~.I~

lI

~,~.[..

...!<,

Life ScienceJournal,3 (2) ,2006 ,Neves, et al, Wheat Milling By-products Fermentation

.t

Wheat Milling By-products Fermentation: PotentialSubstrate for Bioethanol Production

,I

Marcos Antonio das Nevesl, Naoto Shimizul, Toshinori Kimuri, Kiwamu Shiiba3..

1. Graduate School of Life and EnvIronmental Sciences, University of Tsukuba ,

1-1-1 Tennodai, Tsukuba, Japan

2. Graduate School of Agriculture, Hokkaido University, Kita 9, Nishi 9, Kita-ku,

Sapporo. Hokkaido, Japan

3. Nisshin Flour Milling Co. Ltd. 25, Kanda-Nishiki-cho 1-chome,

Chiyoda-ku, Tokyo, Japan

Abstract:An overview on the potential application of wheat milling by-products for bioethanol production is made.The fermentation performance of low-grade wheat flour (LG) and wheat bran (WE) was evaluated and comparedto wheat flour (WF). a-amylase or cellulase was used for liquefaction, followed by simultaneous saccharification andfermentation (SSF) by glucoamylase and Zymomonas mobilis. The final ethanol concentration, overall productivityand yield obtained from LG (51. 4 g ethanollL, 2.72 g ethanollL' hand O. 17 g ethanol/g flour, respectively)were considerably higher compared to WE (18.1 gIL, 1. 09 gIL' hand 0.02 gig). High LG fermentation rates,reaching the highest ethanol productivity (4. 4 gIL' h) within 6 h of SSF, indicated considerable savings on fermen-tation time, compared to current industrial processes. [Life Science Journal. 2006;3(2) :83 - 87J (ISSN: 1097-8135) .

~I

..I

t

~

,~

(Keywords: wheat; flour; by-product; bran; fermentation; ethanol

Abbreviations:LG: low-grade wheat flour; fl.: growth rate; P: ethanol production; Q: ethanol productivity; Qv:overall volumetric ethanol productivity; Rp: solid residue; SSF: simultaneous saccharification and fermentation;WE: wheat bran; WF: wheat flour; ¥L: liquefaction yield; ¥p;s: ethanol yield

1 Introduction

}

With the search for alternative renewable en-ergy sources, biofuels are becoming a viable solu-tion, as they are non-fossil fuels from renewablesources. Of all biofuels, ethanol is already producedon a fair scale worldwide. The bulk of the produc-tion is located in Brazil and the USA.

Various studies on ethanol conversion systemsfrom wheat products have been conducted based onthe utilization of raw wheat flour or damaged wheatgrains (e. g. Montesinos & Navarro 2000, Sureshet al. 1999). However, only a few reports onwheat milling by-products fermentation are avail-able (e. g. Adrados et al. 2005). Thus, the objec-tives of this study were to develop a simultaneoussaccharification and fermentation (SSF) process inbatch mode for wheat milling by-products and to e-valuate the fermentation performance of low-gradewheat flour (LG) and wheat bran (WB), com-pared to wheat flour (WF).

\~I

.I

I

~t,

2 Materials and Methods

2 .1 MaterialsRaw material: LG, WB and WF. Typical

composition was summarized in Table 1.Bacterial cells: Zymomonas mobilis NBRC

13758.Enzymes: a-amylase (51 U/mg, Sigma,

USA); glucoamylase (23 U/mg, Sigma, USA);cellulase (106 U /mg, MP Biochemicals).

Liquefaction: One liter slurries containing 200g dried matterll of LG (pH 6.1), WB (pH 5.9)or WF (pH 5.7) were hydrolyzed separately using400 U a-amylase/gflour (in the caseof LG or WF)or cellulase (for WB) at 55 "C, 100 rpm for 2 h.

SSF: 200 U glucoamylase/g flour and 100 mlstarter culture (3. 1 X 108 viable cells/ml) wereadded to the liquefied slurry and the SSF was con-ducted in a previously sterilized 2 liter jar fermentor(Marubishi) at 35 "C, 100 rpm, pH 4. 5 in anaero-bic environment sparging Nz gas (100 mllmin) .2.2 Analytical methods

Glucose, maltose, and ethanol were deter-

. 83 .

}

JJJJ."'..,,

Life Science]aurnal ,3 (2),2006, Neves, et al, Wheat Milling By-products Fermentation

mined using HPLC (Shiiba et al. 1993). The re-ducing sugars were calculated based on the sugardistribution obtained by HPLC. They were com-posed of monosaccharides (glucose, arabinose andxylose) or disaccharides (maltose and cellobiose).The liquefaction yield ( YL) was calculated as thequotient of the amount of maltose released duringliquefaction by the initial substrate.

The fermentation performance was evaluatedbased on ethanol concentration (P), productivity

(Q), overall volumetric productivity (Qv), yield( YPIs), and solid residue (R p ). Qv was calculat-ed as the quotient of the total ethanol production bythe fermentation time. YPiS was defined as thequotient of the ethanol produced by the initial sub-strate. An aliquot of the final product was cen-trifuged (4, 000 rpm, 20 min), and the solidresidue dried in order to evaluate the residue forma-tion (Rp).

Table1. Fennentation perfonnance of various wheat milling products

WE

oj

1

WF

/

jJJ

I~j

!jJ)I~

Substrate concentration (g/L)a

Protein (%) (wlw)b

Starch (%) (wlw)C

Moisture (%) (wlw)

Ash (%) (wlw)

Final product

Ethanol concentration (P) (gll) d

Supernatant (m!)

Solid residue (Rp) (g dry matter)

Ash (%) (wlw)

Ethanol yield (Yp/s) (g ethanollg flour)

Overall volumetric ethanol productivity (Qv) (g/L' h)

Growth rate (fL)' (llh)e

a 1 liter slurries were prepared for each substrate, separatelyb Protein content (total nitrogen X 5.7); provided by Nisshin Flour Milling Co. Ltd.

C Composition assays perfonned as described elsewhere (Neves et al. 2006). Mean:t S. D. of three experiments

d Before centrifugation (5,000 rpm for 30 min)e Growth rate was calculatedby linear regressionof microbialgrowth curves (Slope=fL/2.303)

Substrate

Raw material

3 Results and Discussion

~

3. 1 LiquefactionThe LG liquefaction yield was O.065 ::tO. 002

g maltoselg flour (Mean ::t S. D., n = 3 repeti-tions), followed by WB (0.010 ::t O. 005 gig),compared to the reference substrate WF (0. 126 ::t0.021 gig).

The YL from LG was nearly six fold that fromWE, evidencing the lower initial starch content inWE. Besides releasing glucose (the major productof WE hydrolysis using cellulase), arabinose andxylose may also be produced, as mentioned in pre-vious literature (Adrados et al. 2004 j Shiiba et al.1993). The sugar content during WE liquefactionfluctuated considerably between replicates. Thisbehavior was likely caused by WE pentosans. Theyare known to have low water solubility, thus de-

creasing their availability for enzymatic hydrolysis(Adrados et al. 2004). For instance, while amy-lopectin is soluble in water, the starch granules it-self as well as amylose are insoluble in cold water.

In the case of LG, maltoseconcentrationwasnearly stable after 2 h liquefaction, indicating theend of starch hydrolysis. This is in line with the re-sults reported by Montesinos et al. (2000). Theydemonstrated that liquefaction during 2 h was nec-essary for efficient hydrolysis of glucose polymerswhen slurries containing 300 g raw wheat flour /Lwere hydrolyzed at 95 "C using thermostable Q-amylase.3.2 Simultaneous saccharification and fermenta-tion

The results of SSF using different wheat prod-ucts are depicted in Figure 1. The increase in glu-cose at the SSF onset implies that glucoamylase ac-

LG

20015.0

15. 6 :t O. 09

14.0:t0.15

2.7:t0.Ol

51. 40:t O.37

600

28. 50 :t 2. 69

3.90:t0.04

0.17:t0.01

2.72:t0.04

0.119

. 84 .

200

13.3

11.7:t0.25

12.2:t0.19

5.6:t0.04

18. 10 :t 3. 17

320

111.20:t8.31

18.70:t0.21

O. 02 :t O. 08

1.09:t0.21

0.142

200

10.4

62. 0 :t O.04

13.l:t0.01

0.6 :t O.15

.~

JI

JJI68.1O:t 1. 43

760

8. 30 :t O. 93

7.70 :t 0.42

0.,30 :t O.03

3 . 64 :t O.08

0.043

IJJ

j

jI~

JI.IJ

~J

J

JJ

l

~I~.I

jI1J

..r~

lI.

I

Life Science Journal, 3 (2),2006, Neves, et al, Wheat Milling By-products Fermentation

tivity was sufficient to release more glucose than Z.mobilis could metabolize. However, in the case ofLG, after 2 h SSF the glucose level decreased con-tinuously, disappearing at c. a. 9 h. Meanwhile,ethanol increased gradually, stabilizing at nearly 18h.

\

~t

..~r 80

~

}

!!.i:I 60 !

.r~ ii A:I - 1:i! ~ 40 .~- ofi~~8 20.at.:I

"\t..

~if

o~.

80

.~.~

or:;~

VJ -..60bD~.51 '-''":I-

i! ~40~ofiai~....'"

.at.:Il

0

80

~

~-

~ 60VJ~

bD~.51 '-'

i 140~ofi.r ~~ 20.at.:I

~I\\ 0

0

(I~II

( a) LG

. . I

,t . . 4

(b) WB

!

+ !,

(c)WF.

"if .6 12 18

Fennentation time (h)24

Figure 1. Time-course profiles of simultaneous saccharificationand fermentation from various wheat products.(a) Low grade flour; (b) Wheat bran; (c) Wheat flour.Symbols: "', Glucose; ., Reducing sugars; ., Ethanol.The bars represent the Standard Deviation (n =3 replicates)

~

~- From Figure 1c it was apparent that besidesglucose, other reducing sugars were released duringWB hydrolysis, mostly arabinose or xylose, ratherthan maltose which was rarely detected. These re-sults are in agreement with Shiiba et al. (1993).Those authors reported that hemicellulose (which ismainly consisted of arabinose and xylose) is the ma-jor component of WB cell wall polysaccharides. Xy-

~\

tI~

l

lose fermentation involves a metabolic pathwaywith two enzymes, xylose reductase and xylitol de-hydrogenase. Z. mobilis is not able to producethese two enzymes, thus the impossibility to fer-ment pentoses. Some other microorganisms, suchas P. stipitis, produce these enzymes naturally(Kodaki et al. 2004). Slight amounts of cel-lobiose, a reducing sugar obtained by partial cellu-lose hydrolysis, were detected in WB fermentationmash; this is a potential carbon source, since it canbe converted to glucose by a-glucosidase, which oc-curs naturally in wheat products.

Experimental microbial growth rate (p.) valuesindicated that Z. mobilis grew faster in WB- orLG-based substrates, rather than in WF. Low p.values obtained for WF (Table 1) suggested that inthis case the sugars contained in the fermentationmash were preferably metabolized into ethanol,rather than used for biomass production, substanti-ated by high P and low Rp values.3.3 Fermentation performance

The fermentation performance was evaluatedusing various parameters (P, Q, Qv, Yp/s, andRp), summarized in Table 1.

Figure 2 shows the ethanol productivity (Q)profiles throughout the SSF process. Q values forWB were lower compared to LG or WF, which waslikely due to the presence of pentose sugars in WE.Incomplete starch hydrolysis during WF liquefac-tion may have caused delay on ethanol production,as well. This was consistent with the hypothesisthat starch hydrolysis is made difficult by the in-creased viscosity during liquefaction, caused bystarch granules swelling as well as water penetra-tion (Montesinos & Navarro 2000).

0 6 12 18Fennentation time (h)

24

Figure 2. Ethanol productivity (Q) from various wheat products(., lowgrade flour; ..., wheatbran; ., wheatflour)Q was calculated by differentiating the experimental ethanol pro-duction data as function of time (Jain et al. 1985).

...

To gain some perspective on the improvementsachieved using different conversion systems and mi-croorganisms, a comparison was provided in Table

. 85 .

6 . .25.4 . ..t 3-c

.& *.@ 2

AJ:: 1 A_A0

~

J

~~I

jI

Life Science Journal, 3 (2),2006, Neves, et al, Wheat Milling By-products Fermentation

2. A combination of high ethanol levels and pro-ductivity is desirable in minimizing processingcosts. It was apparent that ethanol concentrationand yield obtained for LG were relatively low,compared to other systems using WF. Such prob-lems can be overcome, e. g. using fed-batch culturesystems, as reported in previous literature (Robleet al. 2003). Those authors were able to reach up

to 90 g ethanollL using a circulating loop bioreactorfor SSF of raw cassava starch.

In the year 2000 nearly 6. 8 million tons ofWF were produced in Brazil (FIBGE, 2001) re-sulting in about O.34 million tons of LG. Assumingthat this by-product could be fully used as feed-stock, the potential bioethanol production from LGbecomes 78.2 mega liters.

"'t~,i

1

iJJ

i;.JI

1J~of

jIJI

J

jJ~

1

)iJ

4 Conclusion

~

In this work, two wheat milling by-products,LG and WB, were utilized as substrate forbioethanol production. High LG fermentationrates, reaching the highest ethanol productivity(4.4 gIL'h) within 6 h of SSF, indicated consid-erable savings on fermentation time, compared tocurrent industrial processes.

The present system employing commercial en-zymes might be too expensive for commercialbioethanol production, in view of the current mar-ket prices of most amylolytic enzymes. This processwas used to evaluate the fermentation performanceof different wheat milling by-products. Industrialapplication of this system requires modifications inorder to decrease the overall cost, e. g. in situ a-amylase production as well as immobilizing theethanol-producing strain, utilizing fed-batch orcontinuous fermentation process.

In order to use wheat milling by-products forlarge scale fermentation, a system for collectingthese by-products from the milling site has to bedeveloped, which could be done after LG is recog-nized as a low cost feedstock.

AcknowledgmentsWe are thankful to Nisshin Flour Milling Co.

Ltd. for providing LG, WB and WF samples.

IJ

~

Correspondenceto:Marcos Antonio das NevesGraduate School of Life and Environmental Sci-encesUniversity of Tsukuba1-1-1 Tennodai, Tsukuba, JapanTelephone: 81-29-853-7222Fax: 81-29-855-2203Email: maneves2000@yahoo.com

Jj

fJ~I~

jJ

J

LReferences

1. Adrados BP, Choteborska P, Galbe M, Zacchi G.Ethanol production from non-starch carbohydrate ofwheat bran. BioresTechnol 2005; 96: 843 - 50.

2.Adrados BP, Juhasz T, Galbe M, Zacchi G. Hydrolysisof nonstarch carbohydrates of wheat starch effluent forethanol production. BiotechnolProg 2004; 20: 474 - 9.

3. FIBGE Fundaciio Instituto Brasileiro de Geografia eEstatistica. Anuario Estatistico do Brasil 2000 - 2001(in Portuguese) .

4. Jain WK, Diaz IT, Baratti J. Preparation and characteri-zation of immobilized cells of Zymomonas mobilis for

JI

JI

i

]

. 86 .

Table2. Summary of ethanol production using varioussubstrates and microorganisms

Reactor Mode CultureSubstrate Ethanol Productivity Yield

(volume) (g/L) . (g/L) (g/L' h) (gig flour) Source

Jar Batch Commerciala-amylase, Raw wheat flour 67 3.19 0.45" b

fermentor(40 h) glucoamylase (300)

(2 L) and S. cerevisiae

Erlenmeyer Batch a-amylase (from Fine-wheat flour 44 0.49 0.18 c

(500 mL) (90 h) B. subtilis) and Damagedwheat 34 0.38 0.14

S. cerevisiae Damaged sorghum 27 0.30 0.11

(250)

Circulating Fed-batch a-amylase (from Raw cassavastarch 90 1.17 0.45" d

loop (600 h)A. awamori) and (150)

reactor (9 L) S. cerevisiae

Jar Batch Commercial LG (200) 51 2.72 0.17 This

fermentor (48 h) a-amylase, WE (200) 18 1.09 0.02 work

(2 L) glucoamylase WF (200) 68 3.64 0.30

and Z. mobilis

a Ethanolyield(g ethanollgstarch);b(Montesinoset al. 2000);C(Sureshet al. 1999);d(Robleet al. 2003)

..I

..,.'.\

Life ScienceJournal, 3 (2) ,2006, Neves, et al , Wheat Milling By-products Ferrrumtation

~~ ethanol production Biotechnol. Bioeng 1985; 27: 273-

9,5. Kodaki T, Watanabe S, Makino K. Baiomasu no E-

tanoru Hankan. Purotein Kougakuteki Apurochi. EcoIndustry 2004; 9: 38 - 44 (in Japanese) .

6.Montesinos T, Navarro JM. Production of alcohol fromraw wheat flour by amyloglucosidaseand saccharomycescerevisiae. Enzyme Microb Technol 2000; 27: 362-70.

7. Neves MA, Kimura T, Shimizu N, Shiiba K. Produc-tion of alcohol by simultaneous saccharificationand fer-mentation of low-grade wheat flour. BrazilianArch BioiTechno12006; In press.

.'~

~

..I

~

..

\

\I

t

~

~..I

~

k

I~

lI

8. Roble ND, Ogbonna JC, Tanaka HA. novel circulatingl=p bioreactor with cells immobilized in l=fa sponge forthe bio-conversion of raw cassava starch to ethanol. ApplMicrobiolBiotechnol2003; 60: 671 - 8.

9. Shiiba K, Yamada H, Hara H, Okada K, Nagao S. Pu-rification and characterization of two arabinoxylans fromwheat bran. Cereal Chern 1993; 70: 209 -14.

10. Suresh K, Kiransree N, Rao LV. Production of ethanolby raw starch hydrolysis and fermentation of damagedgrains of wheat and sorghum. Bioprocess Eng 1999; 21:165- 8.

Received March 10, 2006

. 87 .