The effect of heating on mineral composition and grain size distribution of flux calcined porcelanites from the Gafsa-Metlaoui basin, southwestern Tunisia

Accepted Manuscript The effect of heating on mineral composition and grain size distribution of flux calcined porcelanites from the Gafsa-Metlaoui basin, southwestern Tunisia Raja Saidi, Ali Tlili, Fakher Jamoussi PII:

S1464-343X(16)30316-8

DOI:

10.1016/j.jafrearsci.2016.09.022

Reference:

AES 2680

To appear in:

Journal of African Earth Sciences

Received Date: 2 April 2016 Revised Date:

14 August 2016

Accepted Date: 21 September 2016

Please cite this article as: Saidi, R., Tlili, A., Jamoussi, F., The effect of heating on mineral composition and grain size distribution of flux calcined porcelanites from the Gafsa-Metlaoui basin, southwestern Tunisia, Journal of African Earth Sciences (2016), doi: 10.1016/j.jafrearsci.2016.09.022. 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.

ACCEPTED MANUSCRIPT 1

The effect of heating on mineral composition and grain size distribution of

2

flux calcined porcelanites from the Gafsa-Metlaoui basin, Southwestern

3

Tunisia

5 6

1

7

Faculty of Science, PB 802, 3038, Sfax Univ. of Sfax, Tunisia,

8

2

9

Emails: [email protected] , [email protected]

RI PT

Raja Saidi 1, Ali Tlili 1 and Fakher Jamoussi 2

: Laboratory Géoressources, Materials, Environment and Global Change, Department of Earth Sciences,

: Technopole Borj Cedria, PB 273, 8020 Soliman, Tunisia,

M AN U

10

SC

4

Abstract.

12

The porcelanite rock of Ypresian phosphatic series of the Gafsa-Metlaoui basin (south-

13

western Tunisia), is composed mainly of opal CT, and presents a variable percentage of

14

carbonates and fibrous clays. This rock is treated with flux calcination at different

15

temperatures in order to prepare a specific filter aid for cleaning melting sulfur which can be

16

used for the production of sulfuric acid. This work presents the effect of heating on the

17

mineralogy and grain size distribution of carbonate-rich porcelanite (Tm1) and clay-rich

18

porcelanite (Gh) compared to flux calcined silica-rich porcelanite (CHM3) and diatomaceous

19

filtration aids. The porcelanite samples used in this work come from three localities of the

20

Gafsa-Metlaoui basin: Kef El Ghis (Gh), Tamarza (Tm1) and Mides (CHM3). Flux

21

calcination at 1000°C provokes a mineralogical transformation on carbonate-rich porcelanite

22

samples. The opal CT transforms to opal C and becomes neater and more stable. The Thermal

23

treatment of porcelanite (Tm1) incites also the apparition of new peaks of wollastonite.

24

However, the structural change of opal CT to opal C by heat treatment is blocked for flux

25

calcination of clay-rich porcelanite. The opal CT of fluxing clay-rich porcelanite becomes

26

more ordered without significant change to opal C. The difference between fluxing carbonate-

AC C

EP

TE D

11

1

ACCEPTED MANUSCRIPT rich porcelanite (Tm1) and fluxing clay-rich porcelanite (Gh) appears also with granulometric

28

distribution histogram of the tow heated samples. All raw samples have unimodal

29

granulometric distribution (1-100µm). After calcination with alkaline flux at 1000°C fluxing

30

carbonate-rich porcelanite displays bimodal granulometric distribution and a new mode

31

appears systematically, between 0.1 µm and 1 µm. This occurs for fluxing silica-riche

32

porcelanite and diatomaceous filtration aids as well and corresponds to the opal C formed

33

after heat treatment. Whereas fluxing clay-rich porcelanite present trimodal granulometric

34

distribution and a third mode appears (100-300µm), which due to silica glass phase. Since, the

35

granulometric rearrangement of porcelanite during thermal treatment may due to

36

mineralogical transformation of opal CT to opal C and crystal grow.

37

Key words: Porcelanite, Gafsa-Metlaoui basin, thermal treatment, mineralogy, grain size

38

distribution

39

Introduction

40

The biogenic silica rocks as diatomite, chert and porcelanite are among the most studied rocks

41

for scientific and economic reasons. The scientific reason poses the problem of genesis of

42

these rocks: origin of the silica, hydrothermal contribution (Jun et al., 2000; Kametaka et al.,

43

2005; Zhou et al., 2006, He et al., 2011; Huang et al., 2013;) and diagenesis transformation of

44

the silica phases (Hesse, 1988; Dixit and Van Cappellen, 1998; Dirk, 2000; Bernoullia and

45

Gunzenhauser, 2001; He et al., 2011;). On the other hand, these rocks can be used in different

46

domains: like filter in several domains (Martinovic et al., 2006) or in industry (Erdem et al.,

47

2005) and in electronics. These rocks are composed of opal A, opal CT and/or quartz. Their

48

use in industrial filtration requires previous purification and thermal treatment in order to get

49

a silica phase rich in cristobalite. Therefore some authors studied the effect of thermal

50

treatment on the mineralogical and physical characterestics of these rocks (Kouteren, 1994;

51

Hadjadj et al., 2005 ; Hadjar et al 2008 ; Yılmaz and Ediz, 2008 ; Ediz et al., 2010 ; Tlili et al.,

AC C

EP

TE D

M AN U

SC

RI PT

27

2

ACCEPTED MANUSCRIPT 2012; Arasuna et al., 2013). The thermal treatment studies confirm generally the formation of

53

the opal C (opal cristobalite) and/or cristobalite above 850°C up to 1400°C. However, the

54

presence of certain impurities or other mineral phases let the structural evolution of silica to

55

opal cristobalite and/or cristobalite become more difficult during heat treatment. On the other

56

hand, many studies have focused on the mineralogical evolution of thermal treatment of silica

57

but a few ones inspect the granulometric distribution transformation after calcinations (Saidi

58

et al., 2012).

59

The porcelanite rock of Ypresian phosphatic series of the Gafsa-Metlaoui basin, composed

60

mainly of biogenic opal CT, is treated with flux calcination at 600°C, 800°C and 1000°C in

61

order to prepare a specific filter aid of melting sulfur filter used for the production of sulfuric

62

acid. This work concerns the thermal treatment of carbonate-rich porcelanite and clay-rich

63

porcelanite compared to flux-calcined silica-rich porcelanite (CHM3). This work presents

64

also the effect of heat treatment on the variation of the granulometric distribution of those flux

65

calcined

66

Chemical Group, These diatomites are imported from French clarcel, Algerian Kieselgur and

67

Spanish diatomite.

68

2. Geological setting and lithological description

69

The Ypresian phosphatic series of the Gafsa-Metlaoui basin corresponds to Chouabine

70

formation. This series contains nine distinct phosphate beds that alternate with the layers of

71

the marls, limestone, carbonates and porcelanite intercalations (Fig 1). Phosphatic beds are

72

named layer CIX to C0. In general, porcelanite intercalation is located between CVI and CVII

73

layers. The number of porcelanite intercalation increases significantly in the western part of

74

the basin (Burollet, 1956; Sassi, 1974; Burollet and Oudin, 1980; Belayouni, 1983; Chaabani,

75

1995), which represents three intercalations in Tarmaza and four intercalations in Mides (Tlili

76

et al., 2012 ; Saidi et al., 2012; Saidi, 2015). The thickness of porcelanite intercalation

M AN U

SC

RI PT

52

AC C

EP

TE D

porcelanite compared with three industrial diatomite filters, used by Tunisian

3

ACCEPTED MANUSCRIPT exceeds 10 m in Mides section. The porcelanite rock of Ypresian phosphatic series of Gafsa-

78

Metlaoui basin is composed mainly of biogenic opal CT (Belayouni, 1983; Chaabani, 1995;

79

Felhi et al., 2008; Henchiri, 2007; Saidi et al., 2014; Sassi, 1974; Tlili et al., 2010, Haj Ahmed

80

et al., 2014).

81

3. Materials and methods

82

3.1. Materials

83

The porcelanite samples used in this work come from three localities of the Gafsa-Metlaoui

84

basin: Kef El Ghiss (Gh), Tamarza (Tm1) and Mides (CHM3) (Fig. 2). The choice of

85

porcelanite samples for the thermal treatment is based on the high percentages of silica and on

86

the mineralogical composition (CT opal associated with phyllosilicates or carbonates). The

87

carbonate-rich porcelanite (Tm1) is sampled from intercalation VI-VIII in Tamarza and the

88

clay-rich porcelanite (Gh) is sampled from intercalation VI-VII in Kef El Ghiss (Fig. 1). The

89

behavior of these samples during thermal treatment is compared with the result obtained with

90

flux calcined porcelanite (CHM3) and industrial diatomite filters imported from French

91

clarcel, Algerian Kieselgur and Spanish diatomite. CHM3 sample comes from intercalation

92

CV-CVI of porcelanite in Mides section. This sample is characterized by high silica content

93

(SiO2% = 84,9%) (Tlili et al., 2012).

94

3.2. Thermal treatment

95

The thermal treatment of Tm1 and Gh porcelanite samples with 5% of alkaline flux (Na2CO3)

96

was achieved at 600 °C and 800 °C and at 1000°C for 1,5 h, 4 h and 6 h (Table 1) in a

97

refractory brick container, using an electric furnace. The granulometry of raw porcelanite

98

starting powder was maintained lower than 100 µm and displays unimodal distribution

99

between 1-100 µm. The result was the outcome of the study of these two samples discussed

100

with the one obtained from calcined porcelanite (CHM3) that has been studied by Tlili et al

101

(2012). This sample (CHM3) is heated also with 5% alkaline flux at 400 °C and 800 °C for 3

AC C

EP

TE D

M AN U

SC

RI PT

77

4

ACCEPTED MANUSCRIPT h, and at 1000 °C for 1.5 h, 3 h and 6 h.

103

3.3. Chemical analysis

104

Chemical analysis of the raw porcelanite was determined by X-ray fluorescence (XRF)

105

technique on fused glass pearls using type Oasis 9900, Thermo-Fisher, in the Control and

106

Quality Laboratory at the Gabes Cement Company. The pearls are formed by a mixture of 14

107

g of sample and 2 g of cellulose and dried at 110 °C during 6 h. A pearl of international

108

standard (norm IN 196/2 or NT 47.30-2) is used for the calibration of the X-ray fluorescence

109

device.

110

3.4. X-ray diffraction (DRX)

111

X-ray diffraction analysis (XRD) was performed using Philips PANalytical X'Pert equipment

112

with a copper anticathode diffractometer (λ= 0.178897 nm) operated at 45 kV, 40 mA

113

diffractometer 40 kV and 40 mA in a continuous scan mode by scanning range of 3−70° (2θ)

114

at 0.017°/s per step. The clay fraction (<2 µm particles) was separated from those samples by

115

sedimentation and centrifugation (Brindly and Brown, 1980; Felhi et al., 2008; Felhi, 2010;

116

Tlili et al., 2010). The bulk rock analysis and three oriented solid particles (untreated,

117

glycolated and heated at 550°C) were used in order to determine the bulk rock mineralogy

118

and the clay mineral assemblages. The percentage of all clay mineral content was determined

119

from the surface area of the peak at 4.45 Å. Where, the percentages of each mineral,

120

determined in bulk rock samples were estimated from the surface area of their strong

121

reflections.

122

3.5. Laser Granulometry

123

Granulometric analysis for untreated porcelanite sample, flux calcined porcelanite sample and

124

industrial diatomite filter were determined with Granulometer Lazer Masteriser S at the

125

Tunisian Chemical Group of Gabes, using wet method. The samples have been analyzed in

126

water suspension. In order to obtain good particle dispersions, a suitable ultrasonic sound time

AC C

EP

TE D

M AN U

SC

RI PT

102

5

ACCEPTED MANUSCRIPT has been applied (120s). In order to compare the granulometric parameters from different

128

samples, medium diameter (MMD) has been calculated. D (v; 0.5) corresponds to the particle

129

size for that 50% of the sample having a lower size and 50% of sample having a higher size.

130

4. Results

131

4. 1. Chemical analysis

132

Chemical analysis shows that all the samples of porcelanites are rich in SiO2 (Table 2). The

133

SiO2 content of clay-rich porcelanite (Gh) (80.37%) is greater than in carbonate-rich

134

porcelanite sample (Tm1) (72.65%). Moreover, Gh porcelanite is richer in Al2O3 (6.49%) and

135

Fe2O3 (4.35%) but contains the smallest amount of CaO (1.49%). Whereas, Tm1 porcelanite

136

presents the highest content of CaO (7.62%). The utmost content of SiO2 belongs to CHM3

137

porcelanite from Mides (83.66%) where the percentages of CaO, Al2O3 and Fe2O3 are

138

respectively 2.01%, 3.5% and 2.06%.

139

4. 2. Porcelanite color change after thermal treatment

140

Thermal treatment changes generally the coloration of all calcined porcelanite with alkaline

141

flux to white (Benda and Paschen, 1993; Hadjadj-Aoul et al.,2005; Martinovic et al., 2006;

142

Hadjar et al., 2008; Ediz et al., 2010; Arasuna et al., 2013). However the difference of color

143

between the calcined carbonate-rich porcelanite and the calcined clay-rich porcelanite is

144

notable. The beige color of raw porcelanite sample, on carbonate (Tm1), becomes yellow at

145

800°C and white brilliant at 1000°C. This color is also gotten with the fluxing porcelanite

146

samples of Mides (CHM3) (Tlili et al., 2012) and the fluxing calcined industrial products

147

(clarcel, kieselgur and diatomite). Whereas the color of the clay rich sample (GH), becomes

148

orange at 800°C and red at 1000°C. This red coloration is due to the oxidation of iron in the

149

clay-rich porcelanite after the breakdown of the clay mineral at high temperature.

150

4.3. X-ray diffraction (DRX)

151

The XRD patterns of tow porcelanite raw materials Tm1 and Gh revealed that they are

AC C

EP

TE D

M AN U

SC

RI PT

127

6

ACCEPTED MANUSCRIPT composed mainly of opal CT (Cristobalite/Tridymite) similar to porcelanite raw material

153

CHM3 (Tlili et al., 2012). The three main reflections of opal CT appears at the vicinity of 4.3,

154

4.1, and 2.5 Å (Figs. 3, A1 and 4, B1) (Hesse 1988; Nagase and Akizuki 1997; Önal and

155

Sarikaya 2007; Önal et al. 2007; Eversuel and Ferrell 2008). The percentage of opal CT in

156

Tm1 and Gh porcelanites attains respectively 85 and 81% (Table 3). The porcelanite of

157

Tamarza (Tm1) contain also carbonate (calcite and dolomite), low phase of clay minerals

158

(3%) and trace amounts of quartz, francolite, feldspar, gypsum and pyrite (Fig. 3, A1). The

159

percentage of carbonate in Tm1 porcelanite, estimated from the surface area of their strong

160

reflections of calcite and dolomite is 8 % (Table 3). Whereas, the percentage of clay minerals

161

in porcelanite of Kef El Ghiss (Gh) attain 15%. The XRD patterns of the Gh porcelanite show

162

also trace amounts of quartz, francolite and feldspar and the absence of carbonate (Fig. 4, B1).

163

The porcelanite of Mides contains a small amount of clay minerals and carbonate (as calcite)

164

and high amount of opal CT (Fig. 5, C1), which attain 91 % (Table 2).The amounts of all clay

165

minerals (ACM) in the three porcelanites (Tm1, Gh and CHM3) are constituted essentially of

166

smectite, sepiolite, and playgorskite (Fig. 6). Thermal treatment generally provokes a

167

mineralogical transformation in porcelanite. After fluxing calcination of the carbonate rich

168

porcelanite (Tm1) at 1000°C, opal-CT, transform mostly to cristobalite (Fig 3, A2) suggesting

169

similar behavior of CHM3 sample fluxed also at 1000°C (Fig. 5, C2). The opal C is the main

170

mineralogical composition of fluxing porcelanites Tm1 and CHM3 as same as diatomite

171

industrial filter aids (Tlili and al., 2012). The thermal treatment of Tm1 shows also the

172

apparition of new peaks (2.98 Å, 3.53 Å and 1.72 Å) that seem to be characteristic peaks of

173

the wollastonite. However, the rich clay mineral sample GH, which is carbonate poor , don't

174

show the same mineralogical transformations gotten after fluxing calcinations of the Tm1

175

samples. After thermal treatment of the porcelanite Gh at 1000°C the opal CT doesn't change

176

to opal C despite the improvement of the sharpness of the reflections of opal CT (Fig. 4, B2).

AC C

EP

TE D

M AN U

SC

RI PT

152

7

ACCEPTED MANUSCRIPT 4.4. Laser Granulometry

178

The granulometry of porcelanite starting powder was lower than 100 µm and displays

179

unimodal distribution between 1-100 µm (Fig. 7, G1). This granulometry changes during the

180

thermal treatment. However, after calcination of carbonate-rich porcelanite (Tm1) with

181

alkaline flux at 1000°C, the histogram of frequency that starts by a shape unimodal for the

182

raw sample becomes bimodal (Fig. 7). A new mode appears, between 0.1 µm and 1 µm after

183

heating during 1.5h, 4h and 6h (Fig. 7: G2, G3 and G4). The increase of the time of

184

calcination didn't change much the new granulometric distribution. This new mode appears

185

also in fluxing clay-rich porcelanite (Gh) (Fig. 8: G6) and fluxing silica-rich porcelanites

186

CHM3 at 1000°C (Saidi et al., 2012) as same as all industrial filter aids Kieselgur and

187

diatomite (Fig. 9: G7 and G8) and clarcel (Saidi et al., 2012). Nevertheless, fluxing clay-rich

188

porcelanite (Gh) at 1000°C during 6h, displays trimodal distribution which is different to that

189

of the fluxing carbonate-rich porcelanite sample (Tm1) for the same conditions.

190

5. Discussion

191

The color of all raw porcelanite of Gafsa-Metlaoui basin is beige brown correspond to the

192

occurrence of iron phase and organic matter. Alyosef et al (2014) reported that red brown

193

color of raw Eocene diatomite from Kasr El-Sagha, El Fayium is due to limonitic phase (FeO

194

(OH)·nH2O). After acid treatment, the color of this diatomite changes to yellowish-white. The

195

color of the Gafsa-Metlaoui porcelanite changes also after calcination with alkaline flux at

196

1000°C similar to flexing Pleistocene diatomite from Kom Osheem locality in El Fayoum,

197

Egypt (Ibrahim and Selim, 2010). The beige brown color of raw porcelanite (Tm1) became

198

white after fluxing carbonate-rich porcelanite (Tm1) and the fluxing silica-rich porcelanite

199

(CHM3) (Tlili et al., 2012), similar to the calcined industrial filter aids: clarcel, Kieselgur and

200

diatomite (Mechling, 2000; Tlili et al., 2012). Though, the color of fluxing clay-rich

201

porcelanite (Gh) becomes red after calcination. Indeed, the flux allows iron oxides to enter a

AC C

EP

TE D

M AN U

SC

RI PT

177

8

ACCEPTED MANUSCRIPT glassy phase where it is colorless (Ediz 2010; Tlili et al 2012). The Fe2O3 content of clay-rich

203

porcelanite (4.35 %) is twice as high as then in silica-rich porcelanite (2.01 %) and carbonate-

204

rich porcelanite (2.34 %). On the other hand the quantity of alkali flux used during all

205

calcination is always 5%. This quantity of alkali flux seems to be insufficient for the clay-rich

206

porcelanite sample to allow iron to enter a glassy phase during fluxing calcinations.

207

Therefore, the red coloration is due to the riches of Gh sample of Fe2O3, especially after the

208

breakdown of the clay mineral at high temperature and iron oxidation.

209

Mineralogical transformations appear after thermal treatment of the Gafsa-Metlaoui

210

porcelanite. The raw porcelanites (Tm1, Gh and CHM3) are composed mainly of

211

microcrystalline opal which corresponds to a mixture of short-distance order of tridymite and

212

cristobalite (Jones & Segnit, 1971; De Jong et al., 1987; Elzea et Riz, 1996; Nagase and

213

Akizuki, 1997; Çolak et al., 2000; Chao and Lu, 2002 ; Aras, 2004; Yuan al., 2004; Kahraman

214

et al., 2005; Önal and Sarikaya, 2007; Önal et al., 2007). Since α-cristobalite alternates

215

randomly, in a disorganized way, with the layers of α-tridymite (Hesse, 1988). The thermal

216

treatment at 1000°C encourages the progressive transformation of the disordered phase of

217

opal CT to the more ordered phase of opal C. This was happened for fluxing carbonate-rich

218

porcelanite and the fluxing silica-rich porcelanite. In fact, during thermal treatment, tridymite,

219

change to cristobalite and their characteristic peaks (d

220

nearly disappear (Fig. 10, A1 (2)). The strong peaks of cristobalite d (101) to 4.06 Å became

221

more developed and their characteristic peaks (3.16 Å and 2.87 Å) increase with the

222

increasing temperature (Fig. 10, A1 (1). This can be observed from the small tridymite

223

reflection near 4.32 Å in association with the development of sharp cristobalite reflection near

224

4.06. According to Arasuna et al (2013) a minor evidence of low-tridymite stacking was

225

evident after heat treatment of synthetic opal at 1400°C and the formation of α-cristobalite.

226

Thermal treatment of Tm1 porcelanite yielded also the formation of wollastonite, like

(404)

to 4.10 Å and d

(112)

to 4.32 Å)

AC C

EP

TE D

M AN U

SC

RI PT

202

9

ACCEPTED MANUSCRIPT Pleistocene diatomite from deposits Kom Osheem in El Fayoum, fired at temperature up to

228

1100°C (Hanna et al., 2014) and fluxing diatomite of Algeria above 900 °C (Hadjadj-Aoul et

229

al., 2005). The excess of carbonates in the Tm1 porcelanites produces the calcium from

230

800°C which contributes to the wollastonite crystallization. Whereas, the silica phase formed

231

by opal CT doesn't turn into opal C after fluxing calcination of clay-rich porcelanite (Gh). The

232

main reflections of tridymite, d(112) and d(404) respectively at 4.32 Å and 4.10 Å become more

233

intense and sharp (Fig. 10, A2 (3)). The XRD patterns of fluxing clay-rich porcelanite Gh at

234

1000°C may show the occurrence of opal T as reported by Eversuel and Ferrell, (2008), but

235

the weak reflections of cristobalite (3.16 Å and 2.87 Å) (Fig. 10, A2 (4)) indicates that the

236

silica phase gotten after thermal treatment is formed by opal CT. However, in the natural

237

deposition environment, the mineralogical transformation of the biogenic opal A to opal-CT

238

and/or to quartz take place more quickly in carbonated environment that in clay environment

239

(Greenwood, 1973; Kastner et al., 1977). In the same way, the presence of clay minerals

240

during the thermal treatment, blocks or delay the transformation of opal CT to opal C and/or

241

to cristobalite. Whereas, this transformation is more easy with carbonate. In order to follow

242

the mineralogical change during the thermal treatment, we have proceeded to calculate the

243

relative intensity (I2/I1+I2) such: I2 is the intensity of the reflection d (101) of cristobalite at

244

4.06 Å and I1 is the intensity of the reflection d

245

intensity of raw carbonate-rich porcelanite begins at 0.7 and decreases after fluxing

246

calcinations at 600°C during 1h 30mn, 3h and 6h. This reduction is may be due to the increase

247

of the intensity of the neighboring reflection of 4.32 Å that can be explained by the augment

248

of the layers of tridymite in the opal CT until 600°C. Then, the relative intensity (I2/I1+I2)

249

increases quickly with the increasing temperature between 600°C and 800°C, and reached the

250

maximum for 1000°C where it comes closer to 1 (Fig. 11, A). This increasing ratio can be

251

explained by the mineralogical rearrangement of the silica phase, following the

(112)

of tridymite at 4.32 Å. The relative

AC C

EP

TE D

M AN U

SC

RI PT

227

10

ACCEPTED MANUSCRIPT transformation of the disordered opal CT to more ordered opal C. This transformation is

253

accompanied with the liberation of the water molecules and the losses of weak links silanols

254

(Si-O-H) at the expense of siloxanes groupings (Si-O-Si). Arasuna et al (2013) reported that

255

heat treatment up to 400 provoke the water molecules losses of synthetic opal due to the

256

dehydration of silanol and the condensation of a new Si-O-Si silioxane groups. The ratio of

257

relative intensities of fluxing carbonate-rich porcelanite (Tm1) at 1000°C are also comparable

258

to those of fluxing silica-rich porcelanite CHM3 at 1000°C and to those of the industrial filter

259

aids clarcel and diatomite (Tlili et al., 2012). On the other hand, the relative intensity

260

(I2/I1+I2) of clay-rich porcelanite (Gh) begins from 0.75 with raw sample and decrease

261

slowly according to the increase of the treatment temperature. This ratio attains a minimum at

262

800°C and shows a maximum of 0.7 after calcination at 1000°C during 3h (Fig. 11, B). The

263

relative intensity of fluxing clay-rich porcelanite (Gh) at 1000°C during 1h30mn, 3h and 6h

264

don't attain the relative intensity ratios gotten with fluxing carbonate-rich porcelanite Tm1,

265

fluxing silica-rich porcelanite CH, and the industrial filter aids (clarcel and diatomite). The

266

silica phase gotten with fluxing clay-rich porcelanite (Gh) remains composed by opal CT and

267

the transformation to opal C is weak or absent. In the natural depositional environment of

268

silica, the clay minerals can delay the formation of opal CT because of their specific surface

269

area which enhances the sorption of opal A and decreases his transformation to opal CT. The

270

transformation of the opal CT to quartz will also be blocked by clay mineral. Hence, the opal

271

CT persists even with the increasing temperature (Williams and Crerar, 1985). It seems to

272

apply to the fluxing clay-rich porcelanite even with the breakdown of an important part of

273

clay minerals during thermal treatment (Fig. 10, A2 (5)). Opal CT of the fluxing clay-rich

274

porcelanite (Gh) seems becomes more organized and the sharpness of the main reflections of

275

tridymite (4.32 Å and 4.10 Å) may indicate that the new silica phase contains more trydimite

276

than cristobalite.

AC C

EP

TE D

M AN U

SC

RI PT

252

11

ACCEPTED MANUSCRIPT In the other hand, the fluxing calcination of the porcelanite influences the granulometry of

278

treated samples, but the granulometric transformation during the thermal treatment is different

279

between fluxing carbonate-rich porcelanite and fluxing clay-rich porcelanite (Gh). The

280

granulometry of raw carbonate-rich porcelanite and clay-rich porcelanite display unimodal

281

distribution between 1-100 µm and have medium diameter MMD equal respectively to 5.71

282

µm and 6.17 µm. Fluxing carbonate-rich porcelanite displays bimodal distribution. A new

283

mode appears systematically between 0.1 µm and 1 µm. However, the MMD after heated

284

temperature at 1000° C during 1.5 h, 4 h and 6 h equal respectively to 23.11 µm, 24.14 µm

285

and 24.12 µm (Fig 7. G2, G3 and G4). This behavior is quite similar to the industrial filter

286

aids: diatomite (MMD = 23.37 µm) (Fig 9. G8). The granulometric distribution of fluxing

287

carbonate-rich porcelanite is also similar to the granulometric distribution of fluxing silica-

288

rich porcelanite (the silica phase is formed mainly of opal C) (Saidi et al., 2012). The alkaline

289

flux encourages the agglomeration of the silica grains within porcelanite and leads to the

290

appearance of coarse agglomeration (1-100µm) and fine agglomeration (0.1-1µm) (Saidi et

291

al., 2012; Tlili et al 2012, Saidi 2015). The granulometric rearrangement of porcelanite during

292

thermal treatment may due to mineralogical transformation and crystal grow. It seems that the

293

cristobalite grains in opal C may have two different sizes (coarse and fine) as observed for

294

industrial filtration aids diatomite. Despite of the increasing granulometry of fluxing clay-rich

295

porcelanite (Gh) (MMD= 18.00 µm), his granulometric distribution is different from those of

296

fluxing carbonate-rich porcelanite fluxing silica-rich porcelanite and industrial filter aids.

297

Fluxing clay-rich porcelanite display trimodal distribution: fine size mode (0.1-1 µm), middle

298

size mode (1-100µm) and coarse size mode (100-300µm). The two first modes occur in

299

fluxing carbonate-rich porcelanite (Tm1), fluxing silica-rich porcelanite CHM3 at 1000°C and

300

in the industrial filter aids correspond to the opal C formed after heat treatment. But the third

AC C

EP

TE D

M AN U

SC

RI PT

277

12

ACCEPTED MANUSCRIPT one appears only in the flux-calcined clay-rich porcelanite and it may be due to silica glass

302

phase before transformation to opal C.

303

6. Conclusion

304

Thermal treatment induces different behavior on carbonate-rich porcelanite (Tm1) and clay-

305

rich porcelanite (Gh). The opal CT in carbonate-rich porcelanite changes to opal C as same as

306

silica-rich porcelanite and industrial filter aids. Since, the carbonate of Tm1 sample encourage

307

the transformation of silica during thermal treatment as same as the natural environment. But,

308

this transformation is blocked within clay rich porcelanite and fluxing clay-rich porcelanite

309

which remains formed by opal CT despite of the improvement of the main reflections. On the

310

other hand, all raw samples have unimodal granulometric distribution (1-100µm). The

311

calcination of carbonate-rich porcelanite Tm1 with alkali flux enhances the particle size and

312

reduces grain sharpness by agglomeration. Furthermore his granulometric distribution

313

becames bimodale and a new mode appear between 0.1 µm and 1 µm as same as fluxing

314

silica-rich porcelanite and industrial filter aid. However, fluxing clay-rich porcelanite display

315

trimodale granulometric distribution and another new mode ranges between 100-300µm due

316

to silica glass before transformation to opal C. Indeed, granulometric distribution is

317

influenced by the mineralogical transformation and crystal grows. The mineralogy and the

318

granulometry of fluxing carbonate-rich porcelanite became similar to those of flexing silica-

319

riche porcelanite and the industrial filter aids. Whereas, the presence of clay minerals prevents

320

the improvement of granulometric and mineralogical characteristics.

SC

M AN U

TE D

EP

AC C

321

RI PT

301

322 323 324 325 13

ACCEPTED MANUSCRIPT Figure captions

327

Fig 1. Lithological sections of the Ypresian phosphatic series in Kef El Ghiss, Mides and

328

Taramza.

329

Fig. 2. Geological map of the study area and position of the studied samples.

330

Fig 3. X-ray diffraction patterns (Cu Kα radiation) of raw and fluxing porcelanite sample of

331

Tamarza (Tm1). A1 : Tm1 raw porcelanite; A2: Tm116 fluxing porcelanite at 1000°C during

332

6h (ACM, all clay minerals; Ca, calcite; Do, dolomite; Fe, feldspar; Fr, francolite; Gy,

333

gypsum; He, hematite; OCT, opal CT; Pa, palygorskite; Py, pyrite; Se, sepiolite; Sm,

334

smectite; Qz, quartz).

335

Fig 4. X-ray diffraction patterns (Cu Kα radiation) of raw and fluxing porcelanite sample of

336

Kef El Ghiss (Gh). B1 : Gh raw porcelanite; B2: Gh16 fluxing porcelanite at 1000°C during

337

6h (ACM, all clay minerals; Fe, feldspar; Fr, francolite; Gy, He, hematite; OCT, opal CT; Pa,

338

palygorskite; Se, sepiolite; Sm, smectite; Qz, quartz).

339

Fig 5. X-ray diffraction patterns (Cu Kα radiation) of raw and fluxing porcelanite sample of

340

Mides (CHM3). A1 : CHM3 raw porcelanite; A2: CHM306 fluxing porcelanite at 1000°C

341

during 6h (ACM, all clay minerals; Ca, calcite; Fe, feldspar; Fr, francolite; OCT, opal CT; Pa,

342

palygorskite; Sm, smectite; Qz, quartz)

343

Fig. 6. X-ray diffraction pattern (CuKα radiation) of three oriented aggregates of

344

fraction <2 µm separated from three samples, CHM3, Tm1 and Gh. Untreated, glycolated,

345

and heated at 550 °C (Cl, clinoptilolite; Pa, palygorskite; Se, sepiolite; Sm, smectite)

346

Fig 7. Granulometic distribution histogram of raw and fluxing carbonate-rich porcelanite

347

(Tm1) at 1000˚C: G1, raw porcelanite; G2, porcelanite heated during 1.5 h; G3, porcelanite

348

heated during 4 h and G4, porcelanite heated during 6 h.

349

Fig 8. Granulometric distribution histogram of fluxing carbonate-rich porcelanite (Tm1) : G5,

350

raw porcelanite; G6, heated porcelanite at 1000˚C during 6 h.

clay

AC C

EP

TE D

M AN U

SC

RI PT

326

14

ACCEPTED MANUSCRIPT Fig 9. Granulometric distribution histogram of industrial filter aids: G7 Kieselgur; G8

352

diatomite.

353

Fig 10. X-ray diffraction patterns of raw and fluxing porcelanite during 6h (A1: Tm1 raw

354

porcelanite, Tm66 heated porcelanite at 600°C, Tm86 heated porcelanite at 800°C and Tm116

355

heated porcelanite at 1000°C (1: Increase of the sharpness of the characteristic reflections of

356

cristobalite; 2: Reduction of the intensity of the characteristic reflections of tridymite) ; A2:

357

Gh raw porcelanite, Gh66 heated porcelanite at 600°C, Gh86 heated porcelanite at 800°C and

358

Gh116 heated porcelanite at 1000°C(3: breakdown of the clay mineral; 4: Increase of the

359

sharpness of the characteristic reflections of tridymite) (ACM, all clay minerals; OC, opal C;

360

OCT, opal CT; Pa, palygorskite; Se, sepiolite; Sm, smectite; Wo, wollastonite)

361

Fig 11. Relative intensity vs. calcination temperature of tow samples of porcelanite at

362

different time (1.5, 4, 6h).

363

porcelanite sample (Gh) (I1: the intensity of the reflection d(112) of tridymite at 4.32 Å, I2: the

364

intensity of the reflection d(101) of cristobalite to 4.06 Å.

367 368 369 370

SC

M AN U

TE D EP

366

A: carbonate-rich porcelanite sample (Tm1); B: clay-rich

AC C

365

RI PT

351

371 372 373 374 15

ACCEPTED MANUSCRIPT

References

376

Alyosef, H.A., Ibrahim, S., Welscher, J., Inayat, A., Eilert, A., Denecke, R., Schwieger, W.,

377

Münster, T., Kloess, G., Einicke, W.D., Enke, D., 2014. Effect of acid treatment on the

378

chemical composition and the structure of Egyptian diatomite, International Journal of

379

Mineral Processing. 132, 17-25.

380

Aras, A., 2004. The change of phase composition in kaolinite- and illite-rich clay-

381

basedceramic bodies. Applied Clay Science. 24, 257-269.

382

Arasuna, A., Okuno, M., Okudera, H., Mizukami, T., Arai, S., Katayama, S., Koyano, M., Ito,

383

N., 2013. Structural changes of synthetic opal by heat treatment, Phys Chem Minerals. 40,

384

747-755.

385

Belayouni, H., 1983. Etude de la matière organique dans la série phosphatée du bassin de

386

Gafsa Metlaoui (Tunisie) Application à la compréhension des mécanismes de la

387

phosphatogenèse. Thèse Doct. ès-Sci. Univ. Orléans, France. 205p.

388

Benda, L.L., Paschen S., 1993. Kieselguhr. ln: Ulmann's Encyclopedia of Industrial hemistry,

389

A23, Weinheim, VCH Verlagsgesellchatt mbH, 607-613

390

Bernoullia, D., Gunzenhauser, B., 2001. A dolomitized diatomite inan Oligocene-Miocene

391

deep-sea fan succession, Gonfolite Lombarda Group, Northern Italy. Sedimentary Geology.

392

139, 71-91.

393

Brindly, G.W, Brown, G., 1980. Crystal structures of clay minerals and their X-ray

394

identification. Mineralogical society. 495 pp.

395

Burollet, P.F., 1956. Contribution à l'étude stratigrafique de la Tunisie centrale. A. mines et

396

Géol; Tunis, N° 18.

397

Burollet, P.F., Oudin, J.L., 1980. Paléocène en Tunisie-Pétrole et phosphate. In : Géologie

398

comparée des gisements de phosphate et de pétrole, mémoire du BRGM. 116 p.

399

Chaabani, F., 1995. Dynamique de la partie orientale du bassin de Gafsa au Crétacé et au

AC C

EP

TE D

M AN U

SC

RI PT

375

16

ACCEPTED MANUSCRIPT Paléogène: Etude minéralogique et géochimique de la série phosphatée Eocène, Tunisie

401

méridionale. Thèse Doc. Etat. Univ. Tunis II. Tunisie, 428p.

402

Chao, CH., Lu, H.Y., 2002. b. β-Cristobalite stabilization in (Na2O+Al2O3)-added silica

403

Metallurgical and Materials Transactions. 33, Issue 8, 2703-2711.

404

Çolak, M., Helvac, C., Maggetti, M., 2000. Saponite from the Emet Colemanite Mines,

405

Kütahya, Turkey, Clays Clay Miner. 48(4), 409-423.

406

De Jong, B.H.W.S., Van Hoek, J., Veeman, W.S., Manson, D.V., 1987. X-ray diffraction and

407

29Si magic-angle-spinning NMR of opals: incoherent long- and short-range order in opal-CT.

408

American Mineralogist. 72, 1195-1203.

409

Dirk, R., 2000. Dissolution kinetics of biogenic silica in marine environments. Thesis

410

modified version. At the Graduate School “Dynamik globaler Stoffkreisläuf im System Erde”.

411

der Abteilung Marine Umweltgeologie des Geomar. Geowissenschaften der Christian-

412

Albrechts-University, 234 p

413

Dixit, S., Van Cappellen, P., 1998. Interactions between biogenic silica and aluminum:

414

Implications for the oceanic silica cycle. E05 Transactions, Am. Geophys. Union, 1998 Fall

415

Meeting, 79 (49), 447.

416

Erdem, E., Çölgeçen, G., Donat, R., 2005. The removal of textile dyes by diatomite earth,

417

Journal of Colloid and Interface Science. 282, p 314–319.

418

Ediz, N., Bentli, İ., Tatar, İ., 2010. Improvement in filtration characteristics of diatomite by

419

calcination International Journal of Mineral Processing. 94, 129-134.

420

Elzea, J.M Rice, S.B., 1996. TEM and X-ray diffraction evidence for cristobalite and

421

tridymite stacking sequences in opal. Clays and Clay Minerals. 44, 492-500.

422

Eversuel, L.G., Ferrell, R., 2008. Disordered silica with tridymite-like structure in the Twiggs

423

clay. Am Mineral 93(4):565–572.

424

Felhi, M., Tlili, A., Montacer, M., 2008. Geochemistry, petrographic and spectroscopic

AC C

EP

TE D

M AN U

SC

RI PT

400

17

ACCEPTED MANUSCRIPT studies of organic matter of clay associated kerogenof Ypresianseries: Gafsa-Metlaoui

426

phosphatic basin, Tunisia. Resource Geology. 59, 428-436.

427

Felhi, M., 2010. Les niveaux intercalaires de la série yprésienne du bassin Gafsa-Métlaoui :

428

Apports de la minéralogie des argiles et de la géochimie de la matière organique résiduelle à

429

la reconstitution paléoenvironnementale. Thèse de Doctorat, Université de Sfax. FSS: p 170.

430

Greenwood, R., 1973. Cristobalite: its relationship to chert formation in selected samples

431

from the Deep Sea Drilling Project. J. sedim. Petrol. 43, 700-708.

432

Hadjadj-Aoul, O., Belabbes, R., Belkadi, M., Guermouche, M.H., 2005.Characterization and

433

performances of an Algerian diatomite-based gas chromatography support. Applied Surface

434

Science. 240, 131-139.

435

Hadjar, H., Hamdi, B., Jaber, M., Brendlé, J., Kessaïssia, Z., Balard, H., Donnet, J.B., 2008.

436

Elaboration and characterisation of new mesoporous materials from diatomite and charcoal.

437

Journal of Microporous and Mesoporous Materials. 107, 219-226.

438

Haj Ahmed, A., Tlili, A., Zalat, A., Jeddoui, Y., 2014. Fossil diatoms from endogangue of the

439

Ypresian phosphatic pellets of the Gafsa-Metlaoui basin: implication on the origin of biogenic

440

silica and depositional environment, Arabian Journal of Geosciences. 8 (2), 1077-1087.

441

Hanna, S.B., Ibrahim, S.S., Wahsh, M.M.S., Mansour, T.S., 2014. Diatomic heat insulating

442

material combined by ceramic bond, African Journal of Engineering Research, ISSN: 2354-

443

2144, 2(2), 26-38.

444

He, J., Zhou, Y., Li, H., 2011. Study on Geochemical Characteristics and Depositional

445

Environment of Pengcuolin Chert, Southern Tibet, Journal of Geography and Geology. 3, 1,

446

p187-188.

447

Hesse, R., 1988. Origin of chert, I. Diagenesis of biogenic siliceous sediment, Geoscience

448

Canada, (Diagenesis). 15, 171-192

449

Henchiri, M., 2007. Sedimentation, depositional environment and diagenesis of Eocene

AC C

EP

TE D

M AN U

SC

RI PT

425

18

ACCEPTED MANUSCRIPT biosiliceous deposits in Gafsa basin (southern Tunisia) Journal of African Earth Sciences. 49,

451

187-200.

452

Huang, H., Du, Y., Yang, J., Huang, H., Tao, P., Huang, Z., Yu, W., Guo, H., (2013).

453

Geochemistry of the Late Paleozoic cherts in the Youjiang Basin: Implications for the basin

454

evolution, Journal of Palaeogeography. 2 (4), 402-421.

455

Ibrahim, S.S., Selim, A.Q., 2010. Producing a micro-porous diatomite by a simple

456

classification calcinations process, The Journal of ORE DRESSING. 12 - Issue 23, 24-32.

457

Jones, J.B., Segnit, E.R., 1971. The nature of opal. Part 1: Nomenclature and constituent

458

phases, Journal of the Geological Society of Australia. 18, N 1, 57-68.

459

Jun, P., Haisheng, Y.I., Wenjie, X., 2000. Geochemical Criteria of the Upper Sinian Cherts of

460

Hydrothermal Origin on the Southeast Continental Margin of the Yangtze Plate. Chinese

461

Journal of Geochemistry. 19 No. 3, 217-226.

462

Kahraman, S., Önal, M., Sarıkaya, Y., Bozdoğan, I., 2005. Characterization of silica

463

polymorphs in kaolins by X-ray diffraction before and after phosphoric acid digestion and

464

thermal treatment. Analytica Chimica Acta. 552, 201-206.

465

Kametaka, M., Takebe, M., Nagai, H., Zhu, S., Takayanagi., Y., 2005. Sedimentary

466

environments of the Middle Permian phosphorite–chert complex from the northeastern

467

Yangtze platform, China; the Gufeng Formation: a continental shelf radiolarian chert;

468

Sedimentary Geology. 174, 197-222.

469

Kastner, M., Keene, J.B., Gieskes, J.M., 1977. Diagenesis of siliceous oozes, I. Chemical

470

controls on the rate of opal-A to opal-CT transformation-an experimental study. Geochim.

471

Cosmochim. Acta. 41, 1041-1059.

472

Kouteren, SV., 1994. Filters and absorbents. In: Carr, D.D. (Ed.), SMME. Colorado. USA,

473

497-507.

474

Martinovic, S., Vlahovic, M., Boljanac, T., Pavlov, L., 2006. Preparation of filter aids based

AC C

EP

TE D

M AN U

SC

RI PT

450

19

ACCEPTED MANUSCRIPT on diatomites, International. Journal of Mineral Processing. 80, 255-260.

476

Mechling, JM., 2000. Formulation de bétons courants avec les Grès du Luxembourg et les

477

kieselguhrs usagés des brasseries. Thèse de Doctorat en Géologie Appliquée au Génie

478

Civil,Université Henri Poincaré, Nancy1. 307p.

479

Önal, M., Sarikaya, Y., 2007. The effect of heat treatment on the paracrystallinity of an opal-

480

CT found in a bentonite. Journal of Non-Crystalline Solids. 353, 4195-4198.

481

Önal, M., Kahraman, S., Sarıkaya, Y., 2007. Differentiation of α-cristobalite from opals in

482

bentonites from Turkey, Applied Clay Science. 35. 25-30.

483

Nagase, T., Akizuki, M., 1997. Texture and structur of opal-C in volcanic rocks. The

484

Canadian Mineralogist. 35, 947-958.

485

Saidi, R., Tlili, A., Fourati, A., Ammar, N., Ounis, A., Jamoussi, F., 2012. Granulometric

486

distribution of natural and flux calcined chert from Ypresian phosphatic series of Gafsa-

487

Metlaoui basin compared to diatomite filter aid, IOP Conf. Series : Materials Science and

488

Engineering. 28 : p. 1-8.

489

Saidi, R., Falhi, M., Tlili, A., Khlil, L., Fourati, A., Kamoun, L., Jamoussi, F., 2014.

490

Depositional environment and stability of the porcelanite within the Ypresian phosphatic

491

series of the Gafsa-Metlaoui basin, southwestern Tunisia, Arab J Geosci. 8, Issue 7, 5223-

492

5237.

493

Saidi, R., 2015. Etudes géochimique minéralogique et pétrophysique des niveaux de

494

porcelanite du bassin de Gafsa-Metlaoui : perspectives de valorisation dans la filtration

495

industrielle. Thèse de Doctorat en Sciences Géologique, Université de Sfax, 286p.

496

Sassi, S., 1974. La sédimentation phosphatée au paléocène dans le Sud et dans le Centre

497

Ouest. PhD Thesis, Université Paris Orsay. 224p.

498

Şan, O., Özgür, C., 2009. Preparation of a stabilized β-cristobalite ceramic from diatomite,

499

Journal of Alloys and Compounds. 484, 920-923.

AC C

EP

TE D

M AN U

SC

RI PT

475

20

ACCEPTED MANUSCRIPT Tlili, A., Felhi, M., Montacer, M., 2010. Origin and depositional environment of palygorskite

501

and sepiolite from the Ypresian phosphatic series, southwestern Tunisia. Clays and Clay

502

Minerals. 58, No. 4, 573-584.

503

Tlili, A., Saidi, R., Fourati, A., Ammar, N., Jamoussi, F., 2012. Mineralogical study and

504

properties of natural and flux calcined porcelanite from Gafsa-Metlaoui basin compared to

505

diatomaceous filtration aids, Applied Clay Science. 62-63, 47-57.

506

Williams, L.A., Crerar, D.A., 1985. Silica digenesis, II. General mechanisms: Journal of

507

Sedimentary Petrology. 55, 312-321.

508

Yılmaz, B., Ediz, N., 2008. The use of raw and calcined diatomite in cement production,

509

Cement & Concrete Composites. 30, 202-211.

510

Yuan, P., Wu D.Q., He, H.P. Lin, Z.Y., 2004. The hydroxyl species and acid sites on

511

diatomite surface: a combined IR and Raman study. Applied Surface Science. 227, 30-39.

512

Zhou, Y., Edward, H.C., Jayanta, G., Lu, H., Tu, G., 2006. Hydrothermal origin of Late

513

Proterozoic bedded chert at Gusui, Guangdong, China: petrological and geochemical

514

evidence. Sedimentology. 41, Issue 3, 605-619.

517 518 519 520

SC

M AN U

TE D

EP

516

AC C

515

RI PT

500

521 522 523 524

21

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

525

22

ACCEPTED MANUSCRIPT

Table 1. Sample preparations and analysis Sample Gh

Thermal treatement with Na2CO3

CHM3

Tm161: 600°C, 1.5h Tm181: 800°C, 1.5h Tm111: 1000°C, 1.5h Tm164: 600°C, 4h Tm184: 800°C, 4h Tm114: 1000°C, 4h Tm166: 600°C, 6h Tm186: 800°C, 6h Tm116: 1000°C, 6h (Tlili et al., 2012) CHM3043: 400°C, 3h CHM3083: 800°C, 3h CHM3101: 1000°C, 1.5h CHM3103: 1000°C, 3h CHM3106: 1000°C, 6h

AC C

EP

TE D

Mides

Tm1

M AN U

Tamarza

SC

Gh61: 600°C, 1.5h Gh81: 800°C, 1.5h Gh11: 1000°C, 1.5h Gh64: 600°C, 4h Gh84: 800°C, 4h Gh14: 1000°C, 4h Gh66: 600°C, 6h Gh86: 800°C, 6h Gh16: 1000°C, 6h

RI PT

Section Kef El Ghiss

ACCEPTED MANUSCRIPT Table 2 Chemical composition of porcelanite rocks Tm1, Gh and CHM3 from the Ypresian phosphatic series of the Gafsa-Metlaoui basin (LOI = loss on ignition). Ref.

SiO2 Al2O3 Fe2O3 CaO MgO P2O5 K2O Na2O SO3 TiO2 LOI

Total

Gh

80.37

6.49

4.35

1.49

2.88

0.9

0.47 0.52 0.12 0.24 4.16 101,99

Tm 1

72.65

4.49

2.34

7.62

2.63

0.67

0.4

CHM3 83.66

3.5

2.06

2.01

1.14

1.04 0.48 0.54 0.75 0.14 6.49 101.81

ElGhis Tamarza

1.1

AC C

EP

TE D

M AN U

SC

Mides

RI PT

Kef

0.46 0.13 8.45 101,57

ACCEPTED MANUSCRIPT Table 3. Mineralogy of bulk rocks sampled from porcelanite of the Gafsa-Metlaoui basin (OCT = opal CT, ACM = all clay minerals).

Mides

Cl

CHM3 1

Phy

OCT

Qz

Fe

Ca

Do

Fr

2

91

1

2

1

2

0

Kef El Gh

0

15

81

1

2

Tm 1

0

3

85

1

0

M AN U TE D EP AC C

Tamarza

1

0

0

8

0

0

SC

Ghiss

RI PT

Ref.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

RI PT SC M AN U TE D EP





Porcelanite is heated with flux to prepare a specific filter aid for cleaning melting sulfur Mineralogy & granulometry of fluxing carbonate-rich and clay-rich porcelanites is different Opal CT transforms to opal C and becomes neater and more stable fluxing carbonate-rich porcelanite Transformation of opal CT to opal C by heat treatment is blocked with flux clay-rich porcelanite Granulometry of fluxing carbonate-rich is bimodal & for clay-rich porcelanites is trimodal

AC C