Extraction of arabinoxylan from corncob through modified alkaline method to improve xylooligosaccharides synthesis

Accepted Manuscript Extraction of arabinoxylan from corn cob through modified alkaline protocol to improve xylooligosaccharides synthesis

Pranati Kundu, Sandeep Kumar, Vivek Ahluwalia, Sushil Kumar Kansal, Sasikumar Elumalai PII: DOI: Reference:

S2589-014X(18)30007-0 https://doi.org/10.1016/j.biteb.2018.01.007 BITEB 8

To appear in: Received date: Revised date: Accepted date:

9 January 2018 29 January 2018 30 January 2018

Please cite this article as: Pranati Kundu, Sandeep Kumar, Vivek Ahluwalia, Sushil Kumar Kansal, Sasikumar Elumalai , Extraction of arabinoxylan from corn cob through modified alkaline protocol to improve xylooligosaccharides synthesis. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Biteb(2017), https://doi.org/10.1016/j.biteb.2018.01.007

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

Extraction of arabinoxylan from corn cob through modified alkaline protocol to improve

2

xylooligosaccharides synthesis

3

Pranati Kundu,1, 2 Sandeep Kumar,1, 2 Vivek Ahluwalia,1 Sushil Kumar Kansal,2 and

5

Sasikumar Elumalai1,†

T

4

Chemical Engineering Division, Center of Innovative and Applied Bioprocessing (CIAB),

CR

1

7

IP

6

8

Mohali, Punjab 140306 India

Dr. S. S. Bhatnagar University Institute of Chemical Engineering & Technology,

US

2

9 10

AN

Panjab University, Chandigarh 160014 India

11

Abstract

13

A modified alkaline protocol involving a combination of NaOH and NH4OH was employed for

14

the isolation of hemicellulose from corncob. During the extraction of hemicellulose, alkaline

15

reagents promoted selective cleavage of ester and ether linkages in corncob biomass. It was

16

possible to obtain a hemicellulose fraction consisted of considerable branching constituents

17

(arabinose and uronic acid, and low lignin). Based on modeling analysis, reaction parameters

18

such as alkali concentration and temperature significantly influenced the amount of total

19

hemicellulose extracted. Subsequent hydrolysis of isolated hemicellulose in the presence of

20

H2SO4 resulted in better conversion (69% wt.) with enriched XOs conc. (73.68% with DP up to

21

4) than other fractions under milder conditions. Advantageously, gas phase NH3 formation was

22

achieved during the reaction, where NaOH and NH4OH mix was used at an equal ratio that

AC

CE

PT

ED

M

12

†

Author correspondence email: [email protected] (Tel: +91-172-5221-444). 1

ACCEPTED MANUSCRIPT 23

could significantly help in reducing the overall processing cost of XOs production (through

24

recycling and reuse) during large-scale manufacture.

25

Key words: xylooligosaccharide, corncob, hemicellulose, arabinoxylan, severity parameter

27

model.

T

26

IP

28

1. INTRODUCTION

30

Xylooligosaccharides (XOs) have been receiving considerable attention in recent years due to

31

their potential applications in food and pharmaceutical industries. They are known for their

32

‘prebiotic’ effects when consumed as a part of the diet. The fast growth of the functional food

33

market and increasing number of other industrial applications has led researcher’s world-over, to

34

explore different sources and develop technologies for the commercial production of XOs with a

35

defined degree of polymerization (DP). XOs are a mixture of xylose oligomers commonly

36

composed of 2 to 10 units, which can be obtained by hydrolyzing β-(1→4)-glycosidic linkages

37

of xylan. The application of these oligosaccharides varies according to the chain length, for

38

example, xylobiose (DP 2) is mainly considered for food applications (Chapla, Pandit, & Shah,

39

2012; Vazquez, Alonso, Domınguez, & Parajo, 2000). Xylan can be extracted from under-

40

utilized lignocellulosic biomass, such as corncob and bagasse as hemicellulose or hetero-xylan.

41

Among the agri-residue biomass, corn cob is known to be an excellent substrate for the

42

production of several industrial enzymes (e.g., cellulase, xylanase), XOs, xylitol, and xylose.

43

Although corncob is rich in hemicellulose (35-40% wt.), yet its potential as an important source

44

of xylan is still challenging due to its heterogeneity (Yuan, Zhang, Qian, & Yang, 2004).

45

Hemicellulose is considered to be a complex component of the plant cell wall due to its close

AC

CE

PT

ED

M

AN

US

CR

29

2

ACCEPTED MANUSCRIPT association with cellulose and lignin constituents through hydrogen and covalent linkages. This

47

hetero-polymer is an amorphous polysaccharide, which has β-(1→4)-linked backbones with an

48

equatorial configuration, consisting of predominantly xylan (xylose units accounts averagely

49

>70%), carbohydrate residues, e.g., arabinose, galactose, glucose, mannose, uronic acid, and an

50

aromatic constituent-lignin. The composition of hemicellulose varies in different plant species.

51

For example, the dominant form of hemicellulose polysaccharide in annual plants like straw and

52

grass is (glucurono) arabino-xylan, which has a backbone of β-(1-4)-xylosyl units linked with a

53

few short side chains of arabinosyl at O-2 and/or O-3 position, and α-(1,2)-glucuronic acid

54

and/or its 4-O-methyl derivative on O-2 and acetyl group on the arabinose or xylose. To some

55

extent, the xylan branch groups can be associated with other groups. Aromatic feruloyl (from

56

ferulic and diferulic acids), and p-coumaryl have been reported to be attached to arabinose

57

residues at the O-5 position (Carvalho, de Oliva Neto, Da Silva, & Pastore, 2013; Wang et al.,

58

2015).

ED

M

AN

US

CR

IP

T

46

Hemicelluloses are generally classified as alkali-soluble material after removal of the

60

pectic substances due to their weak inter-unit linkages between the substituted constituents

61

(Farhat et al., 2017). Till date, a variety of extraction methods have been reported, including

62

steam and alkali extraction as well as other supplementary combinations, such as alkali and

63

hydrogen peroxide, alkali and chlorite solutions or dimethyl sulfoxide. The final composition of

64

hemicellulose is dependent on nature of the chemicals employed (da Silva, de Oliveira, da Silva

65

Neto, Pimentel, & dos Santos, 2015). Much of the extraction studies have used NaOH or KOH as

66

carbohydrate dissolution reagent due to its efficacy and low cost (Qing, Li, Kumar, & Wyman,

67

2013). It has been reported that NaOH treatment effectively disrupts the cell wall by dissolving

68

hemicellulose and lignin, hydrolyzing uronic and acid esters, thereby cause swelling of cellulose

AC

CE

PT

59

3

ACCEPTED MANUSCRIPT (Luo et al., 2012). Alternatively, for the purpose of necessary lignocellulosic depolymerization,

70

recent studies have used aqueous NH3, which is known to cleave the C-O-C bonds in lignin

71

selectively, and ether and ester bonds in lignin-carbohydrate complexes. However, this treatment

72

resulted in limited carbohydrate dissolution and also, it may not be effective for the pretreatment

73

of biomass material having relatively higher lignin content such as wood feedstocks (Gunawan et

74

al., 2017). However, these alkaline treatments cause remarkable morphological changes in the

75

resultant substrate, thereby improving its accessibility of chemicals or enzymes during

76

depolymerization reaction (da Costa Sousa et al., 2016).

CR

IP

T

69

The traditional XOs production methods using lignocellulosic biomass involves the use

78

of either single-step or multi-step processing technique (Carvalho et al., 2013). The commonly

79

preferred techniques in direct XOs synthesis are auto-hydrolysis and treatment of biomass in the

80

presence of acid or enzyme (Akpinar, Erdogan, & Bostanci, 2009a; Surek & Buyukkileci, 2017).

81

However, these techniques normally produce low product yield (≤10% xylan conversion). The

82

resultant hydrolysate in the case of acid hydrolysis consists of a variety of undesirable

83

components, including soluble lignin. In addition, it generates a large amount of mono- and di-

84

saccharide units along with its dehydration products (e.g., furfural) and thus, requires extensive

85

purification steps (Aachary & Prapulla, 2011; Akpinar, Erdogan, & Bostanci, 2009b). However,

86

the two-stage approach involves the extraction of hemicellulose preferentially using alkaline

87

reagent and further hydrolysis in the presence of acid or enzyme. This method is industrially

88

viable due to its high product yield (25-37% XOs conversion) along with the desired degree of

89

polymerization (DP 2 to 6) (Brienzo, Carvalho, & Milagres, 2010; Michelin, Ruiz, Maria de

90

Lourdes, & Teixeira, 2018). However, hemicellulose yield and its characteristics are highly

91

dependent on the employed extraction protocol. Similarly, the purpose and application of the

AC

CE

PT

ED

M

AN

US

77

4

ACCEPTED MANUSCRIPT 92

hemicellulose also vary according to the extraction method (H. Li et al., 2015). The important

93

characteristics of hemicellulose for applications in derived compounds preparations, including

94

XOs,

95

branching/substitution (e.g., sugars units, acetyl groups, and phenolics), residual lignin, solubility

96

and reactivity (Luo et al., 2012; Qing et al., 2013). In the present study, a modified alkaline

97

protocol involving both NaOH and NH4OH was employed during the hemicellulose extraction

98

with the aim of achieving a fraction consisting nominal branched units that allows enriched

99

xylooligosaccharieds synthesis during the secondary reaction in the presence of acid or enzymes

100

(de Figueiredo, Carvalho, Brienzo, Campioni, & de Oliva-Neto, 2017). In order to verify the

101

most influential parameter during the hemicellulose extraction through the modified protocol, an

102

extended severity parameter model was used.

and

biochemicals

are

degree

of

polymerization

(DP),

degree

of

103

M

AN

US

CR

IP

T

furanics,

2. EXPERIMENTAL METHODS

105

2.1 Chemicals

106

Analytical grade chemicals and solvents, including H2SO4, NaOH, NH4OH, and ethanol were

107

purchased from Sigma Aldrich (India). Carbohydrate standards, such as glucose, xylose, and

108

arabinose were also purchased from Sigma Aldrich (India). XOs standards ranging from xylose

109

to xylopentose were purchased from Megazyme Inc., USA. All chemicals used in this study were

110

used as received without any modifications unless otherwise indicated.

AC

CE

PT

ED

104

111 112

2.2 Feedstock preparation and compositional analysis

113

Corncob sample was collected from an agricultural field near Mohali (Punjab), India. It was air-

114

dried to 6% moisture content (dry basis). The particle size was reduced to 0.5 mm (30 mesh)

5

ACCEPTED MANUSCRIPT using a commercial grinder (Kinematica PX-MFC 90 D, Luzern, Switzerland). The powdered

116

material was stored in plastic bags and kept under refrigeration condition until further use. The

117

fractional compositional analysis of corncob sample was done by following standard NREL

118

protocols. In order to determine the carbohydrate contents (glucose, xylose, and arabinose), and

119

other lignin and ash contents of the sample, it was first extracted with ethanol/benzene mixture

120

(Sluiter, Ruiz, Scarlata, Sluiter, & Templeton, 2010). Similarly, glucuronic acid content was

121

determined according to the method of reported method of (Tobimatsu et al., 2012).

CR

IP

T

115

122

2.3 Hemicellulose extraction from corncob

124

The finely ground corncob powder was mixed with the alkaline solution (NaOH plus NH4OH or

125

NaOH or NH4OH) at concentrations ranging between 0.25 and 2.5 M in 250 ml glass bottle (da

126

Silva et al., 2015). The solid to liquid ratio was maintained at 1:20 (w/v). The bottle was tightly

127

closed to avoid any evaporation and mixed using a magnetic stirrer. The corncob slurry was

128

gradually heated to the desired temperature under stirring condition (200 rpm) on a hotplate.

129

After completion of reaction (up to 60 min), the resultant thick slurry was centrifuged at 8000

130

rpm for 15 min at 10°C (Eppendorf 5810R), and the supernatant was collected and cooled to

131

room temperature. The pH of the supernatant was reduced to 5.5 by addition of 6 M HCl at room

132

temperature and followed by addition of ~3 volumes of absolute ethanol. The sample was

133

centrifuged at 8000 rpm for 15 min 10oC to collect pellets rich in hemicellulose. They were

134

washed thoroughly using distilled and freeze dried by using Martin Christ Delta 2–24 LSC plus

135

drier. Finally, the hemicellulosic fractions obtained through different treatments viz. NaOH plus

136

NH4OH, NaOH, and NH4OH were labeled as H1, H2, and H3, respectively. After extraction, the

AC

CE

PT

ED

M

AN

US

123

6

ACCEPTED MANUSCRIPT 137

carbohydrate content of the isolated hemicelluloses was determined according to the standard

138

NREL protocol (Sluiter et al., 2010).

139

2.4 Acidic hydrolysis of isolated hemicellulose fractions

141

About 100 mg sample of isolated hemicellulosic fractions (H1, H2, and H3) were suspended in 10

142

ml 0.25 M H2SO4 solution and placed in an oil bath at 100°C under continuous stirring condition

143

(Bian et al., 2014). Once the reaction was completed, the mixture was cooled down immediately

144

to room temperature using an ice bath, neutralized with calcium carbonate (CaCO3) solution, and

145

centrifuged at 8000 rpm for 15 min at 10°C to remove the resulted precipitate. The supernatant

146

was filtered through 0.45 µm nylon filter. The hydrolyzed products were analyzed using high-

147

performance liquid chromatography (HPLC, Agilent Technologies 1200 infinity series, CA,

148

USA) for XOs determination. The system was equipped with Agilent Hi-Plex Ca column (300

149

mm length, 8 µm porosity) and maintained at 85°C, 0.6 mL/min of HPLC grade water (mobile

150

eluent) on refractive index (RI) detector operated at 40°C. The concentration of XOs was

151

estimated using their respective calibration charts prepared by commercial grade standards. At

152

the same time, traditional thin layer chromatographic (TLC) technique was also employed for the

153

determination of XOs. The solvent system used was ethyl acetate:acetic acid:1-butanol:water

154

mixture prepared in the ratio of 4:3:2:2. 1- naphthyl ethylenediamine dihydrochloride (0.5%) in

155

5 % sulfuric acid : methanol mix was used as spraying agent.

AC

CE

PT

ED

M

AN

US

CR

IP

T

140

156 157

2.5 Enzymatic hydrolysis of isolated hemicellulose

158

Twenty mg sample of isolated hemicellulosic fractions (H1, H2, and H3) were suspended in 10 ml

159

of 50 mM sodium acetate buffer (pH 4.8) in 100 mL Erlenmeyer flask (Huang et al., 2017). The

7

ACCEPTED MANUSCRIPT 160

substrate was loaded with 10 U/ml commercial xylanase (Sigma Aldrich, India) and incubated at

161

50°C under continuous shaking condition (250 rpm). The enzymatic reaction was continued up

162

to 24 h. A ~0.5 ml aliquot was taken at each specific time intervals and analyzed for XOs

163

concentration on HPLC by adapting the same method, as mentioned earlier.

T

164

2.6 Colorimetric ammonia determination

166

Ammonia concentration in the supplemented alkaline solution was determined calorimetrically

167

(Elumalai, Roa-Espinosa, Markley, & Runge, 2014). Briefly, 5 ml sample solution

168

(NaOH+NH4OH) was taken in an Erlenmeyer flask with stopper, and 0.2 ml phenol reagent (20

169

g phenol in 200 ml 95% v/v ethanol) was added. After stirring the mixture for few mins, the

170

following reagents were added sequentially as 0.2 ml of sodium nitroprusside (1.0 g sodium

171

nitroprusside in 200 ml water) and 0.5 ml of oxidizing solution (a mixture of 100 ml alkaline

172

reagent and 25 ml of sodium hypochlorite). The alkaline reagent was prepared by dissolving 100

173

g sodium citrate and 5 g NaOH in 500 ml distilled water. The solution mixture was stirred for

174

few mins and allowed to stand for 1 h at room temperature. Finally, the generated blue color was

175

read at 640 nm on UV/Vis-spectrophotometer (Agilent Cary 60).

CR

US

AN

M

ED

PT

CE

176

IP

165

2.7 Analytical instrumentation

178

Fourier transform infrared spectroscopy (FTIR): Hemicellulose fractions obtained through

179

different alkaline treatments were analyzed on FTIR instrument (Agilent Carry 600). The

180

spectrum were recorded between 400 and 4000 cm-1 at 4 cm-1 nominal resolution under room

181

temperature condition with maximum 10 scans.

AC

177

182

8

ACCEPTED MANUSCRIPT 183

NMR spectroscopy: 2D-HSQC NMR spectra of the hemicellulose fractions (H1, H2, and H3)

184

were acquired on a Bruker AVIII 600 MHz spectrophotometer using ~40 mg sample dissolved in

185

1 mL D2O. Data processing was performed using standard Bruker topspin NMR software.

186

2.8 Alkaline treatment modeling

188

Modified combined severity model developed by Silverstein and co-workers was used for better

189

prediction of yield responses during alkaline treatments (Elumalai, Agarwal, & Sangwan, 2016;

190

Silverstein, Chen, Sharma-Shivappa, Boyette, & Osborne, 2007). The model combines the main

191

processing parameters like alkali concentration, temperature and time into a single factor, in

192

order to relate to reaction recoveries, as represented in Eq. (1):

AN

US

CR

IP

T

187

T −T

r b M0 = t × Cn × exp [( 14.75 )]

(1)

ED

194

M

193

195

where Mo - modified severity parameter; t - residence time (min); C - chemical concentration (%

197

wt.); Tr - reaction temperature (°C); Tb - base temperature (100°C); n - arbitrary constant; 14.75 -

198

conventional energy of activation. Meanwhile, hemicellulose constituents recovery, mainly

199

xylose, arabinose and uronic acid during various alkaline treatments was predicted using second

200

order polynomial model consisting of alkali concentration, temperature and time, as continuous

201

variables is Eq. (2):

AC

CE

PT

196

202 203

Yresponse = β0 + β1 C + β2 T + β3 t + β4 CT + β5 Tt + β6 Ct + β7 C2 + β8 T 2 + β9 t 2

204

9

(2)

ACCEPTED MANUSCRIPT 205

Where β0 is the intercept; β1-n is the interaction coefficients; C is the alkali concentration (% wt.);

206

T is the temperature (°C); t is the reaction time (min). The recoveries data were subjected to

207

ANOVA analysis to examine the adjustment of the model using SPSS software.

208

2.9 Statistical data analysis

210

Statistical regression and ANOVA analyses were performed to the alkaline treatments recovery

211

data using SPSS software v23. The quality of model fit was expressed by the squared correlation

212

coefficient, R2 value. Model terms were evaluated based on the probability (p) value with 95%

213

confidence. Multiple (pairwise) comparisons among alkaline treatment recovery means were

214

carried out using Tukey’s honestly significance difference (HSD) test at α levels of 0.05.

AN

US

CR

IP

T

209

215

3. RESULTS AND DISCUSSION

217

3.1 Alkaline treatment of corncob for hemicellulose isolation

218

The present study employed a modified alkaline protocol, which involved both NaOH and

219

NH4OH in equal proportions for the purpose of hemicellulose extraction from corncob and used

220

for XOs synthesis. With respect to the characteristic nature of these alkaline reagents, it can be

221

reasonably expected that selective cleavage of ester and ether linkages of the branching as well

222

as associated constituents of hemicellulose, thereby rendering it suitable for improved

223

hydrolysis. Initial fractional compositional analyses of native corncob biomass revealed that it

224

consisted 26.8% glucose, 26.2% xylose, 1.8% arabinose, 3.4% glucuronic acid, 20.68% both

225

klason and soluble lignins, and 1.5% ash contents (dry wt. basis). Yield responses of the

226

hemicellulose constituents with respect to the increased severities during alkaline treatment were

227

recorded and presented in Table 1. Apparent results displayed that significantly improved solids

AC

CE

PT

ED

M

216

10

ACCEPTED MANUSCRIPT (relative differences between 8 and 74% wt.) recovery with xylose content (27-46% wt. based on

229

fractional composition analysis of solid precipitate) was achieved at incremented reaction

230

severities (combined severity factor, CS ranging from 2.43 to 3.75). However, increased residual

231

arabinose content (referring L-arabinose) was attained relatively 7-fold throughout the

232

incremented severities (i.e., xylose to arabinose ratio decreased from 9.1 to 6.2). It was also

233

noted that treatment maintained increasing residual acetyl content (referring to glucuronic acid

234

and/or 4-O-methyl glucuronic acid) in the precipitate, i.e., 4-fold relative differences and

235

therefore, the xylose to uronic acid ratio was decreased from 6.8 to 3.7. This characteristic

236

changes in the isolated fraction might be attributed to the disruption and breaking of hydrogen

237

bonds, and selective cleavage of the ester-linked substituents of the hemicellulose and other

238

associated cell wall constituents through the parallel action of alkaline reagents, depending on its

239

distinct nature (S.-L. Sun, Wen, Ma, & Sun, 2013).

M

AN

US

CR

IP

T

228

For comparison, reference treatment involving either NaOH or NH4OH alone was

241

employed for the extraction of hemicellulose from corncob biomass under similar conditions

242

(Table 1). The NaOH treatment resulted in increased xylan yield, i.e., maximum ~2-fold relative

243

differences

PT

ED

240

the

raised

severities

AC

CE

throughout

11

(CS

from

2.10

to

3.70).

ACCEPTED MANUSCRIPT 244 245

Table 1. Alkaline isolation of hemicellulose from corncob biomass under varied severity conditions. Alkali load on solid mass (gm per gm) NaOH NH4OH 2.50 4.24 2.50 4.24 2.50 4.24 5.0 8.48 5.0 8.48 5.0 8.48 5.0 5.0 5.0 10.0 10.0 10.0 8.48 8.48 8.48 16.96 16.96 16.96

t (min)

Modified CS factor, log Mo

80 100 121 80 100 121 80 100 121 80 100 121 80 100 121 80 100 121

60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60 60

2.43 (n=1.5) 3.02 3.64 2.54(n=1.2) 3.13 3.75 2.10(n=1.3) 2.69 3.31 2.49(n=1.3) 3.08 3.70 2.40(n=1.3) 2.99 3.60 2.54(n=1.5) 3.13 3.75

246 247 248

D E

C A

c

Xylose yield (% wt.)

U N

A

M

T P

9.6(14.5) 0.9**(0.4**)

b

12.6 14.3 16.1 36.6 32.1 19.9 49.2 31.8 6.5 17.8 72.9 65.2 0.7 1.7 2.4 0.9 2.5 4.3

T P

I R

C S

E C

Std. error of mean Tukey’s HSD @ p<0.05 a

Solid precipitate recovery (% wt.) 5.3 b(8.0) 7.1(9.5) 9.2(8.6) 14.2(23.0) 13.9(19.1) 15.4(12.1) 17.0(28.5) 18.8(16.2) 12.1(6.0) 26.6(42.7) 26.1(44.4) 26.2(33.4) 0.3(0.4) 0.9(1.1) 1.7(1.8) 0.4(0.6) 1.2(1.6) 2.5(2.9)

a

T (°C)

26.1 0.1**

Xylose to arabinose ratio 9.1 10.3 6.8 7.9 6.9 6.2 13.4 10.4 6.6 17.4 12.7 8.0 8.6 10.2 14.2 7.5 6.6 6.0

Xylose to uronic acid ratio 6.8 3.8 4.0 4.9 6.6 3.7 6.6 4.8 1.5 1.5 5.3 5.6 4.3 3.4 3.2 6.7 4.4 3.8

9.1 0.1*

0.4 0.1*

Modified combined severity factor; Total carbohydrate recovery was calculated based on solids yield with residual sugars content (glucose, xylose, arabinose, uronic acid) to the potential contents; b xylan recovery was calculated based on the potential xylose content. Asterisks (*) and (**) indicate statistical significance and not significance at 95% confidence level.

12

ACCEPTED MANUSCRIPT In contrast to the response attained with modified alkaline treatment, decreasing pattern of

250

residual arabinose was observed with the incremented severities up to 58% relative

251

differences (from 4.07 to 1.74% wt. at 10% loading and between 80-121 °C), due to which

252

the xyl/ara was decreased from 13.4 to 8.0 (at the same time xylose concentration was

253

decreased from 71.8 to 65.2). However, the uronic acid concentration was steadily increased

254

with the corresponding severity levels up to 60% relative differences, resulted in decreased

255

proportionate of xyl/uro ratio from 6.6 to 5.6. Normally, strong alkaline solvents show higher

256

selectivity towards cleavage of weak bonds (ester linkages) connecting arabinose and uronic

257

acid contents, however, the tendency of the release of each of the constituents was dependent

258

of biomass heterogeneity and processing conditions (Luo et al., 2012; R. Sun, Lawther, &

259

Banks, 1995). Whereas, the NH4OH treatment could yield relatively up to 6-fold solid and 9-

260

fold xylan recoveries along with meager residuals (arabinose and uronic acid) concentrations,

261

i.e., < 10% wt. in response to the increased severity levels (CS from 2.40 to 3.75). While the

262

comparison of the yield results, the modified protocol resulted in achieving relatively 12-15%

263

lower and 50-60% higher solids precipitate yield with xylan content than individual NaOH

264

and NH4OH protocols, respectively. However, it extracted the xylan consisting more

265

branched substituents, such as arabinose and uronic acid that is relatively up to 54% and ~5-

266

fold than NaOH and NH4OH treatments, respectively. Further residual lignin content of the

267

solids obtained under maximum severity conditions was determined, and it was estimated

268

that modified protocol yielded 18% wt., whereas NaOH and NH4OH protocols yielded 24%

269

and 6% wt. respectively. This adverse effect is caused due to the characteristic effect of

270

NaOH, which effectively solubilize the hemicellulose and lignin through hydrolyzing the

271

ester linkage constituents, such as arabinose, uronic acid, and acetic acid. It also cleaves the

272

ether linkages between hemicellulose and lignin, and the ester bonds between lignin and/or

AC

CE

PT E

D

MA

NU

SC

RI

PT

249

13

ACCEPTED MANUSCRIPT hemicelluloses and hydroxycinnamic acids (e.g., ferulic and p-coumaric acids) (Bender et al.,

274

2017; S.-L. Sun et al., 2013). Mechanistically, NH4OH reacts selectively with lignin by

275

cleaving C-O-C bonds in lignin and ether and ester bonds in lignin-carbohydrate complexes

276

but limits carbohydrate extraction (da Costa Sousa et al., 2016). Therefore, results suggest

277

that change in the characteristic nature of alkaline solvent greatly influenced the selective

278

cleavage of alkali-labile linkages, thereby enabling the release of carbohydrate sugars along

279

with lignin (Bender et al., 2017). However, the absolute relation between cleavage

280

mechanism of xylan and its branched substituents during modified alkaline isolation is not

281

clear. Overall, maximum yield responses were achieved under higher severity conditions and

282

therefore, it was selected for further validation. However, those fractions exhibited lower

283

carbohydrate recovery might be attributed to the stringent reaction led to the loss of

284

carbohydrates via hydrolysis and further dehydration, and also, lignin. In complementary to

285

the above alterations, it can be expected that these in the substrate fraction could positively

286

reflect the hydrolysis result.

D

MA

NU

SC

RI

PT

273

Additionally, uncontrolled gas phase NH3 formation was observed during alkaline

288

supplementation reaction after prolonged reaction (60 min) due to the continuous supply of

289

external OH- ions by NaOH into the medium consisting NH4OH, according to shift reaction

290

mechanism (see E-supplement data) (Elumalai et al., 2014). In practical, this phenomenon

291

could favorably help in the development of cost-effective biorefining process through

292

recovery and reuse technique.

AC

CE

PT E

287

293 294

3.2 Modeling of alkaline isolation of hemicellulose

295

In order to validate the effectiveness of alkaline solvents on hemicellulose recoveries during

296

corncob treatments, conventional models, such as linear and quadratic algebraic equations 14

ACCEPTED MANUSCRIPT were used. These models consciously relate the reaction parameters and yield responses of

298

dissolution of carbohydrate constituents after alcohol washing, assuming complete recovery

299

of dissolution constituents, as recovery data presented in Table 1. The second order

300

polynomial model consisting of the following dependent variables: alkali concentration (C),

301

temperature (T) and reaction time (t) was used to measure the respective percent recovery of

302

hemicellulosic constituents mainly xylose, arabinose and uronic acid during treatment. The

303

model estimates for recovery of all three constituents were identified from ANOVA analysis

304

as linear, interaction and quadratic terms, and used to quantify the predicted recoveries.

305

Based on ANOVA results, the model equations Eqs. (3-5) represent for all three constituents

306

recovery were in the following form (see E-supplement data):

NU

SC

RI

PT

297

307

309

Xylose recovery, % = −59.09 + 2.25C − 0.16T + 0.11C2 − 0.01𝑇 2 − 0.07C ∗ T + 0.16𝐶 ∗

MA

308

𝑡 + 0.01𝑇 ∗ 𝑡

312

Arabinose recovery, % = −165.55 + 6.59C + 0.155T + 0.32C2 + 0.01𝑇 2 − 0.04C ∗ T +

PT E

311

D

310

0.18𝐶 ∗ 𝑡 + 0.01𝑇 ∗ 𝑡

316

Uronic acid recovery, % = −15.24 + 0.40C + 0.01T + 0.02C2 + 0.01𝑇 2 − 0.02C ∗ T + 0.03𝐶 ∗ 𝑡 + 0.01𝑇 ∗ 𝑡

(5)

AC

315

(4)

CE

313 314

(3)

317

Individual xylose, arabinose, and uronic acid linear models exhibited the correlation values

318

(R2) more than 0.90 with respect to the reaction severities, based on regression analysis.

319

Further the respective values of C, T, and t were incorporated into the polynomial equation in

320

order to validate the model, and consequently, plots between predicted versus experimental 15

ACCEPTED MANUSCRIPT values were made to all hemicellulose constituents, as displayed in Fig. 1. The plots

322

established R2 values of 0.89, 0.96 and 0.82 to the corresponding hemicellulose constituents,

323

establishing the model better predicted the yield responses. Based on statistical analysis, C

324

and T had the most influential impact on hemicellulose constituents recovery among

325

dependent variables and in particular, had the strongest effect on xylose and arabinose yield

326

(p<0.05-0.1) (see E-supplement data). Nearly similar results were observed to the individual

327

reference treatment protocols involving only NaOH or NH4OH, i.e., xylose and arabinose

328

were strongly influenced by C and T, where the R2 value lies in the range 0.50-0.95 and

329

p<0.05-0.1, as data presented in E-supplement data. This is caused due to less available

330

model terms in the equation and multi-step procedure for carbohydrate recovery and

331

therefore, it could not be effectively explained the model for better prediction of the yield

332

responses (Haghi & Zaikov, 2014).

MA

NU

SC

RI

PT

321

Simultaneously, a linear model relating combined severity parameter (single factor)

334

and the model recovery responses was used to represent the yield data (based on Table 1) of

335

modified alkaline protocol, as follows Eqs. (7-9) (according to E-supplement data):

PT E

D

333

336 337

Xylose recovery, % = 3.01 ∗ log Mo + 9.64

338

Arabinose recovery, % = 9.79 ∗ log Mo + 0.78

(7)

339

Uronic acid recovery, % = 0.59 ∗ log Mo + 2.17

(8)

CE

AC

340

(6)

341

16

ACCEPTED MANUSCRIPT 342

T P

I R

C S U

N A

343 344

M

345

Fig.1 Correlation plot of modified severity factor (log Mo) vs. experimentally yield data of xylose, arabinose, and uronic acid through different

346

alkaline protocols using (a) NaOH supplemented with NH4OH, (b) individual NaOH and (c) NH4OH reagents.

D E

T P E

C C

A

17

ACCEPTED MANUSCRIPT While plotting the graph between respective log Mo versus hemicellulose constituents

348

recovery, the R2 value lies between 0.55 and 0.95 (see E-supplement data). The n-values for

349

dissolution of these hemicellulose constituents that provided the best model fits were 0.1, 10

350

and 0.1, respectively and were obtained by data training while keeping the log Mo value

351

positive range. The modified severity parameter model was validated by plotting the

352

predicted values against experimental percent recovery of hemicellulose constituents.

353

Resulted in R2 of the plots exhibited between 0.5 and 0.9, indicating that model fairly

354

predicted the response yield (as provided in E-supplement data). Presumably, this might be

355

attributed to the feedstock (lignocellulose) heterogeneity and inability of the extended

356

severity parameter model to completely capture the dependence of response variables on

357

independent variables (treatment recoveries) in the absence of variables like solid to liquid

358

ratio, treatment time and physical property of feedstock material (e.g. particle size) (Elumalai

359

et al., 2016; Silverstein et al., 2007). Nearly similar results were observed to the reference

360

treatment methods (involving only NaOH or NH4OH) that is model less fairly represented the

361

response yield data (overall R2 values lies between 0.1 and 0.9) while using the

362

correspondingly developed model, based on the data obtained from the respective correlation

363

plot, as displayed in E-supplement data.

RI

SC

NU

MA

D

PT E

CE

364

PT

347

3.3 Characterization of isolated hemicelluloses

366

Typical FT-IR spectroscopic characterization was performed to the isolated hemicelluloses,

367

namely H1, H2, and H3 referring to fractions obtained through modified NaOH plus NH4OH,

368

and individual NaOH and NH4OH protocols, respectively, in order to determine its main

369

functional groups. The spectra of these three individual fractions were collected between 400

370

and 4000 cm-1. The comparative spectrum exhibited nearly identical absorption trend with the

AC

365

18

ACCEPTED MANUSCRIPT samples except for few additional strong absorption intensities with H2 (see E-supplement

372

data). In the fingerprint region, the high absorbance at 1584 cm-1 is attributed to C-H

373

deformation. Likewise, the small band at 1508 cm-1 originated from the aromatic skeletal

374

vibrations of the benzene ring, suggested that minor amount of bound lignin is presented in

375

the hemicellulose precipitates. The bands at 1456 and 1374 cm-1 are assigned to C-H

376

vibrations of polysaccharides. The presence of glucuronic acid was identified mainly by the

377

appearance of symmetric stretching from COO groups at 1420 cm -1. Similarly, the

378

wavenumber at 1258 cm-1 is the indicative of carboxylic acid vibrations, which might be due

379

to attachment of 4-O-methyl-α-glucuronic acid side groups. The spectrum displayed that each

380

particular polysaccharide had a maximum specific band in the region 1200-800 cm-1, and the

381

band at 1149 cm-1 is assigned to C-OH ring vibrations overlapped with stretching vibrations

382

of side groups and glycosidic bond vibrations (C-O-C). In accordance with the literature

383

reports, absorption peaks attained in the range 1175-1140 cm-1 might possibly due to different

384

conformational states of hemicellulose molecules. Whereas, the peaks within 1170-1050 cm-1

385

range are related to the hemicelluloses. In anticipation, H2 exhibited a strong absorption at

386

3343 cm-1 and 2950-2850 cm-1 regions are assigned to hydrogen bond O-H stretching and C-

387

H stretching vibrations, respectively (Li, Sun, Zhou, Peng, & Sun, 2015; H. Li et al., 2015).

CE

PT E

D

MA

NU

SC

RI

PT

371

Additional structural characterization was performed to the hemicelluloses on NMR

389

spectroscopy using 2-D HSQC technique. The NMR result of H1, where the five dominant

390

signals at δC/δH 102.67/4.38, 73.83/3.21, 75.19/3.36, 75.38/3.68, 64.28/3.97+3.31 are

391

typically assigned to C1-H1, C2-H2, C3-H3, C4-H4, C5-H5 of the (1→4)-linked β-D-Xylp units,

392

respectively (see E-supplement data). The cross peaks at 80.84/3.97 (C2) and 86.88/ 4.14 (C4)

393

indicated the presence of α-L-Araf residues, and similarly, the cross peak at 74.22/3.39 is

394

assigned to the internal Xylp unit in heteroxylan backbone. The presence of D-Galp residues

AC

388

19

ACCEPTED MANUSCRIPT located in the side polysaccharide chains was confirmed from the anomeric carbon signals at

396

101.7/4.33 (C1-H1) and 64.08/3.87. Whereas, H2 reference sample exhibited additional

397

signals over H1 at 100.92/4.25 (C1-H1) and 75.77/ 3.26 (C2-H2), which are assigned to (1→4)-

398

β-D-Xylp-2-O-(4-O-Me-D-GlcpA) unit (see E-supplement data). Furthermore, the cross

399

signals at 73.05/3.03 (C2-H2), 73.12/3.46 (C3-H3), 78.3/3.61 (C4-H4) 78.6/3.81 (C5-H5)

400

represented 4-linked β-glucose units. Other cross peaks at 74.02/3.53 and 63.69/3.87 are

401

assigned to (1→4)-linked α-D-Glcp unit. Whereas, signals at 75.97/3.81 (C5-H5) and

402

75.77/3.49 (C4-H4) are assigned to the 4-linked β-mannose unit. Similarly, O-CH3 acetyl

403

group was confirmed by the cross signal at 62.33/3.44. Similarly, H3 exhibited nearly same

404

result as that of others (H1 and H2) except few cross signals spotted at 73.24/3.79 (C2-H2),

405

78.31/3.41 (C3-H3), 82.40/3.2 (C4-H4) and 60.96/3.72 (see E-supplement data) are the

406

indicative of methoxyl groups of 4-O-methyl-D-glucuronic acid residues, which were

407

possibly due to inherent branched substituent 4-O-methyl glucuronoxylan. Results were in

408

good agreement with the literature reports (H.-Y. Li et al., 2015; H. Li et al., 2015; Ma et al.,

409

2016; Wang et al., 2015). Thus, comparative characterization results ensured the presence of

410

carbohydrates content along with its residual impurities and showed that composition of

411

solids varies depending on the employed alkaline solvent.

RI

SC

NU

MA

D

PT E

CE

412

PT

395

3.4 XOs synthesis from isolated hemicelluloses

414

In order to test the effectiveness of the isolated fractions on XOs synthesis, samples H1, H2,

415

and H3 were subsequently introduced to hydrolysis in the presence of dilute H2SO4 (4% wt.)

416

at 100 ºC for 60 min. Comparative results displayed that H1 gave better results over others by

417

achieving highest total XOs yield with DP up to 4 (Fig. 2), i.e., 14.3% xylose (X1), 12.2%

418

xylobiose (X2), 15.5% xylotriose (X3) and 31.7% wt. xylotetrose (X4) (dry wt. basis).

AC

413

20

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

419

420

MA

421

Fig. 2 Comparative result of XOs yield achieved during acid hydrolysis at 4% wt. H2SO4 and

423

121°C for 1 h using corncob hemicelluloses obtained after treatment with different alkaline

424

reagents (a) NaOH supplemented with NH4OH, (b) individual NaOH and (c) NH4OH. Error

425

bar indicates standard error.

PT E

CE

426

D

422

Whereas, others yielded relatively 34 to 37% lower XOs with maximum DP of 3. Although

428

xylopentose elution was spotted during chromatographic analysis, accurate estimation could

429

not be made due to the interference of other carbohydrate molecules based on HPLC and

430

TLC chromatograms, as presented in E-supplement data, respectively. Apparent results

431

established that residual impurities, including arabinose, uronic acid, and lignin greatly

432

influenced the XOs conversion. In a typical reaction, arabinose and lignin normally interfere

AC

427

21

ACCEPTED MANUSCRIPT 433

with acid hydrolysis reactions due to the ester bond similarities (Chemin et al., 2015).

434

Further, the isolated fractions were evaluated for XOs synthesis against enzyme catalysis

435

using commercial xylanase (endo-xylanase) under standard conditions. In resemblance to the

436

acid hydrolysis result, H1 pronounced remarkable differences, i.e., relatively 10% higher than

437

reference fractions (H2 and H3) (Fig. 3).

PT

438

PT E

D

MA

NU

SC

RI

439

441

CE

440

Fig. 3 Comparative XOs yield achieved through enzymatic with xylanase at 50°C up to 24 h

443

using corncob hemicelluloses obtained after alkaline treatment with different alkaline

444

reagents (a) NaOH supplemented with NH4OH, (b) individual NaOH and (c) NH4OH. Error

445

bar indicates standard error.

AC

442

446

22

ACCEPTED MANUSCRIPT However, the overall yield achievement through enzyme hydrolysis was ~3-fold lower than

448

the literature reports after prolonged hydrolysis (up to 24 h) might be caused due to higher

449

residual impurities in hemicellulose and lower enzyme loadings used. In support, several

450

detailed studies have postulated that branched substituents generally prevent the accessibility

451

of xylanase to approach xylan during hydrolysis, but some endoxylanases prefer hydrolysis

452

of more branched xylans (de Figueiredo et al., 2017). It is worthwhile that presence of side

453

groups exhibit a branched XOS, which largely offers diverse biological properties (Aachary &

454

Prapulla, 2011).

SC

RI

PT

447

The major disadvantages of the acidic hydrolysis of xylan are the generation of more

456

monosaccharide units and its derivative products preferentially furfural. Moreover, the

457

distribution of DPs of the XOs products is mainly dependent on the concentration of acid

458

and/or reaction time. Therefore, optimization of the processing conditions during acid

459

hydrolysis was conducted conventionally using the H1 hemicellulosic fraction. The effect of

460

hydrolysis time was studied between 15 and 120 min, while maintaining other parameters

461

constant (4% wt. acid concentration and 100°C) (Fig. 4). The apparent result showed that

462

significantly increased XOs conversion was achieved up to 60 min (~70% wt. total XOs) and

463

further extended reaction resulted in gradually decreasing yield (relatively 10% wt. during

464

each incremented intervals). This is caused due to enhanced cleavage of β-1,4-glycosidic

465

units in xylan that in turn led to yield more short polymeric units (DP<4) under higher

466

severity conditions referring acid concentration, time and temperature.

AC

CE

PT E

D

MA

NU

455

467

23

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

468

Fig. 4 Kinetics of XOs synthesis through acid hydrolysis using corncob hemicellulosic

470

fraction obtained after alkaline treatment using NaOH supplemented NH4OH reagent. Error

471

bar indicates standard error.

PT E

D

MA

469

472

Based on overall process material balance, 5.25% wt. total XOs production could be achieved

474

under the modest reaction conditions through two-stage processing approach using corncob

475

biomass (Fig. 5). However, the process setup enabled to achieve ~2-fold lower XOs yield

476

compared to literature reports involving expensive enzymes and harsh chemicals (NaOH,

477

KOH, H2SO4) under relatively higher severity conditions (Aachary & Prapulla, 2011;

478

Akpinar, Erdogan, Bakir, & Yilmaz, 2010; Akpinar et al., 2009b). This might be due to less

479

chance of the modified alkaline protocol for improved extraction of hemicellulose.

AC

CE

473

480 24

SC

RI

PT

ACCEPTED MANUSCRIPT

NU

481 482

Fig. 5 Overall material balance of XOs production from corncob biomass through sequential

484

isolation and hydrolysis.

MA

483

PT E

D

485

Overall, it can be noted that in contrary to the hypothesis that modified protocol helps in

487

achieving a fraction consisting minimal residual contaminations, which allows higher XOs

488

conversion. Yet, understanding the mechanism of effect of residual contaminations on XOs

489

synthesis would help in properly addressing the process challenges.

AC

490

CE

486

491

4. CONCLUSIONS

492

The present study demonstrated that modified alkaline protocol involving both NaOH and

493

NH4OH stimulates the cleavage of alkali-labile linkages between hemicellulose and other

494

associated constituents, depending on its performance nature. Therefore, it allowed for 25

ACCEPTED MANUSCRIPT enriched xylooligosaccharides synthesis through subsequent hydrolysis reaction in the

496

presence of dilute acid, thereby achieving maximum (73.68% XOs yield with DP up to 4).

497

Whereas, the individual protocols yielded relatively 34 to 37% lower XOs with maximum DP

498

of 3. The results suggested that hemicellulose residual impurities significantly influenced the

499

hydrolysis yield in the presence of acid or enzyme. The combined severity model fairly

500

represented the hemicellulose recovery data might be primarily due to the lignocellulose

501

heterogeneity and inability of the extended model because of the absence of essential other

502

independent variables. In complementary to the lower yield achievement via modified

503

approach, the protocol generates NH3 (gas) during alkali supplementation reaction, which

504

could help positively in the reduction of overall manufacturing cost of XOs through recovery

505

and reuse, in an economic point of view.

MA

506

NU

SC

RI

PT

495

Acknowledgements

508

The authors gratefully thank Department of Biotechnology (Government of India), New

509

Delhi, India for their consistent financial support. S. Elumalai thanks Department of Science

510

and Technology (DST-SERB), New Delhi for the financial support through Grant No.

511

YSS/2014/000031 (2015-18).

512

CE

PT E

D

507

References

514 515 516 517 518 519 520 521 522 523

Aachary, A. A., & Prapulla, S. G. (2011). Xylooligosaccharides (XOS) as an emerging prebiotic: microbial synthesis, utilization, structural characterization, bioactive properties, and applications. Comprehensive Reviews in Food Science and Food Safety, 10(1), 2-16. Akpinar, O., Erdogan, K., Bakir, U., & Yilmaz, L. (2010). Comparison of acid and enzymatic hydrolysis of tobacco stalk xylan for preparation of xylooligosaccharides. LWT-Food Science and Technology, 43(1), 119-125. Akpinar, O., Erdogan, K., & Bostanci, S. (2009a). Enzymatic production of Xylooligosaccharide from selected agricultural wastes. Food and Bioproducts Processing, 87(2), 145-151. doi:https://doi.org/10.1016/j.fbp.2008.09.002

AC

513

26

ACCEPTED MANUSCRIPT

CE

PT E

D

MA

NU

SC

RI

PT

Akpinar, O., Erdogan, K., & Bostanci, S. (2009b). Production of xylooligosaccharides by controlled acid hydrolysis of lignocellulosic materials. Carbohydrate Research, 344(5), 660-666. Bender, D., Nemeth, R., Wimmer, M., Götschhofer, S., Biolchi, M., Török, K., . . . Schoenlechner, R. (2017). Optimization of Arabinoxylan Isolation from Rye Bran by Adapting Extraction Solvent and Use of Enzymes. Journal of Food Science. Bian, J., Peng, P., Peng, F., Xiao, X., Xu, F., & Sun, R.-C. (2014). Microwave-assisted acid hydrolysis to produce xylooligosaccharides from sugarcane bagasse hemicelluloses. Food Chemistry, 156, 7-13. doi:https://doi.org/10.1016/j.foodchem.2014.01.112 Brienzo, M., Carvalho, W., & Milagres, A. M. (2010). Xylooligosaccharides production from alkali-pretreated sugarcane bagasse using xylanases from Thermoascus aurantiacus. Applied biochemistry and biotechnology, 162(4), 1195-1205. Carvalho, A. F. A., de Oliva Neto, P., Da Silva, D. F., & Pastore, G. M. (2013). Xylooligosaccharides from lignocellulosic materials: chemical structure, health benefits and production by chemical and enzymatic hydrolysis. Food Research International, 51(1), 75-85. Chapla, D., Pandit, P., & Shah, A. (2012). Production of xylooligosaccharides from corncob xylan by fungal xylanase and their utilization by probiotics. Bioresource Technology, 115, 215-221. Chemin, M., Wirotius, A.-L., Ham-Pichavant, F., Chollet, G., Perez, D. D. S., Petit-Conil, M., . . . Grelier, S. (2015). Well-defined oligosaccharides by mild acidic hydrolysis of hemicelluloses. European Polymer Journal, 66, 190-197. da Costa Sousa, L., Jin, M., Chundawat, S. P., Bokade, V., Tang, X., Azarpira, A., . . . Uppugundla, N. (2016). Next-generation ammonia pretreatment enhances cellulosic biofuel production. Energy & Environmental Science, 9(4), 1215-1223. da Silva, J. C., de Oliveira, R. C., da Silva Neto, A., Pimentel, V. C., & dos Santos, A. d. A. (2015). Extraction, Addition and Characterization of Hemicelluloses from Corn Cobs to Development of Paper Properties. Procedia Materials Science, 8, 793-801. de Figueiredo, F. C., Carvalho, A. F. A., Brienzo, M., Campioni, T. S., & de Oliva-Neto, P. (2017). Chemical input reduction in the arabinoxylan and lignocellulose alkaline extraction and xylooligosaccharides production. Bioresource Technology, 228, 164170. Elumalai, S., Agarwal, B., & Sangwan, R. S. (2016). Thermo-chemical pretreatment of rice straw for further processing for levulinic acid production. Bioresource Technology, 218, 232-246. Elumalai, S., Roa-Espinosa, A., Markley, J. L., & Runge, T. M. (2014). Combined sodium hydroxide and ammonium hydroxide pretreatment of post-biogas digestion dairy manure fiber for cost effective cellulosic bioethanol production. Sustainable Chemical Processes, 2(1), 12. Farhat, W., Venditti, R., Quick, A., Taha, M., Mignard, N., Becquart, F., & Ayoub, A. (2017). Hemicellulose extraction and characterization for applications in paper coatings and adhesives. Industrial Crops and Products, 107, 370-377. Gunawan, C., Xue, S., Pattathil, S., Sousa, L. C., Dale, B. E., & Balan, V. (2017). Comprehensive characterization of non-cellulosic recalcitrant cell wall carbohydrates in unhydrolyzed solids from AFEX-pretreated corn stover. Biotechnology for Biofuels, 10(1), 82.

AC

524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569

27

ACCEPTED MANUSCRIPT

CE

PT E

D

MA

NU

SC

RI

PT

Haghi, A., & Zaikov, G. (2014). Green nanofibers-production and limits. In C. W, G. G, & V. M. d. M. L (Eds.), Handbook of Research on Functional Materials: Principles, Capabilities and Limitations (pp. 11-94). Canada: CRC Press. Huang, C., Lai, C., Wu, X., Huang, Y., He, J., Huang, C., . . . Yong, Q. (2017). An integrated process to produce bio-ethanol and xylooligosaccharides rich in xylobiose and xylotriose from high ash content waste wheat straw. Bioresource Technology, 241, 228-235. doi:https://doi.org/10.1016/j.biortech.2017.05.109 Li, H.-Y., Sun, S.-N., Zhou, X., Peng, F., & Sun, R.-C. (2015). Structural characterization of hemicelluloses and topochemical changes in Eucalyptus cell wall during alkali ethanol treatment. Carbohydrate Polymers, 123, 17-26. Li, H., Li, Z., Peng, P., She, D., Xu, Q., & Zhang, X. (2015). Characteristics of hemicelluloses obtained from sweet sorghum based on successive extractions. Journal of Applied Polymer Science, 132(46). Luo, Q., Peng, H., Zhou, M., Lin, D., Ruan, R., Wan, Y., . . . Liu, Y. (2012). Alkali extraction and physicochemical characterization of hemicelluloses from young bamboo (Phyllostachys pubescens Mazel). BioResources, 7(4), 5817-5828. Ma, L., Du, L., Cui, Y., Song, P., Jiang, F., Ma, Q., & Xiao, D. (2016). Isolation and structural analysis of hemicellulose from corncobs after a delignification pretreatment. Analytical Methods, 8(41), 7500-7506. Michelin, M., Ruiz, H. A., Maria de Lourdes, T., & Teixeira, J. A. (2018). Multi-step approach to add value to corncob: Production of biomass-degrading enzymes, lignin and fermentable sugars. Bioresource Technology, 247, 582-590. Qing, Q., Li, H., Kumar, R., & Wyman, C. E. (2013). Xylooligosaccharides production, quantification, and characterization in context of lignocellulosic biomass pretreatment. In Aqueous pretreatment of plant biomass for biological and chemical conversion to fuels and chemicals (pp. 391-415). Silverstein, R. A., Chen, Y., Sharma-Shivappa, R. R., Boyette, M. D., & Osborne, J. (2007). A comparison of chemical pretreatment methods for improving saccharification of cotton stalks. Bioresource Technology, 98(16), 3000-3011. Sluiter, J. B., Ruiz, R. O., Scarlata, C. J., Sluiter, A. D., & Templeton, D. W. (2010). Compositional analysis of lignocellulosic feedstocks. 1. Review and description of methods. Journal of agricultural and food chemistry, 58(16), 9043-9053. Sun, R., Lawther, J. M., & Banks, W. (1995). Influence of alkaline pre-treatments on the cell wall components of wheat straw. Industrial Crops and Products, 4(2), 127-145. Sun, S.-L., Wen, J.-L., Ma, M.-G., & Sun, R.-C. (2013). Successive alkali extraction and structural characterization of hemicelluloses from sweet sorghum stem. Carbohydrate Polymers, 92(2), 2224-2231. Surek, E., & Buyukkileci, A. O. (2017). Production of xylooligosaccharides by autohydrolysis of hazelnut (Corylus avellana L.) shell. Carbohydrate Polymers, 174, 565-571. Tobimatsu, Y., Elumalai, S., Grabber, J. H., Davidson, C. L., Pan, X., & Ralph, J. (2012). Hydroxycinnamate conjugates as potential monolignol replacements: in vitro lignification and cell wall studies with rosmarinic acid. ChemSusChem, 5(4), 676-686. Vazquez, M., Alonso, J., Domınguez, H., & Parajo, J. (2000). Xylooligosaccharides: manufacture and applications. Trends in Food Science & Technology, 11(11), 387393.

AC

570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615

28

ACCEPTED MANUSCRIPT 616 617 618 619 620 621

Wang, S., Ru, B., Dai, G., Sun, W., Qiu, K., & Zhou, J. (2015). Pyrolysis mechanism study of minimally damaged hemicellulose polymers isolated from agricultural waste straw samples. Bioresource Technology, 190, 211-218. Yuan, Q., Zhang, H., Qian, Z., & Yang, X. (2004). Pilot‐ plant production of xylo‐ oligosaccharides from corncob by steaming, enzymatic hydrolysis and nanofiltration. Journal of Chemical Technology and Biotechnology, 79(10), 1073-1079.

622

624

Highlights:

625 626 627 628

   

PT

623

AC

CE

PT E

D

MA

NU

SC

RI

Modified alkaline protocol promoted cleavage of ester and ether linked constituents. The protocol permitted achieving enriched XOs through acid hydrolysis. The setup demonstrated a cost-effective process through recovery of reuse of alkali. A net ~5% wt. XOs (X1-X4) can be obtained through the modified approach.

29