Journal of Industrial and Engineering Chemistry 19 (2013) 659–664

Biohydrogen production from palm oil mill effluent using immobilizedmixed culture

Lakhveer Singh a, Muhammad Faisal Siddiqui b, Anwer Ahmad c, Mohd Hasbi Ab. Rahim a, Mimi Sakinah b,Zularisam A. Wahid d,*a Faculty of Industrial Sciences & Technology,Universiti Malaysia Pahang (UMP), Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, Malaysiab Faculty of Chemical and Natural Resource Engineering, Universiti Malaysia Pahang (UMP),Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, Malaysiac Department of Civil and Engineering, King Saud University (KSU), PO Box 800, Riyadh 11421, Saudi Arabiad Faculty of Civil Engineering and Earth Resources,Universiti Malaysia Pahang (UMP), Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, Malaysia

A R T I C L E I N F O

Article history:

Received 14 August 2012

Received in revised form 19 September 2012

Accepted 1 October 2012

Available online 8 October 2012

Keywords:

Biohydrogen

Immobilized sludge

Suspended cell

Palm oil mill effluent

A B S T R A C T

Cell immobilization techniques were adopted to bio-hydrogen production using immobilized anaerobic

sludge as the seed culture. Palm oil mill effluent (POME) was used as the substrate carbon source. It was

found that with a POME concentration of 20 g COD/l in the feed, the suspended-cell containing reactor

was able to produce hydrogen at an optimal rate of 0.348 l H2/(l POME h) at HRT 6 h. However, the

immobilized-cell containing reactor exhibited a better hydrogen production rate of 0.589 l H2/(l POME

h), which occurred at HRT 2 h. When the immobilized-cell containing reactor was scaled up to 5 l, the

hydrogen production rate was 0.500–0.588 l H2/(l POME h) for HRT 2–10 h, but after a thermal treatment

(60 8C, 1 h) the rate increase to 0.632 l H2/(l POME h) at HRT 2 h. The main soluble metabolites were

butyric acid and acetic acid, followed by propionic acid and ethanol.

� 2012 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights

reserved.

Contents lists available at SciVerse ScienceDirect

Journal of Industrial and Engineering Chemistry

jou r n al h o mep ag e: w ww .e lsev ier . co m / loc ate / j iec

1. Introduction

Hydrogen is a viable alternative source to replace conventionalfossil fuel due to its clean, renewable and high energy yield (122 kJ/g) nature [1–3]. This yield is 2.75 times higher than energy yields ofhydrocarbon fuel. When hydrogen is used as a fuel, its maincombustion product is water which can be recycled again toproduce more hydrogen. On the other hand, unlike fossil fuelshydrogen gas is not readily available in nature and the commonlyused production methods are quite expensive [4]. In recent yearswidely uses of hydrogen have been demonstrated viz. hydrogen-fueled transit buses, ships and submarines etc., including chemicaland petrochemical applications. Currently, about 98% of hydrogencomes from fossil fuel [5]. Worldwide, 48% production of hydrogenfrom natural gas or steam reforming of hydrocarbon, 30% from oil,18% from coal, and the remaining 4% via water electrolysis.

Abbreviations: PEG, polyethylene glycol; POME, palm oil mill effluent; UASB,

upflow anaerobic sludge blanket; COD, chemical oxygen demand; HRT, hydraulic

retention times; PEG, polyethylene glycol; TN, total nitrogen; VSS, volatile

suspended solids; TS, total solid; GC, Gas chromatograph; TP, total phosphorus;

TN, total nitrogen; VFAs, volatile fatty acids; HBu, butyric acid; HAc, acetic acid;

EtOH, ethanol; TVFA, total volatile fatty acid; HPr, propionic acid.

* Corresponding author. Tel.: +60 95493002.

E-mail addresses: zularisam@ump.edu.my, zularisam@gmail.com (Z.A. Wahid).

1226-086X/$ – see front matter � 2012 The Korean Society of Industrial and Engineer

http://dx.doi.org/10.1016/j.jiec.2012.10.001

However, these processes involved with electricity and whichcomes from fossil fuel combustion so that they are energyexhaustive, expensive and not always environmental friendly.Biological process for hydrogen production is one of the alternativemethods, can be operated at ambient temperatures, pressures, lessenergy intensive and more eco-friendly compared to conventionalchemical method [6]. This process is not only eco-friendly, but alsoescort to open new path for the exploitation of renewable energyresources which are unlimited [7–10]. Biohydrogen productioncan be achieved by dark fermentation or light driven photofermentation and both routes have been extensively reviewed andreported [11,12]. Dark fermentation of organic waste material hadpresented a promising route of biohydrogen production comparedto photosynthetic routes. The major advantages of dark fermenta-tive process are high rate of cell growth, operation without lightsource and no oxygen limitation problems [13,14]. Variousattempts have been made to generate fermentative hydrogenfrom biomass and wastewater like sugarcane bagasse (SCB) [15],wheat straw [16], market waste [17], cheese whey wastewater [18]and dairy waste [19].

In Malaysia, the palm oil industry annually generates about15.2 million tons of wastewater, known as palm oil mill effluent(POME). POME is considered as high strength complex wastewaterwith total chemical oxygen demand that can reach up to 94 kg m-3

[20]. Most of the study on dark hydrogen production from POME

ing Chemistry. Published by Elsevier B.V. All rights reserved.

L. Singh et al. / Journal of Industrial and Engineering Chemistry 19 (2013) 659–664660

wastewater focused on suspended-cell system [21–23]. Immobi-lized-cell systems have become common alternatives to sus-pended-cell systems in continuous operations and can be operatedat high dilution rates (or low retention times) without encounter-ing washout of cells. The immobilization-cell system also giftedwith a feature of creating a local anaerobic environment, which iswell suited to oxygen-sensitive dark hydrogen fermentation. Inaddition, immobilized cells over free cells include more tolerant toenvironment perturbation, process stability, reusable and higherbiological activity since higher cell density can be applied [24].Recent studies found that single cyanobacteria cultures or acombination of cultures have been immobilized in fiber glass,silica-gels and agar for improvement of H2 production [25–28].Photoproduction of H2 for up to 5 months with maximum rate of0.2 ml H2/mg dry weight/h has been observed in a specificallydesigned laboratory-scale photobioreactor with hollow fiberimmobilized Anabaena variabilis [29]. However, it should berecognized that the photosynthetic bacteria immobilization ofentrapment into inorganic and organic polymers has a drawbackdue to the difficulties in the light penetration and transport ofsubstrate and product [30,31]. There are only a few reports aboutthe immobilization of dark fermentation bacteria for improvingthe bio-hydrogen process [32,33]. Usually, entrapment of cells oncarrier such as ethylene vinyl acetate copolymer [32], polyvinylalcohol [34], polymethyl methacrylate [35], or agar gel [36], etc.Nearly all their work showed that immobilized dark fermentativebacteria can enhance and stabilize hydrogen production process.

Polyethylene glycol (PEG) with additional merits was selectedin this work for entrapment due to its simple immobilizationprocedure, low toxicity, good mechanical properties and highlyporous structure that helps to sustain immobilized cell viability[37]. However, information about utilization of immobilized mixedculture (such as POME sludge) in upflow anaerobic sludge blanket(UASB) reactor for fermentative hydrogen production from realwastewater (POME) is not yet available. In this work, acclimatedPOME sludge was entrapped in PEG for continuous hydrogenproduction using POME as the substrate. The lab scale UASBreactors containing the immobilized biomass were operated atdifferent hydraulic retention times (HRTs) to assess their perfor-mance in hydrogen production and comparison with suspendedcell loaded reactor.

2. Experimental

2.1. POME

Fresh POME was collected from the receiving tank of an oil palmmill in Kilang Sawit Lepar Hilir Pahang, Malaysia and kept at 4 8Cprior to use. The temperature of the discharge POME range from 65to 70 8C and pH 4.0–5.0. The POME was fully characterized, asshown in Table 1.

Table 1Chemical characteristics of palm oil mill effluent used in this study.

Parameter Concentration (mg/l)

Biochemical oxygen demand (BOD) 22,100–54,200

Chemical oxygen demand (COD) 75,100–96,300

pH 4.0–5.0

Total carbohydrate 16,200–20,000

Total nitrogen 820–910

Ammonium–nitrogen 25–30

Total phosphorus 95–120

Phosphorus 14–20

Oil 80,100–10,500

Total solid 35,000–42,000

Suspended solids (SS) 8400–12,000

All values are in mg/l except pH

2.2. Seed sludge

The seed was obtained from a same palm oil mill wastewatertreatment plant used as inoculums. The collected sludge wassieved using 500 mm mesh. Hydrogen productivity of the sludgewas increased by heat treatment at 80 8C for 50 min [38]. The pHwas restored to 6.0 by 0.1 N NaOH and the 2 l volume of filtrate wastransferred to the fermentor. The initial volatile suspended solid(VSS) and total solid (TS) concentration of the sludge were 8.0 and11.4 g/L, respectively. Before subjected to immobilization, the heattreated sludge was acclimated with synthetic wastewater in aUASB reactor at 37 8C.

2.3. Preparation of PEG immobilized H2-producing sludge

Ten gram of PEG, 1 g N,N0-methylenebisacrylamide (MBA)crosslinker and 50 ml of distilled water was carefully heated to40 8C to completely dissolve the PEG. The solution than was cooleddown to below 30 8C. One portion of centrifuge H2-producingPOME activated sludge at 2000 rpm for 15 min (100 ml � 100 g)and one portion of PEG solution (100 ml) mentioned abovethoroughly mixed. To start polymerization, 0.5 g potassiumpersulfate initiator (K2S2O8) was mixed in and the mixture allowedto stand for about 20 min to promote bead formation. The resultingimmobilized sludge bead was cut into 3 mm beads (density 1.42 g/cm3) shown in Fig. 1. The biomass content of the immobilizedbeads was �10 mg VSS/g bead. The experiments were performedunder anaerobic conditions at room temperature.

2.4. Set- up and operations of UASB reactor for hydrogen production

UASB reactors with a (working volume of 500 ml or 5 l) wereused in this study. The reactors were flushed with oxygen-freenitrogen gas for 20 min to established anaerobic condition [22].The pH of the medium was controlled at 5.5 with 1 M NaOH and1 M HCl, constantly mixing and temperature maintained at 37 8Cthrough water jacket. Immobilized sludge beads (100 g or 200 g)were placed in a UASB reactor (500 ml or 5 l) containingaforementioned medium. The feed was pumped into the reactorusing a peristaltic pump (Masterflex L/S, Cole Palmer Instrument,USA). During the reaction phase, the reactor was intermittentlymixed with liquid recirculation to provide better distribution ofbiomass and improve contact of microflora with wastewater andalso prevented inhibitory effect of oxygen from influent on strictlyanaerobic mixed culture in reactor [21]. The gas line wasconnected to the gas holder to measure the daily biogasproduction by water displacement method. The composition ofbiogas and soluble metabolites produced during hydrogen

Fig. 1. PEG-immobilized sludge pallets.

Fig. 2. HRT-dependent profiles of biogas production rates, volatile fatty acid (VFA)

production, and substrate utilization in the effluent with suspended-cell containing

reactor.

L. Singh et al. / Journal of Industrial and Engineering Chemistry 19 (2013) 659–664 661

fermentation was determined at designated time intervals. Thereactor was operated at HRT of 1–12 h by adjusting the volumetricflow rate of the feed.

2.5. Analytical methods

pH, VSS and TS were analyzed according to the Standardmethods [39]. HACH programs, 430 COD LR (low range) and 435COD HR (high range/high plus) was used for COD analysis. A gaschromatograph (GC 8500 PerkinElmer) equipped with a thermalconductivity detector and a 2 m stainless-steel SS350A columnpacked with a molecular sieve (80/100 mesh) was used todetermined the fraction of H2 in biogas using nitrogen as a carriergas at a flow rate of 25 ml/min. The operating temperatures of thedetector, injection port and oven were 150 8C, 100 8C, and 80 8Crespectively. The alcohol and volatile fatty acid (VFA) contents offiltered samples (0.2 mm) were analyzed on the GC using a flameionization detector equipped with a Nukol capillary column. Thedry weight of immobilized cell in immobilized beads was assessedby measuring the difference in dry weight between the biomass-loaded beads and the beads alone.

3. Results and discussion

3.1. Continuous hydrogen production

UASB reactors with a working volume of 500 ml were operatedcontinuously to produce hydrogen from POME wastewater usingsuspended and immobilized-cells. The effects of HRTs on thebiogas production rate, composition of biogas and solublemetabolites, and substrate utilization efficiency were examinedduring the course of hydrogen production. The data for suspendedand immobilized-cell reactors are showed in Figs. 2 and 3,respectively. The correlations between HRTs and hydrogengeneration efficiency as well as substrate utilization efficiencywere summarized in Table 2. In all cases, the biogas primarilyconsisted of CO2 and H2, while CH4 was undetectable. Substrateutilization efficiency was essentially within range of 60–91% at allHRTs used (Table 2). Typical soluble metabolites included butyricacid (HBu), acetic acid (HAc), propionic acid (HPr), and ethanol(EtOH).

3.2. Hydrogen production performance with suspended-cell system

3.2.1. Decrease in HRT from 12 to 6 h

Fig. 2 shows that the suspended-cell reactor was started fromHRT 12 h, at which the steady-state hydrogen generation rate was0.275 l H2/(l POME h). As the HRT was decreased to 6 h, hydrogenproduction rate increased to 0.348 l H2/(l POME h) and thenslightly declined. Production of EtOH dominated the solublemetabolites initially, when hydrogen generation was not signifi-cant, it then sharply decreased as production rate exceeded 0.2 lH2/(l POME h) and VFA became major soluble products. Solventformation is unfavorable to anaerobic production of hydrogen gasdue to loss of electrons protons to the formation of more reducedproduct [40].

3.2.2. Decrease in HRT from 6 to 1 h

As the HRT was decreased from 6 to 2 h and then to 1 h, thehydrogen production rate did not enhance, while CO2 (and totalbiogas) production increased significantly in response to higherorganic loading rate for lower HRTs (Fig. 2). The hydrogen contentin biogas reduced to less than 15% at HRT 1 h, in contrast to 41% ofH2 at HRT 6 h (Table 2). The hydrogen generation rate decreased toas low as 0.192 l H2/(l POME h). Meanwhile, formation of HPr rosedramatically when HRT shifted down from 6 to 1 h, while

composition of HBu and HAc did not vary considerably duringthis period (Fig. 2). Analysis of the variations in composition ofbiogas and soluble metabolites suggests that operation at a highcarbon substrate (POME) loading rate (or a low HRT) led toconversion of POME to CO2 by bacterial populations that wereinefficient in hydrogen production, but somehow dominatedsubstrate utilization to produce CO2 and HPr as the major gasand soluble products. Therefore, HPr was most likely produced bybacterial populations other than the primary hydrogen producer inthe culture. Thus, observation of an overwhelming HPr productionmay be considered as a signal of inefficient hydrogen fermentationwith the sludge culture. In addition, when HRT was adjusted from 2to 1 h, the biomass concentration in the liquid effluent abruptlyincreased, suggesting that significant washout of cells occurred at

Fig. 3. HRT-dependent profiles of biogas production rates, volatile fatty acid (VFA)

production, and substrate utilization in the effluent with immobilized-cell

containing reactor.

Table 2Performance of hydrogen fermentation in suspended and immobilized-cell reactor

at different hydraulic retention times (HRTs).

Reactor system HRT

(h)

Hydrogen

content in

biogas (%)

Hydrogen

production

rate (l H2/l

POME/h)

Substrate

utilization

efficiency (%)

Suspended-cell

system

12 39.0 0.255 58

6 36.5 0.348 60

2 22.5 0.244 57

1a 12.1 0.091 53

6a 23.1 0.312 59

2a 18.2 0.302 55

Immobilized-cell

system

6 39.1 0.257 83

2 37.1 0.589 87

1 33.2 0.102 66

2a 28.4 0.419 87

1a 34.1 0.532 75

Immobilized scale-up

reactor

10 32 0.553 91

6 29 0.587 90

2 26 0.581 83

8 19 0.011 61

2a 28 0.632 81

a Repeated HRT (the same HRT was used earlier in this run).

L. Singh et al. / Journal of Industrial and Engineering Chemistry 19 (2013) 659–664662

HRT = 1 h. Thus, operation at HRT = 1 h appeared to be unstable forthe suspended-cell reactor.

3.2.3. Increase in HRT from 1 to 6 h

Since operation at HRT 1 h resulted in low hydrogengeneration efficiency and a considerable washout of the cells,the HRT was readjusted to 6 h. It was found that production ofhydrogen increased to 0.214 l H2/(l POME h) (Fig. 2). Thehydrogen content also increased as hydrogen accounted for 25%of the biogas during this period (Table 2). In the meantime, theproduction of HBu and HAc increased considerably, accompa-nied by decreases in HPr and EtOH, indicating that the hydrogengenerating population revived. Interestingly, HPr productiondecreased when HRT increased from 1 to 2 h, implying that theactivity of HPr producing bacteria in the suspended-cellcontaining reactor were stimulated only when substrate feedingrate exceeded a high level.

3.2.4. Decrease in HRT from 6 to 2 h

After the reactor reached steady state for a prolonged period atHRT = 2 h, the HRT was subsequently lowered to 1 h to examinehow the reactor would respond to an increased organic loadingrate at this time. Again, production of total biogas and CO2

increased due to faster supply of substrate, but the hydrogenconcentration in biogas became lower (19%), in contrast to 21.3% atHRT 6 h (Table 2). Consequently, the hydrogen generation rate atHRT 2 and 6 h was quite same. Overall assessment of the data inFig. 2 and Table 2 shows that optimal HRT was 6 h for suspendedsludge containing USAB reactors in continuous hydrogen produc-tion. When the reactor was operated at HRT 6 h, the best hydrogenproduction rate was 0.348 l H2/(l POME h) (Fig. 2).

3.3. Performance of immobilized-cell UASB reactor

Fig. 3 shows that the steady-state hydrogen production rate ofPEG immobilized-cell reactor increased from 0.291 to 0.589 l H2/(lPOME h) when HRT was shifted down from 6 to 2 h. During thisstage, the hydrogen content in the biogas was around 40–45%(Table 2) and the major VFA product was clearly HBu. As the HRTwas reduced further to 1 h, a slight increase in CO2 production wasobserved. However, the hydrogen production was not enhancedbut rather decreased slightly to lower hydrogen content (32%). Thistrend is similar to what was observed in suspended cell containingUASB reactor when HRT was decreased to 1 h (Fig. 2), except thatthe increase of CO2 production in immobilized reactor correspond-ing to a HRT drop was much less significant than that occurred insuspended cell containing reactor. Moreover, HPr was not thedominant VFA product as it was in suspended-cell containingreactor. Therefore, the immobilized reactor was relatively stablewhen it was operated a low HRT (1 h). Accidental power failureoccurred at 75 h when the immobilized reactor was operated atHRT 1 h (Fig. 3). The feeding stream was stopped for 15 h, and thereactor was then restarted at HRT 2 h. Evolution of hydrogenrecovered immediately after the feeding was restored andhydrogen production rate reached a steady-state value of 0.249 lH2/(l POME h) (Table 2).

The HRT was then switched to 1 h again and production rateincreased further to a peak value of ca. 0.532 l H2/(l POME h), andthen started to decrease as the operation at HRT of 1 h continued.During the periods of efficient hydrogen production (25–60 h and

L. Singh et al. / Journal of Industrial and Engineering Chemistry 19 (2013) 659–664 663

100–125 h), HBu was the most abundant acid product, followedby HAc and HPr, while production of EtOH was insignificant. Incontrast to hydrogen production rate, the substrate utilizationefficiency did not vary significantly with change in HRT, but stillreached highest level of 87.5% when the reactor was operated atHRT of 6 h and 2 h, respectively, which were higher thansuspended-cell process. Immobilized-cell UASB reactor showeda better performance on hydrogen production in comparison tosuspended-cell reactor (Table 2). Despite operation at a HRT of 2 h,immobilized cell UASB reactor attained a maximum hydrogenproduction rate of 0.589 l H2/(l POME h), which was 55% higherthan suspended-cell reactor in this study. The hydrogen content inbiogas and substrate utilization efficiency values were alsogreater then those obtained from the suspended-cells reactor(Table 2).

Fig. 4. Performance of hydrogen fermentation during scale-up operations of

immobilized cell-containing UASB reactor (working volume = 5 l).

3.4. Scale-up of immobilized cell containing reactor

To assess the feasibility of using PEG-immobilized cellscontaining reactor for large scale hydrogen production, the reactorwas scaled up to 5 l working volume. The scale up experiment wascarried out under identical condition to those used for the smallerreactors. As indicated in Fig. 4 and Table 2 with the 5 l reactor, thehydrogen production rate nearly constant at 0.589 l H2/(l POME h)when the HRT was 6–10 h. Whereas further decrease in HRT to 2 hresulted in a significant decrease in hydrogen production rate andhydrogen content (Table 2). This may be due to an overload ofmedium causing a substrate inhibition effect on hydrogenproducers. As CO2 did not decrease along with the decrease inhydrogen, it is likely that some non-hydrogen producers started todominate the culture at high loading rates (i.e., low HRTs) andconverted the substrate to CO2 without hydrogen production. Thisspeculation was supported by the restoration of hydrogenproduction performance after the culture was subject to a heattreatment, in which the culture was heated to 60 8C for 1 h at HRT2 h to inhibit the non-hydrogen-producing bacteria activity [41].

After treatment hydrogen production rate reach to maximumvalue 0.632 l H2/(l POME h) and hydrogen content in biogas alsoimproved at HRT 2 h. The concentration of soluble metabolitesproduced during conversion of POME to hydrogen in the 5 limmobilized reactor essentially followed the order ofHBu > HAc > HPr > HAc > EtOH, comparable to what wasobtained in the 500 ml immobilized reactor. The total quantityof soluble metabolites decreased considerably as the HRT wasshortened (Fig. 4), probably due to the increase in the dilution rate.This result clearly suggested that using PEG-immobilized cellsUSAB reactor might reduced the operational cost by gaining acomparable hydrogen producing capacity at a low HRT (or a widerange of OLR).

4. Conclusions

Effective hydrogen production from POME wastewater wasattained using immobilized and suspended-cell culture in UASBreactors with optimal hydrogen generation rate of 0.589 l H2/lPOME h (HRT 2 h) and 0.348 l H2/l POME h (HRT 6 h), respectively.The PEG-immobilized biomass loaded reactor seems to be a betterchoice of the two for continuous hydrogen fermentation since itexhibited higher hydrogen production rate, and also showed morestability when it was operated at low hydraulic retention time(HRTs). Over the HRT range 1–12 h, the hydrogen production ratedid not always increase along with decreases in HRT, while theoptimal HRTs for immobilized and suspended-cell reactors were 2and 6 h, respectively. When the reactors were operated at a HRTlower than their optimal values, poor hydrogen productionoccurred due to loss of competition with non-hydrogen-producingmicroorganism for substrates. Thus, immobilized-cell developedhere may have potential for hydrogen production system in largerscale than the production from suspended-cell systems.

Acknowledgements

The authors are thankful to the Postgraduate Research Scheme(PGRS) (Grant No. GRS-110332), Universiti Malaysia Pahang (UMP)for financial supports.

References

[1] N. Rashid, W. Song, J. Park, H.F. Jin, K. Lee, Journal of Industrial and EngineeringChemistry 15 (2009) 498.

[2] W. Chatpaisalsakul, O. Mekasuwandumrong, J. Panpranot, C. Satayaprasert, P.Praserthdam, Journal of Industrial and Engineering Chemistry 15 (2009) 77.

L. Singh et al. / Journal of Industrial and Engineering Chemistry 19 (2013) 659–664664

[3] Y.J. Lee, Y.M. Kim, H.R. Jeong, M.S. Kang, Journal of Industrial and EngineeringChemistry 14 (2008) 655.

[4] I.K. Kapdan, F. Kragi, Enzyme and Microbial Technology 38 (2009) 271.[5] Y. Kalinci, A. Hepbasli, I. Dincer, International Journal of Hydrogen Energy 34

(2009) 8799.[6] J. Wang, W. Wan, International Journal of Hydrogen Energy 34 (2009) 799.[7] J.R. Benemann, International Journal of Hydrogen Energy 22 (1997) 979.[8] K. Sasikala, ChV. Ramana, P.R. Rao, K.L. Kovacs, Advances in Applied Microbiology

10 (1993) 211.[9] K. Miyamoto, O. Wable, J.R. Benemann, Biotechnology Letters 10 (1989) 703.

[10] S. Tanisho, N. Wakao, Y. Kokako, Journal of Chemical Engineering of Japan 16(1983) 529.

[11] S. Manish, R. Banerjee, International Journal of Hydrogen Energy 33 (2008) 279.[12] Y.Z. Tao, Y. Chen, Y.Q. Wu, Y.L. He, Z.H. Zhou, International Journal of Hydrogen

Energy 32 (2007) 200.[13] D.B. Levin, L. Pitt, M. Love, International Journal of Hydrogen Energy 29 (2004)

173.[14] K. Nath, D. Das, Applied Microbiology and Biotechnology 65 (2004) 520.[15] S. Pattra, S. Sangyoka, M. Boonmee, A. Reungsang, International Journal of

Hydrogen Energy 36 (2008) 5256.[16] Y. Chu, Y. Wei, X. Yuan, X. Shi, Bioresource Technology 102 (2011) 3805.[17] L. Jaitalee, S. Dararat, O. Chavalparit, EnvironmentAsia 3 (2010) 115.[18] N. Azbar, F.T.C. Dokgo, T. Keskin, K.S. Korkmaz, H.M. Syed, International.Journal of

Hydrogen Energy 34 (2009) 7441.[19] S. Venkata Mohan, V. Lalit Babu, P.N. Sarma, Enzyme and Microbial Technology 41

(2009) 506.[20] A. Abdul-Raof, A. Ohashi, H. Harada, Journal of Environmental Systems and

Engeering 776 (2004) 115.[21] S. O-Thong, P. Prasertsan, N. Intrasungkha, S. Dhamwichukorn, N.K. Birkeland,

Enzyme and Microbial Technology 41 (2007) 583.[22] M.L. Chong, R.A. Rahin, Y. Shirai, M.A. Hassan, International Journal of Hydrogen

Energy 34 (2009) 764.

[23] I. Ismail, M.A. Hassan, N. Abdul Rahman, S.S. Chen, Biomass and Bioenergy 34(2010) 42.

[24] E. Seol, A. Manimaran, Y. Jang, S. Kim, Y.K. Oh, S. Park, International Journal ofHydrogen Energy 36 (2011) 8681.

[25] T.V. Laurinavichene, A.S. Fedorov, M.L. Ghirardi, M. Seibert, A.A. Tsygankov,International Journal of Hydrogen Energy 3 (2006) 659.

[26] D.J. Dickson, C.J. Page, R.L. Ely, International Journal of Hydrogen Energy 34 (2009)204.

[27] N. Rashid, W. Choi, K. Lee, Biomass and Bioenergy 36 (2012) 241.[28] W. Song, N. Rashid, W. Choi, K. Lee, Bioresource Technology 102 (2011) 8676.[29] S.A. Markov, M.J. Bazin, D.O. Hall, Enzyme and Microbial Technology 17 (1995) 306.[30] A.A. Tsygankov, Y. Hirata, M. Miyake, Y. Asada, J. Miyake, Journal of Fermentation

and Bioengineering 77 (1994) 575.[31] M. Vincenzini, R. Materassi, M.R. Tredici, G. Florenzano, International Journal of

Hydrogen Energy 7 (1982) 725.[32] C.N. Lin, S.Y. Wu, J.S. Chang, J.S. Chang, Bioresource Technology 100 (2009) 3298.[33] A.R. Barros, E.D.C. de Amorim, C.M. Reis, G.M. Shida, E.L. Silva, International

Journal of Hydrogen Energy 35 (2010) 3379.[34] J.O. Kim, Y.H. Kim, J.Y. Ryu, B.K. Song, I.H. Kim, S.H. Yeom, Process Biochemistry 40

(2005) 1331.[35] K.J. Wu, J.S. Chang, Process Biochemistry 42 (2007) 279.[36] H. Yokoi, T. Tokushige, J. Hirose, S. Hayashi, Y. Takasaki, Journal of Fermentation

and Bioengineering 83 (1997) 481.[37] E.J.T.M. Leenen, V.A.P. Dos Santos, K.C.F. Grolle, J. Tramper, R.H. Wijffels, Water

Research 33 (1996) 2985.[38] F.Y. Chang, C.Y. Lin, International Journal of Hydrogen energy 29 (2004) 33.[39] APHA, Standard Methods For Examination of Water and Wastewater, 19th Ed.

Washington, DC, USA, American Public Health Association. 1998, P. 1220.[40] R.T. Yan, C.X. Zhu, C. Golemboski, J.S. Chen, Applied and Environment Microbiol-

ogy 54 (1988) 642.[41] C.Y. Chu, S.Y. Wu, P.C. Hsieh, C.Y. Lin, International Journal of Hydrogen Energy 36

(2011) 14078.