ARTICLE IN PRESS

0961-9534/$ - se

doi:10.1016/j.bi

�Correspond+886 7 6011000

E-mail addr

lay@ccms.nkfu

Biomass and Bioenergy 31 (2007) 250–254

www.elsevier.com/locate/biombioe

Enhanced biohydrogen production from cornstalk wastes withacidification pretreatment by mixed anaerobic cultures

Mao-Lin Zhanga, Yao-Ting Fana,�, Yan Xinga, Chun-Mei Pana,Gao-Sheng Zhanga, Jiunn-Jyi Layb,�

aDepartment of Chemistry, Zhengzhou University, Zhengzhou 450052, PR ChinabDepartment of Safety, Health, and Environmental Engineering, National Kaohsiung First University of Science and Technology, Kaohsiung, Taiwan, ROC

Received 11 January 2006; received in revised form 26 July 2006; accepted 5 August 2006

Available online 19 December 2006

Abstract

Biohydrogen production from the cornstalk wastes with acidification pretreatment was reported in this paper. Batch tests were carried

out to analyze influences of several environmental factors on biohydrogen production from cornstalk wastes. Two predominant bacterial

morphologies, namely spore-forming rod shape bacteria and micrococcus were screened, purified, and identified after enriched from a

hydrogen-producing fermentor with cow dung composts. The maximum cumulative H2 yield of 149.69mlH2 g�1 TVS was obtained at

initial pH 7.0 and substrate concentration 15 g l�1, the value is about 46-fold as compared with that of raw cornstalk wastes. The

maximum hydrogen production rate was 7.6mlH2 h�1. The hydrogen concentration in biogas was 45–56% (v/v) and there was no

significant methane observed in the biogas throughout this study. In addition, biodegradation characteristics of the substrate by

microorganisms were also discussed. During the conversion of cornstalk wastes into hydrogen, the acetate, propionate, butyrate, and the

ethanol were main by-products in the metabolism of hydrogen fermentation. The test results showed that the acidification pretreatment

of the substrate plays a crucial role in conversion of the cornstalk wastes into biohydrogen gas by the cow dung composts generating

hydrogen.

r 2006 Elsevier Ltd. All rights reserved.

Keywords: Biohydrogen production; Acid pretreatment; Alkali pretreatment; Cornstalk; Mixed culture

1. Introduction

Considering global environmental impacts, such asgreenhouse effect and resource recovery, there is a pressingneed to develop non-polluting and renewable energysource. Compared with fossil fuels as traditional energysources, hydrogen is a promising candidate as a cleanenergy carrier in the future because of its high-energy yield(122 kJ g�1) and producing only water instead of green-house gases on burning [1–5].Compared with traditional hydrogen-production process

by physical and chemical methods, microbial conversion ofbiomass, such as agricultural and industrial wastes and

e front matter r 2006 Elsevier Ltd. All rights reserved.

ombioe.2006.08.004

ing authors. Tel./fax: +86 371 67766017;

x2303.

esses: yt.fan@zzu.edu.cn (Y.-T. Fan),

st.edu.tw (J.-J. Lay).

residues, into biohydrogen gas using fermentative bacteria,is an environmentally friendly and energy-saving process,and is attracting increasing interest as a useful way ofconverting biomass to hydrogen [5–7]. Most studies ofbiohydrogen production so far, however, have been limitedto using pure carbohydrates, such as glucose, sucrose andstarch [5–10]. Little information is available on thefeasibility of using the biomass containing crude cellulosesuch as corn stover. Recently, the conversion of wheatstraw and beer into hydrogen gas by mixed anaerobicculture was carried out in our lab [11,12]. However,bioconversion of agricultural residues into biohydrogengas by microorganisms is surprisingly lacking because oftheir complex chemical composition, e.g., cellulose, hemi-cellulose, lignin, protein and fat.It is well known that biomass residues, such as crop

stalks, are persistent in the environment and remains asenvironmental pollutants. It is reported that the annual

ARTICLE IN PRESSM.-L. Zhang et al. / Biomass and Bioenergy 31 (2007) 250–254 251

yields of biomass residues exceed about 220 billion tons inthe world, the value are equivalent to the energy of 60–80billion tons crude oil. In China alone, the annual yield ofbiomass wastes exceed about 0.7 billion tons, among them,the annual yields of cornstalk, wheat straw and strawwastes are around 220, 110 and 180 million tons,respectively [11,12]. Except that some of them were usedto make paper or as feedstuff for livestock, most of themwere set on fire or discarded as environmental pollutants.The biomass residues can, however, be a valuable and vastrenewable resource.Based on this background, our research interest is to

explore the feasibility of conversion of corn stalk residueswith acidification pretreatment into biohydrogen gas by themixed culture. A expected hydrogen yield of149.69mlH2 g

�1 TVS from corn stalk residues withacidification pretreatment was obtained by mixed anaero-bic cultures in this study. To the best of our knowledge, thisis the first report of biohydrogen production fromcornstalk residues by mixed culture so far.

2. Experimental methods

2.1. Hydrogen-producing microflora

Hydrogen-producing microflora was taken from cowdung compost in the suburb of Zhengzhou City, which hadbeen composting for 3 months.

2.2. Acidification pretreatment of the substrate

The cornstalk residues used as substrate were obtainedfrom the suburbs of Zhengzhou City. Before they weredegraded by microorganisms, the raw cornstalks wereground separately by a vegetation disintegrator (FZ-102,250 kw, Beijing Yong Guang Ming Medical Appliance

Factory, China) to pass 20-mesh screen. The grindingsample was employed as substrate of hydrogen productionin the batch experiments, then the mixture of the groundcornstalks and dilute HCl were boiled in a beaker for30min and neutralized to pH 7 with either dilute NaOH orHCl solution. TVS value was determined as follows:

TVS ¼Wdry corn stalk � W ash

Wdry corn stalk� 100%.

Here; TVS ¼ 0:8745Wdried cornstalk.

2.3. Experimental procedure

The batch experiments were carried out with 250mlserum vials as batch reactors filled with 150ml comprisingthe mixture of the composts, the pretreated cornstalks, and3ml of nutrient stock solution. These vials were gassedwith nitrogen gas to remove oxygen and the headspace ofthe reactors to keep the anaerobic environment. The

bottles were incubated at 3671 1C and operated in anorbital shaker with a rotation speed of 90 rpm to providebetter contact among substrates. Each liter of nutrientstock solution contained 80 g of NH4HCO3, 12.4 g ofKH2PO4;0.1 g of MgSO4 � 7H2O;0.01 g of NaCl;0.01 g ofNa2MoO4 � 2H2O;0.01 g of CaCl2 � 2H2O;0.015 g ofMnSO4 � 7H2O;0.0278 g of FeCl2, which was slightlymodified from Lay [5]. The volume of biogas wasdetermined using glass syringes of 5–50ml.

2.4. Analytical methods

The hydrogen gas percentage (H2%) was determined bycomparing the sample biogas with a standard of purehydrogen using a gas chromatograph (GC, Agilent 4890D)equipped with a thermal conductivity detector (TCD) and6-foot stainless-steel column packed with Porapak Q(80/100mesh). The operational temperatures of the injectionport, the oven and the detector were 1001, 801 and 1501,respectively. Nitrogen was used as the carrier gas at a flowrate of 20mlmin�1. The concentrations of the volatile fattyacids (VFAs) and the alcohols were analyzed using anotherGC of the same model with a flame ionization detector(FID) and an 8-ft stainless-steel column packed with 10%PEG-20M and 2% H3PO4 (80/100 mesh). The temperatureof the injection port, the detector and the oven were 220,240 1C and a programmed column temperature of130–175 1C, respectively. Nitrogen was used as the carriergas at a flow rate of 20mlmin�1. The flow rate of hydrogenand air was 30mlmin�1. The pH values inside the digesterswere determined by a microcomputer pH-vision 6071.

3. Results and discussion

3.1. Seed hydrogen-producing microflora and microscopic

observations

A mixture of the cow dung composts and water(v/v ¼ 1:5) was continuously aerated by forced air for 3 hat 501 in order to inhibit the bioactivity of hydrogenconsumers and to harvest high yield hydrogen-producingspore-forming anaerobes.A close examination revealed two predominant bacterial

morphologies, namely spore-forming rod-shaped bacteriaFig. 1(a) and micrococcus Fig. 1(b), that were screened,purified, and identified after enrichment from a hydrogen-producing fermentor with cow dung composts. Morpho-logical, physico-biochemical characteristics and compara-tive sequence analysis of 16S rDNA indicated that twodominant strains belonged to Clostridium sp. Fig. 1illustrated the micrographs of spore-forming rod-shapedbacteria and micrococcus. As illustrated in Fig. 1, the rod-shaped bacteria were in the size range of 2.5–5 mm in lengthand 0.4–0.9 mm in width, which were spore-formingobligate anaerobic bacteria with maximal hydrogen-produ-cing potential of 2.088mol H2mol

�1 glucose. The micro-coccus were in the range of 0.9–2 mm in size, which also

ARTICLE IN PRESS

0

40

80

120

160

0 0.2 0.4 0.6 0.8 1

Cum

ula

tive

H2 (

mL

/gT

VS

)

HCl concentration (%)

Fig. 2. The effects of pretreatment of substrate using dilute HCl on H2

Fig. 1. Micrographs of rod-shaped bacteria (a) and micrococcus (b).

M.-L. Zhang et al. / Biomass and Bioenergy 31 (2007) 250–254252

were spore-forming obligate anaerobic bacteria withmaximal hydrogen-producing potential of 0.637molH2

mol�1 glucose.

yield.

Fig. 4. SEM images of the raw cornstalk (a) and the cornstalk residue

after fermentation producing hydrogen (b).

0

20

40

60

80

0 0.5 1.5 2

NaOH concentration (%)

cum

ula

tive

H2 (

mL

/g T

VS

)

1

Fig. 3. The effects of pretreatment of substrate using dilute NaOH on H2

yield.

3.2. Effect of acidification pretreatment of substrate on

hydrogen yield

Figs. 2 and 3 depicted the effects of the changes in theHCl and NaOH concentrations on H2 yield at the initialpH 7.0 and the substrate concentration of 15.0 g l�1,respectively. As can be seen from Fig. 2, the H2 yieldincreased sharply with the increase of HCl concentration inthe range of 0.04–0.2%, and then decreased sharply withincreasing HCl concentration. Maximum hydrogen yield of149.69mlH2 g

�1 TVS was observed at 0.2% HCl concen-tration for the substrate of acidification pretreatment.As shown in Fig. 3, the change curve of H2 yield by

NaOH pretreatment was different from that by HClpretreatment, the H2 yield increased remarkably with theincrease of NaOH concentration. Maximum H2 yield of56.70mlH2 g

�1 TVS occurred at 0.5% NaOH concentra-tion for the substrate of alkali pretreatment, and then theH2 yield decreased gradually with increasing NaOHconcentration. In contrast, the acidification pretreatmentof the substrate was a more effective method. As far as ourknowledge goes, the direct conversion of raw cornstalk intobiohydrogen gas by microorganisms is very difficultbecause of its complex polymer structure, e.g., the maximalH2 yield of 3.16mlH2 g

�1 TVS was only obtained inconversion of the raw cornstalk into biohydrogen by cowdung compost in our batch tests.In order to explain the mechanism of production of

hydrogen from the corn stalk, the main composition of theraw cornstalk wastes was analyzed as follows: moisture,8.96%; hemicellulose, 20.87%; cellulose, 38.92%; lignin,21.52%; protein, 2.74%; and unidentified materials,6.99%. Compared with the raw cornstalks, the solublesugar content increased from 0.90% to 3.11%, and thecellulose and hemicellulose contents decreased from40.51% and 19.41% to 37.3% and 10.89% for the acidpretreated cornstalk, respectively. The results implied thatthe enhanced H2 yield very nearly coincided with anincrease in the soluble sugar in the composition of thesubstrate. In other words, the acidification pretreatment of

the corn stalk plays an important role in the conversion ofcornstalk into biohydrogen by microorganisms.Fig. 4 illustrated the SEM images of the raw cornstalk

(a) and the hydrogen-producing residue (b). Comparedwith Fig. 4(a), the finely porous cellular textures in thehydrogen-producing residue were observed (b), in whichthe average pore diameter was in the range of 10–20 mm.These results showed that the complex polymer structures

ARTICLE IN PRESS

0

40

80

120

160

4 5 6 7 8 9

Initial pH

Cum

ula

tive

H2 (

mL

/g T

VS

)

Fig. 6. Effect of initial pH on cumulative H2 yield.

0 20 40 60 80 100 120 140 160 180

0

300

600

900

1200

0

80

160

240

0

40

80

120

160

4.5

5.0

5.5

6.0

6.5

7.0

Vola

tile

aci

ds

(mg/L

)

Incubation Time (h)

Acetate

Propionate

Butyrate

Cum

ula

tive

H2

yie

ld (

mlH

2/g

TV

S)

pH

Eth

anol

(mg/L

)

a

b

c

d

Fig. 7. Developments of operating pH, cumulative H2 yield, VFAs and

alcohols in the batch reactor during the conversion of the substrate to

biohydrogen gas.

M.-L. Zhang et al. / Biomass and Bioenergy 31 (2007) 250–254 253

of the raw cornstalks were effectively biodegraded bymicroorganism.

3.3. Effects of substrate concentrations on cumulative H2

yield

Fig. 5 showed the effects of the substrate concentrationsversus cumulative H2 yield at initial pH 7.0 by themicroorganisms. As can be seen from Fig. 5, thecumulative H2 yield increased remarkably with increasingthe substrate concentration from 47.62mlH2

�g�1 TVS atthe substrate concentration of 5 g l�1 to 143.79mlH2 g

�1

TVS at the substrate concentration of 15 g l�1. Thereafter,the cumulative H2 yield slightly decreased as substrateconcentration increased from 135.11mlH2 g

�1 TVS at thesubstrate concentration of 20 g l�1 to 122.80mlH2 g

�1 TVSat the substrate concentration of 40 g l�1. Maximum H2

yield of 143.79mlH2 g�1TVS occurred at the substrate

concentration of 15 g l�1. The results were expected becausethe excessive substrate concentration would result in theaccumulation of VFAs, a fall of pH value in the reactor,and an increase in the partial pressure of hydrogen in thebatch reactor, in which case the microorganisms wouldswitch to alcohol production, thus inhibiting hydrogenproduction [1,3,5].

3.4. Effect of initial pH on cumulative H2 yield

The changes of the cumulative H2 yields at differentinitial pH values from 4.0 to 9.0 were plotted in Fig. 6. Asshown in Fig. 6, the cumulative H2 yield sharply increasedfrom 15.15mlH2 g

�1 TVS at the initial pH 4.0 to141.56mlH2 g

�1 TVS at the initial pH 7.0. Thereafter,the H2 yield slightly decreased with increasing the initialpH from 141.56mlH2 g

�1 TVS at the initial pH 7.0 to123.08mlH2 g

�1 TVS at the initial pH 9.0. When the initialpH value in the reactor was less than 4.0, the hydrogenproduction stopped.The maximum H2 yield of 141.56ml H2 g

�1 TVSoccurred at the initial pH of 7.0. The results were consistentwith what our previous most reports [5,11,12]. Usually the

40

60

80

100

120

140

160

5 10 15 20 25 30 35 40

Substrate concentration (g/L)

Cu

mu

lati

ve

H2 (

mL

/gT

VS

)

Fig. 5. Effect of substrate concentrations on cumulative H2 yield.

pH control could significantly affect the H2 production,and stimulate the microorganisms to produce hydrogenand would achieve the system having a maximum H2 yield,but the activity of hydrogenase would be inhibited by lowor high pH values in overall hydrogen fermentation[3,10–12].

3.5. Biodegradation characteristics of the substrate

Fig. 7 showed the changes of the operating pH value inthe batch reactor (a), the accumulative H2 yield (b), alcohol(c) and VFAs (d) during the conversion of the pretreatedcornstalk to biohydrogen by cow dung compost. As shownin Fig. 7(a) and (b), the hydrogen evolution began to occur

ARTICLE IN PRESSM.-L. Zhang et al. / Biomass and Bioenergy 31 (2007) 250–254254

after culture 11.7 h and the H2 yield increased upto143.81ml g�1 TVS at 96 h. The maximum H2 yield of144.4mlH2 g

�1TVS was observed at 150.5 h. The hydrogenpercentage in the biogas was 45–56% and there was nosignificant methane observed in the batch tests. As shownin Fig. 7(b), the operating pH value in the batch reactordecreased from 7.0 to 4.62 with the progress of hydrogenevolution and substrate degradation, the optimum pHvalue of hydrogen production appeared in the range of5.37–4.62. Hydrogen production was accompanied withthe formation of VFAs and alcohols (Fig. 7(c) and (d)).During this period, acetate, propionate and butyratereached maximum yields of 928.5, 922.02 and1138.96mg l�1 at about 126.5 h, respectively. The ethanolbegan to produce after 25 h and increased upto187.18mg l�1 at 126.5 h. The contents of ethanolachieved the maximum value of 254.24mg l�1 at 174 h,but amounts of propanol and butanol weren’t detected inthe batch tests. Meanwhile, VFAs accounted for 92% oftotal metabolic by-products. Hydrogen production stoppedwhen the available substrate was consumed, while acetate,propionate, butyrate, and ethanol, as significant by-products, were left in the batch reactor.In order to quantitatively describe the cumulative

hydrogen production under the optimal condition ofproducing hydrogen (substrate concentration of 15 g l�1

and initial pH of 7.0), a modified Gompertz equation [3,5]was used to fit the experimental data:

H ¼ P exp � expRme

Pl� tð Þ þ 1

� �� �,

where H is the cumulative hydrogen production (calcd.value , mlH2 g

�1 TVS), l is the lag time (h), P the hydrogenproduction potential (found value , ml H2 g

�1 TVS), Rm isthe maximum hydrogen production rate (ml/h), e is the2.718281828. The corresponding values of P, Rm and lwere 147.68mlH2 g

�1 TVS, 7.6ml H2 h�1 and 11.7 h,

respectively, which were estimated using the solver functionin Excel (version 5.0, Microsoft) with a Newtonianalgorithm. Here, R2 was 0.989047, indicating that theparameters were statistically significant.

4. Conclusions

Effective hydrogen production from cornstalk acidifica-tion pretreatment was carried out at substrate concentra-tion 15 g l�1 and initial pH 7.0 using an anaerobic cowdung compost as seed after forced aeration treatment by airpump for 3 h at 50 1. Two predominant bacterial morphol-ogies, namely spore-forming rod-shaped bacteria andmicrococcus were screened, purified, and identified afterenrichment from a hydrogen-producing fermentor withcow dung composts. The maximum H2 yield of149.69mlH2 g

�1 TVS occurred after pretreating cornstalkswith 0.2% HCl concentration, the value was about 46 fold

higher than that of raw cornstalks. The maximumhydrogen production rate was 7.6mlH2 h

�1. The hydrogenpercent in the biogas was 45–56% (v/v) and there was nosignificant methane observed. During the conversion ofcornstalks into hydrogen, acetate, propionate, butyrate,and ethanol were main metabolic by-products in the batchtests.Our test results showed that the enhanced H2 yield very

nearly coincided with an increase in the soluble sugar in thecomposition of the substrate. In other words, theacidification pretreatment of the corn stalk plays animportant role in the conversion of cornstalk intobiohydrogen by microorganisms.

Acknowledgements

The authors wish to thank the China National Key BasicResearch Special Funds(No.2005CB214501), the NationalNatural Science Foundation of China (No.20471053 and90610001), and the Energy & Technology Program fromZhengzhou University for the financial support of thisstudy.

References

[1] Lin CY, Lay CH. Carbon/nitrogen-ratio effect on fermentative

hydrogen production by mixed microflora. International Journal of

Hydrogen Energy 2004;29:41–5.

[2] Pinto FAL, Troshina O, Lindblad P. A brief look at three decades of

research on cyanobacterial hydrogen evolution. International Journal

of Hydrogen Energy 2002;27:1209–15.

[3] Lay JJ, Li YY, Noike T. A mathematical model for methane

production from landfill bioreactor. Journal of Environmental

Engineering ASCE 1998;124:730–6.

[4] Hart D. Hydrogen power: the commercial future of ‘‘the ultimate

fuel’’. London: Financial Times Energy Publishing; 1997.

[5] Fan YT, Li CL, Lay JJ, Hou HH, Zhang GSh. Optimization of initial

substrate and pH levels for germination of sporing hydrogen-

producing anaerobes in cow dung compost. Bioresource Technology

2004;91(2):189–93.

[6] Fang HHP, Zhang T, Liu H. Effect of pH on hydrogen production

from glucose by a mixed culture. Bioresource Technology

2002;82:87–93.

[7] Ginkel SV, Sung SW, Lay JJ. Biohydrogen production as a function

of pH and substrate concentration, Environ. Science and Technology

2001;35(24):4726–30.

[8] Ueno Y, Haruta S, Ishii M, Igarashi Y. Microbial community in

anaerobic hydrogen-producing microflora enriched from sludge

compost. Applied Microbiology and Biotechnology 2001;57:555–62.

[9] Van Ginkel S, Sung S, Lay JJ. Biohydrogen production as a function

of pH and substrate concentration. Environmental Science and

Technology 2001;35(24):4726–30.

[10] Fang HHP, Li Chl, Zhang T. Acidophilic biohydrogen production

from rice slurry. International Journal of Hydrogen Energy

2006;31:683–92.

[11] Fan YT, Zhang YH, Zhang SF, Hou HW, Ren BZ. Bioresource

Technology 2006;97:500–5.

[12] Fan YT, Zhang GS, Guo XY, Xing Y, Fan MH. Biohydrogen-

production from beer lees biomass by cow dung compost. Biomass &

Bioenergy 2006;30:493–6.