Effects of combined pretreatment with rod-milled and torrefaction on physicochemical and fuel characteristics of wheat straw

Accepted Manuscript Effects of combined pretreatment with rod-milled and torrefaction on physicochemical and fuel characteristics of wheat straw Xiaopeng Bai, Guanghui Wang, Yue Sun, Yan Yu, Jude Liu, Decheng Wang, Zhiqin Wang PII: DOI: Reference:

S0960-8524(18)30919-2 https://doi.org/10.1016/j.biortech.2018.07.022 BITE 20152

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

26 June 2018 4 July 2018 5 July 2018

Please cite this article as: Bai, X., Wang, G., Sun, Y., Yu, Y., Liu, J., Wang, D., Wang, Z., Effects of combined pretreatment with rod-milled and torrefaction on physicochemical and fuel characteristics of wheat straw, Bioresource Technology (2018), doi: https://doi.org/10.1016/j.biortech.2018.07.022

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.

1

Effects of combined pretreatment with rod-milled and

2

torrefaction on physicochemical and fuel characteristics of

3

wheat straw

4 5

Xiaopeng Baia, Guanghui Wanga,*, Yue Suna, Yan Yua, Jude Liub, Decheng Wanga,

6

Zhiqin Wanga

7 8 9

a

Department of Agricultural Engineering, College of Engineering, China Agricultural

10

University, Beijing 100083, China

11

b

12

University Park, PA 16802, USA

13

Email: [email protected] (XB)

Department of Agricultural and Biological Engineering, Pennsylvania State University,

14 15

*Corresponding author:

16

Guanghui Wang

17

Department of Agricultural Engineering, College of Engineering, China Agricultural

18

University, Beijing 100083, China

19

Tel: +86 10 6273 7845; Fax: +86 10 6273 7845; Email: [email protected]

1

20

Abstract:

21

The mechanism of rod-milling combined with torrefaction as well as its effects on

22

physicochemical and fuel properties of wheat straw were investigated. Rod-milling and

23

hammer-milling samples were torrefied under three temperatures (250, 275, and 300 ℃)

24

with a duration time of 30 min. The results indicated that combined rod-milling and

25

torrefaction pretreatment (CRT) significantly elevated carbon content, higher heating

26

value, fuel ratio, and reduced oxygen content and

27

straw. Moreover, CRT significantly reduced cellulose crystallinity, and increased the

28

specific surface area and pore volume of wheat straw, which lowered the wheat straw’s

29

degrading pyrolysis temperature. These peak values appeared under 300 °C.

30

Devolatilization index (Di) was improved by rod-milling pretreatment under identical

31

torrefaction conditions except 275 °C. Therefore, the combination of rod-milling with

32

torrefaction under 300 °C has the advantage of enhancing fuel properties of

33

lignocellulosic biomass materials.

in wheat

34 35

Keywords: Fuel properties, Rod-milling, Torrefaction, Fuel ratio, Devolatilization index

2

36

1.Introduction

37

A massive growth of global energy demand over the last several decades has

38

caused diminishing fossil fuel reserves, serious environmental pollution and high

39

greenhouse gas emission (Chen et al., 2015). To overcome the potential energy crisis

40

and mitigate dependence on fossil fuels, developing and utilizing biomass energy has

41

become a trend worldwide. Straw, as a valuable and abundant renewable resource, could

42

be utilized as a solid fuel for potential clean energy production (Bai et al., 2018).

43

As an important renewable resource in the future, biomass could be converted into

44

high value-added fuels through various technologies, such as pyrolysis,

45

transesterification, fermentation, and saccharification etc. (Erlich and Fransson, 2011).

46

However, the low qualities of raw biomass feedstocks, including high oxygen content,

47

higher water content, hydrophilicity, low energy density and structural heterogeneity,

48

limited the energy utilization of biomass. Thus, the pretreatment for biomass feedstocks

49

was usually necessary to improve their qualities for efficient energy conversion. There

50

were several types of pretreatments including physical, thermal and chemical

51

pretreatments (Kan et al., 2016). Torrefaction is a pretreatment, which is usually

52

selected to convert raw biomass feedstocks to high-valued solid fuel (Chew et al., 2011).

53

Torrefaction is a moderate thermal treatment with a reaction temperature between 200 to

54

300 ℃ in the absence of oxygen atmosphere. During torrefaction, the water and light

55

volatile components were fully removed from raw biomass materials and the oxygen

56

content in biomass was partially reduced. Torrefaction pretreatment can break down the

57

fiber structure, increase energy density, enhance hydrophobicity, decrease atomic O/C 3

58

and H/C ratios, and improve the grind-ability and reactivity (Shang et al., 2012).

59

Torrefied biomass was more reactive than raw biomass, and was generally used as

60

feedstock for combustion or gasification (Chen et al., 2012).

61

However, the conversion efficiency of lignocellulosic biomass during torrefaction

62

was affected by the fiber structure and chemical composition. The cellulose network was

63

generally embedded in hemicelluloses and lignin matrix, which significantly hindered

64

commercial bioconversion of lignocellulose materials (Han et al., 2007). For chemical

65

composition, thermal stability of hemicellulose, cellulose, and lignin in lignocellulosic

66

biomass had great effects on torrefaction process (Gong et al., 2016). Therefore, many

67

studies on combination of pretreatments have been done to improve the efficiency of

68

torrefaction in the literature. The different dehydrations combined with torrefaction

69

pretreatments were studied in our previous research (Yu et al. 2018), where fuel

70

characteristics were improved significantly with sun-cured and 300 ℃ torrefaction

71

pretreatment. Combining water washing and torrefaction could significantly increase the

72

content of laevoglucose (Zhang et al., 2016). Ukaew et al. (2018) found that combined

73

acid and torrefaction pretreatment could also improve fuel characteristics. However, the

74

influence of water washing pretreatment on the quality of biomass was limited due to

75

the chemical structure of biomass was not affected by washing (Chen et al., 2017).

76

Moreover, since the associated high costs, toxicity, corrosive nature, and environmental

77

pollution issues, chemical pretreatment methods using strong acids, alkaline and ionic

78

liquids were very limited (Bai et al., 2018). On the contrary, mechanical pretreatment is

79

an environmentally friendly technique and has become an essential part of the 4

80

bioconversion of lignocellulosic biomass. As an example of mechanical pretreatment,

81

rod-milling could reduce the particle size and cellulose crystallinity, increase the special

82

surface area (SSA) and pore volume (PV) of wheat straw, which could greatly increase

83

pyrolysis efficiency (Bai et al., 2018). Moreover, the ball milling pretreatment could

84

enhance enzymatic hydrolysis of biomass because the crystalline structure of cellulose

85

was destroyed to form amorphous cellulose (Wang et al., 2013; Ji et al., 2016). More

86

valuable chemical resources were obtained undergoing ball milling pretreatment (Khan

87

et al., 2015). In addition, the higher heating value (HHV) of torrefied pine sawdust was

88

enhanced after ball milling (Gong et al., 2016). Although many torrefaction and

89

mechanical pretreatments of biomass feedstocks have been studied, a combination of

90

rod-milling and torrefaction for biomass treating to change fuel properties has rarely

91

been focused on.

92

No reports and published literature were found on the studies of the effects of

93

combination torrefaction and rod-milling pretreatment on physicochemical and fuel

94

characteristics of wheat straw. The objectives of this study were to: 1) analyze the

95

underlying mechanism and understand how the combined with rod-milling and

96

torrefaction influence physiochemical characteristics of wheat straw; and 2) investigate

97

the effect of combination rod-milling and torrefaction on the fuel characteristics of

98

wheat straw based on the alteration of physicochemical properties of the wheat straw.

99 100

2. Materials and methods

101

2.1 Raw materials 5

102

Wheat straw was obtained from Gu’an County, Hebei Province (China) in June,

103

2017. The initial moisture content of WS was 6.46 wt.%. The same methods with our

104

pervious study were used to prepare the hammer-milled wheat straw (HWS) and the

105

rod-milled wheat straw (RWS) samples (Bai et al., 2018). The mean sizes of raw HWS

106

and RWS were 480.7 ± 19.5 μm and 20.2 ± 0.3 μm. The detailed measurement

107

procedures were reported in our previous work (Bai et al., 2018). After milling, all

108

samples were dried at 40 °C for 48 h and then stored in the zip-lock plastic bags for

109

subsequent experiments.

110

2.2 Torrefaction experiments

111

A fixed-bed tubular reactor (SK-G08123K; Tianjin Zhonghuan Experimental

112

Furnace Co. Ltd., China) was used to perform the experiment of torrefaction. The set

113

torrefaction conditions for the samples of HWS and RWS were 250, 275, and 300 ℃,

114

respectively, with duration time of 30 min. After being torrefied, the HWS and RWS

115

were labeled as THWSX and TRWSX, with the value of “X” indicating the torrefaction

116

temperature. The energy density was calculated by the mass yield and the energy yield

117

obtained by weighing before and after samples torrefaction.

118

2.3. Physicochemical analysis

119

An Elementar analyzer (Vario EL cube, Germany) was used to measure the carbon

120

(C), hydrogen (H), nitrogen (N), and sulfur contents (S) of samples. The oxygen (O)

121

content was calculated by difference. The volatile matter (VM) and ash (Ash) content

122

were determined according to the ASTM D3175-89 and ASTM D3174-04, while the

123

fixed carbon (FC) was calculated by difference. The Fuel ratios (FR) of samples were 6

124

attained from the FC analysis and VM analysis and then calculated by Eq.(1) (Huang et

125

al., 2017). The higher heating value (HHV) of all samples was obtained according to

126

ultimate analysis using Eq. (2) (Friedl et al., 2005).

127

FR  fixed carbon volatile matter

(1)

128

HHV  3.55  C 2  232  C  2230  H  51.2  C  H  131 N  20600, kJ/ kg

(2)

129 130

2.4 Specific surface area analysis Specific surface area (SSA) experiments of samples were conducted using the

131

Quadrasorb S1 automated surface area and pore size analyzer (Quantachrome

132

Instruments Co., Ltd,, Florida, USA). For each test, about 100 mg of sample was

133

degassed at 110 °C for 8 h. During the analysis, the nitrogen adsorption and desorption

134

of samples were carried out under 77 K liquid nitrogen environment. The relationship

135

between the equilibrium adsorption pressure and the amount of adsorbed gas was

136

determined. The SSA of samples was obtained by fitting the adsorption curve according

137

to the BET theory.

138

2.5 Particle morphology analysis

139

The morphology of all dried samples was evaluated by scanning electron

140

microscopy (SEM). These samples’ surfaces were sputter-coated with Pt for 5 min using

141

a sputter coater (JFC-1600, JEOL, Akishima, Japan). A scanning electron microscope

142

(JSM-6700F, JEOL) operated at 10 kV was used to obtain the electron micrographs.

143

2.6 X-ray powder diffraction (XRD) analysis

144 145

X-ray diffractograms of samples were obtained by an XD3 X-ray diffractometer (Purkinje General Instrument Co., Ltd, Beijing, China), using Cu Kα radiation at 40 kV 7

146

and 40 mA in the scanning range of 5–40° at a rate of 2 °/min with a increment of 0.2°.

147

Each test was repeated 3 times. Crystallinity index (CrI) were calculated according to

148

the Segal’s method (Segal et al., 1959) using following Eq.(3).

149

CrI % = (I002-Iam)/I002×100 %

(3)

150

where I002 and Iam are the intensities of the diffraction of 002 peak at 2θ ≈ 22.7° and the

151

intensity of amorphous at 2θ ≈ 18°.

152

2.7 Thermogravimetric analysis

153

The Thermogravimetric (TG) analysis was performed on a DTG-60 (Shimadzu,

154

Kanagawa, Japan) thermogravimetric analyzer. Each sample of about 5 mg was heated

155

from 40 ℃ to 800 ℃ at a heating rates 20 ℃/min under nitrogen flow rate of 100 mL/min

156

and held for about 5 min when reached 105 °C. To quantify the effects of combined

157

pretreatment with rod-milled and torrefaction on the performance of VM during

158

pyrolysis process, the devolatilization index (Di) of samples (Wu et al., 2014) was

159

defined, as follow:

160

Di  Rmax (TinTmax T1/2 )

(4)

161

where Rmax is the maximum decomposition rate, Tin is the initial devolatilization

162

temperature and it is corresponding to a weight loss of 5% respect to the final weight

163

loss. Tmax is the maximum mass loss temperature.  T1/2 is the temperature interval when

164

Rd/Rmax equals to 1/2. Rd is the decomposition rate, Rd = dmt/dt, where mt is the mass

165

loss of the raw coal sample at time t. Rmax and Rd can be obtained from the derivative

166

thermogravimetric curve.

167 8

168

3. Results and discussion

169

3.1 Compositional analysis

170

3.1.1 Ultimate analysis

171

Table1 presented the results of the ultimate analysis of raw and torrefied wheat

172

straw samples. The carbon content increased from 44.00 wt.% to 56.94 wt.% for HWS,

173

and from 43.38 wt.% to 64.95 wt.% for RWS after torrefaction. Meanwhile, the

174

hydrogen and oxygen contents decreased from 5.76 wt.% and 48.92 wt.% to 5.04 wt.%

175

and 36.23 wt.%, respectively, for HWS. For RWS samples, hydrogen and oxygen

176

contents declined from 5.59 wt.% and 49.66 wt.% to 3.39 wt.% and 30.36 wt.%,

177

respectively,. Nitrogen and sulfur contents changed slightly during torrefaction due to

178

their very low contents in wheat straw. The reduction of hydrogen and oxygen contents

179

with the increase of the torrefaction temperature was mainly attributed to the

180

decomposition of carbohydrate, and dehydration and decarboxylation reactions from

181

volatile components (Yue et al., 2017; Chiou et al., 2015). These reactions resulted in

182

losing water and releasing gases and light volatiles. Interestingly, more pronounced

183

increases of carbon content and the reduction of hydrogen and oxygen contents were

184

observed from the torrefied RWS. This indicated that the components of RWS were

185

more easily degradable during torrefaction. This phenomenon could be explained that

186

the structure of biomass was changed (Chen et al., 2017). It has been observed in our

187

previous work that rod-milling appreciably reduced the particle size and cellulose

188

crystallinity, and increased the SSA and pore volume (PV) of wheat straw (Bai et al.,

189

2018). Under identical conditions, that would cause greater degrees of decarboxylation, 9

190

dehydration, and depolymerization of the organic portion of the wheat straw after

191

rod-milling.

192

As shown in Table 1, the reduction of atomic O/C and H/C ratios was consistent to

193

the increment of HHV. This means that the dehydration and decarbonization increased

194

the HHV of raw and torrefied samples. The contents of C-C and C-H bonds having

195

higher energy were accumulated after removal of C-O bond as the torrefaction

196

temperature increased (Poddar et al., 2014). This also can be confirmed by our previous

197

results (Bai et al., 2017), which also indicated that the C–O bond diminished if the

198

torrefaction temperature increased. Moreover, the HHV of HWS was enlarged from

199

17.52 to 22.52 MJ/kg with the increase of torrefaction severity, while that of RWS was

200

raised from 17.29 to 24.32 MJ/kg under the same condition. For others pretreatment, the

201

HHV of bamboo sawdust samples was enhanced by 18.4% after undergoing wet

202

torrefaction pretreatment (Wang et al,.2018). The calorific value of biomass was

203

increased after steam treatment, and the spruce had the highest increase of 26% in

204

calorific value (Tooyserkani et al., 2013). In this paper, a 40.66% increase in HHV of

205

RWS was found with increasing torrefaction temperature. In addition, the HHV of RWS

206

was 12.12% higher than that of HWS under 300 ℃, and the HHV of TRWS300 reached

207

to 24.32 MJ/kg, which was very close to the bituminous coal’s, 25-35 MJ/kg (Chen et

208

al., 2015).Therefore, biomass pretreated by combining rod-milling and torrefaction has

209

potential for use as an alternative fuel to coal.

210 211

Fig.1 illustrated the Van Krevelen diagram of raw and torrefied samples, which can be found the changes in elemental compositions of samples. Interestingly, the atomic 10

212

H/C value showed a good linear curve correlation with atomic O/C. The relationship

213

between H/C and O/C could be described as H/C=1.397*O/C+0.375, R2=0.973 for

214

HWS, and H/C=1.778*O/C+0.017, R2=0.998 for RWS. Atomic H/C and O/C ratios of

215

raw and torrefied samples decreased as the torrefaction temperature increased. For HWS,

216

the atomic H/C ratio declined from 1.57 to 1.06, and O/C ratio dropped from 0.83 to

217

0.48. For RWS, the atomic H/C ratio declined from 1.48 to 0.63, and atomic O/C ratio

218

reduced from 0.84 to 0.35, respectively. The moisture and light volatiles that contain

219

more hydrogen and oxygen were removed after undergoing torrefaction, which led to

220

relatively more carbon retained. Interestingly, the slope of H/C of RWS declination was

221

larger than that of HWS with the increase of torrefaction temperature. Moreover, it was

222

clearly found that the TRWSX had generally lower atomic H/C and O/C ratios than

223

those of THWSX. This phenomenon indicated that RWS required relatively lower

224

temperatures in torrefaction process to produce better solid fuels.

225

3.1.2 Proximate analysis and fuel ratio

226

The results of proximate analysis and fuel ratio (FR) of raw and torrefied wheat

227

straw were shown in Table 2. The contents of VM decreased with an increase in

228

torrefaction temperature for all samples, while high FC content appeared in torrefied

229

samples. It was likely that the moisture and light volatiles were released from the

230

materials during torrefaction (Chew et al., 2011). Among chemical composition of

231

wheat straw, lignin is the most difficult constitute to be thermally degraded. Therefore,

232

lignin content increased relatively as the decomposition of hemicellulose and partial

233

depolymerization of cellulose and lignin during torrefaction (Zhang et al., 2015), which 11

234

caused the FC of all samples increased and the VM of all samples decreased as the

235

torrefaction temperature increased. Furthermore, TRWSX showed a faster growth rate

236

in FC content and degraded rate in VM content than THWSX. It indicated that

237

hemicellulose and part of cellulose in RWS were more degradable than HWS during

238

torrefaction. It could be confirmed by the results of CrI. In section 3.4, it showed the

239

faster decline rate in CrI of TRWSX than that of THWSX. Meanwhile, it can be seen

240

from Table 1 and Table 2 that the trends of increasing FC and HHV are consistent with

241

the increase in temperature. This was confirmed by Du et al., (2014) that the FC could

242

hold higher energy. Therefore, RWS requires lower torrefaction temperature to produce

243

the higher quality solid fuel.

244

In addition, a significant upward trend was observed in the Ash content of torrefied

245

samples when increasing torrefaction temperatures. The RWS had higher Ash contents

246

than the HWS. The Ash content of RWS was almost four times higher than that of HWS.

247

That could be explained by the abrasion of the ultrafine vibration rod mill during

248

rod-milling, which led to inorganic substances being mixed into the wheat straw. The

249

increase in Ash content of RWS also resulted in lower carbon content and FC. A

250

significant upward trend was observed in the FR of HWS and RWS during torrefaction,

251

which due to VM was liberated from the samples and FC content increased gradually.

252

The FR was increased from 0.25 to 0.98 for HWS and from 0.23 to 1.56 for RWS at

253

severe condition (300 ℃). It was observed clearly that TRWSX showed a faster growth

254

rate and higher value in FR than THWSX. FR was a very important index and generally

255

used to evaluate the materials in combustion process. It was demonstrated that higher 12

256

FR could generate much less emissions of CO2 in the process of combustion (Granados

257

et al., 2017). It can be concluded that the FR of wheat straw was markedly improved

258

after CRT. Therefore, CRT as a pretreatment has been encouraged to produce cleaner

259

and greener biomass fuel to substitute coal.

260

3.2 Mass yield and energy yield

261

The results of mass yield, energy yield and energy density of raw and torrefied

262

samples were shown in Table 3. The values of mass yield and energy yield all decreased

263

with increase in torrefaction temperatures, while the energy density increased. As the

264

torrefaction temperature increased, the reduction of mass yield mainly attributed to

265

devolatilization and thermal cracking (Crawford et al., 2016). It can be seen from Table

266

3 that the decrease rate of TRWSX’ mass yield remained almost stable, whereas

267

THWSX mass loss rates increased significantly from 275 to 300 ℃. This indicated that

268

RWS had more stable thermal degradation behavior than HWS. Similar changes of mass

269

yield analysis results were also found in our previous studies of torrefied pine sawdust

270

(Gong et al., 2016). Furthermore, it was clearly seen that TRWSX had a lower mass

271

yield than THWSX under the same torrefaction conditions. Therefore, RWS produced

272

more easily degradable components, and contributed to more uniform thermal

273

properties.

274

The energy yield is mainly determined by mass yield and HHV of materials.

275

Although, the HHV of torrefied samples was enhanced with increasing torrefaction

276

temperature, the declining slope of mass loss was greater than the HHV increasing slope.

277

As the torrefaction temperature increased from 250 to 300 ℃, the energy yield declined 13

278

from 88.14% to 63.92% for HWS and from 80.80% to 64.91% for RWS, respectively.

279

Although the HHVs of torrefied samples were enhanced, a large amount of energy was

280

lost due to the loss of sample mass under severe torrefaction temperature. This caused

281

the energy yield to decrease largely. However, the energy yield of RWS was higher than

282

that of HWS. Moreover, with rising torrefaction temperature, the decreased rates of

283

energy yield were lower than those of mass yield. This led to the increased energy

284

density with increasing the torrefaction temperature. The energy density of THWS was

285

enlarged from 1.13 to 1.29 at 300 ℃, while that of TRWS was raised from 1.09 to 1.39

286

under the same condition. The energy density of TRWS300 was enhanced 8% more

287

than THWS300’s. Therefore, the CRT as a pretreatment is recommended to upgrade fuel

288

characteristics of wheat straw for producing renewable biofuels.

289

3.3 SSA and PV analysis

290

The results of SSA and PV of raw and torrefied samples were shown in Table 3.

291

The SSAs of HWS, THWS250, THWS275, and THWS300 were 1.74, 0.94, 1.67, and

292

1.87 m2/g, respectively. The trend decreased firstly in the SSA as the torrefaction

293

temperature reached to 250 ℃ and then increased with torrefaction temperature rising

294

from 250 to 300 ℃. The PVs showed similar trends to the SSAs. This phenomenon was

295

attributed to the fact that the pore of wheat straw was blocked by the tar at torrefaction

296

temperature of 250 ℃ (Chen et al., 2017). This also can be confirmed by SEM

297

(THWS250) image in supplementary data that some pores were blind and closed. That

298

led to the lower SSA and PV for torrefied wheat straw than HWS. However, the rapid

299

release of VM could open and link the blind and closed pores with the increase of 14

300

torrefaction severity, and create new cracks, micropores and mesopores, which led to a

301

significant increase of SSAs (Xu et al., 2010 ; Chen et al., 2017). This was consistent

302

with our previous report (Bai et al., 2017), where results were found that higher

303

torrefaction temperature caused a highly porous structure. Moreover, Wang et al. (2017)

304

investigated that the ground stem wood particles were more porous with massive pores

305

and some open tubular structure after torrefaction at 300 ℃. It was confirmed by SEM

306

(THWS300) images in supplementary data that surface of THWS300 had much more

307

smooth and clean surfaces than HWS, and it had porous structure with many pores and

308

tubular openings.

309

For RWS, the SSA and PV were significantly increased after rod-milling. And their

310

values were 2.86 m2/g and 1.19 cm3•10-2/g, respectively. The SSA of particle contains

311

the outside surface area and the inner pore surface area (Gao et al., 2017). The inner

312

pore surface area is much larger than the outside surface area (Liu et al., 2015). This

313

suggested that the porous structure of wheat straw was disrupted, and the inner pore

314

surface area exposed to the surface after rod-milling, which caused the SSAs and PVs of

315

RWS to increase. Moreover, rod-milling dramatically decrease the particle size, and it

316

increased the SSAs of particles (Ji et al., 2016). The thermal stability of hemicellulose,

317

cellulose, and lignin were reduced by rod-milling (Bai et al., 2018), and leading to much

318

easier to devolatilization, depolymerization, and carbonization reactions of

319

hemicellulose, cellulose and lignin under the same temperature conditions. The

320

TRWSX’s SSAs and PVs also showed the same trends to THWSXs’ after torrefaction

321

pretreatment. However, the SSAs and PVs of TRWS250 were relatively lower than those 15

322

of THWS250, which could be interpreted that the interparticle space reduced after

323

rod-milling, and it impeded seriously the escape of volatiles matters from particle

324

surface to the external gas phase during torrefaction temperature of 250 ℃ (Tian et al.,

325

2016).The SEM (TRWS250) images in supplementary data confirmed this phenomenon

326

that the small particles were glued together by the tar to form bigger particles during the

327

torrefaction process. With increasing torrefaction severity, the blind and closed pores

328

opened largely by the more faster release of VM, and causing the increases of SSAs and

329

PVs, especially for TRWS300. It can be seen from SEM (TRWS275 and TRWS300)

330

images supplementary data that the particles that stuck together were destroyed, the

331

particle size became smaller, the structure was looser, and the porosity was larger.

332

Granados et al. (2017) observed that the pores, cracks and crater were directly related to

333

volatile material release, and it was enhanced with torrefaction temperature. This was

334

consistent with the results of section 3.1.2 about proximate analysis. It showed that VM

335

declined as the torrefaction temperature increased, and the VM of RWS had lower

336

content than HWS under same torrefaction temperature conditions.

337

3.4 XRD analysis

338

-

Five major peaks (101, 101, 021, 002, and 040) characteristics of crystalline

339

cellulose and overlapped shallow peak assigned to the amorphous contribution were

340

observed (Supplementary Data). The intensity of five peaks assigned to crystalline

341

cellulose was reduced as an increase of the torrefaction temperature for HWS and RWS.

342

This indicated that the cellulose crystalline structure was destroyed gradually because

343

the hydrogen bonds of cellulose chains were severely damaged by torrefaction (Gong et 16

344

al., 2016; Yu et al., 2018). However, the peak of amorphous cellulose narrowed down

345

first when the temperature reached to 250 ℃ and then increased with the torrefaction

346

temperature rising. It suggested that the contents of amorphous cellulose decreased

347

firstly and then increased during torrefaction. In addition, the 021 peak in raw HWS was

348

not obvious, while RWS showed a clearly 021 peak. The 002 peak of RWS became

349

disappeared, which could be explained that it was covered by broad amorphous

350

cellulose. This indicated that rod-milling pretreatment immensely disrupted the

351

cellulose crystalline structure (Bai et al., 2018).

352

The cellulose crystallinity indices (CrI) values for raw and torrefied samples under

353

different conditions were presented in Table 3. For all samples, the CrI of WS increased

354

from 51.33% to 54.61% for HWS, from 11.59% to 24.12% for RWS, respectively, when

355

the torrefaction temperature was 250 ℃.Then decreased sharply from 54.61% to 18.87%

356

for HWS, from 24.12% to 2.56% for RWS, respectively, as the torrefaction temperature

357

further increased from 275 to 300 ℃. The changes of CrI resulted from the competitive

358

decomposition of the amorphous cellulose and the crystalline cellulose with the increase

359

of the temperature (Wang et al., 2017). This was consistent with the results of the

360

changing of the peaks of crystallinity and amorphous cellulose. When the torrefaction

361

temperature was 250 ℃, the CrI remain increased because of the recrystallization of

362

amorphous cellulose (Wen et al., 2014). Moreover, that also could be explained that the

363

amorphous cellulose was degraded, and the decomposition of crystalline cellulose had

364

not yet begun when torrefaction was 250 ℃ (Bai et al., 2018). These observations were

365

consistent with the results of Basch et al.(1973) who reported that the decomposition of 17

366

the amorphous cellulose was dominant when the torrefaction temperature was 250 ℃.

367

Interestingly, the change in Crl for TRWS250 during torrefaction was much larger

368

intense than THWS250. This indicated that the degree of recrystallization of amorphous

369

cellulose was enhanced largely. When it exceeded 250 ℃, the conversion of crystalline

370

cellulose into amorphous form was faster than the degradation of amorphous cellulose.

371

That led to the decrease of CrI, especially after torrefaction at 300 °C. This meant the

372

destructions of crystalline cellulose in THWS300 and TRWS300 were the severest. This

373

could be confirmed by the results of SSA, PV and the SEM images. Moreover, TRWSX

374

showed a faster rate of descent in CrI content than THWSX, indicating that wheat straw

375

after rod-milling pretreatment requires lower torrefaction temperature to destroy largely

376

the cellulose structure. That attributed to the thermal stability of crystalline cellulose

377

reduced largely, and it was easier to degrade under the same torrefaction conditions.

378

Thus, the CrI of TRWS300 was the lowest. It can be concluded that CRT as a

379

pretreatment for biomass markedly reduced recalcitrance of lignocellulose biomass to

380

degradation.

381

3.5 TG analysis

382

The differential thermogravimetric (DTG) results of raw and torrefied samples was

383

shown in Fig.2. For raw and torrefied samples, the thermal decomposition process was

384

divided into three distinguished stages according to the weight loss rate (Zhang et al.,

385

2016). The characteristics parameters of thermal degradation for the raw and torrefied

386

samples were presented in Table 4. In Fig.2, the initial slightly decline in weight is due

387

to the evaporation of water and degradation of small organic compounds in wheat straw 18

388

where temperature is below 150 ℃. The percent weight loss of HWS was lower than

389

RWS samples, which mainly because of more water absorbed to the RWS surface than

390

HWS’. This was confirmed by the results of section 3.3 in HWS and RWS samples. It

391

shows that RWS had the greater SSA and PV than HWS samples. Moreover, RWS

392

existed stronger intensity of polar interactions and absorption of oxygen containing

393

groups than HWS, which led to the weight loss of RWS higher than HWS samples (Bai

394

et al.2018). There we found that the intensity of O-H of RWS was enhanced after

395

rod-milling pretreatment.

396

The second stage ranged from 150-500 ℃ was mainly devolatilization process, and

397

it contained the decomposition of hemicellulose, cellulose, and lignin (Zhang et al.,

398

2016). In general, thermal decomposition temperature in range of 150-350 ℃ for

399

hemicellulose, 275-350 ℃ for cellulose, and a broad range of 250-500 ℃ for lignin

400

(Chen et al., 2010). At this stage, it can be clearly seen that the significant differences of

401

DTG curves for samples under different torrefaction conditions. In Fig.2 (a), the peak of

402

hemicellulose degradation was not obvious because of the hemicellulose curve merged

403

with that of cellulose. This was associated with its low content of hemicellulose (17.52

404

wt.%) (Chang et al., 2012). With an increase in torrefaction temperature, the decline in

405

hemicellulose amplitude could be attributed to a decrease in its mass fraction, which

406

was caused by the fact of the lower content of hemicellulose after torrefaction. The

407

degradation of hemicellulose of HWS occurred earlier than THWSX because of the

408

HWS with the high VM (Yu et al., 2018). In Fig.2 (b), the broader hemicellulose region

409

of RWS than HWS were observed, which mean that the thermal degradation of 19

410

hemicellulose occurred more easily during pyrolysis after rod-milling pretreatment. This

411

was consistent with the results of Table 4. It presented the initial temperature of

412

devolatilization (Ti) in HWS at 254.86 ℃, while the Ti in RWS shifted to a lower

413

temperature at 218.41 ℃. This was interpreted by the fact that the structure of the

414

hemicellulose was destroyed led to a lower thermal stability under rod-milling

415

pretreatment (Bai et al., 2018). Therefore, the Tis of TRWSX was lower than those of

416

THWSX (Table 4).

417

The peak with the maximum mass loss temperature represented the thermal

418

decomposition of cellulose for HWS samples. The peak of cellulose degradation slightly

419

shifted to the left (Tmax: from 341.31 to 339.52 °C) as the torrefaction temperature

420

ranged from 0 to 275 ℃, suggesting torrefaction pretreatment decreased degradation

421

temperature of cellulose in samples. As mentioned in XRD analysis, the suitable

422

torrefaction destroyed the crystalline structure of cellulose, which decreased the thermal

423

stability of cellulose. However, the crystallization of cellulose for THWS250 was

424

enhanced under torrefaction temperature of 250 ℃, especially for TRWS250 (Table 3).

425

Previous studies showed similar finds for pine sawdust (Gong et al., 2016). The

426

phenomenon could be explained by the results that smaller crystallization was formed

427

during recrystallized process. In addition, small cellulose crystallization had lower

428

thermal stability than large crystallite (Zhang et al., 2015). When the torrefaction

429

temperature reached to 300 ℃, the hemicellulose and cellulose were mostly degraded,

430

and the lignin content increased sharply. Lignin degradation requires more energy,

431

which led to the Tmax significantly increased as well. In Fig 2 (b), there were similar 20

432

trends that the peak of cellulose degradation slightly shifted to the lower temperature

433

region, and then turned to higher temperature region as the torrefaction temperature

434

increased. However, the inflection point occurred at 275 ℃ but not 250 ℃. This

435

suggested that the cellulose of RWS had been degraded largely when the torrefaction

436

temperature was 275 ℃. Rod-milling pretreatment disrupted the crystallinity of cellulose

437

into amorphous cellulose, and it required less energy to decompose than crystalline

438

cellulose during pyrolysis (Wang et al., 2017). Moreover, the lignin content increased as

439

the degradation of hemicellulose and cellulose under 275 ℃ torrefaction, which led to

440

the thermal degradation of cellulose was overlapped with the degradation of lignin.

441

What’s more, the temperature with the maximum mass loss rate (DTGmax) in RWS

442

(331.82 ℃) was the lowest among HWS and THWSX (341.32 ℃, 340.9 ℃, 339.52℃,

443

and 363.35 ℃), which indicated a lower thermal stability of cellulose for RWS. This

444

phenomenon could also be ascribed to the smaller size, larger SSA and PV of RWS

445

(Section 3.3), which resulted in a better heat transfer to the inner part of cellulose (Khan

446

et al., 2016; Bai et al., 2018).

447

The peak intensities of cellulose were observed around 2.51 wt.% /℃ for HWS;

448

2.67 wt.% /℃ for THWS250; 1.82 wt.% /℃ for THWS275; 0.52 wt.% /℃ for THWS300,

449

respectively (Table 4). The intensity increased initially and later declined with

450

increasing torrefaction temperature, and the width of cellulose degradation peaks

451

decreased. The ascent can be explained by the fact that the content of cellulose was

452

increase, which resulted from the releasing of volatiles, degrading of hemicellulose, and

453

the light degrading of cellulose at 250 ℃. However, the cellulose content gradually 21

454

decreased, and the increasing amorphous cellulose content led to a more concentrated

455

degradation of cellulose with the further increases in torrefaction temperature (Joshi et

456

al., 2015). When the temperature reached to 300 ℃, the peak of cellulose degradation

457

was basically disappeared, and the peak of maximum mass loss temperature moved to

458

higher temperature region (from 339.52 to 363.35 ℃). It showed that cellulose was

459

decomposition dramatically after torrefaction of 300 ℃. Meanwhile, the peak of lignin

460

degradation became protruding, and the thermal degradation of lignin began to degrade

461

at lower temperature. The reason for this has been explained by an increase in content of

462

lignin and the damage in structure of lignin after 300 ℃ torrefaction. The intensity for

463

cellulose in RWS (5.04 wt.%/℃) was higher than that in HWS (2.51 wt.%/℃), which

464

was explained by the results of a more concentrated degradation of amorphous cellulose

465

through rod-milling. This was due to the destruction of crystalline structure and the

466

increase of amorphous cellulose through rod-milling pretreatment. This was confirmed

467

by the results of the section XRD analysis as well. Furthermore, the amplitude for lignin

468

in TRWSX was higher than that in THWSX, which implied that cellulose degraded

469

more readily after rod-milling and thus caused an increase in the proportion of lignin

470

mass during the torrefaction process. This was similar to what has been reported Gong

471

et al. (2016).

472

The last stage ranged from approximate 500 to 800 °C is the carbonization of

473

residues (Zhang et al., 2016). The mass loss and pyrolysis rate gradually declined during

474

this thermal decomposition stage. Table 4 showed that the final residue mass (Mr)

475

increased significantly with the increase of torrefaction temperature for all samples, 22

476

which was caused by the release of some volatile components during torrefaction

477

(Zheng et al., 2015). This was consistent with the results of 3.1.2 that the torrefaction

478

pretreatment increased the Ash. There was no significantly difference about the amount

479

of Mr between HWS and RWS under 250 °C torrefaction. When torrefaction

480

temperature reached to 275 °C and 300 °C, TRWSX showed higher Mr than THWSX.

481

Yu et al. (2018) found that the Mr was affected largely by heat transfer efficiency during

482

pyrolytic reaction. That could be confirmed by the results of section 3.3 that TRWS275

483

and TRWS300 had smaller particle size, larger SSA and PV.

484

The Di of HWS and RWS under different torrefaction temperatures were shown in

485

Fig.3. For all samples, the Di decreased significantly as the torrefaction temperature

486

increased. This could be related to the results of section 3.1.2 that the VM decreased

487

with the rises in torrefaction temperature. Compared with HWS, the higher Di for RWS

488

meant the easier pyrolysis of a certain component occurred after RMP (Fang et al.,

489

2015). The Dis of THWS250 and TRWS250 were close. However, the Di of TRWS275

490

was lower than THWS275’s. This indicated that the pyrolysis reactivity of HWS was

491

higher than RWS’s under 275 ℃. As described in section 3.4 that rod-milling

492

pretreatment disrupted the structure of cellulose and decreased its stability, causing it

493

degrade at lower torrefaction temperature. Moreover, the degradation of cellulose for

494

HWS mainly occurred in 300 ℃, while the cellulose degradation of RWS mainly

495

occurred in 275 ℃. Therefore, the content of lignin for RWS was higher than HWS’s

496

under 275 ℃. Moreover, higher lignin content reduced the reactivity of biomass (Biswas

497

et al., 2011). When torrefaction temperature reached to 300 ℃, the lignin became the 23

498

main part of HWS and RWS samples, which resulted from the release of the most

499

hemicellulose and cellulose. In addition, the Di of 10.28 of TRWS300 was higher than

500

8.34 of THWS300, which meant that lignin of TRWS300 was easier to degrade. This

501

could be explained by the fact that rod-milling pretreatment lower the thermal stability

502

of lignin and the structure of lignin was affected by rod-milling pretreatment. These

503

observations’ explanations can be seen in our previous findings (Bai et al. 2018).

504 505 506

4.Conclusion CRT affected wheat straw’s physicochemical and torrefaction characteristics. CRT

507

significantly elevated torrefied wheat straw’ carbon content, HHV, FR and reduced

508

oxygen content and atomic H/C and O/C ratios. CRT significantly reduced cellulose

509

crystallinity and increased SSA and PV of wheat straw, causing wheat straw to degrade

510

at lower pyrolysis temperatures. Moreover, CRT altered the thermal degradation

511

mechanism of wheat straw, and improved Di under identical torrefaction conditions

512

except 275 °C. These values reached the maximum under 300 °C. In conclusion, the

513

combination of rod-milling with suitable torrefaction as a pretreatment was

514

recommended to enhance fuel properties of lignocellulosic biomass materials.

515 516

E-supplementary data for this work can be found in e-version of this paper online.

517 518 519

Acknowledgement The authors gratefully acknowledge financial support from China Scholarship 24

520

Council (201706355006) and the financial support from China Agriculture Research

521

System (CARS-35). The work performed at Penn State University was supported by the

522

USDA National Institute of Food and Agriculture (NIFA) Federal Appropriations under

523

Project PEN04547.

524 525

Conflicts of interest: none.

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

References 1. Bai, X., Wang, G., Yu, Y., Wang, D., Wang, Z. 2018. Changes in the physicochemical structure and pyrolysis characteristics of wheat straw after rod-milling pretreatment. Bioresour Technol, 250, 770-776. 2. Bai, X., Wang, G., Gong, C., Yu, Y., Liu, W., Wang, D. 2017. Co-pelletizing characteristics of torrefied wheat straw with peanut shell. Bioresour Technol, 233, 373-381. 3. Basch, A., Lewin, M. 1973. The influence of fine structure on the pyrolysis of cellulose. I. Vacuum pyrolysis 4. Biswas, A.K., Umeki, K., Yang, W., Blasiak, W. 2011. Change of pyrolysis characteristics and structure of woody biomass due to steam explosion pretreatment. Fuel Processing Technology, 92(10), 1849-1854. 5. Chen, D., Cen, K., Jing, X., Gao, J., Li, C., Ma, Z. 2017. An approach for upgrading biomass and pyrolysis product quality using a combination of aqueous phase bio-oil washing and torrefaction pretreatment. Bioresour Technol, 233, 150-158. 6. Chen, H., Chen, X., Qin, Y., Wei, J., Liu, H. 2017. Effect of torrefaction on the properties of rice straw high temperature pyrolysis char: Pore structure, aromaticity and gasification activity. Bioresour Technol, 228, 241-249. 7. Chen, W.H., Du, S.W., Tsai, C.H., Wang, Z.Y. 2012. Torrefied biomasses in a drop tube furnace to evaluate their utility in blast furnaces. Bioresour Technol, 111, 433-8. 8. Chen, W.H., Kuo, P.C., 2010. A study on torrefaction of various biomass materials and its impact on lignocellulosic structure simulated by a thermogravimetry. Energy 35 (6), 2580–2586 9. Chen, W.-H., Peng, J., Bi, X.T. 2015. A state-of-the-art review of biomass torrefaction, densification and applications. Renewable and Sustainable Energy Reviews, 44, 847-866.EM 10. Chew, J.J., Doshi, V. 2011. Recent advances in biomass pretreatment – Torrefaction fundamentals and technology. Renewable and Sustainable Energy Reviews, 15(8), 4212-4222. 11. Chiou, B.S., Valenzuela-Medina, D., Bilbao-Sainz, C., Klamczynski, A.K., Avena-Bustillos, R.J., Milczarek, R.R., Du, W.X., Glenn, G.M., Orts, W.J. 2015. Torrefaction of pomaces and nut shells. Bioresour Technol, 177, 58-65. 12. Du, S.W., Chen, W.H., Lucas, J.A. 2014. Pretreatment of biomass by torrefaction and carbonization for coal blend used in pulverized coal injection. Bioresour Technol, 161, 333-9. 13. Erlich, C., Fransson, T.H. 2011. Downdraft gasification of pellets made of wood, palm-oil residues respective bagasse: Experimental study. Applied Energy, 88(3), 899-908. 25

556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599

14. Fang, S., Yu, Z., Lin, Y., Hu, S., Liao, Y., Ma, X. 2015. Thermogravimetric analysis of the co-pyrolysis of paper sludge and municipal solid waste. Energy Conversion and Management, 101, 626-631. 15. Friedl, A., Padouvas, E., Rotter, H., Varmuza, K. 2005. Prediction of heating values of biomass fuel from elemental composition. Analytica Chimica Acta, 544(1-2), 191-198. 16. Gao, C., Xiao, W., Ji, G., Zhang, Y., Cao, Y., Han, L. 2017. Regularity and mechanism of wheat straw properties change in ball milling process at cellular scale. Bioresour Technol, 241, 214-219. 17. Gong, C., Huang, J., Feng, C., Wang, G., Tabil, L., Wang, D. 2016. Effects and mechanism of ball milling on torrefaction of pine sawdust. Bioresour Technol, 214, 242-247. 18. Granados, D.A., Ruiz, R.A., Vega, L.Y., Chejne, F. 2017. Study of reactivity reduction in sugarcane bagasse as consequence of a torrefaction process. Energy, 139, 818-827. 19. Han, K.-H., Ko, J.-H., Yang, S.H. 2007. Optimizing lignocellulosic feedstock for improved biofuel productivity and processing. Biofuels, Bioproducts and Biorefining, 1(2), 135-146. 20. Huang, Y.F., Cheng, P.H., Chiueh, P.T., Lo, S.L. 2017. Leucaena biochar produced by microwave torrefaction: Fuel properties and energy efficiency. Applied Energy, 204, 1018-1025. 21. Ji, G., Gao, C., Xiao, W., Han, L. 2016. Mechanical fragmentation of corncob at different plant scales: Impact and mechanism on microstructure features and enzymatic hydrolysis. Bioresour Technol, 205, 159-65. 22. Joshi, Y., Di Marcello, M., de Jong, W. 2015. Torrefaction: Mechanistic study of constituent transformations in herbaceous biomass. Journal of Analytical and Applied Pyrolysis, 115, 353-361. 23. Kan, T., Strezov, V., Evans, T.J. 2016. Lignocellulosic biomass pyrolysis: A review of product properties and effects of pyrolysis parameters. Renewable and Sustainable Energy Reviews, 57, 1126-1140. 24. Khan, A.S., Man, Z., Bustam, M.A., Kait, C.F., Khan, M.I., Muhammad, N., Nasrullah, A., Ullah, Z., Ahmad, P. 2015. Impact of Ball-Milling Pretreatment on Pyrolysis Behavior and Kinetics of Crystalline Cellulose. Waste and Biomass Valorization, 7(3), 571-581. 25. Liu, H., Gu, X., Wang, X., Han, L., Xiao, W., 2015. Ultrafine grinding and alcolhol liquefaction for corn stover. Trans. Chin. Soc. Agric. Mach. 46 (11), 214–220. 26. Poddar, S., Kamruzzaman, M., Sujan, S.M.A., Hossain, M., Jamal, M.S., Gafur, M.A., Khanam, M. 2014. Effect of compression pressure on lignocellulosic biomass pellet to improve fuel properties: Higher heating value. Fuel, 131, 43-48. 27. Segal, L., Creely, J.J., Martin, A.E., Conrad, C.M., 1959. An empirical method for estimating the degree of crystallinity of native cellulose using the X-ray diffractometer. Text. Res. J. 29 (10), 786–794. 28. Shang, L., Ahrenfeldt, J., Holm, J.K., Sanadi, A.R., Barsberg, S., Thomsen, T., Stelte, W., Henriksen, U.B. 2012. Changes of chemical and mechanical behavior of torrefied wheat straw. Biomass and Bioenergy, 40, 63-70. 29. Tian, B., Qiao, Y.y., Tian, Y.y., Liu, Q. 2016. Investigation on the effect of particle size and heating rate on pyrolysis characteristics of a bituminous coal by TG–FTIR. Journal of Analytical and Applied Pyrolysis, 121, 376-386. 30. Tooyserkani, Z., Sokhansanj, S., Bi, X., Lim, J., Lau, A., Saddler, J., Kumar, L., Lam, P.S., Melin, S. 2013. Steam treatment of four softwood species and bark to produce torrefied wood. Applied Energy, 103, 514-521. 31. Wang, L., Barta-Rajnai, E., Skreiberg, Ø., Khalil, R., Czégény, Z., Jakab, E., Barta, Z., Grønli, M. 2017. Effect of torrefaction on physiochemical characteristics and grindability of stem wood, stump and bark. Applied Energy. 32. Wang, S., Dai, G., Ru, B., Zhao, Y., Wang, X., Xiao, G., Luo, Z. 2017. Influence of torrefaction on the characteristics and pyrolysis behavior of cellulose. Energy, 120, 864-871. 33. Wang, Y., Wu, Q., Dai, L., Zeng, Z., Liu, Y., Ruan, R., Fu, G., Yu, Z., Jiang, L. 2018. Co-pyrolysis of wet torrefied bamboo sawdust and soapstock. Journal of Analytical and Applied Pyrolysis, 132, 211-216. 34. Wang, Z., McDonald, A.G., Westerhof, R.J.M., Kersten, S.R.A., Cuba-Torres, C.M., Ha, S., Pecha, B., 26

600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618

Garcia-Perez, M. 2013. Effect of cellulose crystallinity on the formation of a liquid intermediate and on product distribution during pyrolysis. Journal of Analytical and Applied Pyrolysis, 100, 56-66. 35. Wen, J.-L., Sun, S.-L., Yuan, T.-Q., Xu, F., Sun, R.-C. 2014. Understanding the chemical and structural transformations of lignin macromolecule during torrefaction. Applied Energy, 121, 1-9. 36. Wu, Z., Wang, S., Zhao, J., Chen, L., Meng, H. 2014. Synergistic effect on thermal behavior during co-pyrolysis of lignocellulosic biomass model components blend with bituminous coal. Bioresour Technol, 169, 220-228. 37. Xu, S., Zhou, Z., Yu, G., Wang, F. 2010. Effects of Pyrolysis on the Pore Structure of Four Chinese Coals. Energy & Fuels, 24(2), 1114-1123. 38. Yu, Y., Wang, G., Bai, X., Liu, J., Wang, D., Wang, Z. 2018. Combined different dehydration pretreatments and torrefaction to upgrade fuel properties of hybrid pennisetum (Pennisetum americanum xP. purpureum). Bioresour Technol, 263, 541-547. 39. Yue, Y., Singh, H., Singh, B., Mani, S. 2017. Torrefaction of sorghum biomass to improve fuel properties. Bioresour Technol, 232, 372-379. 40. Zhang, S., Dong, Q., Zhang, L., Xiong, Y. 2016. Effects of water washing and torrefaction on the pyrolysis behaviohanr and kinetics of rice husk through TGA and Py-GC/MS. Bioresour Technol, 199, 352-361. 41. Zhang, Y., Bi, P., Wang, J., Jiang, P., Wu, X., Xue, H., Liu, J., Zhou, X., Li, Q. 2015. Production of jet and diesel biofuels from renewable lignocellulosic biomass. Applied Energy, 150, 128-137. 42. Zheng, A., Zhao, Z., Chang, S., Huang, Z., Zhao, K., Wei, G., He, F., Li, H. 2015. Comparison of the effect of wet and dry torrefaction on chemical structure and pyrolysis behavior of corncobs. Bioresour Technol, 176, 15-22.

27

619

Figure captions

620

Fig. 1. Van Krevelen diagram of raw and torrefied samples.

621

Fig. 2. DTG results of raw and torrefied (a) HWS and (b) RWS at a residence time of 30 min.

622

Fig. 3. Di of raw and torrefied samples.

28

623

Table and Figures

624

Table 1. Ultimate analysis of raw and torrefied samples.

625

a

Samples

C (%)b

H (%)

Oa (%)

N (%)

S (%)

O/C

H/C

HHV (MJ/kg)

HWS

44.00

5.76

48.92

0.94

0.38

0.83

1.57

17.52

THWS250

48.41

5.25

45.04

0.85

0.45

0.70

1.30

19.10

THWS275

52.29

5.20

40.95

1.08

0.48

0.59

1.19

20.64

THWS300

56.94

5.04

36.23

1.28

0.51

0.48

1.06

22.52

RWS

43.38

5.59

49.66

0.93

0.44

0.86

1.55

17.29

TRWS250

48.78

5.03

45.07

0.69

0.43

0.69

1.24

19.17

TRWS275

58.16

4.24

36.38

0.77

0.45

0.47

0.87

22.39

TRWS300

64.95

3.39

30.36

0.76

0.54

0.35

0.63

24.32

By difference. b %, dry basis.

29

626

627

Table 2. Proximate and FR analysis of raw and torrefied samples.

a

Samples

FC (%)a

VM (%)

Ash (%)

FR

HWS

19.47±0.15

77.03±0.25

3.50±0.07

0.25 (<0.01)

THWS250

23.92±0.68

66.13±0.25

9.95±0.22

0.36 (<0.01)

THWS275

24.99±0.29

65.35±0.57

9.66±0.62

0.38 (<0.01)

THWS300

43.11±0.46

44.18±0.26

12.71±0.40

0.98±0.01

RWS

17.13±0.22

74.73±0.48

8.14±0.62

0.23 (<0.01)

TRWS250

23.69±0.79

64.35±0.58

11.96±0.13

0.37±0.02

TRWS275

36.46±0.61

50.99±0.83

12.55±0.41

0.72±0.02

TRWS300

52.94±1.23

33.97±0.24

13.09±0.98

1.56±0.04

%, dry basis. Data are shown as their mean values±standard deviation.

30

628

629

Table 3. Mass yield, energy yield, energy density, CrI, SSA and PV of raw and torrefied samples.

a

PV (cm3•

Samples

Mass yield (%)a

Energy yield (%)

Energy density

SSA (m2/g)

HWS

100.00

100.00

1.00

1.74

0.44

51.33±0.08

THWS250

77.90±0.51

88.14±0.58

1.13

0.94

0.20

54.61±1.22

THWS275

69.39±1.15

77.44±1.32

1.15

1.67

0.40

35.29±1.03

THWS300

49.72±1.35

63.92±1.35

1.29

1.87

0.46

18.87±0.92

RWS

100.00

100.00

1.00

2.86

1.19

11.59±0.62

TRWS250

73.95±0.48

80.80±0.52

1.09

1.55

0.29

24.12±0.54

TRWS275

57.50±1.17

73.48±1.49

1.28

1.66

0.26

8.98±0.24

TRWS300

46.76±0.8

64.91±1.11

1.39

2.98

1.26

2.56±0.54

%, dry basis. Data are shown as their mean values±standard deviation.

31

10-2/g)

CrI (%)

630

631

Table 4 Pyrolysis parameters of the raw and torrefied samples with the heating rate of 20 ℃/min.

a

Samples

Ti (℃)

Tmax (℃)

Rmax ( % min-1)

DTGmax (% ℃-1)

△T1/2 (℃)

Mr(%)a

HWS

254.86

341.31

16.89

2.51

300.61

23.95

THWS250

302.64

340.9

17.92

2.67

317.17

32.97

THWS275

309.43

339.52

12.17

1.82

316.32

39.81

THWS300

344.05

363.35

3.47

0.52

332.77

58.68

RWS

218.41

331.82

15.11

5.04

276.98

23.65

TRWS250

263.98

323.47

15.06

5.02

294.76

31.36

TRWS275

307.11

336.86

9.13

3.04

313.49

47.93

TRWS300

336.04

400.88

4.79

1.26

345.92

59.89

%, dry basis.

32

1.6

Atomic H/C ratio

HWS:H/C=1.397*O/C+0.375, R2=0.973 1.4

1.2

1.0

0.8

RWS:H/C=1.778*O/C+0.017, R2=0.998

HWS THWS250 THWS275 THWS300 RWS TRWS250 TRWS275 TRWS300

0.6

0.3

632 633

0.4

0.5

0.6

0.7

0.8

0.9

Atomic O/C ratio

Figure.1

33

(a)

3.0

HWS THWS250 THWS275 THWS300

Cellulose

DTG, wt.%/℃

2.5

2.0

1.5 Hemicellulose 1.0 Lingin 0.5

0.0 100

200

300

400

500

600

700

800

Temperature, ℃

634

(b)

Cellulose RWS TRWS250 TRWS275 TRWS300

DTG, wt.%/℃

5

4

3

2

Hemicellulose Lingin

1

0 100

300

400

500

600

700

800

Temperature, ℃

635 636

200

Figure.2

34

80

75.27

HWS RWS

70

Di, 10-8 % min-1 ℃ -3

64.59 59.83

60 54.76 50 40

36.62

30

28.15

20 10

8.34

10.28

0 T0

637 638

T250

T275

T300

Torrefaction, ℃

Figure.3

639

35

640

HIGHLIGHTS

641



642 643

torrefaction conditions except 275 °C. 

644 645

Rod-milling improved the devolatilization index of wheat straw under identical

Rod-milling combined with torrefaction changed the thermal degradation mechanism of wheat straw.



Combined pretreatment is conducive to the energy utilization of wheat straw.

646 647 648

36