b i om a s s an d b i o e n e r g y 3 5 ( 2 0 1 1 ) 3 0 9 4e3 1 0 3

Avai lab le at www.sc iencedi rect .com

ht tp : / /www.e lsev ier . com/ loca te /b iombioe

Optimisation of dilute alkaline pretreatment for enzymaticsaccharification of wheat straw

S. McIntosh a,1, T. Vancov a,b,*,1

a Industry and Investment NSW, Wollongbar Primary Industries Institute, NSW, Australiab Primary Industries Innovation Centre, University of New England Armidale, NSW, Australia

a r t i c l e i n f o

Article history:

Received 26 May 2010

Received in revised form

6 April 2011

Accepted 15 April 2011

Available online 13 May 2011

Keywords:

Alkali pretreatment

Biofuels

Enzyme saccharification

Lignocellulose

Wheat straw

* Corresponding author. Industry and Investmfax: þ61 2 6628 3264.

E-mail address: tony.vancov@industry.ns1 Both authors contributed equally to the m

0961-9534/$ e see front matter ª 2011 Elsevdoi:10.1016/j.biombioe.2011.04.018

a b s t r a c t

Physico-chemical pretreatment of lignocellulosic biomass is critical in removing substrate-

specific barriers to cellulolytic enzyme attack. Alkaline pretreatment successfully deligni-

fies biomass by disrupting the ester bonds cross-linking lignin and xylan, resulting in

cellulose and hemicellulose enriched fractions. Here we report the use of dilute alkaline

(NaOH) pretreatment followed by enzyme saccharifications of wheat straw to produce

fermentable sugars. Specifically, we have assessed the impacts of varying pretreatment

parameters (temperature, time and alkalinity) on enzymatic digestion of residual solid

materials. Following pretreatment, recoverable solids and lignin contents were found to be

inversely proportional to the severity of the pretreatment process. Elevating temperature

and alkaline strengths maximised hemicellulose and lignin solubilisation and enhanced

enzymatic saccharifications. Pretreating wheat straw with 2% NaOH for 30 min at 121 �C

improved enzyme saccharification 6.3-fold when compared to control samples. Similarly,

a 4.9-fold increase in total sugar yields from samples treated with 2% NaOH at 60 �C for

90min, confirmed the importance of alkali inclusion. A combination of three commercial

enzyme preparations (cellulase, b-glucosidase and xylanase) was found to maximise

monomeric sugar release, particularly for substrates with higher xylan contents. In

essence, the combined enzyme activities increased total sugar release 1.65-fold and

effectively reduced cellulase enzyme loadings 3-fold. Prehydrolysate liquors contained

4-fold more total phenolics compared to enzyme saccharification mixtures. Harsher

pretreatment conditions provide saccharified hydrolysates with reduced phenolic content

and greater fermentation potential.

ª 2011 Elsevier Ltd. All rights reserved.

1. Introduction 50% by 2025 [1]. Allowing for current rates of production and

The search for sustainable transportation fuels is instigated by

diminishing crude oil supplies, impact on climate change and

green house gas (GHG) emissions. Global petroleum demands

have steadily increased from 57 � 106 barrels/day in 1973 to

82 � 106 barrels/day in 2004 and is anticipated to rise another

ent NSW, 1243 Bruxner

w.gov.au (T. Vancov).anuscript.

ier Ltd. All rights reserve

existing reserves, we will soon approach Hubbert’s predicted

‘peak oil’ levels [2]. In selecting an alternative, consideration

must be given to fuels which serve to combat climate change

and produce cleaner air. Since the industrial revolution

atmospheric CO2 levels have increased fromw275ew380 ppm

owing to the burning of fossil fuels. Consequently,

Highway, Wollongbar, 2477 NSW, Australia. Tel.: þ61 2 6626 1359;

d.

b i om a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 3 0 9 4e3 1 0 3 3095

atmospheric temperatures have risen by 0.6 � 0.2 �C during

the twentieth century. If left unchecked, CO2 levels could

easily surpass 550 ppm by the middle of this century [3].

Biofuels, fuels derived fromplant biomass are currently the

only sustainable class of liquid fuels [4]. First-generation bio-

fuels such as ethanol are currently produced from plants rich

in storage carbohydrates (i.e. sugar and starch). However, as

the demands for the feedstock intensify so will the debate

between ‘food’ versus ‘fuel’. Moreover, 1st generation bio-

ethanol does not significantly diminish green house gas (GHG)

emissions [5]. These shortcomings can be addressed by

producing ethanol from lignocellulosic material (2nd genera-

tion biofuels), such as agricultural and forest waste residues.

Second-generation biofuels are derived from the inedible and/

or unexploited part of the plant (lignocellulose) and can be

sourced from plant residues or organic waste such as crop

straw, forestry thinnings or municipal waste.

Potential agricultural biomass residues from major grain

crops in Australia can be exploited for bioethanol production

and is presently estimatedat 65Mt [6,7].Makingallowances for

current uses like ground cover, soil carbon and fodder, leaves

approximately 45% available for production of 2nd generation

biofuels [6]. The largest individual contributor to this biomass

derives from wheat cultivation. Presently, the total planted

area in Australia stands at approximately 13 million hectares

with an estimated grain yield of 22 Mt [8]. Assuming the

harvest index statement that for every 4 ton of grain at least

6 ton of residue is produced holds true, the total straw yield

from wheat cultivation in Australia would approach 33 Mt.

Utilization of lignocellulosics requires effective pretreat-

ment to free cellulose from its lignin seal and open up the

crystalline structure so that enzymatic hydrolysis can take

place [9]. Although a range of chemical, physical and biolog-

ical processes have been configured to release constituent

sugars from lignocellulose, they suffer several shortcomings

such as cost, infrastructure needs and technological impasses

[10]. Alkali treatment processes are generally very effective in

the pretreatment of agricultural residues such as wheat straw

and herbaceous crops [11,12].

Alkali-based pretreatment efficiently delignifies lignocel-

lulose by disrupting the ester bonds cross-linking lignin and

xylan, leading to fractions enriched in both cellulose and

hemicellulose [9]. Its mechanism of action closely resembles

soda or kraft pulping. Alkali pretreatment processes generally

utilize lower temperatures, pressures and residence times

compared to other pretreatment technologies, the degree

dependant directly on the nature of the biomass feedstock,

mainly its lignin content [13]. Sodium hydroxide, ammonia,

peroxide and lime have received a great deal of attention as

pretreatment agents, owing in part to cost-effective practises

such as chemical and water recycling (processes well devel-

oped in thepulping industry) andpartly because lower enzyme

loads are generally required to convert cellulose to glucose [14].

Several studies have assessed with varying degrees of

success, the utility of pretreating lignocellulosic materials

with dilute alkali followed by enzyme hydrolysis with cellu-

lases [15e18]. For the most part, these studies examined and

reported conditions for hydrolysing the lignocellulosic

biomass without detailing the specific affects of individual

treatment parameters on saccharification yields and changes

in monosaccharide compositions. Moreover, these studies

failed to highlight the fact that they employed extremely high

enzyme dosages, extended reaction and fermentation times

above what would be considered the cost-benefit threshold.

Overcoming the toxic effects of inhibitors generated from

sugar and lignin decomposition (furans, phenols, carboxylic

acids) and salts was the major reason for these economically

unsustainable practises.

In this study, we report sugar yields and profiles from post-

grain harvested wheat straw following dilute alkaline pro-

cessing and low enzyme dose saccharification. Specifically,

we examine and describe three characteristic phases: 1)

ascertaining the function of key pretreatment parameters

(alkaline strength, temperature and residence time) and their

impact on sugar solubilisation, lignin reduction and solid

losses; 2) enzymatic hydrolysis efficacy of pretreated solid

residues and variations in sugar composition with respect to

pretreatment parameters; and 3) the role of individual and

combined enzyme activities and their impact on the rates and

yields of sugar release. Due to the impact of phenolic

compounds on downstream processes we also discuss their

release during pretreatment and saccharification. Under-

standing these key elements will enable further process

optimisation of wheat residues and assist in determining the

efficacy of the conversion strategy.

2. Materials and methods

2.1. Materials

Post-grain harvested wheat straw (Triticum aestivum) was

sourced from the Liverpool plains Northern NSW, Australia.

The wheat straw was dried at 55e60 �C for 48 h, ground in

a rotary mill (Thomas Wiley Laboratory Mill) and passed

through a 1.5 mm screen. All chemicals used were of reagent

or analytical grade and purchased from Sigma Chemical Co.

(St. Louis, MO).

2.2. Pretreatment

The effect of pretreatment parameters including: alkalinity,

residence time and temperature were evaluated using

a 4 � 3 � 2 factorial design. The procedure described by

McIntosh and Vancov [19] was employed in this study. Briefly,

dilute sodium hydroxide (NaOH) at 0.75% 1.0% and 2.0% (w/v)

was used to pretreat milled wheat straw samples at a solid

loading of 10% (w/v). Treatments were performed at 60 �C in

a water bath and in an autoclave at 121 �C (15 psi) with resi-

dence times of 30, 60 and 90 min. Insoluble solids were sepa-

rated from liquor fractions following pretreatment using

a Buchner funnel fitted with glass fibre filters. Solid material

was washed with water until the pH was 7, sealed in plastic

bags and stored at �20 �C.

2.3. Enzyme assays

Cellulase (NS50013), b-glucosidase (NS50010) and xylanase

(NS50030) preparations were kindly supplied by Novozymes

(BagsværdDenmark). Enzymeactivities,asdescribedbysupplier,

b i om a s s an d b i o e n e r g y 3 5 ( 2 0 1 1 ) 3 0 9 4e3 1 0 33096

are 70 filter paper unit (FPU)/g, 250 cellobiase units (CBU)/g and

500 Farvet Xylan Unit (FXU)/g respectively. FXU is measured

relative to a Novozymes fungal enzyme standard. The activity of

the reference standard, Humicola insolens, xylanase batch no. 17-

1194, isdefined tohaveanenzymaticactivityof 3 550FXU/gatpH

6.0 and 50 �C in 30 min reaction time of colour release from

remazol-xylan substrate (from Novozymes Biomass Kit for

conversionof lignocellulosicmaterials). Total cellulaseactivityof

NS50013was confirmed using the filter paper assay as described

by the National Renewable Energy Laboratory (NREL) laboratory

procedure LAP006 [20]. Protein content of liquid enzyme prepa-

rations was determined using a commercial bicinchoninic acid

(BCA) protein assay reagent kit (Pierce Products, USA). Endoglu-

canase, exoglucanase, xylanase and pectinase activities were

individually determined in reactionmixtures (10mL) containing

1%(w/v)carboxymethylcellulose,0.5%(w/v)avicel, 0.5%(w/v)oat

spelt xylan and 0.5% (w/v) citrus pectin, respectively, in 50 mM

citrate buffer (pH 5.2), and appropriately diluted enzyme solu-

tions as described by McIntosh and Vancov [19]. After 30 min

incubation at 50 �C, the reducing sugar liberated in the reaction

mixturewasmeasured by the dinitrosalicylic acid (DNS)method

[21].Oneunit (U)ofeachenzymeactivity isdefinedas theamount

of enzyme that produces 1 mmol of reducing sugar as glucose,

xylose (for xylanase) or galacturonic acid (for pectinase) in the

reaction mixture per minute, per mg protein under the above

specified conditions (Table 1).

b-Glucosidase activities were assayed in reaction mixtures

(1 mL) containing 4 mM p-nitrophenyl b-D-glucoside, 50 mM

acetate buffer (pH 5.0) and appropriately diluted enzyme solu-

tions as described by McIntosh and Vancov [19]. Following

incubationat 50 �Cfor 30min, reactionswere stoppedbyadding

100 uL of ice-cold 100 mM NaOH, and the resulting colour

intensity measured at 405 nm. A unit (U) of enzyme activity is

defined as the amount of enzyme that releases 1 mmol of

p-nitrophenol per minute, per mg of protein in the reaction

mixture under these assay conditions. b-Glucosidase activities

present in commercial preparations are reported in Table 1.

2.4. Enzymatic saccharification

Enzymatic saccharifications were performed according to the

method described by McIntosh and Vancov [19]. Essentially,

Table 1e Specific activity of commercial enzymes used inNaOH pretreated wheat straw saccharification.

Enzymes Specific activity (U/mg protein)a

NS50013 NS50010 NS50030

Endoglucanase 14.20 0.11 0.02

Exoglucanase 1.51 0.07 0.05

Xylanase 7.05 75.00 129.50

b-glucosidase 1.07 10.08 0.04

Pectinase 0.03 0.4 ND

Cellulaseb 70.00 ND ND

Protein (mg/ml)c 135 150 33

ND Not determined.

a At pH 5.0 and 50 �C.b Measured as filter paper units/g protein.

c Concentration of Novozymes preparations.

solid residues at a 5% (w/v) loading were resuspended in

solutions of 50 mM citrate buffer (pH 5.2) and appropriately

diluted enzymes (NS50013, NS50010, andNS50030) as specified

in Section 3.4. Hydrolysis was performed in a shaking water

bath at 50 �C and 150 rpm for up to 72 h and in the presence of

10 mM sodium azide to prevent microbial contaminant

growth. Samples were placed in a centrifuge tube and imme-

diately chilled on ice, centrifuged at 8000g for 5 min, filtered

and stored at �20 �C awaiting sugar analysis. In this paper,

units of enzyme activity are those given by the manufacturer.

2.5. Analytical methods

Neutral detergent fibre (NDF), acid detergent fibre (ADF), acid

detergent lignin (ADL) and acid insoluble ash (AIA) were

determined in untreated wheat straw by Industry and

Investment NSW’s Diagnostic and Analytical Services (Wagga

Wagga, Australia) using ANKOMTechnologyMethods [22e24].

The difference between NDF and ADF provides an estimate of

detergent soluble hemicellulose. Detergent cellulose is calcu-

lated by subtracting the values for ADL plus AIA from ADF.

Carbohydrate content of untreated material was also deter-

mined bymeasuring the hemicellulose (xylan and araban) and

cellulose (glucan) derived sugars in supernatants following

concentrated acid hydrolysis as described by NREL [25]. Acid

insoluble lignin content of untreated and pretreated wheat

straw was determined according to the NREL methods [25].

Water and ethanol soluble sugars were also extracted from

untreated wheat straw and quantified according to NREL

methods [26] (see Table 2).

Sugar composition of prehydrolysate and enzymatic

saccharification liquors were determined using high perfor-

mance liquid chromatography (HPLC) according to the

procedure described by McIntosh and Vancov [19]. The sepa-

ration system consisted of a solvent delivery system

(Controller 600 Waters, Milford, MA) equipped with an auto-

sampler (717, Waters), a refractive index detector (410 differ-

ential refractometer, Waters) and a computer software based

integration system (Empower, Waters).

Sugars were analysed using either a Sugar-Pak 1

(6.5 � 300 mm, Waters) or an IC-Pak Ion-Exclusion 50Ao 7 mm

(7.8 � 300 mm, Waters), both fitted with the IC-Pak Ion Exclu-

sion guard-Pak (Waters). The Sugar-Pak 1 column was main-

tained at 70 �C, and sugars were eluted with degassed Milli-Q

filtered water containing 50 mg/l Ca-EDTA at a flow rate of

Table 2 e Composition of untreated wheat straw.

Component Percentage (%)a

Neutral detergent fibre 69.0

Acid detergent fibre 43.0

Acid detergent lignin 5.9

Acid insoluble ash 0.9

Cellulose 36.0

Hemicellulose 26.0

Acid insoluble lignin 7.6

Water extractives 13.0

Ethanol extractives 5.5

a Composition percentages are on dry-weight basis.

b i om a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 3 0 9 4e3 1 0 3 3097

0.5mL/min. The IC-Pak Ion-Exclusion columnwasmaintained

at 60 �C, and sugars, acetic acid and ethanol were eluted with

degassed Milli-Q filtered water containing 2 mM H2SO4 at

a flow rate of 0.8 mL/min. The refractive index detector was

maintained at 50 �C for all applications. Compounds were

identified by comparison with retention times of standards

(glucose, xylose, galactose, arabinose, mannose, fructose,

sucrose and cellobiose). Quantification was performed using

a calibrationwith external standards of known concentration.

2.6. Xylan extraction

The method was performed according to the procedure

described by McIntosh and Vancov [19]. Wheat straw was

pretreated as described in section 2.2. Liquid fractions

following pretreatment were separated from solids by filtra-

tion using a Buchner funnel and then centrifuged (10,000g for

10 min) to precipitate particulates. Decanted liquor was

adjusted to � pH 4.0 with 6M HCL. After 10 min of continuous

stirring, particulates were precipitated by centrifugation. To

the supernatant, approximately three volumes of cold ethanol

were added whilst stirring for 15 min and the precipitate

(xylan) collected by centrifugation. The xylan was then dried

and weighed.

2.7. Acid-insoluble lignin extraction

The method is as described by McIntosh and Vancov [19].

Particulates were removed from liquors by centrifugation

(10min; 10 000 g), and the supernatants adjusted to pH 2.0with

concentratedH2SO4andheated (60 �C).After 5minof continual

stirring, samples were cooled to ambient temperature and

particulates precipitated by centrifugation. Acid-insoluble

lignin precipitates were washed by mixing with acidified

water (pH 2.0) by gently inversion, collected by centrifugation

(10 min; 10 000 g), dried and their weight recorded.

2.8. Total phenolic determination

The enzymatic method described by McIntosh and Vancov

[19] was used to determine total phenolic content of hydro-

lysates. Samples were centrifuged and filtered (0.45 micron)

prior to analysis. A 25 ml aliquot of an appropriately diluted

phenolic sample was mixed with 225 ml of enzyme-reagent

working solution into 96 well microtitre plate (clear

F-bottom). The enzyme-reagent working solution was

a freshly prepared solution containing 0.1 M potassium

phosphate buffer solution (pH 8.0), 30 mM 4-aminoantipyrine

(4-AP), 20 mM hydrogen peroxide (H2O2) and 6.6 mMHRP. After

15 min at room temperature, the absorbance was read at

540 nm, using a Flurostar (BMG labtechnologies, GmbH) plate

reader. Vanillic acid standards (0e500 ng/mL) were subjected

to the same assay conditions as the samples. Total phenolics

were reported as vanillic acid equivalents.

2.9. Statistical methods

Each set of observations was modelled as a response to the

classifying factors generated by the experimental design. The

data was analysed using analysis of variance which allowed

separation of the variation in the data into components of

temperature, residence time, alkaline strength and interac-

tions between those terms.

The modelling process enabled prediction of the expected

(average) response at each combination of the experimental

factors and a measure of the experimental error. Estimated

experimental error was used to calculate the “Least significant

difference” (l.s.d., p¼ 0.05) between three averages required to

indicate a statistically significant effect. Statistical analysis

and graphical presentation were conducted using software

provided by the R Development Core Team [27].

3. Results and discussion

3.1. Effects of alkaline pretreatment on wheat straw

The composition of post-grain harvested wheat straw used in

this study is presented in Table 2. The holocellulose fraction

totalled 62% of the dry biomass with cellulose being the major

component at 36% and the remaining 26% derived from

hemicellulose. Both acid detergent and acid insoluble lignin

levels were 5.9% and 7.6%, respectively. Water extractive

compounds accounted for approximately 130 mg/g dry straw,

of which 35 mg � 2 mg was identified as a storage trisaccha-

ride (glucoseefructoseefructose) of the steam and leaves [28].

Further solvent extraction with ethanol resulted in 55 mg of

material, presumably composed of oils, pigments and waxes.

The profile and quantities of individual wheat straw compo-

nents is comparable to reported values in the literature [16].

We initially observed that pretreatment of wheat straw

with dilute NaOH resulted in a dark coloured slurry and that

the colour intensity generally increased with pretreatment

severity. Separation of the slurry revealed the insoluble frac-

tion was both reduced in mass and bleached in colour. Others

have reported similar reductions in solids during alkali

pretreatment and attribute the degree of solubilisation with

the severity of temperature, alkali concentrations and resi-

dence time. We found solid losses ranging from 25% under

mild conditions (1% NaOH; 60min; 60 �C) to 57% when pre-

treated in 2% NaOH at 121 �C. Although each pretreatment

variable in this study contributed to the loss of solids, we

found that temperature caused the greatest loss followed by

alkalinity and then residence time. Comparable solid loss and

treatment parameter trends have been reported in related

studies on wheat straws [11]. However, a survey of the liter-

ature reveals a number of studies reporting disparity in

susceptibility to alkali pretreatment between different agri-

cultural crop residues [11,17,29].

These solid losses are equivalent to the amount of solubi-

lised hemicellulose fraction and other components into pre-

hydrolysate liquors. Aside from lignins (discussed later),

several studies have reported the hydrolysis of hemicellulose

and release of oligoxylans (polyoses) of mixed molecular

weights following exposure to alkali-based chemicals during

the pretreatment process [30e32]. Once considered a disad-

vantage of alkaline treatment processes (i.e. reduction in total

fermentable sugar yields) current biorefinery platforms are

exploiting this phenomenon for the recoveryofhighmolecular

weight oligoxylans/arabinoxylans which are desirable high-

Fig. 1 e Total sugar release from alkali pretreated (10%, w/

v) and enzyme saccharified wheat straw. Sugar yields are

expressed as mg/g pretreated material. Displayed data

represents averages of three independent experiments.

The average l.s.d. ( p [ 0.05) [ 25.0.

b i om a s s an d b i o e n e r g y 3 5 ( 2 0 1 1 ) 3 0 9 4e3 1 0 33098

value co-products [15,33,34]. In order to measure hemi-

cellulose degradation, we initially attempted to quantify the

amount of liberated pentose sugars (xylose and arabinose).

However, HPLCanalysis of the prehydrolysate liquors revealed

a complex profile of monosaccharides and mixed higher

molecular weight oligosaccharides. Analysis of these oligox-

ylans showed that the yieldswere proportional to the strength

of the alkaline solution at 121 �C. Pretreating wheat straw at

121 �C for 30min in 0.75% and 2.0%NaOH solutions resulted in

20% and 33% solubilisation, respectively, of the hemicellulose

fraction. This was correspondingly isolated as a crude xylan

precipitant from prehydrolysate liquors as given by the mass

balance outlined in section 3.7. In addition to alkaline strength,

variation in temperature and timemay also have an impact on

the yield of isolated arabinoxylans [15].

3.2. Total sugars in enzyme saccharified hydrolysates

The hydrolysis of both cellulose and hemicellulose in pre-

treated materials via enzymatic action is critical in releasing

monomeric sugars for fermentation to bioethanol. The rate

and extent of enzymatic saccharification of the poly-

saccharides provides a measure of the pretreatment’s effec-

tiveness. The relationship between pretreatment severity and

enzyme saccharification of pretreated solid material was

examined. In an attempt to identify and describe the key

variables, twenty four pretreatment combinations derived by

altering the alkaline concentrations (4 levels), time (3 levels)

and temperature (2 levels), were used to assess wheat straw.

To avert large rapid sugar releases which could potentially

mask any significant pretreatment variable(s), saccharifica-

tions were performed with relatively low enzyme doses

(2.5 FPU cellulase, 2.5 CBU b-glucosidase and 1.5 FXU xylanase

per gram of pretreated solids). That is, the intension of the

enzyme hydrolysis process was not designed to maximise

sugar release but to augment and identify key parameters in

successful pretreatment processes. Sugar yields were quan-

tified and total sugar release was modelled as a response to

pretreatment parameters and expressed as a function of

alkaline strength, temperature and residence time (Fig. 1).

Increases in pretreatment temperature, residence time and

alkali concentration all improve the enzymatic saccharifica-

tion efficiency of recovered solids. Temperature had the

greatest significant ( p < 0.05) effect on enzyme saccharifica-

tion followed by alkaline strength and residence time. Solids

pretreated at 121 �C were more acquiescent to enzymatic

hydrolysis than at 60 �C. Within the 121 �C experiments an

increase in the strength of the alkaline solution from 0 to 2%

resulted in a 5.2 fold increase in total sugar release. Pretreating

wheat straw with 2% NaOH for 30 min at 121 �C, followed by

enzyme saccharification, yielded the highest recorded sugar

releaseof 850mg/gpretreatedmaterial. Extending the reaction

time to 90min failed to liberate any furthermonomeric sugars.

Total sugar release did, however, decline (by 67 mg/g or 7.8%)

when the reaction time was reduced to 60 min. Hu and Wen

[35] and Wang et al. [18] reported similar responses to

temperature and alkaline concentrations, albeit, they recov-

ered significantly less total sugars at elevated NaOH strengths.

At the lower pretreatment temperature of 60 �C sugar yields

were found to rise 5 fold with increasing NaOH concentration

(0e2%). Amaximumyield of 667mg/g pretreatedmaterial was

attained with 2% NaOH followed by saccharification. Under

these conditions statistically similar ( p < 0.05) yields were

obtained from solid materials exposed to an elevated

temperature and reduced hydroxide combination (121 �C/0.75% NaOH). This raises the possibility that under mild alka-

line conditions the optimal pretreatment temperature may be

lower than 121 �C, offering potential power and cost savings in

a commercial process. No discernable differences between the

30 and 60 min treatments were observed at 60 �C, however,

extending the time to 90min improved total sugar yields for all

combinations of alkalinity. In the absence of NaOH, increasing

time did not influence sugar yields but raising the temperature

to 121 �C gave a slight improvement in saccharification.

3.3. Pentose and hexose sugars in enzyme saccharifiedhydrolysates

The effectiveness of enzymatic saccharification on pretreated

material is principally evaluated by the degree of conversion

of cellulose to glucose monomers. For alkaline based

pretreatment processes, this also includes the release of

monomeric pentose (xylose and arabinose) sugars from

preserved hemicellulose. Alkaline pretreatment partially

solubilises the hemicellulose fraction leaving a material

enriched in cellulose [18,29,31]. Thus quantifying individual

sugar components in enzyme treated hydrolysates permits

appraisal of their fermentation potential and assists in

determining the best possible conversion strategy. Mono-

saccharides present in saccharified hydrolysates from pre-

treated solids were quantified and expressed as a function of

alkaline strength, temperature and residence time in Fig. 2.

b i om a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 3 0 9 4e3 1 0 3 3099

Generally, enzymatic hydrolysis of cellulose increasedwith

pretreatment temperature, residence time and alkali concen-

tration. Temperature had the greatest significant ( p < 0.05)

effect, with 121 �C delivering greater cellulose hydrolysis than

60 �C. A maximum glucose yield was recorded (567 mg/g pre-

treated solids) when pretreated at 121 �C for 90 min in 2%

NaOH, though reducing the reaction time to 30 min delivered

comparable yields (552 mg). Within the 121 �C treatments,

elevating alkaline strengths resulted in a significant ( p < 0.05)

increase in glucose recovery for all pretreatment times. Similar

trendswere noted in samples treated at the lower temperature

of 60 �C. Glucose release from wheat straw exposed to 2%

NaOH at 60 �C was found to equal or surpass glucose levels

resulting from pretreatment with 0.75% NaOH at 121 �C for all

pretreatment times. These results suggest that increasing

alkaline strength acts as a trade-off to reducing temperatures.

Increasing pretreatment temperature, residence time and

alkali concentration substantially improved hemicellulose

saccharification and subsequent xylose release from pre-

treated straw. Temperature had a significant ( p < 0.05) effect

with 121 �C producing greater xylose release than 60 �C.Maximum xylose yields were attained when wheat straw was

exposed to 1% NaOH at 121 �C giving a peak yield of 275 mg

after 60mins. As observed for glucose yields (above), a reduc-

tion in alkali strength (0.75%) resulted in significantly

( p> 0.05) lower xylose release suggesting that lower alkalinity

is ineffective in disrupting the lignocellulosic structure. When

wheat straw was pretreated at conditions optimal for glucose

recovery (i.e. 2% NaOH/121 �C/30 min), a significantly

( p < 0.05) lower xylose yield was observed. Others have

reported similar declines in xylose yield which ultimately

correlate with increasing xylan levels in pretreatment

hydrolysates as pretreatment settings are elevated [11,15,17].

In the control samples, xylose release was very small

(30 mge45 mg/g), irrespective of temperature setting.

Lowering the pretreatment temperature to 60 �C led to

a reduction in themaximumxylose yield (227mg/g pretreated

material). However, we noted that xylose levels from enzyme

saccharified wheat straw exposed to 1e2% NaOH at 60 �C for

Fig. 2 e Monosaccharide composition of sugars released

from pretreated (10%, w/v) and enzyme saccharified wheat

straw. Sugar yields are expressed as mg/g pretreated

material. Displayed data represents averages of three

independent experiments. The average l.s.d. ( p [ 0.05)

are: 13.2 (glucose), 11.1 (xylose), 2.7 (arabinose) and 25.0

(total yields).

90 min equalled xylose release from enzyme digested solids

pretreated with 0.75% NaOH at 121 �C. Inadequate hemi-

cellulose hydrolysis at this lower temperature has probably

physically constrained, and therefore impeded, cellulase

breakdown. Supplementing the enzyme mixture with addi-

tional hemicellulase/xylanase activity will improve sacchari-

fication of mildly treated substrates containing higher

amounts of xylan [36]. Pretreatment conditions for maximum

arabinose sugar release correlated with those observed

(above) for xylose sugars at both temperatures. Maximum

yields of z33 mg/g pretreated material were attained under

these conditions. Arabinose yields from solids pretreated in

2% NaOH at 121 �C were also significantly ( p < 0.05) reduced.

Glucose and xylose yields from controls (water treated mate-

rials) were approximately 4 and 6-fold, respectively, lower

than yields resulting from alkali catalysed pretreatment. In

summary, these results firstly confirm the need for an alkali

catalyst and secondly, that the optimal concentration will be

based on whether maximising for glucose recovery or

a combination including hemicellulose sugars.

3.4. Time course of enzyme hydrolysis

For any individual biomass and pretreatment strategy, it is

essential to tailor the saccharification process (enzyme

mixture and conditions) to maximise sugar yields [37]. Addi-

tionally, although commercial cellulase preparations are

abundant in b-endoglucanase and cellobiohydrolyase activity,

the level of b-glucosidase and xylanase activities are generally

inadequate for efficient monomeric sugar release from

substrates containing higher amounts of arabinoxylan [38]. As

shown in Table 1, the Novozymes cellulase preparation

(NS50013) has 10-fold and 18-fold less b-glucosidase and

xylanase activities, respectively, than NS50010 and NS50030

enzyme preparations, hence enzyme blending is necessary. In

Fig. 3aeb, the rate and extent of saccharification in response

to differing enzyme combinations and dosages from NaOH

(1.0% NaOH; 60 min; 121 �C) pretreated wheat straw is exam-

ined. This pretreatment regime was employed to evaluate

pretreated material that has been substantially delignified yet

retained most of its xylan fraction.

The combination of cellulase with b-glucosidase substan-

tially promoted sugar release and was greater than the indi-

vidual preparations. A supplementary experiment (unreported

data) revealed that increasing the ratio of NS50010 to NS50013

from 1:1 to 4:1 (a 4-fold increase in b-glucosidase activity) lead

to a corresponding rise in saccharification. However, beyond

the ratio of 1:1 the gains were neither statistically significant

nor cost-effective for cellulose conversion, and this ratio was

subsequently used in following enzyme trials. Tengborg et al.

[39] also described similar benefits and limitations of b-gluco-

sidases in enzymatic saccharifications of lignocellulosics in

their work on softwoods.

As predicted, the rates of sugar release over 62 h and total

sugar yield improved as the cellulase and b-glucosidase dosage

was raised 6-fold (5 FPUe30 FPU). For example, total sugar

release increased 1.3 fold in the presence of 10 FPU cellulase

plus 10 CBU b-glucosidase, with the main benefit of glucose

rather than xylose release (Fig. 3a,b). However, the final total

sugar yields were lower that anticipated. Given that NS50010

Fig. 3 e Time course of glucose (a) and xylose (b) release by

enzymatic saccharification (5%, w/v; 50 �C, pH 5.2) of

alkaline pretreated (1.0% NaOH; 60 min; 121 �C) wheat

straw using five enzyme combinations. Glucose and xylose

yields are presented as mg/g pretreated material. Data

represents averages of three separate experiments. The

average l.s.d. ( p [ 0.05) [ 12.9 (glucose) and 9.6 (xylose).

b i om a s s an d b i o e n e r g y 3 5 ( 2 0 1 1 ) 3 0 9 4e3 1 0 33100

contains reasonable levels of xylanase (see Table 1), we

assumed that its activity should be sufficient to hydrolyse the

hemicellulose fraction in these pretreated solids. It is well

documented that hemicellulose (xylan) acts as a physical

barrier around the cellulose restricting cellulase access and

decreasing hydrolysis efficiency [40e42]. The hydrolytic effi-

ciency was improved by supplementing the NS50013 and

NS50010 mixture with NS50030. The final combined glucose

andxylose yields increase byanadditional 180mg following the

addition of 1.5 FXU xylanase to the 10 FPU/10 CBU enzyme

mixtures. As shown in Fig. 3a,b approximately 90% of glucose

and xylose was released within 14 h and hydrolysis was virtu-

ally completed by 24 h. There was no advantage in extending

the saccharification after this time; albeit in the absence of

additional xylanase, extended saccharification time is neces-

sary. Both glucose and xylose yields increased by 1.3 and 1.24

fold respectively. More importantly, this enzyme combination

was capable of delivering similar rates of release and total sugar

yields as a 3 fold higher dose of cellulase/b-glucosidase (30 FPU/

30 CBU) mixture lacking the additional xylanase. In essence,

addition of xylanase in the saccharification reaction had the

effect of reducing the cellulase enzyme loadings 3-fold. Further

increases in xylanases (3.0 FXU) failed to promote greater sugar

gains (unreported data).

In a complementary experiment (unreported data), various

enzymatic mixtures were used to saccharify wheat straw

exposed to harsh pretreatment conditions (2% NaOH/90 min/

121 �C). Optimal mixtures (10 FPU/10 CBU/1.5 FXU) released

total sugar yields which peaked at 940 mg/g pretreated mate-

rial, with glucose to xylose ratios approaching 3:1 within the

first 14 h. Increasing the enzyme load 3-fold (30 FPU/30 CBU/

1.5FXU), failed to improve final sugar yield suggesting that

cellulase loading may have reached saturation point. Dispar-

ities in cellulase saturation loading amongst wheat and other

herbaceous straw saccharifications are well documented and

reported to result from variations in enzyme activities and

substrate composition/structure [11,43,44]. Thehighly reactive

nature of alkaline pretreated straw, demonstrated by its near

complete digestion to monomeric sugars and rapid reduction

in volume (80% within 8 h), could prove to be advantageous in

overcoming limitations in solid to liquid load ratios and limi-

tations of dilute sugar streams for fermentation.

3.5. Delignification during mild-alkaline pretreatment

The degree of delignification reflects the effectiveness of the

alkaline pretreatment process. Moreover, it is critical in

improving enzymatic degradation of lignocellulosics and is

ultimately influenced by pretreatment severity [9]. The effect

of NaOH pretreatment on the delignification of wheat straw

was quantified by determining the reduction of acid-insoluble

lignin in pretreated solids as a function of temperature, alkali

concentration and residence time. This was expressed as

a percent reduction compared to original starting material

(Fig. 4). All parameters contributed towards delignification,

particularly temperature, which had the most significant

( p < 0.05) effect. At 121 �C, delignification ranged from 33%

(0.75% NaOH/30 min/121 �C) to a maximum of 72% (2.0%

NaOH/90 min/121 �C). At 2% NaOH/121 �C, similar reductions

in lignin content were attained irrespective of reaction time.

At reduced alkaline strengths, the degree of delignification

between 30min and 90 min became significant ( p < 0.05).

Generally, it was found that increasing alkaline concentration

significantly ( p < 0.05) improved delignification at 121 �C,whereas, responses to increasing time were less pronounced.

Under similar reaction conditions, Varga and co-workers [17]

reported almost complete delignification (>95%) when alka-

line concentrations were raised to 10%, though total recover-

able carbohydrate levels had drastically diminished.

At 60 �C delignification was substantially reduced and

ranged from 15% (0.75% NaOH/30 min) to 42% (2.0% NaOH/

90 min). However, the maximum delignification achieved at

60 �C in 2% NaOH surpasses that at 121 �C in 0.75% NaOH.

Alkaline treatments at 0.75% NaOH were generally ineffective

in delignifying wheat straw irrespective of reaction tempera-

ture and time. The reduced levels of delignification observed

when pretreating in 0.75% NaOH at 60 �C coincides with

reduced sugar yields in saccharified hydrolysates. These

findings show a correlation between delignification and

enhanced enzyme saccharification of pretreated wheat straw.

In fact several studies have demonstrated strong negative

correlations between lignin content and sugar released by

enzymatic hydrolysis [45,46]. Various researchers have

confirmed that lignin directly acts as a physical barrier,

Fig. 4 e Reduction of acid-insoluble lignin in wheat straw

pretreated in 0.75% (white) 1.0% (grey) and 2.0% (black)

NaOH. Results are presented as percent reduction of

untreated wheat straw. The data presented are averages of

three separate experiments. The average l.s.d.

( p [ 0.05) [ 4.77.

b i om a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 3 0 9 4e3 1 0 3 3101

restricting cellulase access to cellulose, and reduces the

enzyme’s activity through non-productive binding [40].

3.6. Total phenolics quantified in hydrolysates

Other compounds formed or released during pretreatment of

lignocellulosic materials are known to have an inhibitory

effect on downstream processes including enzyme sacchari-

fication and fermentation [47,48]. Themain compound groups

are phenols, furans, carboxylic acids and inorganic salts and

their occurrence is related to biomass composition and

pretreatment severity. Neither acids nor inorganic salts were

tested in this study. Wheat, like most monocotyledons, is

typically rich in phenolic acid esters associated with hemi-

cellulose and lignin [49]. Total phenolic release into pretreated

and enzyme saccharified hydrolysates were quantified and

the data presented as a function of changes in pretreatment

temperature, residence time and alkaline strength in Fig. 5.

Excluding the control samples (0.0% NaOH), approximately

7.4-fold more total phenolics were found in pretreatment

Fig. 5 e Total phenolics present in pretreatment and

enzyme saccharified hydrolysates as a function of alkaline

strength, temperature and residence time. Results are

presented as mg/g vanillic acid equivalents. The data

presented are averages of three separate experiments.

liquors (maximum of 1486 mg/g) compared to enzyme

saccharification mixtures (maximum of 200 mg/g). Tempera-

ture had a larger impact than reaction times or alkalinity on

phenolic levels in pretreated liquors. At 121 �C, increasing

pretreatment reaction time and alkaline strength >0.75% did

not generally enhance phenolic release. At 60 �C responses to

changes in time and alkalinity were varied although a net

decrease in total yields is observed.

Most of the phenolics were recovered in the pretreated

liquors particularly in the higher temperature treatments.

Consequently, the total phenolics content found in the enzyme

saccharifiedhydrolysateswere substantially lower. Conversely,

saccharificationmixtures of samples pretreated at 60 �C and all

the water controls contained higher phenolic content. Overall

these results suggest that harsher pretreatment conditions

should produce saccharified hydrolysates with reduced

phenolic content and greater fermentation potential due to

decreased concentrations of inhibitory substances.

3.7. Overall mass balance

A mass balance of the process from pretreatment to enzy-

matic hydrolysis was calculated and is shown in Fig. 6. The

wheat straw at a solid loading of 10% (w/v) was pretreated

under conditions optimised for maximum sugar recovery (2%

NaOH/121 �C/30 min). The remaining insoluble fraction was

separated from the pretreatment hydrolysate prior to enzy-

matic saccharification. The amount of recovered material

corresponded to z 51% (w/w) of the original starting material

and was subjected to saccharification. Enzyme

Fig. 6 e Mass balance of the treatment process. Wheat

strawwas soaked in a 2% NaOH solution at a solid to liquid

ratio of 1:10 and pretreated for 30 min at 121 �C. Insolublefractions were recovered and washed with water and

subjected to enzymatic saccharification at 50 �C for 48 h.

Acid insoluble lignin and xylans were sequential

precipitated from recovered pretreatment hydrolysates.

b i om a s s an d b i o e n e r g y 3 5 ( 2 0 1 1 ) 3 0 9 4e3 1 0 33102

saccharification was performed with a low dose, three

enzyme combination of 2.5 FPU cellulase, 2.5 CBU b-glucosi-

dase and 1.5 FXU xylanase (pre gram of pretreated solids) and

incubated at 50 �C for up to 48 h. Sugar yields were recorded at

279mg of glucose, 136mg of xylose and 15mg of arabinose per

gram of original starting material. The recovered pretreated

hydrolysate was further fractionated through titration with

3M H2SO4. At pH 4.0, 162.6 mg/g of acid insoluble lignin was

recovered as a precipitate. Addition of cold ethanol to the

aqueous phase led to the precipitation of 86 mg/g crude xylan.

The pretreated hydrolysate fraction also contained approxi-

mately 35 mg/g of water extractive storage carbohydrate and

other unquantified polysaccharides, phenolics and degrada-

tion compounds. Finally, the mass balance demonstrates that

combining mild pretreatment conditions with low enzyme

doses gave recoveries of approximately 87% of the theoretical

sugar potential of wheat straw.

4. Conclusion

The dilute alkali pretreatment studies satisfy some of the

more important requisites for an effective pretreatment

process, specifically; excellent delignification, cellulose

enriched fraction that is responsive to enzyme digestion with

high and rapid sugar release and desirable high phenolic

inhibitor losses. Our study also shows that further process

optimisation in pretreatment temperature and/or use of other

enzyme combinations and dosages is possible. Using alkaline

pretreatment to extract oligoxylans and lignins while simul-

taneously improving cellulose hydrolysis can be a means of

consolidating the economic viability of a biorefinery. Consid-

ering its abundance and high sugar potential, wheat straw is

a appropriate feedstock for ethanol production and is ideally

suited to the biorefinery concept.

Acknowledgements

We gratefully acknowledge the financial support provided by

Climate Action Grant (TOC/CAG/013-2007) for this work and

the support of Industry and Investment NSW, Australia. We

express our gratitude to Mr Steve Pepper for technical assis-

tance and Mr Steve Morris for providing advice and assistance

in the presentation of the data.

r e f e r e n c e s

[1] Ragauskas AJ, Nagy M, Kim D, Eckert CA, Hallett JP, Liotta CL.From wood to fuels: integrating biofuels and pulpproduction. Ind Biotechnol 2006;2:55e65.

[2] Campbell CJ, Laherrere JH. The End of Cheap Oil. Globalproduction of convention oil will begin to decline soonerthan most people think, probably within 10 years. Sci Am;1998:78e83.

[3] Zhang YHP. Reviving the carbohydrate economy via multi-product lignocellulose biorefineries. J Ind MicrobiolBiotechnol 2008;35:367e75.

[4] Ragauskas AJ, Williams CK, Davison BH, Britovsek G,Cairney J, Eckert CA, et al. The path forward for biofuels andbiomaterials. Science 2006;311:484e9.

[5] Spatari S, Bagley DM, MacLean HL. Life cycle evaluation ofemerging lignocellulosic ethanol conversion technologies.Bioresour. Technol. 2010;101:654e67.

[6] Dunlop M, Poulton P, Unkovich M, Baldock J, Herr A, Poole M,et al. Assessing the availability of crop stubble as a potentialbiofuel resource. In: Unkovich M, editor. Global issues,paddock action: proceedings of the 14th agronomyconference, Adelaide, Australia, 21e25 Sept 2008. Adelaide:The Regional Institute Limited; 2008.

[7] Fung P, Kirschbaum M, Raison R, Stucley C. The potential forbioenergy production from Australian forests, itscontribution to national greenhouse targets and recentdevelopments in conversion processes. Biomass Bioenergy2002;22:223e36.

[8] Australian Bureau of Agricultural and Resource Economics.Australian crop report. Canberra: Australian Bureau ofAgricultural and Resource Economics; 2009.

[9] Sun Y, Cheng J. Hydrolysis of lignocellulosic materials forethanolproduction: a review.BioresourTechnol 2002;83:1e11.

[10] Antizar-Ladislao B, Turrion-Gomez J. Second-generationbiofuels and local bioenergy systems. Biofuels Bioprod andBiorefin 2008;2:455e69.

[11] Chen Y, Sharma-Shivappa R, Keshwani D, Chen C. Potentialof agricultural residues and hay for bioethanol production.Appl Biochem Biotechnol 2007;142:276e90.

[12] Hsu T. Pretreatment of biomass. In: Wyman CE, editor.Handbook on bioethanol: production and Utilization.Washington, DC: Taylor and Francis; 1996. p. 179e95.

[13] McMillan J. Enzymatic conversion of biomass for fuelsproduction. In: Himmel M, Baker J, Overend R, editors.Pretreatment of lignocellulosic biomass. Washington, DC:American Chemical Society; 1994. p. 292e324.

[14] SendichE, LaserM,KimS,AlizadehH,Laureano-PerezL,DaleB,et al. Recent process improvements for the ammonia fiberexpansion(AFEX)processandresultingreductions inminimumethanol selling price. Bioresour Technol 2008;99:8429e35.

[15] Persson T, Ren JL, Joelsson E, Jonsson A-S. Fractionation ofwheat and barley straw to access high-molecular-masshemicelluloses prior to ethanol production. BioresourTechnol 2009;100:3906e13.

[16] Saha BC, Cotta MA. Ethanol production from alkalineperoxide pretreated Enzymatically saccharified wheat straw.Biotechnol Prog 2006;22:449e53.

[17] Varga E, Szengyel Z, Reczey K. Chemical pretreatments ofcorn stover for enhancing enzymatic digestibility. ApplBiochem Biotechnol 2002;98e100:73e87.

[18] Wang Z, Keshwani DR, Redding AP, Cheng JJ. Alkalinepretreatment of Coastal Bermudagrass for bioethanolproduction. DigitalCommons@University of Nebraska -Lincoln, http://digitalcommons.unl.edu/cgi/viewcontent.cgi?article¼1035&context¼biosysengpres; 2008 [accessed05.08.09].

[19] McIntosh S, Vancov T. Enhanced enzyme saccharification ofSorghum bicolor straw using dilute alkali pretreatment.Bioresour Technol 2010;101:6718e27.

[20] Adney B, Baker J. Measurement of cellulase activity. NRELtechnical report NREL/TP-510e42628, http://www.nrel.gov/biomass/pdfs/42628.pdf; 2008 [accessed 15.09.11].

[21] Miller G. Use of dinitrosalicylic acid reagent forDetermination of reducing sugar. Anal Chem 1959;31:426e8.

[22] ANKOM Technology. Method for determining acid detergentlignin in beakers, http://www.ankom.com/09_procedures/ADL_beakers.pdf; 2005 [accessed 11.11.08].

[23] ANKOM Technology. Acid detergent fibre in feeds- filter bagtechnique method 5, http://www.ankom.com/09_procedures/ADF_81606_A200.pdf; 2006 [accessed 11.11.08].

b i om a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 3 0 9 4e3 1 0 3 3103

[24] ANKOM Technology. Neutral detergent fibre in feeds- filterbag technique method 6, http://www.ankom.com/09_procedures/NDF_081606_A200.pdf; 2006 [accessed 11.11.08].

[25] Sluiter A, Hames B, Ruiz R, Scarlata C, Sluiter J, Templeton D,et al. Determination of Structural carbohydrates and ligninin biomass, http://www.nrel.gov/biomass/pdfs/42618.pdf;2008 [accessed 11.11.08].

[26] Sluiter A, Ruiz R, Scarlata C, Sluiter J, Templeton D.Determination of extractives in biomass, http://www.nrel.gov/biomass/pdfs/42619.pdf; 2008 [accessed 11.11.08].

[27] R Development Core Team. R: a language and environmentfor statistical computing. Vienna, Austria: R Foundation forStatistical Computing; 2009.

[28] Wardlaw I,Willenbrink J. Carbohydrate storageandmobilisationby the culm of wheat between heading and grainmaturity: therelationship to sucrose synthase and sucroseephosphatesynthase. Aust J Plant Physiol 1994;21:255e71.

[29] Silverstein RA, Chen Y, Sharma-Shivappa RR, Boyette MD,Osborne J. A comparison of chemical pretreatment methodsfor improving saccharification of cotton stalks. BioresourTechnol 2007;98:3000e11.

[30] Cheng K-K, Zhang J-A, Ping W-X, Ge J-P, Zhou Y-J, Ling H-Z,et al. Sugarcane bagasse mild alkaline/oxidativepretreatment for ethanol production by alkaline recycleprocess. Appl Biochem Biotechnol 2008;151:43e50.

[31] Varga E, Schmidt A, Reczey K, Thomsen A. Pretreatment ofcorn stover using wet oxidation to enhance enzymaticdigestibility. Appl Biochem Biotechnol 2003;104:37e50.

[32] Zhang Q, Cai W. Enzymatic hydrolysis of alkali-pretreatedrice straw by trichoderma reesei ZM4-F3. Biomass Bioenergy2008;32:1130e5.

[33] Huang H-J, Ramaswamy S, Tschirner UW, Ramarao BV. Areview of separation technologies in current and futurebiorefineries. Sep Purif Technol 2008;62:1e21.

[34] Moure A, Gullon P, Domınguez H, Parajo JC. Advances in themanufacture, purification and applications of xylo-oligosaccharides as food additives and nutraceuticals.Process Biochem 2006;41:1913e23.

[35] Hu Z, Wen Z. Enhancing enzymatic digestibility ofswitchgrass by microwave-assisted alkali pretreatment.Biochem Eng J 2008;38:369e78.

[36] Bura R, Chandra R, Saddler J. Influence of xylan on theenzymatic hydrolysis of steam-pretreated corn stover andhybrid poplar. Biotechnol Prog 2009;25:315e22.

[37] Chang VS, Holtzapple MT. Fundamental factors affectingbiomass enzymatic reactivity. Appl Biochem Biotechnol2000;84e86:5e37.

[38] Berlin A, Maximenko V, Gilkes N, Saddler J. Optimization ofenzyme complexes for lignocellulose hydrolysis. BiotechnolBioeng 2007;97:287e96.

[39] Tengborg C, Galbe M, Zacchi G. Reduced inhibition ofenzymatic hydrolysis of steam-pretreated softwood. EnzymMicrob Technol 2001;28:835e44.

[40] Jeoh T, Ishizawa C, Davis M, Himmel M, Adney W,Johnson D. Cellulase digestibility of pretreated biomass islimited by cellulose accessibility. Biotechnol Bioeng 2007;98:112e22.

[41] Kabel MA, Bos G, Zeevalking J, Voragen AGJ, Schols HA. Effectof pretreatment severity on xylan solubility and enzymaticbreakdown of the remaining cellulose from wheat straw.Bioresour Technol 2007;98:2034e42.

[42] Yang B, Wyman CE. Effect of xylan and lignin removal bybatch and flowthrough pretreatment on the enzymaticdigestibility of corn stover cellulose. Biotechnol Bioeng 2004;86:88e98.

[43] Lloyd TA, Wyman CE. Combined sugar yields for dilutesulfuric acid pretreatment of corn stover followed byenzymatic hydrolysis of the remaining solids. BioresourTechnol 2005;96:1967e77.

[44] Spindler D, Wyman C, Grohmann K. Evaluation of pretreatedherbaceous crops for the simultaneous saccharification andfermentation process. Appl Biochem Biotechnol 1990;24e25:275e86.

[45] Chen F, Dixon RA. Lignin modification improves fermentablesugaryields forbiofuelproduction.NatBiotech2007;25:759e61.

[46] Guo G-L, Hsu D-C, Chen W-H, Chen W-H, Hwang W-S.Characterization of enzymatic saccharification for acid-pretreated lignocellulosic materials with different lignincomposition. Enzym Microb Technol 2009;45:80e7.

[47] Klinke HB, Thomsen AB, Ahring BK. Inhibition of ethanol-producing yeast and bacteria by degradation productsproduced during pre-treatment of biomass. Appl MicrobiolBiotechnol 2004;66:10e26.

[48] Palmqvist E, Hahn-Hagerdal B. Fermentation oflignocellulosic hydrolysates. I: inhibition and detoxification.Bioresour Technol 2000;74:17e24.

[49] Akin D. Plant cell wall aromatics: influence on degradation ofbiomass. Biofuels Bioprod and Biorefin 2008;2:288e303.