. It has advantages
over dark-fermentative H2 production due to its high H2 yields
(w9 mol H2/mol glucose versus w2 mol H2/mol glucose for
dark fermentation) [2–5]. One drawback of the MEC, however,
is that it requires electrical energy input generated from fossil
fuels for H2 generation (i.e., an applied voltage), which can
increase the H2-production cost and lower the net energy
output. A second drawback is that most MECs today include
expensive metal catalysts on the cathode, typically platinum
(Pt) [6–12]. The average cost of Pt is $38 per gram of Pt in 2009
[13]. For the MEC to become a practical H2 producer, high H2-
production rates and yields must be attained with low applied
voltage and without a high cost of precious catalysts. The first
two features also are necessary for the MEC to produce a net
energy benefit.
The applied voltage is one of the most significant factors
controlling energy efficiency (i.e., the H2-production cost) in
an MEC. While the applied voltage does not affect the H2 yield
directly, a large applied voltage lowers the net energy value of
the generated H2. The applied voltage depends on the energy
losses generated by MEC operation. Energy loss is the difference
between the equilibrium electrical potential with no net
current and the potential with a current. The energy losses
increase with increasing current density, which normally is
proportional to H2-production rate in the MEC. Previous MEC
studies reported a wide range of applied voltage (0.3–1.3 V)
when utilizing acetate as the electron donor [6–12,14]; from
* Corresponding author. Tel.: þ1 480 727 0849; fax: þ1 480 727 0889.
E-mail addresses: hyungsool[at]asu.edu (H.-S. Lee), rittmann[at]asu.edu (B.E. Rittmann).
Available at www.sciencedirect.com
journal homepage: www.elsevierlocate/he
i n t e r n a t i onal j o u r n a l o f hydrogen energy 3 5 ( 2 0 1 0 ) 9 2 0 – 9 2 7
0360-3199/$ – see front matter ª 2009 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijhydene.2009.11.040
this, the energy loss can be computed as being in the range of
0.16–1.16 V, based on 0.14 V as the standard potential for the
overall reaction CH3COO þ 3H2O ¼ CO2 þ HCO3
þ 4H2 at pH 7.
The wide range of applied voltages and energy losses occurs
because each study had different current density, biofilmanode
composition and thickness, donor concentration, pH,
electrode material, electrode distance, and membrane type.
Traditional energy losses in chemical fuel cells and electrolysis
cells are divided into ohmic, activation, and concentration
losses [15,16]. Ohmic loss is caused by electrical
resistance to current in conductors (electrodes þ wire) and ion
transfers in the electrolyte (membrane þ medium). The activation
energy loss is the energy required for overcoming
energy barriers across the electrode/electrolyte interface to
generate net current, and it is characterized by the Butler–
Volmer equation [16]. Concentration energy loss is due to
concentration gradients between the bulk liquid and the
electrode surface, which become significant at high current
density in chemical fuel cells [15,16]. Because current density
in a microbial fuel cell (MFC) or an MEC is orders of magnitude
smaller than in chemical fuel cells/electrolysis cell, it may
seem reasonable to assume that concentration losses in the
MFC/MEC are negligible. However, concentration losses can
become significant in the MFC/MEC, because concentration
gradients develop for substrate in any type of biofilm system
[17–19], including the biofilm anode in the MFC/MEC. Only if
the biofilm thickness is thin enough can we assume no
concentration gradients between bulk and electrode surface,
which is called a fully penetrated biofilm [17,19]; activation
and concentration energy losses are sometimes considered
together as electrode energy losses [6].
Even if the anode’s biofilm has no gradients in the
concentrations of the donor substrate and protons, a dualchamber
MFC/MEC using a membrane to separate the anode
from the cathode can present a unique concentration loss due
to [Hþ] or [OH] accumulation in a chamber, since they are net
produced at half reactions on the electrodes [20–22]. The high
concentrations of other ions in the liquid supplied to an MEC/
MFC (e.g., Naþ), compared to [Hþ] or [OH], means that charge
neutrality can be achieved with little transport of Hþ or OH
ions through membrane [6,20], and a strong pH gradient can
develop across the membrane, causing a substantial concentration
energy loss. Rozendal et al. [6] reported that the pH
difference between the two chambers increased the concentration
energy loss up to 0.38 V in a dual-chamber MEC. They
also showed that this was the largest part of the total energy
loss.
Of many possibilities to decrease energy losses, removing
the membrane, which creates a single-chamber MEC, can be
very efficient; however, the lack of membrane accentuates the
need for rapid and efficient hydrogen recovery to counteract
hydrogen scavenging by methanogens [14]. One potential
benefit of a single-chamber MEC is that the concentration
energy loss due to Hþ or OH accumulation should be negligible,
because Hþ produced in the anodic reaction is neutralized
directly by OH produced in a cathodic reaction
alternately or by reacting with electrons at the cathode to
form H2 molecules. The other potential benefit of a singlechamber
MEC can be a substantial reduction in ohmic energy
loss, since the resistance to ion flow through the membrane
can be the main ohmic energy loss [6–10,14,23]. Thus, a singlechamber
MEC can provide a high H2-production rate with
smaller applied voltage as a consequence of energy loss
attenuation. Previous studies using a single-chamber MEC
reported 3.12m3 H2/m3 d (292 A/m3) at an applied voltage of
0.8 V [7], 1.7m3 H2/m3-d (188 A/m3) at an applied voltage of
0.6 V [24], and 0.65m3 H2/m3 d (39 A/m3) at an applied voltage
of 0.6 V [10]. The authors claimed that these high H2-production
rates were achieved with relatively low applied voltage,
compared to dual-chamber studies. However, a study with
a dual-chamber MEC [2] also achieved H2-production rates
and applied voltages similar to the single-chamber MECs.
Thus, the validity of energy loss mitigation in single-chamber
MECs requires more rigorous study.
The energy-conversion efficiency (ECE) is an essential
criterion for MEC sustainability, but the definition of ECE is not
agreed upon. Previous works defined ECE in MECs as the heat
of combustion of captured H2 divided by the input electrical
energy [2,7,10,24]. ECEs were well over 100% (194–351%) using
this definition. They also have added the energy value of the
input substrate to the denominator. With this definition, ECE
declined to 58–86% [2,7,24].
Another approach to computing ECE is to compare the
input and output energy in the same form. The most logical
way to make the energy inputs and outputs consistent is to
use electrical energy for both. Since generation of electrical
energy from output H2 incurs losses, the numerator in this
approach is smaller than that used with the previous
approaches that led to ECE values greater than 100%. For
example, typical efficiency of energy transformation from H2
heat energy to electricity is w55% in hydrogen fuel cells [15],
while the efficiency isw33% if the H2 is combusted to produce
electricity [25].
We performed this work to provide rigorous characterizations
of energy loss and ECE in an upflow single-chamber MEC.
In our previous study [14], we found that the upflow singlechamber
MEC with the cathode placed on top of the MEC
improved the cathodic conversion efficiency (CCE) approximately
two-fold over a conventional MEC having the cathode
alongside the anodes: a CCE of 98 2% at the same time as the
Coulombic efficiency was 60 1%, and negligible CH4 was
generated. Despite having a metal-catalyst-free cathode, the
upflow MEC produced 0.57 0.02m3 H2/m3 d at an applied
voltage of w1 V. For this study, we first estimate the concentration
energy loss from the maximum pH gradient that
developed in the liquid contents of the MEC chamber. Second,
we experimentally measured the cathode/the anode energy
losses, and ohmic energy loss. We then determine which
energy losses were mainly responsible for the applied voltage
and if the concentration energy loss and ohmic energy loss
were mitigated in the single-chamber MEC. Third, we define
and compute the ECE in the upflow MEC, and this allows us to
identify the factors most critical for obtaining the maximum
energy benefit from an MEC.
2. Background on electrode energy loss
Electrode energy loss is the potential difference between the
theoretical chemical potential for a reduction half-reaction of
i n t e r n a t i o n a l journa l o f hydrogen energy 3 5 ( 2 0 1 0 ) 9 2 0 – 9 2 7 921
donor or acceptor substrate and the measured potential of the
electrode when it is generating a certain current; the theoretical
potential can be estimated with the Nernst equation
[16]. Because the electrode energy loss includes the activation
and concentration energy losses, it is a means to lower the
complexity of energy loss analysis, especially for the anode
having an ARB biofilm.
Fig. 1 shows the sigmoidal shape of an anode-polarization
curve for a biofilm anode in an MEC using acetate as substrate.
The standard chemical potential for the acetate half reaction
at pH 7 is 0.28 V, and this Eanode is close to the value for no
current. Eanode becomes more positive when current is
generated: e.g., 0.22 V at current density 2 A/m2 and 0.13 V
at current density 6 A/m2. Anode energy losses for each
current density are at 0.06 V and 0.15 V, respectively, based on
the standard potential 0.28 V. Finally, Eanode reaches 0.02 V
as the current density approaches saturation current density
(8.1 A/m2, or 95% of maximum current density), and the anode
energy loss is 0.26 V. For any Eanode greater than 0.02 V, the
anode energy loss and applied voltage increase, but without
any further increase in current density.
We can produce cathode-polarization curves in the same
manner. One difference is that abiotic catalysts (e.g., Pt) normally
are used for the cathode, instead of ARB. Catalysts
significantly improved cathode energy loss in water- electrolysis
cells, which can achieve a current density of a few
hundred A/m2 [26], which is orders of magnitude larger than
for MECs. However, their catalytic effect on the energy loss is
relatively smaller for the current densities that MECs usually
produce, w a few A/m2. This suggests the possibility that
using no metal catalyst on the cathode might work for MECs
producing moderate ranges of current density.
[attachment=20251]
a b s t r a c t
We characterized electrode energy losses and ohmic energy loss in an upflow, singlechamber
microbial electrolysis cell (MEC) with no metal catalyst on the cathode. The MEC
produced 0.57m3-H2/m3-d at an applied voltage of w1 V and achieved a cathodic conversion
efficiency of 98% and a H2 yield of 2.4 mol H2/mol acetate. Eliminating the membrane
lowered the ohmic energy loss to 0.005 V, and the pH energy loss became as small as
0.072 V. The lack of metal catalyst on the cathode led to a significant cathode energy loss of
0.56 V. The anode energy loss also was relatively large at 0.395 V, but this was artificial, due
to the high positive anode potential, poised at þ0.07 V (vs. the standard hydrogen electrode).
The energy-conversion efficiency (ECE) was 75% in the single-chamber MEC when
the energy input and outputs were compared directly as electrical energy. To achieve an
energy benefit out of an MEC (i.e., an ECE >100%), the applied voltage must be less than
0.6 V with a cathodic conversion efficiency over 80%. An ECE of 180% could be achieved if
the anode and cathode energy losses were reduced to 0.2 V each.
ª 2009 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.
1. Introduction
Although H2 gas has many merits as an energy-carrier, it will
not become a viable renewable option until it is produced from
a non-fossil-fuel source and the cost of its production and
delivery decreases substantially [1]. A microbial electrolysis
cell (MEC) is a biomass-based approach that has the potential
to meet these future energy requirements. It has advantages
over dark-fermentative H2 production due to its high H2 yields
(w9 mol H2/mol glucose versus w2 mol H2/mol glucose for
dark fermentation) [2–5]. One drawback of the MEC, however,
is that it requires electrical energy input generated from fossil
fuels for H2 generation (i.e., an applied voltage), which can
increase the H2-production cost and lower the net energy
output. A second drawback is that most MECs today include
expensive metal catalysts on the cathode, typically platinum
(Pt) [6–12]. The average cost of Pt is $38 per gram of Pt in 2009
[13]. For the MEC to become a practical H2 producer, high H2-
production rates and yields must be attained with low applied
voltage and without a high cost of precious catalysts. The first
two features also are necessary for the MEC to produce a net
energy benefit.
The applied voltage is one of the most significant factors
controlling energy efficiency (i.e., the H2-production cost) in
an MEC. While the applied voltage does not affect the H2 yield
directly, a large applied voltage lowers the net energy value of
the generated H2. The applied voltage depends on the energy
losses generated by MEC operation. Energy loss is the difference
between the equilibrium electrical potential with no net
current and the potential with a current. The energy losses
increase with increasing current density, which normally is
proportional to H2-production rate in the MEC. Previous MEC
studies reported a wide range of applied voltage (0.3–1.3 V)
when utilizing acetate as the electron donor [6–12,14]; from
* Corresponding author. Tel.: þ1 480 727 0849; fax: þ1 480 727 0889.
E-mail addresses: hyungsool[at]asu.edu (H.-S. Lee), rittmann[at]asu.edu (B.E. Rittmann).
Available at www.sciencedirect.com
journal homepage: www.elsevierlocate/he
i n t e r n a t i onal j o u r n a l o f hydrogen energy 3 5 ( 2 0 1 0 ) 9 2 0 – 9 2 7
0360-3199/$ – see front matter ª 2009 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijhydene.2009.11.040
this, the energy loss can be computed as being in the range of
0.16–1.16 V, based on 0.14 V as the standard potential for the
overall reaction CH3COO þ 3H2O ¼ CO2 þ HCO3
þ 4H2 at pH 7.
The wide range of applied voltages and energy losses occurs
because each study had different current density, biofilmanode
composition and thickness, donor concentration, pH,
electrode material, electrode distance, and membrane type.
Traditional energy losses in chemical fuel cells and electrolysis
cells are divided into ohmic, activation, and concentration
losses [15,16]. Ohmic loss is caused by electrical
resistance to current in conductors (electrodes þ wire) and ion
transfers in the electrolyte (membrane þ medium). The activation
energy loss is the energy required for overcoming
energy barriers across the electrode/electrolyte interface to
generate net current, and it is characterized by the Butler–
Volmer equation [16]. Concentration energy loss is due to
concentration gradients between the bulk liquid and the
electrode surface, which become significant at high current
density in chemical fuel cells [15,16]. Because current density
in a microbial fuel cell (MFC) or an MEC is orders of magnitude
smaller than in chemical fuel cells/electrolysis cell, it may
seem reasonable to assume that concentration losses in the
MFC/MEC are negligible. However, concentration losses can
become significant in the MFC/MEC, because concentration
gradients develop for substrate in any type of biofilm system
[17–19], including the biofilm anode in the MFC/MEC. Only if
the biofilm thickness is thin enough can we assume no
concentration gradients between bulk and electrode surface,
which is called a fully penetrated biofilm [17,19]; activation
and concentration energy losses are sometimes considered
together as electrode energy losses [6].
Even if the anode’s biofilm has no gradients in the
concentrations of the donor substrate and protons, a dualchamber
MFC/MEC using a membrane to separate the anode
from the cathode can present a unique concentration loss due
to [Hþ] or [OH] accumulation in a chamber, since they are net
produced at half reactions on the electrodes [20–22]. The high
concentrations of other ions in the liquid supplied to an MEC/
MFC (e.g., Naþ), compared to [Hþ] or [OH], means that charge
neutrality can be achieved with little transport of Hþ or OH
ions through membrane [6,20], and a strong pH gradient can
develop across the membrane, causing a substantial concentration
energy loss. Rozendal et al. [6] reported that the pH
difference between the two chambers increased the concentration
energy loss up to 0.38 V in a dual-chamber MEC. They
also showed that this was the largest part of the total energy
loss.
Of many possibilities to decrease energy losses, removing
the membrane, which creates a single-chamber MEC, can be
very efficient; however, the lack of membrane accentuates the
need for rapid and efficient hydrogen recovery to counteract
hydrogen scavenging by methanogens [14]. One potential
benefit of a single-chamber MEC is that the concentration
energy loss due to Hþ or OH accumulation should be negligible,
because Hþ produced in the anodic reaction is neutralized
directly by OH produced in a cathodic reaction
alternately or by reacting with electrons at the cathode to
form H2 molecules. The other potential benefit of a singlechamber
MEC can be a substantial reduction in ohmic energy
loss, since the resistance to ion flow through the membrane
can be the main ohmic energy loss [6–10,14,23]. Thus, a singlechamber
MEC can provide a high H2-production rate with
smaller applied voltage as a consequence of energy loss
attenuation. Previous studies using a single-chamber MEC
reported 3.12m3 H2/m3 d (292 A/m3) at an applied voltage of
0.8 V [7], 1.7m3 H2/m3-d (188 A/m3) at an applied voltage of
0.6 V [24], and 0.65m3 H2/m3 d (39 A/m3) at an applied voltage
of 0.6 V [10]. The authors claimed that these high H2-production
rates were achieved with relatively low applied voltage,
compared to dual-chamber studies. However, a study with
a dual-chamber MEC [2] also achieved H2-production rates
and applied voltages similar to the single-chamber MECs.
Thus, the validity of energy loss mitigation in single-chamber
MECs requires more rigorous study.
The energy-conversion efficiency (ECE) is an essential
criterion for MEC sustainability, but the definition of ECE is not
agreed upon. Previous works defined ECE in MECs as the heat
of combustion of captured H2 divided by the input electrical
energy [2,7,10,24]. ECEs were well over 100% (194–351%) using
this definition. They also have added the energy value of the
input substrate to the denominator. With this definition, ECE
declined to 58–86% [2,7,24].
Another approach to computing ECE is to compare the
input and output energy in the same form. The most logical
way to make the energy inputs and outputs consistent is to
use electrical energy for both. Since generation of electrical
energy from output H2 incurs losses, the numerator in this
approach is smaller than that used with the previous
approaches that led to ECE values greater than 100%. For
example, typical efficiency of energy transformation from H2
heat energy to electricity is w55% in hydrogen fuel cells [15],
while the efficiency isw33% if the H2 is combusted to produce
electricity [25].
We performed this work to provide rigorous characterizations
of energy loss and ECE in an upflow single-chamber MEC.
In our previous study [14], we found that the upflow singlechamber
MEC with the cathode placed on top of the MEC
improved the cathodic conversion efficiency (CCE) approximately
two-fold over a conventional MEC having the cathode
alongside the anodes: a CCE of 98 2% at the same time as the
Coulombic efficiency was 60 1%, and negligible CH4 was
generated. Despite having a metal-catalyst-free cathode, the
upflow MEC produced 0.57 0.02m3 H2/m3 d at an applied
voltage of w1 V. For this study, we first estimate the concentration
energy loss from the maximum pH gradient that
developed in the liquid contents of the MEC chamber. Second,
we experimentally measured the cathode/the anode energy
losses, and ohmic energy loss. We then determine which
energy losses were mainly responsible for the applied voltage
and if the concentration energy loss and ohmic energy loss
were mitigated in the single-chamber MEC. Third, we define
and compute the ECE in the upflow MEC, and this allows us to
identify the factors most critical for obtaining the maximum
energy benefit from an MEC.
2. Background on electrode energy loss
Electrode energy loss is the potential difference between the
theoretical chemical potential for a reduction half-reaction of
i n t e r n a t i o n a l journa l o f hydrogen energy 3 5 ( 2 0 1 0 ) 9 2 0 – 9 2 7 921
donor or acceptor substrate and the measured potential of the
electrode when it is generating a certain current; the theoretical
potential can be estimated with the Nernst equation
[16]. Because the electrode energy loss includes the activation
and concentration energy losses, it is a means to lower the
complexity of energy loss analysis, especially for the anode
having an ARB biofilm.
Fig. 1 shows the sigmoidal shape of an anode-polarization
curve for a biofilm anode in an MEC using acetate as substrate.
The standard chemical potential for the acetate half reaction
at pH 7 is 0.28 V, and this Eanode is close to the value for no
current. Eanode becomes more positive when current is
generated: e.g., 0.22 V at current density 2 A/m2 and 0.13 V
at current density 6 A/m2. Anode energy losses for each
current density are at 0.06 V and 0.15 V, respectively, based on
the standard potential 0.28 V. Finally, Eanode reaches 0.02 V
as the current density approaches saturation current density
(8.1 A/m2, or 95% of maximum current density), and the anode
energy loss is 0.26 V. For any Eanode greater than 0.02 V, the
anode energy loss and applied voltage increase, but without
any further increase in current density.
We can produce cathode-polarization curves in the same
manner. One difference is that abiotic catalysts (e.g., Pt) normally
are used for the cathode, instead of ARB. Catalysts
significantly improved cathode energy loss in water- electrolysis
cells, which can achieve a current density of a few
hundred A/m2 [26], which is orders of magnitude larger than
for MECs. However, their catalytic effect on the energy loss is
relatively smaller for the current densities that MECs usually
produce, w a few A/m2. This suggests the possibility that
using no metal catalyst on the cathode might work for MECs
producing moderate ranges of current density.