Structural order evaluation and structural evolution of coal derived natural graphite during graphitization

Journal Pre-proof Structural order evaluation and structural evolution of coal derived natural graphite during graphitization Shuai Zhang, Qinfu Liu, Hao Zhang, Rujia Ma, Kuo Li, Yingke Wu, Brian J. Teppen PII:

S0008-6223(19)31122-4

DOI:

https://doi.org/10.1016/j.carbon.2019.10.104

Reference:

CARBON 14755

To appear in:

Carbon

Received Date: 9 July 2019 Revised Date:

30 October 2019

Accepted Date: 31 October 2019

Please cite this article as: S. Zhang, Q. Liu, H. Zhang, R. Ma, K. Li, Y. Wu, B.J. Teppen, Structural order evaluation and structural evolution of coal derived natural graphite during graphitization, Carbon (2019), doi: https://doi.org/10.1016/j.carbon.2019.10.104. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. 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. © 2019 Published by Elsevier Ltd.

1

Structural Order Evaluation and Structural Evolution of Coal

2

Derived Natural Graphite during Graphitization

3

Shuai Zhanga, *, Qinfu Liua, *, Hao Zhanga, Rujia Maa, Kuo Lia, Yingke Wua, Brian J. Teppenb

4

a

5

Technology (Beijing), Beijing, 100083, People’s Republic of China

6

b

7

Michigan, 48824, United States

School of Geosciences and Surveying Engineering, China University of Mining and

Department of Plant, Soil, and Microbial Sciences, Michigan State University, East Lansing,

8 9 10

*Corresponding authors. Tel: +86 18810545318; E-mail: [email protected] Tel: +86 13911683809; E-mail: [email protected]

11 12 13 14 15 16 17 18 19 20 21 22 23 24 1

25

Abstract: The structure of onset phase of coal derived natural graphite (CDNG for short),

26

semi-graphite, and CDNG subjected to different metamorphic grade were investigated using

27

XRD, Raman spectroscopy, and HRTEM to trace their structural evolution during

28

graphitization. The transitional phases with different structural order exist in graphite field. A

29

classification of these transitional phases is of particular importance to achieve the rational

30

utilization of CDNG. The d(002)-spacing is used to distinguish the pre-phases of highly ordered

31

CDNG (ordered CDNG) and highly ordered CDNG in graphite field. The asymmetry index

32

(AI) of (002) reflection of XRD patterns and Raman parameters D1/G area ratio and FWHM

33

of G band correlate well with d(002)-spacing of CDNG in the zone of semi-graphite to graphite,

34

thus they are also proposed to evaluate the structural order of CDNG. For structural evolution

35

of CDNG during graphitization, at the beginning, the BSUs composed of stacking aromatic

36

layers arrange face to face and tend to link together laterally. Then adjacent BSUs link

37

laterally to form long-range wrinkled layers. With graphitization, the wrinkled layers evolve

38

to flat lamellae and turbostratic structure gradually disappears. Finally, the dislocation in

39

stacking and in-plane defect of flat aromatic layers decreases to attain triperiodic graphite

40

structure.

41 42 43 44 45 46 47 48 49 50 51 52 2

53 54 55

1. Introduction

56

Graphite of metamorphic origin hosted in metamorphic rocks is derived from the

57

biological or abiogenic carbon, or graphitic carbon precipitated from hot fluids subjected to

58

regional and contact metamorphism [1-4]. This transformation is called graphitization

59

proceeding through intermediate forms, and reach triperiodic graphite structure [3,5]. When

60

the coal-bearing strata is intruded by large batholith, the coal macerals can be transformed to

61

graphite [6-8]. This is referred to as coal derived natural graphite (CDNG for short), and it is

62

an important component of graphite resources and has relatively small crystalline grain

63

compared to phaneritric flaky graphite. The formation of CDNG requires massive igneous

64

intrusion in coal-bearing strata which provides thermal energy to induce the coal

65

macromolecules to be transformed into a triperiodic graphite structure through contact

66

metamorphism. Depending on their distance relative to the igneous body, anthracite,

67

meta-anthracite, semi-graphite, and graphite are distributed in different metamorphic zone

68

[7,8].

69

The International Committee for Coal and Organic Petrology (ICCP) distinguishes

70

anthracite, meta-anthracite, semi-graphite, and graphite using XRD based d(002)-spacing,

71

maximum reflectance (%Rmax), and H/C atomic ratio [6]. XRD and Raman spectroscopy are

72

the commonly used methods to investigate the structural evolution of carbonaceous materials

73

with high resolution [9-20]. Both of them provide quantitative parameters on the degree of

74

structural organization of carbonaceous materials. XRD parameter d(002)-spacing was

75

employed for evaluating structural order of graphite [10,11]. The full width at half maximum 3

76

(FWHM) of G band and the D1 / (G + D1 + D2) area ratio of Raman spectrum were proposed

77

to distinguish semi-graphite and graphite [17,21]. High resolution transmission electron

78

microscope (HRTEM) images combined with selected area electron diffraction patterns

79

(SAED) can directly visualize the structural order evolution of carbonaceous materials during

80

graphitization [1,17,22]. XRD gives bulk structural information of samples [23,24]. While

81

Raman spectroscopy and HRTEM can provide less averaged information because of high

82

spatial resolution [17]. For the CDNG, there are transitional phases with different structural

83

order existing within the field of semi-graphite to graphite due to subjected to different

84

metamorphic grade [8]. A clear classification of these transitional phases is of particular

85

importance, because the physicochemical properties of CDNG are highly dependent to their

86

structural order that directly influences their industrial utilization [25].

87

In present study, the structure of CDNG with increasing metamorphic grade is

88

investigated using XRD, Raman spectroscopy, and HRTEM to trace their structural evolution

89

during graphitization. The aim is to use XRD and Raman parameters to evaluate the structural

90

order of CDNG especially for samples in the graphite field, which is of significance for

91

rational utilization of CDNG.

92

2. Materials and methods

93

2.1. Samples and sample preparation

94

Nine CDNG samples were collected from active mines in Xinhua and Lutang County in

95

Hunan Province, China, where the coal-bearing strata occurring in Lower Carboniferous

96

Ceshui Formation and Upper Permian Longtan Formation were intruded by Indosinian

97

Tianlongshan granitic pluton and Yanshannian Qitianling granitic pluton, respectively [26,27]. 4

98

The coal was thermally affected by intrusive body forming graphite with different structural

99

order in the contact aureole of granitic pluton [7,8,26]. Here, CDNG were selected from two

100

contact aureole, because highly ordered CDNG samples were not found in the high

101

metamorphic zone very near to the contact of Tianlongshan granitic pluton. Thus, the highly

102

ordered CDNG samples in Lutang mines were also selected to fill the gap from semi-graphite

103

to highly ordered CDNG. The samples collected in Shengli, Baichong, and Shihangli mines in

104

Xinhua County are located at the distance around 1000 m, 600 m, and 300 m, respectively, to

105

the Tianlongshan granitic pluton [26]. They were labeled as SL, BC, and SHL using the

106

abbreviation of the mine names. The No. 3 and No. 5 coal seams are economically minable in

107

the Lower Carboniferous Ceshui Formation in the mining area of Xinhua County. The

108

samples collected from the two coal seams in Baichong mine were labeled as BC-1 and BC-2,

109

respectively. The five samples collected in Lutang mines of Lutang County were labeled as

110

LT-1 ~ LT-5. The distance of these samples to the Qitianling granitic pluton were given in

111

Table 1.

112

The bulk graphite samples were crushed into small pieces, and all the crushed samples

113

were sieved using 120 mesh sieve instead of grinding to minimize the damage to primary

114

structure of CDNG. The samples were demineralized before XRD and HRTEM investigations

115

to avoid the effect of mineral matter on quantitative analysis of their crystalline structure. For

116

demineralization, 5 g of each sample was dispersed in mixed acid of 30 ml of HCl solution

117

(37 wt.%) and 20 ml of HF solution (40 wt.%) and stirred for 3 h under 60

118

repeated 3 times. The samples were filtered and washed with deionized water until a pH of ~7

119

was reached. The samples were filtered and dried in air at ambient temperature. 5

. This was

120

2.2. X-ray diffraction analysis

121

The XRD was carried out using Rigaku D/max-2500PC diffractometer equipped with

122

CuKα radiation, graphite monochrometer, slit system 1º–0.3 mm–1º. The X-ray generator

123

voltage and current were held at 40 kV and 100 mA, respectively. Samples were scanned

124

using continuous sweep method from 10 to 90º in 2θ range at scanning speed of 4º/min. The

125

sampling width was 0.02º during scanning. To determine the stacking height (Lc) and lateral

126

size (La) of graphite crystals using Scherrer equation, the region of (002) reflection in 2θ

127

range of 20 to 30º and the region of (110) reflection in 2θ range of 75 to 80º were rescanned at

128

lower speed of 0.5º/min, respectively. Before XRD scanning, the silicon (16.67% wt.%) as

129

the internal standard was mixed uniformly with the powdered samples. The 2θ values of

130

obtained (002) reflection and (110) reflection were calibrated using the internal silicon

131

standard (2θ: 28.443º and 76.378º), respectively. Because XRD gives the bulk

132

characterization of samples, the measured (002) reflection and (110) reflection are contributed

133

from reflection of turbostratic structure, structure defect, and fully ordered structure in a

134

graphite sample. The (002) and (110) reflections were fitted to isolate γ peak that is the

135

turbostratic structure and structure defect as shown in Fig. S1 in Supplementary Material.

136

For each reflection, the curve fitting was performed three times, and the mean value was

137

adopted. Because the (110) reflections are almost symmetric for highly ordered CDNG

138

(samples LT-2 ~ LT-5), their (110) reflections were not fitted. The pseudo-Voigt function in

139

Jade 5 package was used for fitting the reflections. The d(002)-spacing, peak positions, and full

140

width at half maximum (FWHM) were determined (Table S1 in Supplementary Material).

141

The crystalline structure parameters of CDNG, lateral size (La) and stacking height (Lc), 6

142

were determined using Scherrer equations (1) and (2) [28].

143

La = 1.84λ/βacos(θa)

(1)

144

Lc = 0.89λ/βccos(θc)

(2)

145

where λ is the wavelength (0.154056 nm) of the radiation used; βa and βc are the FWHM of

146

fitted (110) peak and (002) peak, respectively, and θa and θc are the corresponding Bragg

147

angles of each peak.

148

The slit width of XRD could widen the reflections, which should be removed from the

149

fitted FWHM of (002) peak and (110) peak. The FWHM mainly caused by the slit width in 2θ

150

range of 2.5 to 90º was obtained using the XRD pattern of silicon standard obtained under the

151

same measurement condition for the scanning of (002) and (110) reflections at lower scanning

152

speed of 0.5º/min. The XRD pattern of silicon standard was also fitted using the pseudo-Voigt

153

function in Jade 5 package. Then the FWHM mainly caused by the slit width at corresponding

154

2θ of (002) peak and (110) peak were subtracted by the FWHM of (002) peak and (110) peak.

155

2.3. Raman spectroscopy analysis

156

Raman spectra were acquired employing a Renishaw in Via Raman spectrometer

157

equipped with a Leica DMLB microscope, Renishaw helium neon laser (wavelength: 532 nm),

158

and CCD array detector at room temperature using the polished sections made from bulk

159

samples. Laser focusing and sample viewing were performed through a 50× objective lens.

160

The laser power of incident beam on sample was kept below 5 mW to prevent thermal

161

damage to the sample surface. The samples were scanned between 500 ˗ 3200 cm-1 including

162

the first order region (500 ˗ 2000 cm-1) and the second order region (2000 ˗ 3200 cm-1) with

163

data acquisition time of 10 ˗ 30 s. Three different spots for each sample were scanned due to 7

164

the heterogeneity of less ordered CDNG, and each spot was scanned twice to ensure the

165

accuracy of measurement. The Raman spectra parameters (band position, FWHM, intensity,

166

and area) are significantly sensitive to the curve-fitting procedure [29,30]. The procedure

167

developed by Lünsdorf et al. [30] was used for the curve fitting of Raman spectra.

168

2.4. High resolution transmission electron microscopy analysis

169

The CDNG samples for HRTEM analysis were dispersed in an ultrasonic ethanol bath for

170

approximately 30 min. One drop of sample suspension was placed on carbon-coated copper

171

grids and then dried at room temperature. The HRTEM was carried out using a FEI Tecnai G2

172

F30 electron microscopy operated at an acceleration voltage of 300 kV for obtaining TEM

173

images and high-resolution lattice images.

174

3. Results and discussion

175

3.1. XRD

176

XRD is one of the most common methods to study the structural order of graphite using

177

crystallographic parameters such as d(002)-spacing, FWHM of (002) reflection, and crystallite

178

dimension [11,19,31]. XRD patterns of CDNG (Fig. 1 and Table 1) reflect their structural

179

order evolution during graphitization. The XRD patterns of CDNG samples are ordered with

180

increasing metamorphic grade, their structural evolution during graphitization can be easily

181

observed (Fig. 1). The sample SL collected in the mine located at the distance around 1000 m

182

of Tianlongshan granitic pluton was subjected to low metamorphic grade, therefore, its XRD

183

pattern shows broad asymmetric (002) and (110) reflections (Fig. 1b and d). The d(002)-spacing

184

is 0.3395 nm (Table 1), and the (100) and (101) reflections (Fig. 1c) are not well defined. This

185

indicates that the turbostratic structure exists in its crystalline structure [19]. The graphite 8

186

samples closer to intrusive body attained higher metamorphic grade, consequently, their

187

structural order increased as reflected by their (002) and (110) reflections that become

188

gradually sharper and symmetric (samples BC, SHL, and LT-1 ~ LT-5 in Fig. 1b and d), which

189

are similar to the XRD patterns of carbonaceous materials subjected to increasing

190

metamorphic grade in contact and regional metamorphic rocks [9-11,31]. The d(002)-spacing

191

(Table 1) decreases continuously from 0.3395 to 0.3355 nm approaching an ideal crystalline

192

structure of Sri Lanka graphite [9]. Although the (100), (101), (110) and (112) reflections of

193

CDNG have relatively weak intensity compared to the (002) reflection, they provide

194

important crystallographic information such as lateral extent of aromatic sheet and polytype

195

of graphite [19,24]. The (100) and (101) reflections represent thermodynamically stable 2H

196

phase (hexagonal phase) of graphite. A 3R(101) reflection representing rhombohedral phase

197

of graphite exists between (100) and (101) reflections (Fig. 1c) [32,33]. The rhombohedral

198

phase is thermodynamically unstable, and it normally coexists with hexagonal phase in

199

natural metamorphic graphite, which is considered an extended stacking fault in hexagonal

200

graphite [32]. The (112) reflection is obscured in XRD patterns of SL ~ LT-1 at 2θ value of

201

around 83.5º and appears in the XRD patterns of LT-2 ~ LT-5 (Fig. 1e), indicating the

202

formation of ordered three-dimensional crystalline structure by releasing dislocations within

203

graphite lamellae as structural order increases [19,24,34,35].

9

204 205 206

Fig. 1. XRD patterns of CDNG in 2θ range (a) between 10 to 90º, (b) between 24 to 28º, (c) between 42 to 46º, (d) between 77 to 78.5º, (e) between 83 to 84º.

207

According to the International Committee for Coal and Organic Petrology (ICCP)

208

classification for anthracite, meta-anthracite, semi-graphite, and graphite [6], the anthracite

209

has d(002)-spacing ˃ 0.340 nm. The meta-anthracite is classified by d(002)-spacing within the

210

range of 0.338 ˗ 0.340 nm, and semi-graphite is defined by d(002)-spacing within the range of

211

0.337 ˗ 0.338 nm. The d(002)-spacing of 0.3354 ˗ 0.337 nm represent the graphite zone. The

212

d(002)-spacing of samples in the present study (Table 1) indicate that sample SL in Xinhua

213

County is within the meta-anthracite range, which can be treated as the onset phase of

214

graphite, while sample BC-1 is at the boundary between meta-anthracite and semi-graphite,

215

being almost a semi-graphite. On the other hand, sample BC-2 belongs to the semi-graphite

216

category, while the sample SHL in Xinhua County and all samples LT-1 ~ LT-5 in Lutang 10

217

County are graphite. Within the graphite field, the samples SHL and LT-1 ~ LT-5 have

218

different structural order as reflected by the broad and asymmetric degree of (002) reflections

219

(Fig. 1b) and the subtle difference of d(002)-spacing (Table 1). The d(002)-spacing is the standard

220

that represents the crystallographic structure of graphite. Several authors [11,19,31,36] have

221

used the value of d(002)-spacing ˂ 0.336 nm to classify the highly ordered graphite formed

222

under both contact and regional metamorphism. In the current study, this value is also

223

applicable to the highly ordered CDNG samples. The d(002)-spacing within the range of 0.336

224

˗ 0.337 nm represent the transitional phases between semi-graphite and highly ordered

225

graphite in the graphite field. These transitional phases are the pre-phases of highly ordered

226

graphite, thus they can be classified as ordered graphite. The samples LT-2 ~ LT-5 having the

227

d(002)-spacing ˂ 0.336 nm (Table 1) belong to the highly ordered CDNG. The samples SHL

228

and LT-1 have relatively larger d(002)-spacing of 0.3365 and 0.3364 nm, respectively, and their

229

(002) reflections of XRD patterns are broader and more asymmetric compared to the samples

230

LT-2 ~ LT-5, which are classified as ordered CDNG. However, for the graphite samples, the

231

difference of their d(002)-spacing is subtle. Measurement error such as caused by instrument

232

aging may result in the 2θ deviation of XRD patterns. Therefore, it is necessary to use internal

233

standard to calibrate the 2θ of (002) reflection for obtaining the precise d(002)-spacing. The

234

asymmetry index (AI) of (002) reflection is proposed to evaluate the degree of ordering of

235

crystalline structure of the CDNG based on the concept for the asymmetry index of (001)

236

reflection of illite used for evaluating its degree of crystallinity [37]. Fig. 2a shows that the AI

237

of (002) reflection of CDNG is calculated using left FWHM divided by right FWHM due to

238

the asymmetric nature of (002) reflection caused by the turbostratic structure and structure 11

239

defect in the crystalline structure. For the AI value, high value reflects more symmetric (002)

240

reflection of CDNG, which indicates high degree of ordering of crystalline structure. The

241

relationship between AI and d(002)-spacing presents a good correlation in the zone of

242

semi-graphite to graphite (Fig. 2b). Based on their linear relationship, the AI within the range

243

of 0.50 ˗ 0.63 and the range of AI ˃ 0.63 correspond to the d(002)-spacing within the range of

244

0.336 ˗ 0.337 nm and the range of d(002)-spacing ˂ 0.336 nm, respectively. Therefore, the AI

245

within the range of 0.50 ˗ 0.63 and the range of AI ˃ 0.63 can be used to delimit ordered

246

CDNG and highly ordered CDNG in the graphite field. The sample SL (onset phase of CDNG)

247

deviates the correlation (Fig. 2b) probably caused by the high heterogeneity of its crystalline

248

structure.

249 250 251 252

Fig. 2. (a) Illustration of how to calculate asymmetry index (AI) using the FWHM of (002) reflection of CDNG in case of sample LT-1, where r and r refer to the right FWHM and left FWHM, respectively. (b) AI versus d(002)-spacing in the zone of semi-graphite to graphite (data from Table 1). 1

2

253

The highly ordered CDNG (samples LT-2 ~ LT-5) show that the (002) reflections are still

254

slightly asymmetric, which is also observed for the synthetic graphite using Pennsylvania

255

anthracite as starting material [38], indicating dislocation exists along parallel stacking of

256

aromatic layers. The (110) reflections are almost symmetric for highly ordered CDNG as

257

shown in Fig. 1d, suggesting the high degree of ordering of graphite lamellae in lateral 12

258

direction. The relationships of calculated Lc and La versus d(002)-spacing of CDNG (Fig. 3)

259

show that both Lc and La continuous increase during graphitization. In the stage of onset

260

phase to semi-graphite, the Lc and La increase slowly. When entering graphite stage, both

261

crystallite size increase steeply. The growth trend of crystallite size of CDNG during

262

graphitization agrees well with that of carbonaceous material in sedimentary rocks during

263

contact metamorphism [19]. This suggests a continuous transformation of anthracite to

264

CDNG without major breaks, supporting the idea of a continuous graphitization process

265

[9,17,20].

266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282

Fig. 3. Relationships between crystallite size (Lc and La) and d(002)-spacing of CDNG. Table 1 The distance of graphite samples to the pluton and the d(002)-spacing, calculated crystallite size Lc and La, and asymmetry index (AI) of (002) reflection of CDNG. Sample

Distance to pluton (m)

d(002) (nm)

Lc (nm)

Sd

La (nm)

Sd

AI

SL

1000

0.3395

9.61

0.061

34.34

7.377

0.40

BC-1

600

0.3383

12.12

0.209

74.93

1.228

0.39

BC-2

600

0.3371

22.03

1.012

86.91

4.144

0.45

SHL

300

0.3365

26.52

1.920

119.92

1.430

0.51

LT-1

200

0.3364

31.77

0.774

129.88

1.235

0.52

LT-2

150

0.3358

81.12

1.730

176.47

0.69

LT-3

100

0.3356

123.23

1.331

210.35

0.70

LT-4

˂ 100

0.3355

121.68

3.875

212.49

0.70

LT-5

˂ 100

0.3355

129.14

4.386

239.36

0.71

Note: Sd is the standard deviation from the mean of Lc and La calculated from the FWHM and Bragg angles of (002) and (110) peaks fitted three times. 13

283

3.2. Raman spectroscopy

284

Raman spectroscopy is very sensitive to the degree of ordering of carbonaceous materials,

285

being a commonly used method to investigate the structural evolution of coal during

286

coalification and graphitization [12,39-41]. Fig. 4 and Fig. 5 show the Raman spectra of

287

CDNG and peak fitting of Raman spectrum in the first order region in case of sample BC-2.

288

The Raman parameters such as band center, FWHM, and percentage of area obtained from the

289

peak fitting are listed in Table 2. In the first order region (Fig. 4), two common bands at 1320

290

˗ 1350 cm-1 and ~1580 cm-1 corresponding to the D1 and G bands, respectively, are observed.

291

The D2 band appears as a shoulder on high wavelength side (~1620 cm-1) of the G band

292

originating from sp2 bond stretching mode of graphitic carbon in a hexagonal lattice [42]. The

293

D1 and D2 bands are attributed to the structural defect of graphitic carbon [43]. For samples

294

SL, BC-1, and BC-2, the disordered D4 band at 1170 - 1178 cm-1 deriving from sp2-sp3 bonds

295

or C-C and C=C stretching vibrations of polyene-like structures is also observed [44],

296

indicating their less ordered structure. The D4 bands in the Raman spectra of samples SL,

297

BC-1, and BC-2 are hard to be detected in Fig. 4 plotted in stacked style. It can be observed in

298

their individual spectrum in the first order region as shown in Fig. 5 and Fig. S2 in

299

Supplementary Material. As the degree of ordering increases, the D4 band gradually

300

disappears (Fig. S2 in Supplementary Material), the G band becomes sharper, and the D1

301

and D2 bands become weaker (Fig. 4). The Raman spectra of progressively graphitized

302

carbonaceous materials subjected to regional and contact metamorphism and the synthetic

303

graphite with different structural order prepared from anthracite also show the similar

304

tendency [17,19,45]. 14

305

The Raman bands in the second order region attributed to the overtone and combination

306

of disorder-induced bands provide better information about three-dimensional order of

307

carbonaceous materials [46-49]. The prominent feature in the second order region is the

308

~2700 cm-1 band (Fig. 4), which corresponds to the overtone of the D1 band [43]. This band

309

shifts to high frequency with the structural order of CDNG increases as shown in Fig. 4.

310

Additionally, the ~2700 cm-1 band is a single sharp band for samples SL ~ LT-1, and splits for

311

samples LT-2 ~ LT-5. The less intense band at ~2900 cm-1 assigned to the D1 + G bands exists

312

in the Raman spectra of samples SL ~ LT-1 [43], and gradually disappears for samples LT-2 ~

313

LT-5. The splitting of the 2700 cm-1 band and progressive disappearance of the 2900 cm-1

314

band indicate the formation of ordered three-dimensional crystalline structure of graphite

315

(samples LT-2 ~ LT-5) [23], agreeing with the XRD analysis. The Raman spectra of

316

carbonaceous materials in Schistes Lustrés formation, Western Alps subjected to

317

high-pressure metamorphism show the same evolution during graphitization [17].

15

318 319

Fig. 4. Raman spectra of CDNG in the first and second order regions.

320 321

Fig. 5. Peak fitting of Raman spectrum in the first order region in case of sample BC-2.

322

The FWHM of G band was evidenced to be able to distinguish semi-graphite and

323

graphite [17,21]. Here, the D1/G area ratio and FWHM of G band are used to evaluate the

324

degree of ordering of crystalline structure of CDNG. The relationships between D1/G area 16

325

ratio and FWHM of G band and d(002)-spacing of CDNG for the zone of semi-graphite to

326

graphite (Fig. 6) are analogous to the relationship between AI and d(002)-spacing, both D1/G

327

area ratio and FWHM of G band have high correlation with d(002)-spacing for the zone of

328

semi-graphite to graphite. The sample SL also deviates the correlation as in plot of

329

Asymmetry index versus d(002)-spacing (Fig. 2b), probably due to the high heterogeneity of its

330

crystalline structure. According to their linear relationships, the D1/G area ratio within the

331

range of 0.85 ˗ 1.70 and the G FWHM within the range of 23.8 ˗ 30.3 cm-1 correspond to the

332

d(002)-spacing within the range of 0.336 ˗ 0.337 nm, which, therefore, can be used to classify

333

the ordered CDNG in the graphite field. The range of D1/G area ratio ˂ 0.85 and G FWHM ˂

334

23.8 cm-1 correspond to the range of d(002)-spacing ˂ 0.336 nm, which classify the highly

335

ordered CDNG in the graphite field.

336 337 338

Fig. 6. (a) D1/G area ratio and (b) FWHM of G band of Raman parameters versus d(002)-spacing of CDNG in the zone of semi-graphite to graphite.

17

Table 2. Raman parameters (mean values and standard deviation of the band center, FWHM, and percentage of area) obtained from the peak fitting of Raman spectrum per sample.

D4 Sample

Center (cm-1)

Sd

FWHM (cm-1)

SL

1172

9.59

BC-1

1171

BC-2

1178

D1 Sd

Area (%)

Sd

Center (cm-1)

135

12.60

15.02

112

3.92

114

Sd

FWHM (cm-1)

6.32

2.51

1328

2.07

8.83

4.04

0.46

1324

8.21

4.47

0.67

G Sd

Area (%)

Sd

Center (cm-1)

Sd

FWHM (cm-1)

45

5.94

62.40

3.67

1580

0.84

1.02

52

2.03

64.74

4.51

1576

1325

0.99

47

0.97

60.14

4.51

SHL

1326

2.62

40

1.27

48.75

LT-1

1348

0.92

33

0.87

LT-2

1349

LT-3

1347

0.03

37

0.33

51

LT-4

1350

0.74

LT-5

1348

0.68

D2 Sd

Area (%)

38

3.01

4.11

38

1575

0.47

3.16

1574

39.62

1.34

0.53

33.77

5.44

39.02

40

1.53

35

0.77

D1/G area ratio

Sd

1.17

2.55

0.48

6.02

1.73

2.59

0.52

5.35

8.64

3.26

2.52

0.62

18

1.68

5.13

0.88

1.08

0.17

2.53

14

6.11

2.88

0.57

0.69

0.03

1618

3.37

16

3.24

1.95

1.34

0.53

0.04

1618

3.69

33

12.11

3.71

0.77

0.68

0.08

4.40

1618

5.60

10

0.61

1.54

1.35

0.50

0.08

2.20

1617

3.10

23

10.01

3.27

1.76

0.45

0.02

Sd

Center (cm-1)

Sd

FWHM (cm-1)

Sd

Area (%)

Sd

24.94

3.68

1612

2.33

26

2.74

7.69

3.88

25.47

3.78

1607

1.82

24

4.06

33

2.53

24.49

3.94

1607

0.24

28

3.94

25

0.56

45.44

4.60

1610

2.65

1580

1.11

23

1.17

57.50

0.77

1618

1.14

1580

3.11

1578

0.27

21

0.17

64.28

2.47

0.54

22

0.04

57.27

2.34

32.67

3.05

1579

0.12

21

0.68

65.79

29.89

0.44

1579

0.28

21

0.00

66.84

Note: Sd is the standard deviation from the mean calculated by all spectra obtained within the same sample.

18

3.3. HRTEM HRTEM coupled with selected area electron diffraction pattern (SAED) can directly image the microtexture and structural order of carbonaceous materials at nanoscale [17,50-52]. It complements the powdered XRD measurement that gives averaged structural information of carbonaceous materials and is unsuccessful in recognizing the detailed microtexture especially for structurally and microtexturally heterogeneous ones under low-grade metamorphism. The HRTEM images and corresponding SAED patterns of samples SL, BC-2, LT-1, and LT-4 representing the onset phase of CDNG, semi-graphite, transitional phase between semi-graphite and highly ordered CDNG (ordered CDNG), and highly ordered CDNG, respectively, are displayed (Figure 7) to illustrate the structural evolution of CDNG at nanoscale during graphitization. In sample SL, the microporous-like carbonaceous material presenting structural heterogeneity is observed. The basic structural units (BSU) in sample SL are composed of several aromatic layers that have short in-plane dimensions and orient in wavy pattern. The BSUs random arrange, and have a tendency to link laterally. On the local scale, some BSUs have grown into long-range wrinkled graphite lamellae. The SAED pattern shows broad rings, suggesting less structural order of onset phase of graphite. Then polymerization process links the adjacent BSUs to form long-range wrinkled layers stacked in c-axis direction orienting in wavy pattern (sample BC-2). Additionally, the moderate organized carbonaceous material presenting concentric microtexture (onion rings) is also observed, where the aromatic layers in the outer part of onion ring are long and waved, while the aromatic layers in core remain poorly organized. This kind of carbonaceous material was commonly recognized in the metasediments of marine origin [1,17]. Beyssac et al. 19

systematically investigated the graphitization of this type of carbonaceous material [17]. The aromatic layers in the outer part of onion rings reoriented themselves and became longer, and the diameter of rings gradually increased with graphitization proceeding. Then the ordered aromatic layers in the out part of onion rings were dissociated from the internal core of the concentric structure, and subsequently formed triperiodic graphite. However, a small amount of carbonaceous materials with concentric microtexture were recognized in our samples, and the structural evolution of concentric microtexture phase transformed to graphite lamellae was not detected. The SAED pattern of sample BC-2 shows spots and ring patterns, implying presence of turbostratic BSUs in the crystallite of semi-graphite. The HRTEM image and SAED pattern of semi-graphite from greenschist facies metamorphic rocks also show the long-range wrinkled layers with blurred rings in SAED pattern, suggesting less structural order of semi-graphite [21]. With graphitization proceeding, the wrinkled layers become flat and the turbostratic BSUs gradually disappear, simultaneously, the stacking of ordered aromatic layers is increased (Sample LT-1). The SAED rings are thin, while the spot patterns are not fully separated, indicating structural defect exists in graphite crystallite as reflected by the asymmetric (002) reflection of its XRD pattern. The dislocation in stacking and in-plane defect of flat aromatic layers decrease to reach a highly ordered graphite structure with advanced graphitization. The spot patterns are fully separated (sample LT-4), indicating the high structural order of CDNG is achieved. Even in the highly ordered graphite, a few stacking defects still exist in graphite lamellae due to the weak bonding between aromatic layers [1,19]. The HRTEM investigation suggests that the structural evolution of CDNG during natural graphitization is analogous to the industrial procedures [5,53]. 20

Fig. 7. HRTEM images and SAED patterns of CDNG with increasing structural order.

4. Conclusion 21

In this study, the crystalline structure of graphite transformed from coal subjected to contact metamorphism through massive magmatic intrusion were investigated using XRD, Raman spectroscopy, and HRTEM. The structural order of CDNG increases by the position relative to the intrusive body. There are transitional phases with different structural order existing in the graphite field due to subjected to different metamorphic grade. The d(002)-spacing is the standard that represents the crystallographic structure of graphite. The asymmetry index (AI) of (002) reflection of XRD patterns, Raman parameters D1/G area ratio and FWHM of G band correlate well with d(002)-spacing in the field of semi-graphite to graphite. Based on their linear relationships, the AI within the range of 0.50 ˗ 0.63, the D1/G area ratio within the range of 0.85 ˗ 1.70, and the G FWHM within the range of 23.8 ˗ 30.3 cm-1 correspond to the d(002)-spacing within the range of 0.336 ˗ 0.337 nm, which can be used to classify the ordered CDNG in the graphite field. The range of AI ˃ 0.63, D1/G area ratio ˂ 0.85, and G FWHM ˂ 23.8 cm-1 correspond to the range of d(002)-spacing ˂ 0.336 nm, which classify the highly ordered CDNG in the graphite field. The structural evolution of CDNG during nature graphitization is analogous to the laboratory formation of graphite as expected transformation process [5,53]. At the beginning of graphitization, the BSUs composed of stacking aromatic layers arrange face to face and tend to link laterally. Then polymerization process promotes the link between adjacent BSUs forming long-range wrinkled layers. The wrinkled layers become flat and the turbostratic structure gradually disappears with further graphitization. Finally, the dislocation in stacking and in-plane defect of flat aromatic layers decreases to reach highly ordered graphite structure.

22

Appendix A. Supplementary Material The peak fitting of (002) and (110) reflections of XRD of coal derived natural graphite, and the obtained parameters d(002)-spacing, peak positions (2θ), and full width at half maximum (FWHM). Peak fitting of Raman spectra of samples SL, BC-1, BC-2, and SHL in the first order region.

Acknowledgments Research reported in this publication was supported by the National Natural Science Foundation of China (41802189, 41672150) and the China Postdoctoral Science Foundation (2017M620956). The authors would like to thank Dr. Xisheng Lin for his help in peak fitting of XRD reflections using Jade 5 package and Dr. Yan Fan at Beijing Center for Physical & Chemical Analysis for assistance with HRTEM observation. We also thank the three anonymous reviewers, whose comments are very helpful to improve the quality of the manuscript.

23

References [1] P.R. Buseck, B.-J. Huang. Conversion of carbonaceous material to graphite during metamorphism. Geochim. Cosmochim. Acta., 49 (1985), pp. 2003-2016. [2] F.J. Luque, J.-M. Huizenga, E. Crespo-Feo, H. Wada, L. Ortega, J.F. Barrenechea. Vein graphite deposits: Geological settings, origin, and economic significance. Miner. Deposita., 49 (2014), pp. 261-277. [3] O. Beyssac, D. Rumble. Graphitic carbon: A ubiquitous, diverse, and useful geomaterial. Elements, 10 (2014), pp. 415-420. [4] F.j. Luque, J.D. Pasteris, B. Wopenka, M. Rodas, J.F. Barrenechea. Natural fluid-deposited graphite mineralogical characteristics and mechanisms of formation. Am. J. Sci., 298 (1998), pp. 471-498. [5] P.R. Buseck, O. Beyssac. From organic matter to graphite: Graphitization. Elements, 10 (2014), pp. 421-426. [6] B. Kwiecińska, H.I. Petersen. Graphite, semi-graphite, natural coke, and natural char classification—ICCP system. Int. J. Coal Geol., 57 (2004), pp. 99-116. [7] K. Li, S.M. Rimmer, Q. Liu. Geochemical and petrographic analysis of graphitized coals from central Hunan, China. Int. J. Coal Geol., 195 (2018), pp. 267-279. [8] Z. Zheng, J. Zhang, J.Y. Huang. Observations of microstructure and reflectivity of coal graphites for two locations in China. Int. J. Coal Geol., 30 (1996), pp. 277-284. [9] E.S. Grew. Carbonaceous material in some metamorphic rocks of New England and other areas. J. Geol., 82 (1974), pp. 50-73. [10] T. Itaya. Carbonaceous material in pelitic schists of the Sanbagawa metamorphic belt in central Shikoku, Japan. Lithos, 14 (1981), pp. 215-224. [11] G.-F. Wang. Carbonaceous material in the Ryoke metamorphic rocks, Kinki district, Japan. Lithos, 22 (1989), pp. 305-316. [12] B. Kwiecinska, I. Suárez-Ruiz, C. Paluszkiewicz, S. Rodriques. Raman spectroscopy of selected carbonaceous samples. Int. J. Coal Geol., 84 (2010), pp. 206-212. [13] B. Wopenka, J.D. Pasteris. Raman spectra of graphite as indicators of degree of metamorphism. Can. Mineral., 29 (1991), pp. 1-9. [14] K. Angoni. Remarks on the structure of carbon materials on the basis of Raman spectra. Carbon, 31 (1993), pp. 537-547. [15] E.V. Ulyanova, A.N. Molchanov, I.Y. Prokhorov, V.G. Grinyov. Fine structure of Raman spectra in coals of different rank. Int. J. Coal Geol., 121 (2014), pp. 37-43. [16] N.K. Lünsdorf. Raman spectroscopy of dispersed vitrinite — methodical aspects and correlation with reflectance. Int. J. Coal Geol., 153 (2016), pp. 75-86. [17] O. Beyssac, J.-N. Rouzaud, B. Goffé, F. Brunet, C. Chopin. Graphitization in a high-pressure, low-temperature metamorphic gradient: A Raman microspectroscopy and HRTEM study. Contrib. Mineral. Petrol., 143 (2002), pp. 19-31. [18] W. Dun, L. Guijian, S. Ruoyu, C. Shancheng. Influences of magmatic intrusion on the macromolecular and pore structures of coal: Evidences from Raman spectroscopy and atomic force microscopy. Fuel, 119 (2014), pp. 191-201. [19] Y. Nakamura, J. Akai. Microstructural evolution of carbonaceous material during graphitization in the Gyoja-yama contact aureole: HRTEM, XRD and Raman spectroscopic study. J. Miner. Petrol. Sci., 108 (2013), pp. 131-143. [20] B. Wopenka, J.D. Pasteris. Structural characterization of kerogens to granulite-facies graphite: Applicability 24

of Raman microprobe spectroscopy. Am. Mineral., 78 (1993), pp. 533-557. [21] G. Rantitsch, W. Lämmerer, E. Fisslthaler, S. Mitsche, H. Kaltenböck. On the discrimination of semi-graphite and graphite by Raman spectroscopy. Int. J. Coal Geol., 159 (2016), pp. 48-56. [22] Y. Nakamura, K. Oohashi, T. Toyoshima, M. Satish-Kumar, J. Akai. Strain-induced amorphization of graphite in fault zones of the Hidaka metamorphic belt, Hokkaido, Japan. J. Struct. Geol., 72 (2015), pp. 142-161. [23] S. Rodrigues, I. Suárez-Ruiz, M. Marques, I. Camean, D. Flores. Microstructural evolution of high temperature treated anthracites of different rank. Int. J. Coal Geol., 87 (2011), pp. 204-211. [24] S. Rodrigues, M. Marques, I. Suárez-Ruiz, I. Camean, D. Flores, B. Kwiecinska. Microstructural investigations of natural and synthetic graphites and semi-graphites. Int. J. Coal Geol., 111 (2013), pp. 67-79. [25] M. Wissler. Graphite and carbon powders for electrochemical applications. J. Power Sources., 156 (2006), pp. 142-150. [26] X. Zhou, Y. He, H. Liu. Geological characteristics of coal and graphite in Hanpoao mining area in northern section of Lianshao coalfield of Hunan province. Coal. Geol & Explor., 45 (2017), pp. 9-13. [27] C. Li, D. Wang, L. Zhou, H. Zhao, X. Li, W. Qu. Study on the Re-Os isotope composition of graphite from the Lutang graphite deposit in Hunan province. Rock. Miner. Anal., 36 (2017), pp. 297-304. [28] J. Biscoe, B.E. Warren. An X-ray study of carbon black. J. Appl. Phys., 13 (1942), pp. 364-371. [29] R. Hinrichs, M.T. Brown, M.A.Z. Vasconcellos, M.V. Abrashev, W. Kalkreuth. Simple procedure for an estimation of the coal rank using micro-Raman spectroscopy. Int. J. Coal Geol., 136 (2014), pp. 52-58. [30] N.K. Lünsdorf, I. Dunkl, B.C. Schmidt, G. Rantitsch, H. von Eynatten. Towards a higher comparability of geothermometric data obtained by Raman spectroscopy of carbonaceous material. Part I: Evaluation of biasing factors. Geostand. Geoanal. Res., 38 (2014), pp. 73-94. [31] D. Nakamura. Comparison and interpretation of graphitization in contact and regional metamorphic rocks. Isl. Arc., 4 (1995), pp. 112-127. [32] H.A. Wilhelm, B. Croset, G. Medjahdi. Proportion and dispersion of rhombohedral sequences in the hexagonal structure of graphite powders. Carbon, 45 (2007), pp. 2356-2364. [33] H. Shi, J. Barker, M.Y. Saïdi, R. Koksbang. Structure and lithium intercalation properties of synthetic and natural graphite. J. Electrochem. Soc., 143 (1996), pp. 3466-3472. [34] A. Deurbergue, A. Oberlin, J.H. Oh, J.N. Rouzaud. Graphitization of korean anthracites as studied by transmission electron microscopy and X-ray diffraction. Int. J. Coal Geol., 8 (1987), pp. 375-393. [35] M.S. Nyathi, C.B. Clifford, H.H. Schobert. Characterization of graphitic materials prepared from different rank Pennsylvania anthracites. Fuel, 114 (2013), pp. 244-250. [36] C.A. Landis. Graphitization of dispersed carbonaceous material in metamorphic rocks. Contrib. Mineral. Petrol., 30 (1971), pp. 34-45. [37] L. Ren, Y. Chen. On the division of diagenesis stages according to the transformation of clay minerals Oil & Gas. Geol., 5 (1984), pp. 325-334. [38] R.M. Bustin, J.N. Rouzaud, J.V. Ross. Natural graphitization of anthracite: Experimental considerations. Carbon, 33 (1995), pp. 679-691. [39] M. Baysal, A. Yürüm, B. Yıldız, Y. Yürüm. Structure of some western Anatolia coals investigated by FTIR, Raman, 13C solid state NMR spectroscopy and X-ray diffraction. Int. J. Coal Geol., 163 (2016), pp. 166-176. [40] O.O. Sonibare, T. Haeger, S.F. Foley. Structural characterization of Nigerian coals by X-ray diffraction, Raman and FTIR spectroscopy. Energy, 35 (2010), pp. 5347-5353. 25

[41] M. Marques, I. Suárez-Ruiz, D. Flores, A. Guedes, S. Rodrigues. Correlation between optical, chemical and micro-structural parameters of high-rank coals and graphite. Int. J. Coal Geol., 77 (2009), pp. 377-382. [42] F. Tuinstra, J.L. Koenig. Raman spectrum of graphite. J. Chem. Phys., 53 (1970), pp. 1126-1130. [43] M.A. Pimenta, G. Dresselhaus, M.S. Dresselhaus, L.G. Cançado, A. Jorio, R. Saito. Studying disorder in graphite-based systems by Raman spectroscopy. Phys. Chem. Chem. Phys., 9 (2007), pp. 1276-1290. [44] A. Sadezky, H. Muckenhuber, H. Grothe, R. Niessner, U. Pöschl. Raman microspectroscopy of soot and related carbonaceous materials: Spectral analysis and structural information. Carbon, 43 (2005), pp. 1731-1742. [45] D. González, M.A. Montesmorán, A. Isabel Suárezruiz, A.B. Garcia. Structural characterization of graphite materials prepared from anthracites of different characteristics:  A comparative analysis. Energy Fuels, 18 (2004), pp. 365-370. [46] R.J. Nemanich, S.A. Solin. First- and second-order Raman scattering from finite-size crystals of graphite. Phys. Rev. B., 20 (1979), pp. 392-401. [47] Y. Kawashima, G. Katagiri. Fundamentals, overtones, and combinations in the Raman spectrum of graphite. Phys. Rev. B., 52 (1995), pp. 10053-10059. [48] Y. Hishiyama, H. Irumano, Y. Kaburagi, Y. Soneda. Structure, Raman scattering, and transport properties of boron-doped graphite. Phys. Rev. B., 63 (2001), pp. 245406. [49] Y. Sato, M. Kamo, N. Setaka. Raman spectra of carbons at 2600–3300 cm−1 region. Carbon, 16 (1978), pp. 279-280. [50] J.N. Rouzaud, A. Oberlin. Structure, microtexture, and optical properties of anthracene and saccharose-based carbons. Carbon, 27 (1989), pp. 517-529. [51] G.N. Okolo, H.W.J.P. Neomagus, R.C. Everson, M.J. Roberts, J.R. Bunt, R. Sakurovs, J.P. Mathews. Chemical–structural properties of south African bituminous coals: Insights from wide angle XRD–carbon fraction analysis, ATR–FTIR, solid state 13C NMR, and HRTEM techniques. Fuel, 158 (2015), pp. 779-792. [52] J.P. Mathews, A. Sharma. The structural alignment of coal and the analogous case of Argonne Upper Freeport coal. Fuel, 95 (2012), pp. 19-24. [53] A. Oberlin. Carbonization and graphitization. Carbon, 22 (1984), pp. 521-541.

26

Highlights: Transitional phases with different structural order exist in classified coal derived natural graphite field. XRD and Raman parameters can evaluate structural order of coal derived natural graphite. Structural evolution of coal derived natural graphite during graphitization is analogous to laboratory formation of graphite. Transformation of anthracite to coal derived natural graphite is a continuous graphitization process.

Declaration of interests ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: