E3S Web of Conferences 67, 02008 (2018)
3rd i-TREC 2018
https://doi.org/10.1051/e3sconf/20186702008
Improvement of biohydrogen production in microbial
electrolysis cell (MEC) system by tackling methanogenesis
Matthew Hardhi1, Putty Ekadewi1, Rita Arbianti1, Tania Surya Utami1*, and Heri Hermansyah1
1
Department of Chemical Engineering, Faculty of Engineering Universitas Indonesia, Depok 16424, Indonesia
Abstract. The increasingly adverse effects of climate change caused by a variety of fossil-based
fuel demands an alternative to such fuel. Hydrogen is one of the potential renewable fuel that
offers numerous advantages compared to its competitors. However, the dominant hydrogen
production methods are still energy-heavy and dependent on fossil-based resources. Microbial
electrolysis cell or MEC system is one of the leading solution towards replacing conventional
hydrogen production method. A persistent downside to this system in the presence of
methanogens that consumes the hydrogen product. This research proposes alternative biological
method to control the methanogen colony by introducing isolates of denitrifying bacteria to the
system which will act as inhibitor to hydrogenotrophic methanogen. The reactor implemented is a
single-chambered, membrane-less 20-ml reactor. Net hydrogen yield produced in the cathodic
headspace will be analyzed by gas chromatography (GC). Hydrogen yield for reactor with
enriched cathode is expected to be higher in comparison to unenriched reactor, as nitrogen
oxides produced during the metabolism of the denitrifiers were known to inhibit methanogen
growth. Experimental results showed consistent higher H2 yield in inoculated reactor compared to
control reactor, where in the second cycle H2 production increased 100% compared to the control.
1 Introduction
Fossil fuel usage brought more harm than good.
Negative environmental consequences such as climate
change and the rising global temperature level are just
some to name. Clean, renewable energy must take the
stage as the world’s leading energy source in the near
future to minimize the irreversible impacts caused by
centuries-long pollution to our atmosphere.
Hydrogen gas or H2 for short, is one of the most
promising leading candidates of renewable energy of the
future. It emits almost zero carbon dioxide [1] and
nitrous oxide [2] to the atmosphere, has high calorific
value at 120-140 MJ/kg compared to petrol at 44-46
MJ/Kg, and it is not perceived negatively in the politic
and economic sector [3].
Microbial Electrolysis Cell (MEC for short) system is
one of the alternative method for H2 production that has
been gaining traction due to its relatively cheap and easy
to obtain substrate (e.g. wastewater) that has multiple
possible applications, ranging from energy production
(H2 and CH4) to wastewater treatment [4] [5].
A common problem that hinders potential application
of MEC usage to produce H2 is the presence of
methanogenic bacteria in the system that produces CH4
out of the produced H2 [5] [6]. This reduces the possible
yield of H2, and this is very apparent in a membrane-less
reactor [6]. Although this problem can usually be
minimalized by using an ion exchange membrane to
*
separate the cathodic and anodic chamber, such usage of
membrane hinder scale-up application as the membrane
consists a huge portion of MEC manufacturing cost [7]
and must be routinely replaced, making MEC technology
expensive.
One of the method that has yet to be tried in order to
tackle the presence of methanogens is by biological
competition. Denitrifier is a class of bacteria that are
capable of reducing nitrates present in wastewater [8] to
a reduced form (e.g. nitrite, nitrous oxide) that will
ultimately result in N2 gas, in which the process is
accelerated in an anaerobic environment [9]. Research
showed that in a contained system, metabolites produced
by denitrifier is known to inhibit methanogenic activity
[10] [11] [12].
Our team aimed to validate whether or not the
presence of denitrifiers in a MEC system will assist with
H2 production by inhibiting the methanogens. This
method is relatively cheap and environmentally-friendly
as denitrifier is well-known to be employed in
bioremediation efforts [8]. The chosen denitrifier is the
Pseudomonas stutzeri because aside from the relatively
easy access to the culture, it is known to be able to
perform complete denitrification pathway [13] and is
able to grow in an anaerobic condition.
The MEC system that we used is made of 20-ml
glass vial with a working volume of 15 ml. The reactor
will be membrane-less so as to reduce cost [14] for
potential scale-up. The wastewater sludge is obtained
Corresponding author: nana@che.ui.ac.id
© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution
License 4.0 (http://creativecommons.org/licenses/by/4.0/).
E3S Web of Conferences 67, 02008 (2018)
3rd i-TREC 2018
https://doi.org/10.1051/e3sconf/20186702008
from a local wastewater reservoir located in Jakarta.
Denitrifier culture of P. stutzeri is obtained from InaCC
LIPI.
cycles with Ultra-High Purity N2 gas to purge oxygen
out of the system, creating an anaerobic environment.
Flushing and vacuuming is carried out by means of tube
with syringe attached to the end, connected to each
respective sources. For flushing then the tube is attached
to a N2 source and for vacuuming it is attached to a
vacuum pump. Each time flushing is about to be carried,
the syringe will pierce the rubber septum to minimize
gas exchange between reactors and the outside
environment.
The flushing cycles are as follows: firstly vial is
pumped with N2 for 10 s, then the vacuum line is added
to simultaneously pump N2 and vacuum the air for 20 s,
lastly the N2 line is removed and the vial is vacuumed
empty again for 10 s. Whenever N2 is flushed, the vial
will be tilted so as to submerge the needle tip with the
liquid inside in order to flush aqueous oxygen out of the
liquid.
After three cycles of flushing and vacuuming, the
reactor will then be filled again with N2 for 5-7 s to build
positive pressure inside the system to prevent outside air
from being sucked inside the reactor. All reactors will
then be autoclaved to sterilize them. The picture below
illustrates the flushing and vacuuming process. Note the
formed bubbles that indicates gas exchange between
liquid and gas phase.
2 Methods
2.1 MEC Reactor Design
We based our reactor design on a study by Call & Logan
(2011) [15] in which a small-scale, affordable reactor
was assembled using commercially-available materials.
Based on the picture below, we used a single-chambered,
membraneless reactor made of 20 ml clear glass vials
(Agilent) crimped with appropriate rubber septum (E)
used to prevent gas exchange. Anodes (A) used were
isomolded graphite plates with a thickness of 0.125
inches (Grade GM-10; GraphiteStore.com Inc.) cut
manually to an average dimension of 1.5-cm x 1.0-cm (L
x W). Cathodes (B) used were stainless steel mesh cut to
the same dimension of its respective anode pair. In order
to establish electrical connection with the electrodes,
titanium wire (0.08-cm diameter, Grade 2; Ti-Shop.com)
(C) and stainless steel wire (0.08-cm diameter) (D) are
used for the anode and cathode respectively. All
electrodes and wires and polished beforehand with
sandpaper, sonicated for 20-min, and rinsed three times
by Type-1 water. Only for anode, after sonication it is
soaked overnight in 1 N HCl solution.
In order to connect the wire with the electrodes, a
hand drill (0.08-cm diameter drill tip) is used to bore 2
holes on the electrodes. The wire is then inserted into the
first hole, then bent at the end to insert it at the second
hole. The wire is then crimped to the electrode by a plier
in order to create a tight connection. Internal resistances
for the electrodes will then be tested with a multimeter.
Only electrodes with resistances below 2 Ω will be used
in the experiment. The wire side of the electrode will
then be inserted through the rubber septum, and then
crimped with the glass vial to create a finished reactor.
The electrode spacing is kept at a minimum distance of
0.5-cm. The total working volume is 15 ml, providing
headspace for gas to accumulate.
2.3 MEC Reactor Operation
A programmable power supply is used to supply energy
to the system in order to drive the reaction that produces
H2 in the cathode, the required potential being 0.7 V Call
& Logan (2011) [15]. The output is made parallel by
using two sets of 4 test leads soldered together at one
end, then attached to the positive and negative terminal
of the power supply. This is done in order to supply
power simultaneously to more than one reactors. The
supplied current is set at 1 A.
Each test leads will then be connected to the reactors
using alligator clips, with the anode connected to the
positive terminal and the cathode connected to the
negative terminal. Between the power supply and the
anode, a 10 Ω resistor was used in order to record the
voltage of the reactor. A multimeter (model 109N;
APPA Technology Corp.) was used to record the voltage
in every 10-s interval and was connected to a computer
to visualize the graph.
Fig 1. MEC Reactor Design
2.2 Anaerobic Procedure
After crimp sealing the reactor with rubber septum, the
reactor will then be vacuumed and flushed for three
Fig 2. MEC reactor setup
2
E3S Web of Conferences 67, 02008 (2018)
3rd i-TREC 2018
https://doi.org/10.1051/e3sconf/20186702008
3 Result and Discussion
The medium used was formulated according to
Tugtas (2007) [16] but without the vitamin stock and
kept in sterile, anaerobic glass vial (Agilent) that had
been crimped with rubber septum. The procedures for
anaerobic flushing and sterilization are the same akin to
that for the reactor. As the medium contains 0.0002%
w/v methylene blue in its formulation, it gives off a light
blue coloration that will change to colorless in an
environment where oxygen has been depleted [17].
Medium is replaced after each gas chromatography
analysis by using sterile, disposable syringe inside a
laminar hood to minimize risk of contamination in which
depleted medium is sucked with a syringe, then the new
batch of medium is added with a new syringe. Each
replacement of medium adds approximately same
amount of volume as the working volume, which is 15
ml.
The experiment will consist of three reactors running
in parallel, with each reactor being given the same
treatment (ex: medium replacement, current input, etc.).
The first reactor will consists of wastewater sludge,
medium, and inoculum of denitrifier P. stutzeri (B1205,
InaCC LIPI). The second reactor functions as a negative
control, containing only wastewater sludge and medium.
The third reactor also functions as a negative control,
containing only the medium. Inoculum of P. stutzeri
were grown beforehand in the same medium inside a
sterile, anaerobic glass vial before being transferred with
a syringe to the first reactor in a laminar hood.
Wastewater sludge was obtained from Waduk Ria
Rio (Pulo Mas, Jakarta) to serve as the source of mixed
exoelectrogenic culture and stored beforehand in a
falcon tub inside a refrigerator. Wastewater will be
diluted a hundredfold before being used in the
experiment, in which phosphate-buffered saline was
used as the medium of dilution. The first reactor contains
1.5 ml diluted wastewater sludge, 1.5 ml inoculum, and
12 ml medium. The second reactor contains 1.5 ml
wastewater sludge and 13.5 ml medium. The third
reactor only contains 15 ml medium. From hereafter, the
first reactor shall be labelled “A”, the second reactor
labelled “B”, and the third reactor labelled “C”.
3.1 Visual Observation
Visual observation is done for the first two cycles in
order to verify the operation of the reactors. The
observed elements were the medium coloration and the
presence of gas bubbles in the cathode. The color of the
medium will change from light blue to colorless in an
oxygen-free environment, which indicates growth of
microorganism that eats the aqueous oxygen [17].
Cathodic bubbles were observed as it is an indicator of
evolution of H2 gas. The produced H2 will bubble due to
its low solubility in water [18] and rise up to accumulate
in the reactor headspace.
Observation of the first cycle (after ± 22 hrs elapsed
of reactor operation) showed that only reactor A and B
got discoloured in their medium, with reactor C retaining
its light blue methylene blue coloration and reactor A
being more colorless than B. This proves that there was
growth of microorganism in reactors inoculated with
inoculum and wastewater, but not in the reactor
containing only medium. It also indirectly proves that
the sterilization process for the medium was successful.
Furthermore, bubbles of gas were found attached on the
cathode of reactor A and B, meaning that the MEC setup was successful and that hydrogen gas is being
produced. On the contrary, no bubbles were observed in
reactor C which indicates that the voltage used was high
enough to enable H2 production by MEC mechanism but
low enough that H2 is not produced by any other sort of
mechanism such as water electrolysis [15].
Observation of the second cycle (after ± 20 hrs
elapsed of reactor operation) showed that reactor A and
B had gotten completely colorless, with reactor C
retaining its coloration as predicted. Compared to the
observation on the first cycle at around the same interval
(± 22 hrs), the rate of discoloration was much faster. It is
presumed that the microorganism had acclimatized to the
environment of the reactor and thus reduced the
methylene blue at a much faster rate.
3.2 Chromatographic Analysis
2.4 Analysis
In this study, accumulated gas in the headspace will be
analysed after certain intervals using a gas
chromatograph. This is due to the fact that if
accumulated gas is left too long, H2 may get fully
converted into methane by methanogens. The gas
chromatograph used (model 8A, Shimadzu) is equipped
with TCD (thermal conductivity detector) and utilizing
argon as the carrier gas in order to detect H2. In each
analysis, 1 ml of headspace gas was taken with
chromatography syringe to ensure no gas escapes from
the syringe, and then injected into the gas
chromatograph. From the resulting chromatogram,
composition of the gas can be deduced and quantified.
The section below will detail the result of each cycle of
reactor operation.
First Cycle
Fig 3. Chromatogram for the first cycle (69 hours)
3
E3S Web of Conferences 67, 02008 (2018)
3rd i-TREC 2018
https://doi.org/10.1051/e3sconf/20186702008
Analysis of the headspace taken after 69 hrs of
reactor operation is shown in Fig. 3 above.
Chromatogram result for reactor C yielded no peak of
hydrogen. This shows that indeed at the applied voltage
of 0.7 V, no H2 was produced by other mechanism (such
as water electrolysis) other than that of MEC working
principle, meaning that the applied voltage was
appropriate in this study. On the other hand, both reactor
A and B showed peak for hydrogen at around 0.6, with
reactor A’s headspace composed of 29.06% H2 and
reactor B’s headspace composed of 24.63% H2. This
proved that aside from the fact that the sludge used
contained living exoelectrogen culture, H2 was also
produced meaning that the set-up of the experiment was
successful. The small fluctuations seen on the
chromatogram is presumed to be noise and thus not
analysed. However, the predicted production difference
was not yet seen in this cycle, which may be caused by
either long reactor operation time that made H2 evolution
rate at a slower interval ot observed or simply that the
methanogens have yet to be active in the reactor. The
result of the first cycle is summarized in Figure 4 below.
From hereafter, chromatogram result will not be
displayed as the retention time of the peaks for the
consequent analysis remain the same.
proved the premise that the voltage applied was suitable,
for subsequent cycle we did not analyse for reactor C
anymore. On the other hand, both reactor A and B
showed peak for hydrogen at around 0.6, with reactor
A’s headspace composed of 10.85% H2 and reactor B’s
headspace composed of 5.60% H2.
This result proved the hypothesis in which the
presence of the denitrifer inside the reactor is beneficial,
in which in this case the H2 production is increased by
100%. Although the result showed higher production in
reactor inoculated with denitrifier, due to the scope of
this research, the direct mechanism behind these
phenomena is yet to be verified, aside from the
possibility that products of the denitrifier metabolism
inhibited methanogenic activites [10] [11] [12] and thus
reduced methanogen growth, increasing the net H2
harvested in the reactor.
Another thing to note is that the faster analysis time
(39 hrs vs. 69 hrs.) allowed better insight as to which
reactor can produce higher yield of H2 in a shorter
amount of time, something important if H2 production by
means of MEC is to be scaled-up.
Third Cycle
Fig 6. GC result for the third cycle (46 hours)
Fig 4. GC result for the first cycle (69 hours)
Analysis of the headspace taken after 46 hrs of
reactor operation is shown in Fig. 6 above. Only reactor
A showed peak for hydrogen at around 0.6, with reactor
A’s headspace composed of 2.95% H2. H2 peak was not
observed with reactor B. It is presumed that even after
longer operational time compared to the previous cycle
(46 hrs vs. 39 hrs) the H2 production rate has slowed
down or even not produced at all. This could mean that
either the denitrifiers had entered the lag phase due to
absence of nutrient or the exoelectrogenic culture from
the wastewater sludge which had entered lag phase.
Second Cycle
Fourth Cycle
Fig 5. GC result for the second cycle (39 hours)
Analysis of the headspace taken after 39 hrs of
reactor operation is shown in Fig. 5 above.
Chromatogram result for reactor C still did not yield any
peak of hydrogen as predicted. As consequent analysis
4
E3S Web of Conferences 67, 02008 (2018)
3rd i-TREC 2018
https://doi.org/10.1051/e3sconf/20186702008
exoelectrogen, and methanogen) as well as its validity in
a scale-up reactor.
4 Conclusion
The conclusion obtained from this research are that by
using small-scale MEC reactor to produce H2 as the
clean, renewable fuel of the future, it is beneficial to
introduce denitrifier microorganism in order to tackle the
methanogenesis activity present in wastewater that
consumes the produced H2. A reactor inoculated with P.
stutzeri running in parallel with an uninoculated one
showed higher H2 yield and better sustainability of H2
production, in which in the second cyle H2 production
increased 100% compared to control reactor. Further
research is still needed, however, in order to assess the
truly assess the interaction between the two
microorganism and the mechanism that leads to increase
of H2 evolution, as well as explaining the high jump in
H2 production in the last cycle.
Fig 7. GC result for the fourth cycle (71 hours)
Analysis of the headspace taken after 71 hrs of
reactor operation is shown in Fig. 7 above. Although
reactor B consistently yielded no peak for H2 akin to the
previous cycle, hydrogen peak was observed yet again in
reactor A. The headspace analysis for reactor A showed
that the H2 composition was 78.60%, the highest out of
any cycle. It is presumed that although the H2 production
rate slowed down in the previous cycle that may lead to
either denitrifier/exoelectrogen entering lag phase, not
all died. This is proven that by supplying fresh batch of
medium on a smaller number of microorganism (some
must be dead on the lag phase), allowed the
microorganism to produce H2 again by means of the
MEC. However, this large jump in H2 production needs
to be analysed further to exactly determine the real cause
of this huge leap of H2 generated. The overall result of
the experiment is summarized in Figure 8 below.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
Fig 8. Overall Composition of H2 on Reactor Headspace
[10]
Something to note is that, given that all reactors were
given the same kind of treatment, the same amount and
type of food, the same source of wastewater sludge, and
the same interval of headspace analysis, only reactor A
that has been inoculated with denitrifier P. stutzeri was
capable of maintaining its H2 production on subsequent
cycle and even saw an increase in production, while H2
production by reactor B saw a decline between the first
and second cycle. We posit that the presence of
denitrifier in a MEC reactor is highly beneficial if one
wishes to produce H2 by means of MEC as it increases
H2 production yield. Further research may be required in
order to truly assess the interaction between the three
types of organism present in the reactor (denitrifier,
[11]
[12]
[13]
[14]
[15]
[16]
[17]
5
A. Kadier, M.S. Kalil, P. Abdeshahian, K.
Chandrasekhar, A. Mohamed, N.F. Azman, W.
Logroño, Y. Simayi, A.A. Hamid, Renew.
Sustain. Energy Rev. 61 (2016) 501–525.
P. Hoffmann, U.K. Mirza, Electron. Green J.
(2002).
P. Kumar, R. Britter, N. Gupta, J. Fuel Cell Sci.
Technol. 6 (2009) 021009.
H. Liu, S. Grot, B.E. Logan, Environ. Sci.
Technol. 39 (2005) 4317–4320.
T.H.J.A. Sleutels, A. Ter Heijne, C.J.N.
Buisman, H.V.M. Hamelers, ChemSusChem 5
(2012) 1012–1019.
L. Lu, N. Ren, X. Zhao, H. Wang, D. Wu, D.
Xing, Energy Environ. Sci. 4 (2011) 1329.
R.A. Rozendal, H.V.M. Hamelers, K. Rabaey, J.
Keller, C.J.N. Buisman, Trends Biotechnol. 26
(2008) 450–459.
D. Clifford, X. Liu, J. Am. Water Works Assoc.
(1993) 135–143.
W. Zumft, Microbiol Mol Biol Rev 61 (1997)
533–616.
K.C. Chen, Y.F. Lin, Water Res. 27 (1993)
1749–1759.
A.E. Tugtas, S.G. Pavlostathis, Biotechnol.
Bioeng. 96 (2007) 444–455.
A.E. Tugtas, U. Tezel, S.G. Pavlostathis,
Biotechnol. Bioeng. 105 (2010) 98–108.
C.A. Carlson, J.L. Ingraham, Appl. Environ.
Microbiol. 45 (1983) 1247–1253.
L. Rago, Y. Ruiz, J.A. Baeza, A. Guisasola, P.
Cortés, Bioelectrochemistry 106 (2015) 359–
368.
D.F. Call, B.E. Logan, Biosens. Bioelectron. 26
(2011) 4526–4531.
A.E. Tugtas, (2007).
D. Karakashev, D. Galabova, I. Simeonov,
World J. Microbiol. Biotechnol. 19 (2002) 233–
E3S Web of Conferences 67, 02008 (2018)
3rd i-TREC 2018
[18]
https://doi.org/10.1051/e3sconf/20186702008
238.
A. Kadier, Y. Simayi, P. Abdeshahian, N.
Farhana, K. Chandrasekhar, M. Sahaid,
ALEXANDRIA Eng. J. (2015).
6