Factors affecting the longterm stability of biomass and hydrogen productivity in outdoor photofermentation
Post on 08-Feb-2023
0 Views
Preview:
Transcript
ww.sciencedirect.com
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 en e r g y 3 7 ( 2 0 1 2 ) 1 6 4 2 1e1 6 4 2 9
Available online at w
journal homepage: www.elsevier .com/locate/he
Amelioration of photofermentative hydrogen production frommolasses dark fermenter effluent by zeolite-based removalof ammonium ion
Dominic Deo Androga a,*, Ebru Ozgur b, Inci Eroglu b, Ufuk Gunduz c, Meral Yucel c
aDepartment of Biotechnology, Middle East Technical University, Ankara 06800, TurkeybDepartment of Chemical Engineering, Middle East Technical University, Ankara 06800, TurkeycDepartment of Biological Sciences, Middle East Technical University, Ankara 06800, Turkey
a r t i c l e i n f o
Article history:
Received 11 November 2011
Received in revised form
24 February 2012
Accepted 28 February 2012
Available online 28 March 2012
Keywords:
Dark fermentation
Photofermentation
Ammonium
Clinoptilolite
Rhodobacter capsulatus
* Corresponding author. Tel.: þ90 312 210 26E-mail address: dominicdeo@gmail.com
0360-3199/$ e see front matter Copyright ªdoi:10.1016/j.ijhydene.2012.02.177
a b s t r a c t
One of the challenges in the development of integrated dark and photofermentative bio-
logical hydrogen production systems is the presence of ammonium ions in dark fermen-
tation effluent (DFE). Ammonium strongly inhibits the sequential photofermentation
process, and so its removal is required for successful process integration. In this study, the
removal of ammonium ions from molasses DFE using a natural zeolite (clinoptilolite) was
investigated. The samples were treated with batch suspensions of Na-form clinoptilolite.
The ammonium ion concentration could be reduced from 7.60 mM to 1.60 mM and from
12.30 mM to 2.40 mM for two different samples. Photofermentative hydrogen production
on treated and untreated molasses DFE samples were investigated in batch photo-
bioreactors by an uptake hydrogenase deleted (hup�) mutant strain of Rhodobacter capsu-
latus. Maximum hydrogen productivities of 1.11 mmol H2/Lc$h and 1.16 mmol H2/Lc$h and
molar yields of 79% and 90% were attained in the treated DFE samples, while the untreated
samples resulted in no hydrogen production. The results showed that ammonium ions in
molasses DFE could be effectively removed using clinoptilolite by applying a cost-effective,
simple batch process.
Copyright ª 2012, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights
reserved.
1. Introduction biological hydrogen production processes can be improved by
Biological hydrogen production processes, namely bio-
photolysis, dark fermentation and photofermentation, offer
the prospect of producing hydrogen from renewable sources
derived from waste streams and agricultural residues. Low
hydrogen yields and production rates, however, are the major
barriers in developing these technologies [1]. Nevertheless,
studies have demonstrated that the performance of these
96; fax: þ90 312 210 2600.(D.D. Androga).2012, Hydrogen Energy P
integrating them in 2- or 3-stage processes [2e5].
One widely investigated integrated system is the sequen-
tial dark and photofermentation system [6]. In the dark
fermentation stage, anaerobic bacteria like the Clostridium
species break down carbohydrate-rich substrates (sugars) to
produce H2, CO2 and short-chain organic acids such as acetic
acid and butyric acid Eq. (1). The organic acids produced in the
dark fermentation stage can be used in a sequential
ublications, LLC. Published by Elsevier Ltd. All rights reserved.
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 1 6 4 2 1e1 6 4 2 916422
photofermentation stage where photoheterotrophic bacteria
convert them to hydrogen and carbon dioxide Eq. (2) [1,6]. The
photofermentation process occurs under anaerobic and
nitrogen limited conditions and utilizes light energy to supply
ATP that is needed in the complete conversion of the organic
acids to hydrogen and carbon dioxide [1].
C6H12O6þ2H2O/2CH3COOHþ4H2þ2CO2 DG : �206 kJ (1)
2CH3COOHþ4H2O/8H2þ4CO2 DG: 104:6 kJ�2¼209:2 kJ (2)
Theoretically, presuming acetic acid to be the major by-
product of the dark fermentation process, 12 mol of H2
(4 mol of H2 from dark fermentation þ 8 mol of H2 from pho-
tofermentation) can be produced per mol of glucose in the
integrated processes. However, actual yields are lower
because part of the glucose is used for microbial growth
and maintenance [2,6]. Approximately 9 mol of H2/mol of
glucose (75% conversion efficiency) can be attained in an
integrated thermophilic fermentation and photofermentation
process [2].
One of the major challenges in these systems is the pres-
ence of fixed nitrogen in the dark fermentation feed, which
leads to the formation of ammonium ions that are inhibitory
to the photofermentation process [7]. Ammonium ions are
known to repress the nitrogenase enzyme which is respon-
sible for catalyzing hydrogen production in the photosyn-
thetic bacteria [8]. High ammonium concentrations inhibit the
nitrogenase enzyme either directly via feedback inhibition
(reversible) or at the genetic level by the repression of nif genes
that encode subunits of the nitrogenase enzyme [9]. An
ammonium ion concentration not exceeding 2mM in the dark
fermenter effluent has been reported to be suitable for pho-
tofermentative hydrogen production [3,4,8].
Several alternatives have been suggested to address the
issue of inhibitory ammonium concentrations in the dark
fermenter effluent. The commonly applied method is dilution
of the effluent to reduce ammonium concentration
[3,4,10e13]. This technique is effective in reducing ammo-
nium ion inhibition, but it also reduces the concentration of
other essential components in the feed and thus may cut
down hydrogen production efficiency. Also, dilution incurs
extra costs, which are particularly unfavorable for large scale
systems.
Another suggestion is the use of ammonium ion tolerant
photosynthetic bacteria (wildtype or mutant). Various
researchers have applied genetic modification to develop
mutant bacterial strains that can grow and produce hydrogen
at high ammonium concentrations [9,14]. However, great
metabolic and genetic diversity exists among the purple non
sulfur bacteria (PNS) such that further studies are required to
isolate or develop strains that can grow and produce hydrogen
at high rates and yields under high ammonium ion concen-
trations. Moreover, the use of mutant strains in industrial
settings can be problematic and expensive [7].
Another method of circumventing the inhibitory effects of
ammonium ions on photofermentative hydrogen production
is the use of wastewater treatment methods. A variety of
biological and physicochemical methods and technologies
such as air stripping, breakpoint chlorination, nitrification-
denitrification and ion exchange are commonly applied to
remove ammonium from wastewater [15]. Among these
techniques, ion exchange using natural zeolites is more
attractive because of its low cost and simplicity of application
and operation. Moreover, the zeolites can be regenerated for
reuse as fertilizer or animal feed supplement [16]. Previous
studies by Eroglu et al. [17] have demonstrated the successful
application of olive mill wastewater treatment using clay and
zeolite for photofermentative hydrogen production.
Clinoptilolite is a zeolite mineral with an aluminosilicate
tetrahedral crystalline structure that contains exchangeable
cations such as Kþ, Naþ, Mg2þ and Ca2þ. It has a Si/Al > 4 and
its structure is: (Na,K,Mg)6(Al6Si30O7)$24H2O [18]. Owing to its
sorptive properties, clinoptilolite has been applied in the
removal of heavy metals [19], dye contaminants [20] and
ammonium [15,18,21] from wastewater. It was reported to
selectively adsorb cations in the order: Pb2þ > NHþ4 >
Ba2þ > Cu2þ > Zn2þ > Cd2þ > Co2þ and have a maximum
exchange capacity of 2.6 mEq/g [22].
This study aimed to investigate the possibility of using
natural clinoptilolite zeolite to reduce the ammonium ion
content in molasses dark fermenter effluent that has high
ammonium concentrations. The molasses dark fermenter
effluent was treated with zeolite in batch processes, and
parameters such as ammonium ion concentration, pH levels,
color change, and duration of treatment were investigated.
The resulting effluents were analyzed for both metal cation
content and organic acid concentrations and tested for pho-
tofermentative hydrogen production by Rhodobacter capsulatus
YO3 (hup�).
2. Materials and methods
2.1. Clinoptilolite zeolite pre-treatment
Clinoptilolite samples used in this study were taken from
Manisa-Gordes, Western Anatolia region of Turkey. The
zeolite samples were ground and sieved between mesh 30/60
(0.25 to 0.60 mm) particle size and pre-treated with brine
solution. Pre-treatment aimed to replace the exchangeable
cations in the zeolite (Kþ, Mg2þ, Ca2þ) with the more easily
removable ones such as Naþ ions, therefore creating a near
homoionic form. This improved the zeolite’s effective
exchange capacity, subsequently providing better perfor-
mance in ion exchange application [22]. The clinoptilolite
samples were batch pre-treated using the procedure applied
by Bayraktaroglu [23]. Mixtures of 50 g of clinoptilolite and
500 ml of 1 M NaCl solution were stirred using a wrist-action
shaker (Fisher-Kendall mixer) at ambient temperature for 13
days. The NaCl exchange solution was replaced with freshly
prepared 1 M NaCl at 24 h intervals. The pre-treatment
procedure was stopped when the cation concentration (Kþ,Ca2þ and Mg2þ) in the exchange solution effluent reached
saturation. The mixture was then decanted and the zeolite
washed using deionizedwater to get rid of NaCl traces. During
washing, presence of Naþ was checked using 1% (v/v) AgNO3
solution, then it was dried in an oven set at 70 �C for 24 h Table
1 lists the chemical composition of the zeolite before and after
pre-treatment. The analyses were carried out using lithium
metaborate fusion dissolution method [24].
Table 1 e The chemical composition of the Gordesclinoptilolite zeolite.
Component Original (% w/w) Na-form (% w/w)
SiO2 68.12 68.06
Al2O3 11.15 11.10
CaO 1.70 0.40
MgO 1.50 0.90
Fe2O3 3.47 2.46
Na2O 2.32 5.54
K2O 1.98 0.99
Othersa 9.76 10.55
LOIb 10.06 10.14
a Mixture of MnO, TiO2, P2O5, SO3 and H2O.
b Loss if ignition.
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 en e r g y 3 7 ( 2 0 1 2 ) 1 6 4 2 1e1 6 4 2 9 16423
2.2. Ammonium ion reduction in the molasses DFE
Molasses DFEs were delivered by DLO-FBR (Wageningen UR
Food and Biobased Research, The Netherlands). The samples
M1 and M2 were the effluents of two separate thermophilic
dark fermentation processes [3] which were carried out with
different ammonium chloride concentrations. Composition of
the effluents is shown in Table 2. The ammonium concen-
trations of M1 and M2 were 7.60 mM and 12.30 mM,
respectively.
The molasses DFE samples (200 ml) were treated with Na-
form clinoptilolite zeolite (10 g) in an Erlenmeyer flask placed
in a water bath shaker (Clifton NE25) at 25 �C and 70 rpm.
During treatment, ammonium concentration, pH and color of
the liquid samples were measured and recorded at 30 min
intervals.
2.3. Bacteria culture and media
R. capsulatus YO3 (hup�), an ‘uptake hydrogenase’ deleted
strain of R. capsulatus MT1131 [25] was used in this study. The
inoculums were prepared in modified Biebl and Pfennig (BP)
media [26]. Compositions of treated and untreated molasses
DFEs are given in Table 2. The hydrogen production
Table 2 e Comparisons of the change in the organic acidconcentrations, cation concentrations, total nitrogen(TN), total organic carbon (TOC) and the chemical oxygendemand (COD) of the two molasses dark fermentereffluent samples before (M1 and M2) and after treatment(M1T and M2T) with the Na-form clinoptilolite zeolite.
Molasses DFE sample M1 M1T M2 M2T
Ammonium (mM) 7.60 1.60 12.30 2.30
Mg2þ(mg/L) 54.72 70.25 70.75 61.44
Fe3þ(mg/L) 1.87 0.75 1.98 0.52
Lactic acid (mM) 4.40 4.30 1.10 0.90
Formic acid (mM) 0.50 0.40 4.00 4.00
Acetic acid (mM) 85.40 82.60 98.00 95.00
Butyric acid (mM) 1.40 1.30 13.00 13.00
TN (M) 0.026 0.020 0.027 0.016
TOC (M) 0.078 0.078 0.079 0.078
C/N molar ratio 3.00 3.90 2.93 4.88
COD (mg/L) 3380 3340 3180 3120
experiments were carried out using two times diluted
molasses DFE samples. Samples were centrifuged and steril-
ized by autoclaving to remove contaminants and any colloidal
materials that may interfere with light penetration into the
photobioreactors. Untreated molasses DFE that contained
high ammonium concentrations were used as controls. The
DFE sampleswere supplementedwith iron (Fe citrate, 0.1mM)
andmolybdenum (Na2MoO4$2H2O, 0.16 mm) and 5mM sodium
carbonate (Na2CO3) buffer. Hydrogen production experiments
were carried out in duplicate glass bottle photobioreactors
(55 ml) inoculated with freshly grown bacterial cultures (10%).
The photobioreactors were maintained at 30e32 �C in an
incubator. Continuous illumination was provided by 60 W
tungsten lamps adjusted to provide a uniform light intensity
of 170W/m2 at the surface of the reactors. The initial pH in the
photobioreactors was 6.6. Hydrogen production was followed
by water displacement method, using calibrated glass
columns.
The performance of hydrogen production was analyzed
based on the maximum hydrogen production rate and
maximum molar yield. The hydrogen production rate (mmol/
Lc/h) was determined as the number of moles of hydrogen
produced in a given time period by the bioreactor culture
volume. The percent molar yield (%) was calculated as the
ratio of the moles of hydrogen produced from the experiment
to the moles of theoretical hydrogen that would have been
produced if all the acetic acid utilized had been converted to
hydrogen [27].
2.4. Analytical methods
During the clinoptilolite pre-treatment procedure, the
concentrations of Kþ, Mg2þ and Ca2þ were determined using
atomic absorption spectroscopy (Philips, PU9200X) and Naþ
ions using a flame photometer (Jenway Model PFP7). The
ammonium ion concentration, total nitrogen (TN), total
organic carbon (TOC) and chemical oxygen demand (COD)
were determined spectrophotometrically using their respec-
tive Hach-Lange kits and DR/2400 Hach-Lange Spectropho-
tometer, Germany. The bacterial cell concentration, organic
acid analyses, evolved gas analyses and pH were measured as
previously described [27]. The color of the molasses DFE was
analyzed using a DR/2400 Hach-Lange spectrophotometer.
3. Results and discussion
3.1. Pre-treatment of the clinoptilolite zeolite
The sodium content of the clinoptilolite zeolite was success-
fully increased from 2.32% to 5.54% after pre-treatment (Table
1). The exchangeable cations in the clinoptilolite zeolite (Kþ,
Mg2þ and Ca2þ) decreased as theywere replacedwith Naþ ions
from the exchange solution. No further ion exchange was
observed after 10 days of pre-treatment (Fig. 1). Bayraktaroglu
[23] also reported that 10 days were adequate to attain Na-
form clinoptilolite applying the same batch pre-treatment
procedure.
Fig. 1 e The change in the concentration of exchangeable
cations in the clinoptilolite zeolite during pre-treatment
with 1 N brine solution. (A) KD, (-) Mg2D and (:) Ca2D.
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 1 6 4 2 1e1 6 4 2 916424
3.2. Treatment of molasses dark fermenter effluent
Table 2 summarizes ammonium ion, organic acid, cation
concentrations, total nitrogen (TN), total organic carbon
(TOC), carbon to nitrogenmolar ratio and the chemical oxygen
demand (COD) of the molasses dark fermenter effluent
samples (M1 and M2) and the treated samples with the Na-
form clinoptilolite zeolite (M1T and M2T).
Ammonium ion concentration decreased from 7.60 mM to
1.60 mM in sample M1 and from 12.30 to 2.30 mM in sample
Fig. 2 e The variation of ammonium ion concentration, pH and
with different initial NHD4 concentrations using Na-form clinop
([NHD4 ]Initial [ 12.30 mM). (-) Ammonium, (,) pH and (D) color
M2 (Fig. 2A and B). The duration of treatment varied depend-
ing on the initial NHþ4 concentration; sampleM1 took a shorter
time (120 min) compared to sample M2 (180 min). Treatment
was stopped when the NHþ4 concentration decreased to
around 2 mM, which was the threshold concentration for
photofermentative hydrogen production using R. capsulatus
YO3 (hup�) [3]. High ammonium ion removal efficiency
(around 80%) was achieved for both samples. Successful NHþ4
removal obtained in this study confirmed the high selectivity
and suitability of using clinoptilolite for ammonium ion
removal as previously reported in literature [15,18,21]. The
presence of organic compounds in dark fermenter effluent
might enhance the ammonium ion removal by reducing the
surface tension of the aqueous phase to the point of raising
access of the aqueous phase to the macropores of the zeolite
[28].
The real exchange capacity (REC) of the Na-form clinopti-
lolite was determined as 2 mg NHþ4 =g and 3 mg NHþ
4 =g, for
samples M1 and M2, respectively. These values are compa-
rable to the adsorption capacity (2e30 mg NHþ4 =g) of clinopti-
lolite reported for ammonium in literature [29]. However, they
were lower than the calculated theoretical exchange capacity
(TEC) of the pre-treated zeolite which was 45.54 mg NHþ4 =g.
The difference between the REC and TEC could be attributed to
incomplete removal of cations from the zeolite during treat-
ment. That might occur either due to strong bonding of
cations on zeolite structure or interference of competing ions
color during treatment of molasses dark fermenter effluent
tilolite. (A) M1 ([NHD4 ]Initial [ 7.60 mM) and (B) M2
.
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 en e r g y 3 7 ( 2 0 1 2 ) 1 6 4 2 1e1 6 4 2 9 16425
existing in the molasses DFE samples. Also, parameters such
as the cation concentration, pH, temperature and agitation
speed may affect ion exchange [18,21,22].
During the treatment process, pH ranged between 7.0 and
7.6 in sample M1 and between 8.8 and 9.0 in sample M2. In
literature, pH 6e7 was suggested for optimal ammonium
removal [30], however pH was not significant for the removal
of ammonium ion from molasses DFE.
Negligible color change was observed during the treatment
of the DFEs. Color ranged between 2090 and 2120 PtCo APHA
and 2560 and 2780 PtCo APHA in the M1 and M2 samples,
respectively (Fig. 2A and B).
Elemental analysis of themolasses DFE samples before and
after treatment revealed that some of the cations (Kþ, Mg2þ,Ca2þ, Fe3þ) were released from the zeolite while others were
taken up from the molasses DFE. After treatment the
concentration of Mg2þ increased in sample M1, while it
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
6.0
6.4
6.8
7.2
7.6
8.0
8.4
8.8
9.2
9.6
10.0
0 1 2 3 4 5 60.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Bio
mas
s (g
DC
W/L
c)pH
Cum
ulat
ive
volu
me
of H
2 pro
duce
d (L
/Lc)
Tim
Fig. 3 e Biomass growth (A), pH change (B) and cumulative hyd
during photofermentation on treated and untreated molasses d
times. The initial ammonium concentrations of the treated sam
untreated samples (DM1 (B) and DM2 (D)) were 3.8 mM and 6.2
decreased in sample M2 (Table 2). The variation in the uptake
and release of the cations by the zeolite could be attributed to
the interactions between the exchangeable cations and the
differences in pH of the samples. It was reported that pH
affected the surface charges on the zeolite, therefore
impacting cation exchange [31]. Under high pH, OH� ions
could form complexeswith heavymetals, while under low pH,
Hþ ions tend to competewith cations for exchange sites on the
zeolite, thereby decreasing its efficiency [32]. Generally, pH
range between 3 and 11 is applied in metal cation removal
processes using clinoptilolite [33].
It was also observed that the amount of Fe (III) in the
molasses DFE decreased after treatment with clinoptilolite e
circa 62% and 76% in the M1 and M2 samples, respectively
(Table 2). Iron is an important co-factor of the nitrogenase
enzyme that is responsible for hydrogen production in the
photosynthetic bacteria. Supplementation of molasses DFE
7 8 9 10 11 12 13
e (days)
B
A
C
rogen production (C) by Rhodobacter capsulatus YO3 (hupL)
ark fermenter effluents. The samples were diluted two
ples (DM1T(,) and DM2T(>)) were 0.2 mM and 1.2 mM,
mM, respectively.
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 1 6 4 2 1e1 6 4 2 916426
with Fe was reported to improve hydrogen yield and produc-
tivity [3,4]. In practice, although the addition of Fe to the
molasses DFE is to be carried out after the removal of
ammonium using clinoptilolite (before photofermentative
hydrogen production), further investigations are needed to
determine treatment conditions that would prevent Fe uptake
during the ammonium removal process. This would reduce
the amount of Fe to be supplemented to the DFE and save on
raw materials costs in an industrial setting.
The ammoniumremoval procedure using clinoptilolite had
negligible effect on the organic acid concentrations, TOC and
CODof themolassesDFEs (Table 2). Clinoptilolitewas reported
to have a high affinity for cations like ammonium and heavy
metals like lead and cadmium [16,31e33]. This is a desirable
trait for the treatment of the DFEs. The inhibitory effects of
0
10
20
30
40
50
0
10
20
30
40
50
Con
cent
rati
on o
f or
gani
c ac
id (
mM
)
0
10
20
30
40
50
0 1 2 3 4 5 6
0
10
20
30
40
50
Time (d
Fig. 4 e Organic acid and ammonium ion utilization by Rhodoba
molasses dark fermenter effluents. (A) DM1T ([NHD4 ]Initial [ 0.2
([NHD4 ]Initial [ 3.8 mM); (D) DM2 ([NHD
4 ]Initial [ 6.2 mM). (C) Acetic
ammonium ion.
ammonium ions and heavy metals on photofermentation can
be reduced, therefore improving hydrogen production rates
and yield. TN in themolasses DFE decreased as a consequence
of the NHþ4 removal during treatment (Table 2).
3.3. Photofermentation
The hydrogen production experiments were carried out using
two times dilutedmolasses DFE samples. Dilution was carried
out to adjust the concentration of acetic acid in the DFE
samples to around 40 mM, which is reported to be optimum
for photofermentative hydrogen production [34e36]. Higher
biomass concentrations were obtained in the batch experi-
ments using the treated molasses DFE samples (DM1T and
DM2T) compared to the untreated ones (DM1 and DM2). On
0
1
2
3
4
5
6
7 8 9 10 11 12 13
ays)
012345678
Concentration of am
monium
B
A
C
D
cter capsulatus YO3 (hupL) on treated and untreated
mM); (B) DM2T ([NHD4 ]Initial [ 1.2 mM); (C) DM1
acid, (>) formic acid, (6) butyric acid ( ) lactic acid and (:)
Fig. 5 e The average daily variation in total gas
composition during photofermentation by Rhodobacter
capsulatus YO3 (hupL) on treated molasses dark fermenter
effluents DM1T and DM2T with initial ammonium
concentrations 0.2 mM and 1.2 mM, respectively. H2 (-)
and CO2 (,).
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 en e r g y 3 7 ( 2 0 1 2 ) 1 6 4 2 1e1 6 4 2 9 16427
average, maximum biomass concentrations of 0.7 gdcw/Lcand 0.4 gdcw/Lc were observed in the treated and untreated
samples, respectively (Fig. 3A). A rapid increase in pH was
observed as the bacteria grew. In the experiments using
samples DM1T and DM2T, pH increased from 6.6 to around 7.8
on the second day of the experiment and then fluctuated
between 7.6 and 8.1 during the rest of the experiments
(Fig. 3B). However, in the experiments using samples DM1 and
DM2, pH sharply increased from 6.6 to around 9 on the second
day of the experiment. Then, it ranged between 8.4 and 9.6
(Fig. 3B). The high pH (>9) could have led to the low bacterial
growth observed in the untreated DFE samples. During pho-
tofermentation, bacterial cells maintain their membrane
potential by the efflux of negatively charged OH� ions from
the cell to counteract the effect of taking in negatively charged
organic acids [3]. Therefore, high pH in the feed could prevent
the cell from maintaining its membrane potential, therefore
affecting cellular metabolism and eventually hindering cell
growth. PNS bacteria such as R. capsulatus are reported to
optimally grow between pH 6e9 [37].
The findings in this study also showed that in the presence
of high ammonium concentrations (as in samples DM1 and
DM2), 5 mM Na2CO3 buffer used in the photofermentation
experiments was inadequate to maintain pH at tolerable
levels (pH 7e7.5). Further increase in buffer concentration
using 10, 15 and 20 mMNa2CO3 or 22 mM KH2PO4 gave similar
results (data not shown). Hence, higher buffering capacity is
required to maintain pH at the desirable levels.
Maximum hydrogen productivities of 1.11 mmol H2/Lc/h
and 1.16 mmol H2/Lc/h and molar yields of 79% and 90% were
obtained in the experiments using samples DM1T and DM2T,
respectively. No hydrogen production was observed in the
DM1 and DM2 samples (Fig. 3C). This could be because of the
inhibition of the nitrogenase enzyme by the initially high
ammonium concentrations (3.8 mM and 6.2 mM NHþ4 ). In
experiments using batch cultures of Rhodobacter sphaeroides
and growth media containing different NH4Cl concentrations
(1 to 10 mM), Akkose et al. [8] observed no hydrogen produc-
tion in NH4Cl concentration above 2 mM. They attributed this
to the inhibition of nitrogenase by ammonium. Similarly,
Ozgur et al. [3] observed no hydrogen production in batch
cultures of R. capsulatus hup� fed with sugar beet molasses
DFE containing ammonium >2 mM. However, they obtained
hydrogen productivity of 1.37mmol H2/Lc/h andmolar yield of
58% inmolasses DFE samples lacking ammonium. The results
obtained in this study compare to that in literature
[1,3,4,6,10e12].
Organic acid analysis results showed that acetic acid was
well utilized for growth and hydrogen production by the
bacteria in the DM1T and DM2T containing photobioreactors
compared to their DM1 and DM2 counterparts (Fig. 4). In the
DM1T and DM2T experiments, hydrogen production stopped
after the depletion of acetic acid on the 6th day of the exper-
iment (Fig. 4A and B). In contrast, no hydrogen productionwas
observed and around 15 to 20 mM of acetic acid remained at
the end of the DM1 and DM2 experiments (Fig. 4C and D).
Lactic acid and propionic acid were consumed in all the pho-
tobioreactors while butyric acid was only depleted in low
ammonium containing photobioreactors. Accumulation of
formic acid, ranging between 15 and 20mMwas also observed
in the low NHþ4 photobioreactors (Fig. 4C and D). It could have
either been produced by NAD-dependent formate dehydro-
genase which reversibly catalyzes formic acid synthesis from
carbon dioxide and NADH, Hþ or as a fermentation end
product through pyruvate by the action of pyruvate formate
lyase. Avcioglu et al. [11] also observed formic acid production
in fed-batch cultures of R. capsulatus in experiments using
molasses DFE lacking NHþ4 ions.
The analyses of the evolved gas showed that at the
beginning of the experiment (during exponential growth) the
gas composition was 98% hydrogenwith the rest being carbon
dioxide. The gas quality decreased as the bacteria approached
stationary phase (86% hydrogen) and ranged between 71 and
76% hydrogen for the rest of the experiment (Fig. 5).
4. Conclusions
Ammonium ions in molasses dark fermenter effluent can be
effectively removed using natural clinoptilolite zeolite by
applying a cost-effective batch process. The treatment was
successfully carried outwith pH levels ranging from 7 to 9. The
reduction of ammonium ions from the dark fermenter
effluent samples led to the attainment of a high hydrogen
productivity and molar yield in the photofermentation
experiments. This is a significant improvement in integrating
dark and photofermentation processes for biohydrogen
production, which is a sustainable way of producing hydrogen
from renewable resources such as biomass.
Acknowledgments
This research study was supported by the EU 6th Framework
Integrated Project 019825 (HYVOLUTION). The authors would
like to thank Dr Truus de Vrije from DLO FBR (Wageningen UR
Food and Biobased Research, The Netherlands), for providing
the molasses dark fermenter effluent samples, Dr. Yavuz
Ozturk from GMBE TUBITAK-MAM, Gebze, Turkey for
providing the R. capsulatus YO3 (hup�) mutant strain used in
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 1 6 4 2 1e1 6 4 2 916428
this study and Kerime Guney from the Chemical Engineering
Department at the Middle East Technical University, for
carrying out the elemental analysis of the molasses dark
fermenter effluent samples and the component analysis of the
natural clinoptilolite zeolite. Dominic Deo Androga acknowl-
edges the Scientific and Technological Research Council of
Turkey (TUBITAK-BIDEB) for providing financial support
through the PhD Fellowships for Foreign Citizens (Code 2215)
program.
Acronyms
BP Biebl and Pfennig
COD Chemical oxygen demand, mg/L
DFE Dark fermentation effluent
DM1 Two times diluted molasses dark fermenter effluent
sample containing 3.80 mM ammonium ion
DM2 Two times diluted molasses dark fermenter effluent
sample containing 6.20 mM ammonium ion
DM1T Treated (ammonium ion reduced) and two times
diluted molasses dark fermenter effluent sample
containing 0.80 mM ammonium ion
DM2T Treated and two times diluted molasses dark
fermenter effluent sample containing 1.20 mM
ammonium ion
gdcw Gram dry cell weight
hup� Membrane bound uptake hydrogenase deficient,
(mutant)
Lc Liter culture
LOI Loss of ignition
M1 Molasses dark fermenter effluent sample containing
7.60 mM ammonium
M2 Molasses dark fermenter effluent sample containing
12.30 mM ammonium
M1T Treated molasses dark fermenter effluent sample
containing 1.60 mM ammonium ion
M2T Treated molasses dark fermenter effluent sample
containing 2.30 mM ammonium ion
NADH Reduced nicotinamide adenine dinucleotide
NHþ4 Ammonium ion
PNS Purple non sulfur
PtCoAPHA Color unit based on the standards of American
Public Health Association. One color unit is equal
to 1 mg/l platinum as chloroplatinate ion
REC Real exchange capacity, mg NHþ4 =g
TEC Theoretical exchange capacity, mg NHþ4 =g
TOC Total organic carbon, M
TN Total nitrogen, M
r e f e r e n c e s
[1] Argun H, Kargi F. Bio-hydrogen production by differentoperational modes of dark and photofermentation: anoverview. Int J Hydrogen Energy 2011;36:7443e59.
[2] Claassen PAM, de Vrije T, Koukios E, van Niel E, Eroglu I,Modigell M, et al. Non-thermal production of pure hydrogenfrom biomass: HYVOLUTION. J Clean Prod 2010;18:S4e8.
[3] Ozgur E, Mars AE, Peksel B, Louwerse A, Yucel M, Gunduz U,et al. Biohydrogen production from beet molasses bysequential dark and photofermentation. Int J HydrogenEnergy 2010;35:511e7.
[4] Afsar N, Ozgur E, Gurgan M, Akkose S, Yucel M, Gunduz U,et al. Hydrogen productivity of photosynthetic bacteria ondark fermenter effluent of potato steam peels hydrolysate.Int J Hydrogen Energy 2011;36:432e8.
[5] Lo YC, Chen CY, Lee CM, Chang JS. Sequential dark-photofermentation and autotrophic microalgal growth for high-yield and CO2-free biohydrogen production. Int J HydrogenEnergy 2010;35:10944e53.
[6] Keskin T, Abo-Hashesh M, Hallenbeck PC. Photofermentativehydrogen production from wastes. Bioresour Technol 2011;102:8557e68.
[7] Guwy AJ, Dinsdale RM, Kim JR, Massanet-Nicolau J,Premier G. Fermentative biohydrogen production systemsintegration. Bioresour Technol 2011;102:8534e42.
[8] Akkose S, Gunduz U, Yucel M, Eroglu I. Effects of ammoniumion, acetate and aerobic conditions on hydrogen productionand expression levels of nitrogenase genes in Rhodobactersphaeroides O.U.001. Int J Hydrogen Energy 2009;34:8818e27.
[9] Pekgoz G, Gunduz U, Eroglu I, Yucel M, Kovacs K, Rakhely G.Effect of inactivation of genes involved in ammoniumregulation on the biohydrogen production of Rhodobactercapsulatus. Int J Hydrogen Energy 2011;36:13536e46.
[10] Ozgur E, AfsarN, deVrije T, YucelM,GunduzU, Claassen PAM,et al. Potential use of thermophilic dark fermentationeffluents in photofermentative hydrogen production byRhodobacter capsulatus. J Clean Prod 2010;18:S23e8.
[11] Avcioglu SG, Ozgur E, Eroglu I, Yucel M, Gunduz U.Biohydrogen production in an outdoor panel photobioreactoron dark fermentation effluent of molasses. Int J HydrogenEnergy 2011;36:11360e8.
[12] Ozkan E, Uyar B, Ozgur E, Yucel M, Eroglu I, Gunduz U.Photofermentative hydrogen production using darkfermentation effluent of sugar beet thick juice in outdoorconditions. Int J Hydrogen Energy 2012;37:2044e9.
[13] Azbar N, Dokgoz FTC. The effect of dilution and l-malic acidaddition on bio-hydrogen production with Rhodopseudomonaspalustris from effluent of an acidogenic anaerobic reactor. IntJ Hydrogen Energy 2010;35:5028e33.
[14] Li X, Liu T, Wu Y, Zhao G, Zhou Z. Derepressive effect of NH4þ
on hydrogen production by deleting the glnA1 gene inRhodobacter sphaeroides. Biotechnol Bioeng 2010;106:564e72.
[15] Demir A, Gunay A, Debik E. Ammonium removal fromaqueous solution by ion exchange using packed bed naturalzeolite. Water SA 2002;28:329e36.
[16] Leung S, Barrington S, Wan Y, Zhao X, El-Husseini B. Zeolite(clinoptilolite) as feed additive to reduce manure mineralcontent. Bioresour Technol 2007;98:3309e16.
[17] Eroglu E, Eroglu I, Gunduz U, Yucel M. Treatment of olive millwastewater by different physicochemical methods andutilization of their liquid effluents for biological hydrogenproduction. Biomass Bioenergy 2009;33:701e5.
[18] Karadag D, Koc Y, Turan M, Armagan B. Removal ofammonium ion from aqueous solution using natural Turkishclinoptilolite. J Hazard Mater 2006;B136:604e9.
[19] Tao YF, Qiu Y, Fang SY, Liu ZY, Wang Y, Zhu JH. Trapping thelead ion in multi-component aqueous solution by naturalclinoptilolite. J Hazard Mater 2010;180:282e8.
[20] Armagan B, Ozdemir O, Turan M, Celik MS. The removal ofreactive azo dyes by natural and modified zeolites. J ChemTechnol Biotechnol 2003;78:725e32.
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 en e r g y 3 7 ( 2 0 1 2 ) 1 6 4 2 1e1 6 4 2 9 16429
[21] Zorpas AA, Inglezakis VJ, Stylianou M, Irene V. Sustainabletreatment method of a high concentrated NH3 wastewater byusing natural zeolite in closed-loop fixed bed systems. OpenEnv Sci 2010;4:1e7.
[22] Inglezakis VJ, Hadjiandreou KJ, Loizidou MD,Grigoropoulou HP. Pretreatment of natural clinoptilolite ina laboratory-scale ion exchange packed bed. Wat Res 2001;35:2161e6.
[23] Bayraktaroglu K. Multicomponent ion exchange onclinoptilolite. Turkey: Chemical Engineering Department,MiddleEastTechnicalUniversity; 2006,MasterofScience thesis.
[24] Feldman C. Behavior of trace refractory minerals in thelithium metaborate fusion-acid dissolution procedure. AnalChem 1983;55:2451e3.
[25] Ozturk Y, Yucel M, Daldal F, Mandacı S, Gunduz U, Turker L,et al. Hydrogen production by using Rhodobacter capsulatusmutants with genetically modified electron transfer chains.Int J Hydrogen Energy 2006;31:1545e52.
[26] Biebl H, Pfennig N. Isolation of members of the familyRhodosprillaceae. In: Starr MP, Stolp H, Truper HG, Balows A,Schlegel HG, editors. The prokaryotes, vol. 1. New York:Springer; 1981. p. 267e73.
[27] Androga DD, Ozgur E, Gunduz U, Yucel M, Eroglu I. Factorsaffecting the longterm stability of biomass and hydrogenproductivity in outdoor photofermentation. Int J HydrogenEnergy 2011;36:11369e78.
[28] Jorgensen TC, Weatherley LR. Ammonia removal fromwastewater by ion exchange in the presence of organiccontaminants. Water Res 2003;37:1723e8.
[29] Wang S, Peng Y. Natural zeolites as effective adsorbents inwater andwastewater treatment. Chem Eng J 2010;156:11e24.
[30] Ji ZY, Yuan JS, Li XG. Removal of ammonium fromwastewater using calcium form clinoptilolite. J Hazard Mat2007;141:483e8.
[31] Martinez RC, Miranda VM, Rios MS, Sosa IG. Evaluation ofnatural and surfactant- modified zeolites in the removal ofcadmium from aqueous solutions. Sep Sci Technol 2004;39:2711e30.
[32] Sprynskyy M, Buszewski B, Terzyk AP, Namiesnik J. Study ofthe selection mechanism of heavy metal (Pb2þ, Cu2þ, Ni2þ
and Cd2þ) adsorption on clinoptilolite. J Colloid Interface Sci2006;304:21e8.
[33] Payne KB, Abdel-Fattah TM. Adsorption of divalent lead ionsby zeolites and activated carbon: effects of pH, temperatureand ionic strength. J Environ Sci Health Part A Toxic/HazardSubst Environ Eng 2004;39:2275e91.
[34] Ozgur E, Uyar B, Ozturk Y, Yucel M, Gunduz U, Eroglu I.Biohydrogen production by Rhodobacter capsulatus on acetateat fluctuating temperatures. Resour Conserv Recycl 2010;54:310e4.
[35] Boran E, Ozgur E, van der Burg J, Yucel M, Gunduz U,Eroglu I. Biological hydrogen production by Rhodobactercapsulatus in solar tubular photobioreactor. J Cleaner Prod2010;18:S29e35.
[36] Androga DD, Ozgur E, Gunduz U, Yucel M, Eroglu I.Significance of carbon to nitrogen ratio on the long-termstability of continuous photofermentative hydrogenproduction. Int J Hydrogen Energy 2011;36:15583e94.
[37] Sasikala K, Ramana CV, Rao PR, Kovacs KL. Anoxygenicphototropic bacteria: physiology and advances inhydrogen production technology. Adv Appl Microbiol1993;38:211e95.
top related