Optimization of key factors affecting biohydrogen production from microcrystalline cellulose by the co-culture of Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂

Abstract

The key factors (initial pH value, substrate concentration, incubation time, C/N and L-cysteine concentration) that affect biohydrogen production from microcrystalline cellulose in batch fermentation by the co-culture of isolated strains (Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂) were optimized using an orthogonal experiment. The isolated strain, Clostridium acetobutylicum X₉, had high hydrogen yield from microcrystalline cellulose (MCC), and Ethanoigenens harbinense B₂ could produce hydrogen efficiently from monosaccharides directly from microcrystalline cellulose. The optimal parameters were as follows: initial pH value of 6.0, 12 g L⁻¹ substrate concentration, 40 h incubation time, 0.7 g L⁻¹ L-cysteine concentration and a 4 : 1 ratio of C/N. Under the optimum culture conditions, a maximum hydrogen yield rate of 10.4 mmol g-MCC⁻¹ was obtained. This yield is approximately 2.2-fold greater than that of mono-culture Clostridium acetobutylicum X₉. It suggests that the optimal conditions achieved can be applied to the production of hydrogen from microcrystalline cellulose using the co-culture of isolated strains, Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂.

---

1 Introduction

Energy is an essential commodity for increasing productivity in both agriculture and industry. The worldwide energy demand has been increasing rapidly and this energy crisis has serious negative effects on the environment such as global warming.^1,2 As a renewable energy, biomass power has had a rapid growth in the past decade in China.^3,4 Biomass will play an important role in the global energy infrastructure in the future for the generation of power and heat, along with the production of chemicals and fuels.⁵ In recent years, many researchers have already studied biomass conversion to energy such as hydrogen, ethanol and methane.^6–10

Currently, hydrogen is produced, exclusively, by the electrolysis of water or by the steam reformation of methane. Hydrogen production via biological technologies provides a wide range of approaches to generate hydrogen,¹¹ for example biological hydrogen production from renewable lignocellulosic waste has attracted significant attention.¹² The study of Alibardi and Cossu indicated that the bread–pasta fraction in organic wastes has a marked effect on hydrogen potential production.¹³ Ren et al. presented a comprehensive review on the bioconversion of lignocellulosic biomass to hydrogen, which sheds light on the perspectives on lignocellulosic biomass conversion to hydrogen.¹⁴ To generate hydrogen directly from lignocellulose materials using dark fermentation requires expensive pretreatment processes for release of the underlying monomeric sugars such as delignification and hydrolysis.^15–17 Therefore, prior to DF, these biomasses are often subjected to physical, chemical and biological pre-treatment to increase their digestibility.¹⁸ Favaro et al. reported that a properly pre-treated inoculum could be used to improve the hydrogen yield from organic waste.¹⁹ Microcrystalline cellulose (MCC) is a cellulose derived from high quality wood pulp by acid hydrolysis, which is used to remove the amorphous regions.²⁰ It is a purified partially depolymerized non-fibrous form of cellulose that occurs as a white, odorless, tasteless, crystalline powder composed of porous particles.²¹ Therefore, microcrystalline cellulose (MCC) can be effectively utilized as a model substrate to produce hydrogen.

The co-culture of cellulolytic and hydrogen-producing strains, which takes advantage of their specific metabolic capacities, offers a promising new way to enhance the conversion efficiency of cellulose to hydrogen.²² Many studies on co-cultures have been investigated for the enhancement of hydrogen production.^12,23,24 To improve the hydrogen production efficiency, two aspects can be considered: the characterization of the key factors that affect biohydrogen production by co-culture strains and identification of the ecological relationship among the organisms.

Based on this background, the aim of this study is to explore the optimal conditions of the key factors that affect biohydrogen production by the co-culture of isolated strains Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂. In order to determine the optimal combination of factor levels, single-factor experiments and orthogonal experiments were used in this experimental design, and a series of experiments were conducted.²⁵

2 Materials and methods

2.1. Hydrogen-producing strains

The strains Clostridium acetobutylicum X₉ (NCBI: EU434651) and Ethanoigenens harbinense B₂ (NCBI: EU639425) were isolated from activated sludge in a pilot-scale continuous fermentative hydrogen production reactor (working volume: 1.48 m³, substrate: molasses). The procedure was conducted using organic loading rates of 3.11–85.8 kg COD m⁻³ d⁻¹ for over 200 days.²⁶ The strain Clostridium acetobutylicum X₉ has a typical butyrate-type fermentation metabolism with high hydrogen yield and cellulose degradation, whereas Ethanoigenens harbinense B₂, which has a 98% similarity to B₄₉, undergoes the so-called ethanol-type fermentation metabolism with high hydrogen yield.²⁷

2.2. Batch experiments

Batch experiments were carried out anaerobically in 100 mL or 250 mL serum bottles, at 37–40 °C, on an orbital shaker at a rotation speed of 90–130 rpm. The fermentation broth composition was (g L⁻¹): microcrystalline cellulose: 12, peptone: 4.0, beef extract: 2.0, yeast extract: 1.0, NaCl: 4.0, K₂HPO₄: 1.0, MgCl₂: 0.1, FeSO₄: 0.1, and L-cysteine: 0.5. Moreover, 10 mL L⁻¹ medium of vitamins (cyanocobalamin: 0.01 g L⁻¹, ascorbic acid: 0.025 g L⁻¹, riboflavin: 0.025 g L⁻¹, citric acid: 0.02 g L⁻¹, pyridoxine: 0.05 g L⁻¹, folic acid: 0.01 g L⁻¹, 4-aminobenzoic acid: 0.01 g L⁻¹, and creatine: 0.025 g L⁻¹) and micronutrients (MnSO₄·7H₂O: 0.01 g L⁻¹, ZnSO₄·7H₂O: 0.05 g L⁻¹, H₃BO₃: 0.01 g L⁻¹, N(CH₂COOH)₃: 4.5 g L⁻¹, CaCl₂·2H₂O: 0.01 g L⁻¹, Na₂MoO₄: 0.01 g L⁻¹, CoCl₂·6H₂O: 0.2 g L⁻¹, and AlK(SO₄)₂: 0.01 g L⁻¹) were added. Moreover, 1 mL of resazurin (0.2%) was added to verify whether the reaction system was in the anaerobic condition. The microcrystalline cellulose had a cellulose content of 97.2% (v/v, dry basis) and water solubility of 0.1% (w/v).

In the co-culture experiments, the two strains, Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂, were mixed in the same volumes at a total biomass quantity of 6 mL (6%). All experiments were repeated three times.

2.3. Analytical methods

In the fermentation experiments, hydrogen production, end liquid products, cellulose degradation, and quantities of reductive saccharides were measured.

Gas chromatography (SCII, Shanghai Analytical Apparatus, China) was used to determine the hydrogen content in the gas phase using a thermal conductivity detector (TCD) and nitrogen as the carrier gas (70 mL min⁻¹). A 2.0 m stainless column was packed with TDS-01 (60–80 mesh). The column and detector were both kept at 150 °C.

Another gas chromatograph (GC122, Beijing Oriental Fine Hua Yuan Co., China) was employed to detect the volatile fatty acids (VFAs) and alcohol contents in the centrifuged (4000 rpm) fermenting liquor using nitrogen as the carrier gas (flow rate of 60 mL min⁻¹). A hydrogen flame-ionization detector (FID) was employed using hydrogen and air flow rates of 50 and 490 mL min⁻¹, respectively, and a 2.0 m GDX-103 (60–80 mesh) column. The column and all cells were maintained at 190 °C.

Microcrystalline cellulose concentration was determined via the phenol–H₂SO₄ method after removal of cell mass, as described by Minato et al.²⁸ The amount of reducing sugar was determined via the DNS method using xylose (Sigma) as a standard. One unit (U) of xylanase activity was expressed as 1 μmol of reducing sugar (xylose equivalent) released in 1 min.²⁹

Based on the cumulative H₂ production data and MCC substrate content in the batch experiment, the H₂ yield and cellulose degradation were calculated as follows:

H₂ yield = n_H2/(M₁ − M₂) (1)

where n_H2 (mmol) is the maximum cumulative H₂ production, M₁ (g) is the MCC substrate content before fermentation, and M₂ (g) is the MCC substrate content after fermentation. If it is not marked, the data were measured 40 h after inoculation.

3 Results and discussion

3.1. Single-factor experiment

3.1.1. Effects of initial pH. Initial pH is important in increasing the efficiency of cellulose hydrolysis and it significantly affects the cumulative hydrogen production.^30,31 Fig. 1(a) and (c) show that hydrogen yield and cell dry weight increased with the increase in initial pH from 3.0 to 6.0 and then decreased when the initial pH was greater than 6.0. At initial pH 6.0, the maximum H₂ yield of Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂ reached 9.63 mmol g-MCC⁻¹, and cellulose degradation was 81%. Moreover, the cell dry weight reached 0.66 g L⁻¹. Fig. 1(b) shows that the maximum concentration of ethanol, butyrate and acetate in the end liquid products was 2166 mg L⁻¹, 1483 mg L⁻¹ and 1994 mg L⁻¹, respectively. In addition, the reducing sugar under different initial pH values did not accumulate to a detectable quantity in the fermenting liquid.

Fig. 1 (a) Effects of initial pH on H₂ yield/cellulose degradation; (b) effect of initial pH on end liquid products/reduced sugar of X₉ + B₂; (c) effects of pH on cell dry weight.

At initial pH 6.0, the molar ratio (ethanol/butyrate/acetate) of the end liquid products was almost 1.5 : 1.0 : 1.5. In other words, the co-cultured strains underwent the ethanol-type fermentation researched by Ren et al.³² The high content of ethanol in the end liquid products could buffer the ferment-end pH and reduce the inhibition of end products from MCC, which is beneficial for maintaining the stability of microbial growth and hydrogen formation.

3.1.2. Effects of substrate concentration. Substrate concentration, individually, has a significant influence on the optimization of hydrogen yield.^13,33 Hence, the effect of substrate concentration on biohydrogen production should be revealed. As shown in Fig. 2(a–c), hydrogen yield, cellulose degradation, end liquid products and cell dry weight increased with the increase in substrate concentration from 1.0 g L⁻¹ to 12.0 g L⁻¹. At the substrate concentration of 12.0 g L⁻¹, the maximum H₂ yield of Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂ reached 10.2 mmol g-MCC⁻¹, cellulose degradation of 86%, cell dry weight of 0.64 g L⁻¹, and the maximum concentrations of ethanol, butyrate and acetate in the end liquid products were 2320 mg L⁻¹, 1520 mg L⁻¹ and 1949 mg L⁻¹, respectively. Similarly, the reducing sugar did not accumulate to a detectable quantity in the fermenting liquid.

Fig. 2 (a) Effects of substrate concentration on H₂ yield/cellulose degradation; (b) effects of substrate concentration on end liquid products/reduced sugar of X₉ + B₂; (c) effects of substrate concentration on cell dry weight.

Understanding the dependence of substrate concentration on fermentative hydrogen production is a critical step towards optimal control. Whether substrate concentration is moderate or not, it will directly affect the state of the growth and the activation of hydrogen-producing bacteria. In particular, too high or too low substrate concentration will affect the related enzyme secretion and metabolic pathways of bacteria hydrogen production.

3.1.3. Effects of incubation time. Sreela-or et al.³⁴ and Lo et al.³⁵ discussed the effects of incubation time on hydrogen production, enzyme activity and reducing sugar production. Fig. 3(a) and (b) reveal that hydrogen yield, cellulose degradation and end liquid products increased from 26 h (mid-late log phase) to 40 h and achieved a steady state after 40 h of incubation time. At an incubation time of 40 h, the maximum H₂ yield of Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂ reached 10.2 mmol g-MCC⁻¹, and cellulose degradation of 85%. The cell dry weight remained steady after incubation of about 40 h, which led to the maximum ethanol, butyrate and acetate concentrations of 2387 mg L⁻¹, 1510 mg L⁻¹ and 2069 mg L⁻¹, respectively, in the end liquid products. Likewise, the reducing sugar did not accumulate to a detectable quantity in the fermenting liquid. Fig. 3(c) shows that the cell dry weight increased from 26 h to 40 h. At 40 h, multiple microorganisms reached a stable phase.

Fig. 3 (a) Effects of incubation time on H₂ yield/cellulose degradation; (b) effects of incubation time on end liquid products/reduced sugar of X₉ + B₂; (c) effects of incubation time on cell dry weight.

3.1.4. Effects of C/N. Carbon and nitrogen are needed for the growth and metabolism of microorganisms. A proper C/N ratio enhances the material metabolized and bacterial hydrogen produced.^36,37 In this experiment, microcrystalline cellulose (MCC) was used as the sole carbon source, and yeast extract/peptone/beef extract (1 : 1 : 1) was used as the complex nitrogen source.

Fig. 4(a–c) illustrate that the hydrogen yield, cellulose degradation, end liquid products and cell dry weight increased with an increase in C/N ratio, which ranged from 0 to 4.0. At a C/N ratio of 4.0, the maximum H₂ yield rate of Clostridium sp. X₉ + Ethanoigenens harbinense B₂ reached 10.26 mmol g-MCC⁻¹, cellulose degradation of 86%, dry cell of 0.72 g L⁻¹, and the maximum concentrations of ethanol, butyrate and acetate were 2412 mg L⁻¹, 1470 mg L⁻¹ and 2015 mg L⁻¹, respectively, in the end liquid products. Moreover, the reducing sugar also did not accumulate to a detectable quantity in the fermenting liquid.

Fig. 4 (a) Effects of C/N on H₂ yield/cellulose degradation; (b) effects of C/N on end liquid products/reduced sugar of X₉ + B₂; (c) effects of C/N on cell dry weight.

3.1.5. Effects of L-cysteine concentration. Supplementation with a reducing agent such as L-cysteine is an alternative way to maintain an anaerobic environment. Moreover, L-cysteine as a mediator between bacteria and substrate could reduce the oxidation–reduction potential (ORP) values of the fermentation system and increase the growth rate of some bacteria.^38,39 The results of Yuan et al.⁴⁰ showed that L-cysteine could be used as a low-cost and highly efficient bioactive agent to increase dark fermentative hydrogen production.

As shown in Fig. 5(a) and (b), the hydrogen yield, cellulose degradation and end liquid products increased with an increase in L-cysteine concentration, which ranged from 0 to 0.7 g L⁻¹. The H₂ yield of Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂ and cellulose degradation reached a peak of 10.3 mmol g-MCC⁻¹ and 85%, respectively, at the L-cysteine concentration of 0.7 g L⁻¹. Moreover, the maximum concentrations of ethanol, butyrate and acetate in the end liquid products were 2386 mg L⁻¹, 1500 mg L⁻¹ and 2032 mg L⁻¹, respectively. As shown in Fig. 5(c) , the cell dry weight increased with an increase in L-cysteine concentration. Similarly, in the fermenting liquid, the reducing sugar did not accumulate to a detectable quantity.

Fig. 5 (a) Effects of L-cysteine concentration on H₂ yield/cellulose degradation; (b) effects of L-cysteine concentration on end liquid products/reduced sugar of X₉ + B₂; (c) effects of L-cysteine concentration on cell dry weight.

3.2. Orthogonal experiment

Based on single-factor experiments, the factors that influence hydrogen yield were examined through an orthogonal experiment. The design and results of the orthogonal experiment L₁₆ (4⁵) are presented in Tables 1 and 2. The parameter, K, is the statistical average of hydrogen yield at one level (for one factor). The parameter, R, is the statistical range of K₁–K₄ for one factor. The different values of K show the effects of the four levels on hydrogen yield, whereas the different values of R suggest the effects of the five factors on hydrogen yield.⁴¹

Table 1 Results of the L₁₆ (4⁵) orthogonal design

Experiment no.	Parameters					RH2^c (mmol g^−1)
	SC^a (g L^−1)	Initial pH	C/N	Incubation time (h)	LC^b (g L^−1)
a SC represents substrate concentration.b LC represents L-cysteine concentration.c RH2 represents hydrogen yield.
1	10	5.0	2.5	36	0.3	8.5
2	10	5.5	3.3	38	0.5	9.6
3	10	6.0	4.0	40	0.7	10.1
4	10	6.5	5.0	42	1.0	9.7
5	12	5.0	3.3	40	1.0	9.1
6	12	5.5	2.5	42	0.7	9.9
7	12	6.0	5.0	38	0.5	10.3
8	12	6.5	4.0	36	0.3	9.4
9	15	5.0	4.0	42	0.5	8.7
10	15	5.5	5.0	40	0.3	9.7
11	15	6.0	2.5	38	1.0	9.8
12	15	6.5	3.3	36	0.7	9.2
13	18	5.0	5.0	38	0.7	8.3
14	18	5.5	4.0	36	1.0	8.6
15	18	6.0	3.3	42	0.3	9.0
16	18	6.5	2.5	40	0.5	9.3

---

Table 2 Analysis of the L₁₆ (4⁵) experiment results

	Hydrogen yield
	SC (g L^−1)	Initial pH	C/N	Incubation time (h)	LC (g L^−1)
a K represents the average of hydrogen yield of four experiments at one level (for one factor).b R represents the range of K1–K4 for one factor.
K1^a	9.475	8.650	9.375	9.150	9.150
K2	9.675	9.450	9.225	9.275	9.375
K3	9.350	9.800	9.500	9.550	9.475
K4	8.800	9.400	9.200	9.325	9.300
R^b	0.875	1.150	0.300	0.400	0.325
Optimal result	SC2	pH3	C/N3	Time3	LC3

---

According to data analysis in Table 2, the order of influence strength is initial pH > substrate concentration > incubation time > L-cysteine concentration > C/N. The optimum hydrogen yield condition was initial pH 6.0, substrate concentration 12 g L⁻¹, incubation time 40 h, L-cysteine concentration 0.7 g L⁻¹ and C/N 4.0. A verification experiment was carried out under the optimal condition, and the hydrogen yield was 10.4 mmol g-MCC⁻¹.

The stain B₂ and X₉ have different abilities for bio-hydrogen production from cellulose. In the mono-culture test, X₉ achieved a much higher hydrogen yield than of B₂.⁴² In the B₂ and X₉ co-culture test, a maximum hydrogen yield of 10.4 mmol g-MCC⁻¹ was obtained under the optimum condition, which was approximately 2.2-fold greater than the mono-culture Clostridium acetobutylicum X₉.⁴⁰ This reveals that the co-culture of strain B₂ and X₉ can achieve bioaugmentation effects for hydrogen production from cellulose.

Moreover, the efficiency of the X₉ + B₂ co-culture cellulosic H₂ production system is comparable to that reported in the other studies (Table 3). The results indicate that co-culture of Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂ presents a potential approach to converting cellulose into hydrogen energy.

Table 3 Comparison of H₂ production performance using cellulosic material as the substrate^a

Microbe	Substrate	Temperature (°C)	H2 yield (mmol g^−1)	Reference
a MC microcrystalline cellulose, CMC carboxymethyl cellulose.
X9	Stream-exploded corn stover (15 g L^−1)	37	3.4	Wang et al.^43
X9 + B49	MCC (10 g L^−1)	38	8.1	Wang et al.^43
JN4 + GD17	MCC, filter paper or cellobiose (5 g L^−1)	60	18	Liu et al.^44
G1 + B49	MCC (5 g L^−1)	37	2.97	Wang et al.^45
Sludge & Clostridium pasteurianum	CMC (10 g L^−1)	35	1.09	Lo et al.^46
X9 + B9	MCC (12 g L^−1)	37	10.4	This study

---

4 Conclusions

This study explores the optimization of the key factors that affect biohydrogen production from microcrystalline cellulose by the co-culture of isolated strains Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂. In single-factor experiments, hydrogen production, end liquid products, cellulose degradation, and quantities of reductive saccharides were measured versus initial pH, substrate concentration, incubation time, C/N, and L-cysteine concentration. Based on these single-factor experiments, the factors that influence the hydrogen yield were examined through an orthogonal experiment. The influence strength sequence of the factors is initial pH > substrate concentration > incubation time > L-cysteine concentration > C/N. The determined optimal conditions were initial pH 6.0, substrate concentration 12 g L⁻¹, incubation time 40 h, L-cysteine concentration 0.7 g L⁻¹ and C/N 4.0. Under the optimum condition, a maximum hydrogen yield of 10.4 mmol g-MCC⁻¹ was obtained, which was approximately 2.2-fold greater than the mono-culture Clostridium acetobutylicum X₉ in our previous study.⁴² The two strains were isolated from the same habitat and there exists an ecological niche complementarity between them. The corresponding end liquid products were acetate, ethanol, and butyrate. In the stable hydrogen production phase, the ethanol content was 2400 mg L⁻¹, which was 1.6-fold greater than butyrate. The increasing neutral ethanol content and component could avoid the effect of acid products on the microbial metabolic processes. In this stage, the pH range of variation, high cellulose degradation, hydrogen production and microbe activity would be maintained, and the hydrogen production cycle would be prolonged. Hence, the co-culture of the strain Ethanoigenens harbinense B₂ and Clostridium acetobutylicum X₉ can achieve bioaugmentation effects for hydrogen production from cellulose and is more competitive than the mono-culture in cellulose conversion. Our research results indicate that the dark fermentation of cellulosic biomass by the co-culture of Clostridium acetobutylicum X₉ + Ethanoigenens harbinense B₂ should be further developed. It has the potential use for converting cellulose and hemicellulose into hydrogen energy.

Acknowledgements

This study was supported by the National Natural Science Foundation (30470054), the Scientific and Technological Project of Liaoning Province (2001304024), the natural science foundation of Liaoning Province (No. 20120132), the Liaoning province science and the cause of Public Research Fund (No. 20111012), the Bureau of Shenyang city science and Technology Research Foundation (No. F12-277-1-39) and the City State Key Laboratory of water resources and water environment of open fund (No. HC201214).

References

1. I. K. Kapdan and F. Kargi, Enzyme Microb. Technol., 2006, 38, 569–582.
2. A. Midilli, M. Ay, I. Dincer and M. A. Rosen, Renewable Sustainable Energy Rev., 2005, 9, 255–271.
3. Z. y. Zhao and H. Yan, Renewable Energy, 2012, 37, 53–60.
4. J. Li and J. Ge, Procedia Environ. Sci., 2011, 10, 2153–2158.
5. E. Kırtay, Energy Convers. Manage., 2011, 52, 1778–1789.
6. K. Zhang, N. Ren, C. Guo, A. Wang and G. Cao, J. Environ. Sci., 2011, 23, 1929–1936.
7. F. Xu, K. Theerarattananoon, X. Wu, L. Pena, Y. C. Shi, S. Staggenborg and D. Wang, Ind. Crops Prod., 2011, 34, 1212–1218.
8. M. Asgher, Z. Ahmad and H. M. N. Iqbal, Ind. Crops Prod., 2013, 44, 488–495.
9. Q. Zhang, L. Tang, J. Zhang, Z. Mao and L. Jiang, Bioresour. Technol., 2011, 102, 3958–3965.
10. X. Y. Cheng, Q. Li and C. Z. Liu, Bioresour. Technol., 2012, 114, 327–333.
11. D. Levin, Int. J. Hydrogen Energy, 2004, 29, 173–185.
12. Q. Li and C.-Z. Liu, Int. J. Hydrogen Energy, 2012, 37, 10648–10654.
13. L. Alibardi and R. Cossu, Waste Manag., 2015, 36, 147–155.
14. N. Ren, A. Wang, G. Cao, J. Xu and L. Gao, Biotechnol. Adv., 2009, 27, 1051–1060.
15. M. Cui, Z. Yuan, X. Zhi, L. Wei and J. Shen, Int. J. Hydrogen Energy, 2010, 35, 4041–4047.
16. R. Datar, J. Huang, P. Maness, A. Mohagheghi, S. Czernik and E. Chornet, Int. J. Hydrogen Energy, 2007, 32, 932–939.
17. I. A. Panagiotopoulos, R. R. Bakker, T. de Vrije, E. G. Koukios and P. A. M. Claassen, Int. J. Hydrogen Energy, 2010, 35, 7738–7747.
18. A. Ghimire, L. Frunzo, F. Pirozzi, E. Trably and R. Escudie, Appl. Energy, 2015, 144, 73–95.
19. L. Favaro, et al., Int. J. Hydrogen Energy, 2013, 38, 11774–11779.
20. A. M. Adel, Z. H. Abd El-Wahab, A. A. Ibrahim and M. T. Al-Shemy, Carbohydr. Polym., 2011, 83, 676–687.
21. P. Mathew, K. Oksman and M. Sain, J. Appl. Polym. Sci., 2005, 97(5), 2014–2025.
22. A. Geng, Y. He, C. Qian, X. Yan and Z. Zhou, Bioresour. Technol., 2010, 101, 4029–4033.
23. C. H. Chou, C.-L. Han, J.-J. Chang and J.-J. Lay, Int. J. Hydrogen Energy, 2011, 36, 13972–13983.
24. S. Wu, X. Li, J. Yu and Q. Wang, Bioresour. Technol., 2012, 123, 184–188.
25. S. Ma, H. Wang, Y. Wang, H. Bu and J. Bai, Renewable Energy, 2011, 36, 709–713.
26. A. Wang, Int. J. Hydrogen Energy, 2008, 33, 912–917.
27. N. Q. Ren, X. Q. Wang, W. S. Xiang, M. Lin, J. Z. Li and W. Q. Guo, High Technol. Lett., 2002, 8, 21–25.
28. H. Minato, A. Endo, H. Kouriyama and T. Uemura, Nippon Nogeikagaku Kaishi, 1962, 36, 101–106.
29. G. L. Miller, J. Anal. Chem., 1959, 31, 426–428.
30. Y. C. Lo, M. D. Bai, W. M. Chen and J. S. Chang, Bioresour. Technol., 2008, 99, 8299–8303.
31. Y. Fan, Bioresour. Technol., 2004, 91, 189–193.
32. N. Q. Ren, B. Z. Wang and J. C. Huang, Biotechnol. Bioeng., 1997, 54, 428–433.
33. S. V. Ginkel, S. Sung and J.-J. Lay, Environ. Sci. Technol., 2001, 35, 4726–4730.
34. C. Sreela-or, T. Imai, P. Plangklang and A. Reungsang, Int. J. Hydrogen Energy, 2011, 36, 14120–14133.
35. Y. C. Lo, Y. C. Su, C. Y. Chen, W. M. Chen, K. S. Lee and J. S. Chang, Bioresour. Technol., 2009, 100, 5802–5807.
36. C. Lin, Int. J. Hydrogen Energy, 2004, 29, 41–45.
37. Q. Li, D. Xing, N. Ren, L. Zhao and Y. Song, J. Environ. Sci., 2006, 27, 810–814.
38. R. a. Doong and B. Schink, Environ. Sci. Technol., 2002, 36, 2939–2945.
39. Y. Song and B. E. Logan, Water Res., 2004, 38, 1626–1632.
40. Z. Yuan, H. Yang, X. Zhi and J. Shen, Int. J. Hydrogen Energy, 2008, 33, 6535–6540.
41. H. Su, J. Cheng, J. Zhou, W. Song and K. Cen, Int. J. Hydrogen Energy, 2009, 34, 8846–8853.
42. H. X. Bao, W. W. Cai, X. P. Ma, Y. T. Song, M. L. Shen, Z. L. Chen, L. D. Li and N. Q. Ren, Adv. Mater. Res., 2012, 512–515, 1446–1449.
43. A. Wang, N. Ren and Y. Shi, et al., Int. J. Hydrogen Energy, 2008, 33, 912–917.
44. Y. Liu, P. Yu and X. Song, et al., Int. J. Hydrogen Energy, 2008, 33, 2927–2933.
45. A. Wang, L. Gao and N. Ren, et al., Biotechnol. Lett., 2009, 31, 1321–1326.
46. Y. C. Lo, M. D. Bai, W. M. Chen and J. S. Chang, Bioresour. Technol., 2008, 99, 8299–8303.

---