A pilot-scale study of biohydrogen production from distillery effluent using defined bacterial co-culture

international journal of hydrogen energy 33 (2008) 5404–5415

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/he

A pilot-scale study of biohydrogen production from distillery effluent using defined bacterial co-culture T.M. Vatsala*, S. Mohan Raj1, A. Manimaran2 Shri AMM Murugappa Chettiar Research Centre, Photosynthesis and Energy Division, Tharamani, Chennai, India, 600 113

article info

abstract

Article history:

We evaluated the feasibility of improving the scale of hydrogen (H2) production from sugar

Received 15 February 2008

cane distillery effluent using co-cultures of Citrobacter freundii 01, Enterobacter aerogenes E10

Received in revised form

and Rhodopseudomonas palustris P2 at 100 m3 scale. The culture conditions at 100 ml and 2 L

4 July 2008

scales were optimized in minimal medium and we observed that the co-culture of the

Accepted 7 July 2008

above three strains enhanced H2 productivity significantly. Results at the 100 m3 scale

Available online 20 September 2008

revealed a maximum of 21.38 kg of H2, corresponding to 10692.6 mol, which was obtained through batch method at 40 h from reducing sugar (3862.3 mol) as glucose. The average

Keywords:

yield of H2 was 2.76 mol mol1 glucose, and the rate of H2 production was estimated as

Biohydrogen

0.53 kg/100 m3/h. Our results demonstrate the utility of distillery effluent as a source of

Distillery effluent

clean alternative energy and provide insights into treatment for industrial exploitation.

Pilot-scale H2 production

ª 2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.

COD BOD Enterobacter aerogenes Citrobacter freundii Rhodopseudomonas palustris

1.

Introduction

Hydrogen (H2) is a sustainable, non-polluting source of energy. Owing to its potential use, it has been considered as a prominent future fuel and energy vector. The United States and the European countries have already committed to establish an energy economy based on H2 from fully sustainable or renewable sources. Based on the national H2 program of the United States, the contribution of H2 to total energy market will be w8–10% by 2025 [1]. The market for H2 and technologies for its production are potentially huge, and the level of investment into H2 research has increased significantly in recent years,

and H2 may, in the long-term, replace all other fuels. However, cost is currently the biggest impediment to H2 fuel production. H2 is produced from nonrenewable energy sources, such as oil, natural gas, and coal that require expensive energy input [2]. Thus, the development of cost-effective and/or economically viable processes is necessary for increasing the achievable dominance of H2 as an energy source. As a consequence, scientists are pursuing biological systems as a way to meet increasing energy demand. Biological H2 production stands out as an environmentally friendly process carried out under mild operating conditions that include fermentation of sugars, photolysis of water and water gas shift reactions [3–6].

* Corresponding author. Present address: R&D, Hydrolina Biotech Private Ltd., 406, TICEL Biopark, Tharamani, Chennai, India, 600 113. Tel.: þ91 44 2254 1199; fax: þ91 44 2471 8155. E-mail address: [email protected] (T.M. Vatsala). 1 Present address: Department of Chemical and Biochemical Engineering, Pusan National University, Busan 609 735, Republic of Korea. 2 Present address: Department of Biotechnology and Food Technology, Durban University of Technology, Durban 4000, South Africa. 0360-3199/$ – see front matter ª 2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhydene.2008.07.015

international journal of hydrogen energy 33 (2008) 5404–5415

Biological H2 production is the most challenging and exigent prefecture of renewable energy technology development, especially where it involves the treatment of organic waste. H2 can also be produced by bacteria from renewable biomass. Several studies on biohydrogen production from various substrates, including wastes [7–10], have been reported at laboratory level. Vatsala and Seshadri [7] have studied the enzymatic hydrolysis of silk-cotton waste into H2 and volatile fatty acids (VFAs) using Rhodospirillum rubrm ATCC 11170. Kapdan and Kargi [8] have published an excellent review on the use of various substrates, particularly simple sugars, corn starch wastes, sweet potato, food industrial wastes, wastewater, and waste sludge, for H2 production under dark and phototrophic conditions. While, chemical wastewater was used successfully for biohydrogen production at lab-scale [9]. Yu et al. [10] have demonstrated the continuous production of H2 from anaerobic acidogenesis of a highstrength rice winery wastewater by a mixed bacterial flora at 3 L upflow reactor. Various types of bioreactors with different capacities have also been used for H2 production, including batch, fed-batch and continuous-flow, stirred tank reactors [11–14]. The concept of waste utilization through a microbial fermentation process represents an ideal alternative technology for coordinating pollutant elimination and energy production. In this sense, biohydrogen is considered an intriguing ‘‘future fuel’’. The energy required for aerobic wastewater treatment is huge, but can be reduced through economically viable biohydrogen production processes. We have been studying H2 production from various waste materials [15,16]. Among the waste materials examined, distillery effluent was the most suitable substrate for H2 production [17]. H2 production from distillery effluent was investigated at lab-scale and pilot-scale [17]. The seasonal variations on the molasses quality, particularly sugar content, sulfur content, and their influences on H2 production using individual and/or co-culture of Citrobacter freundii and/or Rhodopseudomonas palustris P2 were investigated [17]. We also reported H2 production from distillery effluent at 10 m3 volume and the produced H2 was subsequently tested for fuel cell application [18]. Here, the present investigation focuses on developing a cost-effective, viable process for producing clean H2 on a pilot-scale (100 m3) from distillery effluent. Apart from H2 as an energy source, the treatment of distillery waste was also accomplished, with a drastic reduction in chemical oxygen demand (COD) and biological oxygen demand (BOD). To the best of our knowledge, this is the first report on H2 production at 100 m3 scale.

2.

Materials and methods

2.1.

Materials

The biochemical characterization kits for bacterial identification were purchased from Hi-Media (Mumbai, India). The genomic DNA isolation kit was procured from Promega (Madison, WI, USA). DNA gel purification kit was purchased from Qiagen (Mannheim, Germany). Primers were synthesized by Cosmotech Co. Ltd., Korea. All other chemicals, unless otherwise indicated, were obtained from E-Merck (Mumbai, India) and/or Sigma–Aldrich (St. Louis, MO, USA).

2.2.

5405

Microorganisms and culture conditions

Heterotrophic bacterial strains (HTB) were isolated from the effluent treatment plant, EID Parry (I) Ltd., Nellikuppam, Tamil Nadu, India. Photoheterotrophic bacterial strains (PTB) were isolated from pond soil, Chengalpet, Tamil Nadu, India. Ormerod’s minimal medium [19] was used for the regular maintenance of photoheterotropic isolates. To measure the H2 production efficiency of the PTB isolates, experiments were conducted in serum bottles (125 ml total volume (tv); 100 ml working volume (wv)) at 37  C, under photoheterotrophic conditions at 7000 lux intensity with an incandescent light source. Malate at 50 mM was used as a carbon source for PTB isolates. Heterotrophic isolates were cultured in H2 production medium, which contained (per litre): 1.0 g yeast extract, 1.0 g NaCl, 1.0 g sodium thiosulphate and 10 g glucose, under dark anaerobic conditions for 24 h. This medium was fortified with a potassium phosphate buffer (50 mM; pH 7.0), and the experiments were conducted in serum bottles at 37  C. Before inoculating the seed culture (HTB and/or PTB), the bottles were flushed with argon (Ar) gas (99.9%) for 10 min to ensure that the bottles were completely deprived of oxygen (O2). Bottles were then sealed with butyl rubber septa and aluminum caps. Isolates were tested for H2 production efficacy with specific media. The best H2 producing strains, HTB01, HTB10, and PTB2, were subjected to physiochemical characterization using the methodologies described previously [20]. To further confirm the identification of bacterial isolates, the 16S rRNA genes from the strains HTB01, HTB10, and PTB2 were amplified by using the primers: 27F (50 -AGA GTT TGA TCM TGG CTC AG-30 ) and 1492R (50 -TAC GGY TAC CTT GTT ACG ACT T-30 ) and sequenced by Cosmotech, Co. Ltd., Korea. Sequence alignment and analysis of similarity of the 16S rRNA genes were performed with CLUSTAL W [21] and DNA baser RC2 (v2.9.97) programs. A similarity search was done in the Ribosomal Database Project (RDP) [22]. To determine the efficiency of the selected microorganism for H2 production, several experiments were conducted at a lab-scale with 125 ml serum bottles and a glass bioreactor (Bioengineering, KLF, 2000, SW). Axenic strains were acclimatized on distillery effluent in the laboratory under dark anaerobic conditions to grow and produce H2 for several generations.

2.3.

Bioreactors designing, fabrication and operation

The feasibility of H2 production and effluent treatment was evaluated at the 100 m3 scale. The lab-scale process of H2 production was then scaled-up from 1 L to 100,000 L. As shown in Fig. 1, a sequence of bioreactors was constructed for inoculum preparation for the 100 m3 reactor. The total volumes of the reactors were 0.125, 1.25, 12.5 and 125 m3, and the working volumes were set from 0.1 to 100 m3. The heightto-diameter (H/D) ratio of each reactor was set to 1.28. All the reactors were fabricated using mild steel with epoxy coating to prevent chemical and/or biological corrosion. Temperature and pH were monitored using specialized probes for 10 and 100 m3 reactors. However, temperature and pH were not controlled. A 10% inoculum was used for each reactor. The

5406

international journal of hydrogen energy 33 (2008) 5404–5415

F001

TI

GV-001-50

SW-002-50

F002 F002

TI

pH SG

F003 F003

TI

P-F

P1 P-TE PI

pH SG

PI

SW-001-100

PI

PI pH SG

V-001-50

V-002-25

GV-002-25

V-003-25

SW-003-25

SW-004-25

GV-003-25

PI

PI pH SG

VENT VENT

SW-005-150 V-004-25

GV-004-25

GV GV-005-80

F004

P-F

TI

SPENT WASH

IT-003-50 IT-003-50

D-003-50

D-003-50

D-004-25

D-003-50

P-SW-1

C TO ETP

P-IT

P-F

P I

D-005-150 D-005-150

P-TE

D

P-C

PI

MIXING TANK TANK MIXING

P-F

- Pump for fermentor inlet

C

- Caustic soda

IT

V

- Vent

GV - Gas Vent

P-TE - Pump for treated effluent

D

- Drain

TI

P-T

F

- Fermentor

P-C - Pump for caustic soda

F001 - 0.125 m3 reactor

SG

- Sight glass

PI

- Pressure indicator

F002 - 1.25 m3 reactor

pH

- pH meter

F003 - 12.5 m3 reactor-

SW - Spent wash

- Inoculum transfer - Temperature indicator

- Pump for inoculum transfer

F004 - 125 m3 reactor

Fig. 1 – Schematic representation of biohydrogen production plant layout – reactor design and process diagram.

effluent/medium in the 10 and 100 m3 reactors was circulated using 5 m3 and 50 m3 per hour pumps, respectively, from the bottom of each reactor to the top. For the initial pH adjustment, caustic lye (48%) was used with non-corrosive polypropylene pipes and pumps. The samples were analyzed periodically for sugar, COD, BOD and VFAs. The initial temperature of the effluent was 60  C when loading into the bioreactor and thus the distillery effluent was cooled through internal circulation. The off-gas produced from each reactor was passed through wet gas flow meters (DM3D, DM3G; GH Zeal’s; UK) to quantify evolved gas.

2.4.

Analytical methods

Cell concentration in minimal medium was measured in 10mm-path-length cuvette using a spectrophotometer (Spectronic 20, USA) and biomass concentration was determined by measuring cell dry weight using a pre-determined correlation between OD (1 OD600 ¼ 0.3 g l1 cdw of washed cells) and cell dry weight. Since distillery effluent is dark brown in color, having an excess of suspended and dissolved solids, a total viable count (TVC) by pour plating method was adopted to analyze the microbial load every 6 h at the 100 m3 scale. For TVC analysis, selective HTB and PTB media were used to detect the viability of cells during the fermentation process. One millitre of sample was serially diluted from 101 to 107 with 0.85% sterile saline water and plated on selective media.

HTB agar plates were incubated at 37  C for 24 h under dark and PTB agar plates were overlaid with sterile mineral oil (#161403; Sigma–Aldrich, USA) and incubated at 37  C for 72 h under incandescence illumination with 7000 lux intensity. Glucose or reducing sugars were quantified by the method described by Miller [23]. Ethanol and other organic acids were detected by HPLC using refractive index (RI) and photodiode array (PDA) detectors (HPLC, Agilent 1100 series, USA). The supernatants, obtained by centrifugation of culture samples at 10,000  g for 5 min, were filtered through 0.2 mM Tuffrynmembranes (Acrodisc, Pall Life Sciences) and samples were eluted through a 300 mm  7.8 mm Aminex HPX-87H column (Bio-Rad, USA) at 50  C with 5.4 mM H2SO4. The flow rate was set at 0.5 ml min1 and the flow cell temperature was set to 35  C. Fresh distillery effluent was collected from EID Parry (I) Ltd., Nellikuppam, Tamil Nadu, India. The samples were systematically analyzed physicochemically for pH, COD, BOD, total solids, total volatile solids, total dissolved solids and total suspended solids [24]. Total sugar was analyzed by the method described by Dubois et al. [25]. H2 and CO2 contents were measured using a Gas Chromatograph (Chemito, 3800) equipped with thermal conductivity detector (TCD) and a Porapak Q column (AW, 80–100 mesh; 6 m  100  1800 ). N2 was used as a carrier gas at 30 ml min1. The injector, oven and detector temperatures were set at 70, 50 and 100  C, respectively [26].

5407

international journal of hydrogen energy 33 (2008) 5404–5415

3.

Results and discussion

3.1. Identification of bacterial strains by phenotypic and genotypic characterization Based on the H2 producing ability, only three strains were subjected to phenotypic and genotypic characterization. The results shown in Table 1 indicated that both HTB strains (HTB01 and HTB10) were fit into the family Enterobacteriaceae. In particularly, the distinguishing characteristics, such as growth on salicin, H2S test, lysine decarboxylase, citrate utilization, nitrate reduction, and other IMViC tests (Indole production, methyl red, Voges-Proskauer, and citrate utilization), indicated that the strains HTB01 and HTB10 were similar to C. freundii and Enterobacter aerogenes, respectively [27,28]. The other strain PTB2 was able to grow anaerobically under dark and light. The strain PTB2 exhibits dark reddish brown pigments when light was supplied. The characteristic absorption maxima of living cells of PTB2 at 805 and 875 nm (bacteriochlorophyll (BChl) type a) indicated that this strain belongs to purple bacteria. Also, the principal mode of photosynthesis of strain PTB2 was observed to be photoheterotrophic type, which is a special characteristic of purple non-sulfur bacteria. To further identify the strain PTB2, several biochemical tests were conducted. The results with use of various substrates as carbon source or electron donor suggest that the strain PTB2 shares the similarity to the bacterium R. palustris [29–31]. In order to confirm the bacterial identity, 16S rRNA genes were amplified from the strains and the sequence was determined. The results showed that the strains HTB01 and HTB10 belong to the family Enterobacteriaceae and closest to C. freundii ATCC 8090 (Accession No. AJ 233408.1) (99%), and E. aerogenes (Accession Nos. AB244467.1 and AB244456.1) (100%), respectively. The other strain PTB2 was closely related to the purple non-sulfur phototrophic bacteria, R. palustris CGA 009 (Accession No. BX572608.1) and TIE-1 (Accession No. X 87279.1) (99%). Therefore, the isolated strains HTB01, HTB10, and PTB2 were identified and denoted as C. freundii C01, E. aerogenes E10, and R. palustris P2, respectively.

3.2.

Lab-scale experiments

In order to obtain information about the parameters involved in inoculum development during the scale-up process, several experiments were carried out in serum bottles (125 ml) with a working volume of 100 ml. From the preliminary results, it was observed that, 10 g l1 glucose was optimal for most of the cultures examined. Three strains (C. freundii 01, E. aerogenes E10 and R. palustris P2) were observed to be the best for H2 production (data not shown). Therefore, these strains were utilized in further culture optimization experiments. The results obtained in the batch fermentation are presented in Fig. 2 and Tables 2–6, indicating that 55 mM glucose was completely consumed at 24 h. E. aerogenes E10 yields maximum biomass of 0.79 g l1 at 12 h, whereas C. freundii 01 and R. palustris P2 produced 0.71 and 0.65 g l1, respectively as shown in Fig. 2. On the other hand, although the pattern was little different at log phase culture, glucose consumption was

Table 1 – Biochemical characterization of H2 producing strains Characteristics

HTB01

HTB10

PTB2

Shape Gram’s stain Catalase Oxidase Motility Indole formation Methyl red Voges-Proskauer Simmons citrate Lysine decarboxylase H2S test Salicin Acid and gas from glucose Pigments BChl type Absorption maxima of living cells Nitrate reduction G þ C content (mol%) Aerobic growth Vitamin requirement

Rods  þ  þ (þ) þ  þ  þ  þ   ND

Rods  þ  þ   þ (þ) þ  (þ) þ   ND

þ 51  0.4 þ 

þ 53  0.4 þ 

Rods  þ þ þ ND ND ND ND ND ND ND þ Red to brown a 376, 470,540, 585, 805, and 875 nm  65.5  0.2 þ Biotin

Sugar fermentation Glucose Cellobiose Sucrose Maltose Rhamnose Trehalose Xylose Mannitol Sorbitol Glycerol

þ þ þ þ þ þ þ þ þ þ

þ þ þ þ þ þ þ þ þ þ

þ ND ND ND ND ND ND (þ) þ þ

ND

ND

Electron donor and carbon source H2 Malate Succinate Formate Acetate Tartrate Propionate Butyrate Valerate Pelargonate Pyruvate Lactate Benzoate Methanol Ethanol Arginine Glutamine

þ þ þ þ þ  þ þ þ  þ þ þ (þ) (þ)  þ

The isolated strains HTB01, HTB10 and PTB2 were identified as C. freundii, E. aerogenes, and R. palustris, respectively. ND, not determined; þ, positive; , negative; (þ), weak.

almost the same, at 10 g l1 (55 mM) in 24 h. As shown in Table 2, the maximum H2 yield of 1.62 mol mol1 glucose was obtained with R. palustris P2, while E. aerogenes E10 and C. freundii 01 produced 1.49 and 1.3 mol mol1 glucose, respectively. Therefore, the maximum yield efficiency of 81% was observed for R. palustris P2 rather than for E. aerogenes E10

5408

international journal of hydrogen energy 33 (2008) 5404–5415

1.0

Cell dry mass (g l-1)

0.8

0.6

0.4

0.2

0.0

0

5

10

15

20

25

30

Time (h) Fig. 2 – A time course profile showed the growth pattern of different bacterial strains on glucose under anaerobic condition. Symbols: (;) R. palustris P2, () C. freundii 01, (B) E. aerogenes E10 and (6) co-culture. The values indicate the mean of triplicates.

(74.5%) and C. freundii 01 (65%). However, a maximum rate of H2 production was obtained with E. aerogenes E10 as 10.9 mmol1 g1 cdw h1 and it was slightly higher than the rate obtained with C. freundii 01 (8.78 mmol1 g1 cdw h1) and R. palustris P2 (8.53 mmol1 g1 cdw h1). Since all three strains produce H2 efficiently, a co-culture of these strains was made and the potential on H2 production was evaluated. When combining these three strains, H2 yield slightly improved. The maximum H2 yield of 1.65 mol mol1 glucose was obtained with the highest production rate of 10.6 mmol1 g1 cdw h1. The results obtained with a coculture are advantageous since the experiments with the individual strains exhibited impressive yield or H2 production rate. For instance, although maximum yield of 1.62 mol mol1 glucose was obtained with R. palustris, the H2 production rate was higher with E. aerogenes E10 as 10.9 mmol1 g1 cdw h1. When these strains were combined, both yield and the rate improved considerably. For statistical evaluation of the H2 production data, the one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test was conducted to test the significance of experimental data shown in Table 2. The corresponding ANOVA for H2 production yield and the rate showed that the

F-values were 34.988 and 169.139, respectively. It implies that the data were statistically significant because P < 0.05 (alevel). The coefficient of determination (R2) for the H2 yield and the rate were calculated to be 0.929, and 0.984, respectively (Table 3). It suggests that the data obtained was reliable with the confidence level 93%. The pair wise comparison of bacterial groups for their significance on H2 production rate and yield were also analyzed using Tukey’s test (Table 4). The honestly significant difference (HSD) Q-value is >Q-critical value at an a-level 0.05, indicating that the relevant groups were statistically significant. In particular, the groups M1 vs M4 were statistically significant for both H2 yield and the H2 production rate, while M2 vs M4 groups were significant only for the H2 yield. However, the other groups, M1 vs M2, and M3 vs M4 were significant for the H2 production rate. In general, H2 production is a strictly anaerobic process, so dissolved O2 in the medium must be replaced with either argon (Ar) or nitrogen (N2) gas to achieve maximum yield and conversion efficiency of substrate to H2. However, the use of such gases, although applicable at lab-scale level, is relatively expensive, and therefore uneconomical for commercial scale production of H2 from effluent. E. aerogenes E10 is a facultative anaerobe that produces amounts of H2 comparable to those produced by Clostridium sp. [32–35]. Furthermore, this strain can survive and produce H2 in the presence of low O2 during the initial stage of anaerobic biodegradation [33,34]. We previously studied the co-culture of C. freundii and R. palustris for H2 production [17,18]. Therefore, we attempted to combine E. aerogenes E10 with C. freundii 01 and R. palustris P2 to provide a versatile biological approach for H2 production at pilot-scale. We believe that the lab-scale model, although showed interesting results, it might be very different at pilot-scale. Therefore, the pilot-scale processes need to be studied in detail.

3.3.

Metabolite profile and carbon material balance

As shown in Table 5, ethanol and acetate were the major end products for E. aerogenes E10, while C. freundii 01 produces these metabolites and lactate. On the other hand, R. palustris P2 produces butyric acid and butanol as major byproducts in addition to ethanol, indicating the uniqueness in metabolism compared to other enteric bacteria under dark anaerobic conditions. E. aerogenes E10 accumulated about 26 mM of formate in addition to other usual metabolites. In general, most of the enteric bacteria produce H2 from formate by utilizing the formate H2 lyase (FHL) complex [36–38]. The

Table 2 – Characteristics of anaerobic fermentation of glucose on H2 production H2 production

H2 produced (mmol l1) H2 yield (mol mol1 glucose) H2 production rate (mmol g1 cdw h1)

Bacterial strains C. freundii 01

E. aerogenes E10

R. palustris P2

Co-culturea

69.60  2.62 1.30  0.02 8.78  0.12

83.00  1.50 1.49  0.09 10.90  0.20

89.70  2.71 1.62  0.06 8.53  0.12

92.00  2.78 1.65  0.07 10.60  0.12

cdw: cell dry weight. Errors represent >93% confidence intervals with n ¼ 3 for each analysis. a Co-culture (C. freundii 01, E. aerogenes E10 and R. palustris P2).

5409

international journal of hydrogen energy 33 (2008) 5404–5415

Table 3 – ANOVA for H2 production from glucose under anaerobic condition Parameters a

H2 yield

H2 production rateb

Source of variation Groupsbetween Groupswithin Total Groupsbetween Groupswithin Total

SS

DF

MS

F-value

F-critical

P-value

0.2283 0.0174 0.2457 13.3979 0.2112 13.6091

3 8 11 3 8 11

0.0761 0.0021

34.988*

4.066

6.0E  05

4.4653 0.0264

169.139*

4.066

1.488E  07

SS, sum of squares; DF, degrees of freedom; MS, mean square. F-value > F-critical value (statistically significant (*)). a Coefficient of determination (R2) ¼ SSbetween/SStotal ¼ 0.929. b R2 ¼ 0.984.

reduced H2 yield (1.49 mol mol1 glucose) signified an inadequacy in FHL complex enzymatic activity in this case. If the accumulated formate was oxidized to H2, then it could have attained the theoretical limit of 2 mol mol1 of glucose with enteric bacteria. Moreover, some of the Enterobacter sp. were reported to produce more than 2 mol of H2 mol1 of glucose. Based on dark fermentation stoichiometry, the conversion of 1 mol of glucose into acetate yields 4 mol H2 mol1 glucose, but only 2 mol of H2 mol1 glucose is formed when butyrate is the end product [6]. The highest H2 yield reported from glucose is around 2.0–2.4 mol mol1 [39–41]. Production of butyrate rather than acetate may partially explain the deviations from the theoretical yield. Utilization of substrate as an energy source for bacterial growth is another main reason for obtaining the yields lower than theoretically predicted. Kumar and Das [42] reported that E. cloacae IIT-BY 08 produces 2.2 mol of H2 mol1 glucose. Accumulation of more succinate (18.7 mM) by E. aerogenes E10 in the present investigation indicates the production of large amounts of reducing equivalents, such as NADH, confirming that this strain may operate both H2 producing systems, such as FHL complex and NADH ferrodoxin oxidoreductase, under anaerobic conditions. The production of butyric acid and butanol by R. palustris P2 is similar to that observed for Clostridium butyricum, indicating the operation of both H2 producing enzymes, which contrasts with the metabolic profiles of most enteric bacteria. The production of 2.3-butanediol indicates the uniqueness of Enterobacter metabolism [43].

Table 4 – Tukey’s post hoc test for H2 production from glucose under anaerobic condition HSD Q-value Groups M1 M1 M1 M2 M2 M3

vs vs vs vs vs vs

M2 M3 M4 M3 M4 M4

a

H2 yield

H2 production rateb

1.3389 6.0474* 13.2287* 4.9135* 12.0948* 7.1813*

22.0662* 3.1980 19.4012* 25.2642 2.6650 22.5992*

*HSD Q-value is > HSD-Q-critical value (statistically significant (*)). M1, co-culture; M2, R. palustris P2; M3, E. aerogenes E10; M4, C. freundii C01. For footnotes a and b HSD Q-critical is 4.53 (a-level 0.05). a MSwithin ANOVA ¼ 0.0021; n ¼ 3; DFwithin 8; groups 4. b MSwithin ANOVA ¼ 0.0264; n ¼ 3; DFwithin 8; groups 4.

More than 95% of the carbon input was recovered as various metabolites as shown in Table 6. A maximum of carbon was directed to lactate (26% in C. freundii 01) and around 20% of carbon was converted to ethanol and acetate, whereas 28% of carbon was converted to butyric acid and 23% to butanol in the case of R. palustris P2. However, production of lactate and acetate was greatly decreased in R. palustris P2. Succinate and ethanol were the major end products of 22% of the carbon spent by E. aerogenes E10. When the co-culture of the three strains was made, nearly the same amount of carbon (15–18%) was directed to acetate, ethanol and butyrate formation.

3.4. H2 production and effluent treatment at 10 and 100 m3 scales The proximate analysis of distillery effluent is shown in Table 7. The initial COD, BOD total sugar, reducing sugar and VFAs were measured as 101.2, 58.8, 7.9, 7.5 and 3.52 g l1, respectively. Reducing sugar was completely consumed by 40 h of fermentation. COD and BOD were reduced to 43 g and 0.67 g per liter, respectively, while VFAs were increased to 5.78 g l1 by 40 h. The co-culture of C. freundii 01, E. aerogenes E10 and R. palustris P2 exhibited significant improvement in yield and H2 production rate on the lab-scale, with glucose as a carbon source. Hence, the same co-culture was used while scaling-up the process for H2 production and effluent treatment. Since the growth pattern of these strains cannot be measured spectrophotometrically due to the presence of melanoidin pigment in distillery effluent [44], an indirect method of total viable count (TVC) was used to assess the microbial population. The population of C. freundii 01, E. aerogenes E10 (HTB) and R. palustris P2 (PTB) was analyzed every 6 h and is presented in Fig. 3a. The HTB population increased from 0.7  106 CFU/ml to 23  106 CFU/ml by 12 h of fermentation and decreased to 12.5  106 CFU/ml by 40 h. The sharp decline in HTB from 12 to 40 h could be due to pH and/or substrate depletion. A pH drop to 5.9 at 12 h might have hampered the growth of the HTB, especially on C. freundii 01. Although a slight retardation in growth was observed with R. palustris P2, growth was essentially stable until 40 h. R. palustris has a versatile metabolism that is capable of oxidizing acetate via the glyoxylate shunt under dark conditions [45]. It can also grow well at pH 5.5–9.0 [46, 47]. Presumably, the steady state of PTB growth is due to the availability of oxidizing acetate and/or other organic acids. Our observation

5410

international journal of hydrogen energy 33 (2008) 5404–5415

Table 5 – Anaerobic fermentation of glucose and the yield of H2 and various metabolites by C. freundii 01, E. aerogenes E10, R. palustris P2 and, their co-culture Reactant

Bacterial strains C. freundii 01

E. aerogenes E10

Co-culturea

R. palustris P2

Concentrations (mmol l1) Glucose

53.20

55.50

55.30

55.50

Products H2 CO2 Ethanol Pyruvate Formate Succinate Lactate Acetate Butyric acid n-Butanol 2,3-Butanediol Biomassb

69.60 52.67 31.70 0.24 1.85 4.71 27.53 30.25 0.00 0.00 0.00 0.54

83.00 64.83 38.50 0.00 25.80 18.70 4.45 23.00 0.00 0.00 2.85 0.62

89.70 65.90 23.54 0.00 0.00 1.40 1.25 9.70 23.00 18.80 0.00 0.57

92.00 68.70 27.50 0.00 0.63 1.40 8.67 25.70 15.00 7.10 1.70 0.58

Values given with <5% standard deviation (n ¼ 3). a Co-culture (C. freundii 01, E. aerogenes E10 and R. palustris P2). b Cell dry biomass (g l1).

that R. palustris produce H2 under dark anaerobic condition from acetate present during cultivation in distillery effluent may be due to the presence of residual yeast sludge. Similar observations were made by Uffen and Wolfe [48], who demonstrated production of H2 by R. palustris under dark anaerobic conditions as described above. An industrially attractive R. palustris was isolated and studies were carried out by Lee et al. [47], who showed production of CO-dependent H2 and growth under phototrophic and non-phototrophic

conditions. Thus, phototrophic and heterotrophic bacteria are considered the favorable candidates for large scale H2 production due to their high substrate conversion efficiency and capability of using a wide range of carbon sources for growth and H2 production [49,50]. The results of batch fermentation at 10 and 100 m3 scales are shown in Table 8. A maximum H2 yield of 1.83 kg was obtained at the 10 m3 scale, while the H2 yield at the 100 m3 scale was 21.4 kg. Production of VFAs was measured as 226 kg

Table 6 – Carbon material balance for anaerobic fermentation of glucose by C. freundii 01, E. aerogenes E10, R. palustris P2 and, their co-culture Reactant

Carbon material balance C. freundii 01

E. aerogenes E10

R. palustris P2

Co-culture

a

C (mmol l1)

C (%)

C (mmol l1)

C (%)

C (mmol l1)

C (%)

C (mmol l1)

C (%)

Glucose

319.2

100%

333

100%

331.8

100%

333

100%

Products CO2 Ethanol Pyruvate Formate Succinate Lactate Acetate Butyric acid n-Butanol 2,3-Butanediol Biomassb Total products % Recoveryc

52.67 63.40 0.72 1.85 18.84 82.59 60.50 0.00 0.00 0.00 20.05 300.60

17.17 20.66 0.23 0.60 6.14 26.92 19.72 0.00 0.00 0.00 6.54 – 96

64.83 77.00 0.00 25.8 74.8 13.35 46.00 0.00 0.00 11.40 21.54 334.7

19.03 22.60 0.00 7.57 21.95 3.92 13.50 0.00 0.00 3.35 6.32

65.90 47.08 0.00 0.00 5.60 3.75 19.40 92.00 75.20 0.00 21.28 330.00

19.79 14.14 0.00 0.00 1.68 1.13 5.83 27.63 22.58 0.00 6.39 – 99

68.70 55.00 0.00 0.63 5.60 26.01 51.40 60.00 28.40 6.80 21.69 324.20

20.64 16.52 0.00 0.19 1.68 7.81 15.44 18.02 8.53 2.04 6.52 – 97

– 100

Values given with <5% standard deviation (n ¼ 3). a Co-culture (C. freundii 01, E. aerogenes E10 and R. palustris P2). b Assumed to be CH1.8O0.5N0.2. c (Total carbon in biomass and metabolites)  100/(Total carbon in glucose).

5411

Table 7 – Physiochemical characteristics of distillery effluent during fermentation at 100 m3 scale

a

Characteristics

Total population (10 x 6 CFU ml-1)

international journal of hydrogen energy 33 (2008) 5404–5415

4.3  0.2a

43.00 0.67 67.50 3.84 5.90 12.50 60.50 56.10 0.00 0.00 5.78 8.90 56.50 5.4  0.2

COD, chemical oxygen demand; BOD, biological oxygen demand. Values given with <5% error. a Initial pH of the effluent (adjusted to 7.0  0.2 before fermentation).

5 0

25

b

800

20

600

15 400 10 200 5 0

0

c

7.5

48

7.0

44 40

pH

6.5 36 6.0 32 5.5

d

COD (g l-1)

for 100 m3. The initial VFAs in the effluent 3.52 g l1 increased to 5.78 g l1 after 40 h of fermentation, reflecting the formation of acidic end products. Acetate and butyrate were the major soluble end products of VFAs, along with ethanol (96.6 kg/ 100 m3). As shown in Fig. 3b, the utilization of reducing sugar at 6 h was estimated at 23 kg, and most of the substrate during this period was utilized for the growth of microorganisms. The maximum production of H2 (20 m3) was obtained at 19th hour. H2 production declined after 19 h and the inclination was sustained to 24 h (Fig. 3b) due to decreased temperature. While temperature and pH are essential to fermentative H2 production, the maintenance of temperature/pH for a 100 m3 volume of distillery effluent is not an economically viable prospect for either effluent treatment or H2 production. Hence, control over temperature and pH was not attempted. H2 production after 25 h increased along with temperature to 28  C, suggesting that, although the pH dropped to 5.5, it did not affect H2 production. The optimum growth temperature reported for photosynthetic bacteria was in the range of 30– 37  C [46–52]. Therefore, the production of H2 accelerated when the temperature reached 28  0.2  C. Peck and Gest [53] reported that transcription of the fdhF and hyc genes of FHL complex was pH dependent and only occurs below pH 7.0 in Escherichia coli and other enteric bacteria. In addition, formate is known to be transported into cells via a symport mechanism at an acidic pH, and excreted at alkaline pH [54]. Therefore, the pH gradient across the cytoplasmic membrane determines the formate transport direction and conversion into H2 and CO2 in enteric bacteria. The results obtained in this study demonstrate that, when the pH declined to 5.5, the produced formate was actively transported at an acidic pH where H2 was produced at a high rate (10.9 m3/h). The reported pH range for maximum H2 yield or specific H2 production rate was between pH 5.0 and 6.0 [33,55]. However, some investigators reported that the optimum pH was between 6.8 and 8.0 [43,44], and around pH 4.5 for the

10

Residual sugar (kg)

pH

101.00 58.80 103.00 8.30 23.20 95.00 78.40 56.10 7.90 7.50 3.52 14.46 91.50

15

Temperature (°C)

COD BOD Total solids Total suspended solids Total fixed solids Total dissolved solids Total volatile solids Dissolved volatile solids Total sugar Reducing sugar Volatile fatty acids Total nitrogen Total protein

Final (g l1)

28

5.0

24

110

70

100

60

90

50

80

40

70

30

60

20

50

10

40

0

BOD (g l-1)

Initial (g l1)

20

H2 produced (Litres x m3)

Fermentation

25

30 0

10

20

30

40

50

Time (h) Fig. 3 – A time course of bacterial co-culture fermentation on distillery effluent at 100 m3 scale. (a) Microbial population, () HTB, (B) PTB; (b) H2 production and reducing sugar consumption, () H2 production, (,) sugar consumption; (c) temperature and pH profile, () pH, (B) temperature; (d) reduction of COD and BOD, () COD, (B) BOD. thermophilic culture [56,57]. Most of the studies indicated that the final pH in anaerobic H2 production is around 4.0–4.8 regardless of initial pH [16,19,33,57]. However, in the present investigation, the final pH was in range of 5.2–5.4, as shown in Fig. 3c. A decrease in pH was due to the production of organic

5412

international journal of hydrogen energy 33 (2008) 5404–5415

Table 8 – Production of H2 and other metabolites at pilotscales Substrate and metabolites Fermentation at different scale Substrate consumed and products formed (kg) Reactant Reducing sugar Products Total H2 CO2 Ethanol Volatile fatty acids COD (reduced)

10 m3

100 m3

68.30

675.00

1.83 29.30 ND ND 535.00

21.40 173.00 96.60 226.00 5800.00

ND: not detected. Error: less than 5%.

approximate cost required for the conventional effluent treatment (US $2030/100 m3) by aerobic process, can also be saved through the biohydrogen process.

3.6.

Process economics

The total capital cost of the present system is estimated at $44,444 and the total H2 production cost was estimated at $35.44/100 m3, which includes labor ($8.88), chemicals ($22.66) and power ($3.88). The annual production cost is $5386 (Table 9). The total fermentation run time of this process is

Table 9 – Process economics for biological H2 production (BHP) at 100 m3 scale Process economics

acids [41,58]. Utilization of organic acids, especially acetate under facultative anaerobic condition by R. palustris, could provide better effluent quality in terms of COD, and this was one of the key objectives of the present investigation. As shown in Fig. 3b, the unbalanced state of H2 production was presumably due to temperature variation during the day (32–39  C), and at night (26–32  C). Lack of H2 production after 40 h was due to the unavailability of substrate. The total reducing sugar consumed at 40 h was 675 kg and the remainder was only 0.13 kg. The complete utilization of substrate in a short retention time was encouraging in terms of H2 production and effluent treatment as the COD and BOD levels reduced considerably (Fig. 3d). The average rate of H2 production was estimated as 0.53 kg/100 m3/h and the yield was 2.76 mol mol1 glucose. Yokoi et al. [33] have used dried sweet potato starch residue for H2 production through mixed culture of C. butyricum and E. aerogenes and reported 2.4 mol H2 mol1 glucose from 2% starch residue containing wastewater in long-term repeated batch operations. Ginkel et al. [59] have also studied H2 production from apple and potato processor industrial effluents, and domestic wastewater. Based on the results of our investigation, a sustained state of H2 production can be achieved with parallel treatment of effluent in continuous operation.

3.5.

Effluent treatment process

The initial COD and BOD of the distillery effluent were measured as 101.2 and 58.9 g l1. After the fermentative H2 production process, COD and BOD levels were reduced to 43 g and 0.67 g l1, respectively. These results are highly encouraging because the BOD and COD reductions were 98.3 and 58.2%, respectively (Fig. 3d). Han and Shin [60] have used food waste as medium for H2 production and have reported 58% COD reduction with 70% H2 production efficiency and over 100 L cumulative H2 gas. Anaerobic and aerobic effluent treatment processes have been used for several years in industries and bio-methanation is widely used, but the retention time of waste in the digester is about 15 days [16,26]. Moreover, methane contributes 10% of the green house effect in environmental pollution. On the other hand, the retention time of biohydrogen production process is less than 2 days, and H2 only produces water while burning. In addition, the

Operational cost Caustic soda Power (38.8 kWh  INR 4.5) Labor cost Subtotal Total number of process per year 182 runs/365 days Maintenance time 60 days. Thus, approximate possible runs per year are 152 Total operational cost per year Capital cost Interest on capital @18% Maintenance Total production cost per year Cost gain on H2 per run in a batch processa Cost spent of H2 purification per batchb Cost gain on sludge per run in a batch process Cost gain on pure H2 and sludge per year Cost spent on production per year Thus, net gain from the biological H2 production processc

US$

US$

Net gain (US$)

22.66 3.88 8.88 35.44

5386.00 44,444.00 8000.00 160.00 13,546.00 456.00

168.00 45.00

76,152.00

39,082.00 37,070.00

INR: Indian rupees. a Subject to variation depends upon the process of H2 purification and storage. b Calculated for total CO2, 173 kg/100 m3. NaOH based CO2 absorption method was used for purifying H2. c If the remaining COD level (4275 kg/100 m3) to be further treated to meet the safe disposal level (0.25 g l1) of distillery effluent, the cost w$248 needs to be included. However, by using this process, the effluent treatment cost ($1782) accounted for 100 m3 ($2030– $248 ¼ $1782) can be saved. Therefore, in addition to $37,070 (net gain from pure H2 and sludge), the cost w$270,864/year against effluent treatment can be saved.

international journal of hydrogen energy 33 (2008) 5404–5415

approximately two days, including loading, cooling, and pH adjustment of the effluent for batch fermentation. Thus, on average, 152 batch runs/year can be done, and approximately two months are required for maintenance work in a year. The interest on investment is around $8000 per annum and the maintenance cost is approximately $160. Therefore, in a year, the overall cost of production is $13,546. On the other hand, the cost gain in terms of H2 is approximately $1.9 per m3 as per the current commercial market value of H2 (99.9% pure). Thus, an average gain is $456 from 239.642 m3 H2. In addition, the cost of sludge is around $45/100 m3, with the total gain from the process being $501. Hence, on average, $76,152 will be the gained from this process yearly. On the other hand, the acceptable level of H2 purity for fuel cell application is >99%. Therefore, the cost $168 required to absorb 173 kg of CO2/ 100 m3 should be deduced from the net gain. If the total H2 production cost $13,546 and H2 purification cost $25,536 are deducted from the total gain $76,152 from H2 and sludge, the net gain would be approximately $37,070/year. However, cost factor may vary with mode of operation. If the produced H2 is compared in terms of energy (heat or power equivalent) content, one gallon of H2 is equivalent to 30,000 Btus (British thermal units). One kilowatt-hour is equivalent to 3412 Btus [61]. Therefore, about 8.792 kWh of power can be generated per gallon of H2. When it is compared with gasoline, 1 gallon is equivalent to 120,000 Btus, which is about 35.16 kWh [61–63]. This is 4-fold higher than the H2 energy content, and gasoline costs about $2.50 per gallon. Although, energy content of H2 is 4-fold lower than that of gasoline (on volumetric basis), the cost of H2 produced from this process is approximately $0.9. Since current natural gas prices are high and expected to continue increasing as the available reserves are exponentially consumed the process of biological H2 production from distillery effluent seems to be promising as a viable alternative energy source.

4.

Concluding remarks

In the present investigation, the use of distillery effluent for H2 production was evaluated at the 100 m3 scale. The results obtained from the pilot-scale study demonstrated that coculture of E. aerogenes E10, C. freundii 01 and R. palustris P2 in distillery effluent produced 21.38 kg of H2 in 40 h, which corresponded to the energy of 3.045 GJ, based on the upper combustion value of 7.18 MJ/Nm3. The average yield of H2 was 2.76 mol mol1 glucose, and the rate of H2 production was estimated to be 0.53 kg/100 m3/h. In terms of cost economics, the approximate net gain as $37,070/year can be achieved when using distillery effluent as a source of H2 production. In parallel, a high level of BOD and COD reduction was achieved by 40 h. Hence, this would be a viable process for both distillery effluent treatment and clean alternative energy production.

Acknowledgments This work was supported by Ministry of Non-Conventional Energy Sources (MNES), Govt. of India, and EID Parry (I) Ltd. We

5413

would like to express of our sincere gratitude to MNES and EID Parry (I) Ltd.

references

[1] Armor JN. The multiple roles for catalysis in the production of H2. Appl Catal A Gen 1999;176:159–76. [2] Benemann J. Hydrogen biotechnology: progress and prospects. Nat Biotechnol 1996;14:1101–3. [3] Vatsala TM, Seshadri CV. Microbial production of hydrogen – a review. Proc Indian Natl Sci Acad 1985;B51(2):282–95. [4] Das D, Veziroglu TN. Hydrogen production by biological processes: a survey of literature. Int J Hydrogen Energy 2001; 26:13–28. [5] Nandi R, Sengupta S. Microbial production of hydrogen: an overview. Crit Rev Microbiol 1998;24(1):61–84. [6] Levin DB, Pitt L, Love M. Biohydrogen production: prospects and limitations to practical application. Int J Hydrogen Energy 2004;29:173–85. [7] Vatsala TM, Seshadri CV. Phototrophic breakdown of solid cellulose from silk-cotton to hydrogen and volatile fatty acids by Rhodospirillum rubrum; ATCC 11170. Curr Sci 1988;58: 173–5. [8] Kapdan IK, Kargi F. Bio-hydrogen production from waste materials. Enzyme Microbiol Technol 2006;38:569–82. [9] Venkata Mohan S, Bhaskar YV, Sarma PN. Biohydrogen production from chemical wastewater treatment by selectively enriched anaerobic mixed consortia in biofilm configured reactor operated in periodic discontinuous batch mode. Water Res 2007;41:2652–64. [10] Yu H, Zhu Z, Hu W, Zhang H. Hydrogen production from rice winery wastewater in an upflow anaerobic reactor by using mixed anaerobic cultures. Int J Hydrogen Energy 2002;27: 1359–65. [11] Ramasamy V, Seshadri CV, Vatsala TM, Rakshit SK. Influence of impeller tip speed on microbial hydrogen production by Rhodospirillum rubrum ATCC 11170 from distillery effluent. Aust J Biotechnol 1990;4:287–9. [12] Rachman MA, Nakashimada Y, Kakizono T, Nishio N. Hydrogen production with high yield and high evolution rate by self-flocculated cells of Enterobacter aerogenes in a packed bed reactor. Appl Microbiol Biotechnol 1998;49: 450–4. [13] Oh SE, Iyer P, Bruns MA, Logan BE. Biological hydrogen production using a membrane bioreactor. Biotechnol Bioeng 2004;87(1):119–27. [14] Chang JS, Lee KS, Lin PJ. Biohydrogen production with fixed bed bioreactors. Int J Hydrogen Energy 2002;27:1167–74. [15] Vatsala TM. Hydrogen production from stillage (canemolasses) by Citrobacter freundii and its use in improving methanogenesis. Int J Hydrogen Energy 1992;17:923–7. [16] Venkataraman C, Vatsala TM. Hydrogen production from whey by phototrophic organisms. In: Proceedings of Hawaii hydrogen 90. W.H.E.C 8 hydrogen energy progress VIII, vol. 2. Pergamon Press, NY; 1990, p. 781–8. [17] Meenakshisundaram C, Vatsala TM. Biohydrogen production using Rhodopseudomonas palustris VM 01 and Citrobacter freundii VR 01 – bio-energy from waste. Ph.D. thesis, University of Madras; 2001. [18] Vatsala TM, Manimaran A. Hydrogen production from organic wastes using defined co-culture of bacteria. In: Sayigh AAM, editor. Proceedings of 7th the world renewable energy congress. Cologne, Germany: Pergamon Press; 2002. p. 231–5. [19] Ormerod JG, Ormerod KS, Gest H. Light dependent utilization of organic compounds and photo-production of

5414

[20]

[21]

[22]

[23] [24]

[25]

[26] [27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

international journal of hydrogen energy 33 (2008) 5404–5415

molecular hydrogen by photosynthetic bacteria; relationships with nitrogen metabolism. Arch Biochem Biophys 1961;94:449. Doetsch RN. In: Gerhardt P, Murray RGE, Costilow RN, Nester EW, Wood WA, Krieg NR, Phillips GB, editors. Manual of methods for general microbiology. Washington, DC: American Society for Microbiology; 1981. p. 21–33. Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 1994;22:41–7. Maidak BL, Cole JR, Lilburn TG, Jr CTP, Saxman PR, Farris RJ, et al. The RDP-II (Ribosomal Database Project). Nucleic Acids Res 2001;29:173–4. Miller GL. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Anal Chem 1959:426–9. APHA. Standard methods for the examination of water and wastewater. 19th ed. Washington, DC: American Public Health Association; 1995. Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. Colorimetric method for determination of sugars and related substances. Anal Chem 1956;28:350–6. Vatsala TM, Balaji V. A novel microbial method for augmentation of biogas. Curr Sci 1987;56:898–9. Sakazaki R. In: Krieg NR, Holt JC, editors. Bergey’s manual of systematic bacteriology. Citrobacter (Enterobacteriaceae Werkman CH, Gillen GF 1932. 173AL), vol. 1. Baltimore: Williams & Wilkins; 1984. p. 458–61. Richard C. In: Krieg NR, Holt JC, editors. Bergey’s manual of systematic bacteriology. Enterobacter (Enterobacteriaceae Hormaeche E, Edwards DR. 1960. 72AL), vol. 1. Baltimore: Williams & Wilkins; 1984. p. 465–71. Pfennig N, Truper HG. The phototrophic bacteria. In: Buchanan RE, Gibbons NE, editors. Bergey’s manual of determinative bacteriology. 8th ed. Baltimore: Williams & Wilkins; 1974. p. 24–75. Imhoff JF, Truper HG, Pfennig N. Rearrangement of the species and genera of the phototrophic ‘‘purple nonsulfur bacteria’’. Int J Syst Bacteriol 1984;34:340–3. Imhoff JF, Truper HG. In: Staley JT, Bryant MP, Pfennig N, Holt JC, editors. Bergey’s manual of systematic bacteriology. Purple nonsulfur bacteria (Rhodospirillaceae Pfennig and Truper 1971, 17AL), vol. 3. Baltimore: Williams & Wilkins; 1989. p. 1658–62. Nakashimada Y, Rachman MA, Kakizono T, Nishio N. Hydrogen production of Enterobacter aerogenes altered by extracellular and intracellular redox states. Int J Hydrogen Energy 2002;27:1399–405. Yokoi H, Tokushige T, Hirose J, Hayashi S, Takasaki Y. Hydrogen production from starch by mixed culture of Clostridium buytricum and Enterobacter aerogenes. Biotechnol Lett 1998;20:143–7. Tanisho S, Ishiwata Y. Continuous hydrogen production from molasses by bacterium Enterobacter aerogenes. Int J Hydrogen Energy 1994;19:807–12. Yokoi H, Saitsu AS, Uchida H, Hirose J, Hayashi S, Takasaki Y. Microbial hydrogen production from sweet potato starch residue. J Biosci Bioeng 2001;91:58–63. Jung GY, Kim JR, Park J, Park S. Hydrogen production by a new chemo heterotrophic bacterium Citrobacter sp. Y19. Int J Hydrogen Energy 2002;27:601–10. Kim S, Seol E, Raj SM, Park S, Oh YK, Ryu DDY. Various hydrogenase and formate dependent hydrogen production in Citrobacter amalonaticus Y19. Int J Hydrogen Energy 2008;33: 1509–15. Oh YK, Kim HJ, Park S, Kim MS, Ryu DDY. Metabolic-flux analysis of hydrogen production pathway in C. amalonaticus Y19. Int J Hydrogen Energy 2008;33:1471–82.

[39] Ueno Y, Haruta S, Ishii M, Igarashi Y. Characterization of a microorganism isolated from the effluent of hydrogen fermentation by microflora. J Biosci Bioeng 2001;92:397–400. [40] Fang HHP, Liu H. Effect of pH on hydrogen production from glucose by mixed culture. Bioresour Technol 2002;82:87–93. [41] Morimoto M, Atsuko M, Atif AAY, Ngan MA, Fakhru’l-Razi A, Iyuke SE. Biological production of hydrogen from glucose by natural anaerobic microflora. Int J Hydrogen Energy 2004;29: 709–13. [42] Kumar N, Das D. Enhancement of hydrogen production by Enterobacter cloacae IIT-BT 08. Process Biochem 2000;35: 589–93. [43] Podest´a JJ, Navarro AMG, Estrella CN, Esteso MA. Electrochemical measurements of trace concentrations of biological hydrogen produced by Enterobacteriaceae. Inst Pasteur 1997;148:87–93. [44] Wedzicha BL, Kaputo MT. Melanoidins from glucose and glycine: composition, characteristics and reactivity towards sulphite ion. Food Chem 1992;43:359–67. [45] Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, et al. Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nat Biotechnol 2004;22:55–61. [46] Oh YK, Seol EH, Lee EY, Park SH. Fermentative hydrogen production by new chemoheterotrophic bacterium Rhodopseudomonas palustris P4. Int J Hydrogen Energy 2002;27: 1373–9. [47] Lee TH, Park JY, Park SH. Growth of Rhodopseudomonas palustris under phototrophic and non-phototrophic conditions and its CO-dependent hydrogen production. Biotechnol Lett 2002;24:91–6. [48] Uffen RL, Wolfe RS. Anaerobic growth of purple nonsulfur bacteria under dark conditions. J Bacteriol 1970;104:462–72. [49] Vatsala TM. Degradation of cellulose by the phototrophic bacterium Rhodospirillum rubrum ATCC 11170. Ind J Expt Biol 1989;27:963–6. [50] Vatsala TM. Influence of nutritional factors on hydrogen production by R. rubrum ATCC 11170. J Microbial Biotechnol 1987;2:127–31. [51] Kim MS, Baek JS, Lee JK. Comparison of hydrogen accumulation by Rhodobacter sphaeroides KD131 and its uptake hydrogenase and PHB synthase deficient mutant. Int J Hydrogen Energy 2006;31:121–7. [52] Zhu H, Ueda S, Asada Y, Miyake J. Hydrogen production as a novel process of wastewater treatment – studies on tofu wastewater with entrapped R. sphaeroides and mutagenesis. Int J Hydrogen Energy 2002;27:1349–57. [53] Peck HD, Gest H. Formic dehydrogenase and the hydrogenlyase enzyme complex in coli-aerogenes bacteria. J Bacteriol 1957;73:706–21. [54] Rossmann R, Sawers G, Bock A. Mechanism of regulation of the formate-hydrogenlyase pathway by oxygen, nitrate, and pH: definition of the formate regulon. Mol Microbiol 1991;5: 2807–14. [55] Lay JJ, Lee YJ, Noike T. Feasibility of biological hydrogen production from organic fraction of municipal solid waste. Water Res 1999;33:2579–86. [56] Khanal SK, Chen WH, Li L, Sung S. Biological hydrogen production: effects of pH and intermediate products. Int J Hydrogen Energy 2004;29:1123–31. [57] Kanai T, Imanaka H, Nakajima A, Uwamori K, Omori Y, Fukui T. Continuous hydrogen production by the hyperthermophilic archaeon, Thermococcus kodakaraensis KOD1. J Biotechnol 2005;116:271–82. [58] Shin HS, Youn JH, Kim SH. Hydrogen production from food waste in anaerobic mesophilic and thermophilic acidogenesis. Int J Hydrogen Energy 2004;29:1355–63.

international journal of hydrogen energy 33 (2008) 5404–5415

[59] Ginkel SV, Oh SE, Logan BE. Biohydrogen production from food processing and domestic wastewaters. Int J Hydrogen Energy 2005;30:1535–42. [60] Han SK, Shin HS. Biohydrogen production by anaerobic fermentation of food waste. Int J Hydrogen Energy 2004;29:569–77. [61] Braun H. Calculating hydrogen production costs, http:// evworld.com/article.cfm?storyid¼502; March 2003.

5415

[62] Morris D. The carbohydrate economy, biofuels and the net energy debate, http://www.newrules.org/agri/ netenergyresponse.pdf; Aug 2005. [63] Simbeck D, Chang E. Hydrogen supply: cost estimate for hydrogen pathways – scoping analysis, http://www.nrel. gov/docs/fy03osti/32525.pdf; Nov 2002. NREL/SR-54032525.