Sie sind auf Seite 1von 8

international journal of hydrogen energy 34 (2009) 764771

Available at

journal homepage:

Biohydrogen production by Clostridium butyricum EB6

from palm oil mill effluent

Mei-Ling Chonga, Raha Abdul Rahima, Yoshihito Shiraib, Mohd Ali Hassana,*
Department of Bioprocess Technology, Faculty of Biotechnology and Biomolecular Sciences,
Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia
Graduate School of Life Sciences and System Engineering, Kyushu Institute of Technology,
808-0196 Hibikimo 2-4, Wakamatsu-ku, Kitakyushu-shi, Fukuoka, Japan

article info abstract

Article history: A hydrogen producer was successfully isolated from anaerobic digested palm oil mill
Received 4 October 2007 effluent (POME) sludge. The strain, designated as Clostridium butyricum EB6, efficiently
Received in revised form produced hydrogen concurrently with cell growth. A controlled study was done on
23 June 2008 a synthetic medium at an initial pH value of 6.0 with 10 g/L glucose with the maximum
Accepted 23 October 2008 hydrogen production at 948 mL H2/L-medium and the volumetric hydrogen production rate
Available online 11 December 2008 at 172 mL H2/L-medium/h. The supplementation of yeast extract was shown to have
a significant effect with a maximum hydrogen production of 992 mL H2/L-medium at 4 g/L
Keywords: of yeast extract added. The effect of pH on hydrogen production from POME was investi-
Biohydrogen gated. Experimental results showed that the optimum hydrogen production ability
Clostridium butyricum EB6 occurred at pH 5.5. The maximum hydrogen production and maximum volumetric
Palm oil mill effluent (POME) hydrogen production rate were at 3195 mL H2/L-medium and 1034 mL H2/L-medium/h,
respectively. The hydrogen content in the biogas produced was in the range of 6070%.
2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights

1. Introduction chemical method could be carried out to eliminate the pres-

ence of methanogens as most of the hydrogen producer could
Biohydrogen has been well documented in recent years and produce endospore that survive at harsh environment [3,4].
has generated increased attention from researchers due to its Recently, experiments have been carried out to study the
characteristics as an ideal, clean and sustainable energy possibility of hydrogen production using organic wastes from
resource for the future. In nature, biohydrogen is produced various industries in combination with the wastewater
during acidogenic waste treatment process where acid form- treatment strategy [5,6]. For example, in Malaysia, the palm oil
ing bacteria produces organic acid compound, hydrogen and industry annually generates about 15.2 million tons of
carbon dioxide [1]. However, no hydrogen is bubbling out due wastewater, known as palm oil mill effluent (POME). Due to
to the coexisting bacteria that readily consume hydrogen as the nature of POME, with high cellulose and lignocellulosic
a source of reducing power. Therefore, a specific environment material, it takes a long time to degrade the organic
needs to be created to support the growth of hydrogen substances. Previously, research has been carried out to use
producer and reduce the number of hydrogen consumer [2]. POME sludge as an inoculum, and it has produced a promising
For this aim, pretreatment on sludge sample by heat and level of hydrogen production [7]. However, the cultivation of

* Corresponding author. Tel.: 60 3 89467590; fax: 60 3 89467593.

E-mail address: (M.-L. Chong).
0360-3199/$ see front matter 2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.
international journal of hydrogen energy 34 (2009) 764771 765

a single strain using POME as a substrate has not yet been procedure was repeated three times to ensure the purity of
studied in detail. the strains.
Biohydrogen production can be achieved by light driven
photosynthesis or dark fermentation and both routes have 2.2. Morphological characteristic and scanning electron
been extensively reviewed and reported [8,9]. Dark fermen- microscopy
tation of organic waste materials had presented a promising
route of biohydrogen production compared to photosynthetic Morphological examination was observed by a light micro-
routes [10]. The major advantages of dark fermentation are scope with the cells gram-stained by crystal-violet [19]. Scan-
high rate of cell growth, no light energy required, no oxygen ning electron microscopy (SEM) observations were made as
limitation problems and ability to run on low capital cost [10 follows. Cell pellets were fixed using a standard procedure. The
12]. Fermentative bacteria such as Clostridium sp. (obligate samples were transferred to a specimen basket and were dried
anaerobe) and Enterobacter sp. (facultative anaerobe) are the in a critical dryer for 30 min. The dried samples were mounted
main genera with the ability to produce hydrogen, where on SEM stabs, sputter-coated with gold and photographed with
Clostridium sp. has been recognized as an effective hydrogen a JEOL-6-400 microscope at various magnifications.
producer [13]. Numerous investigations of biohydrogen
production by Clostridium sp. have been reported, which 2.3. 16s rDNA sequencing and phylogenetic analysis
generally gives a yield of 1.42.8 mol H2/mol glucose [1315].
However, due to its sensitivity to minute amounts of dissolved 2.3.1. DNA extraction
oxygen, it is difficult to cultivate a single culture to produce Cells were grown overnight in RCM, 150 rpm at 37  C. Next,
hydrogen and mixed culture is selected as seed inoculum 50 mL of culture were centrifuged at 5000 rpm for 5 min at 4  C,
rather than single culture [16,17]. This is because the presence followed by washing the pellets with sterile distilled water.
of facultative bacteria would create an anaerobic condition for Pellets were resuspended in 10 mL 0.5 M EDTA (pH 8.5) and
the anaerobic bacteria. Therefore, some researches have incubated at room temperature for 10 min. Afterwards, 10 mL
investigated the performance of co-culturing Clostridium sp. lysis buffer (10 mM Tris, 1 mM EDTA pH8, 2 mg/mL lysozyme)
and Enterobacter sp. [18] without addition of reducing agent in was added and incubated at 37  C for 1 h. After addition of 1 mL
culture medium. This would definitely reduce the cost of 10% sodium dodecyl sulphate (SDS) and incubation at 70  C for
hydrogen producing using Clostridium alone. 15 min, the solution became viscous and clear. Equal volumes
In this work, we isolated a hydrogen producer, namely of phenol/chloroform (1:1) was added to the solution and mixed
Clostridium butyricum EB6 from POME sludge. Single strain by inversion before centrifugation at 3000 rpm for 10 min. The
cultivation studies were carried out to investigate the effi- upper (aqueous) phase was transferred to a new clean tube and
ciency of hydrogen gas production. Feasibility studies were treated twice with the phenol/chloroform. One milliliter of 3 M
conducted to examine the possibility of hydrogen gas NaOAc at pH 5.2 was added to the upper phase and followed by
production from raw POME at different pH levels and an equal volume of cold 96% isopropanol. The mixtures were
temperatures. Dynamic relationships of cell growth, hydrogen kept at 20  C for 15 min, followed by centrifugation at
evolution and pH levels were monitored and documented. 10,000 rpm for 15 min. Pellets were rinsed with 4 mL 70%
ethanol by centrifugation at 10,000 rpm for 3 min. After being
air-dried, the pellet was dissolved in 1 mL TrisEDTA buffer and
2. Materials and methods 1 mL RNase (10 mg/mL). The solutions were incubated at 37  C
for 20 min and then the DNA was stored at 4  C.
2.1. Isolation of the bacterial strain from anaerobically
digested POME sludge 2.3.2. PCR amplification, sequencing of 16s rDNA genes and
phylogenetic analysis
The POME sludge from an anaerobic digester pond was 16s rDNA fragments were amplified with a set of universal
collected from Seri Ulu Langat Palm Oil Mill, Dengkil, and primers pA and pH0 , which are specific for universally
Selangor. The POME sludge was pretreated at different conserved bacterial 16s rDNA [20]. The pA is 50 -AGAGTTT-
temperatures; 70  C, 85  C and 100  C, for 10 min. Then 1 mL GATCCTGGCTCAG-30 and pH0 is 50 -AAGGAGGTGATC-
of pretreated sludge was inoculated into 20 mL Reinforced CAGCCGCA-30 . Each PCR mixture (total volume, 25 mL)
Clostrial Medium (RCM) in serum bottle and cultured in consisted of 2.5 mL of 10 PCR buffer, 2.5 mL of MgCl2 (25 mM),
anaerobic condition at 37  C using an anaerobic jar. The RCM 0.5 mL of dNTPs (10 mM for each), 0.5 mL of each primer (10 mM),
was composed of (g/L): meat extract, 10; peptone, 5; yeast 0.2 mL of Taq polymerase (5 U/mL), and 0.5 mL of the above
extract, 3; D glucose, 5; starch, 1; sodium chloride, 5; template and the adequate amount of distilled water added to
sodium acetate, 3; L-cystenium chloride, 0.5; and agar, 0.5. reach the desired volume. A hot-start PCR was performed at
The pH of the medium was adjusted initially to 6.5. After 94  C for 3 min and touchdown PCR was performed as follows.
24 h of cultivation, each sample was serially diluted in The denaturing step of each cycle was carried out at 94  C for
a dilution medium and cultured in RCM with 15 g/L of 1 min, the annealing temperature was set at 58  C, and primer
bacteriological agar in a Hungate tube to isolate a single extension was carried out at 72  C for 1 min. The denaturing-
colony. The dilution medium consists of (g/L): meat extract, primer annealing-primer extension process was carried out
10; peptone, 5; yeast extract, 3; L-cystenium chloride, 0.5; for 29 cycles. The final extension step was carried out at 72  C
and agar, 0.5. Single colonies obtained in agar were cut out for 10 min. The PCR amplification product was purified using
and transferred to RCM and cultured for 24 h. This a QIAgen gel extraction kit. The 16s rDNA PCR products were
766 international journal of hydrogen energy 34 (2009) 764771

sequenced directly by a cycle sequencing system. The nucle- The carrier gas used was nitrogen and the column was packed
otide sequences were determined on both strands. The with Porapak Q (80/100 mesh). The temperatures at the
sequence was compared with reference sequences available stainless column and the injection/ detector point were 50  C
in the NCBI database using the Basic Local Alignment Search and 100  C, respectively. The standard curve of hydrogen was
Tool (BLAST) search. The closest 16s rDNA sequences of plotted using standard hydrogen gas.
reference sequences were retrieved from the database,
aligned and checked manually using the BioEdit version 2.7. Kinetic modeling Phylogenetic trees were constructed using the
neighbor-joining method with MEGA 3.1. Bootstrap re- The cumulated hydrogen production in the batch experi-
sampling analysis for 1000 replicates was performed to esti- ments followed the modified Gompertz equation [21].
mate the confidence of tree topologies.   
Rm e
H P exp  exp l  t 1 (1)
2.4. Hydrogen production in batch fermentation
H is the cumulative hydrogen production (mL), l the lag time
The isolates were pre-cultured on RCM medium, and the late (h), P the hydrogen production potential (mL), Rm the
stationary phase culture was transferred to a 450-mL bioreactor maximum hydrogen production rate (mL/h), and e is
containing 250 mL of media. The media were sparged with 2.718281828. The values of P, Rm and l for each batch were
nitrogen gas for 15 min. The media consist of (g/L): glucose, 20; estimated using the nonlinear estimation function in STA-
KH2PO4, 0.75; K2HPO4, 0.75; MgSO4$7H2O, 0.8; MnSO4$4H2O, 0.2; TISTICA (version 6).
FeSO4$7H2O, 0.03; NaCl, 0.2; Yeast Extract, and 6; L-Cysteine Hydrogen gas production was calculated from bioreactor
HCl, 0.5. The initial pH of the medium was adjusted to 5.5 with headspace measurements of gas composition and the total
1 M NaOH, and the temperature for batch growth was kept at volume of biogas produced at each time interval using the
37  C. The cell concentration, pH level and biogas production followed equation.
were monitored during the course of the experiments. VH;i VH;i1 CH;i VG;i  VG;i1 VH CH;i  CH;i1 (2)

VH,i and VH,i1 are cumulative hydrogen gas volumes at the

2.5. Hydrogen production in batch fermentation using
current (i) and previous (i1) time intervals, VG,i and VG,i1 the
POME as substrate
total biogas volumes in the current and previous time inter-
vals, CH,i and CH,i1 the fraction of hydrogen gas in the head-
One hundred milliliters of the pre-cultured C. butyricum EB6
space of the bottle measured using gas chromatography in the
was inoculated in a 3-L fermenter (B. Braun, Germany) with 1 L
current and previous intervals, and VH the total volume of
of raw POME. Raw POME was collected from the same palm oil
headspace in the reactor [22].
mill (Table 1). To study the effect of pH on hydrogen produc-
tion, the pH value was controlled at 5.0, 5.5, 6.5, 7.5 and 8.5 by
automatic addition of 1 M NaOH at 37  C. The initial anaerobic
condition was established by sparging oxygen free nitrogen
3. Results and discussion
for 20 min. In another set of experiments, the pH value was
In all the experiments, the biogas produced contained
controlled at 5.5, with the temperature controlled at 30  C,
hydrogen (6070%) and carbon dioxide (3040%). No methane
37  C and 55  C. Exit gas was collected in a gas bag.
gas was detected. All the cumulative hydrogen production
data fit Eq. (1) well with R2 > 0.99.
2.6. Analytical method

The exit gas was analyzed using a Shimadzu GC-8A gas

chromatography (GC) with a thermal conductivity detector.

Table 1 Characteristics of POME.

Item Unit Value

Physical characteristics
Moisture content % 93.4
Ash % 1.06
Total solid g 6580
Fiber (soluble) % SS 1.12
Fiber (non-soluble) % TS 3.9

Chemical property
COD mg/L 15,000100,000
BOD5 mg/L 10,25043,750
pH 3.85.0
Protein % TS 12.31
Fig. 1 Scanning electron microscopy photo of Clostridium
Fat % TS 12.95
butyricum EB6, growing in POME at pH 6.5 and 37 8C.
international journal of hydrogen energy 34 (2009) 764771 767

C. butyricum CGS6 (AY540110)

C. butyricum CGS2 (AY540106)
C. butyricum CGS3 (AY540107)
C. butyricum CGS1 (AY540105)
68 C. butyricum CGS4 (AY540108)
C. butyricum CGS5 (AY540109)
78 EB6
C. butyricum ATCC43755 (X68176)

96 83 C. butyricum NCIMB8082 (X68178)

88 C. butyricum DSM2478 (X68177)
C. saccharobutylicum P262 (U16147)
C. beijerinckii NCIMB9362 (X68180)
100 90
C. acetobutylicum (X81021)
Clostridium sp. B902-1b (AB114264)
C. diolis (AJ458417)
C. paraputrificum DSM 2630 (X73445)

100 C. botulinum (X68317)

C. argentinense (X68316)
C. acidisoli (AJ237756)
C. roseum (Y18172)

33 50 Clostridium sp. 44a-T5zd (AY082483.1)

51 Clostridium sp. BL-22 (DQ196626.1)

100 35 Clostridium sp. HPB-46 (AY862516.1)

Clostridium sp. HPB-21 (AY862509.1)
69 EB7
Clostridium sp. HPB-25 (AY862510.1)
Clostridium sp. HPB-2 (AY862514.1)
30 Clostridium sp. HPB-16 (AY862512.1)
Clostridium sp. HPB-26 (AY862511.1)
47 Clostridium sp. HPB-1 (AY862515.1)
53 Clostridium sp. HPB-4 (AY862513.1)
Bacillus subtilis (AB301025)


Fig. 2 Phylogenetic tree of the two biohydrogen-producing strains and their close relatives, based on the almost fully
sequenced 16s rDNA, was constructed. The tree, based on JukesCantor distance, was constructed using a neighbor-joining
method with 1000 bootstrappings. Bacillus soli was selected as an outgroup species. The scale bar represents 0.02
substitutions per nucleotide position. Bootstrap values are indicated at the nodes. Reference sequences in the dendogram
were obtained from NCBI (accession number in parentheses).

3.1. Characterization of isolated hydrogen producer meanwhile the highest similarity of EB7 was obtained
towards Clostridium sp. 44a-T5zd (98%) as compared to the
Two anaerobic hydrogen producers, namely EB6 and EB7, sequences in the NCBI database. The phylogenetic tree was
were successfully isolated from POME anaerobic sludge. Both constructed for both of the strains and those known Clos-
strains were shown to be rod shaped (Fig. 1) bacteria tridium species (Fig. 2). EB6 fell in the cluster with the known
(examined under light microscope, SEM) and endospore C. butyricum species while EB7 fell in the cluster of Clostridium
forming (SchaefferFulton Method). The 16s rDNA gene for sp. Based on the physiological analysis and 16s rDNA
both strains was sequenced and deposited in the NCBI sequence comparison, strains EB6 should belong to C. butyr-
database with the accession number of EU183474 (EB6) and icum. C. butyricum, strictly anaerobic bacteria, was well
EU163273 (EB7). Based on the sequence, both strains were studied as a hydrogen [13] and butyric acid producer [23]. A
highly similar to Clostridium genus with at least 80% identity, preliminary study was carried out on the hydrogen produc-
however they were different species. The highest similarity tion potential of EB6 and EB7. It was found that EB6 has
of EB6 was obtained towards C. butyricum CGS5 (99%), a higher hydrogen production rate than EB7 (data not shown).
768 international journal of hydrogen energy 34 (2009) 764771


Hydrogen production rate (ml/h)

70 14
a b

Cell concentration (g-DCW)

Glucose concentration (g/l)
60 13 2.2

50 12 DCW 2.0
40 11
30 10
20 9
glucose consumption
10 8 1.4

0 7 1.2

400 5.4
c d
Cumulative gas (ml)

300 Biogas production

200 4.8
150 Hydrogen production
0 4.2
0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22
Time (hour) Time (hour)

Fig. 3 Batch hydrogen fermentation of Clostridium butyricum EB6 with the 6 g/L of yeast extract as a nitrogen source with
10 g/L of glucose as a carbon source. (a) Profile of hydrogen production rate (mL/h), (b) profile of cell growth and glucose
consumption, (c) profile of hydrogen gas production and total biogas production, and (d) profile of pH changes.

For this reason, EB6, designated as C. butyricum EB6, was inhibit hydrogen formation. This occurred when nonpolar
selected for further study. undissociated acid penetrated the cell membrane, and disso-
ciated due to the higher intracellular pH, leaving more protons
3.2. Hydrogen production in synthetic medium within the cell. This caused an increase in the energy
requirement that involved coenzyme A and the phosphate
C. butyricum is a known effective hydrogen producer [18]. pools for maintaining neutrality. As an immediate effect, the
However, the information on the culture conditions, essential flux of glucose through glycolysis was reduced, which was
minerals and the inter-relationship of the growth to hydrogen reflected by the consumption of glucose by almost 50% (Fig. 3b
gas production was limited. Fig. 2 shows the profile of and d). Cell and hydrogen production reached a maximum
hydrogen production during batch fermentation of C. butyr- (when the pH level decreased), even though the remaining
icum EB6 on synthetic medium supplemented with iron, glucose was still high. Therefore, it is suggested that pH
magnesium and manganese iron. The gas production began
simultaneously with the cell growth, with the maximum
Total accumulated hydrogen gas (ml)

hydrogen production at 948 mL H2/L-med. This suggests that

the hydrogen production is a primary metabolite that is 240
correlated with the growth of the cell. This phenomenon is 220
explained as, during the primary metabolism, the excess 200
proton was reduced to molecular hydrogen to dispose of the 180
reducing equivalent. In other words, during oxidation of the 160
organic substrate for cell building blocks and metabolic 140
energy, it generates electrons that need to be disposed of to
maintain the electrical neutrality. Hydrogen started to evolve
almost immediately after inoculation and reached 60
a maximum level when the cell entered the stationary phase. 40
The lag phase was 0.6 h and the maximal hydrogen produc- 20
tion rate was 172 mL/h/L, occurring at 10 h (Fig. 3ac). As pH 0
0 2 4 6 8 10 12 14 16 18 20 22 24 26
dropped from 5.2 to 4.4, the hydrogen production rate
Time (hr)
decreased from 172 mL H2/h/L to 20 mL H2/h/L. The pH level
decreased drastically immediately after the cells started to Fig. 4 Cumulative hydrogen production at different ratios
accumulate, suggesting the accumulation of acidic metabo- of yeast extract to ammonium sulphate. B 0:6, , 2:4,
lites [24]. However, the undissociated form of acid is found to > 4:2, and O 6:0.
international journal of hydrogen energy 34 (2009) 764771 769

3500 kinetic parameters decreased. The effect of yeast extract was

Total accumulated hydrogen gas (ml)

found to be significant as the maximum hydrogen production

3000 increased simultaneously with the increase in yeast extract. It
is suggested that the nitrogen content was important for the
building blocks of the cell, and that yeast extract, as an organic
nitrogen source, supplied other trace minerals that might
have a combined effect on hydrogen production.
3.4. Effect of pH on hydrogen production in POME
Fig. 5 illustrates the plots of hydrogen production fitted to the
500 modified Gompertz equation for pH values ranging from 5.0 to
8.5, treating 100,000 mgCOD/L of raw POME with C. butyricum
0 EB6. A controlled experiment was carried out with no pH
0 2 4 6 8 10 12 14 16 18 20 22 24 26
Time (hour) adjustment being performed. This allowed an investigation of
the effect of undissociated acid accumulated in the culture
Fig. 5 Cumulative hydrogen production at different pH broth. The pH range was selected based on a literature search
values, ranging from pH 5.5 to pH 8.5. B pH 5.5, , pH 6.5, [25,26]. The initial pH of POME was found to be in the range of
> pH 7.5, and O pH 8.5. 3.85.0. Therefore, pH adjustment was done initially using 1 M
NaOH before sterilization. The result shows that hydrogen
production was strongly inhibited by pH values as low as pH
controlled at an optimal level may be required to improve 5.0. This finding differed from previously published results
hydrogen production. where hydrogen production was maximized at a pH value as
low as pH 4.5 with mixed microflora. This suggests that
3.3. Effect of yeast extract concentration growing C. butyricum EB6 was almost impossible in raw POME
at pH lower than 5.0. The highest total hydrogen was obtained
The effect of nitrogen on hydrogen production was not well at 3195 mL H2/L-POME when culturing C. butyricum EB6 in pH
studied previously, and information was limited. However, 5.5 controlled raw POME, with the volumetric hydrogen
hydrogen production was growth associated; therefore, it is production rate at 1034 mL H2/L-POME/h at 24 h (Table 2). This
hypothesized that the nitrogen source would have a signifi- result agrees with those obtained by Atif et al. [7]. POME sludge
cant effect on hydrogen production. Fig. 4 depicts the cumu- produced a yield of 4708 mL H2/L-POME in 38 h at pH 5.5, with
lative hydrogen production at different ratios of yeast extract maximum volumetric hydrogen production at 454 mL H2/L-
to ammonium sulphate. The results show that the highest POME/h. Table 2 shows the hydrogen production was greatly
maximum hydrogen production was achieved by 4 g/L of yeast affected by pH level, varying from as low as 5.58.5. Maximum
extract and 2 g/L of ammonium sulphate. The kinetic param- hydrogen production and volumetric hydrogen production
eters were: P, 248 mL; Rm, 37.3 mL/h; l, 0.6 h. As the yeast rate decrease as the pH increases in the sequence of 5.5, 6.5,
extract to ammonium sulphate reduced to 2:4 and 0:6, the and 7.58.5. Maximum hydrogen productions for pH 6.5, 7.5

Table 2 Biohydrogen production by Clostridium butyricum EB6, using raw POME as sole substrate.
pH Temp. FeSO4$6H2O H2 CO2 Total hydrogen Volumetric H2 production Hydrogen yield
( C) (g/L) content content production (mL) rate (mL/h/L) (mL H2/g COD)
(%) (%)
Overalla Maximumb Overallc Maximumb

5 37  C None 0 0 0 0 0 0 0
5.5 62 38 3345 3195 278.8 1034.7 31.95
6.5 66 34 3062 3022 251.8 790.8 30.2
7.5 70 30 1929 1959 160.8 735.1 19.6
8.5 68 32 587 618 41.3 201.1 6.18
Uncontrolled 65 35 2249 2253 112.5 296.4 22.53

5.5 30  C None 64 36 2795 2858 116.5 314.8 26.38

37  C 62 38 3345 3195 278.8 1034.7 31.95
55  C 0 0 0 0 0 0 0

5.5 37  C None 62 38 3345 3195 278.8 1034.7 31.95

0.25 56 44 2648 2638 189.1 498.9 22.53

a Overall hydrogen production at i time.

b Maximum hydrogen production at i time and maximum volumetric production rate calculated based on the modified Gompertz equation.
c Overall volumetric hydrogen production rate calculated by dividing the maximum cumulative hydrogen production (Vi) over by the time
required to reach a maximum.
770 international journal of hydrogen energy 34 (2009) 764771

Table 3 Comparison of the hydrogen production obtained in this study to those cited in the literature.
Microorganism Substrate pH Temperature H2 yield Reference

Mixed culture POME 5.5 60 4708 mL H2/L-POME [7]

Mixed culture Cellulose 7.0 60 3325 mL H2/L-med [27]
Mixed culture Starch 6.0a 55 92 mL H2/g carbohydrate [28]
C. thermolacticum Lactose 7.0 58 2.48 mmol H2/mol lactose/h [29]
C. butyricum CGS5 Sucrose 5.5 37 3088 mL H2/L-med [13]
C. butyricum EB6 Glucose 6.0a 37 948 mL H2/L-med This study
C. butyricum EB6 POME 5.5 37 3345 mL H2/L-POME This study
Mixed culture Rice slurry 4.5 37 346 mL H2/g carbohydrate [30]

a Initial pH.

and 8.5 were 3022 mL H2/L-POME, 1959 mL H2/L-POME and will cause reduction in hydrogen production [31]. An experi-
618 mL H2/L-POME, respectively. The lag phase of hydrogen ment was carried out with 0.25 g/L FeSO4$6H2O to prelimi-
production by C. butyricum EB6 was found to be in the range of narily investigate its effect. When the ambient temperature is
1.86 h, with pH 8.5 as the outlier. The lag phase was found to relatively low, the bacteria need more ferrous ions to activate
be much lower than the published data at different fermen- the hydrogenase so that it can oxidize reduced ferredoxin to
tation condition [7,13]. The discrepancy may be due to the produce more molecular hydrogen and vice versa. It was found
rapid hydrogen production by single culture and acclimati- that the maximum hydrogen production with an iron supple-
zation of culture before inoculation. The comparison of ment was 2638 mL H2/L-POME, compared with no supplement
hydrogen gas production was tabulated in Table 3. at 3195 mL H2/L-POME. It is suggested that the addition of iron
does not increase hydrogen production at 37  C. This result
3.5. Effect of temperature on hydrogen production at pH was not in line with the published data that increasing FeCl2 up
5.5 in POME to 800 mg/L would increase hydrogen production [32]. This
might be because iron is present in POME that additional of it is
Fig. 6 illustrates the plot of hydrogen production fitted to the in excess and no effect on hydrogen production.
modified Gompertz equation at temperatures 30  C and 37  C
at pH 5.5. Duplicate experiments were carried out on the
hydrogen gas production at 55  C. However, there was no gas 4. Conclusion
produced at this temperature, showing that C. butyricum EB6
could not grow and produce gas under thermophilic condi- C. butyricum EB6, with the ability to produce hydrogen, was
tions. The kinetic parameters at 30  C were P, 2858 mL; Rm, isolated from anaerobic treated POME sludge. This bacterium
314.8 mL/h; l, 7.02 h; and at 37  C they were P, 3195 mL; Rm, was characterized as gram positive, rod shaped bacteria, with
1034.6 mL/h; and l, 3.86 h. The lag phase for the experiment at the ability to form endospores. Obtained results show that C.
30  C was longer than at 37  C. It is therefore suggested that the butyricum EB6 produced more hydrogen from medium sup-
bacteria need a longer time to adapt to the lower culture plemented with 4 g/L of yeast extract. Experiments were
temperature. The maximum hydrogen production and volu- carried out to examine the feasibility of hydrogen production
metric hydrogen production rates were higher at 37  C. This from POME. Experimental results show that cell growth and
experiment results are comparable with the published data as hydrogen production were inhibited at pH 5.0, while the
during mesophilic condition, further lowering temperature highest hydrogen production was found at pH 5.5, 37  C with
no iron supplemented. The lag phase was found to be much
3500 shorter than the published data at different fermentation
Total accumulated hydrogen gas (ml)

conditions, suggesting that the hydrogen production from C.

3000 37C butyricum EB6 was associated with the growth of cell.

2500 30C

2000 Acknowledgements

1500 The authors gratefully thank the financial support by the

Ministry of Science, Technology and Innovation, Malaysia,
1000 University Putra Malaysia, Kyushu Institute of Technology
(KIT) and Japan Society for Promotion of Science (JSPS).

0 references
0 2 4 6 8 10 12 14 16 18 20 22 24 26
Time (hour)

Fig. 6 Cumulative hydrogen production at temperatures [1] Angenent LT, Karim K, Al-Dahhan MH, Wrenn BA,
B 30 8C and , 37 8C. Domiguez-Espinosa R. Production of bioenergy and
international journal of hydrogen energy 34 (2009) 764771 771

biochemicals from industrial and agricultural wastewater. [16] Kim SH, Han SK, Shin HS. Feasibility of biohydrogen
Trends Biotechnol 2004;22:47785. production by anaerobic co-digestion of food waste and
[2] De Vrije E, Claassen PAM. Dark hydrogen fermentation. In: sewage sludge. Int J Hydrogen Energy 2004;29:160716.
Reith JH, Wijffels RH, Barten H, editors. Status and [17] Van Ginkel SW, Oh SE, Logan BE. Biohydrogen gas production
perspectives of biological methane and hydrogen from food processing and domestic wastewaters. Int J
production. Dutch Biological Hydrogen Foundation; 2003. p. Hydrogen Energy 2005;30:153542.
10323. [18] Yokoi H, Tokushige T, Hirose J, Hayashi S, Takasaki Y. H2
[3] Cheng CC, Lin CY, Lin MC. Acidbase enrichment enhances production from starch by a mixed culture of Clostridium
anaerobic hydrogen production process. Appl Microbiol butyricum and Enterobacter aerogenes. Biotechnol Lett 1998;20:
Biotechnol 2002;58:2248. 1437.
[4] Zhang YF, Liu GZ, Sheng JQ. Hydrogen production in batch [19] Bergey DH, Holt JG, Krieg NR, Sneath PHA. Bergeys manual of
culture of mixed bacteria with sucrose under different iron determinative bacteriology. 9th ed. Lippincott Williams &
concentrations. Int J Hydrogen Energy 2005;30:85560. Wikins; 1994.
[5] Yu HQ, Zhu ZH, Hu WR, Zhang HS. Hydrogen production [20] Hales BA, Edwards C, Ritchie D, Hall G, Pickup RW,
from rice winery wastewater in an upflow anaerobic reactor Saunders JR. Isolation and identification of methanogen
by using mixed anaerobic cultures. Int J Hydrogen Energy specific DNA from blanket bog peat by PCR amplification and
2002;27:135965. sequence analysis. Appl Environ Microbiol 1996;62:66875.
[6] Han SK, Kim SH, Shin, Hang SS. UASB treatment of [21] Lay JJ, Li YY, Noike T. The influences of pH and moisture
wastewater with VFA and alcohol generated during content on the methane production in high solids sludge
hydrogen fermentation of food waste. Process Biochem 2005; digestion. Water Res 1997;31:151824.
40:2897905. [22] Logan BE, Oh SE, Van Ginkel SW. Biological hydrogen
[7] Atif AAY, Fakhrua-Razi A, Ngan MA, Morimoto M, Iyuke SE, production measured in batch anaerobic respirometers.
Vesiroglu NT. Fed batch production of hydrogen from palm Environ Sci Technol 2002;36:25305.
oil mill effluent using anaerobic microflora. Int J Hydrogen [23] He GQ, Kong Q, Chen QH, Ruan H. Batch and fed-batch
Energy 2005;30:13937. production of butyric acid by Clostridium butyricum ZJUCB. J
[8] Manish S, Banerjee R. Comparison of biohydrogen Zhejiang University Sci 2005;6B:107680.
production processes. Int J Hydrogen Energy 2008;33:27986. [24] Ginkel VS, Logan B. Inhibition of biohydrogen production by
[9] Tao YZ, Chen Y, Wu YQ, He YL, Zhou ZH. High hydrogen undissociated acetic and butyric acid. Environ Sci Technol
yield from a two-step process of dark and 2005;39:93516.
photo-fermentation of sucrose. Int J Hydrogen Energy [25] Fang HHP, Liu H. Effect of pH on hydrogen production from
2007;32:2006. glucose by a mixed culture. Bioresour Technol 2002;82:8793.
[10] Levin DB, Pitt L, Love M. Biohydrogen production: prospects [26] Khanal SK, Chen WH, Li L, Sung SW. Biological hydrogen
and limitations to practical application. Int J Hydrogen production: effects of pH and intermediate. Int J Hydrogen
Energy 2004;29:17385. Energy 2004;29:112331.
[11] Nath K, Das D. Improvement of fermentative hydrogen [27] Ueno Y, Kawai T, Sato S, Otska S, Morimoto M. Biological
production: various approaches. Appl Microbiol Biotechnol production of hydrogen from cellulose by mixed anaerobic
2004;65:5209. microflora. J Fermen Bioeng 1995;79:3957.
[12] Hallenbeck PC, Benemann JR. Biological hydrogen [28] Zhang T, Liu H, Fang HHP. Biohydrogen production from
production: fundamentals and limiting processes. Int starch in wastewater under thermophilic condition. J
J Hydrogen Energy 2002;27:118593. Environ Manage 2003;69:14956.
[13] Chen WM, Tseng ZJ, Lee KS, Chang JS. Fermentative [29] Collet C, Adler N, Schwitzgu1ebel JP, P1eringer P. Hydrogen
hydrogen production with Clostridium butyricum CGS5 production by Clostridium thermolacticum during continuous
isolated from anaerobic sewage sludge. Int J Hydrogen fermentation of lactose. Int J Hydrogen Energy 2004;29:
Energy 2005;30:106370. 147985.
[14] Lin PY, Whang LM, Wu YR, Ren WJ, Hsiao CJ, Li SL, et al. [30] Fang HHP, Li CL, Zhang T. Acidophilic biohydrogen production
Biological hydrogen production of the genus Clostridium: from rice slurry. Int J Hydrogen Energy 2006;31:68392.
metabolic study and mathematical model simulation. Int [31] Zhang YF, Shen JQ. Effect of temperature and iron
J Hydrogen Energy 2007;32:172835. concentration on the growth and hydrogen production of
[15] Levin DB, Islam R, Cicek N, Sparling R. Hydrogen production mixed bacteria. Int J Hydrogen Energy 2006;31:4416.
by Clostridium thermocellum 27405 from cellulosic biomass [32] Lee YL, Miyahara T, Noike T. Effect of iron concentration on
substrates. Int J Hydrogen Energy 2006;31:1496503. hydrogen fermentation. Bioresour Technol 2001;80:22731.