Firmicutes with iron dependent hydrogenase drive hydrogen production in anaerobic bioreactor using distillery wastewater

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 6 ( 2 0 1 1 ) 8 2 3 4e8 2 4 2

Avai lab le at www.sc iencedi rect .com

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

Firmicutes with iron dependent hydrogenase drive hydrogenproduction in anaerobic bioreactor using distillery wastewater

S. Venkata Mohan a, Leena Agarwal b, G. Mohanakrishna a, S. Srikanth a, Atya Kapley b,Hemant J. Purohit b,*, P.N. Sarma a

aBioengineering & Environmental Centre, Indian Institute of Chemical Technology, CSIR, Hyderabad 500607, IndiabEnvironmental Genomics Division, National Environmental Engineering Research Institute, CSIR, Nehru Marg, Nagpur 440020, India

a r t i c l e i n f o

Article history:

Received 17 January 2011

Received in revised form

29 March 2011

Accepted 4 April 2011

Available online 11 May 2011

Keywords:

Distillery

hydA

Hydrogen

Iron dependent hydrogenases

Microbial diversity

Wastewater

* Corresponding author. Tel.: þ91 712 224988E-mail address: [email protected]

0360-3199/$ e see front matter Copyright ªdoi:10.1016/j.ijhydene.2011.04.021

a b s t r a c t

Distillery wastewater rich in organics is an inexpensive renewable resource for making first

generation biofuel. Distillery wastewaters are mostly treated via the biomethanation route;

however, in this study the conditions in sequential batch reactor (SBR) are being set to

develop and analyze the microbial community that opted for hydrogen production. An

optimum performance condition for a bioreactor was achieved after 40 days of operation,

which gave substrate degradation rate of 0.72 kg/m3-day with volumetric hydrogen

production of 0.32 mol H2/m3-day. Study proposes that the dominant Delftia sp., a hydrogen

oxidizing bacterium has been replaced during hydrogen production mode with dominant

Anaerofilum sp., an anaerobic Firmicute and the iron dependent hydrogenases dominated

as functional gene for hydrogen production. Future studies are required where process-

engineering interventions could be applied to improve the hydrogen driving biochemical

process.

Copyright ª 2011, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights

reserved.

1. Introduction the distilleries for hydrogen production would affect the

Due to high-energy yield (122 kJ/g) and non-polluting nature,

hydrogen (H2) is deemed to be a promising fuel for future; and

is produced by the reactions of natural gas or light oil fractions

with steam at high temperatures [1]. Alternatively, using

renewable resources and with microbial capacity for biolog-

ical routes of H2 production are gaining importance. Among

them dark-fermentation process was reported to have

advantages due to the feasibility of utilizing wastewater as

substrate [2e4]. One such source of wastewater is molasses-

based distilleries, which generate 8e15 L of wastewater per

liter of ethanol produced [5]. Ethanol is emerging as one of

future biofuels, in that scenario the utilization of waste from

3.(H.J. Purohit).

2011, Hydrogen Energy P

overall economics of this industry.

Microbes can produce hydrogen by either fermentation [6]

or photosynthesis [7]. Hydrogen production by fermentation

has been extensively studied on several pure cultures [8,9] viz.

Firmicutes [10e14] and Enterobacteria [15,16]. However, in

case of wastewater where the carbon source could be diverted

for secondary metabolism, it is difficult to control the biolog-

ical processes. In this study operational conditions like HRT

and organic loading has been optimized for maximum

production of hydrogen by the microbial community being

acclimatized on distillery waste water. The hydrogen

producing microbial community was analyzed by culture

independent approach using 16S rDNA sequencing and

ublications, LLC. Published by Elsevier Ltd. All rights reserved.

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 6 ( 2 0 1 1 ) 8 2 3 4e8 2 4 2 8235

tracking of hydrogen producing genotype (hydA) by quantita-

tive PCR method.

Hydrogen is one of the promising alternative energy sour-

ces in future and of late microbial hydrogen production from

mixed microbial cultures using wastewater is gaining impor-

tance. This study targets 2 important points (a) Producing

alternative energy source i.e. Hydrogen and (b) Producing

hydrogen energy using distillery waste water, hence gener-

ating energy from the waste. Therefore, in this study optimi-

zation of various operational parameters viz. HRT and organic

loading has been carried out for enhancing hydrogen

production. The microbial diversity as well as functional

diversity (hydA gene) of the SBR (Sequential batch reactor) at

different point of operation and optimization has been

analyzed to enrich the knowledge about the mixed microbial

cultures responsible for hydrogen production. Based on this

study, it is possible that design and maintenance of similar

bioreactors can be planned for hydrogen production.

2. Materials and method

2.1. Reactor design

The bioreactor was designed based on our earlier work [4].

Wastewater samples were collected regularly from industrial

premises and stored until further use at w3�1 �C after the

initial cultivation pHwas adjusted to 7.0� 0.1. Anaerobicmixed

consortiawas acquired as seed for bioreactor fromanoperating

full-scale anaerobic plant treating composite wastewater from

industrial and domestic sources. Bench scale reactor with

suspended growth configuration was fabricated using ’perplex’

material with leak proof sealing along with proper inlet and

outlet arrangements (total volume, 4 L; working volume,

3.6 L).Sequential batch reactor was operated in up-flow mode

with a total cycle period of 24 h consisting of 0.25 h of fill phase,

23.00 h of reaction (anaerobic) phase, 0.50 h of settling phase

and 0.25 h of decant phase. Various sequence phase operations

(feed, reaction, settle and decant) were controlled by pre-

programmed timers employing peristaltic pumps. Initially the

bioreactor was operated with distillery wastewater at an

organic loading (OL) of 3.9 kg COD/m3. After achieving stable

performance, the OL was changed (Table 1). The system was

Table 1 e Consolidated data of bioreactor performance duringproduction and substrate degradation.

SamplingTime(months)

OL(kg COD/m3)

HRT(days)

CHP(mmol)

VHP(mol/m3-day)

COD

R (g/l)xCOD

1st 3.90 3 1.09 0.09 1.45 37.

2nd 9.80 5 2.78 0.14 4.34 44.

3rd 24.50 10 9.91 0.25 6.70 27.

4th 24.50 10 12.76 0.32 8.00 32.

5th 24.50 15 13.64 0.23 8.81 35.

Anaerobic mixed culture; suspended growth; batchmode up-flow operation; feedin

production; VHP, volumetric H2 production; CODR, Total COD removal per cycle;

specific H2 production [¼CHP/CODR � 1000/1000]; VFAmax, maximum VFA pro

operated until steady state performance with respect to COD

removal efficiency and H2 production was achieved. The

reactor was operated with retention time of 24 h at a recircu-

lation rate of 240 ml/h at a temperature 28 � 2 �C under

microaerophilic to anaerobic environment.

2.2. Analysis

H2 gas generated during the experiments was estimated using

a microprocessor based pre-calibrated H2 sensor (ATMI GmBH

Inc., Germany). pH, COD (closed refluxing-titration), suspended

solids, BOD5 (5 days at 20 �C), TDS and volatile fatty acids (VFA)

were determined according to Standard Methods [17].

2.3. Isolation of metagenome

After the bioreactor was stabilized on distillerywastewater for

maximum production of hydrogen, samples were withdrawn

for metagenome extraction. Three samples at an interval of

a month, at third, fourth and fifth month of operation, were

collected from the operating reactor and metagenome was

extracted using a protocol reported earlier [18]. For each

sampling month, randomly ten independent samples of

100mg wet weight of biomass were collected. DNA from all 10

replicateswas pooled to represent a homogeneousmixture for

each month, and stored at �20 �Cuntil required.

2.4. Construction and screening of 16S rDNA cloneLibrary

100 ng of metagenomic DNA was used as template to amplify

16S rDNA gene using universal 16S rDNA primers 27F forward

primer 50-AGAGTTTGATCMTGGCTCAG-30 and 1492 reverse

primer 50-TACGGYTACCTTGTTACGACTT-30. Three indepen-

dent PCR reactions were performed for each month and

amplification products were pooled. The 1.5 kb amplicons

were gel-purified using the gel extraction kit from Qiagen and

cloned into the pDrive vector (Qiagen PCR Cloning Plus Kit,

Germany) as per instruction provided by supplier. Trans-

formants were grown overnight in 5ml LBmedium containing

100 mg/ml kanamycin and used for plasmid DNA purification

using Qiagen Q-20tips (Germany). The 16S rDNA insert was

excised by EcoRI digestion and gel-purified. Total 57 clones

treatment of distillery wastewaters as a function of H2

(%) SDR(kg CODR/m

3-day)SHP

(moles H2/Kg CODR)pHshift

VFAmax

(mg/l)

18 0.44 0.75 7.0 to 7.1 1215

29 0.78 0.64 7.0 to 7.2 1852

35 0.61 1.48 7.0 to 6.8 3690

65 0.72 1.60 7.0 to 7.3 3365

92 0.53 1.55 7.0 to 7.3 3025

g pH, 7; OL, organic load; HRT, hydraulic retention time; CHP, cumulative H2

xCOD, COD removal efficiency ; SDR, Substrate (COD) degradation rate ; SHP,

duced per cycle.

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 6 ( 2 0 1 1 ) 8 2 3 4e8 2 4 28236

with 16S rDNA insert from three sampling time points were

selected randomly for sequencing. Partial sequencing was

carried out from the 50 end using the 27F primer. The sequence

data was analyzed by BLAST to identify the corresponding

clones. Sequences were deposited in GenBank and accession

numbers [from FJ842899 to FJ842955] were indicated in the

supporting data and can be viewed at http://www.ncbi.nlm.

nih.gov/.

2.5. Cloning hydrogenase gene from hydrogen producingmetagenome

Hydrogenase gene (hydA) was amplified using PCR conditions

and primer set (HGf 50-AAGAAGCTTTAGAAGATCCTAA-30)/HGr 50-GGACAACATGAGGTAAACATTG-30) and degenerative

primers BHf (50-TCACCWCAACAAATWTTTGG-30)/BHr (50-GCWGCTTCCATWACTCCACC-30) reported earlier [19]. The

amplicons were cloned into pDrive cloning vector (Qiagen PCR

Cloning Plus Kit, Germany) as described earlier. Cloning,

screening and sequencing of clones for hydrogenase genewas

carried out in similarmanner as described above for 16S rDNA

clones. Eight clones (with hydA gene insert) from each month

were randomly selected for sequencing. Sequences of

hydrogenase gene clones were submitted in GenBank and

accession numbers [from GU253925 to GU253931, GU253936 of

4th month and GU253939 to GU253946 of 5th month] are indi-

cated in the supporting data and can be viewed at http://www.

ncbi.nlm.nih.gov/.

2.6. PCR and quantitative/real time PCR analysis

Quantification of hydrogenase gene from the metagenome

was performed by Real-time PCR assay using an iCycler IQ

System (BioRad, USA). PCR reactions were set up using 100 ng

of extracted metagenomic DNA as template, 1X iQ SYBR

Green Supermix (BioRad, USA), and hydA primers. Regular

PCR was also done using the same conditions without SYBR

Green.

3. Results

3.1. Biohydrogen production

Bench-scale anaerobic reactor was operated continuously for

five months under neutral pH, microaerophilic environment

by varying substrate/organic loading (OL) condition and

hydraulic retention time (HRT) to arrive at optimum hydrogen

producing condition as depicted in Table 1. The substrate

loading was changed once the system showed stable perfor-

mance with respect to both H2 production and substrate

degradation. Initially the reactor was fed with distillery

wastewater at organic loading (OL) of 3.9 kg COD/m3 and

operated at hydraulic retention time (HRT) of 3 days. During

this phase of operation, the system showed volumetric

hydrogen production (VHP) of 0.09mol H2/m3-day at stabilized

conditions. Subsequently, OL was increased to 9.8 kg COD/m3

and operated initially at HRT of 3 days during which the

system yielded lower H2 production. After operating for three

cycles, the HRT was increased to 5 days and the reactor

performance was studied by keeping OL undisturbed. During

stabilized phase of operation at this condition, the system

showed enhanced H2 production (0.14 mol H2/m3-day).

Subsequently, the systemwas operated by almost doubling OL

(24.5 kg COD/m3) and HRT (10 days), wherein H2 production

also doubled. With time, H2 production showed consistent

improvement and stabilized at VHP of 0.32 mol H2/m3-day.

Further, the system was operated at higher HRT (15 days) by

keeping OL constant to evaluate role of HRT on the system

performance. Even though the system showed higher H2

production, volumetric production showed decreasing trend

with increasing the HRT at same substrate loading condition.

This might be attributed to the activity of methanogenic

bacteria persisting in the system where the produced H2 get

used in the formation of CH4 in the process of consuming

soluble acid intermediates, [volatile fatty acids (VFA)] gener-

ated from the acidogenic process during the longer retention

time. Irrespective of the retention time, the system showed

marked improvement in H2 production with subsequent

increase in substrate loading and retention time.

substrate loading conditions and HRT also showedmarked

influence on the substrate degradation rate (SDR). At 3.9 kg

COD/m3-day, the system showed COD removal efficiency

(xCOD) of 37.18% with SDR of 0.44 kg CODR/m3-day. With

increase in substrate loading to 9.8 kg COD/m3, substrate

degradation also increased significantly (44.29%; 0.781 kg

CODR/m3-day).

However, at applied higher substrate load of 24.5 kg COD/

m3 (at 10 days HRT), the substrate degradation decreased

(27.35%; 0.61 kg CODR/m3-day) and approached maximum

stabilized performance after 40 days of operation (32.65%;

0.72 kg CODR/m3-day). Longer period of adaptation observed

under these loading conditions may be attributed to system

inhibition due to shift to higher carbon load. The system

when operated at higher HRT (15 days) showed higher

degradation efficiency (35.92%; 0.53 kg CODR/m3-day)

compared to lower HRT studies in spite of suppressed H2

production. The higher degradation efficiency observed at

higher HRT might be attributed to longer retention of the

substrate in the bioreactor, which is the key for activation of

methanogenic bacteria facilitating additional substrate

degradation. This is in concurrence with the lower VFA

concentration observed at higher HRT when compared to

lower HRT (10 days).

Formation of soluble metabolites (VFA) in dark-

fermentation process is crucial in defining efficacy of H2

metabolic pathway. During dark-fermentation, the generation

of VFAwas in good agreementwith substrate loading andHRT.

Higher substrate load generated higher VFA concentration.

While increasing HRT by keeping carbon load constant resul-

ted in low VFA concentration at the end of cycle period. This

phenomenon observed may be due to its consumption by

methanogenic bacteria in the formation of methane utilizing

H2. VFA and pH are integral expressions of acid-base condi-

tions of anaerobic microenvironment, which provide infor-

mation pertaining to the balance between two important

microbial groups (acidogenic and methanogenic) within the

system. Production of acids gradually reduces buffering

capacity of the system, which in turn, resulted in a concomi-

tant decline in the system pH due to accumulation of organic

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 6 ( 2 0 1 1 ) 8 2 3 4e8 2 4 2 8237

acids leading to process inhibition [12].The shift of pH was

observed towards basic condition in most of the cases studied

due to neutral operation. The system retained more or less

neutral condition during system operation in spite of VFA

generation,may be because of good systembuffering capacity.

3.2. Bacterial community composition dynamics atdifferent time points

This study was directed towards analysis of microbial commu-

nity, its function, and relation to production of hydrogen under

different bioreactor process parameters. Microbial community

analysis is widely studied by targeting 16S rDNA [20]. After two

months of functionally stable condition for bioreactor the total

DNA (metagenome) at 3rd, 4th, and 5thmonthwas extracted. 16S

rDNAwas PCR amplified and cloned. Total 57 clones i.e. 18 from

3rd month, 19 from 4th month and 20 from 5th month were

selected based on EcoRI digestion and sequenced. The third

month sample was dominated by the presence ofDelftia sp., an

aerobic genera, indicating that at this time point anaerobic or

microaerophilic conditions were yet to be achieved in the

bioreactor.The fourthandfifthmonthsamplesweredominated

by the anaerobe, Anaerofilum sp. Table 1 demonstrates the

increase in hydrogen production with the shift in microbial

community from aerobic to anaerobic/microaerophilic. Phylo-

genetic trees can be viewed in the supporting data, (Figure S1).

Sequence analysis of 16S rDNA clones of third month sample

demonstrated two clusters, C3-I and C3eII (Figure S1a). C3-I

cluster represents 83.3% of the total clones and showed high

similarity with Delftia sp. (accession no. EF421406) while the

remainder of clones (16.7%) formed cluster C3eII showed high

similarity with Sphingomonas sp. ATCC 53159 (accession no.

AF503283). All the clones isolated at this time grouped with

b-proteobacteria (C3eI) and a-proteobacteria (C3eII). Meta-

genome from the fourth month of operation, when the biore-

actor was producing 1.60 molH2/Kg COD, showed 16S rDNA

sequences distributed in three clusters viz. C4-I, C4eII and

C4eIII (Figure S1b). 85% of the total clones were identified as

Anaerofilum sp. and are observed to cluster with the genus

Ruminococcus. Both genera belong to the Lachnospiraceae family

from the phylum Firmicutes. The second and third cluster in

Fig. 1 e Hydrogenase gene (hydA) amplification and quantificat

the third, fourth and fifth month of operation. Increase in targe

course of operation. The analysis is based on relative fluorescen

was no significant hydrogen production in the first two months

reference to the results observed during the second month of o

this sample contains only one clone each, C4eII is represented

by ‘uncultured bacterium clone’ and C4eIII by ‘uncultured

Clostridiaceae bacterium clone’. Uncultured bacterium clone

[FJ842931] showed 100% similarity with Delftia sp. and Uncul-

turedClostridiaceaebacteriumclone[FJ842932]wasclassifiedas

Ruminicoccaceae by RDP classifier available at http://rdp.cme.

msu.edu/classifier/classifier.jsp but probably due to short

sequence lengthwould not group with Delftia acidovorans strain

EEZ23 and Ruminococcus clade respectively.

In the fifthmonth of operation, the HRTwas increased from

10 days to 15 days. The microbial community profile did not

change suggesting stability, but a drop in specific hydrogen

production rate was observed. Figure S1c demonstrates the

bacteria community profile at this time point. It can be seen

that the community is dominantly represented by Anaerofilum

sp. Two clusters C5-I and C5eII are observed. Despite being

identified as Anaerofilum sp., C5eII did not align with other

clones. This could possibly be due to short sequence length as

compared to others.

3.3. Quantification and analysis of hydrogenase (hydA)gene

hydA codes for the hydrogenase gene, responsible for the

production of hydrogen. The study also demonstrated the

detection of hydA genotype in all the three extracted meta-

genomes. Fig. 1 shows the amplification pattern of hydA

genotype. Samples analyzed from the metagenome of the

third month of operation, show very faint hydA product. Since

there was significant increase in hydrogen production in the

4th and 5th month quantitative PCR was carried out to assess

the increase in copy number of hydA genotype. The figure

demonstrates 5 fold increase in hydrogen production geno-

type in the fourth and fifth month of operation. The hydA

genotype was also cloned in the fourth and fifth month of

reactor operation and random clones were sequenced. BLAST

analysis of sequence data demonstrated homology with the

reported Fe hydrogenase gene. Two phylogenetic trees were

constructed with hydA gene clones isolated at 4th and 5th

months of hydrogen production. Reported Ni, Fe hydrogenase

gene sequences were downloaded from GenBank and used in

ion from metagenome of bioreactor producing hydrogen at

t genotype* (hydA gene locus) in the bioreactor during its

ce units (RFU) of three independent analysis. * Since there

, these points are not included. ** These values are in

peration of the reactor.

Fig. 2 e Relationship of hydA gene clones constructed at (a) fourth and (b) fifth of operation of bioreactor with other

hydrogenase genes reported in literature. Bootstrap confidence values obtained with 1000 iterations are given at branch

points. Metal free hydrogenase from Methanoscarcina mazaei (Accession no. 1479567) was used as an out group. Reference

strains used in tree construction are given in ‘italic’ type face followed by their sequence accession numbers in paranthesis.

Details describing the reference sequences used in tree construction are given in Table S3. Clones generated in this study

are shown in ‘bold’ type face followed by accession no. in parenthesis.

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 6 ( 2 0 1 1 ) 8 2 3 4e8 2 4 28238

tree construction. A metal free hydrogenase was used as out-

group. Fig. 2a and b demonstrate the clustering of hydA clones.

All clones generated from the metagenome at two different

time points cluster with Fe dependent hydrogenase, which

were reported for hydrogen production.

4. Discussion

The search for raw material to produce new-generation bio-

energy is a priority area of research to support any economy

[21]. Distilleries producingmolasses spent wash could be used

to generate biogas. India is the second largest producer of

alcohol in Asia using molasses as raw material [22] and

presently 295 distilleries are operating across the country,

which produces about 3198 million liters of ethanol per

annum [23] where India generates approximately 10e15 L of

molasses spent wash (MSW) per liter of alcohol produced [24].

Mostly, the wastewater generated by these industries gets

partially utilized for biomethanation.

Several studies with mixed cultures, co-cultures, sludge

compost [25e28] and pure cultures [29e36] have been reported

for microbial hydrogen production. Patel et al. [36], reported

a 2e2.5 fold enhancement in hydrogen production using

mixed cultures immobilized on ligno-cellulosic bio-wastes. In

addition, optimization of process parameters led to reduced

incubation period from 4 days in batch culture to 2 days in

daily fed conditions. This study demonstrates that by

Fig. 2 e (continued).

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 6 ( 2 0 1 1 ) 8 2 3 4e8 2 4 2 8239

adjusting the process parameters, optimum yield of hydrogen

could be achieved by mixed microbial community that has

been acclimatized to distillery wastewater.

At the initial stage when the bioreactor was running under

microaerophilic condition i.e up to 3rd month of acclimatiza-

tion, it was dominated by aerobic population affiliated with

uncultured Delftia sp. clones and uncultured Sphingomonas sp.

clones. At the 3rd month of operation, Delftia sp. (a hydrogen

oxidizingbacterium)dominated in themicrobial community of

the reactor resulting in low hydrogen production. Microbial

community structure completely changed or shifted to anaer-

obic population at 4th and 5th month of operation. 85% of the

microbial community in the fourth month of operation was

represented by Anaerofilum sp., while 10% of the population

constituted Clostridiaceae bacterium clones and 5% by uncul-

tured bacterium. At the 5th month, the microbial population

dynamics was stabilized with 100% uncultured Anaerofilum sp.

clones. This difference in community structure is undoubtedly

reflected in sequential change of hydrogen yield from

9.91 mmol at 3rd month to13.64 mmol at 5th month.

The microbial diversity of any biological reactor is an

important issue in terms of qualitative and quantitative

assessment of efficiency of a bioreactor. Microbes produce

hydrogen using the enzyme hydrogenase, that catalyses

either the production or uptake of hydrogen. Presence of

metal cofactor at the active site categories them into three

groups: (a)- Iron containing hydrogenases, [Fe] (b)- Nickel,

Iron and selenium (in some) hydrogenases, [NiFe] or [NiFeSe]

(c)- Metal free [37]. [NiFe] hydrogenases are widely distributed

class of hydrogenases that consume hydrogen while Fe

hydrogenases, produce hydrogen and have been reported in

anaerobes and some green algae [38e40]. Since, Fe hydoge-

nases are responsible for hydrogen production they can be

used as biomarker for detection of hydrogen producing

bacteria. Hydrogenases are coded by the gene hyd; Fe

hydrogenases have two subunits hydA (large) and hydB

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 6 ( 2 0 1 1 ) 8 2 3 4e8 2 4 28240

(small) [41,42]. Metal free hydrogenases do not produce

hydrogen and are reported in some methanogens [37]. With

the available microbial capacities, the microbial hydrogen

production has not yet developed as an economically viable

option. There are various levels in process development,

where it demands further investigations including improve-

ment in hydrogen production by metabolic engineering

[43,44]. The key factor in economics of hydrogen production

lies in selection of raw material followed by process option.

The anaerobic option of hydrogen production depends on the

available microbial biochemistry which has mostly been

explored in Clostridium spp. Clostridium is reported as a major

component of microbial community in activated sludge

responsible for hydrogen production [19,47]. It belongs to

Firmicute phylum and has been reported in hydrogen

production using various substrates like glucose [47], sucrose

[45], xylose [46], starch [47], lactose [48] and sweet potato

resides [49]. Clostridium sporosphaeroides F52, C. tyrobutyricum

F4, C. pasteurianum F40, have been used as co-cultures to

enhance hydrogen production [50,51]. In the present studies,

molasses spent wash generated from distilleries has been

used as raw material.

With adjustment in HRT and organic loading, during

process optimization, it has been observed in the present

studies that the microbial population became anaerobic at 4th

and 5th months of operation thus providing suitable condi-

tions for anaerobic hydrogen production. This data was

further supported by quantitative amplification of hydA gene

for microbial community in these months. The hydrogenase

reported in the study is iron dependent, was also been

reported in acidophilic ethanol co-producing system [41] and

saline microbial mats [42] .The hydA gene level in C. para-

putrificum M-21 has been reported to have direct correlation

with hydrogen production [52]. Two species of Anaerofilum

viz. Anaerofilum pentosovorans gen. nov., sp. nov., and Anae-

rofilum agile sp. nov isolated from anaerobic bioreactor, are

acidogenic bacteria, use glucose to form acetate, formate,

lactate, ethanol, and carbon dioxide but are not known to

produce hydrogen [53]. They are strictly anaerobic, non spore

producer, rod shaped member of Firmicutes group. Anaerofi-

lum reported in this study formed a separate clade that arose

from Ruminococcus node; a known hydrogen producer [21].

Also the anaerobic firmicutes are known for production of

various organic acids (C3, C4 acids) such as propionic acid,

butyric acid, and their derivatives [54].

5. Conclusion

With the consideration of ethanol as future biofuel, the

molasses spent wash from distillery will emerge as raw

material for hydrogen production, that will affects positively

the process economics. The study proposes that the process

parameters play an important role in eliminating thehydrogen

oxidizing community and enriches with the Fe hydrogenase

gene pool. The study identifies Anaerofilum as a candidate

bacterium for anaerobic hydrogen production. It shows that

with the adjustment in process parameter, the metabolic flux

in hydrogen production has been supported by regular

generation of organic acids, which could provide under

reductive metabolism state the flow of proton to generate the

hydrogen. Study also shows that there is still an accumulation

of volatile fatty acids suggesting that there is further need for

process optimization so that microbial metabolism in such

systems can be engineered in desired way.

Acknowledgements

The authors thank the Directors of IICT, Hyderabad and

NEERI, Nagpur for supporting the collaborative work. Funds

from the Council of Scientific and Industrial Research (CSIR),

New Delhi, are gratefully acknowledged. The work has been

supported by CSIR Network project NWP-19-4.1

Appendix. Supplementary material

The supplementary data associated with this article can be

found in the on-line version at doi:10.1016/j.ijhydene.2011.04.

021.

r e f e r e n c e s

[1] Logan BE. Biologically extracting energy from wastewater:biohydrogen production and microbial fuel cells. Environ SciTechnol 2004;38:160Ae7A.

[2] Yu H, Zhu Z, Hu W, Zhang H. Hydrogen production from ricewinery wastewater in an upflow anaerobic reactor by usingmixed anaerobic cultures. Int J Hydrogen Energy 2002;27:1359e65.

[3] Venkata Mohan S, Bhaskar YV, Murali Krishna P, Rao NC,Babu LV, Sarma PN. Biohydrogen production from chemicalwastewater as substrate by selectively enriched anaerobicmixed consortia: influence of fermentation pH andsubstrate composition. Int J Hydrogen Energy 2007a;32:2286e95.

[4] Venkata Mohan S, Mohanakrishna G, Ramanaiah SV,Sarma PN. Simultaneous biohydrogen production andwastewater treatment in biofilm configured anaerobicperiodic discontinuous batch reactor using distillerywastewater. Int J Hydrogen Energy 2008a;33:550e8.

[5] Tewari PK, Batra VS, Balakrishnan M. Water managementinitiatives in sugarcane molasses based distilleries in India.Resour Conserv Recycling 2007;52:351e67.

[6] Taguchi F, Yamada K, Hasegawa K, Taki-Saito T, Hara K.Continuous hydrogen production by Clostridium sp. StrainNo. 2 from cellulose hydrolysate in an aqueous two-phasesystem. J Ferment Bioeng 1996;82:80e3.

[7] Tadashi M, Hatano T, Yamada A, Matsumoto M.Microaerobic hydrogen production by photosyntheticbacteria in a double-phase photo bioreactor. BiotechnolBioeng 2000;68:647e51.

[8] Heyndrickx M, de vos P, Thibau B, Stevens P, de Lay J. Effectof various external factors on the fermentative production ofhydrogen gas from glucose by Clostridium butyricum strains inbatch culture. Syst Appl Microbiol 1987;9:163e8.

[9] Taguchi F, Chang JD, Takiguchi S, Morimoto M. Efficienthydrogen production from starch by a bacterium isolatedfrom termites. J Ferment Bioeng 1992;73:244e5.

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 6 ( 2 0 1 1 ) 8 2 3 4e8 2 4 2 8241

[10] Ueno Y, Kawai T, Sato S, Otsuka S, Morimoto M. Biologicalproduction of hydrogen from cellulose by mixed anaerobicmicroflora. J Ferment Bioeng 1995;79:395e7.

[11] Lay JJ. Modeling and optimization of anaerobic digestedsludge converting starch to hydrogen. Biotechnol Bioeng2000;68:269e78.

[12] Fang HHP, Liu H. Effect of pH on hydrogen production fromglucose by mixed culture. Bioresour Technol 2002;82:87e93.

[13] Fang HHP, Liu H, Zhang T. Characterisation of a hydrogen-producing granular sludge. Biotechnol Bioeng 2002;78:44e52.

[14] Kalia VC, Jain SR, Kumar A, Josh AP. Fermentation of bio-waste to H2 by Bacillus licheniformis. World J MicrobiolBiotechnol 1994;10:224e7.

[15] Rachman MA, Nakashimada Y, Kakizono T, Nishio N.Hydrogen production with high yield and high evolution rateby self-flocculated cells of Enterobacter aerogenes ina packedbed reactor. Appl Microbiol Biotechnol 1998;49:450e4.

[16] Kumar N, Das D. Enhancement of hydrogen production byEnterobacter cloacae IIT-BT 08. Process Biochem 2000;35:589e93.

[17] The American Public Health Association. Standard methodsfor the examination of water and wastewater. 18th ed.Washington, DC: American Public Health Association; 1998.

[18] Purohit HJ, Kapley A, Moharikar A, Narde G. Extraction ofactivated biological sludge for PCR compatible DNA fromeffluent treatment systems. J Microbiol Methods 2003;52:315e23.

[19] Wang MY, Olson BH, Chang JS. Improving PCR and qPCRdetection of hydrogenase A (hydA) associated with Clostridiain pure cultures and environmental sludges using bovineserum albumin. Appl Microbiol Biotechnol 2007;77:645e56.

[20] ZhangT, FangHHP.Applications of real-timepolymerase chainreactionforquantificationofmicroorganismsinenvironmentalsamples. Appl Microbiol Biotechnol 2006;70:281e9.

[21] Kalia VC, Purohit HJ. Microbial diversity and genomics in aidof bioenergy. J Ind Microbiol Biotechnol 2008;35:403e19.

[22] Subramanian KA, Singal SK, Saxena M, Singhal S. Utilizationof liquid biofuels in automotive diesel engines: an Indianperspective. Biomass Bioenergy 2005;29:65e72.

[23] Naik NM, Jagadeesh KS, Alagawadi AR. Microbialdecolorization of spentwash: a review. Indian J Microbiol2008;48:41e8.

[24] Ruiz G, Jeison D, Chamy R. Development of denitrifying andmethanogenic activities in USB reactors for the treatment ofwastewater: Effect of COD/N ratio. Process Biochem 2006;41:1338e42.

[25] Sparling R, Risbey D, Poggi-Varaldo HM. Hydrogenproduction from inhibited anaerobic composters. Int JHydrogen Energy 1997;22:563e6.

[26] Ueno Y, Haruta S, Ishii M, Igarashi Y. Microbial community inanaerobic hydrogen-producing microflora enriched fromsludge compost. Appl Microbiol Biotechnol 2001;57:555e62.

[27] Lin CY, Lay CH. A nutrient formulation for fermentativehydrogen production using anaerobic sewage sludgemicroflora. Int J Hydrogen Energy 2005;30:285e92.

[28] Yang P, Zhang R, McGarvey JA, Benemann JR. Biohydrogenproduction from cheese processing wastewater by anaerobicfermentation using mixed microbial communities. Int JHydrogen Energy 2007;32:4761e71.

[29] Shin HS, Youn JH, Kim SH. Hydrogen production from foodwaste in anaerobic mesophilic and thermophilicacidogenesis. Int J Hydrogen Energy 2004;29:1355e63.

[30] Zhang T, Liu H, Fang HHP. Biohydrogen production fromstarch in wastewater under thermophilic conditions. JEnviron Manage 2003;69:49e56.

[31] Minnan L, et al. Isolation and characterization of a high H2-producing strain Klebsialle oxytoca HP1 from a hot spring. ReMicrobiol 2005;156:76e81.

[32] Kim MS, Baek JS, Lee JK. Comparison of H2 accumulation byRhodobacter sphaeroides KD131 and its uptake hydrogenaseand PHB synthase deficient mutant. Int J Hydrogen Energy2006;31:121e7.

[33] Koku H, Ero¢glu I, Gundu ZU, Yucel M, Turker L. Kinetics ofbiohydrogen production by the photosynthetic bacteriumRhodobacter spheroids O.U. 001. Int J Hydrogen Energy 2003;28:381e8.

[34] Fang HHP, Liu H, Zhang T. Phototrophic hydrogen productionfrom acetate and butyrate in wastewater. Int J HydrogenEnergy 2005;30:785e93.

[35] de Vrije T, de Haas GG, Tan GB, Keijsers ERP,Claassen PAM. Pretreatment of Miscanthus for hydrogenproduction by Thermotoga elfii. Int J Hydrogen Energy 2002;27:1381e90.

[36] Patel SKS, Purohit HJ, Kalia VC. Dark fermentative hydrogenproduction by defined mixed microbial cultures immobilizedon ligno-cellulosic waste materials. Int J Hydrogen Energy;2010. doi:10.1016/j.ijhydene.2010.03.025.

[37] Vignais PM, Billoud B, Meyer J. Classification and phylogenyof hydrogenases. FEMS Microbiol Rev 2001;25:455e501.

[38] Cammack R. Hydrogenase sophistication. Nature 1999;397:214e5.

[39] Vignais PM, Billoud B. Occurrence, classification, andbiological function of hydrogenases: an overview. Chem Rev2007;107:4206e72.

[40] Posewitz MC, Mulder DW, Peters JW. New frontiers inhydrogenase structure and biosynthesis. Curr Chem Biol2008;2:178e99.

[41] Xing D, Ren N, Rittmann BE. Genetic diversity of hydrogen-producing bacteria in an acidophilic ethanol-h2-coproducingsystem, analyzed using the [Fe]-hydrogenase Gene. ApplEnviron Microbiol 2008;74:1232e9.

[42] Boyd ES, Spear JR, Peters JW. FeFe hydrogenase geneticdiversity provides insight into molecular adaptation ina saline microbial mat community. Appl Environ Microbiol2009;75:4620e3.

[43] Maeda T, Sanchez-Torres V, Wood TK. Enhanced hydrogenproduction from glucose by metabolically engineeredEscherichia coli. Appl Microbiol Biotechnol 2007;77:879e90.

[44] Maeda T, Sanchez-Torres V, Wood TK. Metabolic engineeringto enhance bacterial hydrogen production. MicrobBiotechnol 2008;1:30e9.

[45] Lin CY, Chang RC. Fermentative hydrogen production atambient temperature. Int J Hydrogen Energy 2004;29:715e20.

[46] Taguehi F, Mizukami N, Hasegawa K, Saito-Taki T.Microbial conversion of arabinose and xylose to hydrogenby a newly isolated Clostridium sp. no. 2. Can J Microbiol1994;40:228e33.

[47] Liu G, Shen J. Effects of culture medium and mediumconditions on hydrogen production from starch usinganaerobic bacteria. J Biosci Bioeng 2004;98:251e6.

[48] Collet C, Adler N, Schwitzguebel JP, Peringer P. Hydrogenproduction by Clostridium thermolacticum during continuousfermentation of lactose. Int J Hydrogen Energy 2004;29:1479e85.

[49] Yokoi H, Saitsu A, Uchida H, Hirose J, Hayashi S, Takasaki Y.Microbial hydrogen production from sweet potato starchresidue. J Biosci Bioeng 2001;91:58e63.

[50] Hsiao CL, Chang JJ, Wu JH, Chin WC, Wen FS, Huang CC, et al.Clostridium strain co-cultures for biohydrogen productionenhancement from condensed molasses fermentationsoluble. Int J Hydrogen Energy 2009;34:7173e81.

[51] Ding J, Liu BF, Ren NQ, Xing DF, Guo WQ, Xu JF, et al.Hydrogen production from glucose by co-culture ofClostridium butyricum and immobilized Rhodopseudomonasfaecalis RLD-53. Int J Hydrogen Energy 2009;34:3647e52.

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 6 ( 2 0 1 1 ) 8 2 3 4e8 2 4 28242

[52] Morimoto K, Kimura T, Sakka K, Ohmiya K. Overexpressionof a hydrogenase gene in Clostridium paraputrificum toenhance hydrogen gas production. FEMS Microbiol Lett 2006;246:229e34.

[53] Zellner G, Stackebrandt E, Nagel D, Messner P, Weiss N,Winter J. Anaerofilum pentosovorans gen. Nov., sp. nov., and

Anaerofilum agile sp. nov., two New, strictly anaerobic,mesophilic, acidogenic bacteria from anaerobic bioreactors.Int J Syst Bacteriol 1996;46:871e5.

[54] Ren N, Wang B, Huang JC. Ethanol-type fermentation formcarbohydrate in high rate acidogenic reactor. BiotechnolBioeng 1997;54:428e33.