Fermentative hydrogen production by new marine Clostridium amygdalinum strain C9 isolated from offshore crude oil pipeline

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 5 ( 2 0 1 0 ) 6 6 6 5e6 6 7 3

Avai lab le a t www.sc iencedi rec t .com

journa l homepage : www.e lsev ier . com/ loca te /he

Fermentative hydrogen production by new marine Clostridiumamygdalinum strain C9 isolated from offshore crudeoil pipeline

H.S. Jayasinghearachchi a, Sneha Singh a, Priyangshu M. Sarma a, Anil Aginihotri b,Banwari Lal a,*aEnvironmental and Industrial Biotechnology Division, The Energy and Resource Institute, Darbari Seth Block, Habitat Place,

Lodhi Road, New Delhi, 110 003, IndiabCorporate HSF, Oil and Natural Gas Corporation, New Delhi, India

a r t i c l e i n f o

Article history:

Received 22 January 2010

Received in revised form

6 April 2010

Accepted 8 April 2010

Available online 21 May 2010

Keywords:

Biohydrogen

Clostridium amygdalinum

Pentoses

Starch

* Corresponding author. Tel.: þ91 11 2468210E-mail address: [email protected] (B. L

0360-3199/$ e see front matter ª 2010 Profedoi:10.1016/j.ijhydene.2010.04.034

a b s t r a c t

The present study investigated hydrogen production potential of novel marine Clostridium

amygdalinum strain C9 isolated from oil water mixtures. Batch fermentations were carried

out to determine the optimal conditions for the maximum hydrogen production on xylan,

xylose, arabinose and starch. Maximum hydrogen production was pH and substrate

dependant. The strain C9 favored optimum pH 7.5 (40 mmol H2/g xylan) from xylan, pH

7.5e8.5 from xylose (2.2e2.5 mol H2/mol xylose), pH 8.5 from arabinose (1.78 mol H2/mol

arabinose) and pH 7.5 from starch (390 ml H2/g starch). But the strain C9 exhibited mixed

type fermentation was exhibited during xylose fermentation. NaCl is required for the

growth and hydrogen production. Distribution of volatile fatty acids was initial pH

dependant and substrate dependant. Optimum NaCl requirement for maximum hydrogen

production is substrate dependant (10 g NaCl/L for xylose and arabinose, and 7.5 g NaCl/L

for xylan and starch).

ª 2010 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.

1. Introduction considering its high evolution rate in the absence of any light

Renewable energy resources have received considerable

attention due to the depletion of fossil fuel and environmental

pollution [1]. Among which, hydrogen is considered to be an

ideal energy carrier with a high energy content of 122 KJ/g as it

produces only water when it is combusted as a fuel or con-

verted to electricity [1].

Hydrogen can be obtained via non-biological and biological

processes. However, biological hydrogen production

processes are friendlier to environment and less energy

intensive than chemical and electrochemical processes.

Further, dark fermentation is a more promising method

0; fax: þ91 11 24682144.al).ssor T. Nejat Veziroglu. P

source and versatility of the substrates used than photosyn-

thetic hydrogen production [2].

At present, the dominant cost element in fermentative

hydrogen production is the substrate [3]. Therefore, it is

necessary to find low cost feedstock for commercial purposes.

Renewable energy sources, such as cellulose, lignocelluloses

or starch containing biomass constitute an abundant, inex-

pensive and reliable raw material for biohydrogen production

and offer considerable advantages [4,5]. Further, xylan, the

major portion of hemicellulose of plant cell walls, and is the

second most abundant renewable hemicellulosic poly-

saccharide after cellulose [6]. Therefore, fermentative H2

ublished 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 5 ( 2 0 1 0 ) 6 6 6 5e6 6 7 36666

production from cellulosic feedstock or from lignocellulosic

wastes could be competitive with fossil fuel-derived H2,

providing a plausible approach to practical biohydrogen

production in the future [7].

Further, it is interestingly noted that any studies have been

conducted to screen hydrogen producing bacteria from fresh

water fields. But few studies have been identified salt-tolerant

strains from fresh water fields. Hydrogen producing bacterial

strains can be directly screened from marine environments,

but few such bacteria can be cultivated in the laboratory.

Therefore, aim of the present study was to report moderately

halophilic novel marine strain of Clostridium amygdalinum

strain C9 which can effectively utilize xylan, xylose and arab-

inose to produce hydrogen under mesophilic condition, iso-

lated fromoil watermixtures. Hence, the presentworkmainly

focuses on the optimization of the efficiency of fermentative

hydrogen production under different pH, temperature, NaCl

and different substrate concentrations.

2. Methods

2.1. Oil water samples collection

Oil water mixture samples (97% oil, 3% water) were collected

from offshore sea buried crude oil pipelines in Bombay High,

India into 50 ml of anaerobic sterilized serum bottles con-

taining 1 ml of 2% Na2S. Samples were collected during

December, 2007eJanuary, 2008.

2.2. Medium preparation

PeptoneYeast (PY) liquidmediumcomposedof (g/L);Peptone20,

yeast extract 5,NaCl 5,Na2S2O3 0.03, ferric citrate 0.03,was used

for the isolation of the strain C9 Cysteine-hydro chloride was

added to a final concentration of 300 mg/L. The pH of the

medium was adjusted to 7.5 with 1N KOH at an ambient

temperature before autoclaving. The PYB medium was boiled

under a stream of O2-free N2 gas and aliquots of 6 ml were

anaerobically dispensed into 10 ml serum bottles. The bottles

were then sealed and autoclaved for 20 min at 120 �C. Prior toinoculation, 0.1ml of 2%filter sterilizedNa2S.9H2O and 0.1ml of

10% filter sterilized Na2CO3were injected separately by syringe.

2.3. Isolation of the strain C9

Samples (0.1 ml) were inoculated in liquid PY medium. All the

culture bottleswere incubated at 40 �C for 3e5 days depending

on the growth. Sub culturing was done in the same medium

after 3 days of inoculation. All the sub cultures were incubated

at 40 �C under atmospheric pressure.

The sub cultures were serially diluted and 10 mL of serially

diluted cultures (10�7 and 10�8 dilutions) were plated on

anaerobically prepared PY solid medium that was solidified

with 12% agar. Culture plates were incubated at 40 �C in the

anaerobic chamber (N2:H2:CO2; 5:5:90) for 2e4 days till colo-

nies appeared [8]. Culture plateswere observed daily andwell-

isolated colonies were transferred to the liquid PYP medium

with the help of sterile loop. Liquid cultures were incubated at

40 �C as described above. Gram staining was routinely

performed to check the purity of the cultures. Cells were

observed under a light microscope (Olympus, Japan) and pure

strain C9 was routinely cultivated in anaerobically prepared

Trypticase Soy Broth (TSB).

2.4. Strain identification

GenomicDNAfromthepure isolateC9wasextractedaccording

to the previously described procedures [9]. The 16S rRNA gene

of the isolatewas amplifiedby PCRusingMicroSec full GeneKit

(Applied Biosystems, UK) as per the manufactures instruc-

tions. Amplified product was sequenced using the DyeDeoxy

Terminator Cycle sequencing Kit (Applied Biosystems, UK) as

directed in the manufacturer’s protocol with an automatic

DNA sequencer (Model 300; Applied Biosystems, USA).

The sequence of the strain C9 was compared with other

related sequences available in GenBank using BLAST pro-

gramme [10]. Further, the sequence of strain C9 was aligned

with closely related sequences found in GenBank, using

CLUSTAL W, and pair wise evolutionary distances were

computed using Jukes-Canter Model [11]. Phylogenetic anal-

ysis was performed using Phylip version 3.67 [12] and

a phylogenetic tree was constructed using the neighbour-

joining algorithm [13]. Confidence in the tree topologies was

evaluated by re-sampling 100 bootstrap trees [14].

2.5. Study of the growth properties of the strain C9

The strain was cultured in liquid basal anaerobic medium

composed of (g/L): NH4Cl2 0.5; Yeast 5; K2HPO4 0.25; KCl 0.002;

MgCl2 .6H2O 0.125; NH4HCO3 0.4; Peptone 1; NH4H2PO4 0.4;

NaH2PO4 0.5 and 1ml of trace element solution. Trace element

solution was composed of (mg/L): CaCl2 50; MgCl2.6H2O 100;

FeCl2 25; NaCl 10; CoCl2.6H2O 5; MnCl2.4H2O 5; AlCl3 2.5;

(NH4)6MO7O25 15; H3PO4 5; NiCl2 5; CuCl2.5H2O 5; ZnCl2 5.

Temperature profile (35, 40, 45 and 55 �C) and NaCl require-

ment (0.0e7%) (w/v) for the growth of the strain C9 was

studied in anaerobically prepared liquid basal medium sup-

plementedwith 5 g/L glucose as carbon source. Further, ability

of utilization of different carbon sources for the growth of the

strain C9 was tested in liquid basal medium described above.

Following carbon sources were used at a final concentration of

5 g/L: sucrose, fructose, xylose, xylan, ribose, raffinose, ram-

nose, arabinose, xylan, rhamnose, galactose, cellobiose,

starch and cellulose. All the growth experiments were per-

formed in triplicate, using 60 ml serum bottles containing

10 ml of anaerobic basal medium.

2.5.1. Optimization of hydrogen production on differentcarbon sources in batch fermentationBatch dark fermentation experiments were carried out in 1 L

anaerobic bioreactor with 50 ml of liquid anaerobic basal

medium. Reducing agent such as Na2S or L-cysteine hydro-

chloride was not used during this study. Pre-grown inoculum

acquired during exponential growth phase was added at 5%

(v/v) using air tight sterilized syringe.

Influence of different initial pH levels for the hydrogen

production on different substrates (xylan, xylose, starch and

arabinose) was studied. The initial pH levels ranging from 5.5

to 9.5 with 0.5 increments were selected for this study. The pH

Fig. 1 e Neighbour-joining tree showing the phylogenetic

position of isolated strain, Clostridium spp. C9, based on the

16S rRNA sequences of the genus Clostridium. The numbers

at the nodes indicate the levels of bootstrap percentages

based on the neighbour-joining of 100 replicates. Bar

indicates the nucleotide substitution per site. GenBank

accession numbers are given in parenthesis.

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 5 ( 2 0 1 0 ) 6 6 6 5e6 6 7 3 6667

of the mediumwas adjusted with 1M HCl or 1M NaOH prior to

dispensing the medium into the 1 L anaerobic bioreactors.

Influence of different NaCl concentration on growth and

hydrogen production was also studied. Different NaCl

concentrations testedwere 2.5, 7.5, 10, 20 and 25, 35 and 45 g/L.

Further, effect of initial substrate concentration on

hydrogenproductionwas investigated inbatchexperimentsas

described above. Optimal pH levels exhibited by the strain C9

for themaximumhydrogenproduction ondifferent substrates

wereusedas initial culturepH. Initial substrate concentrations

testedwere5, 10, 15, 20and25g/L for each individual substrate.

Distribution of secondary metabolites during fermentation

on different substrates at different pH levels studied. Further,

influenceof initial substrateconcentrationonthedistributionof

various secondary metabolites was also analysed in this study.

2.6. Invivo hydrogen evolution assay undermicroaerophilic conditions

In vivo hydrogen evolution assay under microaerophilic

condition was measured by using a gas chromatograph as

described below. Methylviolegen reduced by sodium dithion-

ite was used as an electron donor. Ten millilitres of serum

vials with butyl rubber stoppers were used for the assay. The

assay mixture contains: 1080e1380 mL of 100 mM potassium

phosphate (K2HPO4/KH2PO4) pH 6.8e7; 200 mL of 1 M sodium

dithionite; 200 mL of 10% Triton X-100; 200e500 mL of anaero-

bically grown culture (OD600nm 1.2). The reaction vessel was

afterwards closed up with the rubber stopper and crimpled

with aluminium caps. Then the serum bottles were removed

from the anaerobic chamber. The sealed bottles were vortex

for 10 s. Different amounts of air (1e4 ml) were added to the

reaction vessel with a syringe and incubated reaction vessels

in a water bath at 37 �C for 40 min. After incubation, 500 mL of

headspace gas samples were analysed. The assay was con-

ducted in triplicate.

2.7. Analytical methods

Bacterial cell mass in each individual bioreactor was deter-

mined by monitoring optical absorbance at OD600 nm in the

Hitachi U-2000 spectrophotometer (Tokyo, Japan) [20].

The headspace gas composition in each bioreactor was

directly analysed by the gas chromatography (GC 6890N, Agi-

lent, USA) equipped with thermal conductivity detector and

HP PLOTQ column (15M� 530 mm� 40 mm) as described in our

previous study [15].

Secondary metabolic products (Acetate, butyrate, iso-

butrate, valerate, isovalerate and propionate) in liquid phase

were measured with GC 6890N (Agilent, USA) equipped with

flame ionization detector and DB-WAXetr column

(30 m � 530 mm � 1 mm). High Performance Liquid Chroma-

tography (HPLC, Agilent 1100 series, USA) equipped with

Sugar-PAK.1 column (Water Research, USA) was used for the

detection of remaining sugars and ethanol production [15].

Further HPLC (Agilent 1100 series, USA) equipped with Ani-

nexR HPX-87H, (300mm� 7.8mm) column (BIO RAD, CA, USA)

was used to quantify succinate, formate and oleic acid

production. Sulphuric acid (0.0008 N) was used as the mobile

phase at a flow rate of 0.6 ml/min.

3. Results and discussion

3.1. General properties of the strain C9

The strain C9 was a facultative anaerobe and able to grow

under microaerophilic condition. The strain C9 grew from 30

to 42 �C with 37 �C for optimum growth. Weaker growth was

observed at 45 �C and no growth was observed beyond 45 �C.The strain C9 was moderately halophilic and able to grow up

to 6% NaCl (w/v) with optimum of 2e2.5% NaCl (w/v) in the

presence of glucose as a sole carbon source. The strain C9

could utilize glucose, sucrose, fructose, arabinose, xylan,

xylose, starch and ribose, raffinose, rhamose, cellobiose and

cellulose. The stain C9 could not utilized pectin.

3.2. Identification of the strain C9 based on the 16SrRNA gene sequence analysis

The nucleotide sequence of 16S rRNA gene of the strain C9was

deposited in the GenBank under accession of EU862317.

Consequently, the16S rRNAgenesequenceof thestrainC9was

compared with other sequences available in GenBank using

BLAST algorithm. BLAST search against the GenBank database

revealed that the strain C9 was belonging to the genus Clos-

tridium with all three algorithms used (neighbor-joining,

maximum parsimony and maximum likelihood) (Fig. 1).

Further the strain C9 was closely related to the C.

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 5 ( 2 0 1 0 ) 6 6 6 5e6 6 7 36668

saccharolyticum (DSM2544) andC. amygdalinum (ATCCBAA-501)

with 98% similarity. Moreover, the strain C9 is closer to the

Clostridium xylanolyticum (ATCC 4963), Clostridium aerotoleras

(DSM 5434) and Clostridium celerecrescenswith 96% similarity.

3.3. Optimization of pH on hydrogen production ondifferent substrates under mesophilic condition

3.3.1. Hydrogen production from xylan at different pHHydrogen production on xylan was pH dependent (Fig. 3).

When the initial pH rose from 5.5 to 6.5 or 8.0, there was

a significant increase in total hydrogen production from 52 to

120 m mol/L. Further increase in initial pH from 8 to 9.5

resulted in a decrease of total hydrogen production. Thus, the

strain C9 favored optimum hydrogen production (120 m mol/

L) at pH 7.5. Moreover, analysis of growth curve data under

optimal conditions shown that the strain had a comparatively

short lag phase of growth (1 h) and evolution of hydrogen

appeared to start with the cell growth entered to the expo-

nential phase of growth (4 h) (Fig. 2). Hydrogen production

occurred in batch reactor till 60 h. Further fermentation did

not yield significant hydrogen production under present

experimental condition.

During xylan fermentation distribution of acetate and

butyrate were influenced by initial culture pH. At pHs ranging

from 5.5 to 7.5, acetic acid was the major end product (Fig. 3).

However, low amounts of butyric acid were also detected

during this pH range. Further, Bu/Ac ratio ranging from 0.17 to

0.21 was observed during this pH range. On the other hand at

alkaline pHs, high butyric acid production was detected with

low hydrogen production. The Bu/Ac ratio at alkaline pH was

ranging from 0.48 to 0.51. This could be due to the change of

fermentation pathway from acetate to butyrate pathway at

alkaline pH during xylan fermentation. Theoretically when

the fermentation occurs through butyrate pathway, 1 mol of

glucose can produce 2 mol of hydrogen. In addition, trace

amounts of iso-butyrate and succinate, formate and valerate

were detected. Further, small amounts of ethanol up to 2e3%

Fig. 2 e Growth of Clostridium amygdalinum strain C9 on under

and substrate concentration) on xylose, arabinose, xylan and s

were also detected during xylan fermentation. Propionate

production, which consumes hydrogen, was not detected

during xylan fermentation with all the pHs tested. Initial

substrate concentration played a significant role and total

biomass accumulation of the strain C9 was increased when

the xylan concentration further increased up to 20 g/L, but rate

of hydrogen evolution was decreased as compared to that of

10 g/L (Table 2).

There were few studies available on hydrogen production

from xylan. Among them, Taguchi et al. [16] have demon-

strated that Clostridium sp. strain X 53, isolated from termites

showed optimum xylanase activity and hydrogen production

(1254 ml/24 h/10 g/L medium) occurred at pH 6.0 and at 40 �C.But results of this study showed that strain C9 exhibited

optimum hydrogen at pH 7.0 with higher efficiency as

compared to that of strain X 53.

3.3.2. Hydrogen production from arabinoseFermentative hydrogen production from arabinose would be

very interesting since arabinose is a one of the most common

pentose and a component of various hemicellulosic and plant

polysaccharides [17]. Previous reports on hydrogen produc-

tion from arabinose were scarce. Taguchi et al. [18] reported

that optimum yields obtained for the Clostridium strain No. 2

was 2.2mol H2/mol arabinose. As shown in Fig. 4, the strain C9

could utilize arabinose for the hydrogen production within

wide range of pH from 5.5 to 9.5 with optimum hydrogen

production (75mmol/L or 1.78mol/mol arabinose) in the range

of pH8-8.5 (OD600 nm 2.763).

Further, when pH below 7.5, hydrogen production was low,

and average hydrogen production was around 50 mmol/L at

pHs ranging from 5.5 to 7.0. Different types of inocula have

been tested for hydrogen production from arabinose and

demonstrated different pH optima ranging from 6.0 to 8.0 [19].

During arabinose fermentation by strain C9, both acetate and

butyrate were detected (Fig. 4). However, acetate production

was dominant during arabinose fermentation. Further, initial

pH did not have a significant influence on butyrate production

optimum conditions (initial culture pH, NaCl, temperature

tarch during 60 h of incubation.

Fig. 3 e Hydrogen production and distribution of acetate, butyrate production of Clostridium amygdalinum strain C9 during

24 h on xylan fermentation at different initial pH levels.

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 5 ( 2 0 1 0 ) 6 6 6 5e6 6 7 3 6669

as observed in xylan fermentation. Therefore, there is no

significant variation in Bu/Ac ratio at different pHs tested.

Trace amounts of iso-butyrate, isovalerate and formate were

detected during arabinose fermentation. Succinate or propi-

onate production was not detected. Further, high concentra-

tion of valerate production has been observed during

arabinose fermentation when usedmixed cultures [17]. But in

this study, valerate production was not observed.

Further, optimum hydrogen production was obtained with

10 g/L arabinose and higher concentrations did not produce

significantly high hydrogen yield. Moreover, as the initial

concentration of arabinose increases, total hydrogen produc-

tion decreased and VFA formation increased. Excess VFA

formation at high initial arabinose concentration may have

inhibited hydrogen formation. With consistent of the present

results, Danko et al. [17] reported that maximum hydrogen

production did not differ significantly for arabinose concen-

tration between 10 and 50 g/L (Table 2).

3.3.3. Hydrogen production from xyloseD-Xylose, amajor constituent of hardwood hemicelluloses [20]

is a potential feedstock for hydrogen production by xylose-

fermenting microorganisms. Xylose-utilizing hydrogen

Table 1 e Influence of different concentrations of NaCl onhydrogen production on different carbon sources tested.

NaCl (g/L) Total H2 production potential (mmol/L) of strainC9 after 24 h fermentation on different carbon

sources

Xylose Arabinose Xylan Starch

0 21.12 � 0.58 18.87 � 0.26 14.81 � 0.83 22.85 � 0.39

2.5 54.93 � 0.39 41.25 � 0.34 106.22 � 0.42 71.89 � 0.64

7.5 93.24 � 0.87 64.78 � 0.73 119.85 � 0.37 84.69 � 0.21

10 107.08 � 0.42 74.31 � 0.32 108.63 � 0.51 68.45 � 0.43

20 94.83 � 0.74 64.32 � 0.71 89.52 � 0.62 51.36 � 0.51

25 60.92 � 0.31 32.14 � 0.33 37.68 � 0.37 34.69 � 0.62

35 42.52 � 0.25 29.87 � 0.73 24.36 � 0.62 22.32 � 0.43

45 38.96 � 0.61 16.52 � 0.24 13.58 � 0.72 16.52 � 0.27

producing cultures have been reported to produce hydrogen

well at pH values of 4.53e9.33 [21]. Results of this study also

showed that the strain C9 was able to produce hydrogen on

xylose within a wide range of pH from 5.5 to 9.5 (Fig. 5).

However, the strain C9 favored moderately alkaline pHs

(7.5e8.5) for optimum hydrogen production

(ca.141e149 mmol/L) on xylose which is equivalent to

2.2e2.5 mol H2/mol xylose. This was achieved with the initial

xylose concentration of 10 g/L. High concentrations of xylose

tested (15, 20, 25 g/L), did not give significantly higher yields as

compared to that achieved with 10 g/L xylose. This suggests

that the xylose-utilizing hydrogen fermentation might have

a substrate concentration inhibition [22]. Moreover, a consid-

erable amount of ethanol (3e4%) was detected during xylose

fermentation. Further, ethanol production had increased

approximately two-fold at initial concentration of xylose 20 g/

L (Table 2). This suggests that the metabolic pathway may

shifts with the reaction being prolonged [23].

During xylose fermentation, hydrogen production

occurred through mixed-type fermentation and Bu/Ac ratio

was ranging from 0.7 to 0.9. As shown in Fig. 5, the pH above 7,

butyrate production was increased. Theoretically, xylose can

be converted to hydrogen with a maximum yield of 3.33 mol-

H2/mol xylose when the acetate is produced. But when the

fermentation occurred through butyrate pathway, theoreti-

cally, xylose can be converted to hydrogen with a maximum

yield of 1.67 mol H2/mol xylose [24]. Therefore, results of this

study demonstrated that the strain C9 produced 2.2e2.5 mol

H2/mol xylose through mixed type fermentation.

3.3.4. Hydrogen production from starchStarch is a major composition of agricultural by products or

crops such as corn, rice, potato, sweet potato, etc., [25]. The

starch-fermenting bacteria, such as C. butyricum and Enter-

obacter aerogenes have been well studied for the fermentative

hydrogen production [26]. The novel strain C9 isolated in this

study could efficiently utilized starch to produce hydrogen.

Results showed that the hydrogen production occurred at all

levels of pH (5.5e9.0) tested (Fig. 6). The highest hydrogen

production (82 m mol/L) was detected at pH 7.5 (Fig. 6).

Table 2e Influence of different substrates concentrations on total H2 production by C. amygdalinum strain C9 under optimalpH and temperature.

Substrate concentration (g/L) Total H2 production (mmol/L) from each substrate Tested

Xylan Xylose Arabinose Starch

5 98.42 � 0.91 109.32 � 0.29 62.34 � 0.12 56.32 � 0.32

10 120.53 � 0.42 148.52 � 0.18 75.21 � 0.31 75.43 � 0.42

15 124.34 � 0.6.1 147.43 � 0.32 77.42 � 0.41 81.32 � 0.27

20 125.22 � 0.14 145.25 � 0.27 72.72 � 0.43 79.81 � 0.51

25 123.51 � 0.42 71.49 � 0.48 68.53 � 0.61 142.39 � 0.11

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 5 ( 2 0 1 0 ) 6 6 6 5e6 6 7 36670

However, growth curve data analysis showed that the strain

showed long lag phase during starch fermentation (4e5 h) and

hydrogen evolution started after 8 h of incubation (Fig. 2).

Further, during 48 h of fermentation, hydrogen production

reached approximately 52 mmol/L at optimum pH.

Results of this study showed that under optimal batch

experimental conditions, the highest total hydrogen produc-

tion was ca. 81 mmol H2/L, which is equivalent to ca. 390 ml

H2/g starch. This was achieved with the initial starch

concentration of 15 g/L (Table 2). However, as compared to the

previous studies, the novel strain C. amygdalinum strain C9

could be useful for hydrogen production from starch [26].

Further, the optimum hydrogen production (390 ml H2/g

starch) was achieved with the initial starch concentration

20 g/L under optimum conditions [25]. Further increase in

starch concentration up to 25e30 g/L decreased of hydrogen

production to 75 mmol/L. However, it has been shown that

high starch concentrations (20e40 g/L) had no remarkable

effect on hydrogen production [23]. Further, the optimum

hydrogen production (390 ml H2/g starch) was achieved with

the initial starch concentration 20 g/L under optimum condi-

tions [25]. Further increase in starch concentration to 25e30 g/

L resulted in a decrease of hydrogen production to 75mmol/L.

However, it has been shown that high starch concentrations

(20e40 g/L) had no remarkable effect on hydrogen production

[23]. This could also be possibly due to the shifting of bacterial

metabolism towards the volatile fatty acid formation rather

than hydrogen formation [4].

0

20

40

60

80

100

5.5 6 6.5 7

Initi

oitcudorpnegordy

Hn

)L/lo

mm(

Acetate (mButyrate (Hydrogen

Fig. 4 e Hydrogen production and distribution of acetate, butyr

24 h on arabinose fermentation at different initial pH levels.

During starch fermentation, acetate was detected as

a major volatile fatty acid. Low amounts of butyrate produc-

tion were detected irrespective of initial culture pH. The Bu/Ac

ratio was ranging from 0.11 to 0.15 at pH below 7.5. Moreover,

it was observed that there is a tendency of increasing Bu/Ac

ratio with increasing pH from 8 to 9. In addition, small quan-

tities of iso-butyrate (<25 mg/L) and succinate and formate

(<75 mg/L) were also detected. Further, butyrate production

was slightly increased at alkaline pHs (8.0e9.0). Therefore,

results of the distribution of fatty acid data showed that

hydrogen production on starch occurred mainly through

acetate pathway, which is leading to the high hydrogen

production. However, total hydrogen production (390 ml H2/g

starch) from starch achieved in this study was lower than the

theoretical maximum 550 ml H2/g starch, which occurred

through the acetate pathway [4].

3.3.5. Hydrogen production under different concentrations ofNaClSalt is essential for cells, composed primarily of sodium

chloride. Generally the salt concentrations of natural water or

various waste waters were either fresh or marine [27]. Results

of this study showed that sodium chloride was essential for

the optimum growth and hydrogen production of the strain

C9. Addition of NaCl enhanced the hydrogen production

potential of the strain C9 on all the carbon sources tested

(xylan, xylose, arabinose and starch). Optimum NaCl

concentration for the maximum hydrogen production is

7.5 8 8.5 9

al pH

0

2

4

6

8

10

12

14

16

18

20

/lom

m(noitcudorp

sdicA

L)

mol/L) mmol/L) production (mmol/L)

ate production of Clostridium amygdalinum strain C9 during

0

40

80

120

160

5.5 6 6.5 7 7.5 8 8.5 9

Initial pH

oitcudorP

negordyH

n)

L/lom

m(

0

2

4

6

8

10

12

14

16

18

20

/lom

m(noitcudorp

sdicA

L)

Acetate (mmol/L) Butyrate (mmol/L) Hydrogen production (mmol/L)

Fig. 5 e Hydrogen production and distribution of acetate, butyrate production of Clostridium amygdalinum strain C9 during

24 h on xylose fermentation at different initial pH levels.

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 5 ( 2 0 1 0 ) 6 6 6 5e6 6 7 3 6671

substrate dependant and the strain C9 showed maximum

hydrogen production on xylose and arabinose in the presence

of 10 g NaCl/L. Further increase of NaCl concentration in the

medium reduced the hydrogen production potential of the

strain (Table 1). The strain required low concentration of NaCl

(7.5 g/L) for the optimum hydrogen production on xylan and

starch as compared to that of xylose and arabinose. Since the

requirement of high NaCl concentration for the optimum H2

production, the novel strain would be very important to study

it’s potential for the treatment of marine waste.

3.3.6. Hydrogen production under microaerophilic conditionsIt was interestingly noted that the strain could produce

hydrogen under microaerophilic conditions. This was

confirmed by the methylviologen assay. The results of the

methylviolegen assay showed that when 50% of the head-

space replaced with air, the hydrogen evolution was

Fig. 6 e Hydrogen production and distribution of acetate, butyra

24 h on starch fermentation at different initial pH levels.

approximately 2500 nmol/400 mL of culture (2.2 OD at 600 nm).

However, as compared to that of control (under strict anaer-

obic condition), the strain exhibited only 30% reduction in H2

evolution. This suggests that the hydrogenase system present

in strain C9 could be oxygen tolerant. Since oxygen tolerance

hydrogenases have a great potential in future biotechnolog-

ical use, the novel strain C9 would be very important in future

hydrogen production research. Further investigations are

going on to evaluate the actual potential of the hydrogease

enzyme system to produce hydrogen under miceoaerophilic

condition in vitro.

Overall, the final pHs in all batch reactors were within the

range of pH 4.0e4.6 at the end of 24 h fermentation, regard-

less of the initial culture pH or the type of substrate used.

Hence the strain C9 showed low rate of hydrogen evolution

after 24 h during this study. Further, with low hydrogen

production, gradual increase in ethanol production was

te production of Clostridium amygdalinum strain C9 during

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 5 ( 2 0 1 0 ) 6 6 6 5e6 6 7 36672

detected with xylose and xylan fermentation. This suggests

that microbial shift from favorable hydrogen/acid production

phase to energy-consuming ethanol production phase. This

shift may occur when the pH was further down to 4.5 or

below [24].

4. Conclusion

This study reported the hydrogen production potential of

moderately halophilic novel marine strain of C. amygdalinum

C9 isolated from sea buried crude oil pipelines. The strain was

identified based on 16S rRNA gene sequence. Hydrogen

production by strain C 9 was pH and substrate dependant.

Further, the strain favored alkaline pHs for optimum

hydrogen production of xylose (pH 7.5e8.5), xylan (pH 7.5),

arabinose (pH 8.0) and starch (pH 7.5). NaCl is essential for

optimum growth and hydrogen production. The strain C9

preferred high NaCl concentrations for optimum hydrogen

production on xylose and arabinose (10 g NaCl/L), and Xylan

and starch (7.5 g NaCl/L). Finally, it is very interestingly to note

that the strain had a great potential to produce hydrogen

under microaerophilic conditions. Therefore, the results dis-

cussed in this study could be used to assess the feasibility of

using pure bacterial strain in converting renewable energy

resources such as pentose or starch to clean H2. Hence this

study provides useful information for the design and opera-

tion of a fermentation processes using highly abundant

renewable energy resources like xylan, xylose, arabinose and

starch.

Acknowledgements

We are indebted to the R&D centre of Hindustan Petroleum

Cooperation Ltd., Mumbai and the Corporate Health Safety

Environment division of Oil and Natural Gas Corporation for

financial support of this study and also Department of

Biotechnology, Govt. of India for partial support. The

authors are thankful to Dr R. K. Pachauri, DG, TERI, New

Delhi, for providing infrastructure facility to carry out the

present study. We thank Abu Swealsh, Laboratory techni-

cian, Microbial Biotechnology Division, TERI for collection of

samples.

r e f e r e n c e s

[1] Kim D-H, Kim S-H, Shin S-H. Sodium inhibition offermentative hydrogen production. Int J Hydrogen Energy2009;34:3295e304.

[2] Yuan Z, Yang H, Zhi X, Shen J. Enhancement effect of L-cysteine on dark fermentative hydrogen production. Int JHydrogen Energy 2008;33:6535e40.

[3] Ren N-Q, Cao G-L, Guo W-Q, Wang A-J, Ahu Y-H, Liu B-F,Xu J-F. Biological hydrogen production from corn stove bymoderately thermophile Thermoanaerobacteriumthermosaccharolyticum W16. Int J Hydrogen Energy 2010;35:2708e12.

[4] Argun H, Kargi F, Kadapdan IK, Oztekin R. Batch darkfermentation of wheat starch to hydrogen gas: effects of theinitial substrate and biomass concentration. Int J HydrogenEnergy 2008;33:6109e15.

[5] KotaySM,DasD. Biohydrogenas a renewableenergy resources-prospectsandpotentials. Int JHydrogenEnergy2008;33:258e63.

[6] Mohana S, Shah A, Divecha J, Madamwar D. Xylanaseproduction by Burkholderia sp. DMAX strain under solid statefermentation using distillery spent wash. BioresourceTechnol 2008;99:7553e64.

[7] Saratale GD, Chen S-D, Lo YC, Saratale RG, Chang S J-.Outlook of biohydrogen production from lignocellulosicfeedstock using dark fermentation- a review. J Sci Ind Res2008;67:962e79.

[8] Miller TL, Wollin MJ. A serum bottle modification of theHungate technique for the cultivating obligate anaerobes.Appl Microbiol 1974;27:985e7.

[9] Sambrook J, Fritsch EF, Maniatis T. Molecular cloninig:a laboratory manual. 2nd ed. Cold Spring Harbor, NY:ColdSpring Harbor Laboratory; 1989.

[10] Altschul SF, Madden TL, Schaffer AA, Zhang Z, Miller W,LipmanDJ. GappedBLASTandPSI-BLAST: a newgeneration ofprotein database search programme. Nucl Acids Res 1997;25:3389e402.

[11] Jukes TH, Cantor CR. Evolution of protein molecules. In:Munro HN, editor. Mammalian protein metabolism. NewYork: Academic Press; 1969. p. 211e32.

[12] Felsenstein J. PHYLIP (Phylogeny Inference Package) version3.67. Seattle: Department of Genome Sciences, University ofWashington; 2007.

[13] Saitou N, Nei M. The neighbor-joining method: a newmethod for reconstructing phylogenetic trees. Mol Biol Evol1987;4:405e25.

[14] Felsenstein J. Confidence limits on phylogenies: an approachusing the bootstrap. Evolution 1995;30:783e91.

[15] Jayasinghearachchi HS, Sarma PM, Singh S, Aginihotri A,Mandal AK, Lal B. Fermentative hydrogen production by twonovel strains of Enterobacter aerogenes HGN-2 and HT 34isolated from sea buried crude oil pipelines. Int J HydrogenEnergy 2009;34:7197e207.

[16] Taguchi F, Hasegawa K, Saito-Taki T, Hara K. Simultaneousproduction of xylanase and hydrogen using xylan in batchculture of Clostridium sp. strain X53. J Ferment Bioeng 1996;81:178e80.

[17] Danko AS, Abreu AA, Alves MM. Effect of arabinoseconcentration of dark fermentation hydrogen using differentmixed cultures. Int J Hydrogen Energy 2008;33:4527e33.

[18] Taguchi F, Mizukami N, Hasegawa K, Saito-Taki T. Microbialconversion of arabinose and xylose to hydrogen by a newlyisolated Clostridium sp. no. 2. Can J Microbiol 1994;40:228e33.

[19] Abreu AA, Danko AS, Costa JC, Ferreira EC, Alves MM.Inoculum type response to different pHs on biohydrogenproduction from L-arabinose, a component of hemicellulosicbiopolymers. Int J Hydrogen Energy 2009;34:1744e51.

[20] Lo YC, Chen WM, Hung CH, Chen SD, Chang JS. Darkfermentation from sucrose and xylose using H2-producingindigenous bacteria: feasibility and kinetic studies.Water Res 2008;42:827e49.

[21] Taguchi F, Mizukami N, Saito-Taki T, Hasegawa K. Hydrogenproduction from continuous fermentation of xylose duringgrowth of Clostridium sp. strain no. 2. Can J Microbiol 1990;41:536e40.

[22] Lin CY, Cheng CH. Fermentative hydrogen production fromxylose using anaerobic mixed microflora. Int J HydrogenEnergy 2006;31:832e40.

[23] Yang H, Shen J. Effect of ferrous iron concentration onanaerobic bio-hydrogen production from soluble starch. Int JHydrogen Energy 2006;31:2137e46.

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 5 ( 2 0 1 0 ) 6 6 6 5e6 6 7 3 6673

[24] Ren N, Li J, Li B, Wabg Y, Liu S. Biohydrogen production frommolasses by anaerobic fermentation with a pilot-scalebioreactor system. Int J Hydrogen Energy 2006;31:2147e57.

[25] Lin C-Y, Chang C-C, Hung C-H. Fermentative hydrogenproduction from starch using natural mixed cultures. Int JHydrogen Energy 2008;33:2445e53.

[26] Yokoi H, Tokushige T, Hirose J, Hayashi S, Takasaki Y. H2

production from starch by amixed culture of ClostridiumbutyricumandEnterobacteraerogenes. BiotechnolLett1998;10:43e7.

[27] Zhu D, Wang G, Qiao H, Cai J. Fermentative hydrogenproduction by new marine Pantoea agglomerans isolated fromthe mangrove sludge. Int J Hydrogen Energy 2008;33:6116e23.