Mineralogy and Geochemistry of Permian Coal Seams of the Sydney Basin, Australia, and the Songzao Coalfield, Sw China

Home , Black coal equivalent, Coal dust, Coal preparation plant, Funginite, Illawarra Coal Measures, List of coalfields

MINERALOGY AND GEOCHEMISTRY OF PERMIAN COAL SEAMS OF THE SYDNEY BASIN, AUSTRALIA, AND THE SONGZAO COALFIELD, SW CHINA

Lei Zhao B.E. (Environmental Engineering) M.Sc. (Environmental Science)

Supervisors: Professor Colin Ward Dr Ian Graham Dr David French

A dissertation submitted in fulfillment of the requirement for the degree of Doctor of Philosophy

School of Biological, Earth and Environmental Sciences University of New South Wales, Sydney, Australia 2012

ACKNOWLEDGEMENTS

I would like to thank the following people and organisations for their assistance during my study and in completing this thesis.

I sincerely appreciate my supervisor, Prof. Colin Ward, and co-supervisors, Dr. Ian Graham and Dr. David French, for providing training opportunities, consistent professional advice, and the immeasurable time they committed towards this work. Thank you for your assistance and encouragement throughout this study and the writing-up of the thesis.

I would like to gratefully acknowledge the China Scholarship Council for financial support during this study. I am indebted to Dr. Chen-Lin Chou of Illinois State Geological Survey, and Prof. Kuilli Jin, Prof. Shifeng Dai and Prof. Longyi Shao of China University of Mining and Technology (Beijing), for their support and advice on my PhD study. I also would like to thank Dr. Zhongsheng Li of CSIRO for his professional support and technical assistance during the duration of my study.

Thanks are expressed to CSIRO Energy Technology, Prof. Shifeng Dai of China University of Mining and Technology (Beijing), and Peter Krempin of the Austar coal mine, for providing samples and other relevant data for conducting the investigation.

Thanks are also expressed to Rad Flossman and Joanne Wilde of UNSW, for preparation of the polished sections and thin-sections, to Irene Wainwright, Dorothy Yu, and Yu Wang of the Mark Wainwright Analytical Centre, UNSW, and Owen Farrell of CSIRO, for chemical analyses and technical assistance in the XRD analysis, and to Eugene White of the Electron Microscope Unit, Mark Wainwright Analytical Centre, UNSW, for technical assistance in the SEM analysis. Thanks are also expressed to the technical and administrative staff, especially Michael De Mol and Jonathan Russell, for their assistance with different aspects of the research program.

Postgraduate students of School of BEES, UNSW, especially Justin Ugbo, Asep Permana, and Kaydy Pinetown, are thanked for their general advice and encouragement. Students of University of Mining and Technology, Beijing, especially Xibo Wang, Yanfeng Lu and Xingwei Zhu, are thanked for collecting samples from the Songzao underground coal mines.

The comments on the REE minerals from Dr. Vladimir Seredin of Russian Academy of Sciences are highly appreciated. I would also like to thank the reviewers, Dr. Robert Finkelman of U.S. Geological Survey and Prof. David Spears of University of Sheffield, for their careful review and constructive comments on the manuscript.

Finally, special thanks are expressed to my parents for their great understanding, patience and encouragement throughout the duration of my study. I am forever indebted to you for your endless love and support.

ABSTRACT

This study is an investigation of the abundance and modes of occurrence of the mineral matter and trace elements in the Permian coal seams of the Sydney Basin, eastern Australia and the Songzao Coalfield, SW China, as well as the relationships between trace elements and mineral matter components within the different parts of the coal seams. A range of analytical techniques have been used to obtain relevant data, including optical microscopy, electron microscopy/microprobe analyses, quantitative X-ray diffraction, geochemical techniques (ICP-MS/OES, CV-AFS, HG-AFS and Eschka method), and Laser Raman spectroscopy analysis.

The Greta coal is a high-volatile bituminous coal and typically contains a high proportion of liptinite. The upper section of the Greta seam has several different indicators of marine influence, such as anomalously low vitrinite reflectance and abundant syngenetic pyrite, in the top part of the seam. Pyrite typically comprises 40 to 56% of the mineral assemblage in the coals from the marine-influenced upper section. In contrast, the mineral matter in the lower section contains minor pyrite, and relatively abundant dawsonite, which may have been formed by reactions between earlier-precipitated kaolinite and Na2CO3-or

NaHCO3-bearing fluids. The minerals, including most of the clay minerals, pyrite, siderite and quartz, within most of the Greta coal plies are largely of authigenic origin. Authigenic Na-rich I/S may have been syngenetically precipitated, probably after the peat was accumulated, with abundant Na and relatively minor K ions being supplied by the marine water.

Coals from the Great Northern and Bulli seams are mainly high volatile A bituminous and medium volatile bituminous in rank, respectively. The mineral fractions of the coals, especially in the middle parts of the seams, are dominated by authigenic kaolinite with a very low abundance of quartz and carbonate minerals. Apart from tonstein bands in the Great Northern seam, authigenic processes therefore appear to be the dominant mechanism of mineral matter formation in both coal seams. Authigenic K-feldspar also occurs in the lower part of the Great Northern seam, with a variety of unusual modes of occurrence. A late syngenetic low-temperature hydrothermal fluid injection process is suggested for formation of this feldspar component.

iii

Both quartz and non-kaolinite clay minerals are also abundant in the lowermost ply of the coal seams, suggesting that the immediate base of the peat bed in each case was made- up of organic matter admixed with the same detrital sediment as supplied to the basin. K- feldspar, which is present in the coals and non-coal bands in the lower metre of the Great Northern seam section, is not present in the Bulli seam. This may reflect deposition of the Bulli seam at a greater distance from the sediment source, which was located in the New England Fold Belt.

The coals from the Songzao Coalfield are mainly high ash, high sulphur semianthracites. XRD analysis indicates that minerals within the Songzao coals are mainly kaolinite, pyrite (or marcasite in some cases), and quartz, with minor proportions of carbonates, feldspar, anatase and sulphate minerals. Some of the illite and I/S is Na-rich in some of Datong coal samples. The I/S in the Songzao coals is mainly an alteration product of the original dispersed volcanic ash, due to the availability of necessary ions (e.g. K, Na, Mg) in the marine-influenced coal swamp. Organically-bound Na, which was expelled from the organic matter with coal rank advance, especially with anthracitization, may have supplied additional Na for the formation of Na-rich illite and I/S. Authigenic I/S also commonly occurs in a Tonghua coal ply that is overlain by a mafic bentonite and underlain by an alkali tonstein. K, Na and Mg for the formation of the I/S were probably derived from the leaching of the adjacent alkali tonstein and mafic bentonite. Although the marine water was also a possible supplier of the alkali elements, authigenic I/S is rare in other coals that occur further away from the altered volcanic layers.

REE minerals, which occur as fracture infillings in a Tonghua coal sample, were probably crystallized from ascending hydrothermal fluids carrying high REE concentrations, which may in turn have been associated with contemporaneous volcanic activity. Two groups of REE minerals, probably REE-hydroxides or oxyhydroxides, and REE-carbonates, were tentatively identified.

Tonstein bands in the Great Northern seam consist essentially of kaolinite. The occurrence of idiomorphic crystals of K-feldspar, which may represent members of the anorthoclase-sanidine series or a sodic sanidine, indicates an acid to intermediate volcanic ash input. Two tonstein and one K-bentonite bands in the Songzao coal seams have kaolinite and I/S as the dominant clay minerals, respectively. The volcanic ash layers in the peat swamp may have been originally converted to smectite, which was in turn altered to I/S and illite during diagenesis and/or rank advance, assuming that necessary

iv ions (e.g. K, Na and Mg) were available from the marine water percolating in the peat swamp. Na-rich I/S may also have been formed in the claystones, with the additional Na probably being released from the organic matter during the coal’s rank advance. The thin tonstein layers were formed, with relevant ions having been largely removed, probably due to a greater leaching efficiency.

In the relatively low-ash coals of the Greta, Great Northern, and Bulli seams, the concentrations of most trace elements are lower than that of average worldwide coals. By contrast, the high-ash Songzao coals have relatively high concentrations of most trace elements compared to averages for worldwide coals.

In the sulphur-rich Songzao and Greta coals, most of the chalcophile trace elements show either poor or negative correlations with total iron sulphide contents. Only Hg and Se in the Songzao coals and Hg, Tl and As in the Greta coals are positively correlated with iron sulphides, respectively. This may be because the pyrite in the Songzao and Greta coals is mostly of syngenetic origin. Some chalcophile elements are correlated with Al2O3,which most likely indicates a common source. The absence of traditional pyrite-metal associations may reflect wide variations in the concentrations of these elements in individual pyrite/marcasites, or simply poor retention of those elements in the pyrite/marcasite of the relevant coals. In addition to the lithophile elements, chalchophile elements in the Great Northern coals, including Se, Pb and Cu, also appear to be associated with kaolinite, and more likely a common source as well.

The geochemistry of the coals has been affected by the adjacent tonstein/bentonite bands. The relatively immobile elements enriched in the altered volcanic ashes also tend to be enriched in the adjacent coal plies, possibly due to leaching by groundwaters. The coals near the alkali tonstein bands in the Tonghua and Yuyang sections of the Songzao Coalfield are high in Nb, Ta, Hf, Ga, Th, U, and REE. Coal samples overlying the mafic bentonite in the Tonghua section are high in TiO2, V, Cr, Zn and Cu. However, the influence of the acid to intermediate tonstein layers in the Great Northern seam on the geochemistry of the adjacent coals is not as significant as in the Songzao coal seams. Trace elements, such as Li, Th, and U, are relatively high in most of the Great Northern coal plies adjacent to the tonstein bands.

CONTENTS

LIST OF FIGURES ...... xi

LIST OF TABLES ...... xviii

CHAPTER 1 INTRODUCTION...... 1

1.1 Overview ...... 1

1.2 Coal seams in the study area ...... 2

1.3 Research objectives ...... 3

1.4 Thesis outline ...... 4

CHAPTER 2 MINERAL MATTER AND TRACE ELEMENTS IN COAL SEAMS 7

2.1 Mineral matter in coal ...... 7

2.1.1 Origin and modes of occurrence of mineral matter in coal ...... 8

2.1.2 Minerals in coal...... 9

2.1.3 Intra-seam volcanic claystones...... 22

2.1.4 Summary ...... 30

2.2 Trace elements in coal...... 31

2.2.1 Origin of trace elements in coal...... 33

2.2.2 Modes of occurrence of trace elements in coal ...... 40

2.2.3 Geochemistry of intra-seam volcanic claystones ...... 50

2.2.4 Summary ...... 54

CHAPTER 3 GEOLOGICAL BACKGROUND OF THE STUDY AREA ...... 55

3.1 Geological setting of the Sydney Basin...... 55

3.1.1 Tectonic framework of the Sydney-Gunnedah-Bowen Basin ...... 56

3.1.2 Basin structure...... 56

3.1.3 Regional stratigraphy and depositional environment...... 57

3.1.4 Structural geology...... 59

3.1.5 Coal-bearing sequences and coal seams in the present study ...... 61

3.1.6 Tuffs and tonsteins in the coal measures of the Sydney Basin...... 65

vi 3.1.7 Summary...... 65

3.2 Geological background of the Songzao Coalfield ...... 66

3.2.1 Broad-scale tectonics ...... 66

3.2.2 Stratigraphy and depositional environment in SW China ...... 69

3.2.3 Local geology of the Songzao Coalfield...... 71

3.2.4 Summary...... 74

CHAPTER 4 SAMPLING AND ANALYTICAL TECHNIQUES ...... 75

4.1 Sampling...... 75

4.1.1 Samples from the Sydney Basin, Australia...... 75

4.1.2 Samples from the Songzao Coalfield, SW China...... 82

4.2 Sample preparation...... 82

4.3 Analytical techniques...... 83

4.3.1 Low-temperature oxygen-plasma ashing, X-ray diffraction (XRD) and Siroquant analysis...... 83

4.3.2 Clay fraction preparation and clay mineral analysis ...... 85

4.3.3 Petrographic analysis ...... 86

4.3.4 Scanning electron microscopy (SEM)...... 87

4.3.5 Electron microprobe analysis...... 87

4.3.6 Laser Raman spectroscopy...... 88

4.3.7 Proximate analysis ...... 88

4.3.8 X-ray fluorescence (XRF) spectrometry...... 88

4.3.9 Inductively coupled plasma-mass spectroscopy/optical emission spectroscopy (ICP-MS/OES)...... 89

4.3.10 Eschka method (B and Cl)...... 89

4.3.11 Hydride generation-atomic fluorescence spectroscopy (HG-AFS) (As and Se) ...... 90

4.3.12 Inductively coupled plasma-mass spectroscopy (ICP-MS) (As and Se)...... 91

4.3.13 Cold vapour-atomic fluorescence spectroscopy (CV-AFS) (Hg) ...... 91

4.3.14 DMA-80 direct Hg analyser (Hg)...... 92

4.3.15 Pyrohydrolysis/fluoride ion-selective electrode (ISE) (F)...... 92

vii 4.3.16 Data compilation and processing ...... 93

4.4 Summary...... 93

CHAPTER 5 MINERALOGY AND GEOCHEMISTRY OF THE GRETA SEAM 95

5.1 Proximate analysis ...... 95

5.2 Coal petrology ...... 98

5.2.1 Maceral composition...... 98

5.2.2 Vitrinite reflectance ...... 103

5.3 Mineralogy of the Greta seam...... 104

5.3.1 Minerals in roof, floor and other clastic strata...... 105

5.3.2 Minerals in Partings ...... 107

5.3.3 Coal mineralogy...... 116

5.4 Geochemistry of the Greta seam ...... 131

5.4.1 Mineralogical and chemical analysis data ...... 132

5.4.2 Vertical variations in major elements...... 137

5.4.3 Trace element geochemistry...... 139

5.4.4 Geochemical associations and element affinity...... 142

5.4.5 Distribution and affinity of REE and Y ...... 145

5.5 Summary...... 150

CHAPTER 6 MINERALOGY AND GEOCHEMISTRY OF THE GREAT NORTHERN SEAM...... 153

6.1 Coal characteristics ...... 153

6.2 Coal petrology ...... 153

6.3 Mineralogy of the Great Northern seam...... 158

6.3.1 Minerals in roof and floor samples ...... 158

6.3.2 Minerals in non-coal partings ...... 162

6.3.3 Minerals in coal samples...... 172

6.4 Geochemistry of the Great Northern seam ...... 180

6.4.1 Mineralogical and chemical analysis data ...... 180

6.4.2 Major element oxides in coals...... 184

viii 6.4.3 Associations of element components and minerals ...... 184

6.4.4 Selected elements in non-coal samples ...... 188

6.4.5 Distribution and affinity of REE and Y...... 190

6.5 Summary...... 192

CHAPTER 7 MINERALOGY AND GEOCHEMISTRY OF THE BULLI SEAM 195

7.1 Coal characteristics...... 195

7.2 Coal petrology...... 196

7.3 Mineralogy of the coals and associated strata...... 198

7.3.1 Mineralogy of the non-coal strata ...... 198

7.3.2 Mineralogy of the coal samples ...... 200

7.4 Geochemistry of the Bulli seam...... 203

7.4.1 Mineralogical and chemical analysis data...... 203

7.4.2 Major element oxides in coals ...... 204

7.4.3 Selected trace elements...... 206

7.4.4 Selected elements in non-coal samples ...... 207

7.4.5 Distribution and affinity of REE and Y...... 210

7.5 Summary...... 212

CHAPTER 8 MINERALOGY AND GEOCHEMISTRY OF THE SONGZAO COAL SEAMS ...... 215

8.1 Coal quality and chemistry ...... 215

8.2 Mineralogy of the coal seams...... 217

8.2.1 Minerals in roof and floor strata ...... 220

8.2.2 Minerals in Partings...... 221

8.2.3 Minerals in coal samples ...... 230

8.3 Geochemistry of the Songzao coal seams ...... 248

8.3.1 Relation between mineralogical and ash chemical data...... 248

8.3.2 Geochemical associations in coal samples...... 252

8.3.3 Associations of major elements in the coals ...... 258

8.3.4 Selected trace elements in coal...... 260

ix 8.3.5 Selected elements in non-coal samples...... 266

8.3.6 Distribution and affinity of REE and Y ...... 268

8.4 Summary...... 273

CHAPTER 9 COMPARISON AND INTERGRATION...... 277

9.1 Cross-checking of different analytical techniques ...... 277

9.1.1 XRF and XRD...... 277

9.1.2 XRD and SEM-EDS...... 278

9.1.3 ICP-MS/OES and XRF...... 279

9.2 Processes of mineral formation ...... 283

9.2.1 Detrital processes...... 283

9.2.2 Syngenetic processes...... 284

9.2.3 Epigenetic processes...... 286

9.2.4 Other minerals formed by diagenetic processes ...... 288

9.3 Mineralogy and geochemistry of tonstein/K-bentonite bands ...... 289

9.4 Factors influencing the geochemistry of the coal seams...... 290

9.4.1 Ash yield...... 290

9.4.2 Nature of the sediment source...... 290

9.4.3 Influence of volcanic ashes...... 291

9.4.4 Modes of occurrence of trace elements ...... 292

9.4.5 REE distributions ...... 293

CHAPTER 10 CONCLUSIONS...... 295

10.1 Mineralogy of the coal seams ...... 295

10.2 Geochemistry of the coal seams...... 298

10.3 Significance of intra-seam volcanic claystones ...... 299

REFERENCES...... 301

APPENDIX 1: Coefficients among ash yield, minerals and elements in the Greta coals APPENDIX 2: Coefficients among ash yield, minerals and elements in the Songzao coals APPENDIX 3: List of publications APPENDIX 4: Reprint of “Mineralogy of the volcanic-influenced Great Northern coal seam in the Sydney Basin, Australia” – a paper published on International Journal of Coal Geology, Vol. 113, 2012.

x LIST OF FIGURES

Figure 3.1 Location of the Sydney-Gunnedah-Bowen Basin system of eastern Australia (After Tadros, 1995)...... 55 Figure 3.2 General stratigraphy of the Sydney Basin (modified from Tadros, 1995, and Stewart and Alder, 1995)...... 58 Figure 3.3 Major structures within the Sydney Basin (after Stewart and Alder, 1995)...... 60 Figure 3.4 Stratigraphy of the Greta Coal Measures in the Lochinvar Anticline area (after Van Heeswijck, 2001)...... 62 Figure 3.5 Stratigraphic units in the Newcastle Coalfield of the Sydney Basin (left) with additional detail on units in the Moon Island Beach subgroup (right) (after Agnew et al. 1995)...... 63 Figure 3.6 Stratigraphy of the Illawarra Coal Measures in the Southern Coalfield (After Hutton, 2009)...... 64 Figure 3.7 Location of coal-producing provinces in SW China...... 67 Figure 3.8 (A) Sketch map showing the main tectonic units of China (NCB=North China Block; SCB=South China Block); (B) Map showing the South China Block, consisting of the Yangtze and Cathaysia Blocks, separated by the Jiangnan Orogen. After Duan et al. (2011)...... 68 Figure 3.9 Tectonics of the western Yangtze Block in the Hercynian period (€=Cambrian, O=Ordovician, S=Silurian; D=Devonian) (after Wang, 1996)...... 68 Figure 3.10 Lithostratigraphic correlations in Guizhou and Yunnan Provinces, SW China (after Wang et al., 2011)...... 69 Figure 3.11 Late Permian Longtan-stage palaeogeographic map of Southwest China (modified from Wang et al., 1995), showing the location of the Songzao Coalfield...... 71 Figure 3.12 Main structures in the Songzao Coalfield, with indication of mining areas (modified from Songzao Coalfield report)...... 72 Figure 3.13 Sedimentary sequences of the Songzao Coalfield, showing the location of the coal seams (modified from Songzao Coalfield report)...... 73 Figure 4.1 Map of Sydney Basin and locations of seam sections of the Great Northern seam and the borehole of the Greta seam...... 77 Figure 4.2 Lithologic columns of the Greta seam at Cessnock No. 1 Colliery (A) and Austar Coal Mine (B). .. 78 Figure 4.3 Lithologic column sections of the Great Northern seam at Newvale colliery (A) and Catherine Hill Bay (B)...... 79 Figure 4.4 Map of Sydney Basin and locations of the Bulli seam at Coal Cliff Colliery...... 81 Figure 4.5 Lithologic column sections of the Bulli seam at Coal Cliff Colliery...... 81 Figure 4.6 Lithologic columns of coal seams in the Songzao coalfield. (A) No. 7 seam of the Datong Coal Mine; (B) Seam K2b in the Tonghua Coal Mine; (C) No. 11 seam in the Yuyang Coal Mine...... 82 Figure 4.7 A scheme of sample preparations and analyses...... 84 Figure 5.1 Variation of proximate analysis data and total sulphur through the Greta seam of the Cessnock and Austar sections...... 97 Figure 5.2 Photomicrographs showing typical macerals in the Greta coals (Austar). Oil immersion, reflected light. (A) Telocollinite (tc) (upper field) and collodetrinite (cd) (lower field). Other macerals illustrated are semifusinite (sf) and funginite (fg) of fungal hyphae. (B) Cutinite (c) embedded in collotelinite. Also illustrated are syngenetic pyrite and funginite of fungal spores. (C) Fusinite (f), semifusinite (sf) and sporinite (sp). (D) An agglomeration of sporinite. (E) Gelinite (g) infilling cell cavities of semifusinite (F) Thick-walled cutinite. Micrinite and inertodetrinite are incorporated in the cutinite...... 100 Figure 5.3 Photomicrographs of liptinite macerals in the Greta coals (Austar). Oil immersion, reflected light and blue-violet fluorescence. (A) Resinite (rs) and sporinite (sp) in collodetrinite. (B) Same view as (A) under fluorescent illumination. Other fluorescing matter is sporinite and cutinite (across the field). (C) Exsudatinite

xi filling vein-like structure (horizontal), associated with clay mineral-filled cleats (vertical). (D) Same view as (C) under fluorescent illumination. (E) Concentrates of liptodetrinite (intense fluorescence) and clay minerals (weak fluorescence) in vitrinite. (F) Same view as (A) under fluorescent illumination. Other fluorescing matter in the matrix of collodetrinite is mainly sporinite, liptodetrinite and probably some clay minerals...... 101 Figure 5.4 Photomicrographs showing typical inertinite macerals in the Greta coals (Austar). Oil immersion, reflected light. (A) An agglomeration of funginite. Cutinite forms long, dark bands stretching across the upper field. Note compression effect around secretinite. (B) Non-vesicular secretinite showing internal notches. Note compression effect around secretinite. (C) Funginite (fg) of fungal spores and disseminated micrinite (mi), inertodetrinite (id) in a matrix of collodetrinite. (D) Funginite (fg) of fungal hyphae in the centre of the view. (E) Vesicular secretinite (se). (F) Fusinite (f) grading to semifusinite (sf). Funginite (f) are incorporated in the fusinite. Fusinite grades upwards into semifusinite...... 102 Figure 5.5 Plots showing vertical variation of vitrinite reflectance and abundance of major minerals in the Greta seam (Austar)...... 105 Figure 5.6 Minerals in the roof and floor samples. Transmitted light with crossed polars, (A) Spherulitic siderite grains in a floor sample. (B) Dawsonite (D) and framboidal and cubic pyrite (P) in a roof sample...... 106 Figure 5.7 Photomicrographs of non-coal partings. Plane polarised light (PPL). (A) Graupen to clastic textures in sample G20. (B) Graupen to clastic textures in sample G22. (C) Volcanic rock fragment with a weathering rim in sample G22. Feldspar phenocryst is noted. (D) Branching root structure in sample G6...... 108 Figure 5.8 SEM images of minerals in claystone sample G-20. (A) Euhedral quartz (Q) and goyazite; the laminae are I/S. (B) a rounded pellet of kaolinite, enclosed are subhedral to euhedral quartz (Q), dolomite (D) and goyazite (G). (C) Enlargement of (B) showing kaolinite aggregates (K), well-shaped goyazite (G) and dolomite (D)...... 110 Figure 5.9 SEM images of minerals in claystone sample G-22. (A) Diagenetic quartz and goyazite in clay mineral matrix. (B) Enlargement of (A) showing the association of goyazite (G) and quartz (Q)...... 111 Figure 5.10 SEM images of minerals in claystone sample G-20. (A) Dolomite (D) infills the cracks within diagenetic kaolinite (K). (B) Kaolinite enclosing dolomite...... 111 Figure 5.11 SEM images of anatase crystallites in claystones. (A) Ti-rich inclusions in cracks of kaolinite matrix. (B) Enlargement of image (A). (C) Ti-rich inclusions in cracks of I/S matrix. (D) Enlargement of image (C)...... 112 Figure 5.12 Plots showing vertical variation of abundance of major minerals (normalised to pyrite-free) in the Greta seam (Austar)...... 117 Figure 5.13 Photomicrographs showing clay minerals in the Greta coals (Austar). (A) Aggregates of clay minerals in the collodetrinite matrix. In air, reflected light. (B) Clay (dark) and carbonate minerals in a cleat. Oil immersion, reflected light...... 118 Figure 5.14 SEM images of kaolinite in the Greta coals. (A) Kaolinite (grey) and pyrite (bright) occurring as cell infillings, G-18. (B) Aggregates of kaolinite booklets in cell cavities. Bright area is pyrite (P), G-23...... 119 Figure 5.15 SEM images of sodium-bearing illite or I/S in the Greta coals. (A) I/S in cell cavities, G-21. (B) I/S in cell cavities, G-11. (C) I/S or paragonite in pore space or crack of coal maceral, G-21...... 121 Figure 5.16 SEM images of quartz in the Greta coals. (A) Detrital quartz, G-18. (B) A detrital quartz (Q) grain in the centre of the image that appears to be leached. More authigenic quartz (Q) occurs in cell cavities or pore spaces in the lower right of the image, G-18...... 122 Figure 5.17 Photomicrographs showing modes of mineral occurrence in the Greta coals. (A) Pyrite framboids and euhedral crystals, G-7, oil immersion (B) Pyrite in cell cavities of inertinite, G-37, oil immersion. (C) Pyrite replacement of wood structure, G-7, oil immersion. (D) Pyrite replacement of syngenetic siderite, G-28, in air...... 123 Figure 5.18 SEM images showing carbonate minerals in the Greta coals. (A) Sr-bearing calcite (C), along with I/S, in cracks of organic matter, G-5. (B) Ferroan dolomite (D) and quartz (Q) in cell cavities, G-29...... 125

xii Figure 5.19 SEM images showing the modes of occurrence of dawsonite in the Greta coal samples. (A) Dawsonite-filled cleats, thin section under PPL (plane polarized light), G-26. (B) Kaolinite and dawsonite co- precipitates in a cleat, G-5. (C) Enlargement of image (A), showing intergrowth of dawsonite crystals (D) in kaolinite (K). (D) Layer of dawsonite (D) in crushed cell cavities. Bright area is ankerite (A) and all the other cell infillings are Na-illite, G-21...... 127 Figure 5.20 SEM images of albite in coal sample G-5. (A) Albite (A) filling crushed cell cavities. (B) Cell-filling albite and quartz...... 128 Figure 5.21 SEM images of goyazite (G) occurring as infillings of pore space in coal sample G-21...... 130 Figure 5.22 Comparison between proportions of major element oxides in coal LTAs and non-coal strata from Greta seam (Austar), inferred from Siroquant and determined by XRF. The diagonal line represents equality in each plot. Relevant trendlines and squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case...... 136 Figure 5.23 Comparison between proportions of major element oxides in coal LTAs and non-coal strata from Greta seam (Cessnock), inferred from Siroquant and determined by XRF. The diagonal line represents equality in each plot. Relevant trendlines and squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case...... 137

Figure 5.24 Variation of major elements of the Greta seam (Austar), normalised to a Fe2O3- and LOI- free basis...... 139 Figure 5.25 Dendrogram developed from cluster analysis on the geochemical data of the Greta coals from the Austar Coal Mine (cluster method, centroid clustering; interval, Pearson correlation; transform values, maximum magnitude of 1)...... 143 Figure 5.26 Correlation of selected elements (Li, Pb, Cu, Sn and Sb) with alumina and kaolinite in the Greta coal samples. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case...... 145 Figure 5.27 Correlation of Tl, As, Hg and Se with pyrite in the Greta coal samples. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case...... 146 Figure 5.28 Correlation coefficients between mean individual REE and Y with LTA% in the Greta coal samples...... 146 Figure 5.29 Distribution patterns of REE in the Greta seam. REE are normalized to Upper Continental Crust (UCC) data from Taylor and McLennan (1985). (A) Coal samples G-5 to G-14; (B) Coal samples G-18 to G- (30-32); (C) Coal samples G-34 to G-41; (D) Roof samples G-(1,2,4) and G-3, floor sample G-(42-45), parting sample G-6, and other clastic rock sample G-(16-17)...... 147 Figure 6.1 Photomicrographs showing typical macerals in the Great Northern coals (Newvale). Oil immersion, reflected light. (A) Collotelinite (ct) bands. Sample 21383. (B) Telinite (t) grading upward to collotelinite (ct). Sample 21388. (C) Collotelinite (t) bands with thin-walled cutinite (c) layers. Sample 11212. (D) Telinite (t) with well preserved cells. Sample 21388. (E) Telinite (t) with the cell cavities filled with gelinite (g). Sample 11210. (F) Corpogelinite (co). Sample 21388...... 156 Figure 6.2 Photomicrographs of liptinite macerals in the Great Northern coals (Newvale). Oil immersion, reflected light and blue-violet fluorescence. (A) Resinite (rs) and sporinite (sp) in collodetrinite. Sample 11210. (B) Same view as (A) under fluorescent illumination. (C) Thick-walled sporinite (sp) and cutinite (c). Sample 11205. (D) Same view as (C) under fluorescent illumination. (E) Alginite (?). Sample 11208. (F) Same view as (E) under fluorescent illumination...... 157 Figure 6.3 Photomicrographs showing typical inertinite macerals in the Great Northern coals (Newvale). Oil immersion, reflected light. (A) Fusinite (f) grading to semifusinite (sf). Sample 11210. (B) Macrinite (ma). Sample 11208. (C) Secretinite (se). Sample 21383. (D) An agglomeration of funginite (fg). Sample 21381. 158 Figure 6.4 (A) Vein probably made-up of K-feldspar cutting well-preserved plant tissue in the siltstone floor of Catherine Hill Bay section, thin-section under crossed polars. (B) SEM image of K-feldspar veins in the

xiii siltstone floor of Newvale section. (C) SEM image of quartz veins in the siltstone floor of Newvale section. (D) SEM image of quartz with crushed fusinite in the siltstone floor of Newvale section...... 160 Figure 6.5 Vertical column section showing variations in clay mineralogy for the Great Northern seam at Newvale...... 164 Figure 6.6 SEM images of kaolinite in claystone partings from Newvale. (A) Graupen to vermicular kaolinite in 21382. (B) Graupen kaolinite in sample 21384. (C) Typical vermicular kaolinite with inclusions of anatase in sample 21384. (D) Platy kaolinite with fine anatase inclusions in sample 21389...... 165 Figure 6.7 SEM images of K-feldspar in two uppermost partings from Newvale. (A) Euhedral K-feldspar (F) (? sanidine) in parting 21382. (B) Euhedral K-feldspar (F) and quartz (Q) in 21384. (C) Etched K-feldspar in 21382. Fine white grains are anatase. (D) K-feldspar (F) fragments in clastic kaolinite (K) matrix in 21384.. 168 Figure 6.8 SEM images of K-feldspar in lowermost claystone parting 21389 from Newvale. (A) K-feldspar (F) and kaolinite (K) in banded or massive aggregates. (B) K-feldspar (F) veins cross-cutting organic stringers. Note a volcanic quartz (Q) in the upper left of the image. (C) Fusinite with mineral infillings. (D) Enlargement of an area in (C) showing K-feldspar (F), quartz (Q) and kaolinite (K). (E) K-feldspar and quartz in thin bands. (F) Enlargement of an area in (E) showing an intergrowth of K-feldspar (K) and quartz (Q)...... 169 Figure 6.9 Elemental maps of Al, Si, K and Na in K-feldspar veins cross-cutting organic matter in claystone parting 21389...... 170 Figure 6.10 SEM images of anatase in claystone parting 21384. (A) Anatase grains (A) and K-feldspar (F). (B) Apatite (Ap) and anatase (A) in a kaolinite matrix. (C) Anatase (A) replacing macerals. (D) Enlargement of an area in (C)...... 171 Figure 6.11 (A) SEM image of Fe, Mn, Mg, Ca phosphate (point 1) and K-feldspar (point 2) in kaolinite matrix of claystone parting 21384. (B) EDS spectrum of point 1...... 172 Figure 6.12 Kaolinite in coal sample 21388 from Newvale. (A) Vermicular kaolinite. (B) Cell infillings of kaolinite. (C) Cleat infillings of kaolinite. (D) Probable kaolinite pseudomorphs after biotite...... 174 Figure 6.13 XRD traces obtained from <2 μm fractions of (A) coal sample 21383 and (B) 21388 after glycol saturation and heating at 400°C for 2h...... 174 Figure 6.14 Minerals in coal samples from Newvale. (A) Quartz (Q) along with kaolinite (K) in cell lumens in coal 21388. (B) Quartz in maceral micropores in coal sample 21388. (C) Detrital quartz grains in collodetrinite in coal sample 21388. (D) Quartz-rich rock fragment in coal sample 21380. (E) Enlargement of (D) showing chlorite (C) and quartz (Q) crystals. (F) K-feldspar in coal sample 21390...... 177 Figure 6.15 (A) Zoned euhedral siderite crystals in coal sample 21388. (B) Cell-infilling siderite overgrown by a later siderite in coal sample 21390. (C) Goyazite-crandallite (G) and kaolinite (K) cell infillings in coal sample 21386. (D) EDS spectrum of point 1 from (C)...... 180 Figure 6.16 Comparison between proportions of major element oxides in coal LTAs and non-coal strata from Newvale, inferred from Siroquant and determined by XRF. The diagonal line represents equality in each plot. Relevant trendlines and squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case...... 183 Figure 6.17 Correlation of elements (Li, Se, Pb and Cu) with kaolinite in the Great Northern coals in the Newvale section. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case...... 187

Figure 6.18 Comparison of element data of the Great Northern coal samples. (A) Na2O against K2O. (B) Rb against K2O. (C) Sc against SiO2. (D) Ag against Nb. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case...... 188

Figure 6.19 Plots of elements for the Great Northern non-coal samples. (A) Comparison of TiO2 and Al2O3 concentrations. The upper and lower diagonal lines represent TiO2/Al2O3 values of 0.07 and 0.02, respectively.

(B) Plot of Zr/TiO2 against Nb/Y ratios. Magma source discrimination diagram of Winchester and Floyd (1977)...... 190

xiv Figure 6.20 Distribution patterns of REE in the Great Northern seam (Newvale). REE are normalized to Upper Continental Crust (UCC) data from Taylor and McLennan (1985). (A) Coal samples 21380 to 21390; (B) Non- coal samples including roof sample 21379, claystones 21382, 21384, and 21389, and floor 21391...... 192 Figure 6.21 Correlation coefficients between mean individual REE and Y with LTA% in the Great Northern coal samples (Newvale)...... 192 Figure 7.1 Photomicrographs showing typical macerals in the Bulli coals. Oil immersion, reflected light...... 197 Figure 7.2 Vertical column sections showing variations in clay mineral compositions of the Bulli seam...... 199 Figure 7.3 SEM images of minerals in coal sample 8123. (A) Siderite concretion, containing minor Ca, Mg and Mn. (B) Fluorapatite (A) co-existing with kaolinite (K) in pore space. Weak Al and Si peaks are from the intimate mixture with the kaolinite. (C) Ca, Ba-bearing goyazite in pore spaces of kaolinite (K) aggregates. 202 Figure 7.4 Comparison between proportions of major element oxides in the Bulli coal ash and non-coal samples inferred from Siroquant and determined by XRF. The diagonal line represents equality in each plot. Relevant trendlines and the squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case...... 205 Figure 7.5 Comparison of elements to kaolinite and ash yield of the Bulli coal samples. (A) Li against kaolinite. (B) Be against the ash yield. (C) As against the ash yield. (D) Tl against the ash yield. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case...... 207

Figure 7.6 Comparison of elements in the Bulli coal and non-coal samples. (A) Li against Al2O3. (B) Be against

Al2O3. (C) Ga against Al2O3. (D) As against Al2O3. (E) Rb against K2O. (F) Sc against SiO2. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case...... 208

Figure 7.7 Plots of elements for the Bulli non-coal samples. (A) Comparison of TiO2 and Al2O3 concentrations.

The upper and lower diagonal lines represent TiO2/Al2O3 values of 0.07 and 0.02, respectively. (B) Plot of

Zr/TiO2 against Nb/Y ratios. Magma source discrimination diagram of Winchester and Floyd (1977)...... 210 Figure 7.8 Distribution patterns of REE in the Bulli seam. REE are normalized to Upper Continental Crust (UCC) data from Taylor and McLennan (1985). (A) Coal samples 8118 to 8124; (B) Non-coal samples including roof sample 8117, claystone sample 8119, and floor sample 8125...... 211 Figure 7.9 Correlation coefficients between mean individual REE and Y with LTA% in the Bulli coal samples...... 212 Figure 8.1 Variation of proximate analysis and vitrinite reflectance data and total sulphur through three seam sections of the Songzao Coalfield...... 217 Figure 8.2 Plots showing vertical variation of abundance of major minerals (normalised to carbonate-free) in the Songzao seam sections...... 218 Figure 8.3 Column section showing vertical variations in clay mineralogy for the three seam sections, Datong (left), Tonghua (centre) and Yuyang (right), in the Songzao Coalfield...... 220 Figure 8.4 Powder (lower trace) and oriented-aggregate XRD traces of Tonghua floor sample (th-k2b-7), showing peaks due to poorly-ordered kaolinite, regularly interstratified I/S, quartz, albite and anatase, plus aluminum from the sample holder; the upper two traces show more detailed characteristics of the clay minerals...... 221 Figure 8.5 Thin section photomicrographs of claystone samples. (A) Elongated pellets in th-k2b-3, PPL. (B) A homogeneous texture shown in yy-11-6. (C) Volcanic quartz in lower part of image, yy-11-6, PPL. Note the resorbed angular texture. (D) Chloritised biotite in th-k2b-3, XPL...... 222 Figure 8.6 X-ray powder diffractograms showing minerals in claystone samples (th-k2b-3, th-k2b-5 and yy-11- 6), including poorly-ordered kaolinite (K), I/S, anatase (An), albite (Alb) and pyrite (py). Aluminum (Al) was derived from the sample holder...... 223 Figure 8.7 Kaolinite in claystone sample th-k2b-5. (A) Thin section photomicrograph of vermicular kaolinite, PPL. (B) SEM image of kaolinite of tabular structure in an I/S matrix. Bright area is Ti-rich material...... 225

xv Figure 8.8 XRD traces obtained from clay fractions of claystones. (A) Claystone th-k2b-3, showing regularly interstratified I/S, kaolinite (K) and chlorite (C). (B) Claystone th-k2b-5, showing regularly interstratified I/S and kaolinite (K). (C) Claystone yy-11-6, showing kaolinite (K) and a trace proportion of I/S...... 226 Figure 8.9 SEM images showing the modes of occurrence of anatase in claystone sample th-k2b-3. (A), (B), (C) and (D): Anatase appears to be replacement of pumice or other volcanogenic components...... 228 Figure 8.10 SEM image showing the modes of occurrence of anatase in claystone sample th-k2b-5...... 229 Figure 8.11 SEM images showing the modes of occurrence of anatase in claystone sample yy-11-6. (A), (B) and (C): Anatase appears to be replacement of glass shards and other volcanogenic components...... 229 Figure 8.12 SEM image showing REE- and Ba- phosphates, probably gorceixite, in claystone sample th-k2b-5...... 230 Figure 8.13 XRD traces obtained from clay fractions of coal LTAs. (A) dt-7-5, showing kaolinite (K), illite (I), and a trace of I/S. (B) th-k2b-1, showing kaolinite (K) and smectite (S). (C) th-k2b-4, showing kaolinite (K), smectite (S) and chlorite (C). (D) yy-11-5, showing kaolinite (K) and probably a trace of I/S...... 231 Figure 8.14 SEM images of clay minerals in coal samples. (A) Detrital kaolinite (K) laminae and cleat- and fracture-filling kaolinite (K). Also shown are syngenetic pyrite (Py) and detrital quartz (Q). yy-11-1. (B) Smectite or I/S in pore spaces. Also shown are vermicular kaolinite (K), pyrite (py) and detrital quartz (Q). th-k2b-4.. 233 Figure 8.15 SEM images showing clay minerals in coal sample dt-7-5. (A) Na-rich I/S, or paragonite in an I/S matrix. The bright area is probably phosphate. (B) Na-rich I/S, or paragonite, showing regular flakes...... 234 Figure 8.16 SEM images of chamosite in coal sample th-k2b-4. (A) Kaolinite and chamosite occurring in fractures probably formed in different stages. The earlier formed fractures are parallel to each other, and are confined in probably a vitrinite band. Note the displacement of the vitrinite band later formed during tectonic deformation. (B) Enlargement of (A), showing chamosite intergrown with kaolinite. (C) Enlargement of (A), showing fracture-filling chamosite and kaolinite. (D) Enlargement of (C), showing chamosite intergrown with kaolinite...... 236 Figure 8.17 SEM images showing modes of occurrence of chlorite in coal sample th-k2b-4. (A) Chlorite (C), ankerite (A) and quartz (Q) in fracture. (B) Enlargement of (A) showing chlorite intergrown with ankerite. (C) Chlorite (C), ankerite (A) and quartz (Q) in fracture. (D) Enlargement of (C) showing chlorite intergrown with quartz...... 237 Figure 8.18 Photomicrographs of quartz in Songzao coal samples. In air, reflected light. (A) Abundant detrital quartz grains in dt-7-1. (B) Euhedral quartz in th-k2b-4...... 238 Figure 8.19 Modes of occurrence of pyrite in Songzao coal samples. (A) Clustered and isolated framboidal pyrite. yy-11-4. In air, reflected light. (B) Euhedral pyrite crystals. yy-11-1. In air, reflected light. (C) Cell-filling pyrite. th-k2b-4. In air, reflected light. (D) Cleat-filling pyrite, dt-7-1. Oil immersion, reflected light. (E) SEM image of framboidal pyrite and isolated pyrite crystals in a clay matrix. dt-7-2. (F) SEM image showing framboidal cemented by later-formed massive pyrite. dt-7-2...... 239 Figure 8.20 SEM images showing the association of pyrite and marcasite in coal sample dt-7-3. (A) Multistage aggregates of pyrite and marcasite. (B) Enlargement of (A) showing pyrite framboids with different density. (C) Enlargement of (B) showing that pyrite framboids consisting of pyrite microcrystals...... 240 Figure 8.21 Photomicrographs showing modes of occurrence of pyrite and maracasite in coal samples. In air, reflected light. (A) Massive marcasite. dt-7-3 (B) Marcasite with bladed morphology. th-k2b-4. (C) Different associations of pyrite and marcasite. dt-7-3. (D) Association of pyrite (Py) and marcasite (Ma). dt-7-3...... 240 Figure 8.22 SEM images showing modes of occurrence of carbonates in coal sample th-k2b-2. (A) I/S and dolomite coexisting in cell cavities. Cell-filling calcite (Ca) is also indicated. (B) Ankerite (A) enclosing earlier- formed pyrite (Py)...... 241 Figure 8.23 SEM images showing gorceixite in coal sample th-k2b-4. (A) Gorceixite in the matrix of band rich in I/S. (B) Gorceixite (G) in a matrix of kaolinite within pores of the organic matter...... 242

xvi Figure 8.24 SEM images showing REE-phosphates in coal sample th-k2b-4. (A) Fine-grained REE- phosphates, some of which probably contain Ba, in the matrix of crack-filling I/S. (B) Fine-grained REE- phosphates in the matrix of I/S bands...... 243 Figure 8.25 SEM images showing modes of occurrence of anatase in coal sample th-k2b-4. (A) Euhedral anatase crystals associated with Ti-bearing kaolinite. (B) Probably an intimate mixture of anatase and kaolinite. (C) Flattened circular bodies of anatase possibly replacing glass spherules. (D) Enlargement of (C) showing detail of the possible replacement texture; the cavities are filled with I/S. (E) Anatase possibly replacing shell or wood fragments. (F) Enlargement of (E). The fine-grained, bright particles are probably Nb-bearing zircon (Zr)...... 246 Figure 8.26 SEM images of REE-bearing minerals in coal sample th-k2b-4. (A) Fracture-filling REE minerals. (B) Two different REE minerals in fracture (see EDS images). Grey area contains Ca, and bright areas are without Ca...... 247 Figure 8.27 Comparison between proportions of major element oxides in coal ashes and non-coal strata from three seam sections in the Songzao Coalfield, inferred from Siroquant and determined by XRF. The diagonal line represents equality in each plot. Circled points are discussed in the text. Relevant trendlines and squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case...... 251 Figure 8.28 Dendrogram developed from cluster analysis on the geochemical data of the coals from three seam sections in the Songzao Coalfield (cluster method, centroid clustering; interval, Pearson correlation; transform values, maximum magnitude of 1)...... 256

Figure 8.29 Correlations between selected elements in the Songzao coal samples. (A) TiO2 against Al2O3. (B)

P2O5 against MnO. (C) Sr and Ba against P2O5. (D) P2O5 against CaO. (E) Ba against Al2O3. (F) Ga against

Al2O3. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case...... 259 Figure 8.30 Correlations between selected elements in the Songzao coal samples. (A) Cr and Cu against V. (B) V, Cr and Cu against the sum of illite and I/S, on a whole-coal basis. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case...... 261 Figure 8.31 Correlation of selected elements (As, Mo, Hg, Se, Tl and Ge) with the sum of pyrite and marcasite in the Songzao coal samples, on a whole-coal basis. (A) As against iron sulphides (the sum of pyrite and marcasite). (B) Mo against iron sulphides. (C) Hg against iron sulphides. (D) Se against iron sulphides. (E) Tl against iron sulphides. (F) Ge against iron sulphides. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case...... 262 Figure 8.32 Plots showing vertical variation of HTA percentage and selected trace elements in the Songzao seam sections ...... 265

(B) Plot of Zr/TiO2 against Nb/Y ratios using the magma source discrimination diagram of Winchester and Floyd (1977)...... 267 Figure 8.34 Correlation coefficients between mean individual REE and Y with ash yield in the Songzao coal samples...... 271 Figure 8.35 Distribution patterns of REE in the three seam sections. REE are normalized to Upper Continental Crust (UCC) data from Taylor and McLennan (1985). (A) Coal samples in the Datong section. (B) Rock samples in the Datong section. (C) Coal samples in the Tonghua section. (D) Rock samples in the Tonghua section. (E) Coal samples in the Yuyang section. (F) Rock samples in the Yuyang section...... 272 Figure 9.1 of Sr (A), Y (B), Cu (C), Nb (D), Th (E), Pb (F), Zn (G), and Zr (H), determined by ICP and XRF, XRF Pro-trace techniques. All data in ppm. Relevant trendlines and the squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case...... 281

xvii Figure 9.2 Comparisons of V (A), Cr (B), Ga (C), Th (D), La (E), Ce (F), Nd (G), and Sm (H), determined by ICP and XRF, XRF Pro-trace techniques. All data in ppm. Relevant trendlines and the squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case...... 282

LIST OF TABLES

Table 2.1 Principal minerals identified in coal (modified from Ward, 2002)...... 11 Table 4.1 Samples for the present study...... 75 Table 4.2 Standard brightness categories (Standards Australia, 1993, 2007b) ...... 76 Table 5.1 Proximate analysis and total sulphur of the Greta coal samples in the Cessnock seam section (data from CSIRO reports) ...... 96 Table 5.2 Classification of maceral groups, sub-groups and macerals according to the ICCP System 1994... 99 Table 5.3 Mineralogy of Austar LTAs and associated non-coal samples by XRD and Siroquant (wt. %). LTA percentages were not determined for the non-coal samples...... 113 Table 5.4 Mineralogy of Cessnock LTAs and associated non-coal samples by XRD and Siroquant (wt. %) . 116 Table 5.5 EDS micro analyses of authigenic I/S in the Greta coals (data from 21 points)...... 120 Table 5.6 Major element analyses of Greta coal ash and non-coal samples from the Austar Coal Mine (%).132 Table 5.7 Major element analyses of Greta coal ash and non-coal samples from the Cessnock Coal Mine (%)...... 133 Table 5.8 Trace element analyses of the Greta coal and associated non-coal samples from the Austar Coal Mine (all elements but Hg in ppm, Hg in ppb, on whole-coal basis) ...... 140 Table 5.9 Broad classification of elements according to the results from cluster analysis and correlation coefficients (R) between the content of individual elements (E) in coal and LTA% and other minerals in the LTAs (Correlation coefficients for all inter-element and ash relationships are presented in Appendix 1)...... 143 Table 5.10 Rare earth elements in coal samples and associated strata from the Greta seam in the Austar section (REE concentrations in ppm, on whole-coal basis)...... 148 Table 6.1 Proximate analysis and vitrinite reflectance data for the Great Northern coal and associated strata from Newvale No.1 Colliery (see Chapter 4). Proximate analysis data from CSIRO reports...... 154 Table 6.2 Maceral analysis (vol., %, mineral-free basis) of Great Northern coal samples from Newvale No.1 Colliery (bdl = below detection limit)...... 155 Table 6.3 Mineralogy of Great Northern coal LTAs and non-coal rock samples from Newvale No.1 Colliery by XRD and Siroquant (wt. %) ...... 161 Table 6.4 Mineralogy of Great Northern coal LTAs and non-coal rock samples from Catherine Hill Bay by XRD and Siroquant (wt. %)...... 161 Table 6.5 Mineralogy of <2 μm fraction of coal LTAs and non-coal strata using oriented aggregate XRD techniques (wt. %)...... 162 Table 6.6 EDS micro analyses of K-feldspar in Great Northern non-coal samples from Newvale No.1 Colliery (%, O by difference) ...... 166 Table 6.7 Major element analyses of Great Northern coal ash and non-coal samples from Newvale No.1 Colliery (%), as determined by XRF analysis (n.d. = not detected)...... 181 Table 6.8 Major element analyses of Great Northern non-coal samples from Catherine Hill Bay (%), as determined by XRF analysis...... 181

xviii Table 6.9 Major oxide and trace element analyses of Great Northern samples in the Newvale section (Major oxides in %, recalculated from the data by XRF analysis; trace elements in ppm; all data on whole-coal basis)...... 186 Table 6.10 Rare earth elements in coal samples and associated strata from the Great Northern coal and associated non-coal samples in the Newvale section (REE concentrations in ppm, on whole-coal basis). ... 191 Table 7.1 Proximate analysis and vitrinite reflectance data of Bulli coal and associated strata. Proximate analysis data from CSIRO reports...... 195 Table 7.2 Maceral analysis (vol., %, mineral-free basis) of the Bulli coal samples (bdl = below detection limit) ...... 196 Table 7.3 Mineralogy of the Bulli coal LTAs and non-coal rock samples by XRD and Siroquant (wt. %)...... 198 Table 7.4 Mineralogy of <2 μm fraction of coal LTAs and non-coal strata using oriented aggregate XRD techniques (wt. %)...... 198 Table 7.5 Major element analyses of Bulli coal ash and non-coal samples (%), as determined by XRF analysis (bdl = below detection limits; n.d. = no data)...... 204 Table 7.6 Major oxide and trace element analyses of samples from the Bulli seam (Major oxides in %, recalculated from the data by XRF analysis, on whole-coal basis; Trace elements in ppm, unless otherwise indicated, on whole-coal basis; bdl = below detection limit)...... 209 Table 7.7 Rare earth elements in coal samples and associated strata from the Bulli coal and associated non- coal samples (REE concentrations in ppm, on whole-coal basis) ...... 211 Table 8.1 Proximate analysis and forms of sulphur (selected samples) and mean maximum vitrinite reflectance value of the Songzao coal samples (%, air-dried basis, unless indicated)...... 216 Table 8.2 Mineralogy of the Songzao LTA and associated non-coal samples by XRD and Siroquant (wt. %) ...... 219 7DEOH 0LQHUDORJ\ RI ȝP IUDFWLRQ RI FRDO /7$V DQG QRQ-coal strata using oriented-aggregate XRD techniques (wt. %)...... 222 Table 8.4 Averages and ranges of concentrations of REE and Ca determined by electron microprobe analyses for REE mineral veins in coal sample th-k2b-4 (wt%)...... 247 Table 8.5 Major element analyses of Songzao coal ash and non-coal samples from three seam sections (%), as determined by XRF analysis...... 250 Table 8.6 Major oxide and trace element analyses of the Songzao coal and associated non-coal samples from three seam sections (Major oxides in %, trace elements in ppm, unless otherwise indicated. All data on a whole-coal basis. Major element oxides recalculated from the data by XRF analysis. Trace elements determined by ICP-MS/OES analysis) ...... 254 Table 8.7 Broad classification of elements according to the results from cluster analysis and correlation coefficients (R) between the content of individual elements (E) and Al2O3 in coal (Correlation coefficients for all inter-element and ash relationships are presented in Appendix 2)...... 257 Table 8.8 Rare earth elements in the Songzao coal samples and associated strata (REE concentrations in ppm, on whole-coal basis)...... 269

xix

CHAPTER 1 INTRODUCTION

1.1 Overview

Coal is an organic sedimentary rock, essentially composed of lithified plant debris interspersed with a smaller proportion of mineral matter (Francis, 1961; Ward, 1984). Mineral matter in coal embraces discrete crystalline minerals and other non-crystalline inorganic matter (Ward, 1989; Standards Australia, 1995, 2000b; Ward, 2002). Knowledge of the mineral matter in coal is of great significance in solving industrial problems, including difficulties associated with materials handling, boiler erosion, ash formation, and slagging in coal processing or utilization (e.g. Ward, 1984; Gupta et al., 1999). Knowledge of the mineral matter is also important in understanding the inorganic processes associated with coal formation (e.g. Finkelman, 1994; Ward et al., 2001), and thus provides important information about the depositional conditions and the geological history of coal-bearing sequences, along with the regional sedimentary and tectonic history (e.g. Ren, 1996; Ward, 2002). In addition, minerals in coal may have direct impacts on human health during mining, preparation and combustion (e.g. Tian et al., 2008; Large et al., 2009; Huang et al., 2006).

Depending on their nature and abundance, particular elements (e.g. As, Se, Hg) in the coal may represent a potential environmental hazard, either from the coal itself, or from the associated preparation wastes or combustion products. However, some trace elements may also represent a significant source of material for industrial purposes. Elements such as Ge, Se, Ga, U, REEs, and PGEs, have been found in sufficient concentrations in coals in some parts of the world to represent deposits of economic interest (Seredin, 1996, 2004; Dai et al., 2005b; Dai et al., 2006a; Zhuang et al., 2006; Qi et al., 2007b, a; Seredin and Finkelman, 2008).

Tonsteins, and less commonly bentonites and K-bentonites, are known to occur in many different coal-bearing formations throughout geological time. Tonsteins are thought to provide geochronological markers, with a widespread distribution that enables stratigraphic correlation. The primary minerals in tonsteins, if preserved, may also provide opportunities for absolute age determination.

Leaching of tonsteins or the precursor volcanic ash by ground waters and organic acids in the peat-forming environment would be expected to result in enrichment of some elements that were released from the ash and accumulated as minerals in the coal (e.g. Crowley et al., 1989; Hower et al., 1999a). The occurrence of such minerals may therefore be indicative of volcanic influence during peat accumulation and coal formation. Crowley et al. (1989) concluded that one of the primary factors responsible for trace-element enrichment in coals was leaching of soluble components from tonsteins.

1.2 Coal seams in the study area

Coal from the Mid Permian Greta seam and the Late Permian Great Northern and Bulli seams of the Sydney Basin, eastern Australia, are used as fuel sources in domestic and international markets. Tuffs and tonsteins are very common throughout the Late Permian sequences of the Sydney Basin (e.g. Loughnan and Corkery, 1975; Diessel, 1983; Agnew et al., 1995; Creech, 2002). Whole rock chemical fingerprinting of tuffs and tonsteins, usually represented by the concentrations of relatively immobile elements and their ratios, has been successfully used in several studies on correlations in the coalfields of the Sydney Basin, Australia (Kramer et al., 2001; Grevenitz, 2003).

The Songzao Coalfield is located in Qijiang County, SW Chongqing. The city of Chongqing, along with other cities in the SW China, has a serious acid rain problem, with 90% of the rain water having a pH < 5.6 and the average pH being as low as 4 (Tang et al., 2009). The sulphate in the rain water is mainly attributed to emissions from combustion sources, particularly the local high-sulphur coal (Larssen and Carmichael, 2000; Liu, 2006; Zhang, 2007).

Altered volcanic ash layers are also widely spread in the Permian strata of SW China (e.g. Zhou et al., 1992; Zhou et al., 1994; Zhou et al., 2000). Although the geochemistry of tonsteins in the Late Permian coals of SW China indicates an origin from silicic volcanic ash fallout (Zhou et al., 1982), alkali tonsteins that developed in the early part of the Late Permian in SW China have also been reported (Zhou and Ren, 1994; Zhou, 1999; Zhou et al., 2000). A recent study of the mineralogy and geochemistry of the Songzao coals (Dai et al., 2007b) has indicated that coals from one of the Songzao coal seams (the No. 11 seam) is significantly enriched in some alkaline elements such as Nb, Ta, Zr, Hf, and REE, and these geochemical anomalies have been mainly attributed to synsedimentary alkaline

Chapter 1 Introduction volcanic ashes. Dai et al. (2011) more recently distinguished three types of tonsteins bands (silicic, mafic and alkali) in the Songzao Coalfield based on their distinctive chemical compositions.

The coal seams evaluated in this study were formed under a range of contrasting geological conditions. The Bulli and Great Northern seams, for example, are inertinite-rich coals with low sulphur contents, and were probably deposited in partly dry freshwater environments, with little, if any, marine influence (e.g. Diessel, 1992; Agnew et al., 1995). Both seams are sources of coal for domestic and export markets. They are of similar age, but were deposited on different sides of the basin, distal and proximal, respectively, to the main sediment source area. The Greta seam and the Songzao coals, on the other hand, are richer in vitrinite and have relatively high sulphur contents, and were deposited under more marine conditions (e.g. Diessel, 1992; Dai et al., 2007b). Although deposited on different continents, they both have pyritic mineral matter, and appear to have been formed under similar environmental conditions.

The magmas responsible for the tuffs (and tonsteins) associated with the Sydney Basin coals were probably rhyodacitic to dacitic in composition (Kramer et al., 2001). However, in addition to similar acid tonsteins, mafic and alkali tonsteins have also been found in the Songzao Coalfield (Dai et al., 2011).

1.3 Research objectives

The work program in this thesis aimed to investigate the abundance, modes of occurrence, and origins of the mineral matter in the coal and associated non-coal strata of the Sydney Basin and the Songzao Coalfield, and also to discuss the relationships between trace elements and mineral matter components within the coals. Although some of the coals in the study are from mines that have now been closed, the study provides an opportunity to assess the influence of the different depositional environments on the coals’ mineralogical and geochemical characteristics.

Within this framework, the objectives of this thesis are to compare and integrate the results obtained from the different analytical techniques, to evaluate the role of the different inorganic processes and sedimentary inputs that were associated with coal formation, and to evaluate different depositional and post-depositional conditions and

3 geological factors that may have controlled the abundance and distribution of minerals and elements, including trace elements, within the various coal seams. Most of the Sydney Basin samples were taken from a sample bank maintained by CSIRO Energy Technology, providing an opportunity to evaluate the mineral matter in relation to a range of other, previously-determined coal properties.

The study of the mineralogy and geochemistry of coals from the Songzao Coalfield is also expected to provide significant information for optimizing coal preparation methods to minimize the acid gas emissions during coal utilization processes.

1.4 Thesis outline

The thesis consists of ten chapters, the contents of which are described below:

Chapter 1 provides an introduction to the thesis, including an overview of the significance of mineral matter and trace elements in coal, a brief introduction to the coal seams in the study area, and the objectives and outlines of the thesis study.

Chapter 2 provides a review of the literature on mineral matter and trace element geochemistry of coal, as a background of the main thesis study. The first part of the chapter reviews the definitions and significance of mineral matter, and, more specifically, the origin and modes of occurrence of mineral matter in coal. It also reviews related research on the mineralogy of the altered volcanic ash layers in coal seams. The second part reviews the definition and significance of trace elements, and, more specifically, the origin and modes of occurrence of trace elements in coal, as well as the geochemistry of the altered volcanic ash layers found in many coal seams.

Chapter 3 provides a summary of the regional geological and tectonic settings of the Songzao Coalfield and the Sydney Basin, with a special focus on the coal seams evaluated in the study.

Chapter 4 describes the sampling and analytical methodology used in the project, including the sources and preparation of the coal and non-coal samples for the study and the different analytical techniques, as well as the computer software used for processing and presenting the analytical results.

Chapter 1 Introduction

Chapters 5 to 7 provide the data obtained from petrological studies of the Sydney Basin coals, and also data on the abundance, origin, and modes of occurrence of mineral matter and inorganic elements in the Greta (Chapter 5), Great Northern (Chapter 6) and Bulli (Chapter 7) coal seams. Chapter 6 also includes a discussion on the origin of the tonstein bands in the Great Northern seam, and the influence of the volcanic ashes on the mineralogy and geochemistry of the adjacent coal beds.

Chapter 8 provides data on the abundance, origin, and modes of occurrence of mineral matter and elements in three coal seams of the Songzao Coalfield. It also includes discussions on the influence of the volcanic ashes on the mineralogy and geochemistry of the adjacent coal layers.

Chapter 9 provides an integration of the findings from these individual studies, and attempts to draw more general conclusions from the work program. More specifically, it includes a comparison of the results obtained from the different analytical techniques for trace element determination, as well as an evaluation of the role of the different inorganic processes and sedimentary inputs that were associated with the formation of the coal seams.

Chapter 10 provides a brief summary of the main conclusions arising from the research program, based on the material presented.

CHAPTER 2 MINERAL MATTER AND TRACE ELEMENTS IN COAL SEAMS

2.1 Mineral matter in coal

Mineral matter in coal, as defined by Standards Australia (1995, 2000b), refers to minerals and other inorganic matter in and associated with coal. According to Ward (1989, 2002), mineral matter embraces three forms of components, namely: discrete crystalline mineral particles (and in some cases amorphous phases), inorganic elements incorporated within coal macerals, and dissolved ions and other inorganic material in pore water. Discrete crystalline minerals are generally the dominant form of mineral matter in coal, but non- crystalline mineral matter of the second and third forms (non-mineral inorganics) may also occur, especially in low-rank coals (Ward, 2002).

As opposed to mineral matter, “ash” does not exist as such in coal, but is the non- combustible inorganic residue that remains after coal combustion (Francis, 1961; Ward, 1984). In other words, ash is derived from the incineration of the mineral constituents (Schopf, 1956), or more correctly the total mineral matter. This process involves dehydroxylation of clay, loss of CO2 from carbonates (Diessel, 1992), oxidation of S from sulphides, and formation of a range of products as a result of the interaction of non- mineral inorganics with organic elements (e.g. sulphur) (Ward, 2002). It can also result in the formation of new minerals or amorphous phases due to structural changes (Speight, 1983).

Knowledge of the mineral matter in coal is of great significance in understanding geological processes and solving industrial problems. It provides important information about the depositional conditions and thus the geological history of coal-bearing sequences, along with the regional sedimentary and tectonic history (e.g. Ren, 1996; Ward, 2002). Knowledge of the mineral matter in coal, as opposed simply to the elemental composition of the coal ash, is important in understanding both the inorganic processes associated with coal formation (e.g. Finkelman, 1994; Ward et al., 2001) and aspects such as materials handling, boiler erosion, ash formation, and slagging in coal processing or utilization (e.g. Ward, 1984; Gupta et al., 1999). Profiles of mineral concentrations may also be useful for seam correlation in coalfields (Rao and Walsh, 1999; Ward et al., 2001). Knowledge of mineral matter is also useful, and sometimes necessary, to assess

7 environmental and health impacts of coal preparation and utilisation (e.g. Large et al., 2009; Huang et al., 2006), as well as to determine the potential for economic by-product recovery from coal or coal products (e.g. Seredin and Finkelman, 2008).

2.1.1 Origin and modes of occurrence of mineral matter in coal

A range of different processes can be involved in the formation of mineral matter during coal formation. Francis (1961) classified the minerals in coal into inherent and adventitious mineral matter on a genetic basis, referring to inorganic constituents that were derived from coal-forming plants and added after peat deposition, respectively. However, distinction is not always possible, especially when mineral matter was precipitated from solutions which may have been derived from either the plants or ions transported to the peat (Diessel, 1992). According to another genetic classification of minerals in coal by Mackowsky (1982), discrete minerals in coal may be of detrital, syngenetic, or epigenetic origin. Both syngenetic and epigenetic minerals are authigenic, that is they formed in situ. However, syngenetic minerals formed during peat accumulation and epigenetic mineals after peat consolidation. However, the distinction between syngenetic and epigenetic minerals is not always clear (Damberger et al., 1984). The origins and precipitation processes involved in mineral formation in coal are usually indicated by the modes of mineral occurrence, and the association of the mineral and organic matter.

Detrital minerals were washed or blown into the peat swamp, incorporated into the peat, and then subjected to diagenesis. They occur either as dispersed grains, or concentrated microscopic laminae or dirt bands, depending on the balance between rates of detrital influx and organic accumulation (Spears, 1987). The formation of dirt bands may result from sudden influxes of sediment-laden water into the swamp possibly related to the formation of crevasse splays in a fluviatile environment or widespread flooding (Spears, 1987). Tonsteins are generally regarded as dirt bands that were originally composed of airborne volcanic ash, and then altered in the coal-forming environment (Spears, 2012). The mineralogy and geochemistry of tonsteins are reviewed later in this chapter.

Syngenetic minerals were precipitated during or shortly after peat accumulation, during the period of biochemical change (Francis, 1961), and thus are generally intimately associated with the organic matter. They are disseminated in the coal matrix, or infill cell lumens. Syngenetic minerals may be derived from either the inorganic constituents of coal-forming plants or solutions that transported the mineral-forming constituents to the

Chapter 2 Mineral Matter and Trace Elements in Coal Seams depositional site. However, inorganic elements may also be removed in solution from parental plant material during compaction and dehydration (Davis et al., 1984; Diessel, 1992). Common syngenetic minerals in coal include framboidal pyrite, concretionary siderite, and cell or pore fillings of pyrite, clay minerals, quartz, or carbonates (calcite, dolomite and ankerite).

Epigenetic minerals are precipitated from evolving pore fluids during burial diagenesis (Spears and Caswell, 1986), and are typically deposited in cleats (joints in coal), fractures or other shrinkage fissures. Unlike syngenetic minerals, they are not intimately associated with the organic matter, since much compaction, solidification and diagenesis has already been completed. The original fluid may be related to volcanic activity or igneous intrusions, or to low-temperature hydrothermal fluids circulating in porous and fissured coals that have undergone thermal metamorphism (e.g. Seredin and Finkelman, 2008). Pyrite, carbonates, and clay minerals are the most common epigenetic minerals.

Non-mineral inorganics in coal, as noted earlier, are inorganic elements (all elements except C, H, N, O and S) bonded in various ways with the organic components and dissolved in the pore water. Although measurable concentrations of Al and Ca have been detected by Ward et al. (2005) in the macerals of bituminous coals using electron microprobe techniques, non-mineral inorganics occur in more significant proportions in low-rank coals. For example, Li et al. (2007b) found up to 1.5% Ca, 0.5% Al and 0.7% Fe in macerals of some low-rank coals. Non-mineral inorganics occurring in low-rank coals may be derived from post-depositional ion migration (Brockway and Borsaru, 1985; Ward, 2002). They may also be further expelled from the organic matter with rank advance, by processes such as dehydration, decarboxylation and dehydroxylation (Li et al., 2010), and may in turn be reprecipitated as epigenetic minerals (Ward, 2002). As indicated by Susilawati and Ward (2006), non-mineral inorganic elements in low-rank coals may also interact with existing minerals to form new minerals when the coal is affected by igneous intrusions.

2.1.2 Minerals in coal

The most common minerals in coal are silicates, silica phases, sulphides and carbonates. Sulphates and phosphates are less common, but may occur in significant proportions in some cases. Other accessory minerals are not always abundant enough to be identified from XRD analysis of whole coal or low temperature ashes. Table 2.1 lists the minerals

9 that have been identified in coal. The common minerals in coal are reviewed below in respect to their origin, formation and modes of occurrence.

2.1.2.1 Clay minerals

Clay minerals are the most common group of minerals in coal and associated lutites. They may be detrital, occurring in bands, laminae and lenticles, authigenic as cell infillings, or as late diagenetic veins. The source of the clay minerals may be weathering of feldspars and micas (Mackowsky, 1982). They may also be derived from the decomposition of volcanic materials, and occur as dispersed crystallites in coal, or be more concentrated, such as in tonstein bands. The clay mineralogy of tonstein bands will be discussed later in this review. The most common clay minerals in coal are kaolinite, illite and mixed layer illite/smecitite (I/S).

Kaolinite Kaolinite is the dominant mineral in many coals, occurring in both well- and poorly-ordered forms. Detrital kaolinite in coal, especially in coal that is adjacent to the roof or floor of the seam, usually occurs in bands and laminae. Although some detrital kaolinite may be involved, much of the kaolinite found in coal is well-ordered and appears to be the result of in situ leaching and reprecipitation processes within the peat swamp (Ward, 1989); such kaolinite typically occurs in the form of cell or pore-infillings. The solubility of aluminium increases markedly as the pH falls, whereas silica solubility is independent of pH below 9 (Spears, 1987). Authigenic kaolinite in coal may be formed by interaction of alumina and silica in solution at higher pH levels (Ward, 2002). Low-ash coals are often dominated by diagenetic kaolinite in the mineral assemblages (Spears, 2000); this has been demonstrated in some Australian coals, especially in the middle part of the coal beds where quartz is subordinate (Ward, 1989). However, poorly-ordered kaolinite, along with illite and I/S, and usually abundant quartz, tends to be common in the basal part and near the top of coal beds (e.g. Ward, 1989; Ward and Christie, 1994).

Kaolinite may also originate from diagenetic alteration of pre-existing alumino-silicates within the peat (Davis et al., 1984), such as feldspars, micas and volcanic ash debris. Such kaolinite may appear book-shaped or tabular in form, and also display a well- ordered structure. Kaolinite is also common as epigenetic cleat/fracture infillings, examples of which have been reported in many coals (e.g. Spears and Caswell, 1986).

Chapter 2 Mineral Matter and Trace Elements in Coal Seams

Table 2.1 Principal minerals identified in coal (modified from Ward, 2002). Mineral Formula Mineral Formula Silicates Oxides and hydroxides

Clay minerals Quartz SiO2

Kaolinite Al2Si2O5(OH)4 Chalcedony SiO2

Illite K1.5Al4(Si6.5Al1.5)O20(OH)4 Anatase TiO2

Smectite Na0.33(Al1.67Mg0.33)Si4O10(OH)2 Rutile TiO2

Chlorite (MgFeAl)6(AlSi)4O10(OH)8 Diaspore Al2O3+2O

Mixed layer Variable Hematite Fe2O3 clays

Feldspars Magnetite Fe3O4

Plagioclase (Na,Ca)Al(Al,Si)Si2O8 Goethite Fe2O3+2O

Albite NaAlSi3O8 Ilmenite FeTiO3

Anorthite CaAlSi3O8 Boehmite AlOOH

Orthoclase KAlSi3O8 Gibbsite Al(OH)3

Sanidine KAlSi3O8 Carbonates

Tourmaline Na(MgFeMn)3Al6B3Si6O27(OH)4 Calcite CaCO3

Analcime NaAlSi2O6+2O Dolomite CaMg(CO3)2

Clinoptilolite (NaK)6(SiAl)36O72+2O Ankerite (Fe,Ca,Mg)CO3

Heulandite CaAl2Si7O18+2O Siderite FeCO3

Zircon ZrSiO4 Dawsonite NaAlCO3(OH)2

Sulphides Strontianite SrCO3

Pyrite FeS2 Witherite BaCO3

Marcasite FeS2 Alstonite BaCa(CO3)2

Sphalerite (Zn,Fe)S Aragonite CaCO3 Galena PbS Sulphates

Pyrrhotite Fe(1-x)S Gypsum CaSO4+2O

Stibnite SbS Anhydrite CaSO4

Millerite NiS Bassanite CaSO4+2

Phosphates Barite BaSO4

Apatite Ca5F(PO4)3 Coquimbite Fe2(SO4)3 +2O)

Crandallite CaAl3(PO4)2(OH)5+2O Copiapite Fe5(SO4)6(OH)2 +2O)

Goyazite SrAl3(PO4)2(OH)5+2O Ferricopiapite Fe5(SO4)62 2+ +2O)

Gorceixite BaAl3(PO4)2(OH)5+2O Szomolnokite FeSO4+2O

Monazite (REE,Th)PO4 Jarosite KFe3(SO4)2(OH)6

Xenotime (Y,Er)PO4 Natrojarosite NaFe3(SO4)2(OH)6

Florencite (REE)Al3(PO4)2(OH)6 Hexahydrite MgSO4+2O

Glauberite Na2Ca(SO4)2

Thenardite Na2SO4

Tschermigite NH4Al(SO4)2+2O

Illite The chemical composition of illite is similar to that of muscovite, but illite contains less K+ and more SiO2 and H2O (Carroll, 1970). Because of this, the X-ray diffraction peaks of illite are usually similar though more diffuse and broader than those of pure micas.

As noted above, non-kaolinite clay minerals are generally less abundant in coal than in the associated non-coal strata. Illite, when present in North American coals, is dominantly of detrital origin (Finkelman, 1981). According to Ward (Ward, 1978), illite is rarely seen in Australian coals, and where present is degraded (depleted in potassium generally due to leaching or weathering) or partly interlayered with expandable clay minerals. However, well-ordered illite may occur in marine-influenced coal seams, where the marine water may have provided sufficient K+ to stabilize any degraded illite that may have been introduced (Ward, 1978).

Epigenetic illite, along with kaolinite and illite, has been found as a precipitated material in cleats of coals from the Bowen Basin, Australia (Faraj et al., 1996). Ammonian illite

(tobelite) has an [001] peak between around 10.35 Å (NH4 end-member of illite) (Eugster and Munoz, 1966) and 10.0 Å (K end-member of illite), depending on the proportion of + NH4 in the total interlayer cations. Ammonian illite has been recorded in coals, occurring in the organic matrix or on cleats/fractures (Daniels and Altaner, 1993) and in the associated underclay (Juster et al., 1987). It has generally been suggested that NH4-illite + + formed by the substitution of NH4 for K in K-illite due to rank advance (e.g. Juster et al., 1987; Ward and Christie, 1994). Some studies (Daniels and Altaner, 1993; Permana et al.,

2010; Dai et al., 2012d), however, instead suggest that NH4-illite formed by interaction of kaolinite with nitrogen, released from the organic matter. In most cases, the development of NH4-illite is due to more localised metamorphism caused by hot fluid injection or igneous intrusions. A similar conclusion that NH4 came from thermal maturation of organic matter was made by Nieto (2002), who described NH4-illite in organic-rich shales associated with coal seams in north Portugal, using a combination of SEM, TEM, AEM and other techniques.

Illite may also be diagenetic as a weathering product of feldspars (Vassilev and Vassileva, 1996). As observed by Burger et al. (1990), it occurs as crystals of columnar, tabular and vermicular forms, and as pseudomorphs after biotite and feldspar in tonsteins. Illite in high rank coal may be a diagenetic replacement of montmorillonite and/or kaolinite (Burger et al., 1990; Diessel, 1992). Burger et al. (1990) found that the proportion of illite increases in

Chapter 2 Mineral Matter and Trace Elements in Coal Seams tonsteins that are associated with coal of decreasing volatile matter (VM) percentages, accompanying chloritization when VM is less than 8%.

Although the increasing crystallinity of illite cannot be precisely correlated with individual stages of anthracitization (Teichmuller and Teichmuller, 1979), illite crystallinity in sedimentary rocks has been found to be associated with their metamorphic grade in a number of published works. For example, Maynard et al. (2001) report that the crystallinity of illite increases with increasing vitrinite reflectance in shales from Cuba, and suggest that illite may be recrystallised from detrital illite. Hower et al. (1976) noted that the proportion of illite layers in I/S increases with increasing burial depth in shales of Oligocene-Miocene sediments from the Gulf Coast of the US, due to burial metamorphism. Hower et al. (1976) also assumed that other K-bearing minerals supplied K and Al in a process of illitisation of smectite layers, represented by the reaction: smectite + K+ + Al3+ĺLOOLWH 6L4+. In addition, the duration of metamorphism has been proved to be an important variable that influences clay mineralogy (e.g. Hillier and Clayton, 1989).

The correlation between illite crystallinity and vitrinite reflectance is not always observed in either rock or coal samples (e.g. Uysal et al., 2000). In fact, illite crystallinity may depend on a number of variables besides organic maturity. According to Bayan and Hower (2012), a lack of K+, an excess of Ca2+ or Mg2+, a closed system from which fluids cannot escape, or an association of organic matter with clays may retard the development of illite crystallinity.

Smectite The smectite minerals are a group of clay minerals that are composed of two tetrahedral

(SiO4) and one octahedral (Al(OH)6) layer, with loosely held and exchangeable cations and water between the unit layers. Smectites have varying chemical compositions due to complex isomorphic substitutions within the tetrahedral and octahedral layers (e.g. Drits, 2003). The basal d-spacing of about 15 Å characteristically swells to about 17-18 Å on ethylene glycol saturation, and the original 15 Å collapses to 9-10 Å after heating at 300 qC for at least 1 hour (Carroll, 1970).

Smectites are common constituents of some Australian coal seams, where they are concentrated in inter-seam tuffs or tuff-derived clay bands (Diessel, 1992) or incorporated in the coal as detrital constituents.

Ward et al. (1989) described smectite of pyroclastic origin in an Australian coal that was converted to illite due to igneous intrusions; however, only the smectite in samples that were in direct contact with the intrusive body were affected, even though the organic matter further away had an elevated vitrinite reflectance. Ward et al. (1989) suggested that only the smectite that obtained K+ from direct contact with the intrusive body was illitised. A similar observation was made by Uysal et al. (2000) for samples from the Bowen Basin coal measures, where thin igneous intrusions have very limited influence on illitisation of I/S. Uysal et al. (2000), among others, also point out that clay mineral reactions are too slow to record the effect of extremely short-lived heating events.

Interstratified clay minerals Illite, smectite and to a lesser extent chlorite and vermiculite are reportedly involved in interstratifications of clay minerals in coal. Both regular and irregular interstratifications can occur in the interstratified structure. Rectorite, for example, is a specific name given to a regular interstratified mineral of mica-like and montmorillonite-like minerals (Brown and Weir, 1963).

Most of the expandable-lattice components in either regular or irregular interstratified structures are of a smectitic nature. Mixed-layer illite/smectites (I/S) are the most common among all the mixed-layer clay minerals in coals. As noted by Ward (1978) in some coals from the Newcastle Coal Measures, the presence of expandable clays in coal that do not fully expand to a d-spacing of 17.4 Å on glycolation indicates an interlayering of smectite and illite, rather than a smectite structure.

Like smectite, mixed-layer clay minerals are also common constituents of Australian coals, where they are concentrated in tuff-derived clay bands (Diessel, 1992). Both regular and irregular (randomly) interstratified illite/smectite (I/S) occur in Australian coals (Ward, 1978; Ward and Christie, 1994). Susilawati and Ward (2006) described rectorite in some Indonesian coals, and suggested that the rectorite was the result of interaction of kaolinite with non-mineral inorganics (e.g. K and Na) in the original lower-rank coal, formed during thermal metamorphism generated by igneous intrusions.

As noted above, diagenetic transformation of smectite or smectite layers in I/S to illite may be associated with burial diagenesis or rank increase (Perry and Hower, 1970; Hower et al., 1976). Hillier et al. (1995) found that the expandability of I/S in mudrock samples from the Pannonian Basin of central Europe decreased with increasing burial depth, and

Chapter 2 Mineral Matter and Trace Elements in Coal Seams random I/S changed to ordered I/S at a certain depth (2500 m). Renton (1982) also suggested that the diagenetic conversion of I/S to illite in coal may take place during metamorphism or coalification. However, the conversion of I/S to illite during coal diagenesis is yet to be quantitatively demonstrated (Kimura et al., 1994).

2.1.2.2 Chlorite

Two species of chlorite are naturally common, clinochlore (Mg-rich) and chamosite (Fe- rich), although compositional variations occur within each. XRD patterns with even reflections ([002] and [004] peaks) dominating over odd ones ([001] and [003] peaks), indicate chlorites with high Fe contents. By contrast, chlorites that have strong odd peaks and weak even peaks are Mg-rich. However, it may be difficult to distinguish iron-rich chlorite from kaolinite if the chlorite is present in small quantities (Ward, 1978).

Chlorite can be present in coal that formed under either freshwater or marine conditions, although its proportions are usually higher in coals that have a marine influence (Renton, 1982). Although a minor constituent of the clay mineral assemblage, chlorite with a range of occurrences has been found in coal.

Chlorite can be epigenetic, formed by hydrothermal fluid injection (Huang et al., 2007). Faraj et al. (1996) report chlorite, along with kaolinite and illite, in cleats of coals from the Bowen Basin, Australia. Dai et al. (2010a) described Fe-rich chlorite, or chamosite, replacing kaolinite in coal cell cavities from the Songzao coalfield, China, and suggested a reaction of kaolinite with Fe-Mg-rich fluids during early diagenesis. A similar occurrence of chamosite was also noted by Wang (2009).

Some chlorite of authigenic origin may be related to coal seam rank advance. Ward and Christie (1994) suggested that some fine-grained chlorite in semi-anthracite from the Bowen Basin, Australia, was possibly due to the effect of rank advance. Permana et al. (2010) also reported chlorite, probably formed by hydrothermal alteration, in another Bowen Basin coal deposit. As with illite, chlorite crystallinity has also been observed to be related to the grade of metamorphism in sedimentary rocks. For example, correlation between chlorite crystallinity and vitrinite reflectance was observed in rock samples by Maynard et al. (2001) and Uysal et al. (2000), indicating authigenic growth of chlorite.

2.1.2.3 Quartz

Quartz is the most common oxide mineral occurring in coal. It largely occurs as discrete grains of detrital origin, as biogenic or authigenic cell or pore infillings, or as veins in the form of crystalline quartz. Dissolved silica to form these infillings may be derived from the weathering of feldspars, micas or volcanic ash.

A high proportion of the quartz commonly occurs in coal plies near the top and the bottom of coal beds. In most cases this material is probably detrital, originating from the same material that makes-up the roof and floor strata. Although detrital quartz is the dominant form of silica in coal, quartz of authigenic or biogenic origin in coal is not rare. Ward (1991) noted a hard siliceous siltstone floor of a Tertiary coal seam in the Mae Moh Basin in Thailand, and suggested that an accumulation of authigenic or biogenic silica may partly be involved in its formation. Sykes and Lindqvist (1993) described diagenetic quartz and amorphous silica of different forms in Tertiary coals from a number of New Zealand coalfields. They suggested that silicification of some coals was due to the sub-horizontal infiltration of the peat bed by silica-saturated groundwater and crystallization of quartz at greater depth; the silica may in turn have been derived from leaching of the basement rocks or from siliceous phytoliths within the coal-forming plant material. Authigenic quartz may also have been derived from the alteration of volcanic ash, which may lead to the formation of porcellanite or chert-like rocks (cf. Loughnan and Ray, 1978).

Dai and Chou (2007) described authigenic quartz coexisting with chamosite as cell infillings in chamosite-rich coals from the Zhaotong Coalfield, China. They found that quartz, kaolinite and chamosite are closely related in origin, and suggested that the quartz was a product of desilicification of kaolinite by reaction with a Fe-Mg-rich fluid.

Quartz of detrital origin generally has particles of silt- to sand-size (Kemezys and Taylor, 1964). Ren (1996) has shown that quartz of detrital origin in many Chinese coals usually has a silt size (0.0625 - 0.0039 mm). Dai et al. (2008c) stated that ultra fine-grained (<0.5 μm) quartz is unlikely to be of detrital origin. They also regarded quartz with a grain-size of <20 μm scattered in collodetrinite in the coal from Xuanwei, China, as being of authigenic origin. Tian et al. (2008) and Large et al. (2009) reported nano-sized quartz from the Late Permian coals of Yunnan Province, SW China, and hypothesized that such nano-quartz was associated with increased lung cancer risk.

Chapter 2 Mineral Matter and Trace Elements in Coal Seams

High-temperature quartz (beta-quartz) has been recorded in coals formed by deposition of volcanic ash-falls (e.g. Dai et al., 2008b). However, not all quartz with a euhedral or bipyramidal form indicates a volcanic origin. Doubly-terminated bipyramidal quartz has been noted in coals without volcanic material in the associated sequences (Ward, 1992; Sykes and Lindqvist, 1993).

Conclusive identification of the genesis of quartz can be made by cathodoluminescence studies. Non-luminescent, blue-luminescent and orange-luminescent quartz particles are indicative of authigenic, volcanic and low temperature/metamorphic (detrital) origin, respectively, as demonstrated by Ruppert et al. (1985), Ruppert et al. (1991) and Ruppert and Moore (1993) on coal and shale samples from Pennsylvania, USA.

2.1.2.4 Sulphides

Pyrite and, to a lesser extent, marcasite are the most important iron sulphides in coal, and can be especially abundant in coal seams formed under marine influence. They may have either syngenetic or epigenetic genesis with a wide range of morphologies, formed during all stages through peat accumulation to post-coalification. A gradation between syngenetic and epigenetic sulphides may also be present in some cases (Ward, 2002).

Syngenetic pyrite is typically present in the form of isolated or clustered framboids, euhedral crystals, or cell infillings. Clustered pyrite framboids are frequently embedded in clay mineral bands. Massive pyrite usually cements pyrite framboids and euhedral crystals, or coats marcasite spherulites. It is thought by Querol (1989) to have formed in the early diagenetic stage, based on multiple phases of sulphide mineralization in a series of Spanish coals. Lowers et al. (2007) suggest that pyrite of massive texture in the Bengal Basin sediments may reflect slower growth relative to framboidal and euhedral pyrite, a factor attributed to limited sulphate availability.

Pyrite also occurs as epigenetic cleat and fracture infillings. Pyrite has been noted as an alteration product of other syngenetic minerals, such as siderite (Smyth, 1966) and ferroan-dolomite/ankerite (Goodarzi, 2002), and sometimes occurs as a replacement of cellular or wood structures (e.g. Wiese and Fyfe, 1986; Taylor et al., 1998) or bacteria and algae (e.g. Kostova et al., 1996; Dai and Chou, 2007).

Marcasite is intimately associated with pyrite in some coals. It may exhibit radiating crystalline growths that precipitated on framboidal pyrite nuclei, as described by Querol et

17 al. (1989) and Wiese and Fyfe (1986). Both massive marcasite cementing pyrite framboids (Querol et al., 1989) and epigenetic marcasite on cleats (Ward, 2002) are also described.

The formation of syngenetic pyrite requires the availability of dissolved ferrous iron and

H2S from bacterial reduction of sulphate in the peat swamp. Renton and Bird (1991) suggest that a higher pH (>4.5) is favoured for production of disulphide ions, resulting in the formation of iron disulphide minerals. The factors controlling precipitation of marcasite rather than pyrite are not yet well-defined (Ward, 2002). Marcasite is suggested to be precipitated with lower pH values in the micro-environment (Querol et al., 1989). A number of experimental results (e.g. Murowchick and Barnes, 1986; Schoonen and Barnes, 1991; Benning et al., 2000) have also indicated that marcasite formation is strongly favoured over pyrite at low pH (approximately 4 to 5).

Most of the sulphur is derived from seawater sulphate, and thus the presence of syngenetic pyrite in coal is generally an indicator of a marine-influenced environment in the precursor peat. For pyrite of epigenetic origin this is not necessarily the case, with the presence of pyrite determined by factors less related to the peat. Syngenetic pyrite can also be present in non-marine peat or coal beds. Chagué-Goff et al. (1996), for example, reported pyrite of framboidal and anhedral forms associated with a high-sulfur, freshwater peat in Alberta, Western Canada, although the pyrite in that case does not appear to be very abundant. The authors suggest that sulphate-rich groundwater seeping through the peatland may be involved in the pyrite formation.

2.1.2.5 Carbonates

A wide range of carbonate minerals have been found in coal. The most common are siderite, calcite, dolomite and ankerite. These carbonates are mostly of authigenic origin with various chemical compositions. They can form during deposition, and also in association with coalification or post-coalification processes (Taylor et al., 1998).

Siderite in coal is mostly a syngenetic mineral, typically occurring as spheroidal nodules, some of which may have a radiating structure. Siderite in the form of euhedral crystals, cell infillings, and, to a lesser extent, replacement of maceral components is also common.

Chapter 2 Mineral Matter and Trace Elements in Coal Seams may be replaced by pyrite, a process that may be the result of environmental changes, such as changes from weakly acid to weakly alkaline conditions (Kortenski and Kostova, 1996).

Most other carbonates, especially calcite, dolomite and ankerite, occur more commonly as cleat and fracture infillings, precipitated from pore fluids containing ions such as Ca, Mg and Fe during the epigenetic stage (Spears and Caswell, 1986). These ions, as noted above, may be derived from the non-mineral inorganics released from the organic matter during coal rank advance. Ward et al. (1996; 1999a) have reported abundant sub- horizontal veins of carbonates in the coal at the very top of two Australian coal seams, which may be a result of crystallisation under pressure following fluid migration.

Veins of carbonate minerals commonly show significant chemical variation. Patterson et al. (1995) describe the paragenesis of carbonate minerals in a range of Australian coal seams. Three stages of carbonate precipitation were found: nodular siderite with minor proportions of Mn, Mg and Ca; Mg-rich siderite occurring as overgrowths on siderite nodules and as cell lumen infillings; veins of ferroan dolomite-ankerite with a wide range of Mg and Fe contents; and essentially pure calcite cleat infillings.

A sequence of calcite precipitation as multi-stage overgrowths and cross-cutting veins was found in Illinois Basin coals by Kolker and Chou (1994), using a combination of cathodoluminescence and geochemical analysis. Fe contents (determined by X-ray fluorescence synchrotron microprobe) ranged from moderate in the earlier-formed calcite to high in the late formed calcite. Dai et al. (2005a) noted at least three stages of ankerite veins that were precipitated from fluids, based on the different Ca/Sr and Fe/Mn ratios of a anthracite from the Dafang Coalfield, Guizhou, China. Golab et al. (2007) reported carbonates ranging from calcite to ankerite and dawsonite, which were precipitated from multiple fluid invasions which were derived either from igneous intrusion or the coal that was intruded.

Dawsonite, though not common in other coal seams around the world, is relatively widespread in the coals of the Sydney Basin, Australia (Goldbery and Loughnan, 1977). It is mostly an authigenic mineral occurring as veins or void infillings (Loughnan and Goldbery, 1972; Loughnan and See, 1976), or as a replacement of other minerals, such as sanidine (Dai et al., 2008b). Golab et al. (2006) used isotopic data to investigate cleat dawsonite, and suggested that the dawsonite is the result of reactions between Na2CO3-or

NaHCO3-bearing fluids and earlier-precipitated kaolinite in the cleat, and that there was a magmatic carbon source for the dawsonite.

2.1.2.6 Phosphates

Apatite and a range of alumino-phosphate minerals are the most common phosphates in coal, and have been identified in a number of coal seams around the world, sometimes with significant proportions in particular sub-sections of the coal beds (e.g. Ward et al., 1996; Rao and Walsh, 1997, 1999; Dai et al., 2006a). The apatite is generally identified as fluorapatite (Ca5(PO4)3F), but other varieties such as chlorapatite (Ca5(PO4)3Cl), hydroxy- apatite (Ca5(PO4)3(OH)) and carbonate-apatite also occur (Ward et al., 1996).

Aluminophosphate minerals include end-members of goyazite (SrAl3(PO4)2(OH)5+2O), gorceixite (BaAl3(PO4)2(OH)5·H2O), crandallite (CaAl3(PO4)2(OH)5+2O) and florencite

(CeAl3(PO4)2(OH)6), together with a range of intermediate (solid solution) phases.

Apatite and Sr-, Ba-, Ca- aluminophosphate minerals, sometimes coexisting with clay minerals, mainly occur in the form of cell and pore infillings. Ward et al. (1996) also noted that apatite and gorceixite may occur as cleat and fracture fillings in Australian coals.

The phosphorous for formation of phosphate minerals may be derived from the decomposition of plant material (Rao and Walsh, 1997). The cations Ca, Ba and Sr within the aluminophosphate minerals are attributed to the alteration of volcanic ash in mires (Rao and Walsh, 1997; Brownfield et al., 2005). High phosphorus concentrations are commonly found at particular horizons of coal seams; this may be controlled by local hydrogeochemical factors or by introduction of additional sources of phosphorus, e.g. volcanic debris, shells or faecal matter (Ward et al., 1996; Rao and Walsh, 1999).

As discussed by Ward et al. (1996), whether formation of apatite or aluminophosphates took place was probably controlled by local hydrogeochemical factors and the availability of alumina. Apatite is precipitated in neutral to alkaline conditions, whereas aluminophosphate minerals are less soluble at acid pH levels; aluminophosphate minerals would be expected from intra-seam precipitation if Al was also available in reactive form at the site of phosphate deposition, and apatite if Al was not available to react with the precipitated phosphatic material.

Authigenic rare earth element phosphate minerals, such as monazite, have been noted in coal (e.g. Hower et al., 1999a; Wang, 2009). These are typically deposited in cracks of

Chapter 2 Mineral Matter and Trace Elements in Coal Seams coal plies which are adjacent to tonstein bands, and were probably precipitated from the leachate of volcanic ash.

2.1.2.7 Sulphates

Sulphates such as coquimbite, jarosite (or natrojarosite), copiapite (or ferricopiapite) and szomolnokite are common components of coal or coal LTAs as secondary minerals (Susilawati and Ward, 2006; López and Ward, 2008; Oliveira et al., 2012). These are most likely derived from oxidation of sulphide minerals during coal storage (Ward, 2002). Another range of sulphate species, such as bassanite, hexahydrite and alunogen, are usually formed as artefacts, produced by interaction of inorganic elements and organic sulphur during the low-temperature ashing process (Frazer and Belcher, 1973). Gypsum may be present in lower-rank coal (Koukouzas et al., 2010; Dai et al., 2012c), either as an authigenic mineral in the coal or by precipitation of Ca and SO4 in the pore waters during drying. It may also be produced by the reaction between calcite and the sulphuric acid which is produced by oxidation of pyrite with sample exposure and storage (Pearson and Kwong, 1979). Partial dehydration of gypsum during sample drying or low-temperature ashing may also result in the formation of bassanite (Ward, 2002).

2.1.2.8 Feldspars

Feldspar minerals are frequently present in coal as minor phases, but rarely in concentrations >1 or 2% (Renton, 1982). Albite and orthoclase are mostly of epiclastic origin (e.g. Vassilev and Vassileva, 1996). Feldspar minerals (e.g. sanidine, albite and anorthoclase) in coal may also be relic pyrogenic minerals. Sanidine, for example, has been observed in coal that has been affected by volcanic ash falls (e.g. Ruppert and Moore, 1993; Dai et al., 2008b). Dai et al. (2008b) have also noted albite replacement of sanidine in a volcanic-influenced high organic sulphur coal from the Yanshan Coalfield, China, where the albite appears to be derived from hydrothermal fluids.

2.1.2.9 Other minerals

As noted above, accessory minerals generally occur in low concentrations. Some may not be detected by XRD analysis, and only observed during optical and electron microscopy. However, sometimes accessory minerals may be in significant proportions in some subsections of coal seams.

Sulphide and related minerals in coal, other than iron sulphides, include sphalerite, galena (Spears and Caswell, 1986; Hower et al., 2001), millerite (Lawrence et al., 1960), stibnite (Karayigit et al., 2000), chalcopyrite (Cressey and Cressey, 1988), clausthalite (Kolker and Finkelman, 1998; Hower and Robertson, 2003), pyrrhotite (Vassilev et al., 1994) and alabandite (Dai et al., 2007b). These minerals are typically precipitated as epigenetic veins that formed later in the diagenetic stage, and also syngenetically in cell and pore lumens, and as discrete crystals in macerals.

Bauxite-group minerals that have been identified in coals include boehmite, diaspore and gibbsite. Gibbsite may be precipitated from Al-rich solutions in the peat (Ward, 2002). A common model of formation of gibbsite involves removal of silica and other ions from clay minerals in an acidic condition or by acidic hydrothermal fluids (e.g. Howard and Fisk, 1988). However, gibbsite is soluble at pH<4 (Loughnan, 1969), and would react with silica to form authigenic kaolinite (Ward, 2002) or dehydrated to form boehmite at a higher temperature (e.g. Deer et al., 1992). For example, diaspore was noted in high-rank coal from the South Walker Creek area, Bowen Basin, Australia, possibly due to intense leaching of clay minerals during hydrothermal activity (Permana et al., 2010).

Dai et al. (2006a; 2008a) suggested that abundant boehmite in medium to high-volatile bituminous coal from the Jungar Coalfield, Ordos Basin, China may be formed from partial dehydration of gibbsite which was initially derived from oxidized bauxite in the sediment source region. Abundant diaspore in the Adaohai Coalfield, Inner Mongolia, China, as suggested by Dai et al. (2012d), was also formed due to dehydration of gibbsite at high temperatures that may have caused by igneous intrusions, with the gibbsite also being derived from the exposed and weathered bauxite in the sediment source region.

2.1.3 Intra-seam volcanic claystones

2.1.3.1 Definitions of clay rocks

Intra-seam clayrock layers with a volcanic ash origin are common in coal-bearing sequences, and have been described as tonsteins or bentonites in the literature. Although the actual definition of tonsteins is still regarded in some cases as controversial, Spears (2012), among others, define tonsteins as “thin, widespread clay-altered layers of volcanic ash, dominated by kaolinite, that are commonly found in coals and associated sediments”.

Chapter 2 Mineral Matter and Trace Elements in Coal Seams

Bentonite, regarded by Spears (2012) as being comparable with tonstein but formed in a more marine environment, was originally described and named by Knight (1898), and the term is well established as representing any clay which is composed dominantly of a smectite group mineral (e.g. Wright, 1968). Although some definitions indicate an airborne volcanic ash origin for bentonites, according to Grim and Güven (1978), bentonite can also be produced by hydrothermal or deuteric alteration of volcanic rocks other than tuff, and formed in certain sedimentary environments without association with any volcanic activity.

K-bentonite, another frequently used term in the literature, was intended by Weaver (1953) to reflect a relatively high I/S content to differentiate it from the Ca-, Na- and Mg-rich smectitic bentonites. The high I/S content in K-bentonite is attributed to potassium fixation, which transformed smectite into I/S during the diagenetic process (Huff, 1983). As suggested by Spears et al. (1999a), tonstein-related clayrocks mainly consisting of I/S may be more appropriately referred to as K-bentonites.

As mentioned above, controversy exists on the definition and use of the terms “tonstein” and “bentonite”. Bohor and Triplehorn (1993) regard all altered volcanic-ash layers occurring in non-marine environment as tonsteins, and restrict the term bentonite to deposits formed in more marine settings, regardless of their clay mineralogy (no matter whether kaolinite-rich or smectite-rich). However, other authors restrict the definition of bentonites to smectite-rich deposits irrespective of their genetic origin or sedimentary settings (e.g. Grim and Güven, 1978), and tonsteins to kaolinite-rich altered volcanic ash deposits with no specific sedimentary setting (Lyons et al., 1994). For example, smectite- rich clay bands in coal seams are referred to as bentonites by some authors (Pevear et al., 1980; Senkayi et al., 1984). The definition of tonstein suggested by Spears (2012) is adopted in this thesis. Following Lyons et al. (Lyons et al., 1994), claystones of volcanic ash origin with a kaolinite content greater than 50% are regarded as tonsteins. Likewise, the claystones are referred to as bentonites and K-bentonites, respectively, when smectite or I/S exceeds 50% of the clay mineral assemblage. When more advanced diagenesis, such as pronounced illitisation and possibly chloritization of kaolinite takes place in the original tonstein, metatonstein is possibly a better term to use (Admakin, 2002; Spears, 2012).

Tonsteins, and less commonly bentonites and K-bentonites, are known to occur worldwide in many different coal-bearing formations throughout geological time according to

23 numerous literature sources such as those referred to herein. Tonsteins are thought to provide geochronological markers, with a widespread distribution that enables stratigraphic correlation, and hence they have been used to correlate coal beds in many coalfields. The primary minerals in tonsteins, such as sanidine, if preserved, may provide opportunities for absolute age determination.

The appearance of individual tonstein bands may vary laterally or vertically. Dense (fine grained) tonstein occasionally grades into normal shale or material of sandy appearance (Diessel, 1965). However, a tonstein band may appear to be dense at the bottom and top where intimately associated with coal (mixed with humic compounds), but coarser in the middle of the section (Diessel, 1965). This suggests that volcanic materials tend to be decomposed more completely than otherwise when affected by peat-forming environments.

“Flint clay”, a term commonly used in the field, has also been used in the literature. Flint clays are dense, non-plastic claystones that lack bedding, break with a conchoidal fracture, and are essentially composed of well-ordered kaolinite (Loughnan, 1971b). Tonsteins that appear to be flinty are described as dense tonsteins in some literature. Tonsteins may appear flinty with a matrix dominated by fine-grained materials and have a conchoidal fracture, or they may be granular with the granules composed of kaolinite (Spears and Lyons, 1995). Flint clays in some coal-bearing sequences are recognized as tonsteins, such as the so-called Fire Clay tonsteins of the Appalachian Basin in the USA (e.g. Lyons et al., 1992; Greb et al., 1999). However, as pointed out by Bohor and Triplehorn (1993), not all flint clays, especially thick ones, have a volcanic origin, and most tonsteins do not have the texture of flint clay.

It should thus be noted that, although tonsteins of apparent volcanic origin have been frequently reported in coal seams around the world, other processes may also be responsible for the thin beds of non-coal sediment commonly found within coal seams (Ward, 2002). Not all thin claystone layers in coal are necessarily tonsteins. Tonsteins that have undergone extensive diagenetic and post-diagenetic alteration may not necessarily contain apparently diagnostic indicators such as volcanogenic minerals and other typical petrographic textures. Caution must therefore be exercised in interpreting the origin of inorganic layers in coal, based not only on macroscopic appearance, but also on mineralogical, textural and chemical characteristics (Ruppert and Moore, 1993).

Chapter 2 Mineral Matter and Trace Elements in Coal Seams

2.1.3.2 Clay mineralogy of volcanic claystones

The clay composition of tonsteins is often monomineralic and dominated by kaolinite (Loughnan, 1978). Although smectite (Addison et al., 1983), I/S (Addison et al., 1983; Wang, 2009; Dai et al., 2011) and illite (Kisch, 1966; Burger et al., 1990; Bieg and Burger, 1992) have been reported as the principal constituent, rather than kaolinite, in some materials described as tonsteins, bentonite and K-bentonite are perhaps better terms when the sediments have more smectitic and illitic compositions, respectively. Lateral transitions from one mineral assemblage to another have also been recorded (e.g. Price and Duff, 1969; Spears and Duff, 1984).

Kaolinite In contrast to the poorly-ordered structure of kaolinite in lutites forming the roof and floor materials of the same coal seams, the kaolinite in tonsteins typically has a well-ordered crystal structure, as indicated by X-ray diffraction analyses. it is mainly derived from the devitrification of volcanic glass, but may also be formed by alteration of primary minerals such as biotite, feldspar, amphibole and pyroxene (Bohor and Triplehorn, 1993). Kaolinite in tonsteins shows a variety of morphologies, as described in the extensive literature (e.g. Diessel, 1992; Bohor and Triplehorn, 1993; Knight et al., 2000). It can occur as grains of vermicular, prismatic or tabular crystals and pseudomorphs after feldspars or micas, set in a cryptocrystalline kaolinite matrix. Some tonsteins also contain grains with a graupen texture, with the graupen being spheroidal aggregates of cryptocrystalline to microcrystalline kaolinite.

Vermicular kaolinite, although not restricted to tonsteins, is often used as evidence that the sediment is a tonstein (Spears, 1971; Ruppert and Moore, 1993), and the presence of visible kaolinite vermicules may aid in the field identification of tonsteins (Bohor and Triplehorn, 1993). Graupen texture of tonsteins was ascribed by Diessel (1983) to the kaolinitisation of biotite, which leads to the development of subspherical shapes instead of the original lath-shape.

Smectite Smectite can occur in tonsteins in various proportions, although bentonite is a more appropriate term when smectite is the principal clay mineral. Whether the volcanic ash altered to kaolinite or to smectite may be related to the original ash composition (Bohor and Triplehorn, 1993). Bohor and Triplehorn (1993) suggest that the original ash of

25 smectite-dominant tonsteins was commonly intermediate in composition. The clay mineral formation is also a function of the water chemistry during and after deposition (Spears, 1971). Volcanic ash tended to alter to smectite-dominant claystones if deposited in a marine environment, where the marine water provides sufficient concentrations of relevant active ions.

The clay mineralogy of tonsteins may also reflect the efficiency of leaching. Kaolinite is formed instead of I/S or smectite as a result of more extreme alteration (Spears, 1971). An intermediate smectitic stage may be involved in the leaching process, although volcanic ash can also alter to kaolinite directly (Triplehorn and Bohor, 1986). Bohor and Triplehorn (1993) noted that thin partings and the margins of thick tonsteins contain greater proportions of kaolinite. A similar observation was made by Diessel (1992) in Australian tonsteins, suggesting that kaolinitisation has progressed furthest where the volcanic ash has had the closest contact with organic matter.

Mixed-layer illite-smectite (I/S) Mixed-layer illite-smectite (or I/S) sometimes dominates thin claystone beds deposited in coal-forming environments, forming material described as K-bentonite by Spears et al. (1999a). X-ray diffraction analyses show that the I/S in K-bentonite is typically regularly interstratified, with various percentages of illite and smectite layers (e.g. Spears, 1971; Pevear et al., 1980; Altaner et al., 1984), although randomly interstratified I/S has also been reported (e.g. Huff et al., 1998).

The I/S in K-bentonites is generally thought to be developed from smectite during burial diagenesis (Spears, 2012). Altaner et al. (1984) suggest that the I/S in K-bentonite was produced by reaction of smectite in the original bentonite with K in the pore fluid, with the K being derived from breakdown of K-bearing minerals (e.g. micas and K-feldspar) in the host rock. However, exceptions may occur. For example, Huff and Tuerkmenoglu (1981) noted that K-fixation occurred in an American K-bentonite without the material having been subjected to the necessary burial conditions. They considered that illitisation during burial diagenesis appears to require a precursor of beidellitic smectite, while K-fixation in response to substitution of Mg2+ for Al3+ in the montmorillonite precursor in their samples may readily take place before burial diagenesis.

Chapter 2 Mineral Matter and Trace Elements in Coal Seams

Illite Illite occurs in materials described as tonsteins, as noted by Burger et al. (1990) and Bieg and Burger (1992), often as crystals with pillar, tabular and vermicular forms, pseudomorphs of biotite and feldspars, or graupen forms, as well as in the cryptocrystalline matrix. The texture of the illite crystals commonly resembles that of kaolinite crystals in tonsteins, which has led authors such as Kisch (1966) and Burger (1992) to suggest an illitisation process of kaolinite pseudomorphs after feldspar and biotite.

Kisch (1966) describes a chlorite-illite tonstein associated with an Australian semianthracite. He ascribes the formation of illite and chlorite to the alteration of kaolinite during diagenesis, under conditions with available K (for illite), and Fe and Mg (for chlorite). Similarly, Burger et al. (1990) found that the proportion of illite increases in tonsteins that are associated with coals of lower volatile matter (VM) percentages (i.e. higher rank), accompanied by chloritization when VM is less than 8%.

2.1.3.3 Volcanogenic minerals

Volcanogenic minerals, also referred to as “primary” minerals, are those that survive the formation of tonsteins as phenoclasts. Common volcanogenic minerals in tonsteins are quartz, biotite, sanidine, zircon, apatite, ilmenite, fayalite and magnetite. The occurrence of these minerals is indicative of a volcanic origin. The presence of relict volcanic minerals, such as zircon and sanidine, in tonsteins may enable absolute dating for chronostratigraphic correlation, and this has been reflected in a number of works (e.g. Turner et al., 1983; Lyons et al., 1992; Lyons et al., 2006; Guerra-Sommer et al., 2008).

Quartz Quartz is the most common volcanogenic mineral in tonsteins. However, the quartz content of tonsteins is generally less than that of the associated strata (e.g. Senkayi et al., 1984). Volcanic quartz, as noted previously, is luminescent in the blue range. It commonly occurs as broken fragments or flakes, sometimes as beta-forms with corrosion embayments (e.g. Zhou et al., 1982; Knight et al., 2000), and occasionally in bipyramidal forms (Bohor and Triplehorn, 1993; Lyons et al., 1994). Some of the quartz may contain glass inclusions, microprobe analysis of which can enable the determination of the original ash composition for the tonsteins (Lyons et al., 1994).

Zircon Zircon is extremely resistant to abrasion and alteration, and in tonsteins almost always occurs as euhedral crystals (Bohor and Triplehorn, 1993). Elongated or equant prismatic crystals with pyramidal terminations are common (e.g. Spears, 1971; Spears and Lyons, 1995). Kowallis and Christiansen (1989), cited by Bohor and Triplehorn (1993), suggested that zircon morphology may be useful for correlation of pyroclastic rocks, and provide a basis for assessing the composition and temperature of the parent magma. Zhou (1999) has described different crystal habits and morphological parameters of zircon grains from several tonstein horizons and from claystones of more normal sedimentary origin, and discussed the source of the materials and the mechanism of transport and deposition based on zircon studies.

Biotite Biotite in tonsteins is not as common and well-preserved as in bentonites, and also shows various degrees of kaolinitisation. Alteration of biotite stacks in tonsteins proceeds from the edges inward along the basal cleavage planes, with bleaching of the dark colour and leaching of Fe, Mg and K from the rims (Diessel, 1983; Bohor and Triplehorn, 1993).

Apatite Like biotite, apatite is uncommon in tonsteins, and where present occurs only in small proportions as fragmental and euhedral crystals (e.g. Price and Duff, 1969; Zhou et al., 1982; Burger et al., 2000; Knight et al., 2000). Triplehorn and Bohor (1983) noted that apatite is absent in kaolinitic volcanic ash layers of the Cretaceous Dakota Group from Colorado, but is common in the laterally correlative smectite-rich layers (bentonites) deposited in a marine shale.

2.1.3.4 Other secondary minerals

Various secondary minerals can be formed in tonsteins from direct replacement of primary minerals and volcanic glass, and also as recrystallisation products of the volcanic material. The clay minerals mentioned above are the most abundant secondary minerals in tonsteins.

Sr-, Ca-, Ba- and REE-aluminophosphate minerals, represented by goyazite, crandallite, gorceixite and florencite respectively, are common mineral constituents of tonsteins (e.g. Price and Duff, 1969; Loughnan, 1970; Hill, 1988). In most cases, they are not present as

Chapter 2 Mineral Matter and Trace Elements in Coal Seams any pure end member but as a solid solution of the end-member phases. Triplehorn and Bohor (1983) have described euhedral to subhedral crystals of goyazite, sometimes with an intergrowth texture, in kaolinitised tuff from Colorado, which clearly indicates precipitation from solution at an early diagenetic stage. Crandallite replacement of glass shards and at the junction of glass bubbles has also been noted in tonsteins (Bohor and Triplehorn, 1993).

Triplehorn and Bohor (1993) also noted that, although minerals of the goyazite series are present in altered tuff which was less marine-influenced, they are not present in the equivalent bentonite horizons. Spears et al. (1988) further stated that the most important phosphorus source is growing organic matter, and suggested that material with abundant phosphates was precipitated in a non-marine environment in which there was reduced competition for the necessary cations from sulphate and other anions.

Zeolite minerals appear to occur in many bentonites (e.g. Mutakyahwa, 2002; Caballero et al., 2005) and smectite-rich tonsteins, generally as alteration products of volcanic glass and feldspars. Loughnan (1966) found abundant analcime in altered tuff samples from the Newcastle Coal Measures, with the principal clay mineral being I/S. Another zeolite, clinoptilolite, has been noted by Senkayi et al. (1984) in an Eocene tonstein band in Texas, with the authors suggesting that clinoptilolite formed after kaolinizaton of the initial bentonite. Coombs (1961) suggested that zeolite formation is favoured by devitrification of volcanic glass in saline conditions. Grim and Güven (1978) suggested that the formation of clinoptilolite appears to be favoured by alkaline conditions and low intensity of leaching, whereas kaolinite is the stable diagenetic mineral under intense leaching.

Other secondary minerals, including anatase (Price and Duff, 1969), halloysite (Ward and Roberts, 1990) and monazite (Lyons et al., 1992; Zhou et al., 1994), have also been recorded in tonsteins or reworked tuffs associated with coal seams.

Normal clastic sediment can be included in the tonsteins, reflected by the presence of detrital quartz, other clay minerals, and heavy minerals. These can be identified by their crystal habits, X-ray diffraction characteristics, or appearance under cathodoluminescence. Diagenetic minerals such as pyrite and carbonates are also common in tonsteins.

2.1.3.5 Indicators of volcanic influence in coal

Leaching of tonsteins or the incorporated volcanic ash by ground waters and organic

29 acids in the peat-forming environment would also be expected to result in enrichment of some elements that have been released from the leachate and accumulated as minerals or organometallic compounds in the coal. The occurrence of such material may therefore be indicative of volcanic influence during peat accumulation and coal formation. Crowley et al. (1989) have concluded that one of the primary factors responsible for trace-element enrichment in coals was leaching of soluble components from tonsteins.

Goodarzi (2002), among others, has indicated that Ca-, Ba-, Sr- and REE- aluminophosphate minerals in coal may provide an indication of volcanic activity. Ruppert et al. (1993) have also noted rare earth crandallite group minerals in an Indonesian peat deposit, which was thought to indicate phosphate formation before most other diagenetic processes. Senkayi et al. (1984) suggested that the formation of clinoptilolite in both lignitic and associated non-lignitic strata was due to the leaching of Na, K and silica from the overlying smectite-rich volcanic layer.

In some cases the volcanic ash was not sufficiently abundant to form visible ash layers in the coal beds (Dewison, 1989), or only clayey micro-sized bands (around 100 microns in thickness) are developed (Dai et al., 2007b). Nevertheless the volcanic material may still have an important impact on the coal mineralogy and geochemistry. A special case is the presence of discrete, clay-free volcanic ash partings in coal seams, although these are not necessarily regarded as tonsteins. Triplehorn et al. (1991) have identified this kind of material in the Paleocene Wyodak-Anderson coal bed of the Powder River Basin, and suggested that the absence of clay minerals was due to extensive leaching and alteration.

2.1.4 Summary

Mineral matter in coal includes discrete crystalline mineral particles and non-mineral inorganic elements, other than N and S, in exchangeable or organometallic forms associated with in the organic matter, or dissolved in the pore water. Non-mineral inorganic elements are important in low-rank coals, but are negligible in higher-rank coals, where discrete crystalline minerals are the dominant.

Minerals can be present in coal with a range of mode of occurrence. Detrital minerals typically occur in the form of band (clay minerals). Authigenic minerals typically occur in the forms of framboids (pyrite), concretions (siderite), cell/pore infillings (kaolinite, pyrite and carbonates), cleat/fracture infillings (kaolinite, pyrite and carbonates). Some minerals

Chapter 2 Mineral Matter and Trace Elements in Coal Seams

(bassanite, gypsum) present in coal or the LTA may be the artificial product of the ashing or storage process. As the recent advance in the direct analysis of minerals, more minerals with low proportions can be identified.

Tonsteins and bentonites in coal seams are altered volcanic ash layers. The clay composition of tonsteins is often dominated by kaolinite, while bentonite and K-bentonite are used to represent the sediments having more smectitic and illitic compositions. A range of volcanogenic and diagenetic minerals may be present in the altered volcanic ash layers.

2.2 Trace elements in coal

Coal contains most elements within the periodic table in major, minor or traces concentrations (Swaine, 1990). Major elements usually dealt with in the literature are Al, Si, Na, Mg, S, Fe, K, Ca, Ti, and P (Gluskoter, 1975), as well as C, O, H, and N, the concentrations of which are normally above 0.01% (Dai et al., 2005c). Trace elements in coal are generally regarded as those elements with concentrations up to around 1000 ppm (Lindahl and Finkelman, 1986; Swaine, 1990). Despite being classified as trace elements under this definition, some of these elements may be present in some coals in higher concentrations. An understanding of trace elements in coal is important from genetic, economic and environmental points of view (e.g. Spears and Zheng, 1999; Dai and Chou, 2007).

Trace elements may be mobilized during coal mining and storage, and redistributed in the air, or in nearby surface and underground waters (Swaine, 1990). Elements in coal are redistributed into fly ash, bottom ash, boiler slag, emission gases and waste products in other forms, during combustion processes. Trace element partitioning and redistribution during combustion of coal (e.g. Hower et al., 1996; Hower et al., 1999b; Spears and Martinez-Tarrazona, 2004; Li et al., 2005; Vassilev et al., 2005; Dai et al., 2010b) and the potential environmental impact of the use and disposal of coal combustion products (e.g. Karuppiah and Gupta, 1997; Drakonaki et al., 1998; Wang et al., 1999b; Dutta et al., 2009; Ward et al., 2009) have been extensively studied. Some volatile elements, such as V and Ni, can cause pitting of metal surfaces (Lindahl and Finkelman, 1986). High concentrations of halogens, such as chlorine and fluorine in the feed coal, may cause technological problems such as etching, fouling, slagging and corrosion in the combustion

31 systems (Sowa et al., 1989; Hower et al., 1997a; Vassilev et al., 2000). In addition, some technological concerns such as poisoning of catalysts, may also be caused by trace elements in coal.

Human health problems have been connected with environmental hazards caused by coal combustion in different parts of the world. Finkelman et al. (1999) and Ding et al. (2001) reported endemic arsenosis in Guizhou Province, China, caused by domestic combustion of high-As coal using unvented stoves. Endemic selenosis was found in Enshi County, Hubei Province, southern China, where high Se (up to a few thousand ppm) occurs in so- called “stone coal” (e.g. Song, 1989; Yang, 1983). Endemic fluorosis in western Guizhou, however, is not attributed to combustion of high-F coal but to high-F clay binders (Finkelman et al., 1999; Dai et al., 2004, 2007a). Mercury, a potential threat to human health, is primarily released as atmospheric emissions during coal combustion (Yudovich and Ketris, 2005b). Radioactive elements like U and Th in coal, and more importantly in coal ashes, have also generated concern because of possible radiation hazard (e.g. Zielinski and Budahn, 1998).

Although particular elements in the coal may represent a potential environmental hazard, either from the coal itself or the associated preparation wastes or combustion products, they may also represent a significant source of elements for industrial purposes, depending on their nature and abundance. Valuable elements, such as Ge, Se, Ga, rare earth elements (REEs), and platinum group elements (PGEs), have been found in significant concentrations and may even form deposits of economic interest in coals in some part of the world; some elements are also being recovered or can be potentially recovered from coal ashes (e.g. Seredin, 1996, 2004; Dai et al., 2006a; Zhuang et al., 2006; Qi et al., 2007a, b; Huang et al., 2008; Seredin and Finkelman, 2008).

Knowledge of the abundance and modes of occurrence of trace elements in coal is significant in predicting their behaviour and environmental impacts during coal mining, storage, preparation and utilization. Furthermore, knowledge of trace element geochemistry may help to understand the depositional conditions, diagenetic and post- diagenetic processes, and the regional tectonic history of different coal basins (Dai et al., 2012b).

Chapter 2 Mineral Matter and Trace Elements in Coal Seams

2.2.1 Origin of trace elements in coal

The trace element geochemistry of a particular coal is the result of the interaction of the original peaty material with water- or air-borne detrital input, and solutions that circulated within the coal basin (Swaine, 1990; Kolker and Finkelman, 1998), influenced in different ways by the botanical, biochemical and geological factors that acted throughout the long- term process of coal formation (Bouška, 1981; Swaine, 1990; Hower et al., 1991).

2.2.1.1 Plant-derived trace elements

Major elements such as C, O, H and N in coal are mainly inherited from phytogenic precursors of coal, including cellulose, lignin, proteins and lipids (Diessel, 1992). The percentage of trace elements inherited from the original plants may largely depend on the nature of those original plants (Fu et al., 2004). Plant-derived trace elements in coal originate from the initial concentrations in the tissues of growing peat-forming plants, and combined with further degradation of the material (e.g. Nicholls, 1968; Cecil et al., 1982).

For example, Lyons et al. (1989) noted that the concentrations of certain elements (particularly Ca, Ce, Co, La, Sc and Sr) in the vitrinite concentrate from the Pittsburgh coal are similar to that in the tree-fern xylem, and suggested that these elements may be entrapped during the peatification stage. Boron and small amounts of Ni and Cu can be incorporated into coal due to the biological concentrations, or by sorption by organic matter from groundwaters that circulated through the peat (Nicholls, 1968).

Some ions may have been removed from the peat swamp by outgoing solutions, and biogenic minerals can be modified or dissolved in the advanced stages of peatification and coalification (López-Buendía et al., 2007). The enrichment or removal of elements in the peat swamp are controlled by biological and geochemical factors (e.g. pH, Eh) (Bouška, 1981), the nature of the ions and the organic matter (Nissenbaum and Swaine, 1976). Deul (1958) noted that plant ashes are enriched with K, Ca, Mg, Na, Li, Ba, and Si. Such elements, however, are notably depleted in peat ashes, probably because they were relatively soluble and were lost during degradation in the early stages of peatification.

Plant-derived trace elements are predominantly associated with organic matter at the early stages of coalification, but decrease as the rank of coal increases (Swaine, 1992). Deul (1955) speculated that plant-derived mineral matter is usually less than 2% of the ash yield for bituminous coals. Although plant-derived inorganic matter could account for

33 some authigenic minerals in coal (e.g. Sykes and Lindqvist, 1993), geological factors play the key role in more general enrichment of trace elements.

2.2.1.2 Geological factors responsible for trace element enrichment

In addition to derivation of trace elements from the plant material, the trace elements in coal may also have been derived from minerals introduced into the peat-forming environment, or from solutions passing through the peat or coal bed at different stages of its depositional and post-burial history. Geological factors that may influence such elements in coal have been discussed by many workers (e.g. Nicholls, 1968; Ren et al., 1999; Dai et al., 2012b). According to Goodarzi (1995), the principal geological factors are the environment of coal deposition, the nature of the country rock, the tectonic setting and the hydrologic conditions, age and rank of the coal. However, no single process can account for the accumulation of all the elements that occur in coal (Nicholls, 1968).

Paleoenvironment The inorganic geochemistry of coal has been used by many workers (e.g. Chou, 1984; Banerjee and Goodarzi, 1990; Eskenazy et al., 1994) to interpret the palaeoenvironment. Sulphur and boron are the most important elemental palaeosalinity indicators, and the concentrations of these elements are usually related to the degree of marine influence during the early stage of coalification.

A higher concentration of B in coal is attributed to greater marine influence (Swaine, 1983). Goodarzi and Swaine (1994) have suggested the following relationships between B concentration ranges and paleosalinity of the coal-forming environment: B concentrations of of <50 ppm, between 50-110 ppm, and >110 ppm in coal are indicative of fresh water- influenced, mildly brackish-influenced and brackish water-influenced settings, respectively. Boron has also been used in ratios with other elements. The B/Be ratio was proposed as a potential paleosalinity indicator by Dominik and Stanley (1993), who noted that the B/Be ratio was consistent with paleosalinity and palaeoclimate results obtained by other independent methods in the Nile delta peat, while the concentrations of B were observed to be controlled by grain size.

However, the enrichment of B is not always an indication of marine influence. In some cases, it may be an indicator of secondary enrichment due to groundwater enriched in brine passing through the seam, and may also be controlled by the nature of the country

Chapter 2 Mineral Matter and Trace Elements in Coal Seams rock (Goodarzi, 1995). Some boron may also be allochthonous, being associated with detrital clay minerals, especially illite (e.g. Bohor and Gluskoter, 1973; Hower et al., 2002).

Additional elements that have been found to be rich in coal formed within a marine- influenced environment include Cr, Cu and Ga. However, most of these are not universally applicable as marine indicators. Ratios of several elements may also useful in interpreting the palaeoenvironment. For example, Th/U ratio variation within the coals may be indicative marine influence (Gayer et al., 1999; Spears and Tewalt, 2009). Source rocks The sediment source rock is a significant contributor to the trace elements within coal, since it supplies detrital minerals and solutions to the peat swamp and influences the surface and groundwater composition (Swaine, 1990; Goodarzi, 1995; Ren et al., 1999). Especially for small-scale coal basins, detrital materials can be transported to the interior of the basin (Dai et al., 2012b). For large-scale coal basins, however, other geological factors may also be important.

The geochemical characteristics of the source rocks may be comparable to and reflected by the geochemistry of the coal basins. For example, Eskenazy and Stefanova (2007) observed that chondrite-normalized patterns of REE in subbituminous coals from southwestern Bulgaria were similar to the associated shales and non-coal partings and also to the same patterns in the volcanic rocks in the source area. The authors indicated that the volcanic host rocks, which have been hydrothermally altered, are also a source of other trace elements, including Sb, Mo, U, Th, As, Li and Rb in the coals.

Studies on Late Permian coals in the Southwest Coal Basin of China (Zhang et al., 2002, 2005b; Dai et al., 2008c) indicated that trace elements (e.g. Ti, V, Co, Cr, Co Ni, Cu and Zn) enriched in these coals also have high concentrations in the basaltic rocks within the sediment source region. Dai et al. (2006a) reported high concentrations of Ga, Sr, Zr and Th, as well as abundant boehmite, in coals from the Jungar Coalfield, Northern China, and indicated that these elements were derived from weathered bauxite in the palaeocrust of the sediment source area.

Surface and ground waters Surface and ground waters may carry a range of trace elements in the form of suspended detrital minerals and/or soluble ions. In addition, surface waters may contain high concentrations of trace elements due to leaching of ore deposits or tuff beds (Seredin and

Finkelman, 2008). Ruppert et al. (1996) observed high Ni and Cr concentrations in a Serbian lignite. The Ni and Cr are primarily associated with Ni- and Cr-bearing detrital minerals, which were probably transported into the original swamp by rivers that drained serpentinized ultramafic ores on the rims of the basin.

The geochemical characteristics of ground waters is controlled by their physical and chemical parameters, and also by their interaction with coal (Seredin and Finkelman, 2008). Changes in hydrological conditions may influence trace element contents in coal by changing the supply of ions and fine particles (Taylor et al., 1998).

Circulation of groundwaters can introduce ions carried in brines into coal beds, and also remove elements by leaching processes (Goodarzi, 1987a; Beaton et al., 1991). Chou (1991) states that Cl in coal is in equilibrium with that of the groundwater, and thus was influenced by the fluid circulation in the basin throughout its geologic history. Beaton et al. (1991) noted anomalously high concentrations of Na, Be, B and Ca in a non-marine Paleocene lignite of Canada, and ascribed these concentrations to the leaching from roof and parting materials by groundwater circulation. Eskenazy (2009) indicated that the enrichment of Cl, Br, B and Sr in coal from the Dobrudza Basin of Bulgaria was due to highly mineralized groundwaters.

Another example of groundwater influence on enrichment of trace elements in coal is the leaching of tuffs or tonsteins. Crowley et al. (1989) noted enrichment of Zr, Nb, Th and Ce in coal directly above and below tonsteins in the C coal bed of the Emery Coal Field, Utah. They suggested that the mechanism of enrichment for some elements in the coal was leaching of volcanic ash by groundwater and subsequent incorporation in organic matter or authigenic minerals, or, alternatively, the incorporation of volcanic ash in peat. Hower et al. (1999a) detected highly elevated concentrations of Zr, Y and REE in the coal directly underlying a tonstein in the Fire Clay coal bed, Kentucky. These authors suggested that the tonstein had been leached by groundwater, and that Y and REE enrichments could be attributed to the authigenic minerals (rare-earth phosphate minerals) precipitated from the Y and REE-rich leachate. Similar observations suggesting enrichment of elements due to leaching of volcanic ash beds by groundwater in coals was also made by Wang (2009). Dai et al. (2006a) noted that REE are abundant in coal plies but are distinctly depleted in the adjacent claystone partings, which are of normal sedimentary origin, in a bituminous coal seam of Northern China. Dai et al. (2006a) also ascribed the enrichment of REE to the leaching of claystone partings by groundwater.

Chapter 2 Mineral Matter and Trace Elements in Coal Seams

Groundwater circulation within coal beds may also be responsible for element enrichment near the margins of coal beds (Nicholls, 1968). Enrichment of Ge in the margins of coal beds and thin coal beds has been recorded, at least in some cases, and was thought to be mainly due to the circulation of Ge-rich groundwater (e.g. Ruppert et al., 1996 2012; Hower et al., 2002; Du et al., 2009). In a study of the Amos Coal bed, Western Kentucky, Hower et al. (2002) noted that Ge concentrations are higher in the lower benches than in the top benches, suggesting that there was infiltration of Ge-rich groundwater from the soil rather than from the roof rocks.

Hydrothermal fluids Hydrothermal fluid activity is one of the most important epigenetic processes that can result in significant concentrations of trace elements in coal (Goodarzi and Cameron, 1990; Ren et al., 1999; Dai et al., 2005a). Dai et al. (2012b) classified the hydrothermal fluid activity that has influenced Chinese coals into magmatic, low-temperature hydrothermal and submarine exhalation activity. In many cases, regional faults play an important role in the migration of solutions of different origin, which are in turn carriers of trace elements (Eskenazy, 2009).

Magmatic hydrothermal activity Coals affected by igneous intrusions are not uncommon around the world. Intrusive igneous rocks not only affect the physical and chemical properties of the organic components (e.g. Galushkin, 1997; Meyers and Simoneit, 1999; Stewart et al., 2005; Cooper et al., 2007; Mastalerz et al., 2009), but also the geochemistry of the inorganic constituents in the coal seams (e.g. Ward et al., 1989; Wang et al., 1999a; Dai and Ren, 2007; Zheng et al., 2007).

The changes in inorganic chemistry of the thermally altered coals are due to the thermal decomposition of macerals and minerals, and also to the input of elements from the intrusive rocks (Goodarzi and Cameron, 1990). Finkelman et al. (1998) concluded that three mechanisms can account for the variation in geochemistry of such altered coal: “removal by volatilization; residual concentration in refractory phases; and addition or removal of elements by fluids directly derived from the intrusion or from a hydrothermal system generated at the intrusion-host boundary”.

High concentrations of Ca, Mg, Fe, Mn, and Sr are usually detected in the coked zone (Golab and Carr, 2004), and roughly increase towards the contact in some studies (Merritt,

1990; Finkelman et al., 1998). These elements are typically associated with carbonate minerals, which are often developed in the coal/coke around the intrusive bodies, due to

CO and CO2 released during coking of the coal by magmatic fluids from the intrusion (Kisch and Taylor, 1966; Ward et al., 1989; Querol et al., 1997a; Finkelman et al., 1998).

Finkelman et al. (1998) noted that almost all of the REEs are concentrated in the coked zone of a bituminous coal in Colorado intruded by a felsic porphyry dyke. Studies by Goodarzi and Cameron (1990) on a coal thermally altered by an alkali basaltic dyke showed some highly enriched REE (La, Ce, Nd, Sm, Eu) in the coke at the contact with the dyke, but a general depletion of REE in the rest of the coked material. The high concentration of these elements at the contact was attributed by Goodarzi and Cameron (1990) to input from the intrusive rocks. However, no obvious relationship was observed by Golab and Carr (2004) between the concentrations of most REE and the relative distance from the contacts of two dykes in intruded coals of the Sydney Basin, Australia. The REE in this case were attributed to aluminosilicates in the coals themselves.

Elements such as Cu, Zn and Fe, which have an affinity for sulphide minerals, are sometimes enriched in thermally altered coals (Finkelman et al., 1998). Precipitation of sulphide minerals, primarily pyrite, has been detected in some thermally metamorphosed coals (Karayigit and Whateley, 1997; Querol et al., 1997a). Goodarzi and Cameron (1990), however, noted that some sulphide-associated elements (As, Fe, Mo, S and Sb) are relatively depleted at the intrusive contact, but are concentrated in the coke 25 cm from the contact.

Most of the volatile elements associated with macerals can be easily driven off by magmatic heat, leading to the depletion of those elements in the altered coal (Dai et al., 2012b). For example, Goodarzi and Cameron (1990) noted that the organically-bonded elements Cl, Br, H, N, O and Sorg are depleted in the coke zone due to thermal alteration. Finkelman et al. (1998), however, observed that the volatile elements F and Hg were not depleted in the coal and coke samples of their study that were exposed to the highest temperature. They ascribed the presence of these elements to secondary enrichment following volatilization of elements inherent in the coal.

Epigenetic hydrothermal activity Epigenetic fluids can carry metals and metalloids in solution and from which they are deposited in a coal host; this may result in the enrichment of these elements in coals

Chapter 2 Mineral Matter and Trace Elements in Coal Seams influenced by epigenetic fluids (Martinez-Tarazona et al., 1994). Epigenetic hydrothermal activity may be one of the major factors for the local enrichment of trace elements in some coals around the world (Querol et al., 1992; Spears and Tewalt, 2009).

A number of trace elements, especially environmentally significant elements, are typically associated with epigenetic sulphide minerals in coals. Finkelman (1994) indicated As and Hg in coal are mainly associated with late-stage cleat- and fracture-filling pyrite, and only in some cases are these elements associated with syngenetic pyrite. For example, As is below detection limits in most framboidal and massive iron sulphides (pyrite and marcasite), but may be observed in significant concentrations in radiating iron sulphide forms in some Kentucky coals by EDS or electron microprobe analyses (Hower et al., 1997b; Ruppert et al., 2005). Querol and Chenery (1995), however, noted a greater association of As, as well as other metals such as Co, Cu, Ni and Zn with early-stage sulphides than with late diagenetic sulphides. Examples of the association of trace elements with epigenetic sulphide minerals (e.g. pyrite, marcasite, getchellite) include: As and Hg in the Springfield coal of the Illinois Basin (Minkin et al., 1984; Hower et al., 2005), As and Hg in coal from western Washington (Brownfield et al., 2005), As and Pb in coal from the South Wales coalfield (Gayer et al., 1999), As, Sb, Hg, and Tl in coal from Guizhou, SW China (Dai et al., 2006b). However, not all epigenetic sulphides are necessarily enriched with these elements, and some of these elements may be detected in early-stage pyrite in some cases (e.g. Ruppert et al., 1992).

A range of potentially toxic elements in coals from the Black Warrior Basin of Alabama were examined in a detailed study by Diehl et al. (2004) using LA-ICP-MS and electron microprobe techniques. Diehl et al. (2004) observed that the later formed pyrite (veins and overgrowths on the earlier generations of pyrite) were particularly enriched with Hg, As and Tl, compared with earlier-formed pyrite (framboids and cell-fillings), and attributed the enrichment of Hg, As and Tl to hydrothermal fluid migration probably related to tectonic activity. However, they also noted that Se was associated in the earlier cell-infilling pyrite other than the veined pyrite in some samples. A similar observation of As enrichment in later formed pyrite overgrowths in coal was made by Kolker et al. (2009).

Epigenetic minerals, other than sulphides, may also locally control the trace element contents in coal. For example, Dai et al. (2005a) noted abundant vein ankerite and quartz in an anthracite from the Dafang Coalfield, SW China. Epigenetic chalcopyrite, sphalerite, and Se-galena were also observed in quartz veins. Dai et al. (2005a) indicated that the

39 vein ankerite and epigenetic sulphide minerals were the major sources of Mn, Cu, Ni, Pb, and Zn in the coal.

2.2.2 Modes of occurrence of trace elements in coal

According to Finkelman (1994), the mode of occurrence of a trace element refers to “how the element is chemically bound and physically distributed throughout the coal”. The elements may occur in both the organic and inorganic constituents of coal, but with the proportions in each form varying from element to element (Swaine, 1990). Therefore, elements have association and affinities with different phases in coal (Vassilev and Vassileva, 1997).

2.2.2.1 Determination of elemental modes of occurrence

A variety of direct and indirect methods have been applied to analysis of the forms of trace elements in coal. Indirect methods include gravity separation such as float-sink and sequential leaching tests. The degree of organic affinity of elements is established, for example, by determining the partitioning of elements between sink and float fractions separated in float-sink tests. Sequential leaching tests are used to infer the modes of occurrence of elements in coal, based on their leaching behaviour under different chemical conditions. Statistical analysis is another indirect method. The correlation between the concentration of an element on a whole-coal basis and the ash yield or mineral matter content has been widely used to determine the organic/mineral affinity of the element. Generally, a positive correlation indicates that the element has an inorganic affinity. Elements that are mainly organically bonded in coal would decrease in concentration with increasing ash yield. However, the data from indirect methods may be misinterpreted, due to the unverified inferences and assumptions which are often inherently involved in the method (Huggins, 2002).

The correlations between trace element abundance and the abundance of particular minerals in the coal (e.g. Ward et al., 1999a) provides a better indication of the mode of occurrence for some elements through element-mineral associations than correlation of element abundance to ash yield, or even correlations between the concentrations of particular trace and major elements in the coal.

Direct determination of the forms of trace elements in coal can be achieved using various

Chapter 2 Mineral Matter and Trace Elements in Coal Seams quantitative micro-analytical techniques. The most widely used of these is scanning electron microscopy and energy dispersive spectrometry (SEM-EDS) (e.g. Finkelman and Stanton, 1978). Electron microprobe analysis (EPMA) (Ruppert et al., 1992), transmission electron microscopy and energy dispersive spectrometry (TEM-EDS) (Faraj and Mackinnon, 1993), synchrotron-based X-ray fluorescence (SXRF) (White et al., 1989), scanning proton microprobe (Micro-PIXE) analysis (Hickmott and Baldridge, 1995; Hower et al., 2008), laser-ablation inductively coupled plasma mass-spectrometry (LA ICP-MS) (Querol and Chenery, 1995; Spears et al., 2007), X-ray absorption-fine structure spectroscopy (XAFS) (Huggins et al., 1993; Huggins and Huffman, 1996; Riley et al., 2012), all of which have lower detection limits than SEM-EDS, have also been used in studies on elemental modes of occurrence in coal. However, most of these techniques are not readily available for routine analysis.

2.2.2.2 Modes of occurrence of trace elements

There are extensive studies on the modes of occurrence of trace elements in coal around the world. Finkelman (1994, 1995) reviewed the environmentally-sensitive trace elements individually in terms of their modes of occurrence, and estimated rank of confidence in the different associations. However, as pointed out by Finkelman (1994, 1995), some information regarding the elemental modes of occurrence in the literature is inconsistent or contradictory, not only because of the indirect methods adopted in most literature, but also because of the various and complex occurrences of the same elements in different coals.

Most elements may have a mixed organic and inorganic affinity, or show either organic or inorganic affinity in different coals. Generally, inorganic elements tend to be associated with organic matter in greater proportions in lower-rank coals, although discrete minerals are also important carriers of some elements, even in the lower-rank materials. The organically-associated inorganic elements largely occur as ion-exchangeable forms that are associated with carboxylates or other functional groups and organometallic complexes in the coal organic structure (Benson and Holm, 1985). In high rank coals, most elements tend to have relatively strong inorganic affinity (Kolker and Finkelman, 1998), occurring either in minerals or adsorbed on minerals (Swaine, 1990). Finkelman (1980) suggested that trace elements which have a mineral affinity may be directly attributed to the occurrence of specific trace minerals, rather than to incorporation in major/minor mineral components.

Beryllium Beryllium has been suggested to have a strong organic affinity in coal (e.g. Swaine, 1990; Querol et al., 1992; Finkelman, 1995). Beryllium may also occur in silicates by substitution for Al or Si (Kolker and Finkelman, 1998). It is sometimes distributed in similar proportions in the different fractions from float-sink experiments, which is indicative of a mixed organic/inorganic affinity (Eskenazy and Valceva, 2003). Eskenazy (2006) noted that Be is predominantly organically bound in high-Be coals, but is mainly in inorganic form in coals where the Be concentration approximates Clarke values. However, different observations were made by Dai et al. (Dai et al., 2012c) for high-Be coal from the Wulantuga lignite deposit of China, where the Be is mainly associated with carbonates and clay minerals. No obvious relationship was observed by Eskenazy (2006) between Be and coal rank in a series of Bulgarian coals.

Fluorine Fluorite in coal is mainly associated with clay minerals, and F-bearing minerals such as fluorapatite, fluorite, tourmaline, biotite, and micas (Francis, 1961; Finkelman, 1980; Godbeer and Swaine, 1987; Swaine, 1990), and occasionally with amphiboles (Finkelman, 1980), although it may also have some organic affinity (Bouška et al., 2000). Martinez- Tarazona et al. (1994) observed a strong correlation between F and ash yield in bituminous coals and anthracites from the Asturian Central Basin, Spain, indicating a prevalent mineral infinity. Dai et al. (Dai et al., 2008a) suggest that F may partially occur in boehmite as a substitute for OH- and O2-, rather than kaolinite, and partially in the organic matter in a boehmite-rich coal from the Haerwusu Mine, Inner Mongolia, Northern China. An association of F with gorceixite was indicated by SEM-EDS data and a positive correlation between P and F in coals from the Adaohai Mine, Inner Mongolia, China (Dai et al., 2012d).

Boron Boron in coal is mainly organically bound (Swaine, 1990). This is indicated in many studies by float-sink experiments (Finkelman, 1980; Solari et al., 1989), a negative correlation with the ash yield of the coal (e.g. Goodarzi and Van Der Flier-Keller, 1988) or a postive correlation with organic carbon (Querol et al., 1997a; Querol et al., 1997b). The abundance of B is generally not related to coal rank (Finkelman, 1995). Minor mineral- associated B may also occur in coal. For example, Bohor and Gluskoter (1973) noted that minor boron is associated with illite in some Illinois coals. A high B concentration of a bituminous coal from Guizhou, China, was mainly ascribed to the presence of tourmaline

Chapter 2 Mineral Matter and Trace Elements in Coal Seams

(Zhuang et al., 2000), although tourmaline is unlikely to be an important mode of occurrence for boron in coal (Finkelman, 1995). Eskenazy et al. (1994) noted that organically bound B is the main mode of occurrence for the the low-ash coals and B-rich coals from Bulgaria. More detailed disscussion on the modes of occurrence of B can be found in Boyd (2002).

Phosphorous Phosphorous in coal mainly occurs in apatite and phosphate minerals (e.g. Brown and Swaine, 1964; Finkelman, 1980). Significant concentrations of apatite, as well as aluminophosphates such as goyazite-, gorceixite- or crandallite- group minerals have been described (Ward et al., 1996; Dai et al., 2012a, 2012d), both of which may be responsible for abundant P in coal seams. Other phosphate minerals such as monazite and xenotime also occasionally contributed to the P enrichment (e.g. Finkelman, 1980). Results from some float-sink studies also show the inorganic affinity for P (e.g. Querol et al., 2001; Li et al., 2011). However, P in coal tends to be concentrated in the low-density fractions in some other experiments cited by Finkelman (1995) or show statistically negative correlation with ash yield (e.g. Liu et al., 2001). Nevertheless, direct evidence of the occurrence of organically bound P is absent. This may indicates the occurrence of P- bearing minerals as petrifactions within the maceral components, which contributes to the apparent organic affinity.

Chlorine and bromine Chlorite and Bromine in coal are negatively correlated with ash yield in many high-rank coals (Goodarzi, 1987b; Grieve and Goodarzi, 1993; Liu et al., 2001). Chlorine has been considered to be held in the coal pore waters, indicated by a positive correlation between Cl and moisture contents (e.g. Caswell et al., 1984a, b; Fynes et al., 1988). Chou (1991) further suggested that Cl in Illinois Basin coals occurs in the form of dissolved NaCl in pore water and as Cl- adsorbed on the inner surfaces of micropores in the macerals. XAFS studies also indicated that Cl in some US coals is primarily inorganically bound as Cl- in the moisture associated with the macerals, regardless of coal rank or location (Huggins and Huffman, 1995, 1996). Significantly reduced Cl contents were noted in high- rank coals (e.g. with >86% carbon content), primarily due to the reduced pore surface area in the microstructure (Skipsey, 1974, 1975). As concluded by Yudovich and Ketris (2006b), Cl content tends to decrease in coals with a rank higher than low volatile bituminous A. Ward et al. (1999) also noted Cl in association with the organic matter in

Gunnedah Basin coals, with the Cl apparently being replaced by S (as SO4) due to ion exchange processes.

Fellowes (1979), cited by Spears (2005), noted significant correlations between water- soluble Cl, Br and Na in two coal seam sections from the Yorkshire–Nottinghamshire Coalfield, indicating a close relationship between Cl and Br. Although the relationship between Br and moisture observed by Spears and Zheng (1999) is not as significant as that between Cl and moisture, this may be because of the variable concentrations of Br in pore waters. Huggins and Huffman (1995; 1996) suggested that Br occurs as Br- anions, which is a similar mode of occurrence to Cl, as indicated by their similar k-edge XAFS spectra. Br was also noted to be very low in abundance in high rank coals from major UK coalfields (Spears and Zheng, 1999).

Chromium Chromium in coal has been reported to have mixed affinities (Lyons et al., 1989; Palmer and Lyons, 1990). XAFS spectroscopic studies on bituminous coals revealed the occurrence of Cr as a poorly crystallized chromium oxyhydroxide (CrOOH) mineral trapped within organic components, and as Cr3+ associated with illite (Huggins et al., 1993; Huffman et al., 1994; Huggins and Huffman, 2004). A dominant form of Cr3+ associated with illite and a subordinate chromium oxyhydroxide phase with the organic fraction was also observed for Cr in some Canadian subbituminous and bituminous coals (Goodarzi and Huggins, 2005).

Other inorganic forms of Cr may sometimes be predominant in coals. Brownfield et al. (1995) and Ruppert et al. (1996) described high-Cr coals from Washington, USA and Serbia, respectively, in both of which the Cr is attributed to relatively abundant Cr-bearing detrital mineral grains, especially the spinel group minerals, introduced into the peat mires from ultramafic deposits in the sediment source area.

Vanadium Vanadium and Cr have similar modes of occurrence in coal (Huggins et al., 2009). They may occur with both clays and the organic matter in coals (Finkelman, 1995). Vanadium has a tendency towards an organic association, and may be enriched in vitrinite concentrates (Lyons et al., 1989). XAFS studies of float and sink fractions of Kentucky and Illinois coals, by Huggins and Huffman (1996), indicated that V occurs as both V3+ and

Chapter 2 Mineral Matter and Trace Elements in Coal Seams

4+ V in illite, and is suspected to occur as poorly crystalline oxide (V2O4) or oxyhydroxide

(VO(OH)2) species which is intimately associated with the coal macerals.

Nickel Nickel is commonly associated with iron sulphides, probably substituting for Fe (Huggins and Huffman, 1996). The element is detected in iron sulphides (pyrite and marcasite) at up to 2600 ppm in coals from the UK (White et al., 1989). Kolker (2012) noted that Ni is preferentially concentrated in early-stage framboidal pyrite. Millerite (NiS) occasionally occurs in coal (e.g. Lawrence et al., 1960; van Der Flier-Keller and Fyfe, 1988). Other sulphides that Ni have been reported in are galena, sphalerite, clausthalite (Finkelman, 1995). A statistically significant correlation between Ni and illite, as well as pyrite, was observed by Huggins et al. (2009), who suggested that Ni may also have a lithophilic association.

An organic association of Ni is indicated in some studies, most of which are based on indirect methods. Results from float-sink experiments showed that Ni is enriched in lighter fractions of a Dakota lignite (Miller and Given, 1986). Ruppert et al. (1996), using a supercritical fluid extraction technique, noted minor organically bound Ni (up to 11%) in lignites from the Kosovo Basin, and that the inorganic Ni is mainly associated with spinels.

Copper The mode of occurrence of Cu can be very diverse in coal. As concluded by Finkelman (1980), chalcopyrite is the main carrier of Cu in coal, but organically bound Cu may be important in low rank coals. It has also been suggested that Cu in coal is, at least in part, organically bound (Nicholls, 1968; Swaine, 1990). Indirect methods, including float-sink experiments and leaching extraction, have indicated that Cu is organically bound in some coals (e.g. Miller and Given, 1986; Querol et al., 2001) and has an iron sulphide- association in others (e.g. Querol et al., 1996). Copper is also present in pyrite (e.g. Spears and Martinez-Tarazona, 1993; Kolker, 2012) and carbonates (e.g. Dai et al., 2005a), and occassionally as Cu sulphides and oxides (Finkelman, 1980).

Zinc Zinc occurs dominantly in sphalerite, which is relatively common in coal (e.g. Hatch et al., 1976; Swaine, 1990; Hower et al., 2000; Mastalerz and Drobniak, 2007; Dai et al., 2008a). It is sometimes associated with pyrite (e.g. Finkelman, 1995), clay minerals, specifically illite, and carbonate minerals (Palmer and Lyons, 1996; Spears and Zheng, 1999).

Huggins et al. (2009) noted that, besides sphalerite, which is a main occurrence of Zn, 1/3 of the Zn in an Illinois coal is in a non-sulphide form which, however, remains unidentified. Zinc is also noted to show a partial organic affinity (Palmer and Lyons, 1990).

Gallium Finkelman (1980) concluded that the affinity of Ga shows significant variation. Ga has been suggested to have a mixed organic and inorganic affinity (Nicholls, 1968), or a strong organic affinity (Zubovic, 1966). Gallium is assumed to have a dominant inorganic affinity in most coals, as it commonly shows a positive correlation with ash yield. The element can substitute for Al in Al-bearing minerals, such as clay minerals (Mastalerz and Drobniak, 2012), boehmite (Dai et al., 2006a, 2008a) and diaspore (Dai et al., 2012d). It may also substitute for Zn in sphalerite (Swaine, 1990).

Germanium Germanium is generally concluded to have a predominant organic affinity (Lindahl and Finkelman, 1986). It is usually reported to have an inverse, although not necessarily significant, correlation with ash yield in coal (Spears and Zheng, 1999; Qi et al., 2007a; Dai et al., 2012d). Germanium also tends to be enriched in the light fractions of coals during density fractionation studies (e.g. Miller and Given, 1986; Querol et al., 1995; Zhuang et al., 2003). Inorganic forms of Ge are rarely indicated in the literature, and when present in coal, inorganic Ge is most likely associated with sphalerite or with the clay minerals (Finkelman, 1980). Mastalerz and Drobniak (2012) noted that Ge in some Indiana coals have either clay mineral or pyrite associations, indicated by results from statistical analyses.

Arsenic Arsenic has been suggested to be organically bound in lower-rank coals (Goodarzi, 1987a; Goodarzi and Van Der Flier-Keller, 1988; Swaine, 1990; Kolker et al., 2000). Direct evidence of organic association of As in lower-rank coals is derived from XANES spectroscopic studies of some US and Canadian sub-bituminous coals, where As occurs as As3+, bound to the organic matrix (Huggins et al., 1996). As indicated in many studies of higher-rank coals, The most important carrier of As is pyrite (Ruppert et al., 1992; Hower et al., 1997b; Ward et al., 1999a; Hower et al., 2000; Kolker et al., 2000), and less commonly other sulphides such as marcasite (Querol and Chenery, 1995; Ruppert et al., 2005) and getchellite (Dai et al., 2006b). SEM study of 32 coal samples from Xingren Country, Guizhou Province indicated that organically-associated As is probably the

Chapter 2 Mineral Matter and Trace Elements in Coal Seams dominant form of As in those coals (Ding et al., 2001). However, another SEM study on Xingren coals by Dai et al. (2006b) suggested that the predominant carrier of As, as well as Hg, Sb, and Tl, is epigenetic getchellite. XAFS spectroscopic investigations have indicated that As occurs as a substitute for S in pyrite, rather than as arsenopyrite (FeAsS), which rarely occurs in the US coals (Huggins et al., 1993; Huffman et al., 1994). 3- In addition to As in sulphides, significant amounts of arsenic occurring as arsenate (AsO4 ), formed due to pyrite oxidation, have been observed by XAFS studies in a range of U.S. bituminous coals (e.g. Huggins et al., 1993; Kolker et al., 2000).

Selenium Organically bound Se may be the dominant form in low-S coals (Yudovich and Ketris, 2006a). Selenium in some US coals has been inferred to have a chiefly organic association, based on results from leaching experiments (e.g. Palmer et al., 1989; Dreher and Finkelman, 1992). Selenium may also be in solid solution in pyrite (Diehl et al., 2004; Kolker, 2012) or enriched in pyritic concentrates of coal (e.g. Palmer and Lyons, 1996), or occur as a substitute for S in the pyrite structure (Huggins and Huffman, 1996). Relatively strong correlation of Se with pyrite was observed in some British coals (Spears and Zheng, 1999). The Se-bearing mineral clausthalite (PbSe) was tentatively identified in coal by 2- Hower and Robertson (2003). Selenium may also occur as selenate (SeO4 ) as a result of oxidation (Huggins and Huffman, 1996; Riley et al., 2007; Huggins et al., 2009). Coleman et al. (1993), based on data from almost 9000 US coals, concluded that Se appears to be chiefly associated with the organic fraction, probably substituting for organic sulphur, with other forms of being selenium-bearing pyrite and galena, and lead selenide (clausthalite). Studies of 6 coals from several coal basins in eastern Australia by Riley et al. (2007), indicated that Se is primarily organically bonded in the coal structure in both low-Se and high-Se coals, although Se associated with pyrite and in a selenite oxidation state were also detected in the high-Se coal by XAFS spectroscopy.

Molybdenum The association of Mo with iron sulphides is indicated by a significant positive correlation between Mo and S or Fe in several studies (e.g. Spears et al., 1999b; Spears and Zheng, 1999). However, such significant correlations were not observed in other studies (e.g. Ward et al., 1999a). Mo commonly has a high correlation coefficient with Se in coals (Spears et al., 1999). Association of Mo with pyrite in coal has been observed by micro analytical techniques (e.g. Chou, 1984). LA-ICP-MS studies of coals from the Black

Warrior Basin, USA, by Diehl et al. (2004), indicated that varying but significant concentrations of Mo (up to 582 ppm) are present in pyrite.

Float-sink experiments conducted by Li et al. (2011) showed that the Mo in a subbituminous coal from the Lincang Mine, Yunnan, China, is mainly enriched in the lighter density fractions, while Mo is relatively evenly distributed among the different density fractions of another sub-bituminous coal from the Wulantuga Mine, Inner Mongolia, China. Molybdenum shows a mixed affinity in other float-sink studies (e.g. Querol et al., 1992). Selective leaching studies on five Australian coals indicate that Mo in most the coals is mainly associated with the organic matter, or that it is shielded (encapsulated with organic matter) or present in resistate minerals (Riley et al., 2012).

Antimony Antimony in coal may be associated with both sulphides and the organic fraction (Swaine, 1990; Finkelman, 1995). Pyrite is the most important carrier of Sb in many coals (e.g. Spears et al., 1999b; Spears and Tewalt, 2009), but the concentration of Sb in coal has also been directly found to be associated with other epigenetic sulphides, such as stibnite (Karayigit et al., 2000) and getchellite (Dai et al., 2006b). Results from some float-sink experiments (e.g. Querol et al., 2001; Li et al., 2011) showed that Sb appears to be enriched in the light fractions, but others (Zhuang et al., 2003) showed it to be mainly enriched in the dense fractions. Studies on representative coals from Bulgaria by Eskenazy (1995) indicated that Sb is mostly of organic affinity, and only a small proportion of the Sb is associated with pyrite and clay minerals.

Rare earth elements In coals with normal REE contents, the REE are mainly associated with both detrital minerals and clay minerals (Seredin, 1996), reflected by a generally positive correlation between the concentration of REE and ash yield. In a majority of REE-rich coals, however, the REE are organically bound and/or associated with authigenic minerals (Seredin and Dai, 2012). Float-sink studies of a few lower-rank coals from the Russian Far East by Seredin (1996) also indicated that REE are generally enriched in the light and medium fractions than in the heavy fraction. Seredin and Shpirt (1999) noted a predominant organic form of occurrence of REE in two brown coals from the Russian Far East region, with the REE primarily being enriched in the humic substance isolated from the coals.

Chapter 2 Mineral Matter and Trace Elements in Coal Seams

Common REE bearing minerals in coal are REE phosphates and aluminophosphates, such as monazites (Hower et al., 1999a; Qi et al., 2007b) and crandallites (e.g. Crowley et al., 1993; Ruppert and Moore, 1993; Seredin, 1996). Seredin (1998) described REE hydrous carbonates (kimuraite and lanthanite) and fluorocarbonates (bastnaesite) in basic dikes hosted in a Russian coal seam. Seredin (2004), cited by Seredin and Dai (2012), also described lanthanite and bastnaesite in Russian REE-rich coals. However, such REE carbonates are rarely reported in coal elsewhere. Yttrium, which is closely associated with lanthanide elements in nature, has a similar mode of occurrence to the REE (e.g. Hower et al., 1999a; Seredin and Dai, 2012).

Mercury The majority of the Hg in coal is thought to be associated with sulphide minerals, commonly pyrite, most likely in solid solution (Finkelman, 1994). The association of Hg with pyrite has been noted in many studies (e.g. Ding et al., 2001; Diehl et al., 2004; Kolker et al., 2009). Mercury may also be present in other sulphides, such as marcasite, clausthalite (Hower and Robertson, 2003), getchellite (AsSbS3) (Dai et al., 2006b), cinnabar (HgS), kleinite, tiemannite (HgSe), and rarely as native mercury (Brownfield et al., 2005; Kolker et al., 2009). Results from sequential extraction or density separation studies commonly also confirm the sulphide-Hg association (e.g. Feng and Hong, 1999; Querol et al., 2001).

However, Yudovich and Ketris (2005a) concluded that Hg may have a mixed organic and sulphide affinity in low-S coals. For example, no relationship between Hg and S or pyritic S was observed for low-S Danville coal from Indiana, USA (Mastalerz and Drobniak, 2005), suggesting a possible organic association of Hg. Zheng et al. (2008) reviewed Hg in low-S Chinese coals, and concluded that, besides organic and pyritic forms, Hg may also be associated with silicate minerals. In low-S and lower-rank coals, organic forms of Hg may be dominant. Senior et al. (2000), using sequential leaching methods, found that almost all the Hg in a low-S sub-bituminous coal from the Powder River Basin, USA, was in organic form and/or as shielded minerals.

Lead Lead in coal is mainly associated with sulphides, most commonly iron sulphides (Brown and Swaine, 1964; Finkelman, 1995). The element is concentrated in the sink fractions in most sequential leaching and float-sink experiments (e.g. Miller and Given, 1986; Querol et al., 2001; Li et al., 2011), indicating a dominant inorganic affinity, while Pb appears to

49 have an intermediate affinity in some other studies (e.g. Querol et al., 1992). Statistical analyses also show significant correlation between Pb and pyrite or pyritic sulphur (e.g. Hower et al., 1997b; Wang et al., 2007). Lead may also occur in coal as galena and occasionally as clausthalite (e.g. Hower and Robertson, 2003).

Thorium and Uranium Uranium in coal may be organically bound and may also have an inorganic association (Swaine, 1990). Uranium seems to be abundant in lower-rank coals, and usually has a dominant organic association in these coals (e.g. Querol et al., 1996). However, in other coals, U may show a mainly inorganic affinity (e.g. Querol and Chenery, 1995).

Finkelman (1980) found that most of the Th in two bituminous coals from the Appalachian Basin is inorganically bound, probably in monazite, with minor Th being associated with xenotime, zircon, and perhaps clay minerals. Thorium and U often have a close relationship. Results from sequential leaching experiments conducted by Riley et al. (2012) on some Australian coals indicate that Th and U are present in a resistate Th/U mineral phase in all the test coals, except for one coal where the U is probably mainly associated with the organic fraction.

2.2.3 Geochemistry of intra-seam volcanic claystones

As with a normal kaolinite-rich rock, the major element geochemistry of a typical tonstein is dominated by alumina, and shows a lower abundance of free silica (mainly quartz) and most other major elements (Spears, 2012). The development of a tonstein from the original volcanic ash involves a loss of mobile elements and concentration of immobile elements (e.g. Spears and Lyons, 1995). The abundance of elements in a tonstein is not only a function of the original ash composition, but also the mobility of the elements during the alteration process. In a study of the Felix coal bed of the Powder River Basin, USA, Zielinski (1985) observed substantial depletion of alkali elements (Na, K, Rb and perhaps Cs), and alkaline earth elements (Mg, Ca), Fe and Mn in a tonstein, suggesting that these elements were soluble in the acid coal-forming environment and were likely to be highly mobile. Aluminum, Ti, Ga, Zr and Hf are usually the most immobile elements (Spears and Rice, 1973; Zielinski, 1985). Thorium and U, which are usually retained in tonsteins developed from acidic ash (e.g. Spears, 1970; Spears and Lyons, 1995) and tuff or tonsteins developed from alkali ash (Dai et al., 2010c; Dai et al., 2011) are also often thought to be immobile. Other elements that are generally considered relatively immobile

Chapter 2 Mineral Matter and Trace Elements in Coal Seams in tonsteins include Sc, V, Zr, Hf, Nb, Ta, Co, Cr and REE (Zhou et al., 2000).

Although Spears and Lyons (1995) suggested that, for general application, element immobility should be independent of specific diagenetic conditions, some differences in element mobility have been shown in tonstein/bentonite studies. In the study of Zielinski (1985), Th, Ta, Nb, REE and Y, which are relatively immobile in most low-temperature alteration environments (Wedepohl, 1978), are also depleted, and only Al, Ti, Ga, Zr and Hf, which best resist alteration, are immobile elements. However, REE, Th, Ta and heavy metals show less mobility during bentonite formation (Zielinski, 1982). Crowley et al. (1989), in a study of the C coal bed, Emery Coal Field, Utah, noted that in addition to Nb, Th and Ce, Zr and Hf are also leached from the tonsteins but abundant in the coal plies directly above and below the tonstein beds. Elements may also show different mobilities in tonsteins in low-S and high-S coal sequences (Zielinski, 1985).

The relatively immobile elements are considered to be present in resistate minerals, such as ilmenite (Ti), and zircon (Zr, Hf), and diagenetic minerals such as kaolinite (Al) and anatase (Ti) (Zielinski, 1985). Spears and Rice (1973) suggested that Ga and Th can be accommodated in kaolinite, and U and probably Y in zircon. Zircon may also contain Nb, Ta and Th (Crowley et al., 1989). However, zircon was rarely identified in tonsteins from SW China (Zhou, 1999), where high concentrations of Zr, Hf, Nb and Ta were detected.

2.2.3.1 Parent magma composition

Relatively immobile elements in tonsteins or bentonites can be diagnostic of the original volcanic ash composition. The normalized values of immobile elements, and their ratios, served, for example, as reliable indicators of magma affinities and volcanotectonic settings in bentonite studies (Roberts and Merriman, 1990; Huff et al., 1993).

(1979), tonsteins with TiO2/Al2O3 values of <0.02 and >0.07 are grouped to indicate parent magmas of acid and basic ash composition, respectively; those with values in between are thought to represent intermediate ash material.

A number of plots using additional immobile elements and their ratios have also been used. The geochemical discrimination plot of Winchester and Floyd (1977), using ratios of

Zr/ TiO2 and Nb/Y, has been widely applied in bentonite and tonstein studies.

Zhou et al. (2000) identified silicic and alkali tonstein layers in the Permian coal bearing strata of SW China, based on a combination of normalised immobile elements. Dai et al. (2011) more recently distinguished three types of tonstein bands (silicic, mafic and alkali) in the Songzao Coalfield, SW China using a similar approach. The tonsteins described as alkali tonsteins in the studies of Zhou et al. (2000) and Dai et al. (2011) are characterised by abundant Nb, Ta, Zr, Hf, REE and Ga.

2.2.3.2 Geochemistry of volcanic-influenced coal seams

As with tonsteins, the enrichment of certain elements in the coal seams themselves may also result primarily from airborne volcanic ash deposited in the original peat swamps. In the study of tonsteins in the C coal bed of Utah, Crowley et al. (1989) attributed the enrichment of a group of elements in the coals overlying or underlying the tonsteins, including Zr, Hf, Ga, Nb, Th and REEs, which are relatively immobile in other studies (e.g. Zielinski, 1985), to three possible processes: (1) leaching of volcanic ash by groundwater and subsequent uptake by organic matter, (2) leaching of volcanic ash by groundwater and subsequent incorporation in minerals, and (3) incorporation of volcanic ash minerals directly into the peat bed.

Similar mechanisms were used by Hower et al. (1999a) to explain elevated Zr, Y and REE in the coal directly underlying a tonstein in the Fire Clay coal bed of Kentucky, where Zr may be incorporated in authigenic REE-phosphates. More abundant volcanogenic minerals, derived from the alteration of volcanic ash in coal, can result in the enrichment of certain elements. For example, high concentrations of Ba, F, P, Sr, and Zr noted in coals from the Green River district, Western Washington, USA, were mainly ascribed to the presence of volcanogenic minerals such as apatite, feldspar, zircon and crandallite group minerals (Brownfield et al., 2005).

The incorporation of volcanic ash in coal may be reflected by the geochemistry of the coal, sometimes even without the formation of visible tonstein layers or volcanogenic minerals. Dewison (1989) observed strong correlations between kaolinite and relatively immobile elements such as Zr, Y and Nb. This suggested a genetic relationship between kaolinite and those elements: kaolinite was thought to have been recrystallised from dispersed

Chapter 2 Mineral Matter and Trace Elements in Coal Seams volcanic ash falls and the elements were liberated by the same leaching process. Dai et al. (2007b) suggested that alkaline volcanic ash was responsible for the enrichment of alkaline elements such as Nb, Zr, Ga, Hf and REE in an anthracite from the Songzao coalfield, SW China, where only clayey micro-sized bands were observed. However, minerals (e.g. zircon) mainly made-up of those elements were rarely found in the coal (Dai et al., 2007b).

The original magma composition can be directly determined from chemical analyses of glass inclusions in the volcanic minerals of tonsteins. Microprobe analyses of glass inclusions in volcanic quartz of the Fire Clay tonstein in North America and tonsteins from UK and Germany, except one English tonstein, all show the composition of a high-silica rhyolite (Lyons et al., 1992; Lyons et al., 1994). Such results are broadly consistent with interpretations derived from immobile element geochemistry (Lyons et al., 1992; Lyons et al., 1994).

Successful correlations based on the immobile elements in the whole rock of Ordovician K-bentonites in North America have been demonstrated in a number of studies (e.g. Kolata et al., 1987; Huff and Kolata, 1990). A larger scale correlation was further proposed by Huff et al. (1992), who used a similar approach in correlation between eastern North America and Europe.

Like bentonites/K-bentonites, tonsteins can also be used as an aid to regional correlations in coalfields on the basis of their chemical composition (e.g. Hill, 1988; Zhou et al., 2000), especially when there is a lack of other persistent diagnostic indicators over large areas. For example, whole rock chemical fingerprinting of tuffs and tonsteins, usually represented by the concentrations of relatively immobile elements and their ratios, have been successfully used in several studies on correlations in the coalfields of the Sydney Basin, Australia (Kramer et al., 2001; Grevenitz, 2003).

Analyses of glass inclusions in quartz in tonsteins and bentonites, which are compositionally different, may also be useful for intra- and inter-basinal correlations (e.g. Delano et al., 1994; Lyons et al., 1994). The composition of phenocrysts of bentonites/K- bentonites (e.g. Batchelor and Jeppsson, 1994; Haynes et al., 1995) has also shown potential for use in stratigraphic correlations.

2.2.4 Summary

Trace elements in coal are generally regarded as those elements with concentrations up to around 1000 ppm, although some of trace elements may be present in some coals in higher concentrations. Although particular trace elements (e.g. As, Se, Hg) in the coal may represent a potential environmental hazard, some of them may also represent a significant source of material for industrial purposes. Geological factors, such as paleoenvironment and nature of source rock, surface and ground waters, and hydrothermal fluids, play the key role in more general enrichment of trace elements.

Despite of extensive studies on the modes of occurrence of trace elements in coal around the world, the elemental modes of occurrence indirectly inferred or directly determined in the literature may be inconsistent or contradictory, not only because of the indirect methods adopted in most literature, but also because of the various and complex occurrences of the same elements in different coals.

Relatively immobile elements in intra-seam claystones can be diagnostic of the original volcanic ash composition. The normalized values of immobile elements, and their ratios, serve as reliable indicators of magma affinities and volcanotectonic settings in bentonite studies. Leaching of volcanic ash layers or the incorporation of volcanic ash within coal may have an influence on the geochemistry of the coal beds.

CHAPTER 3 GEOLOGICAL BACKGROUND OF THE STUDY AREA

3.1 Geological setting of the Sydney Basin

The Sydney Basin is part of the Sydney-Gunnedah-Bowen Basin system, a larger structural complex in eastern Australia that extends from coastal southern New South Wales to Central Queensland (Figure 3.1). The Sydney Basin is confined to the east by the New England Fold Belt and to the west by the Lachlan Fold Belt, and extends offshore to the continental shelf (Tadros, 1995) (Figures 3.1, 3.2). It is approximately 350 km long by an average 100 km wide, covering a total area of approximately 52 km2 on shore and 15 km2 offshore (Stewart and Alder, 1995; Alder et al., 1998).

Figure 3.1 Location of the Sydney-Gunnedah-Bowen Basin system of eastern Australia (After Tadros, 1995).

3.1.1 Tectonic framework of the Sydney-Gunnedah-Bowen Basin

The Sydney-Gunnedah-Bowen Basin has a complex tectonic history, with different mechanisms operating during successive stages of basin development. Various tectonic models have been proposed for the tectonic evolution of the Sydney-Gunnedah- Bowen Basin, or to particular time intervals of basin evolution. These models have been summarized by Murray (1990) and Scheibner and Basden (1998).

A thermal theory was proposed by Brownlow (1981) for the evolution of the Sydney Basin, which involves mantle diapirs emplacement into the crust in the Late Carboniferous, and subsidence of its eastern part in the Early Permian accompanied by igneous activity. Murray (1990), among others, suggested an early Permian volcanic rift stage through a mid Permian sag phase to a final late Permian-Middle Triassic foreland basin stage. Scheibner and Basden (1998), based partly on the models of Murray (1990) and others, suggested the following model for the tectonic evolution of the Sydney-Gunnedah- Bowen Basin: a. The basin commenced as a tensional volcanic rift between the Lachlan and New England Fold Belts during the latest Carboniferous to the early Permian. It was transitional to a volcanic arc in the north (the Bowen Basin). b. The rift system was followed by a thermal collapse and thinning phase. Collapse and thinning of the crust occurred during the extension, most probably with a combination of simple and pure shear. The Meandarra Gravity Ridge reflects emplacement of relatively dense igneous rocks at various levels of the crust. c. Following this, foreland loading by overthrusting of the New England Fold Belt after the rifting and thinning phase occurred. The basin became a foreland basin of the New England Fold Belt from the mid Permian onwards. A retro-arc foreland basin setting might have developed in Queensland.

3.1.2 Basin structure

The structural boundaries and major structural trends within the Sydney Basin are shown in Figure 3.3. The outline of the Sydney Basin is defined by a number of tectonic lineaments (Stewart and Alder, 1995). The Sydney Basin was originally proposed to be

Chapter 3 Geological Background of the Study Area separated structurally from the Gunnedah-Bowen Basin by the Mount Coricudgy Anticline (Bembrick et al., 1973). However, West and Bradley (1986), based on the stratigraphy of Permian coal measures and geophysical studies, suggested that the Liverpool Range was a more appropriate boundary. To the northeast, the Sydney Basin is bounded by the Hunter-Mooki Thrust system, a major structural lineament probably formed during the Hunter-Bowen Orogeny from the Middle Permian to Middle Triassic (Shi and McLoughlin, 1997). The Sydney Basin extends offshore (eastwards) to the edge of the continental shelf, which is a rifted boundary developed during opening of the Tasman Sea in the Late Cretaceous. The western boundary is marked by the uncomfortable onlap of Permian basin sediments onto older Lachlan Fold Belt sequences (Stewart and Alder, 1995).

3.1.3 Regional stratigraphy and depositional environment

The sediments in the Sydney Basin are represented in outcrop by strata ranging from Permian to Triassic in age. The general stratigraphy of the Sydney Basin is shown in Figure 3.2. The stratigraphy and depositional development of the Sydney Basin have been documented by Herbert (1980), Lohe (1992a) and Tadros (1995), among others, and much of the material included here has been summarized from their work.

Early Permian The earliest-formed sediments in the Sydney Basin consist of isolated fluvial, coastal plain and marine sediments. These are represented by the Talaterang Group in the south, and the fluvio-lacustrine Seaham Formation in the Hunter and Newcastle regions. Basin subsidence caused by a major tectonic movement then created accommodation for predominantly marine sediments, represented by the Shoalhaven Group in the south and west, and the Dalwood Group in the north. The Dalwood Group in the Hunter region mostly consists of marine sediments and volcanics.

The Greta Coal Measures were deposited in response to a temporary marine regression, prograding southwards from the tectonically active New England area, and wedging out further south within the basin. The marine conditions continued uninterrupted in the south, with deposition of the Snapper Point Formation and Wandrawandian Siltstone. Deposition in the western part of the basin commenced at this stage, also forming the Snapper Point Formation.

Figure 3.2 General stratigraphy of the Sydney Basin (modified from Tadros, 1995, and Stewart and Alder, 1995).

Mid to late Early Permian Renewed thermal subsidence led to a transgression inundating the Greta Coal Measures, and the deposition of the Branxton Formation of the Maitland Group in the north. The marine incursion was interrupted by regressive/transgressive sand sheets, represented by the Muree Sandstone in the north and Nowra Sandstone in the south. These beds were overlain by the Mulbring Siltstone in the north and Berry Siltstone in the south and west, all formed under marine conditions.

Marine sedimentation terminated in the Late Permian following Mid Permian activation of the Hunter-Mooki Thrust system and uplift of the New England area. Terrestrial sedimentation then prevailed across the basin, with a vast wedge of terrestrial lithic clastics and coals.

Chapter 3 Geological Background of the Study Area

Late Permian Rapid subsidence, associated with continuing uplift and volcanism in the New England region, led to the formation of the Tomago Coal Measures in the Newcastle area and its equivalent Wittingham Coal Measures in the Hunter area. The terrestrial deposition was interrupted briefly by two marine incursions. The first incursion resulted in the formation of the Kulnura Marine Tougue and its equivalent, the Bulga and Archerfield Formations in the north and the Erins Vale Formation in the south. Another marine incursion was also short-lived, but inundated the basin, and resulted in the deposition of the Waratah Sandstone/Dempsey Formation, and its equivalent the Watts Sandstone/Denman Formation sequences in the north, the Baal Bone Formation in the west, and the Bargo Claystone in the south.

Regression conditions were resumed after the short-lived incursions, with southward progradation of the coal measures. This regression resulted in the deposition of the Newcastle Coal Measures and the Wollombi Coal Measures in the north, and the upper part of the Illawarra Coal Measures in the west and south. In the Late Permian, coal measure sedimentation was terminated by the renewed uplift in the New England Fold Belt, and the activation of the Hunter-Mooki Thrust system. An unconformity is present over the Lochinvar Anticline, where the Newcastle-Tomago Coal Measures have been completely eroded.

Triassic The Narrabeen Group was deposited on the eroded surfaces of either the coal measures or the underlying Permian strata, with the sediments largely derived from the north or west. The Narrabeen Group is in turn overlain by the Hawkesbury Sandstone, with the sediment being from the southwest or west (Ward, 1972). This change in sediment transport direction was due to the uplift of the Lachlan Fold Belt, and erosion of the tilted Late Permian and Early Triassic sediments in the south. The Middle Triassic Wianamatta Group marks the uppermost stratigraphic unit in the Sydney Basin, but occurs only in a restricted area.

3.1.4 Structural geology

On a regional scale, post-depositional structural deformation in the Sydney Basin is generally not intense, and the basinal sediments usually have dips of 10°or less (Lohe, 1992a). The Sydney Basin is marked by a relative stable platform facies along its western

59 and southern boundaries, where it onlaps basement (Figure 3.3). The basin deepens significantly to the east and north from these stable areas, across a series of hinge lines and monoclines. The main structural features of the Sydney Basin are broad folds and monoclinal features associated with several periods of compressional tectonics and extensional tectonics (Stewart and Alder, 1995). Major structural trends, including gentle folds, monoclines, and basin hinge lines, have north-south to north-northeast trending fold axes, and appear to be sub-parallel to the edge of the continental shelf.

Figure 3.3 Major structures within the Sydney Basin (after Stewart and Alder, 1995).

Chapter 3 Geological Background of the Study Area

Normal faulting is common in the Sydney Basin. Reverse faults also occur, but are relatively minor, and not penetrative on a regional scale. The major regional thrust structure is the Hunter-Mooki Thrust system that defines the northern margin of the Sydney Basin. It has a broad northwest trend between the Lochinvar and Muswellbrook Anticlines, but turns to a more meridional orientation to the east and north of Muswellbrook. A series of low amplitude folds near the leading edge of the Hunter Thrust sheet developed in response to ramping of the Hunter Thrust (Glen and Beckett, 1989). Strike-slip faults, which are located in several areas of the basin, are notable in the Southern Coalfield (Lohe, 1992a).

3.1.5 Coal-bearing sequences and coal seams in the present study

Greta Coal Measures and the Greta seam The Greta Coal Measures were formed during a short-lived marine regression, developed to the south of the rising New England Fold Belt, and lie conformably between the marine Dalwood and Maitland groups (Agnew et al., 1995) (Figure 3.2). The coal measures crop- out along the flanks of the Lochinvar Anticline and in the Cranky Corner Basin, which is a small outlier basin a short distance to the north, separated from the Sydney Basin by the Hunter Thrust fault system. The coals of the Greta Coal Measures are typified by complex splitting (Hutton, 2009). Worked seams included the Greta and Homeville seams, together with their various splits.

The Greta seam is the upper coal unit of the Greta Coal Measures (Figure 3.4), and is separated from the underlying Homeville seam by coarse clastics (the Kurri Kurri Conglomerate). The Greta seam is split by the Kearsley Lens, an eastward-thickening mudstone to fine-grained sandstone, which probably represents the edge of contemporaneous clastic sedimentation. Thin tuffaceous claystone bands are also widespread within the seam (Agnew et al., 1995). The Greta coal in the Lochinvar Anticline area is typically a banded bright and dull coal with an overall high sulphur content, especially in the upper section.

Figure 3.4 Stratigraphy of the Greta Coal Measures in the Lochinvar Anticline area (after Van Heeswijck, 2001).

Newcastle Coal Measures and the Great Northern seam The Late Permian Newcastle Coal Measures is one of several stratigraphic units in the northeastern Sydney Basin (Figure 3.5) from which significant resources of bituminous coals are currently extracted (Figure 3.2). The sequence is dominantly fluvial, with a maximum thickness of 450 m (Hutton, 2009). As discussed more fully by Agnew et al. (1995), the uppermost part of the sequence, the Moon Island Beach Subgroup (Figure 3.5), includes a number of coarse pebble conglomerates, with associated sandstones and shales, derived from erosion of older strata immediately to the north, in what was then the tectonically active orogen of the New England Fold Belt (Figure 3.1), and deposited by alluvial fan and braided river systems in a foreland basin setting (Branagan and Johnson, 1970; Herbert, 1980; Bocking et al., 1988; Agnew et al., 1995; Diessel and Hutton, 2004). However, the upper part of the Newcastle Coal Measures (including the Moon Island Beach Sub-group) also contains a number of tuffaceous units (Figure 3.5), representing products of contemporaneous volcanism associated with the New England Orogen, transported into the basin by ash fall and ash flow processes (Diessel, 1965; Loughnan and Ray, 1978; Diessel, 1983; Agnew et al., 1995; Kramer et al., 2001; Grevenitz, 2003).

Chapter 3 Geological Background of the Study Area

Figure 3.5 Stratigraphic units in the Newcastle Coalfield of the Sydney Basin (left) with additional detail on units in the Moon Island Beach subgroup (right) (after Agnew et al. 1995).

The Great Northern seam is one of the principal economic coal seams in the Moon Island Beach Subgroup. It is exposed in coastal outcrops at Catherine Hill Bay, 30 km SSW of Newcastle, and is mined from a number of underground collieries in the area to the north and west. The seam is underlain by the Awaba Tuff, an extensive tuffaceous unit which commonly shows cross-stratification (Kramer et al., 2001) and may in part represent ash flow or reworked ash fall material. It is partly interbedded with alluvial channel deposits, and overlain mainly by the Teralba Conglomerate, a sequence of sandstone and conglomerate also formed by fluvial channel processes (Agnew et al., 1995). Another volcanic unit, the Booragul Tuff, also overlies the Great Northern seam in the eastern part of the coalfield, occurring in places between the coal and the Teralba Conglomerate interval.

Illawarra Coal Measures and the Bulli seam The Late Permian Illawarra Coal Measures contain all the economic coals seams of the Southern Coalfield. The coal measures are conformably underlain by the marine

Shoalhaven Group, and erosionally overlain by the fluvial Narrabeen Group of Triassic age (Figure 3.2). The coal measures have an average thickness of 200 m, and reach a maximum thickness of over 400 m in the northeastern part of the coalfield (Armstrong et al., 1995). The Illawarra Coal Measures are separated by an erosional surface into two subgroups. The lower unit, the Cumberland Subgroup, comprises lithic sandstone and latite formations, and minor thin coal seams. The upper unit, the Sydney Subgroup, includes sandstone, siltstone, shale, conglomerate and claystone (tuffaceous) units, and a number of economic coal seams (Figure 3.6).

Figure 3.6 Stratigraphy of the Illawarra Coal Measures in the Southern Coalfield (After Hutton, 2009).

The Bulli seam is stratigraphically the uppermost seam in the Sydney Subgroup of the Illawarra Coal Measures. It rests on a thick sandstone (the Loddon Sandstone). The base of the Bulli seam is marked by a carbonaceous claystone containing abundant root traces (Armstrong et al., 1995). The Bulli seam is typically 2 to 3 m thick, with a maximum of 4 m near Camden, and consists of interbanded bright and dull coal, with minor carbonaceous mudstone interbeds and sideritic penny bands (Armstrong et al., 1995). It can be traced throughout much of the Southern Coalfield, and thins-out only in the most southerly

Chapter 3 Geological Background of the Study Area regions, where the Wongawilli Coal is the uppermost coal in the Illawarra Coal Measures (Lohe, 1992b). This may be due to the erosion of the top section of the Illawarra Coal Measures during the formation of the overlying Narrabeen Group (Bowman, 1980). In some areas of the Southern Coalfield (e.g. south of Bargo), the Bulli seam merges with the underlying Balgownie seam where the Loddon Sandstone thins-out.

3.1.6 Tuffs and tonsteins in the coal measures of the Sydney Basin

The Permian coal-bearing sequences of the Sydney Basin contain numerous volcanic ash deposits, including many inter-seam tuffs, and intra-seam tonstein or bentonite bands in the coal seams. The Newcastle Coal Measures contains relatively abundant pyroclastic sediments, with 20% of the sequence consisting of rhyodacitic tuff and tuff-derived material (Diessel, 1983). These tuffs and tonsteins have been correlated with those in the Wollombi Coal Measures in the Hunter Coalfield (Kramer et al., 2001), and the Illawarra Coal Measures in the Southern Coalfield (Grevenitz, 2003).

Two possible sources for the tuffs in the Sydney Basin have been suggested: a volcanic source located several tens of kilometres to the east and northeast of the present coastline (e.g. Diessel, 1983), and an offshore uplift to the east of the Sydney Basin (e.g. Herbert, 1994; Kramer et al., 2001).

3.1.7 Summary

The Sydney Basin evolved from a rift basin to a foreland basin in the Early Permian. The basin has generally been affected by only mild deformation, and the sediments usually have dips of 10°or less due to local structures. Two main coal-bearing intervals were formed during the Permian in the Sydney Basin. In the early Permian the Greta Coal Measures was deposited in the north. In the Late Permian more widespread coal measures were deposited across the entire Sydney Basin; these are the Tomago and Newcastle Coal Measures, and their equivalent Wittingham and Wollombi Coal Measures in the north, and the Illawarra Coal Measures in the south and west.

The composition of many coal deposits in the Sydney Basin has been influenced by penecontemporaneous volcanism, with tuffs and tonsteins being widespread in the Permian coal-bearing sequences of the Sydney Basin. The coal seams of the Permian coal measures were thus deposited under conditions that were influenced by two different

65 types of non-coal sediment input, namely pyroclastic material and more normal epiclastic material, with the latter being mainly from a sediment source area in the uplifted New England Fold Belt.

The Early Permian Greta seam and the Late Permian Bulli and Great Northern seams are some of the most economically important coal seams in the Sydney Basin. The mineralogical and geochemical characteristics of these coals and the associated strata were affected by both regional tectonics and the coal-forming environment, and have been evaluated as part of the present study.

3.2 Geological background of the Songzao Coalfield

Late Permian coal-bearing sequences are well-developed in SW China. These coal deposits cover western Guizhou, eastern Yunan, southern and eastern Sichuan provinces and the city of Chongqing (Figure 3.7).

3.2.1 Broad-scale tectonics

These Late Permian coal basins are located in the western part of the Yangtze Block, a relatively stable tectonic unit in south China (Figure 3.8). The more extensive South China Block (Figure 3.8A) is a composite continental block which is tectonically divided into two major blocks, the Yangtze Block and the adjacent Cathaysia Block to the southeast (Figure 3.8B) These were two separate landmasses, and amalgamated into the South China Block through collision and related orogenesis (the Jiangnan orogeny) during Neo- proterozoic time. The South China Block is separated from the North China Block to the north by the Qinling-Dabie orogen (Li et al., 2007a; Duan et al., 2011).

Chapter 3 Geological Background of the Study Area

Figure 3.7 Location of coal-producing provinces in SW China.

As discussed by Wang et al. (1989) and Wang (1996), the main tectonic evolution in SW China involved three stages:

(a) During the late stage of the Early Palaeozoic (Caledonian), regional elevation and subsidence resulted in the development of large-scale uplift and depressional structures in the western Yangtze Block.

(b) Tensional fault-depression has taken place since the Devonian. Basaltic eruption started to occur on the sea-floor, and became more voluminous in the middle Carboniferous. The pull-apart movement reached its climax in the Permian, with basaltic eruption that overflowed to the western margin of the Yangtze Block. The western Yangtze Block has evolved into a fault subsidence basin (Figure 3.9).

(c) During the Mesozoic and Cenozoic, the western Yangtze Block was intensively deformed by compression-shear activity (Yenshan and Himalayan movements), which substantially reformed the early tectonic expressions. The western Yangtze Block has become a continental basin since the Triassic. The Sichuan Basin, situated in the NW part of the Yangtze Block, has been developed into a compressional tectonic basin.

Different models have been proposed for the tectonic movement, accompanied by voluminous basaltic eruptions (Emeishan basalts), in the western Yangtze Block during the Mid Palaeozoic to Triassic. Rift-based (without plume) models were earlier proposed

67 by Luo (1981) and Luo et al. (1990). They suggested that taphrogenesis (intracontinental rifting) began in Devonian times, climaxed with the Late Permian basalt eruption, and ended in mid Triassic times.

Figure 3.8 (A) Sketch map showing the main tectonic units of China (NCB=North China Block; SCB=South China Block); (B) Map showing the South China Block, consisting of the Yangtze and Cathaysia Blocks, separated by the Jiangnan Orogen. After Duan et al. (2011).

Figure 3.9 Tectonics of the western Yangtze Block in the Hercynian period (€=Cambrian, O=Ordovician, S=Silurian; D=Devonian) (after Wang, 1996).

Chapter 3 Geological Background of the Study Area

Mantle plume models were more recently proposed by Chung and Jahn (1995), based on geochemical studies of the Emeishan basalts. Such models have been further supported by the sedimentary evidence for a rapid crustal doming prior to the Emeishan flood volcanism (He et al., 2003), and other evidence based on the spatial distribution of different rock types and on geophysical data (e.g. Xiao et al., 2004; Xu et al., 2004; Xu et al., 2007).

3.2.2 Stratigraphy and depositional environment in SW China

The sediments in SW China are represented by strata mainly ranging from Permian to Triassic in age. During the Caledonian movement in the middle or late Palaeozoic, the western Yangtze Platform was gradually rising from west to east. It later underwent denudation, which resulted in the erosion of the Silurian and Devonian strata in most areas. The general stratigraphy of SW China is shown in Figure 3.10.

Early Permian The western Yangtze Block had undergone a long period of erosion until the Early Permian, when marine water transgressed from the southwest of the block and submerged the whole of the South China Block. This extensive transgression resulted in the development of thick carbonate sequences (Qixia and the overlying Maokou Formation) in SW China.

Figure 3.10 Lithostratigraphic correlations in Guizhou and Yunnan Provinces, SW China (after Wang et al., 2011).

Late Permian The Late Permian coal measures in SW China were deposited within a range of facies associations, spanning from fluvial to marine carbonate platform deposits (Wang et al., 2011) (Figure 3.10).

Thick basaltic sequences, the so-called Emeishan Basalts, unconformably rest on the Maokou Formation, and cover most areas to the east of the Chuandian Oldland, including eastern Yunnan, western Guizhou, and southern Sichuan (He et al., 2003). There was a regression, related to the so-called Dongwu Movement, a regional uplift event, which exposed the Yangtze Block and resulted in a depositional hiatus. Extensive transgression, probably during or after the main eruption phase of the Emeishan basalts in the Early Permian (He et al., 2003), resulted in the development of thick carbonate and siliciclastic formations, the Xuanwei and Longtan Formations in eastern Yunnan and western Guizhou, respectively. These strata unconformably overlie the Emeishan Basalts, except in localised areas of eastern Chongqing and western Guizhou, where the Longtan Formation directly rests on the erosional surface of the Maokou Formation.

The Xuanwei Formation in eastern Yunnan and southern Sichuan comprises terrestrial siliciclastic rocks, including interbedded sandstone, siltstone, mudstone, and coal. It is dominated by conglomerates in the basal part, which rest unconformably on the Emeishan Basalt in some areas of eastern Yunnan (Wang et al., 2011).

The Longtan Formation in western Guizhou and southern Sichuan consists of paralic siliciclastic rocks intercalated with limestones and coal seams. In this stage of geological development, the marine conditions continued uninterrupted in eastern and southeastern Guizhou, the eastern part of southern Sichuan and Chongqing, and the southern part of eastern Yunnan. The Wujiapin Formation was formed in these areas, and consists of carbonates and siliciclastic rocks (Wang et al., 2011).

The Changxing Formation, which was formed by the continuing transgression, conformably overlies the Longtan and Wujiaping Formations. The Changxing Formation is a predominant carbonate unit widespread in the eastern part of SW China; in eastern Yunnan and southern Sichuan, however, the upper Xuanwei Formation, which consists of terrestrial siliciclastic rocks, was formed. Between these two areas, in western Guizhou and southern Sichuan, the strata consist of paralic siliciclastic rocks intercalated with

Chapter 3 Geological Background of the Study Area limestones and coal seams. The Changxing Formation is in turn overlain by the Triassic sediments (Feixianguan Formation).

Figure 3.11 Late Permian Longtan-stage palaeogeographic map of Southwest China (modified from Wang et al., 1995), showing the location of the Songzao Coalfield.

3.2.3 Local geology of the Songzao Coalfield

The Songzao Coalfield is located in Qijiang County, SW Chongqing (Figure 3.11). It is 39.5 km long (S-N) and 1.1-11km (E-W) wide, with a total area of 140.8 km2. It includes eight mines, the Datong, Yuyang, Shihao, Songzao, Tonghua, Fengchun, Zhangshiba and Liyuanba mines (Figure 3.12). The coals from the Songzao Coalfield are all high-sulfur anthracites and rich in methane. The coalfield is located on the northwestern flanks of the Jiudianya, Jiulongshan and Sangmuchang Anticlines. The main structures within the Songzao Coalfield are four folds (Figure 3.12). The Yangchatan Anticline, which is a major fold, is an asymmetric anticline, with the eastern limb dipping at 30-40° and the western limb dipping at <10°. Faults are minor, and only those distributed in the main folds affect mining conditions.

Figure 3.12 Main structures in the Songzao Coalfield, with indication of mining areas (modified from Songzao Coalfield report).

The coal-bearing sequence in the coalfield is the Late Permian Longtan Formation, consisting of limestone, sandstone, silty mudstone, mudstone, coal seams and tuffaceous sediments (Figure 3.13). It was deposited in a tidal flat system along the western margin of the epicontinental sea basin (Dai et al., 2010). The Kangdian Oldland in the west was the major sediment source for the coalfield (Figure 3.9). The coal-bearing sequence has an average thickness of 71.8 m. The strata contain 6-11 coal seams, among which the No. 8 coal is workable through the entire coalfield, and the Nos. 6, 7, 8, 11, and 12 are locally workable. A mafic tuff layer with a thickness of 0.2-21m, mostly 3-5m, light-grey or light grey-white in colour, lies unconformably between the Longtan and Maokou Formations in the Songzao Coalfield (Dai et al., 2010a). This mafic tuff is commonly locally described as “bauxite” in field lithologic descriptions. The Longtan Formation is disconformably underlain by the Maokou Formation, which is a limestone unit of Early Permian age.

Intra-seam claystones (tonsteins) are very common within the coal seams of the Songzao Coalfield (Dai et al., 2011), as well as other coalfields in SW China (e.g. Zhou et al., 2000). They are thought to represent air-borne contemporaneous volcanic ash, which was

Chapter 3 Geological Background of the Study Area deposited into the peat swamps. According to Dai et al. (2011), the tonsteins and tuffs in the Songzao Coalfield probably resulted from waning activity of the mantle plume. Geochemical data indicate that the tonsteins were probably derived from different mantle sources that underwent not only low-degree partial melting, but also fluid fractionation and contamination by lithospheric mantle material. Their source magmas had an alkali-basalt composition, and were similar to that of ocean island basalts.

Figure 3.13 Sedimentary sequences of the Songzao Coalfield, showing the location of the coal seams (modified from Songzao Coalfield report).

3.2.4 Summary

The Yunnan, Sichuan and Guizhou provinces of SW China contain economically important coal resources. These coal basins situated in the western Yangtze Block, developed on a basement that was crystallized and metamorphosed in the Proterozoic. The Songzao Coalfield is located in SW Chongqing. The coal-bearing sequence in the coalfield is the Late Permian Longtan Formation, which was developed in a tidal flat system along the western margin of the epicontinental sea basin.

Intra-seam tonsteins are common in the coal-bearing sequences of the Songzao Coalfield. The tonsteins probably resulted from waning activity of the mantle plume, with the source magmas having an alkali-basalt composition. The geochemistry and mineralogy of the Songzao coal and non-coal strata, including the tonsteins, were apparently affected by the regional geological factors of the Songzao Coalfield. This will be discussed further in the present study.

CHAPTER 4 SAMPLING AND ANALYTICAL TECHNIQUES

4.1 Sampling

A total of 106 coal and non-coal samples were collected from the Sydney Basin and the Songzao Coalfield (Table 4.1), as the basis for a comparative study of the minerals and trace elements in these two Permian coal basins. The individual samples represent each separately identifiable layer (bench or ply) of coal or non-coal materials within the seam at the location studied, and were taken to provide a mineralogical and geochemical profile through the relevant seam section.

Table 4.1 Samples for the present study Number of Coal seam Location Sample source samples Sydney Basin Great Northern seam Newvale No.1 Colliery (1962-1994) Seam section 14 Great Northern seam Catherine Hill Bay Outcrop 5 Bulli seam Coal Cliff Colliery (1878-1991) Seam section 9 Greta seam Cessnock No. 1 Colliery (1921-1961) Seam section 9 Greta seam Austar coal mine Borehole 45 Songzao Coalfield No. 7 seam Datong Mine Seam section 7

K2b seam Tonghua Mine Seam section 8 No.11 seam Yuyang Mine Seam section 9

4.1.1 Samples from the Sydney Basin, Australia

Greta seam A series of nine coal and non-coal samples from the Greta seam at the Cessnock No. 1 Colliery, located approximately 9.6 km SW of Cessnock, originally collected in the 1950s, were also supplied from the sample bank of CSIRO Energy Technology. Cessnock No.1 Colliery, previously known as Kalingo Colliery, was closed in 1961. It was amalgamated into Pelton Colliery in the late 1960s, which was in turn amalgamated into Southland Colliery (now Austar Mine) in 1998. The coals, however, were sampled in 1953 and have been oxidized during the long-term storage. This is indicated by the relatively abundant iron sulphate minerals with minor pyrite remaining in the coals, which will be discussed later in the present study.

A complete drill core through the Greta seam was provided by the Austar coal mine, from a site approximately 10 km SW of Cessnock in the Lochinvar Anticline area of the Newcastle Coalfield (Figure 4.1). As indicated above, the Austar mine was previously known as Southland Colliery, and was formed from an amalgamation of several older collieries. The core was drilled in December 2009, and donated to the project in May 2010. The seam in the core is separated into two sections by a sedimentary rock interval comprising sandstone and interbedded sandstone and shale. A total of 45 samples from the drill core, including coal and associated strata, were selected on the basis of lithology for mineralogical and geochemical analysis. For each interval, a representative portion of the core was samples, and the rest proportion was kept for later reference if required. This was intended to allow a more extensive study of the marine-influenced coal, especially the upper section of the coal seam.

Figure 4.2 shows the lithologic profile of the Greta seam section at each location, with indication of macroscopic appearance of individual coal plies from Austar Coal Mine. Symbolic representation for coal lithotypes were used in accordance with Australian Standards (Standards Australia, 1993, 2007b), based on the relative abundance of dull and bright coal lithotypes within the coal interval (Table 4.2). The Greta coals in the Austar Coal Mine are dominated by banded bright and dull lithotypes (BD and BB) (Figure 4.2B).

Table 4.2 Standard brightness categories (Standards Australia, 1993, 2007b)

Brightness designation Coal description DD Dull (less than 1% bright) DM Dull with minor bright bands (1% to 10% bright)

DB Mainly dull with frequent bright bands (10% to 40% bright)

BD Interbanded dull and bright (40% to 60% bright) BB Bright with dull bands (60% to 90% bright) BR Bright (greater than 90%)

Abundant visible pyrite particles occur in the roof sample G3. Minor sulphate minerals are also visible in a small coal band (G8) in the upper sections of the core. Carbonate veins are visible in coal hand specimens from the lower section, as cleat or fracture infillings. The veins commonly cut through coals of different megascopic lithotypes, but terminate at non-coal intervals.

Chapter 4 Sampling and Analytical Techniques

Figure 4.1 Map of Sydney Basin and locations of seam sections of the Great Northern seam and the borehole of the Greta seam.

Figure 4.2 Lithologic columns of the Greta seam at Cessnock No. 1 Colliery (A) and Austar Coal Mine (B).

Chapter 4 Sampling and Analytical Techniques

Great Northern seam A series of 13 coal and associated non-coal rock samples from a section (channel sample) through the Great Northern seam at Newvale No.1 Colliery, located approximately 35 km SSW of Newcastle (Figure 1), originally collected in the 1960s, were supplied by a sample bank maintained by CSIRO Energy Technology. The basic properties of this section had previously been analysed on a ply-by-ply basis as part of a national coal quality assessment. The Newvale No.1 Colliery was closed in 1994, but the seam is still mined at other locations in the area. An additional set of 18 samples, including coal, claystone partings and roof and floor materials, was taken from an outcrop of the Great Northern seam at Catherine Hill Bay (Figure 4.1) some during the early 1990s (processed at that time and the mineral matter stored) and the remainder in 2011, to enable the features of the seam to be evaluated in a wider geological context. The outcrop samples were recovered from a significant depth within the seam, approximately 30 cm from the surface. The seam section at each location is illustrated in Figure 4.3, with indication of the variation in relative abundance of dull and bright coal lithotypes.

Figure 4.3 Lithologic column sections of the Great Northern seam at Newvale colliery (A) and Catherine Hill Bay (B).

The contact between the conglomerate and the underlying coal seam at Catherine Hill Bay is disconformable, and in one part of the outcrop a thick layer of mudstone overlain by a thin bed of coal occurs between the conglomerate roof and the main part of the coal seam. This section is illustrated separately (Figure 4.3B). A tentative correlation between the two sections is also indicated in Figure 4.3. The Great Northern seam at Newvale is thinner than that exposed at Catherine Hill Bay; correlation of the two sections suggests that this is probably due to contemporaneous erosion of the upper part of the coal bed by processes associated with deposition of the Teralba Conglomerate.

Although the organic matter of the coal is affected by weathering, the exposure at Catherine Hill Bay allows a more extensive study of the sediments deposited in the area before and after the peat swamp was established.

Bulli seam A series of nine coal and non-coal rock samples from a section through the Bulli seam at Coal Cliff Colliery, located approximately 23.5 km NNE of Wollongong (Figure 4.4), originally collected in the 1960s, were also supplied from the sample bank of CSIRO Energy Technology. Coal Cliff Colliery was closed in 1991, but the same seam is still worked at West Cliff and other mines in the area in the Southern Coalfield. The lithotype profile of the Bulli seam section at the Coal Cliff Colliery is illustrated in Figure 4.4.

Chapter 4 Sampling and Analytical Techniques

Figure 4.4 Map of Sydney Basin and locations of the Bulli seam at Coal Cliff Colliery.

Figure 4.5 Lithologic column sections of the Bulli seam at Coal Cliff Colliery.

4.1.2 Samples from the Songzao Coalfield, SW China

A total of 24 coal and non-coal samples were collected from seam sections (channel samples) taken at the underground working faces of three operating coal mines in the Songzao coalfield, Datong (DT), Tonghua (TH) and Yuyang (YY) (Figure 3.11). Channel samples were taken, 15 cm across and 10 cm deep, in accordance with Chinese Standard GB/T 482-2008 (2008). The individual samples from these sections were differentiated from each other on the basis of their megascopic lithology (Figure 4.6). No visible oxidation is apparent in the Songzao coals.

Figure 4.6 Lithologic columns of coal seams in the Songzao coalfield. (A) No. 7 seam of the Datong Coal Mine; (B) Seam K2b in the Tonghua Coal Mine; (C) No. 11 seam in the Yuyang Coal Mine.

4.2 Sample preparation

Large chips were selected at random from each coal and non-coal sample for preparation of polished sections and/or thin-sections, and also kept for later reference if required. The remainder of each sample was crushed into smaller chips (1 cm) using a rock hammer. The chips representing each sample were ground to less than 200 mesh (< 0.2 mm) using either a zirconia mill (Newvale, Coal Cliff, Austar, Songzao Coalfield) or a ceramic mortar and pestle (Catherine Hill Bay), and split into representative portions for analysis. Use of the zirconia mill was intended to facilitate study of the trace element components, which would avoid introducing elements such as W, Co, and Nb (e.g. Hickson and Juras, 1986), from a tungsten carbide grinding mill.

Chapter 4 Sampling and Analytical Techniques

Epoxy-mounted block samples were made from large chips (those before crushed into smaller chips) of all the coal and non-coal samples for petrographic and/or electron microscopy/microprobe analyses, except the samples from the CSIRO sample bank. Gain mounts (<2 mm) that had previously been made from duplicates of the Great Northern, Bulli and Greta coal seam samples taken from the CSIRO sample bank were also made available for the study. Those grain mounts had been previously subjected to proximate and microlithotype analyses at CSIRO. Further petrographic analysis of these samples, however, as well as the sections of the other materials, was carried-out for the present study.

The 45 samples of the Greta samples from the drill core were subjected to petrographic and mineralogical analyses (see below). A number of samples covering thin intervals that were adjacent to each other and had similar mineralogical characteristics were then combined, reducing the 45 samples to 23 samples, in order to provide enough material for more detailed geochemical study. To avoid further oxidation, the Greta samples, including the cores, coal chips, pulverized coal and polished sections, both for analysis and to be kept for reference, were stored in a freezer. Samples from other locations, including coal, rock and coal LTA samples were kept at room temperature, with minimal change to the mineralogical and geochemical characteristics.

4.3 Analytical techniques

A combination of techniques for mineralogical and geochemical analysis was applied to the samples, as indicated in Figure 4.7. Standard procedures (e.g. Standards Australia, or ICCP) were used for the individual analyses, where required.

4.3.1 Low-temperature oxygen-plasma ashing, X-ray diffraction (XRD) and Siroquant analysis

Low-temperature oxygen-plasma ashing was carried out on the powdered coal samples, using an IPC 4-chamber oxygen plasma asher at the School of Biological, Earth and Environmental Sciences (BEES), University of New South Wales (UNSW), and following procedures described by Standards Australia (Standards Australia, 2000b). The percentage of low-temperature ash (LTA) was determined for each coal sample.

The resultant LTAs, and also the powdered non-coal samples, were subjected to X-ray diffraction (XRD) analysis, using a Philips PW 1830 diffractometer system with Cu-.Į radiation and a graphite monochrometer, and with a tube voltage of 40 kV and current of P$7KH;5'SDWWHUQZDVUHFRUGHGRYHUDșLQWHUYDORI–60°, with a step size of 0.04° and a count time of 2 seconds per step. The minerals in each sample were identiefied using the Xplot software program (M.D. Raven, CSIRO), which allows display of the XRD data, search/match and display of mineral patterns based on data in the ICDD (International Centre for Diffraction Data) Powder Diffraction File.

Figure 4.7 A scheme of sample preparations and analyses.

The XRD pattern obtained from each coal LTA and non-coal sample was analysed using the Rietveld-based Siroquant™ software package, developed by Taylor (1991b), to obtain the quantitative mineral proportions. In each Siroquant analysis, the background was removed from the pattern and a calibration function applied to allow for the geometry of the sample holder. A synthetic XRD pattern was then generated, based on the minerals present, and the crystallographic parameters (unit cell dimensions, linewidths, and orientation) for each phase interactively adjusted to match the observed XRD pattern. The final output, after the best possible match had been obtained, included an estimate of the

Chapter 4 Sampling and Analytical Techniques error associated with each individual phase determination, (equivalent standard deviation or ESD) and an estimate of the overall goodness of fit, referred as a global chi-squared value. The overall error associated with each mineral percentage is indicated by the ESD for that phase multiplied by the square root of the global chi-squared value. Additional details of the operation of the Siroquant software are given by Ward et al. (1999b).

4.3.2 Clay fraction preparation and clay mineral analysis

As discussed by Ruan and Ward (2002), powder XRD patterns of bulk samples alone may provide an inadequate basis for identification of clay minerals, especially those that are extremely fine-grained (<2 μm) and poorly crystallized, and also the expandable lattice species. In order to obtain a more precise identification and quantification of the clay minerals, the <2 μm fraction was separated from each powdered bulk (LTA and rock) sample, and prepared as oriented-aggregates following the methods described by Gibbs (1971). Small amount of samples were dispersed in beakers with deionized water, and sodium hexametaphosphate (Calgon) was added in order to avoid flocculation. The suspensions were then placed in an ultrasonic bath for 5 minutes to ensure complete dispersion of the clay materials. After settling for 2 hours, the suspension in the beakers was removed from a depth of around 2 cm, and centrifuged at 2000 rpm for 20 minutes. The supernatant water was decanted after centrifuging, and the clay concentrates at the bottom were re-mixed with a small amount of water and transferred using a pipette to pre- labelled glass slides. The slides were then left to dry in the laboratory atmosphere.

XRD data were obtained for the clay slides after exposure to ethylene glycol vapour for approximately 24 hours, and after heating at 400 qC in a muffle furnace for 2 hours. The ;5'SDWWHUQVZHUHUHFRUGHGRYHUDșLQWHUYDORI–30°, with a step size of 0.04° and a count time of 1 second per step, with other parameters the same as for the powder samples. The clay minerals were identified by reference to works such as Moore and Reynolds (1997), based on comparison of the glycolated and heated patterns.

The method of Griffin (1971) was used to provide quantitative mineralogical data on the clay fraction, based on the intensity of the peaks at 7 Å and 10 Å in the oriented aggregate diffractograms after glycol saturation and after heating at 400 qC. The XRD patterns were processed using Philips APD software, as described by Ruan and Ward (2002), to obtain the relevant peak intensities above background. The proportion of kaolinite (+chlorite), illite and expandable clay minerals was calculated for each sample according to the

85 following formulae (K=kaolinite, I=illite, E=expandable clay minerals, h=heated, g=glycoled) (Griffin, 1971):

%K(+C)=(7Åh/2.5)/(7Åh/2.5+10Åh)*100 %I=(10Åg/10Åh)/(7Åh/7Åg))*(100-%K+C) %E=100-%K+C-%I

The height of 7 Å peak in each case was divided by 2.5 to allow for the difference in the diffraction intensity between kaolinite and the illite and expandable clay components. The nature of the expandable clay minerals was also evaluated, based mainly on data from Moore and Reynolds (1997).

4.3.3 Petrographic analysis

Petrographic analysis of the coal polished sections, including mean maximum vitrinite reflectance (Rv, max) or mean random vitrinite reflectance (Rv, r) measurement and maceral analysis using a Zeiss Axioplan reflected-light microscope, equipped with white (100 W halogen) and blue violet (HBO) light sources, and a Zeiss MPM photometer. Maceral classifications are after International Committee for Coal and Organic Petrology (1998, 2001).

Mean maximum or random vitrinite reflectance was measured on the polished surfaces of the epoxy grain mounts or epoxy-impregnated block samples in oil immersion, using monochromatic light (wavelength 546 nm), according to Standards Australia (2000a). The maximum vitrinite reflectance value for each particle measured was recorded by the photometer as the object stage was rotated through 360q. The maximum or random reflectance was measured at a number of points (30-50 measurements) on telovitrinite, or detrovitrinite when telovitrinite was not available. The mean maximum or random reflectance (Rv, max or Rv, max) value of the vitrinite in the coal sample was then calculated.

Grain mounts of selected coal samples (Bulli and Great Northern seams) were subjected to maceral analysis using a point counter, following procedures described by Australian Standards (1998b). Although the quantitative measurement of the maceral abundance was not done for coal samples from other locations, the maceral characteristics and the association of the mineral matter with the maceral components were also assessed from the polished blocks.

Chapter 4 Sampling and Analytical Techniques

Polished thin-sections of the rock samples were subjected to petrographic analysis using a Leica DM 23500P polarizing microscope equipped with a Leica LAS digital imaging software. This allowed examination the texture, the modes of occurrence of the different minerals, and the association of the minerals and organic components in intra-seam tonsteins and other non-coal rock samples.

4.3.4 Scanning electron microscopy (SEM)

The main purpose of using SEM in the present study was to identify the modes of occurrence of the different mineral components within the coal and non-coal strata, as well as to investigate any variations in mineral composition. Selected coal and rock polished sections were also studied using Hitachi 3400-X and Hitachi 3400-I scanning electron microscopes in the Electron Microscope Unit, Mark Wainwright Analytical Centre, UNSW. Both SEMs were fitted with secondary and backscatter electron detectors that allowed for topographic and compositional surface imaging. The Hitachi 3400-X and 3400- I were also equipped with Thermal Fisher energy-dispersive X-ray spectrometers (EDS), which respectively allowed qualitative and semi-quantitative elemental analysis of particular micro-areas. For the coal and other carbonaceous rock samples in the present VWXG\WKHDFFHOHUDWLQJYROWDJHZDVN9DQGWKHEHDPFXUUHQWZDVZLWKLQíP$ during the SEM operation. Prior to examination, specimens were coated with carbon to avoid charging effects, and then mounted on a small aluminium stub, with carbon tape being used to ensure good electrical contact between the specimen and the stub.

4.3.5 Electron microprobe analysis

Electron microprobe analysis was applied to a polished block sample of a Songzao coal, to obtain precise quantitative chemical analysis (particularly for rare earth elements, REE) of REE-bearing mineral veins occurring in that coal. This analysis also helped to better understand the nature of the REE carbonate minerals, which rarely occur in coal. A JEOL JXA-8600 Super-probe, fitted with wavelength-dispersive spectrometry (WDS), was used at the School of Natural Sciences, University of Western Sydney. The microprobe was operated at 15 kV, 15 nA. Standards employed were CaWO4 for Ca, LaB6 and CeO2 for La and Ce, and all pure metals for the other rare earth elements.

4.3.6 Laser Raman spectroscopy

Laser Raman spectroscopy is a non-destructive, fast technique for investigation of molecular structure. A Renishaw Raman spectrophotometer was used at the University of Sydney, to obtain laser Raman spectra from the REE-bearing minerals occurring in the Songzao coal.

4.3.7 Proximate analysis

Proximate analysis was done for the Songzao coals at SGS Australia Pty Ltd, including analysis of inherent moisture, ash yield, volatile matter, fixed carbon, total sulphur content. Forms of suphur were also analysed, but only for selective Songzao coal samples. This is partly to investigate the forms of sulphur in the coals, and partly to cross check the mineralogical data, especially pyrite and sulphate minerals. The proximate analysis data of the Bulli, Great Northern and Greta coals from the CSIRO sample bank were supplied in the CSIRO report.

4.3.8 X-ray fluorescence (XRF) spectrometry

Major element concentrations in the coal and rock samples were determined by X-ray fluorescence techniques, at the Mark Wainwright Analytical Centre, UNSW. All the coal samples, except the Great Northern coals from Catherine Hill Bay, were ashed at 815°C. These ashes, and also samples of the powdered associated non-coal rocks, were then fused into borosilicate glass discs (Norrish and Hutton, 1969). The borosilicate discs were analysed using a PANalytical (formerly Philips) PW2400 XRF spectrometer, equipped with a wavelength dispersive (WD) detection system. SuperQ software was used for to determine the concentrations of major elements present.

A range of 38 trace elements in coal samples was also determined using PW 2400 XRF spectrometer, with PANalytical Pro-Trace software. Each coal sample was mixed with boric acid as a binder and the mixture was then pressed into pellets for analysis. Pro- Trace calibration algorithm enables an accurate determination of matrix, background and spectral interference (line overlap) correction factors. This enables analysis of accurate and reliable trace element of sub-ppm levels. The use of Pro-Trace XRF was intended to provide a check on trace elements data derived from other analytical techniques (e.g. ICPMS). Although the detection limit may not necessarily be as low as other techniques

Chapter 4 Sampling and Analytical Techniques

(e.g. ICPMS), Pro-Trace XRF is a relatively simple, non-destructive and cost-effective method of analysis of trace elements.

4.3.9 Inductively coupled plasma-mass spectroscopy/optical emission spectroscopy (ICP-MS/OES)

Trace element concentrations in the coal and rock samples were determined by ICP-MS or ICP-OES (also referred to as ICP-AES). These techniques were used because of their low detection limits, which are typically in the ppb to ppt range for most trace elements.

Prior to the ICP-MS/OES analysis, the samples were digested for the study by CSIRO Energy Technology. This procedure used in this case was to break down the sample matrix and leave the elements of interest in solutions which were ready for ICP analysis. Two separate digestion procedures were carried-out, to accomplish total decomposition of the samples and to ensure the total content in the sample reflected in the resultant digests. One procedure involved ashing the coal and rock samples at 450°C. The ashes were then subjected to microwave dissolution in a mixed acid (HCl and HNO3). The other procedure involved fusion with a mixture of lithium tetraborate (Li2B4O7) and lithium metaborate

(LiBO2) flux, done on the relevant samples without an ashing process.

Following microwave-assisted acid digestion or borate fusion, the resultant digests were analysed by ICP-MS/OES at the Mark Wainwright Analytical Centre, UNSW. All the determined values were then calculated as concentrations in the original coal or rock samples.

4.3.10 Eschka method (B and Cl)

Boron was determined using Eschka method, following method described by Standards Australia (1998a), at CSIRO Energy Technology. Each coal or rock sample was ignited in intimate contact with Eschka mixture in an oxidizing atmosphere at 800°C to decompose the organic matter. Eschka mixture is a mixture composed of two parts by mass of light magnesium oxide and one part by mass of anhydrous sodium carbonate. The residue was then extracted with hydrochloric acid and heated on a water bath at 60°C to 70°C for 1 h. the resultant solution was the test solution. The procedure also involved preparation of standard solutions from a known anhydrous boric acid solution and a blank solution. Boron concentration in each solution was determined by ICP-OES.

Chlorine was determined using Eschka method, following method described by Standards Australia (1999). Each coal or rock sample was ignited in intimate contact with Eschka mixture in an oxidizing atmosphere at 675°C for 2 hours, to remove the organic material and to convert all the chlorine to chloride. This was then transferred to a beaker and reacted with nitric acid. This was agitated until the solution of the soluble material is complete. The resultant solution was then filtered into a conical beaker. The filtrate was then mixed with silver nitrate solution and n-hexanol. Ammonium iron sulphate

((NH4)2SO4Fe2 (SO4)3.24H2O) indicator solution, titrate with potassium thiocyanate solution was then added until the solution became faintly orange-pink in colour. The difference of volume of potassium thiocyanate used in test and the blank solutions was then used, to calculate the Cl concentration in the test solution.

4.3.11 Hydride generation-atomic fluorescence spectroscopy (HG-AFS) (As and Se)

Arsenic and selenium in the Bulli and Great Northen coals were analysed using HG-AFS technique, following procedures described by Standards Australia (1998c), at CSIRO Energy Technology. Each coal or rock sample was ignited in intimate contact with Eschka mixture in an oxidizing atmosphere at 800°C to decompose the organic matter. The residue is then extracted with hydrochloric acid on a water bath at 60°C to 70°C for one hour. The resultant solution was the test solution. A calibration blank solution was prepared by dissolving Eschka mixture with hydrochloric acid and deionised water. Three calibration standards were also prepared for As and Se, respectively, by combining the blank solution and certain volumes of working As (50 μg/L) or Se (50 μg/L) standards. For As analysis, each of the calibration standard, test solution, and the blank solution was mixed with hydrochloric acid and potassium iodide solution in a vial, left for 20 min. For Se analysis, each of the calibration standard, test solution, and the blank solution was mixed with hydrochloric acid, heated at 90°C for 15 to 20 minutes to reduce selenate to selenite. The resultant solution derived from each procedure was then for ready hydride generation. The analytes were then converted to their volatile hydrides in a hydride generation apparatus, and transported into the atomisation device in the light path of the spectrometer. Fluorescence signal (peak height) at certain wavelength (As at 193.7 nm, and Se at 196 nm) was recorded for each solution.

A calibration graph of concentration of analyte versus instrument reading was constructed, from the standard solutions of As and Se, respectively. This calibration graph was then

Chapter 4 Sampling and Analytical Techniques used to convert the reading obtained on the test and blank solutions to the concentrations of As or Se. All the determined values were calculated as concentrations in the original coal or rock samples.

4.3.12 Inductively coupled plasma-mass spectroscopy (ICP-MS) (As and Se)

Arsenic and selenium in the Greta and Songzao coals were analysed using ICP-MS technique, following the method described by Dai et al. (2011, 2012c), at the China University of Mining and Technology, Beijing (CUMTB). Arsenic and Se were determined by ICP-MS using collision cell technology (CCT) in order to avoid interference from polyatomic ions. For ICP-MS analysis, samples were digested using an UltraClave Microwave High Pressure Reactor (Milestone). The basic load for the digestion tank was composed of 330 ml Mili-Q H2O, 30 ml 30% H2O2 (Metal-Oxide-Semiconductor, MOS reagent), and 2 ml 98% H2SO4 (Guaranteed reagent, GR). Initial nitrogen pressure was set at 50 bars and the highest temperature was set at 240°C for 75 minutes. The reagents digestion of for each 50-mg coal sample were 5 ml 65% HNO3, 2 ml 40% HF, and 1 ml 30%

H2O2. The Guaranteed-Reagent HNO3 and HF for sample digestion were further purified by sub-boiling distillation. Multi-element standards (Inorganic Ventures: CCS-1, CCS-4, CCS-5, and CCS-6) were used for calibration of trace element concentrations.

4.3.13 Cold vapour-atomic fluorescence spectroscopy (CV-AFS) (Hg)

Mercury in the Bulli and Great Northern coals was determined using CV-AFS technique, following procedures described by Standards Australia (2007a), at CSIRO Energy Technology. Each coal or rock sample was digested in a closed vessel with a mixture of concentrated nitric and hydrochloric acids, and the digestion process was kept on a water bath of 80°C for 90 minutes. The mercury in the digest was then completely oxidized with potassium permanganate solution. The resultant solution is decolourized with hydroxylamine hydrochloride. The procedure also involved preparation of standard solutions which were diluted in the same matrix as the test solutions and decolourized in exactly the same manner. An aliquot is reacted with stannous chloride to reduce the ionic mercury species to the elemental state. Mercury vapour is carried into an atomic fluorescence spectrometer by a controlled flow of gas. The fluorescence signal was read for all the test standard and blank solutions. A calibration graph of fluorescence versus Hg content was prepared from the standard solutions. The concentrations of Hg in test and

91 blank solutions were then calculated using the graph. All the determined values were calculated as concentrations in the original coal or rock samples.

4.3.14 DMA-80 direct Hg analyser (Hg)

Mercury in the Greta and Songzao coals was analysed using a Milestone DMA-80 Hg analyser, at CUMTB. The analysis is compliant with ASTM method D-6722-01 (Standard test method for total mercury in coal and coal combustion residues by direct combustion analysis). Each coal or rock sample was weighed into a quartz boat and loaded onto the instrument autosampler. The sample was first dried and then decomposed in a oxygen- rich furnace. Mercury in the sample was released and carried to the catalyst section of the furnace, selectively trapped, and then flown via the carrier gas into the optical path of the spectrophotometer. The atomic absorption at 253.65 nm was measured. The DMA-80 is matrix independent and a liquid mercury standard reference was used.

4.3.15 Pyrohydrolysis/fluoride ion-selective electrode (ISE) (F)

Fluorine in the Greta and Songzao coals was determined at CUMTB using a pyrohydrolysis/fluoride ion-selective electrode technique, following procedures described by Chinese National Standard GB/T 4633-1997. Each coal or rock sample was mixed with silica and then combusted in a tube furnace at approximately 1200 °C in an atmosphere of oxygen and water vapour for 15 minutes. The volatilised fluorine compounds are absorbed in a suitable solution with a volume of 100 ml, and processed (with sodium citrate, potassium nitrate), for determination by ISE techniques. The procedure also involved preparation of standard solutions from a known sodium fluoride solution and a blank solution.

The results of two consecutive determinations carried out in the same laboratory by the same operator using the same apparatus were not allowed to differ by more than either 15 ppm (if the total F concentration of coal is less than 150 ppm) or 10% (relative; if the total F concentration of coal is more than 150 ppm). The standard reference materials GB 11121 and GB 11123 (two Chinese coals) were analysed with each batch of samples.

Chapter 4 Sampling and Analytical Techniques

4.3.16 Data compilation and processing

The data derived from different analytical techniques (e.g. proximate, mineralogical, and chemical analysis) were tabulated in Excel spreadsheets, which were then compiled for each set of samples, set out to compare parameters, and produce relevant tables and graphic plots. For example, graphic plots showing reflectance and mineralogical profiles of coal seams were created using Excel formulas and functions.

Lithologic logs were created using Sedlog 2.1.4, which is a free software package for creating graphic sedimentary logs. Data in spreadsheet format were imported into SedLog, and used as the basis for drawing graphic logs. The log data were then exported in Scalable Vector Graphics (SVG) format, and imported into graphic software (Coredraw X3) for further modification and editing.

Cluster analysis was performed to distinguish different groups of trace elements in each set of coal samples, using statistical analysis software SPSS 16.0 (now IBM SPSS). Data in Excel spreadsheet format was loaded into SPSS “Data Editor” which is a data spreadsheet. Elements were used as variables. Hierarchical clustering was performed using the Pearson correlation coefficients. The dendrogram derived from cluster analysis was also imported and edited with Coredraw.

Results from other analytical techniques (e.g. optical and electron microscope images, XRD patterns) were integrated and processed using other softwares such as Photoshop, Grapher, Microsoft Office PowerPoint and Corel PaintShop. Coredraw X3 was also used to re-draw and modify geological map cited from publications.

4.4 Summary

A series of analytical techniques was applied to obtain comprehensive mineralogical and geochemical data for the present study. Reflected and transmitted light microscopes were used for petrographic analysis of coal and rock samples, respectively. Quantitative mineralogy of coal and rock samples was obtained by using XRD and Siroquant software, which in turn allowed study of the vertical variation in the mineral assemblage through each coal seam. The oriented aggregate technique was used for more accurate identification and quantification of clay minerals. Modes of occurrence of minerals were evaluated by optical and electron microscope analysis. EDS provided micro analysis of

93 chemical composition, and allowed to identification of those minerals which were present in concentrations below the detection limit of the XRD system. Several bulk elemental analysis techniques were used to determine the concentrations of major and trace elements in coal and rock samples according to Australian standards. The analytical results were processed using a range of computer graphic and statistic softwares, and presented as tables, maps, graphs, images, etc. in the present thesis.

CHAPTER 5 MINERALOGY AND GEOCHEMISTRY OF THE GRETA SEAM

This chapter discusses the maceral characteristics, maceral-mineral associations, and variation in vitrinite reflectance through the marine-influenced Greta seam. It also includes a discussion on the modes of occurrence of the mineral matter and trace elements in the coal and associated non-coal strata, and the relations between trace elements and mineral matter components.

As mentioned in Chapter 4 and discussed in this chapter, the study involves samples from the Greta seam at two sections (Austar and Cessnock). However, in the coals from the Cessnock section, the pyrite that originally formed the most abundant mineral in the coal has been oxidized to a large extent. This caused some analytical difficulty in quantitative mineralogical analysis, not only due to the complex secondary iron sulphate minerals produced, but also an overall low XRD intensity, probably arising from the water in the hydrous sulphates. For that reason, the investigation was mainly based on the samples from the Austar section, where the bulk of samples had not been subjected to such oxidation.

The purposes of this study were to examine the vertical mineralogical and geochemical variations through the coal seam, and to evaluate the associations between the different groups of minerals and trace elements within the coal. Such a study was also expected to provide an opportunity to evaluate the geological factors responsible for the mineralogical and geochemical characteristics of the coal seam, and provide a basis for comparison to the other seam sections included in the research program.

5.1 Proximate analysis

The proximate analysis data, as well as total sulphur and calorific value of samples from the Cessnock seam section, representing the lower and more workable part of the seam, are listed in Table 5.1. Figure 5.1 shows the variation of proximate analysis parameters within the Cessnock section, as well as the variation in LTA percentages for the coals of the Austar section, for which proximate analysis data are not available. The LTA percentage, which closely represents the proportion of mineral matter in the coal, is

95 generally a little greater than ash yield from 132) proximate analysis. Also listed in Table 5.1 are the proximate analysis data for the composite subsection (samples 2097 to 2101) of the Cessnock seam section.

The two seam sections (Austar and Cessnock) are discussed together, although only the lower part of the seam was sampled at the Cessnock location. Based on the fixed carbon percentage and the calorific value, the Greta coal is classified as a high volatile A bituminous coal under the ASTM classification (ASTM, 2007).

The LTA percentage of the Greta seam in the Austar section generally increases upwards through the profile, with the exception of the uppermost and lowermost coals, and some small dull bands in the lower section. This is mainly due to abundant pyrite in the upper part of the seam section, and to the presence of relatively abundant dawsonite in the upper part of the lower section. The Greta seam in the Cessnock section has an overall low ash yield, ranging from 3 to 9% (air-dried basis). This is consistent with the LTA profile of the lower part of the Austar section, where the LTA percentage of the coals varies from 4 to 14.4%, with the exception of a few small dull bands. Relatively high ash yields or LTA percentages occur in the upper few metres of both sections.

Table 5.1 Proximate analysis and total sulphur of the Greta coal samples in the Cessnock seam section (data from CSIRO reports)

Thickness Moisture Ash Fixed carbon Volatile matter Calorific value Total S Sample (m) (%, ad) (%, ad) (%, ad) (%, ad) (% daf) (Btu/1b, daf) (%, ad) 2092 0.50 1.72 5.85 44.48 47.95 50.00 15190 2.86 2094 0.60 2.04 8.66 44.40 44.90 48.00 15090 3.42 2095 0.05 1.65 20.22 38.63 39.50 50.60 15100 4.33 2097 0.51 1.59 6.73 48.08 43.60 47.60 15120 3.80 2098 0.18 1.56 8.87 44.92 44.65 48.10 15060 4.07 2099 1.27 2.02 3.23 52.55 42.20 44.50 15170 1.08 2100 1.17 2.48 4.04 49.67 43.81 44.20 15240 0.97 2101 0.41 2.12 3.16 51.32 43.40 54.80 15280 1.03 2102 1.00 1.93 7.00 49.67 41.40 55.40 15150 1.25 ad=air-dried basis; daf=dry ash-free basis; CV=calorific value; -=no data.

The moisture content of the coal samples ranges from 1.56 to 2.48% (air-dried basis), with the higher values occurring in the low-ash coals in the lower metre or so of the section (samples 2099, 2100, and 2101). This may be associated with the presence of higher

Chapter 5 Mineralogy and Geochemistry of the Greta Seam proportions of clay minerals (Table 5.4) which contain moisture attached to their crystal structures, or it may simply reflect the greater abundance of (micro-porous) organic matter.

The fixed carbon of the coal is the percentage of carbonaceous material which is left after the volatile matter is driven-off. The fixed carbon varies between 44.4 and 52.55%, with 47.08% on average (dry, ash-free basis). The fixed carbon percentage appears to be negatively correlated with ash yield and volatile matter (Figure 5.1). The volatile matter ranges between 44.2% and 55.4%, with an average value of 49.24% (dry, ash-free basis). The highest volatile matter value occurs the lower two samples (2101 and 2102). The total sulphur content in the Cessnock section is low in the lower half of the profile, but increases to a relatively high level in the top few samples (Figure 5.1). This indicates that some marine or brackish water percolated into the peat at the top of the Cessnock section, although it was not as significant as in the peat at the actual top of the Greta seam.

Figure 5.1 Variation of proximate analysis data and total sulphur through the Greta seam of the Cessnock and Austar sections.

5.2 Coal petrology

As described in Chapter 4, the macroscopic appearance of the Greta coals is dominated by banded bright and dull lithotypes. Maceral characteristics and the association of the mineral matter with the individual organic components of coal (macerals) were assessed from polished block samples of the Greta coal in the Austar section. Evaluation of the macerals was carried-out on the whole-coal block samples, rather than crushed grain mounts, in order to better examine the relations among individual macerals and mineral matter components. However, the abundance of the different macerals was not quantitatively measured, due to the unavailability of the grain mount samples. The classification of macerals for this study was based on the ICCP standard, which is indicated in Table 5.2. The mineral matter was also examined at a microscopic scale, and is discussed separately in the coal mineralogy section of this chapter.

5.2.1 Maceral composition

5.2.1.1 Vitrinite

The vitrinite group macerals in the Greta coals are mainly telocollinite and collodetrinite, along with trace amounts of gelinite. Collotelinite occurs as structureless and homogeneous bands (Figure 5.2A). Some collotelinite is impregnated with in situ inclusions, such as resinite and syngenetic minerals (Figure 5.2B), or cut by mineral-filled veins (Figure 5.3C). Collodetrinite occurs as a matrix which is impregnated with other in situ macerals, such as inertodetrinite, sporinite, cutinite, and liptodetrinite (Figures 5.2A, B). Gelinite in the Greta coals is rare, and when present, occurs as infillings of the open cell structure of inertinite (mainly fusinite or semifusinite) (Figure 5.2E). Although subject to errors resulting from the representativeness of the polished block samples, the Greta coals rarely contain telinite, which is a maceral of the telovitrinite subgroup with preserved cell structures, in spite of the coal’s high volatile bituminous characteristics.

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Table 5.2 Classification of maceral groups, sub-groups and macerals according to the ICCP System 1994 Maceral group Subgroup Maceral Maceral group Subgroup Maceral Vitrinte Telovitrinite Telinite Liptinite Collotelinite Sporinite Detrovitrinite Vitridetrinite Cutinite Collodetrinite Resinite Gelovitrinite Corpogelinite Fluorinite Gelinite Alginite Inertinite Bituminite Fusinite Exsudatinite Semifusinite Liptodetrinite Funginite Secretinite Macrinite Micrinite Inertodetrinite

5.2.1.2 Liptinite

The Greta coals have relatively high proportions of liptinite. The liptinite group macerals occurring in the Greta coals are mainly sporinite, cutinite and, to a lesser extent, resinite and exsudatinite. Sporinite typically occurs as discrete bodies (Figure 5.2C) in collodetrinite, along with inertodetrinite, or as aggregations (Figure 5.2D). Cutinite was formed from cuticles, and commonly appears folded, toothed and embedded in collotelinite (Figure 5.2B), or straight, long and densely packed (Figure 5.3A). Thick-walled cutinite also occurs (Figure 5.2F), which is distinguished from the common thin-walled cutinite.

Resinite in the Greta coal occurs as oval-shaped bodies, with an orange to reddish- orange colour under incident blue light excitation (Figure 5.3B). The resinite commonly displays orange tinged or red internal reflections (Figure 5.4C). Exsudatinite in the Greta coal occurs as infillings of vein-like structures, with a greenish-yellow colour under incident blue light excitation (Figures 5.3C, D). Liptodetrinite is relatively common (Figures 5.3E, F).

Figure 5.2 Photomicrographs showing typical macerals in the Greta coals (Austar). Oil immersion, reflected light. (A) Telocollinite (tc) (upper field) and collodetrinite (cd) (lower field). Other macerals illustrated are semifusinite (sf) and funginite (fg) of fungal hyphae. (B) Cutinite (c) embedded in collotelinite. Also illustrated are syngenetic pyrite and funginite of fungal spores. (C) Fusinite (f), semifusinite (sf) and sporinite (sp). (D) An agglomeration of sporinite. (E) Gelinite (g) infilling cell cavities of semifusinite (F) Thick-walled cutinite. Micrinite and inertodetrinite are incorporated in the cutinite.

100

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Figure 5.3 Photomicrographs of liptinite macerals in the Greta coals (Austar). Oil immersion, reflected light and blue-violet fluorescence. (A) Resinite (rs) and sporinite (sp) in collodetrinite. (B) Same view as (A) under fluorescent illumination. Other fluorescing matter is sporinite and cutinite (across the field). (C) Exsudatinite filling vein-like structure (horizontal), associated with clay mineral-filled cleats (vertical). (D) Same view as (C) under fluorescent illumination. (E) Concentrates of liptodetrinite (intense fluorescence) and clay minerals (weak fluorescence) in vitrinite. (F) Same view as (A) under fluorescent illumination. Other fluorescing matter in the matrix of collodetrinite is mainly sporinite, liptodetrinite and probably some clay minerals.

101

Figure 5.4 Photomicrographs showing typical inertinite macerals in the Greta coals (Austar). Oil immersion, reflected light. (A) An agglomeration of funginite. Cutinite forms long, dark bands stretching across the upper field. Note compression effect around secretinite. (B) Non-vesicular secretinite showing internal notches. Note compression effect around secretinite. (C) Funginite (fg) of fungal spores and disseminated micrinite (mi), inertodetrinite (id) in a matrix of collodetrinite. (D) Funginite (fg) of fungal hyphae in the centre of the view. (E) Vesicular secretinite (se). (F) Fusinite (f) grading to semifusinite (sf). Funginite (f) are incorporated in the fusinite. Fusinite grades upwards into semifusinite.

102

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

5.2.1.3 Inertinite

The liptinite group macerals occurring in the Greta coals are mainly semifusinite and, to a lesser extent, fusinite. Both semifusinite and fusinite largely occur as bands and lenses, with a more or less well-preserved cellular structure (Figures 5.2A, C). Fusinite may grade into semifusinite (Figure 5.4F). Other minor inertinite macerals are secretinite, inertodetrinite (Figure 5.4B), macrinite (Figure 5.4C), micrinite (Figure 5.4B), funginite, and secretinite.

Funginite is relatively common in the Greta coals. Two forms of funginite are recognized: one consisting of oval-shaped bodies with cell structure (fungal spores) (Figures 5.2B, 5.4A, B), and the other consisting of hyphae of fungal bodies (Figures 5.2A, 5.4D). Fungal spores also occur in fusinite (Figure 5.4F), although in a more subtle form. Secretinite is distinguished from macrinite commonly by occurring as round bodies without obvious plant structure. It is also distinguished from corpogelinite by higher reflectance and the presence of internal notches (Figure 5.4C) or vesicles (Figure 5.4E).

5.2.2 Vitrinite reflectance

The mean maximum reflectance of vitrinite (mainly telocollinite) was measured on 24 coal samples from the Greta seam in the Austar section. The vitrinite reflectance is generally higher in the lower part of the seam than in the upper part, with the reflectance decreasing from approximately 0.74% in the lower split of the seam section studied to approximately 0.62% at the top (Figure 5.5). The lowest sample of the upper coal bed, however, has the highest vitrinite reflectance among the samples studied, and also alow sulphur content. As indicated by Diessel and Gammidge (1998), the anomalously low (suppressed) reflectance of the vitrinite in the upper part of the Greta seam is due to the overlying marine transgression. In the lower section, where the marine influence is minor, there is a general increase in vitrinite reflectance with increasing depth, except for some of the coal bands.

Diessel (1992) and Diessel and Gammidge (1998) noted that the mean random reflectance of telocollinite in the Greta seam decreases from approximately 0.7% at the bottom to approximately 0.55% at the top. The decrease is also accompanied by an increase in vitrinite fluorescence intensity. The study by Diessel (1992) also indicated identical trends in the overlying Pelton seam, which is in turn directly overlain by the

103 marine Braxton Formation. The marine influence is more significant at the Pelton seam, with vitrinite reflectance decreasing from approximately 0.6% at the bottom to approximately 0.45% at the top of the profile. Reflectance suppression has also been reported in coal sequences due to marine influence elsewhere (Diessel, 1992; Gurba and Ward, 1998; Holz et al., 2002). As discussed by Diessel and Gammidge (1998), anaerobic bacteria-generated lipids are rich in humic degradation products, due to increased bacterial activity in the low-acid peat; the reflectance suppression, as well as the fluorescence enhancement, result from the syn- and epigenetic absorption by the vitrinite nuclei of hydrogen donated, presumably, by anaerobic bacteria-generated lipids. The occurrence of bacterial lipids and finely dispersed liptinitic material in vitrinite has been confirmed by transmission electron microscopy studies (e.g. Taylor, 1991a).

5.3 Mineralogy of the Greta seam

Hand specimen observation indicates the occurrence of visible pyrite particles in the roof sample of the Austar core, and secondary sulphate minerals representing pyrite oxidation products in some coal bands at the top part of the section. Abundant cleat- and fracture- filling carbonates occur in the lower part of the section. However, no macroscopic cleat or fracture fillings are observed in the upper section above the clastic interval. Quantitative XRD data on the mineral assemblage in the coal LTA and non-coal samples of the Austar and Cessnock seam sections are given in Tables 5.3 and 5.4, respectively. Figure 5.5 illustrates the vertical trends in abundance of the different minerals in the coal LTA and associated non-coal strata in the section from the Austar Coal Mine.

104

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Figure 5.5 Plots showing vertical variation of vitrinite reflectance and abundance of major minerals in the Greta seam (Austar).

5.3.1 Minerals in roof, floor and other clastic strata

The immediate floor material underneath the coal is a grey, fine carbonaceous sandstone, containing plant fragments, mainly plant roots. Beneath this are, with increasing distance to the bottom of the coal, light grey, massive sandstone, interlaminated sandstone and shale, and carbonaceous shale. The floor sediments mainly comprise equal proportions of quartz, illite and interstratified illite/smectite (I/S), with minor kaolinite and traces of carbonate minerals, pyrite and anatase (Table 5.3).

The abundance of kaolinite drops dramatically in the floor strata, accompanied by an increase in the proportions of non-kaolinite clay minerals (I/S and illite) and quartz. A small difference between the two immediate floor strata and the underlying sediments is that the former has a slightly higher kaolinite content. This may indicate that organic acids percolated from the peat, and had an influence on the underlying floor strata. Traces of siderite occur persistently in all the floor sediments, mainly as spherulitic grains (Figure 5.6A). Trace proportions of dawsonite also occur, but only in the immediate floor stratum.

The immediate seam roof is a thin (1 mm), mid grey carbonaceous shale, having a relatively sharp contact with the coal. Above the shale there is a granule conglomerate with abundant visible pyrite particles. The roof material has a higher proportion of poorly-

105 ordered kaolinite and pyrite than the floor samples, with I/S exceeding illite in the clay fractions and with minor carbonate minerals. The roof samples also contain minor dawsonite, which commonly occurs as a cement or as nodules (Figure 5.6B). The pyrite in all the roof strata occurs as syngenetic framboids and euhedral crystals (Figure 5.6B). It is especially abundant in the conglomerate band, which may reflect the significant porosity and permeability of the conglomerate, allowing marine water to percolate through.

Dawsonite is relatively widespread in the Permian non-coal rock strata of the Sydney Basin. The earliest record of dawsonite in the Sydney Basin was as a cement in sandstones of the Greta Coal Measures, described by Loughnan and See (1967). Apart from a cementing material or nodules, dawsonite also occurs as late diagenetic veins, replacements of other minerals (e.g. quartz and feldspars), and a daughter mineral in fluid inclusions in rock strata of different ages in the Sydney Basin, as described by Loughnan and Goldbery (1972) and Goldbery and Loughnan (1977).

Figure 5.6 Minerals in the roof and floor samples. Transmitted light with crossed polars, (A) Spherulitic siderite grains in a floor sample. (B) Dawsonite (D) and framboidal and cubic pyrite (P) in a roof sample.

The clastic interval in the middle of the seam section (G-15 to G-17) mainly consists of sandstone and fine interbanded carbonaceous shale and sandstone, with a thin carbonaceous shale (G-15) of 2 cm thickness adjacent to the lowermost coal of the upper section. These strata have similar mineralogical characteristics to the roof and floor materials. The presence of poorly ordered kaolinite, along with increased quartz, I/S and illite, indicates that they are epiclastic in origin. In hand specimen, the uppermost shale contains abundant sub-horizontal carbonate veins and vitrain laminae. Carbonates with a similar mode of occurrence have been reported by Ward et al. (1996, 1999a) at the very top and/or bottom of Permian coal seams, and may be a result of crystallisation under pressure following lateral fluid migration.

106

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

5.3.2 Minerals in Partings

Apart from three mineralised horizons (G-9, G-24 and G-33), three non-coal (claystone) partings (G-6, G-20 and G-22) occur in the Austar section. Most of these have a thickness between 1 and 2 cm, except for the uppermost parting, the thickness of which is 6 cm. XRD analysis shows that these claystones have similar mineralogical assemblages. However, the abundance and origin of the minerals vary among the different samples.

Sample G-20 is composed of kaolinite (52.7%) and I/S (34.4%), with minor proportions of quartz (2.5%), dolomite (6.5%) and goyazite (1.1%). If diagenetic dolomite and goyazite are excluded, the sample is entirely composed of clay minerals. Sample G-22 contains a similar mineral assemblage, except for more abundant quartz (8.6%) and less I/S (18.7%). Unlike those of the roof and floor sediments, the XRD patterns show that the kaolinite in both claystones is well-ordered. Both G-20 and G-22 display graupen to clastic textures under the microscope, with the aggregates probably consisting of cryptocrystalline to microcrystalline clay minerals (Figures 5.7A, B). SEM-EDS examination indicates that the clay pellets are mainly kaolinite. The kaolinite is probably diagenetic, and formed as an alteration product of the original volcanic glass, rock or mineral particles. Some cleat-filling kaolinite is also present, which represents the product of a later diagenetic process. Unlike some other intra-seam claystones (e.g. Zhao et al., 2012), samples G-20 and G-22 contain relatively abundant I/S. The I/S appears to occur commonly as laminae associated with the organic component (Figure 5.8A), and may be of epiclastic origin.

Quartz is a relatively minor component of parting G-20, only making up 2.5% of the mineral assemblage. The proportion of quartz in this parting is lower than that in the LTA of the adjacent coal plies. Although there appear to be some low-quartz coals in the upper part of the seam section, this is not the case if allowance is made for dilution by the abundant pyrite. The quartz in sample G-20 mainly occurs as subhedral crystals (Figures 5.8A, B), with a grain size up to 100 μm. These authigenic quartz gains are commonly confined within the clay minerals, rather than in the organic matter. The quartz may be late-stage, formed by epigenetic processes. Relatively higher proportions of quartz occur in the lower claystone, sample G-22 (8.6%). Some authigenic quartz was observed closely associated with authigenic goyazite (Figure 5.9) in the I/S matrix. It appears from the association and texture of the two phases that the goyazite predated the quartz. Such quartz may thus represent a later precipitation product.

107

Goyazite, which only occurs in one coal sample, occurs in the two middle non-coal partings (G-20 and G-22). Although referred to as goyazite, the Sr end member of the aluminophosphate mineral series, the XRD patterns do not match exactly with that mineral, or with the other end-members (Ba or Ca end member). It is most likely a solid solution of the end-member phases. Apart from Al and P, EDS data show that Sr is the dominant cation; Ba and in some cases, trace amounts of Ca also occur (Figures 5.8, 5.9), and hence it is more precisely identified as a barian goyazite.

Figure 5.7 Photomicrographs of non-coal partings. Plane polarised light (PPL). (A) Graupen to clastic textures in sample G20. (B) Graupen to clastic textures in sample G22. (C) Volcanic rock fragment with a weathering rim in sample G22. Feldspar phenocryst is noted. (D) Branching root structure in sample G6.

The goyazite in the claystones is variable in its habit. It largely occurs as subhedral to euhedral crystals incorporated in the kaolinite or I/S matrix (Figures 5.8, 5.9), indicating an authigenic origin. Discrete goyazite also occurs in cracks within the organic matter.

The goyazite in the claystone bands resembles that described by Triplehorn and Bohor (1983) in a kaolinised tuff from Colorado, where the goyazite probably formed concurrently with the kaolinite. As discussed in Chapter 2, goyazite, as well as gorceixite

108

Chapter 5 Mineralogy and Geochemistry of the Greta Seam and crandallite (other end members of the aluminophosphate series), may be common accessory authigenic minerals in tonsteins or altered tuffs. The Sr and Ba may have been partly retained from solution of the original volcanic ash (Triplehorn and Bohor, 1983; Rao and Walsh, 1997; Brownfield et al., 2005), while the P may have been derived from decomposition of plant material, as well as the volcanic debris (Ward et al., 1996; Rao and Walsh, 1997).

XRD analysis indicates that dolomite occurs only in small proportions. EDS data show that some of the dolomite contains a minor proportion of Fe. Figure 5.10A shows that dolomite infills cracks within the kaolinite, and thus appears to post-date the kaolinite. Figure 5.10B, on the other hand, shows that kaolinite may enclose some dolomite particles, which may indicate that the dolomite was formed before the kaolinite. Cleat or fracture-filling dolomite is not observed in the claystone samples.

XRD data also indicate the occurrence of trace proportions of anatase in the claystone bands G-20 and G-22. SEM-EDS examination shows the presence of Ti-rich inclusions in the kaolinite or I/S matrix (Figure 5.11). Based on the XRD data, these Ti-rich inclusions may be anatase crystallites, precipitated in place in cracks or pore spaces of the clay minerals. Some anatase shows a close genetic relationship with the clay minerals (Figure 5.11D). The goyazite was probably formed concurrently with the clay mineral material.

In contrast to the lower two partings, XRD analysis indicates that the uppermost claystone (G-6) consists of poorly-ordered kaolinite, I/S and more abundant quartz (17.1%). Roots are frequently observed penetrating the band (Figure 5.7D). The poorly ordered kaolinite, along with the abundant quartz, suggest that G-6 has characteristics similar to the epiclastic sediment of the roof/floor strata of the coal seam, although some altered volcanic rock fragments also occur (Figure 5.7C). Similar quartz and other detrital material in the thin partings (G-20 and G-22), if originally present, may have been more efficiently leached in the acid swamp environment than in the thick parting (G-6). The relatively higher pH in the original peat of the upper section, due to the percolation of more marine water, may also have decreased the leaching efficiency. The mineral assemblage in G-6 thus indicates that it formed from the deposition of epiclastic sediment, mostly as water- deposited detritus.

Other non-coal horizons in the seam profile are mineralised bands, mainly of authigenic origin. These consist of pyrite and marcasite in the top part of the seam, and siderite or

109 siderite and pyrite in the lower part. This compositional contrast may reflect different sulphate contents in the top and lower parts of the seam profile, respectively, during the early stages of diagenesis. Sample G-33 contains both pyrite/marcasite and siderite, which may have been precipitated at different stages of diagenesis.

Figure 5.8 SEM images of minerals in claystone sample G-20. (A) Euhedral quartz (Q) and goyazite; the laminae are I/S. (B) a rounded pellet of kaolinite, enclosed are subhedral to euhedral quartz (Q), dolomite (D) and goyazite (G). (C) Enlargement of (B) showing kaolinite aggregates (K), well-shaped goyazite (G) and dolomite (D).

110

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Figure 5.9 SEM images of minerals in claystone sample G-22. (A) Diagenetic quartz and goyazite in clay mineral matrix. (B) Enlargement of (A) showing the association of goyazite (G) and quartz (Q).

Figure 5.10 SEM images of minerals in claystone sample G-20. (A) Dolomite (D) infills the cracks within diagenetic kaolinite (K). (B) Kaolinite enclosing dolomite.

111

Figure 5.11 SEM images of anatase crystallites in claystones. (A) Ti-rich inclusions in cracks of kaolinite matrix. (B) Enlargement of image (A). (C) Ti-rich inclusions in cracks of I/S matrix. (D) Enlargement of image (C).

112

Table 5.3 Mineralogy of Austar LTAs and associated non-coal samples by XRD and Siroquant (wt. %). LTA percentages were not determined for the non-coal samples

Thickness LTA Sample Qtz Kao I/S I M Py Mar Cal Dol Ank Sid Daw (m) (%) G-1 0.02 43.9 11.6 19.2 14.9 6.4 0.4 2.9 G-2 0.07 28.5 19.6 44.4 5.1 1 G-3 0.06 44.9 15.2 13.6 3.1 20.9 1.3 G-4 0.01 36.4 17 43.3 1.7 G-5 0.13 9.8 11.1 19 17 39.7 1.4 1.7 1 G-6 0.08 17.1 37.4 27.9 9.5 4.6 0.8 G-7 0.5 12.3 6.9 28.2 51.9 1.9 1.8 0.6 G-8 0.04 18.8 0.6 12 59.8 8.1 0.7 G-9 0.015 63.5 4.3 29.9 34.8 G-10 0.355 17.3 1.4 19.8 47.9 G-11 0.25 23.8 12.2 26.3 5 45.7 3.7 0.2 0.7 0.7 G-12 0.14 14.1 19.2 24.8 46.9 0.9 5.5 G-13 0.35 26.6 3.8 17.6 67 G-14 0.4 9.3 14.5 54 12.9 3.2 3.7 0.2 4.3 G-15 0.02 25.3 13.1 13.8 5.3 41.5 G-16 0.57 39.9 25.2 10.9 13.9 8.9 0.7 G-17 0.14 39.7 16.2 13 20.3 8.9 1.7 G-18 0.07 41.6 53.2 13.2 6.9 9.4 0.6 2.1 13.9 G-19 0.55 13.1 12.6 57.6 18.9 3.2 0.8 0.2 0.7 5.7 G-20 0.02 2.5 52.7 34.4 6.5 G-21 0.55 14.4 5.7 64.3 13.5 6.1 0.5 0.5 0.9 2.8 G-22 0.015 8.6 67.2 18.7 1.7 1.9 G-23 0.415 10.6 11.8 43.6 21.4 0.5 11.9 10.8 G-24 0.01 8 6.9 1.8 0.7 82.6

113

Table 5.3 (Continued)

Sample Alb Ana Rut Gyz Fap Gp Anh Bas Jar Coq Cop Sz Vo Ro

G-1 0.7 G-2 0.4 1.1 G-3 0.9 G-4 1 0.5 G-5 3.9 0.3 0.5 3.4 0.9 G-6 1 0.2 0.6 1 G-7 3.3 0.6 1.3 3.4 G-8 4.3 0.9 1.1 2.9 7.2 2.3 G-9 2.7 23.3 4.9 G-10 4.6 2.1 3.6 12.1 8.6 G-11 0.8 4.6 G-12 0.8 1.9 G-13 0.9 0.8 1.8 5 3 G-14 0.4 0.2 1.1 4.6 0.8 G-15 0.6 0.4 G-16 0.6 G-17 0.2 G-18 0.7 G-19 0.1 0.4 G-20 1.6 1.2 1.1 G-21 3.6 1.4 0.6 G-22 0.5 0.7 0.7 G-23 G-24

114

Table 5.3 (Continued)

Thickness LTA Sample Qtz Kao I/S I M Py Mar Sp Cal Dol Ank Sid Daw Alb Ana Rut Bas Jar (m) (%) G-25 0.04 14.1 11.6 29.3 37.2 3.8 0.6 13.7 3.9 G-26 0.01 45.7 14.5 41.8 32.7 0.3 8.5 1.1 0.6 0.6 G-27 0.4 6.6 7.2 65.9 8.2 2.5 0.7 3.1 9.7 2.6 G-28 0.06 36.7 23.1 48.8 23.1 0.2 1.2 0.6 2.5 0.5 G-29 0.14 9.5 21.6 42.4 6 10.5 18.3 1.1 G-30 0.26 5.6 9 72.3 5.5 0.9 0.6 0.8 8 3 G-31 0.04 48.8 8.8 79.5 5.3 2.9 0.1 0.4 0.5 1.6 1 G-32 0.41 9.1 17.3 57.9 7.1 0.7 0.3 0.3 14.8 1.5 0.2 G-33 0.005 1.4 5.4 22.7 29.9 39 1.5 G-34 0.155 11.5 22.4 52.3 18.1 5.2 1.5 0.5 G-35 0.23 6.5 11.9 58.6 14.9 2 6.5 0.5 0.6 2.3 2.7 G-36 0.55 4 10.8 63.2 14.9 2.6 0.5 1.3 2.6 2.3 0.1 0.3 1.3 G-37 0.44 6.9 7 66.3 17 3.1 0.5 2.4 3.7 0.2 G-38 0.69 4.8 15 63.9 14.6 1.1 0.6 1.6 2.7 0.3 G-39 0.31 4.7 14.3 59.9 14.6 2.3 0.9 0.5 0.9 3.9 2.4 0.1 G-40 0.02 57.6 45.4 14.7 36.9 1.9 1.2 G-41 0.23 14.3 22.9 48.9 7.4 6.5 0.4 4.1 9.7 G-42 0.12 40.5 7.4 12.9 31.5 4.4 2.1 0.5 0.7 G-43 0.05 30.6 8.4 35 25.1 0.4 0.5 G-44 0.17 32.3 3.8 35.4 27.2 0.4 0.5 0.5 G-45 0.03 48.2 3.1 24.9 21.7 1 0.5 0.6 Qtz = quartz; Kao = kaolinite; I = illite; I/S = mixed-layer illite/smectite; M = muscovite; Py = pyrite; Mar = marcasite; Sp = sphalerite; Cal = calcite; Dol = dolomite; Ank = ankerite; Sid = siderite; Daw = dawsonite; Alb = albite; Ana= anatase; Rut= rutile; Gyz = goyazite; Fap = fluorapatite; Gp = gypsum; Anh = anhydite; Bas = bassanite; Jar = jarosite; Coq = coquimbite; Cop = copiapite; Sz = szomolnokite; Vo = voltaite; Ro = rozenite.

115

Table 5.4 Mineralogy of Cessnock LTAs and associated non-coal samples by XRD and Siroquant (wt. %)

2092 2094 2095 2097 2098 2099 2100 2101 2102 LTA 10.2 11.8 27.9 12.5 15.9 4.8 5.5 4.5 9.3 Quartz 1.6 3.2 5.8 0.9 8.3 1 3.9 4 17.6 K 24.3 31 46.7 31.8 25.3 38.2 43.7 56.1 44.9 I/S 20 22.9 13 21.2 18 6 I 3.7 4.3 9.4 7.6 3.4 8.5 11 13.6 Py 2.5 1.9 3.1 2.1 1.4 2.6 Cal 2.4 2.5 Dol 2.2 2.2 Ank 4.9 3.7 4.9 1.8 Sid 2.4 0.6 Daw 1.9 4.5 1.9 4.3 Alb 7.1 13.7 9.3 7.6 6.6 4.7 Ana 1.7 2.7 2.4 2.3 0.8 1.2 Rut Bas 3.9 4.4 2 2.4 5.1 2.9 5.3 2.5 (Na-)jar 4.8 6.1 7 3.7 1.4 2.5 3.2 2.6 Coq 7.1 8.1 12.5 21.6 8.1 2 2 3.2 2.7 (Fe-)cop 9.6 7.9 9.8 6.5 8.8 2.6 1.5 Sz 6.6 6 4.2 16.4 13.5 4.7 2.9 4.7 2.4 (Na-)jarosite = jarosite or natrojarosite; (Fe-)cop = copiapite or ferricopiapite; Other abbreviations same as Table 5.3.

5.3.3 Coal mineralogy

As indicated in the mineralogical profiles (Figure 5.5), the mineralogy of the coal LTA residues is distinctly different in the upper and lower splits of the Greta seam. The marine- influenced upper split contains abundant pyrite, clay minerals (kaolinite, illite and I/S), and minor carbonate minerals (calcite, dolomite, ankerite, siderite and dawsonite). Clay minerals are less abundant in the upper section than in the lower section, apparently due to dilution arising from the greater abundance of pyrite in the former. Small amounts of hydrous iron sulphates are also present in a few pyrite-rich samples, due to oxidation of pyrite during sample storage.

The proportions of minerals were also recalculated and normalized after the pyrite was excluded from the mineral assemblage in each sample, and the results were expressed as graphic plots (Figure 5.12). This was intended to investigate more directly the distribution patterns of the clay minerals and other carbonate minerals within the coal seam, without dilution from the pyrite, and the mechanisms of mineral matter formation by other processes at the different stages of coal formation.

116

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

The normalised mineralogical profiles on a pyrite-free basis (Figure 5.12) indicate that the distribution patterns of quartz and kaolinite are roughly uniform in the coal samples. Decreased proportions of kaolinite, coupled with increased proportions of quartz and non- clay minerals, occur at the detrital sediment horizons. Dawsonite occurs in negligible proportions in the upper section, but is relatively abundant in the lower section. Minor dolomite and ankerite occur throughout the seam, but the proportions are almost negligible in the lower section, even on a pyrite-free basis. Anatase appears to occur in more significant amounts in the coals of the upper section. Siderite only occurs in a couple of coal samples.

Figure 5.12 Plots showing vertical variation of abundance of major minerals (normalised to pyrite-free) in the Greta seam (Austar).

5.3.3.1 Clay minerals

The principal clay minerals in the LTA of the Greta coal samples are kaolinite, I/S, and illite. Under the optical microscope, the clay minerals occur as thin bands, laminae and aggregates (Figure 5.13A) within the vitrinite macerals, cell infillings in fusinite and semifusinite, and, in a minor case, as the sole mineral or along with other minerals, infilling cleats cutting mainly through telocollinite (Figure 5.13B). SEM-EDS studies, together with XRD analysis, allow a more precise identification and estimation of the different clay minerals, as well as their relation to other mineral species.

117

Kaolinite Kaolinite is the dominant clay mineral in the Greta coals. If allowance is made for the occurrence of pyrite, the normalised percentage of kaolinite shows little variation within the LTA of the coal itself. The proportion of kaolinite, however, is typically reduced in coals near horizons with abundant epiclastic sediment (roof, floor, and epiclastic partings), where the abundance of non-kaolinite clay minerals and quartz is generally increased.

The XRD patterns of the coal LTA residues show a well-ordered structure of the kaolinite in all the coals, except for the coal plies (G-40 and G-41) near the seam floor. Well- ordered kaolinite in coal appears to be the result of in situ leaching and reprecipitation processes within the peat swamp (Ward, 1989). Optical microscope and SEM studies indicate that the kaolinite mainly occurs as infillings of cell and pore cavities in the inertinite macerals. The kaolinite infillings are commonly aggregates of kaolinite booklets (Figure 5.14). Cleat kaolinite is rare. This suggests that kaolinite in the Greta coals is primarily a syngenetic precipitate, formed during the early stages of coal formation.

Figure 5.13 Photomicrographs showing clay minerals in the Greta coals (Austar). (A) Aggregates of clay minerals in the collodetrinite matrix. In air, reflected light. (B) Clay (dark) and carbonate minerals in a cleat. Oil immersion, reflected light.

The LTAs of the Greta coals appear to be not as rich in kaolinite as those of many other Australian bituminous coals, such as those studied by Ward (1978; 1989). Ward (1989) noted that the LTA of the vitrinite-rich, marine-influenced coals of the Illinois Basin generally contain less abundant kaolinite than that of some inertinite-rich, fluvial coals of the Sydney Basin. He suggested that the cause may be either atmospheric oxidation of the peat in the Sydney Basin coals, which was more favourable for kaolinite development, or the influence of marine water in the Illinois Basin coals, which was less favourable.

118

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Marine water may provide less favourable conditions for kaolinite formation, due to higher concentrations of relevant ions for precipitation of non-kaolinite clay minerals.

Figure 5.14 SEM images of kaolinite in the Greta coals. (A) Kaolinite (grey) and pyrite (bright) occurring as cell infillings, G-18. (B) Aggregates of kaolinite booklets in cell cavities. Bright area is pyrite (P), G-23.

Illite and Na-rich I/S As described above, illite appears to only occur in the coals near the roof, floor and the detrital horizons, and is almost absent in the coals away from those parts (Figure 5.12). Illite in the Greta coals is thought to be essentially of detrital origin, and, when present, may reflect the greater amount of clastic influx in that particular part of the original peat swamp.

As suggested by Ward (1989) and Ward and Christie (1994), I/S generally shows a similar distribution pattern in many Australian bituminous coal seams. In those coal seams, I/S and illite largely disappear in most coals, and well-ordered kaolinite is essentially the sole constituent in the clay fraction. The distribution pattern of I/S in the Greta coals, however, is rather uniform (Figure 5.12).

SEM analysis indicates the relatively common presence of a Na-rich clay mineral, the EDS spectrum of which shows the presence of Na and in some cases, K, as well as Al and Si, with Al peaks being similar to or a bit higher than those of Si (Figure 5.15). The dominance of Na over K and the slightly higher concentration of Si than Al (Table 5.5) indicate that such material is probably a Na-rich I/S. The possibility of paragonite (Na-rich mica) may be excluded, as XRD analysis indicates that coals contain rare illite and not paragonite in the coal samples.

119

As indicated by SEM-EDS studies, this Na-rich I/S is widely distributed in the Greta coal seam. It mainly occurs as microcrystalline infillings of cell cavities (Figure 5.15A, B), pore spaces or cracks (Figure 5.15C) within the macerals, which indicates a clearly authigenic, rather than detrital origin.

Table 5.5 EDS micro analyses of authigenic I/S in the Greta coals (data from 21 points)

Al (%) Si (%) Na (%) K (%) Al/Si max min mean max min mean max min mean max min mean max min mean

21.8 15.3 17.8 27 18.9 23.4 3.9 0.4 1.7 1.9 0 0.5 0.7 0.8 0.8

Diagenetic I/S or paragonite is commonly present in non-coal rock or high-rank coals that have been thermally metamorphosed. Kisch (1966) described a chlorite-illite tonstein associated with an Australian semi-anthracite. He ascribed the formation of illite and chlorite to the alteration of kaolinite during diagenesis, under conditions with available K (for illite), and Fe and Mg (for chlorite). A rectorite-like mineral has been described by Susilawati and Ward (2006) in claystones associated with coal seams at Bukit Asam, Indonesia, with a vitrinite reflectance ൒1.3%, but is absent from lower rank parts of the Bukit Asam deposit.

However, there seem to be no indication of any heat sources associated with the Greta coals. The Na-rich I/S may have been syngenetically precipitated, probably after the peat was accumulated, with abundant Na and relatively minor K ions being supplied by the marine water. This is different from the situation in a more normal acid swamp environment, where the formation of kaolinite as the only syngenetic clay mineral is favoured, due in part to low cation availability.

5.3.3.2 Quartz

Quartz makes up less than 15% in the majority of the LTA residues of the Greta coals (Table 5.3). High quartz proportions, coupled with poorly ordered kaolinite and relatively abundant non-kaolinite clay minerals, usually occur in the coals near the epiclastic sediments, and in several small coal bands within the seam. In such coals, quartz and other sediments represent essentially detrital material, and are the major contributors to the high LTA percentages of the coals in question.

Detrital quartz in the Greta coals commonly occurs as relatively large grains (Figure 5.16). Some detrital quartz fragments appear to have been leached in the peat swamp (Figure

120

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

5.16B). Authigenic quartz also occurs, typically as infillings of cell cavities or pore spaces (Figure 5.16B). Such quartz probably represents reprecipitation products of biogenic (e.g. siliceous sponge spicules and diatoms) or detrital material, formed during early diagenesis (Ward, 2002). Late stage quartz in cleats or fractures was not observed in the studied coals. Authigenic quartz is probably the dominant form of quartz in the coals away from the epiclastic horizons.

Figure 5.15 SEM images of sodium-bearing illite or I/S in the Greta coals. (A) I/S in cell cavities, G-21. (B) I/S in cell cavities, G-11. (C) I/S or paragonite in pore space or crack of coal maceral, G-21.

121

Figure 5.16 SEM images of quartz in the Greta coals. (A) Detrital quartz, G-18. (B) A detrital quartz (Q) grain in the centre of the image that appears to be leached. More authigenic quartz (Q) occurs in cell cavities or pore spaces in the lower right of the image, G-18.

5.3.3.3 Pyrite/marcasite

Pyrite typically makes up from 40 to 56% of the LTA in coals from the upper part of the seam section, with the exception of one sample immediately above the clastic horizon separating the two splits. The minerals in the lower part of the seam, however, contain much less pyrite (less than 6.5%) in the LTA residues. Marcasite occurs occasionally as a minor phase in the coal samples.

Pyrite in the coal samples, particularly in the high-sulphur upper split, typically occurs as euhedral crystals, framboids (Figure 5.17A), pore infillings (Figure 5.17B), and a massive form with a porous structure. All of these forms indicate that the pyrite is largely a syngenetic precipitate, formed during peat deposition or shortly after peat accumulation. Pyrite in some cases occurs as a replacement of wood structures (Figure 5.17C). Both forms of pyrite were also probably formed in the early stages of diagenesis. Earlier-formed pyrite enclosed by syngenetic siderite (Figure 5.17D) is also relatively common. Late stage cleat-filling pyrite, however, is rare in the Greta coals.

The formation of syngenetic pyrite requires the availability of dissolved ferrous iron, and also of H2S from bacterial reduction of sulphate in the peat swamp. The sulphate may be supplied by the swamp water, or be introduced from waters that penetrate into the peat later after the peat accumulation (Diessel, 1992; Ward, 2002). Thus, syngenetic pyrite in coal is often regarded as an indicator of marine influence during coal formation. The Greta coal seam, as the lower coal seam in the Greta Coal Measures, is thought to have been influenced by the marine water after the peat was buried.

122

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Syngenetic pyrite, however, may also be present in non-marine coals (e.g. Chagué-Goff et al., 1996). In this case, sulphate-rich surface or ground water, which seeped through the peat, may be involved in the pyrite formation (Ward, 2002). Pyrite and, to a lesser extent, marcasite, are concentrated in the upper part of the seam in the present study; this suggests that marine water penetrated to a certain depth into the upper part of the coal bed. Trace amounts of pyrite in the lower part of the section are probably due to the reduced availability of sulphate ions.

Figure 5.17 Photomicrographs showing modes of mineral occurrence in the Greta coals. (A) Pyrite framboids and euhedral crystals, G-7, oil immersion (B) Pyrite in cell cavities of inertinite, G-37, oil immersion. (C) Pyrite replacement of wood structure, G-7, oil immersion. (D) Pyrite replacement of syngenetic siderite, G-28, in air.

5.3.3.4 Carbonates

With the exception of dawsonite, carbonate minerals are minor components in the LTA residues of the Greta coals, although they may be very abundant in some of the non-coal horizons.

123

Siderite Siderite occurs mainly in the lower section of the seam, but is almost absent in the upper section. Optical and electron microscope studies indicate that the siderite in the Greta coals commonly occurs as concretionary nodules. The surrounding inflected bedding indicates that such siderite is clearly a syngenetic product, formed during or shortly after peat deposition.

Syngenetic siderite formation represents the interaction of iron and dissolved CO2, and it may only be abundant when the activity of reduced sulphur species is too low for pyrite production (Ward, 1984; Spears, 1987). The absence of siderite in the upper section of the Greta seam is probably due to increased competition for iron from sulphide species. As noted above, however, syngenetic siderite sometimes encloses pyrite (Figure 5.17D), a process that may be the result of environmental changes, such as changes from weakly alkaline to weakly acid conditions (Kortenski and Kostova, 1996).

Calcite, ankerite and dolomite These carbonate minerals are distributed throughout most of the seam, but in minor proportions and without any preferred pattern. All of these carbonates have similar modes of occurrence, occurring largely in cracks and cell or pore cavities within the macerals. EDS analysis, however, indicates that they may show chemical variation. For example, calcite sometimes contains detectable Sr by EDS analysis (Figure 5.18A), and dolomite and ankerite may have various concentrations of Fe (Figure 5.18B) and Mn respectively.

Cell filling carbonates may indicate either a syngenetic or an epigenetic origin. Cleat systems are not very well-developed, especially in the upper section of the Greta seam, and these carbonates were rarely observed as late-stage cleat fillings. However, later- formed carbonates in cracks were probably precipitated from epigenetic fluids.

124

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Figure 5.18 SEM images showing carbonate minerals in the Greta coals. (A) Sr-bearing calcite (C), along with I/S, in cracks of organic matter, G-5. (B) Ferroan dolomite (D) and quartz (Q) in cell cavities, G-29.

Dawsonite As mentioned above, dawsonite is abundant in the lower section of the seam, but is negligible in the upper seam section. In hand specimen, dawsonite occurs in the Greta coals as abundant visible cleat and fracture infillings. The dawsonite veins commonly cut through coals of different megascopic lithotypes. However, dawsonite was not frequently observed in the coal samples under optical and electron microscopes. Only Al, O and Na were shown in the EDS spectra. Dawsonite in the Greta coal samples occurs mostly as fracture and cleat infillings (Figures 5.19A). It also occurs as thin bands, within inertinite macerals (Figure 5.19D). In some cases, dawsonite shows intergrowth with kaolinite in cleats (Figures 5.19B, C). Such forms of dawsonite indicate that it was precipitated epigenetically from fluid at a relatively late stage.

Although not common in other coal seams around the world, dawsonite is relatively widespread in the Permian non-coal (Goldbery and Loughnan, 1977; Baker et al., 1995) and coal-bearing strata (Golab and Carr, 2004; Golab et al., 2006) of the Sydney Basin. It

125 is mostly an authigenic mineral occurring as veins or void infillings (Loughnan and See, 1967; Loughnan and Goldbery, 1972). Dai et al. (2008b) noted dawsonite as a replacement of pyroclastic sanidine in the Yanshan coals, SW China, and ascribed the dawsonite to the reaction of sanidine with Na-rich fluid, Al and carbonate ions.

Aluminosilicates such as clay minerals and feldspars may serve as sources of Al and/or

Na (Baker et al., 1995). Baker et al. (1995) suggested a magmatic CO2 source in the Bowen-Gunnedah-Sydney Basin system for the widespread formation of dawsonite in the sedimentary rock strata. Golab et al. (2006) used isotopic data to investigate cleat dawsonite in the coals from the Late Permian Wittingham Coal Measures, Sydney Basin, and suggested that the dawsonite was the result of reactions between Na2CO3- or

NaHCO3-bearing fluids and earlier-precipitated kaolinite in the cleat. Golab et al. (2006) also suggested a magmatic carbon source for the dawsonite. A similar mechanism may have been involved in the formation of dawsonite in the Greta seam. The intergrowth texture of kaolinite and dawsonite in the cleat may suggest that the kaolinite was probably a precursor of the dawsonite. Na released from organic matter may not have supplied enough material, as the coal had not been subjected to substantial rank advance.

Aluminium hydroxides, such as nordstrandite (Al(OH)3), co-exist with dawsonte in some coal or non-coal strata (Goldbery and Loughnan, 1970; Loughnan and Goldbery, 1972; Goldbery and Loughnan, 1977; Baker et al., 1995). Nordstrandite, which can be formed from the loss of soda of dawsonite, was not detected in the Greta coals, however, by XRD analysis.

126

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Figure 5.19 SEM images showing the modes of occurrence of dawsonite in the Greta coal samples. (A) Dawsonite-filled cleats, thin section under PPL (plane polarized light), G-26. (B) Kaolinite and dawsonite co- precipitates in a cleat, G-5. (C) Enlargement of image (A), showing intergrowth of dawsonite crystals (D) in kaolinite (K). (D) Layer of dawsonite (D) in crushed cell cavities. Bright area is ankerite (A) and all the other cell infillings are Na-illite, G-21.

127

5.3.3.5 Albite

XRD analysis indicates that albite occurs in minor but significant proportions in a few of the Greta coals. The occurrence of albite was also confirmed by the EDS spectra of this material, which shows the presence of an enhanced peak of Na, and obviously higher peak of Si relative to Al (Figure 5.20). Albite in the Greta coals was observed as infillings of cell cavities (Figure 5.20B). In some cases the cells with albite fillings appear be crushed (Figure 5.20A). This may suggest that the albite was precipitated from hydrothermal fluids after the cell wall was crushed later in the coal’s burial history, rather than in the syngenetic stage.

Figure 5.20 SEM images of albite in coal sample G-5. (A) Albite (A) filling crushed cell cavities. (B) Cell-filling albite and quartz.

Authigenic albite is not common in coal. When present, albite usually occurs as an epiclastic mineral (e.g. Vassilev and Vassileva, 1996). Albite produced during thermal metamorphism of clays was noted by Golab and Carr (2004) in a dyke in a bituminous coal from the Sydney Basin, Australia. Albite replacement of sanidine in a volcanic- influenced high organic sulphur coal from the Yanshan Coalfield, China, was described by

128

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Dai et al. (2008b), and the albite in that case appears to have been derived from hydrothermal fluids.

The coexistence of dawsonite and albite in the Greta coals, both of which were formed epigenetically at a late stage of diagenesis, may indicate that the albite was derived from the same Na-rich hydrothermal fluid as the dawsonite, and precipitated in cavities or cracks of macerals. Active Al and Si ions may have been released from dissolution of detrital minerals (e.g. quartz and feldspars) or biogenic fossils (e.g. siliceous sponge spicules and diatoms). The formation of dawsonite, as noted above, may require a kaolinite precursor.

5.3.3.6 Phosphates

XRD analysis indicates that fluorapatite and goyazite only occur in a few of the Greta coal samples, with proportions of <5% and around 1% in the LTA residues, respectively. They may sometimes coexist in the same sample. SEM studies indicate the occurrence of pore- filling goyazite in macerals, with a grain size <1 μm, commonly containing EDS detectable Ca and S (Figure 5.21).

The phosphorous for formation of phosphate minerals in coal may be derived from the decomposition of plant material (Ward et al., 1996; Rao and Walsh, 1997). A high level of phosphorus in particular beds, however, may also be controlled by local hydrogeochemical factors or by introduction of additional sources of phosphorus, e.g. volcanic debris, shells or faecal matter (Ward et al., 1996; Rao and Walsh, 1999). The cations Ca, Ba and Sr within the aluminophosphate minerals are attributed to the alteration of volcanic ash in the coal-forming mires (Rao and Walsh, 1997; Brownfield et al., 2005).

The goyazite present in the claystones described above, however, does not contain any EDS detectable sulphur, and has Ca instead of Ba. The sulphur may be incorporated in the goyazite structure as a substitute for P in the goyazite lattice, during early diagenesis of coal. The source of this sulphur may be the original plant material, or, more likely, the sulphate ions in the marine water.

As discussed by Ward et al. (1996), whether formation of apatite or aluminophosphates took place was probably controlled by local hydrogeochemical factors and the availability of alumina. Aluminophosphate is precipitated under neutral to alkaline conditions, and 129 aluminophosphate minerals are less soluble at acid pH levels. Aluminophosphate minerals would be expected when Al was also available in reactive form at the site of phosphate deposition, and apatite if Al was not available to react with the precipitated phosphatic material. Although fluorapatite was not observed under the optical microscope or SEM, an authigenic origin is likely, as it occurs at similar horizons to goyazite, and may be precipitated from same phosphorus source in the early diagenesis stage.

Figure 5.21 SEM images of goyazite (G) occurring as infillings of pore space in coal sample G-21.

5.3.3.7 Anatase/rutile

Although of minor occurrence, both anatase and rutile are persistently present in almost all the coal and associated non-coal samples. The occurrence of these minerals in the coal was not observed under the optical microscope or SEM, due to their trace proportions and the low LTA percentages of the coals. As noted below, a significant positive correlation exists between TiO2 and Al2O3 in the Greta coal samples. This may indicate that the Ti-bearing minerals are associated with clay minerals, as observed in the claystone partings. Thus at least some anatase/rutile may be authigenic, as most clay minerals in the coal samples are of authigenic origin.

5.3.3.8 Sulphates

Small amounts of hydrous iron sulphates, including jarosite, coquimbite, copiapite and szomolnokite, are present in a few of the pyrite-rich samples, due to oxidation of pyrite during sample storage. Apart from this series of minerals, the Greta coals from the Austar section, which have been stored for a shorter period of time, also contain voltaite

(K2Fe5Fe4(SO4)12·18H2O), and rozenite (FeSO4·4H2O).

130

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Another range of sulphate minerals, such as bassanite, anhydrite and gypsum, is also present in the Greta coals, especially in the upper part of either or both sections. Such sulphates are usually formed as artefacts, produced by interaction of inorganic elements and organic sulphur during the low-temperature ashing process (Frazer and Belcher, 1973). The presence of organically associated Al, Ca and S in the Greta coal has been confirmed in electron microprobe studies by Ward et al. (2007). Gypsum may be present in lower-rank coal (Koukouzas et al., 2010; Dai et al., 2012d), either as an authigenic mineral in the coal or by precipitation of Ca and SO4 in the pore waters during drying. It may also be produced by the reaction between calcite and the sulphuric acid which is produced by oxidation of pyrite with sample exposure and storage (Pearson and Kwong, 1979). Partial dehydration of gypsum during sample drying or low-temperature ashing may also result in the formation of bassanite (Ward, 2002).

The presence of bassanite, anhydrite and gypsum mainly in the pyrite-rich, upper section of the Greta seam, may also suggest a relatively high organic sulphur level in the upper coals to that in the lower coals. Ward et al. (2007) found that the organic sulphur content of vitrinite in the Gretra seam increased from approximately 1% in the lower part of the seam to around 2.5% in the upper part of the seam section, based on electron microprobe studies. A correlation between organic and pyritic sulphur may exist in the Greta coals, although relevant data are not available for the sections examined in the present study.

5.4 Geochemistry of the Greta seam

As noted in Chapter 4, a number of the Austar samples covering thin intervals that were adjacent to each other and had similar mineralogical characteristics were combined, reducing the original 45 samples to 23 samples, in order to provide enough material for a more detailed geochemical study. Major element chemical data for the high-temperature (815 qC) ashes of these samples are given in Table 5.6, and those of the 815 qC ashes for the samples in the Cessnock section are given in Table 5.7. In both cases the results have been expressed to an LOI- and SO3-free basis, to allow better comparison to the mineralogical data.

131

5.4.1 Mineralogical and chemical analysis data

The relation between ash chemistry and mineralogy for the Greta seam samples from both sites was studied to check the reliability of the quantitative XRD data, following the procedure described by Ward et al. (1999). The inferred chemical composition of the mineral assemblages determined from XRD analysis and Siroquant was calculated and compared with the actual chemical composition of the same samples as determined by XRF analysis (Tables 5.6 and 5.7). The inferred chemical composition was adjusted for + each LTA sample by deducting the CO2 and H2O to derive an equivalent to a coal ash analysis. The actual chemical composition determined by XRF was also normalized to an - SO3 free basis, to allow for differences in SO3 retention by the low- and high-temperature ashing processes.

Table 5.6 Major element analyses of Greta coal ash and non-coal samples from the Austar Coal Mine (%).

Sample HTA SiO2 Al2O3 TiO2 Fe2O3 MgO CaO Na2O K2O MnO P2O5 SO3 LOI G-r (1,2,4) 53.87 13.46 0.85 2.93 0.26 0.08 0.41 1.91 0.02 0.06 3.02 23.13 G-3-r 61.60 8.95 0.41 9.55 0.16 0.03 0.37 1.21 0.01 0.04 7.82 9.85 G-5 9.78 35.81 17.82 1.18 31.44 1.29 4.40 1.06 0.62 0.11 0.34 4.52 1.40 G-6-p 43.12 19.65 1.00 6.52 1.67 2.28 0.36 1.60 0.08 0.23 5.50 17.98 G-7 12.32 24.53 14.14 0.88 51.11 1.05 2.95 0.40 0.28 0.12 0.70 2.95 0.89 G-10 17.75 12.48 9.77 0.79 69.97 0.75 2.42 0.25 0.19 0.08 1.06 1.85 0.41 G-11-13 23.29 21.28 11.28 0.80 60.65 0.42 2.18 0.18 0.30 0.08 1.09 1.55 0.19 G-14 9.29 51.31 29.81 1.63 4.91 1.09 4.37 0.31 0.93 0.02 2.33 2.04 1.23 G-16-17-f 70.88 17.64 0.86 1.18 0.71 0.59 0.32 2.73 0.01 0.06 0.77 4.26 G-18 41.62 70.86 18.36 0.94 1.09 0.59 0.77 3.76 1.06 0.01 0.06 2.77 -0.30 G-19 13.13 52.89 33.49 2.20 1.18 0.50 0.66 2.17 1.37 0.01 0.42 2.21 3.07 G-21 14.39 52.25 39.32 2.33 0.87 0.34 0.51 1.12 1.02 0.01 0.20 0.92 1.10 G-23 10.64 45.43 31.11 1.19 13.73 0.97 0.81 1.75 1.10 0.39 0.26 1.56 1.70 G-27-29 10.27 52.04 32.81 2.85 5.47 0.33 0.36 2.37 1.00 0.05 0.14 0.94 1.64 G-30-32 10.02 52.57 34.53 2.59 3.80 0.43 0.43 2.14 0.81 0.04 0.10 1.06 1.51 G-34 11.47 60.12 30.70 1.95 2.75 0.44 0.36 0.30 1.19 0.06 0.06 0.47 1.60 G-35 6.47 49.89 31.72 1.56 11.40 1.19 0.54 0.60 0.65 0.03 0.08 0.64 1.68 G-36 4.05 51.18 34.93 2.08 3.79 0.74 2.07 1.17 0.41 0.03 0.82 1.63 1.14 G-37 6.91 51.44 36.66 1.75 3.07 0.61 0.95 1.25 0.92 0.04 0.10 1.54 1.66 G-38 4.8 55.91 35.07 2.28 2.27 0.64 0.94 1.39 0.77 0.03 0.09 1.02 -0.41 G-39 4.72 55.01 35.63 1.79 2.32 0.47 0.56 1.35 0.78 0.02 0.09 0.77 1.18 G-41 14.35 52.27 29.40 1.28 2.20 1.64 2.80 3.31 1.23 0.02 0.06 4.29 1.50 G-f (42-45) 64.00 18.42 0.89 2.82 0.71 0.11 0.35 3.81 0.04 0.05 0.63 8.18

132

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Table 5.7 Major element analyses of Greta coal ash and non-coal samples from the Cessnock Coal Mine (%).

Sample HTA SiO2 Al2O3 TiO2 Fe2O3 MgO CaO Na2O K2O MnO P2O5 SO3 LOI 2092 5.66 24.41 18.52 1.18 26.49 1.87 7.54 1.30 0.37 0.121 0.28 10.95 15.92 2094 8.11 28.29 23.41 1.50 26.86 1.81 4.91 1.44 0.44 0.092 2.02 5.92 7.98 2095 17.71 43.12 26.41 3.49 18.60 0.64 1.58 0.56 0.84 0.027 0.16 2.33 4.18 2097 6.36 18.49 17.60 0.86 43.32 1.38 3.29 1.69 0.22 0.067 0.72 7.29 10.19 2098 8.25 25.54 20.29 1.18 38.27 1.38 2.99 1.46 0.34 0.075 0.88 4.45 6.14 2099 3.15 30.15 30.53 1.83 7.35 3.25 7.54 3.52 0.42 0.078 2.14 8.54 9.52 2100 4.19 36.32 31.87 2.10 6.94 1.74 4.58 2.58 0.37 0.054 1.41 5.93 8.72 2102 6.80 51.84 28.20 1.66 7.22 0.94 1.96 1.08 0.31 0.026 0.24 2.24 3.78 Data were not obtained for sample 2101 due to the sample unavailability.

The percentages of each element indicated by both sets of data were plotted against each other (Figures 5.22 and 5.23), to provide a basis for comparing the XRD results to the chemical analysis data for the same coal or parting samples. As discussed for other materials by Ward et al. (1999a), the respective data sets are presented as X-Y plots, with a diagonal line on each plot indicating where the points would fall if the estimates from the two different techniques were equal.

SiO2 and Al2O3

The plots for the bulk of the oxides in the ash, SiO2 and Al2O3, of the Austar section, show that all points fall very close to the relative diagonal equality lines (Figure 5.22). This indicates that the Siroquant results for the major minerals (quartz, clay minerals and dawsonite) in the coal and non-coal samples are quite consistent with the actual ash chemistry determined by the XRF analysis.

The proportions of SiO2 and Al2O3 in most samples from the Cessnock section (Figure 5.23), however, appear to be over-estimated by Siroquant. This is accompanied by considerable under-estimation of Fe2O3 in those samples. Although attempts were made, not all the secondary iron sulphate minerals, the oxidation products of pyrite produced during sample storage, could be identified in the LTAs of the sulphur-rich coals. Further resolution of these minerals, however, is beyond the scope of the present study. The difficulties in identification of these minerals has resulted in determination of higher proportions of aluminosilicate minerals by Siroquant, and consequently higher proportions of SiO2 and Al2O3 (especially the former) inferred from the Siroquant data.

Fe2O3, CaO and MgO

Like SiO2 and Al2O3, the plot for Fe2O3 in the Austar section also shows a high degree of correlation between the Fe2O3 proportions from the Siroquant data and those indicated by

133 the ash chemistry (Figure 5.22). However, the majority of the points for Fe2O3 fall slightly below the equality line, indicating an under-estimation of Fe2O3 inferred from Siroquant, relative to the actual values. This slight difference may reflect substitution of Fe for Mg in the dolomite, or Fe for Al in the illite lattice, neither of which was allowed for in the stoichiometric calculations. In addition to these reasons, the under-estimation of Fe2O3 in the Cessnock section (Figure 5.23) is mainly due to identification difficulties for the secondary iron sulphates as noted above.

Both CaO and MgO data from two methods show a scattered correlation in both sections. MgO in the majority of samples appears to be under-estimated by the interpretation of the Siroquant data. Most of the coal samples contain small but measurable proportions of calcite, and dolomite or ankerite. Such minerals explain the bulk of CaO and MgO contents in the coals, with the exception of one sample (G-10) which contains gypsum and anhydrite.

As mentioned in Chapter 2, Ca and Mg may also be incorporated in the organic components of coals, particularly in lower rank coals, as non-mineral inorganic matter. Non-mineral inorganic Ca has been identified by electron microprobe in some Greta coals (Ward et al., 2007). The presence of bassanite in quite a few coal samples also suggests the occurrence of organically-associated Ca. Such organically associated elements, if not transformed to crystalline phases, may still be retained in non-crystalline form. Although not detected by XRD, these would nevertheless contribute to the ash chemistry of the coal. Ca also occurs in goyazite in various concentrations, which was also not allowed for in the calculations.

K2O and Na2O

Comparison of K2O data from the Austar section shows an overall good level of agreement, especially the data for those non-coal samples with abundant K-bearing minerals (illite and mixed-layer I/S). The sample with the lowest chemically determined

K2O values, however, plots the farthest from the equality line. This sample (G-10) is a highly pyritic coal, with the pyrite being partly oxidized. The K2O inferred from the Siroquant data is mainly contributed by voltaite, an oxidation product of pyrite. Siroquant may have over-estimated the proportion of voltaite in this case, accompanied by an underestimation of other secondary Fe-sulphates.

134

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

The proportion of Na2O inferred from the Siroquant data also shows a somewhat scattered correlation with the proportion by chemical analysis. Some samples have lower proportions of Na2O inferred from the Siroquant data relative to the actual proportions, probably mainly due to the presence of Na-rich I/S, rather than K-illite in those samples. The substitution of other elements for Na in the dawsonite lattice may also have caused an overestimation of Na by Siroquant (Ruan and Ward, 2002). Ruan and Ward (2002) also considered that some Na may be volatilized from the dawsonite at high temperatures, reducing the proportion of Na2O remaining in the 815 °C ash samples for XRF analysis.

TiO2

The plot for TiO2 shows a broad scatter, with almost all the points falling below the equality line (Figure 5.22). The proportion of Ti-bearing minerals appears to be underestimated by the XRD analysis. Ti-bearing minerals were observed to occur as fine- grained inclusions within the clay minerals. Such fine material, with possibly poor crystallinity and small proportions in the LTA residues, is inherently difficult to evaluate by XRD analysis.

135

Figure 5.22 Comparison between proportions of major element oxides in coal LTAs and non-coal strata from Greta seam (Austar), inferred from Siroquant and determined by XRF. The diagonal line represents equality in each plot. Relevant trendlines and squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case.

136

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Figure 5.23 Comparison between proportions of major element oxides in coal LTAs and non-coal strata from Greta seam (Cessnock), inferred from Siroquant and determined by XRF. The diagonal line represents equality in each plot. Relevant trendlines and squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case.

5.4.2 Vertical variations in major elements

The major elements in the Greta seam are dominated by SiO2, Al2O3 and Fe2O3 (Tables 5.6, 5.7), the major carriers of which are apparently quartz, clay minerals and pyrite. In additional to clay minerals, dawsonite also contains a significant proportion of alumina. The proportions of major oxides for the Austar section were also recalculated and normalized after removal of the Fe2O3, which mainly occurs in pyrite, and the results are expressed as graphic plots in Figure 5.24. This was intended to provide an indication of the variation in the non-pyrite components.

137

The normalised Fe-free alumina and silica show two very opposite vertical trends through the seam section, which reflect generally opposing variations in the abundances of quartz and the clay minerals. Normalised MgO, CaO and MnO are roughly correlated, and are all elevated at the top of the seam. This is due to a slightly higher proportion of carbonates, excluding dawsonite, in the top part of the seam.

The normalised K2O appears to be relatively high in the roof, floor and other epiclastic horizons. Relatively high normalized Na2O percentages occur in the majority of the coal plies, most of which contain high proportions of dawsonite and/or a small proportion of albite. In contrast to K2O, the normalised TiO2 appears to be more abundant in the coal samples, than in the epiclastic non-coal materials.

High concentrations of P are commonly found at particular horizons in the coal seam (from G-10 to G-14, and G-36). This high level of P is consistent with the occurrence of fluorapatite and goyazite at the same horizons. The high concentrations of P are also coupled with high CaO, both of which occur in apatite and goyazite. In the low P horizons, however, the CaO is mainly attributed to carbonates. Apart from the decomposition of organic matter, which contributes to the background level, phosphorus, especially when at high concentrations, may also be controlled by local hydrogeochemical factors, or by introduction of additional sources of phosphorus, e.g. volcanic debris, shells or faecal matter (Ward et al., 1996; Rao and Walsh, 1999). The high correlation between Ti and Al suggests that Ti bearing minerals may coexist with the clay minerals.

138

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Figure 5.24 Variation of major elements of the Greta seam (Austar), normalised to a Fe2O3- and LOI- free basis.

5.4.3 Trace element geochemistry

Trace element data for the Austar samples are given in Table 5.8. In general, the Greta coal has overall low concentrations of most trace elements. This is especially prominent for lithophile elements, which are usually associated with ash yield of coals.

Chalcophile elements in the Greta coals are also relatively low to those in typical pyrite- rich coals. Chalcophile elements Hg, As and Tl have elevated concentrations at the top of the section, with concentrations up to 0.8 ppm, 12.58 ppm and 0.44 ppm respectively. Other chalcophile elements such as Pb, Se, Mo and Ni, however, do not show similar trends. Epigenetic sulphides have been noted to be important carriers of chalcophile elements in coal. Although exception occurs, some studies show that chalcophile elements are often present in lower concentrations in early-stage sulphides, relative to epigenetic sulphides (e.g. Finkelman, 1994; Diehl et al., 2004; Diehl et al., 2012). The syngenetic forms of pyrite in the Greta coals may not carry significant concentrations of chalcophile elements other than Hg, As and Tl.

139

Table 5.8 Trace element analyses of the Greta coal and associated non-coal samples from the Austar Coal Mine (all elements but Hg in ppm, Hg in ppb, on whole-coal basis)

F Li Be Sc V Cr Co Ni Cu Zn Ga Ge As Se Rb Y Zr Nb

G-R 236 61.3 2.45 44.7 109 115 11.0 34.7 40.3 132 26.0 1.82 5.8 1.66 59.9 19.6 294 3.24 (1,2,4) G-3-r 276 74.2 1.66 47.3 41.2 55.6 9.49 31.6 20.5 37.7 16.8 1.82 2.76 1.48 39.1 12.7 178 2.15 G-5 197 5.49 0.99 4.71 25.7 4.71 23.7 3.16 10.2 14.4 6.66 3.90 4.64 3.29 1.63 13.7 73.2 0.91 G-6-p 575 88.3 2.75 42.4 56.8 24.7 12.1 32.8 41.1 55.3 90.8 1.16 12.6 1.90 60.8 27.0 189 2.71 G-7 86 5.77 0.29 3.55 7.77 4.92 1.30 3.66 8.74 5.63 4.02 0.66 2.47 2.15 1.05 3.92 2.72 0.64 G-10 98 5.47 0.23 2.24 8.17 3.65 1.18 3.37 8.86 5.27 3.71 0.28 2.87 2.20 0.84 2.72 1.51 0.46 G-11-13 204 10.8 0.40 4.85 12.1 9.10 0.87 7.57 13.7 6.29 5.37 0.41 11.4 3.13 2.03 4.45 2.82 1.51 G-14 158 15.3 1.68 7.00 24.5 11.3 3.02 12.4 11.1 17.8 7.69 3.97 3.05 1.52 3.37 12.4 0.29 1.32 G-16-17 322 89.6 2.23 55.8 75.1 61.1 5.54 15.6 15.5 64.3 32.7 1.28 2.38 1.79 97.3 21.4 263 6.78 G-18 102 43.7 1.79 24.9 32.4 22.7 5.58 21.4 19.9 36.8 11.0 2.16 0.86 1.19 14.3 12.7 137 1.88 G-19 71 24.6 1.44 8.02 29.3 9.57 3.49 11.1 19.8 26.5 11.1 6.58 0.84 2.43 5.35 8.55 59.6 1.01 G-21 51 37.5 0.69 8.14 37.6 6.95 2.17 5.18 12.5 16.5 6.93 3.42 0.31 2.66 4.09 9.33 69.7 1.14 G-23 39 19.8 0.45 4.64 16.1 6.39 1.78 4.84 10.8 9.65 6.14 12.3 0.74 2.52 3.47 6.37 47.6 0.89 G-27-29 41 23.8 0.51 5.46 21.4 9.18 1.76 7.40 13.3 6.10 6.06 12.2 0.80 3.33 3.18 6.97 42.0 1.14 G-30-32 28 23.9 0.45 5.92 18.9 8.89 1.15 7.32 12.6 5.81 5.26 5.36 0.74 2.61 2.40 6.03 35.5 1.14 G-34 55 27.3 0.40 7.24 18.3 23.3 0.98 8.67 17.6 7.59 5.57 0.91 0.30 2.67 5.13 4.59 39.9 1.59 G-35 37 20.5 0.27 3.00 6.28 5.65 0.59 6.46 11.1 5.44 2.46 0.78 2.36 5.71 1.30 2.29 18.0 0.77 G-36 21 7.99 0.32 2.21 6.70 5.47 0.54 6.44 8.92 4.15 2.45 6.49 0.32 1.74 0.52 2.45 3.91 0.73 G-37 14 19.3 1.32 3.68 10.2 8.64 0.74 8.97 10.2 7.09 3.11 17.9 0.49 1.89 1.87 4.66 27.2 0.73 G-38 24 11.0 1.26 2.67 8.19 10.5 0.60 10.6 11.0 4.51 2.64 23.5 0.17 1.47 0.99 3.01 16.4 0.95 G-39 21 10.4 1.72 2.96 9.93 14.9 0.76 14.2 8.31 8.19 2.79 35.6 0.14 1.67 1.12 3.13 17.4 0.67 G-41 49 25.5 3.66 13.9 32.5 50.7 3.32 24.5 14.2 11.6 7.07 33.8 0.25 2.03 6.16 13.3 42.5 1.54 G-F(42-45) 394 102.2 4.03 53.5 86.8 446 27.8 229.6 36.6 84.1 26.5 1.74 4.54 0.87 137 19.2 222 5.30

140

Table 5.8 (Continued)

Hg Mo Ag Cd Sn Sb Te Cs Ba Hf Ta W Re Tl Pb Bi Th U (ppb) G-R 1.66 0.59 0.36 1.43 0.32 <1 3.06 171 7.25 <1 0.77 <0.1 1.96 213 9.79 <0.2 5.91 2.68 (1,2,4) G-3-r 1.26 0.63 0.23 0.99 0.28 <1 1.86 134 4.67 <1 0.37 <0.1 2.46 209 7.76 <0.2 4.83 1.67 G-5 0.69 0.41 0.12 0.60 0.07 <0.1 0.27 125 1.28 <0.2 1.15 <0.01 0.14 811 1.87 0.16 0.75 0.26 G-6-p 1.61 0.54 0.42 3.26 0.82 <1 5.37 1464 9.01 <1 0.87 <0.1 1.20 453 11.5 <0.2 7.46 2.18 G-7 0.37 0.13 0.09 0.33 0.03 <0.1 0.13 68 <0.1 <0.2 0.22 <0.01 0.17 577 2.22 0.13 1.01 0.32 G-10 0.42 0.13 0.09 0.27 0.05 <0.1 0.13 53 <0.1 <0.2 0.14 <0.01 0.43 785 1.51 0.11 0.63 0.31 G-11-13 0.66 0.21 0.16 0.43 0.07 <0.1 0.23 61 <0.1 <0.2 0.16 <0.01 0.44 787 2.66 0.14 1.19 0.49 G-14 0.47 0.29 0.10 0.32 0.16 <0.1 0.37 109 0.17 <0.2 0.39 <0.01 0.08 67.5 2.73 0.14 1.27 0.49 G-16-17 0.76 0.60 0.45 2.67 0.32 <1 5.83 230 7.61 <1 0.88 <0.1 0.24 13.4 17.6 <0.2 11.6 3.03 G-18 0.68 0.56 0.26 1.01 0.16 <0.1 0.89 78 3.20 <0.2 0.25 <0.01 0.09 21.0 5.01 0.16 2.56 0.83 G-19 0.66 0.28 0.15 0.55 0.19 <0.1 0.35 132 1.66 <0.2 0.23 <0.01 0.05 20.8 3.66 0.14 1.21 0.35 G-21 0.45 0.36 0.12 0.48 0.15 <0.1 0.39 36 2.10 <0.2 0.34 <0.01 0.04 32.4 3.77 0.10 1.25 0.34 G-23 0.39 0.27 0.16 0.31 0.15 <0.1 0.20 51 1.47 <0.2 0.45 <0.01 0.04 10.7 3.84 0.06 1.01 0.31 G-27-29 0.73 0.19 0.10 0.57 0.14 <0.1 0.31 23 1.23 <0.2 0.75 <0.01 0.10 60.4 3.96 0.16 1.68 0.48 G-30-32 0.38 0.15 0.09 0.63 0.11 <0.1 0.26 29 1.15 <0.2 0.50 <0.01 0.06 85.6 3.42 0.14 1.92 0.52 G-34 0.36 0.18 0.09 0.69 0.14 <0.1 0.48 23 1.28 <0.2 0.30 <0.01 0.04 11.0 4.99 0.19 2.16 0.62 G-35 0.43 0.09 0.11 0.28 0.03 <0.1 0.08 13 0.56 <0.2 0.13 <0.01 0.19 189 3.05 0.15 1.10 0.27 G-36 0.29 0.08 0.09 0.22 0.04 <0.1 0.04 16 0.14 <0.2 0.24 <0.01 0.03 29.0 2.39 0.12 0.89 0.27 G-37 0.35 0.14 0.11 0.31 0.05 <0.1 0.13 12 0.85 <0.2 0.25 <0.01 0.04 23.5 2.21 0.10 0.96 0.26 G-38 0.31 0.09 0.06 0.27 0.05 <0.1 0.07 8 0.50 <0.2 0.32 <0.01 0.03 10.1 2.20 0.13 1.13 0.28 G-39 0.31 0.08 0.07 0.30 0.05 <0.1 0.13 10 0.52 <0.2 0.33 <0.01 0.03 9.1 2.13 0.10 1.04 0.27 G-41 0.31 0.19 0.09 0.54 0.15 <0.1 0.79 23 1.35 <0.2 0.18 <0.01 0.06 14.6 4.12 0.11 2.54 0.69 G-F(42-45) 1.17 0.50 0.16 1.95 0.39 <1 5.86 128 6.53 <1 1.00 <0.1 0.53 46.8 13.2 <0.2 9.40 2.65

141

5.4.4 Geochemical associations and element affinity

The geochemical results of the Greta coal samples were analysed using cluster analysis, to identify groups of associated elements. Hierarchical clustering was performed using the Pearson correlation coefficients. The likely organic/mineral affinity of the elements in the Greta coals is indicated by the statistical correlation of trace element concentrations with the high-temperature ash percentage (HTA). Elements with a strong inorganic affinity would be expected to show a positive correlation to the ash percentage, and those with a strong organic affinity would show a negative correlation. Likewise, the possible modes of occurrence of the different elements can be also inferred based on the correlation of their concentrations with particular mineralogical abundances. To allow for variations in mineral matter percentage, the relationships between trace element geochemistry and mineralogy of the coal samples were evaluated on a whole-coal basis. Elements which are the most correlated are firstly linked, and then elements or element groups with decreasing correlation are further linked, until a dendrogram is achieved.

Associations of elements in the Greta coals of the Austar section are broadly indicated by the resulting dendrogram (Figure 5.25). Some five groups, and also the statistical correlation coefficients between selected elements and LTA%, are shown in Table 5.9. Apart from the inter-correlation among elements in the same group, each group may also include elements of different sub-groups that have different correlations with HTA percentage, proportion of Al2O3 or abundance of minerals.

142

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Figure 5.25 Dendrogram developed from cluster analysis on the geochemical data of the Greta coals from the Austar Coal Mine (cluster method, centroid clustering; interval, Pearson correlation; transform values, maximum magnitude of 1).

Table 5.9 Broad classification of elements according to the results from cluster analysis and correlation coefficients (R) between the content of individual elements (E) in coal and LTA% and other minerals in the LTAs (Correlation coefficients for all inter-element and ash relationships are presented in Appendix 1).

Group Al2O3 (0.74), SiO2 (0.87), K2O (0.81), Na2O (0.82), Cd (0.84), Sn (0.71), Cs (0.71), Hf RE-HTA

A (0.88), Sc (0.82), MgO (0.7), Rb (0.81), TiO2 (0.67), Cu (0.61), Zr (0.63), Nb (0.63), Li (0.52), Pb (0.46), Th (0.5), U (0.7), La (0.62), Ce (0.59)

Group V (0.45), Ga (0.64), Y (0.42), Sb (0.38), Pr (0.61), Nd (0.52), Sm (0.57), Eu (0.63), Gd RE-LTA B (0.73), Zn (0.68), Ag (0.71), Tb (0.55), Dy (0.59), Ho (0.5), Er (0.52), Tm (0.47), Yb (0.5), Lu (0.44)

Group Fe2O3 (0.97), P2O5 (0.74), Tl (0.91), As (0.87), CaO (0.7), F (0.83), MnO (0.28), Hg (1.0), RE-pyrite C Se (0.2)

Group Co (0.1), Mo (0.53), Ba (0.34), W (-0.18), Bi (0.23) RE-LTA D

Group Be (0.16), Cr (0.27), Ni (0.4), Ge (-0.34) RE-LTA E

Group A comprises the largest number of elements. These elements generally have high correlation coefficients with LTA%. High correlations also generally exist between these elements and alumina and silica, indicating their affinity with the aluminosilicate minerals.

For example, Li is more strongly correlated with Al2O3 (R=0.93, Figure 6A) than with kaolinite (R=0.77, Figure 5.26B). Lead, Cu and Sn, which are usually regarded as

143 chalcophile elements, are associated with Al2O3, having correlation coefficients of 0.81, 0.78 and 0.79 (Figures 5.26C, D, E), respectively. A similar positive correlation is also observed between these elements and dawsonite, although the correlations are generally a little less significant. Therefore, many of the chalcophile elements in the Greta coals appear to be associated with the clay minerals or other Al-bearing minerals, rather than with the pyrite. Alternatively, the correlations between these chalcophile elements and the clay minerals may indicate a common source.

Group B also includes elements that have relatively high correlation coefficients with

LTA%, Al2O3 and SiO2. Like Pb, a high correlation coefficient (R=0.77, Figure 5.26F) also exists between Sb and Al2O3, indicating that Sb is associated with the aluminosilicates, mainly clay minerals, rather than with pyrite.

Group C includes Fe2O3, Tl, Hg, As, Se, P2O5, CaO and MnO. In this association, Tl, As and Hg have a pyrite affinity. They are positively correlated with pyrite in the coal, with correlation coefficients of 0.84, 0.52 and 0.97 (Figures 5.27A, B, C) respectively. No correlation was found between pyrite and other chalcophile trace elements, such as Pb, Mo, Sn, Sb, Co and Cu, although correlations of these elements with pyrite have been observed frequently in other coals around the world. The other sub-group comprises

Fe2O3, CaO and MnO. These elements are correlated with each other, mainly due to their association with carbonate minerals (mainly calcite, dolomite and ankerite). Ca and P are closely linked due to their occurrence in apatite and goyazite. Selenium is not correlated with pyrite (R=0.2) (Figure 5.27D) or any of the carbonate minerals, although it is clustered in this group.

The other two associations are Group D (Co, W, Mo, Ba and Bi) and Group E (Be, Ni, Cr and Ge), which include elements that have correlation coefficients with LTA% of between -0.34 and 0.53. These elements have no obvious correlation with the LTA%, and lack an obvious correlation with any of the identified minerals. These elements may have a mixed organic/mineral affinity. However, no clear explanation can be given for the association.

It is worth noting that a high correlation coefficient may sometimes be developed between two elements, or an element and a mineral, controlled by data from single or minor samples, even though a real correlation does not exist. For this reason, not all the results indicated by the dendrogram can be reasonably explained in terms of the correlation coefficients alone.

144

Chapter 5 Mineralogy and Geochemistry of the Greta Seam

Figure 5.26 Correlation of selected elements (Li, Pb, Cu, Sn and Sb) with alumina and kaolinite in the Greta coal samples. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case.

5.4.5 Distribution and affinity of REE and Y

7KHWRWDOUDUHHDUWKHOHPHQWFRQWHQWV 5((< LQWKH*UHWDFRDOVDPSOHVW\SLFDOO\UDQJH between 10 to 30 ppm (Table 5.10). Coal G-18, which has the highest HTA percentage (41.62%), is also the most REE-enriched (48.7ppm) coal. The REE concentrations were normalised against the Upper Continental Crust (UCC) (Taylor and McLennan, 1985) for each coal and associated non-coal sample, in order to obtain a more clear indication of the distribution patterns (Figure 5.29).

145

Figure 5.27 Correlation of Tl, As, Hg and Se with pyrite in the Greta coal samples. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case.

Figure 5.28 Correlation coefficients between mean individual REE and Y with LTA% in the Greta coal samples.

The correlation coefficients between individual rare earth elements (REE), including Y, and LTA% vary from 0.42 (Y-LTA%) to 0.73 (Gd-LTA%) (Figure 5.28). This indicates a higher affinity of the light and middle rare earths (LREE and MREE) to the mineral matter of the coal than that of the heavy rare earths (HREE) and Y. The HREE and Y may have a greater organic affinity.

The normalized REE distributions in all the Greta coal samples have various LaN/LuN values, all are less than 1 (Table 5.10), indicating heavy REE enrichment (H-type)

146

Chapter 5 Mineralogy and Geochemistry of the Greta Seam compared to the UCC (Seredin, 2001), although some coals display relatively flat REE patterns. As mentioned above, The LREE in the Greta coals generally have a greater mineral affinity than the HREE. Eskenazy (1999) ascribed the enrichment of HREE relative to LREE to the formation by HREE of complexes with organic matter in coal, which would also increase the stability of the HREE relative to the LREE. As discussed by Seredin (2001), H-type distribution patterns may be attributed to the influence of marine water.

With exception of several coal samples which show slight negative Eu and Y anomalies, most of the coals and all the associated non-coal samples show slight to pronounced positive Eu anomalies and negative Y anomalies (Table 5.10, Figures 9.29A, B, C), with Eu/Eu* and Y/Y* values ranging from 1.04 to 1.56 and 0.73 to 0.99, respectively. The Ce/Ce* values of the coal samples are between 0.9 and 1.07, indicating no pronounced Ce anomaly.

Figure 5.29 Distribution patterns of REE in the Greta seam. REE are normalized to Upper Continental Crust (UCC) data from Taylor and McLennan (1985). (A) Coal samples G-5 to G-14; (B) Coal samples G-18 to G- (30-32); (C) Coal samples G-34 to G-41; (D) Roof samples G-(1,2,4) and G-3, floor sample G-(42-45), parting sample G-6, and other clastic rock sample G-(16-17).

147

Table 5.10 Rare earth elements in coal samples and associated strata from the Greta seam in the Austar section (REE concentrations in ppm, on whole-coal basis).

G-(1,2,4) G-3-r G-5 G-6 G-7 G-10 G-(11-13) G-14 G-(16-17) G-18 G-19

La 15.88 13.55 3.58 10.05 2.29 1.21 2.82 3.35 28.37 8.41 6.26 Ce 34.9 30.14 7.12 30.42 5.31 2.7 5.97 8.49 60.1 17.92 14.59 Pr 4.46 3.82 0.91 4.66 0.58 0.33 0.74 1.15 7.53 2.17 1.78 Nd 17.06 13.93 3.88 19.79 2.17 1.5 2.86 4.62 27.26 7.8 8.21 Sm 4.01 3.27 1.12 6.1 0.52 0.41 0.74 1.23 5.77 1.98 1.69 Eu 1.15 0.88 0.31 1.53 0.13 0.11 0.19 0.34 1.44 0.61 0.46 Gd 3.5 2.63 1.31 4.89 0.56 0.48 0.96 1.06 3.49 2.14 1.48 Tb 0.63 0.45 0.3 1.08 0.11 0.09 0.16 0.28 0.79 0.38 0.3 Dy 4.08 2.66 2.05 6.85 0.71 0.6 1.02 1.89 4.85 2.65 1.71 Ho 0.94 0.59 0.56 1.44 0.16 0.12 0.21 0.5 1.03 0.61 0.41 Y 19.65 12.71 13.7 26.99 3.92 2.72 4.45 12.41 21.42 12.7 8.55 Er 2.81 1.97 1.55 4.19 0.48 0.34 0.62 1.35 2.91 1.76 1.16 Tm 0.41 0.3 0.25 0.59 0.07 0.04 0.09 0.22 0.4 0.25 0.17 Yb 2.55 1.99 1.54 4.25 0.41 0.28 0.56 1.54 2.64 1.76 1.14 Lu 0.4 0.34 0.28 0.6 0.07 0.04 0.09 0.24 0.44 0.26 0.18 5(( 92.78 76.52 24.76 96.44 13.57 8.25 17.03 26.26 147 48.7 39.54 Eu/ Eu* 1.44 1.41 1.19 1.32 1.12 1.15 1.04 1.4 1.49 1.38 1.37 Ce/ Ce* 0.94 0.95 0.9 0.96 1.05 0.97 0.94 0.97 0.94 0.96 0.99 Y/Y* 0.76 0.77 0.97 0.65 0.88 0.77 0.73 0.97 0.73 0.76 0.78

(La/Lu)N 0.42 0.43 0.14 0.18 0.35 0.32 0.33 0.15 0.69 0.35 0.37

(La/Sm)N 0.59 0.62 0.48 0.25 0.66 0.44 0.57 0.41 0.74 0.64 0.56

(Gd/Lu)N 0.74 0.65 0.39 0.69 0.67 1.01 0.9 0.37 0.67 0.69 0.69 Enrichment type H H H H H H H H H H H

148

Table 5.10 (Continued)

G-21 G-23 G-(27-29) G-(30-32) G-34 G-35 G-36 G-37 G-38 G-39 G-41 G-(42-45)

La 2.45 3.4 4.92 4.14 3.85 2.18 2.85 2.71 2.27 2.74 4.24 24.42 Ce 7.04 7.6 10.78 9.82 8.89 5.31 5.6 6.43 5.22 5.89 10.11 49.04 Pr 1.1 0.93 1.29 1.19 1.03 0.59 0.58 0.76 0.57 0.62 1.31 5.87 Nd 4.94 3.75 4.89 4.59 3.9 2.19 2.27 2.95 2.11 2.62 5.21 20.32 Sm 1.47 0.93 1.25 1.17 0.87 0.49 0.51 0.73 0.5 0.6 1.35 4.27 Eu 0.39 0.24 0.33 0.28 0.24 0.13 0.13 0.2 0.13 0.16 0.42 1.19 Gd 1.29 0.73 1.01 0.9 0.72 0.33 0.38 0.49 0.31 0.49 1.44 2.91 Tb 0.3 0.17 0.23 0.19 0.15 0.07 0.08 0.13 0.08 0.11 0.32 0.61 Dy 1.83 1.23 1.35 1.33 0.94 0.5 0.47 0.9 0.53 0.61 2.02 4.07 Ho 0.44 0.29 0.33 0.29 0.22 0.11 0.11 0.2 0.13 0.13 0.51 0.88 Y 9.33 6.37 6.97 6.03 4.59 2.29 2.45 4.66 3.01 3.13 13.29 19.25 Er 1.32 0.88 0.95 0.84 0.62 0.28 0.31 0.63 0.37 0.37 1.46 2.64 Tm 0.19 0.14 0.13 0.12 0.09 0.04 0.04 0.08 0.05 0.05 0.19 0.38 Yb 1.26 0.98 0.86 0.77 0.51 0.28 0.28 0.59 0.38 0.31 1.22 2.52 Lu 0.21 0.15 0.13 0.12 0.09 0.04 0.04 0.1 0.06 0.05 0.19 0.39 5(( 24.23 21.42 28.45 25.75 22.12 12.54 13.65 16.9 12.71 14.75 29.99 119.5 Eu/ Eu* 1.33 1.37 1.38 1.28 1.42 1.51 1.38 1.56 1.53 1.39 1.41 1.58 Ce/ Ce* 0.93 0.97 0.97 1 1.02 1.07 0.99 1.02 1.05 1.03 0.97 0.93 Y/Y* 0.79 0.81 0.79 0.74 0.77 0.74 0.82 0.84 0.87 0.84 0.99 0.77

(La/Lu)N 0.12 0.24 0.4 0.37 0.46 0.58 0.76 0.29 0.4 0.58 0.24 0.67

(La/Sm)N 0.25 0.55 0.59 0.53 0.66 0.67 0.84 0.56 0.68 0.69 0.47 0.86

(Gd/Lu)N 0.52 0.41 0.65 0.63 0.67 0.69 0.8 0.41 0.44 0.83 0.64 0.63 Enrichment type H H H H H H H H H H H H

Subscript N indicates values normalized by the average REE content of Upper Continental Crust (Taylor and Mclennan, 1985). Eu/Eu*=2EuN/(SmN+GdN); Ce/Ce*=2CeN/(LaN+PrN);

Y/Y*=2Y/(DyN+HoN).

149

The roof, floor, parting and other non-coal rock samples show very similar REE patterns, with negative Eu and Y anomalies (Table 5.10, Figure 5.29D), which is also similar to most of the coal samples, especially those in the middle part of the seam (Figure 5.29B). This may indicate a similar REE source for the coals and the non-coal samples.

5.5 Summary

The coals in the Cessnock section, which represents the lower section of the Greta seam, and in the Austar borehole, through a more complete section the Greta seam, were evaluated in this chapter. The Greta coal is a high-volatile bituminous coal and typically contains a high proportion of liptinite. Although the Greta seam is not immediately overlain by the marine strata, the percolation of marine water is indicated by the petrological, mineralogical and geochemical characteristics, which resemble those of sequences with a marine roof. The upper section of the Greta seam has several different indicators of marine influence, such as anomalously low vitrinite reflectance and abundant syngenetic pyrite, in the top part of the seam profile.

The coals have contrasting mineralogy in the upper and lower sections of the seam. Pyrite typically comprises 40 to 56% of the mineral assemblage in the coals from the marine- influenced upper part of the Austar section. Small amounts of hydrous iron sulphates, including jarosite, coquimbite, copiapite and szomolnokite, also exist, especially in the coals of the Cessnock section, due to oxidation of pyrite during sample storage. The lower section, however, contains much less pyrite (typically <5%, organic-free basis), and also relatively abundant dawsonite (up to 14%, organic-free basis). Kaolinite appears to be less abundant in the upper section than in the lower section, which is probably due to the dilution by pyrite in the former. Other minor minerals, such as carbonates (calcite, ankerite, dolomite and siderite) and other clay minerals (illite and I/S), appear to show no preferred vertical distribution.

The most abundant detrital minerals in the Greta coals are quartz, poorly ordered kaolinite, illite and I/S. These detrial minerals occur mainly in the floor, roof and other epiclastic horizons of the seam, which reflects greater amount of clastic influx in those parts of the original peat-forming swamp. As indicated by optical and electron microscope analysis, however, detrital minerals are rare in the coals away from the floor, roof and other epiclastic horizons. This may be due almost complete sediment by-passing of the

150

Chapter 5 Mineralogy and Geochemistry of the Greta Seam depositional system, if peat formation took place in front of the slowly encroaching sea and under conditions of continued subsidence (Diessel, 1992). Alternatively, the bulk of the detrital minerals, if they were originally present, may have been leached by diagenetic or post-diagenetic processes. This may have resulted in the formation of soluble ions that were removed from the system, or in the deposition of authigenic minerals within the peat bed.

The minerals within most coal plies themselves are largely of authigenic origin. Syngenetic pyrite is the most abundant, occurring as framboids, euhedral crystals and cell or pore space infillings. In some cases, pyrite also occurs as a replacement of earlier- formed siderite. Clay minerals (mainly kaolinite and Na-rich I/S) and quartz in the coal samples largely occur as cell or pore space infillings, and were probably precipitated in the early diagenetic stage. Siderite occurs as nodules, which also indicate a relatively early, syngenetic origin.

Minor Ti-bearing minerals, anatase or rutile, may occur as microcrystalline particles in a matrix of clay mineral, and were probably co-precipitated with the clay minerals in the early diagenetic stage. Subordinate phosphate minerals, fluorapatite and goyazite, are present at certain levels of the seam profile. Goyazite was observed as fine grained particles in the vitrinite matrix. The goyazite in the sulphur-rich coals contains variable concentrations of sulphur, which was probably available in the early-diagentic stage in the form of sulphate ions. This, along with the fine grain size, indicates that the goyazite was formed during the early-diagenetic stage.

The occurrence of some other minerals in the coal samples generally indicates late-stage precipitation. Abundant cleat- and fracture-filling dawsonite occurs in the lower part of the seam, which was formed at an epigenetic stage of seam development. Intergrowth of kaolinite and dawsonite in the cleat suggests that the kaolinite was probably a precursor of dawsonite. The formation of the dawsonite may be the result of reactions between earlier-precipitated kaolinite and Na2CO3- or NaHCO3-bearing fluids. Minor albite occurs in some cell or clearly broken cavities; this may also be epigenetic, probably precipitated from the same Ca-Al bearing fluids that formed dawsonite.

Other carbonates (calcite, dolomite and ankerite) are minor components. Although the occurrence of these carbonates as late-stage cleat-fillings is very rare, at least some of them may also be epigenetic, precipitated in broken cell cavities of inertinite.

151

The elements in the coal can be grouped into five associations based on cluster analysis.

Group A (Al2O3, SiO2, K2O, Na2O, Cd, Sn, Cs, Hf, Sc, Rb, TiO2, Cu, Zr, Nb, Li, Pb, Th, U, La and Ce) and Group B (V, Ga, Y, Sb, Zn, Ag, Y, REE except La and Ce) include elements that are strongly or relatively strongly associated with LTA% and the aluminosilicate minerals. Group C (Fe2O3, Se, Tl, As, Hg, Fe, P2O5, CaO, MgO and MnO) embraces elements that are associated with pyrite and carbonates. Group D (Co, W, Mo, Ba and Bi) and Group E (Be, Ni, Cr and Ge) include elements that have no obvious correlation with any of the minerals in the seam section.

With the exception of Hg, Tl and As, no traditional association between pyrite and chalcophile elements (e.g. Se, Pb, Sb) is apparent in the Greta coals. This may be because the pyrite in the Greta coal is essentially of syngenetic origin. Although some of the chacophile trace elements may occur in syngenetic pyrite, there is a lack of clear correlation for others. The other chalcophile elements, however, are more strongly correlated with Al2O3, indicating an association with the clay minerals or other Al-bearing minerals, rather than pyrite.

152

CHAPTER 6 MINERALOGY AND GEOCHEMISTRY OF THE GREAT NORTHERN SEAM

This study aims to investigate the characteristics and vertical variation in the mineral matter of the Great Northern seam and associated strata from the northern Sydney Basin, including the immediate roof and floor rocks and a number of thin intra-seam claystone bands, as well as the different individual coal plies within the seam section. Such a study provides an opportunity to assess the mineralogical and geochemical characteristics of both the coal and non-coal strata in a major and economically-important coal seam having a lesser degree of marine influence than the Greta section, with an emphasis on evaluating the role of the different inorganic processes and sedimentary inputs that may be associated with coal formation under different depositional conditions.

6.1 Coal characteristics

A summary of the properties of each sample from the Newvale Colliery section (see Chapter 4) is given in Table 6.1. According to ASTM D388-05 (ASTM, 2007), the coals are high volatile A bituminous in rank. The individual coal plies mainly have low ash yields (<12%), except for two high-ash plies (samples 21380 and 21390) immediately below the roof and above the floor of the seam. All the coals are low in total sulphur (<0.4%), with the sulphur being mainly organic in form. The mean random vitrinite reflectance is typically around 0.82%.

6.2 Coal petrology

As described in Chapter 4, the macroscopic appearance of the Great Northern coals is dominated by banded bright and dull lithotypes. Maceral characteristics and the abundance of the different macerals were evaluated from crushed grain mounts of samples from the Newvale section. The classification of macerals for this study was based on the ICCP standard, as indicated in Table 5.2.

153

Table 6.1 Proximate analysis and vitrinite reflectance data for the Great Northern coal and associated strata from Newvale No.1 Colliery (see Chapter 4). Proximate analysis data from CSIRO reports.

Thickness M Ash VM FC CV Total S Rv, ran Sample (%, (%, (%, (%, (%, (Btu/1b, (m) (%, ad) (%, ad) (%) ad) ad) daf) ad) daf) daf) 21379 0.1 0.58 95.85 - - - - - 0.06 - 21380 0.02 1.02 40.62 - - - - - 0.16 - 21381 0.2 2.59 12.06 34.19 40.1 51.16 59.9 14550 0.36 0.81 21382 0.03 2.57 51.93 21.40 - - - - 0.21 - 21383 0.37 2.68 8.84 32.52 36.8 55.96 63.2 14570 0.35 0.85 21384 0.01 2.4 68.92 23.30 - - - - 0.11 - 21385 0.14 2.7 10.06 28.72 32.9 58.52 67.1 14570 0.33 0.83 21386 0.6 2.65 6.67 36.86 40.6 53.82 59.4 14750 0.39 0.80 21387 0.5 2.82 11.11 32.27 37.5 53.8 62.5 14620 0.36 0.80 21388 0.52 2.86 9.57 34.27 39.1 53.3 60.9 14520 0.39 0.80 21389 0.01 2.69 62.45 - - - - - 0.17 - 21390 0.14 2.44 28.5 26.50 38.4 42.56 61.6 14700 0.41 0.82 21391 0.08 1.98 57.56 - - - - - 0.2 - ad = air-dried basis; daf = dry ash-free basis; M = inherent moisture; FC = fixed carbon; TC = total sulphur; SS = sulphate sulphur; PS = pyritic sulphur; OS = organic sulphur; - = no data.

Like other seams in the upper Newcastle Coal Measures (Warbrooke, 1987), the coal of the Great Northern seam is relatively rich in inertinite. Agnew et al. (1995), indicated around 43% vitrinite, 53% inertinite and 4% liptinite for typical Great Northern seam products on a mineral-free basis.

Table 6.2 lists the maceral analysis results for the individual coal samples from the Newvale section, except for the high-ash coal of sample 21380. On a mineral-free basis, the individual plies have 44-88% vitrinite, 11-55% inertinite (mainly semifusinite) and up to 5% liptinite. The coal plies in the upper part of the seam, especially samples 21383 and 21385, have higher inertinite percentages than those in the lower part, possibly reflecting greater exposure and oxidation of the peat during seam formation.

Vitrinite The vitrinite group macerals in the Great Northern coal samples are mainly collotelinite, collodetrinite and telinite, along with trace amounts of gelinite and corpogelinite. Collotelinite occurs as structureless and homogeneous bands (Figures 6.1A, B, C). Some collotelinite is impregnated with in situ inclusions, such as cutinite (Figure 6.1C) and syngenetic minerals. Telinite consists of well-preserved cell walls (Figure 6.1D) and some telinite contains gelinite within the cell lumens (Figure 6.1E). Telinite sometimes grades to collotelinite (Figure 6.1B). Corpogelinite is relatively rare (Figure 6.1F).

154

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

Table 6.2 Maceral analysis (vol., %, mineral-free basis) of Great Northern coal samples from Newvale No.1 Colliery (bdl = below detection limit)

Maceral 21381 21383 21385 21386 21388 21390 Telinite 10.6 6 12 16.1 30.3 4.2 Collotelinite 21.4 23.5 22.4 31.2 32.4 10.2 Collodetrinite 32.6 18.3 8.2 23.1 17.3 71.1 Vitrodetrinite 0.4 0.2 0.4 0.4 0.2 0 Corpogelinite 0 0.2 0 1 2.2 2.1 Gelinite 0.2 0.8 1 bdl bdl 0 Total vitrinites 65.3 49 44.1 71.8 82.5 87.5

Fusinite 3.1 3.2 6.4 3.1 4.1 2.1 Semifusinite 21.8 36.3 40.5 17.1 7.7 4.2 Funginite 0 0 0 bdl 0.2 0 Inertodetrinite 3.5 2.6 5.2 1.8 1.4 3.9 Macrinite 1.9 2.2 1.8 1 0.6 0.7 Micrinite 1.9 2 1.2 1.4 0.2 0.2 Secretinite 0 0 0.2 0.2 0 0 Total inertinites 32.2 46.2 55.3 24.7 14.3 11.1

Sporinite 0.8 1.2 0.2 1.4 0.6 0.5 Cutinite 0.8 1.2 bdl 1.2 2 0.7 Resinite 0 0.4 0.2 bdl bdl 0 Liptodetrinite 0.8 2 0 0.8 bdl 0.2 Suberinite 0 0 0 0 0.4 0 Fluorinite 0 0 0.2 0 0 0 Bituminite 0 0 0 0 0.2 0 Total liptinites 2.5 4.8 0.6 3.5 3.3 1.4

Liptinite The liptinite group macerals occurring in the Great Northern coals are mainly sporinite and cutinite, and to a lesser extent, resinite and alginite. Sporinite occurs as aggregated or discrete bodies (Figures 6.2A, B) in collodetrinite. Thick-walled sporinite also occurs (Figures 6.2C, D). Cutinite generally has lower fluorescence intensity than sporinite (Figures 6.2C, D). Resinite rarely occurs (Figure 6.2A, B).

Inertinite The inertinite group macerals occurring in the Great Northern coal samples are mainly semifusinite and, to a lesser extent, fusinite (Figure 6.3A). Other minor inertinite macerals are macrinite (Figure 6.3B), inertodetrinite (Figures 6.3B, C), secretinite (Figure 6.3C) and funginite (Figure 6.3D).

155

Figure 6.1 Photomicrographs showing typical macerals in the Great Northern coals (Newvale). Oil immersion, reflected light. (A) Collotelinite (ct) bands. Sample 21383. (B) Telinite (t) grading upward to collotelinite (ct). Sample 21388. (C) Collotelinite (t) bands with thin-walled cutinite (c) layers. Sample 11212. (D) Telinite (t) with well preserved cells. Sample 21388. (E) Telinite (t) with the cell cavities filled with gelinite (g). Sample 11210. (F) Corpogelinite (co). Sample 21388.

156

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

Figure 6.2 Photomicrographs of liptinite macerals in the Great Northern coals (Newvale). Oil immersion, reflected light and blue-violet fluorescence. (A) Resinite (rs) and sporinite (sp) in collodetrinite. Sample 11210. (B) Same view as (A) under fluorescent illumination. (C) Thick-walled sporinite (sp) and cutinite (c). Sample 11205. (D) Same view as (C) under fluorescent illumination. (E) Alginite (?). Sample 11208. (F) Same view as (E) under fluorescent illumination.

157

Figure 6.3 Photomicrographs showing typical inertinite macerals in the Great Northern coals (Newvale). Oil immersion, reflected light. (A) Fusinite (f) grading to semifusinite (sf). Sample 11210. (B) Macrinite (ma). Sample 11208. (C) Secretinite (se). Sample 21383. (D) An agglomeration of funginite (fg). Sample 21381.

6.3 Mineralogy of the Great Northern seam

Quantitative results from XRD analysis and Siroquant interpretation for the coal LTAs and for the non-coal rocks in the Newvale and Catherine Hill Bay sections are given in Tables 6.3 and 6.4, respectively; oriented-aggregate XRD data are given in Table 6.5.

6.3.1 Minerals in roof and floor samples

The Great Northern seam in the Newvale Colliery section rests on a floor of silicified carbonaceous siltstone, containing abundant quartz, a significant proportion of K-feldspar, and a lesser proportion of illite (Table 6.3). A similar siltstone is present below the seam in the exposure at Catherine Hill Bay, although the material at this location contains less

158

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam quartz, higher proportions of kaolinite and smectite, and minor proportions of both K- feldspar and a plagioclase component (Table 6.4).

The siltstone at the base of the seam at Catherine Hill Bay ranges from 0.06 to 0.2 m in thickness (Figure 4.3). Thin-section examination shows a framework consisting mainly of detrital quartz, K-feldspar and kaolinized biotite, supported by abundant silica cement. Dispersed plant fragments with silica-impregnated tissues are also present. Minor veins of K-feldspar cut across the sedimentary fabric, including the plant fragments (Figure 6.4A). SEM examination shows that the K-feldspar occurs in laminae nearly parallel to bedding in the siltstone floor of the Newvale section (Figure 6.4B). Minor detrital microperthite and albite are also present, consistent with the XRD data.

Quartz also occurs as veins (Figure 6.4C) and an infilling of crushed fusinite-like material in this sample (Figure 6.4D), as well as in more massive forms. The siltstone floors at both localities resemble ganisters, which are hard, compact very fine to medium-grained quartz arenites (cf. Folk, 1974), cemented by silica, that have gone through pedogenesis (Retallack, 1977). They commonly occur below coal seams and contain carbonaceous traces (e.g. Hemingway, 1968; Besly and Fielding, 1989). However, quartz in the siltstone floors of both sections is not as abundant as that within true ganisters, which, as Percival (1983) suggested, should be >95%. Correspondingly, clay minerals are less depleted in the floor samples of the present study than in ganisters that have gone through leaching processes in a palaeosol profile.

The siltstone at Catherine Hill Bay is underlain by a mudstone containing abundant smectite and significant proportions of kaolinite and quartz, along with a minor proportion of feldspar (probably an albite-rich plagioclase) and muscovite. Oriented-aggregate XRD analysis (Table 6.5) confirms that smectite is more abundant than in the overlying siltstone, and is the main expandable clay mineral present. Similar material may also occur below the siltstone in the Newvale section, but this was not available for sampling from the mine workings.

159

Figure 6.4 (A) Vein probably made-up of K-feldspar cutting well-preserved plant tissue in the siltstone floor of Catherine Hill Bay section, thin-section under crossed polars. (B) SEM image of K-feldspar veins in the siltstone floor of Newvale section. (C) SEM image of quartz veins in the siltstone floor of Newvale section. (D) SEM image of quartz with crushed fusinite in the siltstone floor of Newvale section.

The roof of the seam in both sections is a framework-supported polymictic conglomerate, with a variety of different lithic fragments representing chert and acid volcanic materials (cf. Ward et al., 1986) derived from the older strata in the New England Fold Belt. At the Catherine Hill Bay exposure this unit is around 25 m in thickness (Diessel and Hutton, 2004). Mineralogically, the material forming the immediate roof of the Newvale section contains abundant quartz and plagioclase feldspar (albite), together with some muscovite and small percentages of chlorite and siderite (Table 6.3). This represents a combination of the minerals in both the framework particles and the finer-grained matrix material. As discussed above, the identification of albite as the dominant feldspar is confirmed by the high proportion of Na2O indicated by XRF analysis.

The mudstone that is exposed between the conglomerate and the main part of the coal bed in one part of the Catherine Hill Bay was also evaluated. The mudstone in this part of the sequence consists almost entirely of smectite (Table 6.4), with minor proportions of kaolinite, quartz and possibly a trace of quartz-feldspar. Oriented aggregate XRD data

160

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam shows that the <2 micron fraction consists almost entirely of smectite, along with trace kaolinite.

Table 6.3 Mineralogy of Great Northern coal LTAs and non-coal rock samples from Newvale No.1 Colliery by XRD and Siroquant (wt. %)

Dol K- Sample LTA Qtz Kao I I/S M Chl Sid Alb Ana Goy Fap Bass /Ank feld

21379 - 48 10.8 4.3 7.9 2.8 1.3 24.9 21380 68.2 3.2 2.3 1.4 83.5 9.2 0.4

21381 14.4 6.3 83.6 6.2 1.8 2

21382 - 2.5 87.1 0.5 9.1 0.7

21383 10.6 8.8 77.3 8.2 3.3 2.4

21384 - 4.9 83.8 9.7 1.6

21385 11.6 15 74.6 2.9 3.3 2.9 1.3

21386 8 11.2 59.8 12 5.7 1.2 2 2.8 5.4

21387 12.8 14 61.9 8.5 1.3 1.7 10.4 0.7 1.5

21388 10.2 19.8 28.4 30.6 1.6 12.1 6.8 0.7

21389 - 13.1 44.8 6.9 34.3 0.9

21390 28.1 53.9 1.9 16.7 1.3 26.3

21391 - 65.9 6.1 10.9 17.2

Table 6.4 Mineralogy of Great Northern coal LTAs and non-coal rock samples from Catherine Hill Bay by XRD and Siroquant (wt. %)

Sample LTA Qtz Kao I I/S S M Chl K-feld Alb Ana Alun Mudstone roof - 1.3 7.7 90.3 0.7 GN27 16.2 51.7 35.3 11.6 1.3 GN26 16.6 28 59.1 13 GN25 14.1 23.3 61.9 2.4 10 2.4 Claystone 1 - 2 89 8.6 0.4 GN23 19.2 14 72.1 11.9 2 GN22 11.1 15 70.4 12.1 2.6 Claystone 2 - 2.3 77.2 18.7 1.8 GN19 14.2 22.3 62.3 2.7 9.9 2.9 GN17 13.9 23.1 62.8 11.4 2.7 GN16 18.6 24.9 57.2 8.4 9.4 GN11 12.2 34 39.4 20.7 5.9 GN10 15.4 66.5 21.7 8.1 1.8 2 GN9 23.4 57.2 19.9 10.6 5.8 6.5 GN7 24.2 58.2 28.9 12.8 GN6 53.3 86.9 9.3 3.8 Siltstone floor - 34.9 33.6 1.4 19.8 4.8 5.4 Mudstone floor - 20 24.2 46.1 5.1 4.6 Abbreviations same as in Table 6.3.

161

Table 6.5 Mineralogy of <2 μm fraction of coal LTAs and non-coal strata using oriented aggregate XRD techniques (wt. %)

Expandable clay Sample K(+Chl) (%) illite (%) Nature of expandable clay minerals minerals (%) Great Northern seam in Newvale No.1 Colliery 21379 54.4 30 15.6 Regular I/S 21380 74.6 11.5 13.9 Regular I/S 21381 93.3 6.7 Regular I/S 21382 100 - 21383 93.5 6.5 Regular I/S+smectite (minor) 21384 96.7 3.3 21385 94.9 5.1 Smectite+ regular I/S (minor) 21386 100 - 21387 87.5 4.9 7.7 Smectite 21388 56.9 9.6 33.5 Smectite 21389 84.9 3.7 11.3 Smectite 21390 31.8 11 57.2 Smectite 21391 28.9 38.5 32.6 Irregular I/S

Great Northern seam at Catherine Hill Bay Mudstone roof 1 99 Smectite Claystone1 100 Claystone 2 100 Siltstone floor 72.6 4.5 22.9 Smectite Mudstone floor 41.7 58.3 Smectite

6.3.2 Minerals in non-coal partings

The intra-seam partings in the Newvale section are represented by carbonaceous mudrocks with ash yields of 52-69% (Table 6.1). XRD analysis (Table 6.3) indicates that the upper two partings (samples 21382 and 21384) both have similar mineral assemblages, with dominant kaolinite, minor quartz and K-feldspar, and traces of anatase and possibly siderite. The two claystone partings analysed in the section at Catherine Hill Bay, correlating with these two horizons (Figure 4.3) also have a similar mineralogy. The mineral assemblage in the lowermost parting at Newvale, however (sample 21389), has a lesser proportion of kaolinite and more quartz, as well as abundant K-feldspar and some mixed-layer I/S.

6.3.2.1 Kaolinite

As with the LTA of the coal plies, the XRD patterns show that the upper two partings at Newvale (21382 and 21384) are dominated by well-ordered kaolinite. The oriented- aggregate XRD data indicate that kaolinite makes up 96.7% of the clay mineral assemblage in the upper two claystones (Table 6.5). A graphical profile of the clay

162

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam mineralogy at Newvale, based on XRD data from oriented-aggregates of the LTAs and non-coal rocks, is given in Figure 6.5. However, kaolinite in the lowermost parting (21389) is mostly disordered in nature, and is similar to that of the LTA from the coal samples in the lower part of the seam.

SEM observations show that the kaolinite in samples 21382 and 21384 is represented mostly by graupen to vermicular forms (Figures 6.6A, B). A vermicular texture in matrix kaolinite is often used as evidence that the sediment is a tonstein (Spears, 1971; Ruppert and Moore, 1993). As defined by Bohor and Triplehorn (1993), among others, such materials generally represent altered air-fall volcanic ash layers deposited in a non-marine, commonly coal-forming environment.

Numerous papers have been published on tonsteins interbedded with coal seams (Addison et al., 1983; Hill, 1988; Burger et al., 1990; Burger et al., 2000; Knight et al., 2000; Burger et al., 2002; Dai et al., 2011), and the mineralogical and geochemical relationships between them and the associated coals (Crowley et al., 1989; Crowley et al., 1993; Hower et al., 1999a; Brownfield et al., 2005). In some cases volcanic ash was not intensive enough to form visible ash layers in the coal beds (Dewison, 1989; Dai et al., 2008b), or only clayey micro-sized bands (around 100 microns in thickness) are developed (Dai et al., 2007b). Nevertheless, the volcanic material may still have an important impact on the coal geochemistry.

The common presence of vermicular kaolinite in the partings of the Great Northern seam provides evidence of in situ alteration (Zhou et al., 1982), and the partings may thus have formed by extensive leaching of easily degraded volcanic ash material. The vermicular kaolinite in the tonstein partings appears to be rich in titanium, probably in the form of anatase inclusions (Figure 6.6C). Kaolinite with crystallites of anatase was identified in Indonesian tonsteins by Ruppert and Moore (1993). The kaolinite in sample 21389 is mostly in the form of clasts, with rare platy crystals, but nevertheless also contains anatase inclusions (Figure 6.6D).

163

Figure 6.5 Vertical column section showing variations in clay mineralogy for the Great Northern seam at Newvale.

Tabular or vermicular kaolinite phenocrysts have been regarded as kaolinized biotite by some researchers (e.g. Knight et al., 2000). However, the diagenetic textures of the vermicular kaolinite crystals, and their relationship, if any, with the original volcanic texture, still need to be established (Spears, 2006).

Tonsteins, perhaps more precisely described as kaolinite-tonsteins, have also been reported in other Late Permian coal seams of the Sydney Basin (Loughnan, 1971a; Loughnan and Ward, 1971; Loughnan and Corkery, 1975; Creech, 2002), and several intervals of ash-fall and ash-flow tuff are known to occur in other parts of the coal-bearing Sydney Basin sequence (Diessel, 1965; Loughnan and Ray, 1978; Diessel, 1983; Agnew et al., 1995; Kramer et al., 2001; Grevenitz, 2003). However, although tonsteins of apparent volcanic origin have been reported to occur in coal seams around the world, other processes may also be responsible for the thin beds of non-coal sediment commonly found within coal seams (Ward, 2002). Evidence of a volcanic origin, such as the textural features described above, are therefore significant in determining the origin of the kaolinite-dominated bands in the present study.

164

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

Figure 6.6 SEM images of kaolinite in claystone partings from Newvale. (A) Graupen to vermicular kaolinite in 21382. (B) Graupen kaolinite in sample 21384. (C) Typical vermicular kaolinite with inclusions of anatase in sample 21384. (D) Platy kaolinite with fine anatase inclusions in sample 21389.

6.3.2.2 Quartz

Despite the relative abundance of K-feldspar, only very small proportions of quartz (1.3% and 4.9%) are present in the upper two partings at both locations (samples 21382 and 21384 from Newvale and claystone-1 and claystone-2 from Catherine Hill Bay). The quartz in the upper two partings at Newvale mostly occurs as euhedral crystals (Figure 6.3B). It may be beta-form quartz, but that would need to be confirmed by cathodoluminescence emission analysis. Quartz is more abundant (13.1%) in the lowermost parting at Newvale (sample 21389). Although mainly made-up of epiclastic material, this sample contains occasional volcanic quartz grains (Figure 6.4B).

The proportion of quartz appears to be lower in the partings compared to the LTA of the adjacent coal plies, especially if allowance is made for dilution of the silicates in the coal plies by authigenic carbonate components. Although a lesser proportion of quartz than in normal mudrocks is not necessarily a characteristic specific to tonsteins, it serves in conjunction with other evidence to identify a volcanic origin (Bohor and Triplehorn, 1993).

165

Together with the modes of quartz occurrence as cell infillings in the coal, this may indicate that the volcanic ash that formed the partings, especially the upper two partings, contained little if any free quartz. Such material may also have been admixed with the peat itself as the seam developed, but if so additional quartz was possibly added to the peat by precipitation of silica in the pores of the maceral components. Some of this silica may have been released into the peat waters by break-down of volcanic glass in the coal- forming mire environment.

6.3.2.3 K-feldspar

K-feldspar is present in all partings sampled from the seam at Newvale and Catherine Hill Bay. The XRD patterns of these samples display peaks at d-spacings of 3.29 Å, 3.25 Å, 2.90 Å and 2.60 Å that may be attributed to sanidine, anorthoclase or orthoclase. No single mineral, however, matches all of these peaks, which may indicate a mixture of two or more K-feldspar components, or possibly a range of feldspar structures (sanidine- anorthoclase series) in the parting samples.

Although comparison with the XRF data (see discussion above) confirms that the material is essentially a K-feldspar, EDS analysis of individual particles within the parting samples (Table 6.6) indicates that the feldspars also contain a small proportion of Na. Deer et al. (1992) noted that similar proportions of Na may be found in K-feldspars generally. The Na may exsolve and be represented by micro-perthite in feldspars of plutonic origin, but may remain incorporated in the lattice of more quickly-cooled volcanic feldspars such as sanidine.

Table 6.6 EDS micro analyses of K-feldspar in Great Northern non-coal samples from Newvale No.1 Colliery (%, O by difference)

Parting 21382 Parting 21384 Parting 21389 Siltstone floor

Element (17 points) (15 points) (5 points) (18 points)

mean max min mean max min mean max min mean max min

Al 10.6 11.3 9 7.9 9.5 7 9.8 10.5 9.3 9.9 10.5 9

Si 28.4 30.1 26.1 24.1 30 20.9 32.5 33 32.1 32.3 34.3 30.9

O 48.5 53.1 43.1 60.8 65.6 48.7 44.4 45.2 42.5 44.4 45.7 43.5

K 9.6 14.7 8.4 5.4 11.3 3.2 12.9 14.7 11.8 13.3 15.2 11.7

Na 2.9 3.5 1.4 1.8 3.1 0.3 0.3 0.7 0 0.1 0.3 0

K/(K+Na) 0.8 0.9 0.7 0.7 1 0.5 1 1 0.9 1 1 1

166

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

SEM examination shows that the feldspar in the two upper partings at Newvale (21382 and 21384) mostly occurs as idiomorphic crystals (Figures 6.7A, B). As discussed above, EDS data indicates a range in compositions from nearly pure potassium feldspar to sodium-bearing K-feldspar with Na being up to 3.5% in single K-feldspar crystals (Table 6.6). This range may represent an anorthoclase-sanidine series or a sodic sanidine material, indicating an acid to intermediate volcanic ash input to form the upper two partings in the coal seam.

The SEM study also shows that some K-feldspar appears to have been etched (Figure 6.7C), especially in sample 21384. Based on EDS data, the etched K-feldspar appears to be free of Na, but some un-etched K-feldspar particles (not all of them) contain both K and Na. Etched feldspar and quartz grains in pelitic layers of pyroclastic origin have also been observed by Ruppert and Moore (1993), who suggested that they had been leached in the acid-rich mires that formed the associated coal seams. Etched K-feldspar was also observed in high-sulphur coals from Yanshan, Yunnan, southwestern China (Dai et al., 2008a).

A small amount of K-feldspar is also present as fragmented aggregates in the upper two partings (Figure 6.7D). Hence, although the common presence of anorthoclase-sanidine series material is suggestive of a dominant volcanic origin, the presence of detrital K- feldspar indicates that the upper two partings could be mixed with minor epigenetic clasts. Mixing of mineral sources could have occurred if the relevant ash-fall had temporarily slowed or halted peat accumulation (Ruppert and Moore, 1993).

Unlike that in the upper two partings, K-feldspar in the form of detrital grains is dominant in the lowermost parting at Newvale (sample 21389), along with poorly-ordered kaolinite in banded or massive aggregates (Figure 6.8A). This may indicate that K-feldspar formation was mainly by epiclastic input from a sediment source close to the peat deposit. Na occurs in much lower concentrations in the K-feldspar of this bed than in the upper two partings (Table 6.6).

Some K-feldspar in the lower parting at Newvale, however, occurs as distinctive veins cross-cutting organic stringers (Figure 6.8B) and cell-infillings (Figures 6.8C, D). EDS elemental maps of Al, K and Na (Figure 6.9) confirm that the veins consist essentially of pure K-feldspar with a negligible amount of Na, if any. EDS also indicates very rare albite coexisting with K-feldspar in the cleat. No feldspar was detected with a composition falling

167 between those of albite and K-feldspar. Although authigenic K-feldspar has been reported in sedimentary formations as overgrowths on detrital grains, as cementing materials and as veins (e.g. Hagen et al., 2001; Liu et al., 2003; Sandler et al., 2004), vein and cell- infilling K-feldspars have not been reported in coal deposits before. The apparent precipitation of K-feldspar in shrinkage fissures of vitrinite appears to indicate hydrothermal activity in the late syngenetic stage of peat development. Adularia, a low- temperature K-feldspar and a polymorph of orthoclase (Klein and Dutrow, 2007), has been found to be associated with hydrothermal mineralization, although most commonly in epithermal deposits (e.g. Simpson et al., 2001; Zhang et al., 2010).

Figure 6.7 SEM images of K-feldspar in two uppermost partings from Newvale. (A) Euhedral K-feldspar (F) (? sanidine) in parting 21382. (B) Euhedral K-feldspar (F) and quartz (Q) in 21384. (C) Etched K-feldspar in 21382. Fine white grains are anatase. (D) K-feldspar (F) fragments in clastic kaolinite (K) matrix in 21384.

Veins consisting mainly of K-feldspar also occur in thin bands or laminae that are parallel to the bedding of the organic matter (Figure 6.8E) in sample 21389. This texture is similar to that of the K-feldspar laminae in the siltstone floor of the Newvale section. Figure 6.8F shows an intergrowth of K-feldspar with lesser quartz. This intergrowth indicates that both minerals were precipitated contemporaneously from the same fluid.

168

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

Both Simmons and Browne (1997) and Craw (1997) suggest temperatures of 220°C- 300°C for adularia crystallisation from hydrothermal fluids. Since no conversion has been observed in the clay minerals surrounding the veins in the floor samples (Figure 6.8B), the veins most likely originated from low-temperature hydrothermal fluids in the late syngenetic stage, before the peat underwent diagenesis. The origin of the hydrothermal fluid is uncertain, but was probably associated with contemporaneous volcanic activity.

Figure 6.8 SEM images of K-feldspar in lowermost claystone parting 21389 from Newvale. (A) K-feldspar (F) and kaolinite (K) in banded or massive aggregates. (B) K-feldspar (F) veins cross-cutting organic stringers. Note a volcanic quartz (Q) in the upper left of the image. (C) Fusinite with mineral infillings. (D) Enlargement of an area in (C) showing K-feldspar (F), quartz (Q) and kaolinite (K). (E) K-feldspar and quartz in thin bands. (F) Enlargement of an area in (E) showing an intergrowth of K-feldspar (K) and quartz (Q).

169

Figure 6.9 Elemental maps of Al, Si, K and Na in K-feldspar veins cross-cutting organic matter in claystone parting 21389.

6.3.2.4 Anatase

Small proportions of anatase occur in all partings from both the Newvale (0.7-0.9%) and Catherine Hill Bay sections (0.4-1.8%). The anatase appears to be embedded between the kaolinite platelets; it also commonly occurs as crack infillings in the kaolinite matrix and as fine grains in kaolinite aggregates (Figures 6.8C, D). Discrete crystals of anatase (Figures 6.10A, B) also occur. Sample 21384 contains anatase that appears to have replaced coal maceral components (Figures 6.10C, D).

Anatase is a common secondary mineral in tonsteins (Spears and Kanaris-Sotiriou, 1979), and is inferred to represent reprecipitation products of chemically leached labile components in the original ash material (Triplehorn et al., 1991). The titanium may have been derived from the breakdown of Ti-rich volcanic glass, ilmenite, magnetite or rutile (Ruppert and Moore, 1993).

170

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

Figure 6.10 SEM images of anatase in claystone parting 21384. (A) Anatase grains (A) and K-feldspar (F). (B) Apatite (Ap) and anatase (A) in a kaolinite matrix. (C) Anatase (A) replacing macerals. (D) Enlargement of an area in (C).

6.3.2.5 Other minerals

Although not in sufficient concentration to be detected by XRD analysis, apatite was identified by SEM studies in the kaolinite matrix of sample 21384 (Figure 6.10B). Euhedral apatite observed by other authors (Knight et al., 2000), however, was not identified. A phosphate particle containing Fe, Mn, Mg and Ca was detected in sample 21384 (Figures 6.11A, B), but not in any of the other partings studied.

No other phases reported elsewhere to be diagnostic indicators of volcanic origin (e.g. biotite, zircon, pyroxene and glass shards) were observed in the partings of the seam section studied, presumably because of the extensive alteration involved.

171

Figure 6.11 (A) SEM image of Fe, Mn, Mg, Ca phosphate (point 1) and K-feldspar (point 2) in kaolinite matrix of claystone parting 21384. (B) EDS spectrum of point 1.

6.3.3 Minerals in coal samples

The coal plies at Catherine Hill Bay have higher LTA percentages than those sampled at Newvale, possibly due in part to oxidation of the organic matter in the outcropping coal seam. Similar quartz and clay mineral assemblages, however, are present in the LTAs of the individual coal plies at both locations (Tables 6.3, 6.4).

The LTA percentage of the lowermost coal plies in each section is very high, with a little over 50% at the immediate base of the seam at Catherine Hill Bay and 25-30% in the overlying parts of both seam sections. The coal plies in the middle and upper parts of the seam in both sections have significantly lower LTA percentages. An exception is the topmost ply in the Newvale section, which has an LTA yield of almost 70%. However, as discussed further below, the LTA of this ply consists almost entirely of authigenic carbonate minerals (dolomite/ankerite), in contrast to the clay-dominated assemblages of the other ply samples.

6.3.3.1 Clay minerals

As indicated in Table 6.3, the clay minerals in the coal plies in both sections are represented by a combination of kaolinite and mixed-layer illite-smectite (I/S). With the exception of the lowermost coal ply at Newvale (sample 21390), kaolinite is the most abundant of these components. It is the only clay mineral identified by powder XRD techniques above the ply represented by sample 21387 in the Newvale section, and is the

172

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam dominant component (>60% of the LTA) in the upper plies of the Catherine Hill Bay seam section.

As noted for other coals in the Sydney Basin (Ward, 1989), the kaolinite in the coal plies, especially those in the upper part of the seam, has a powder XRD pattern indicating a well-ordered crystal structure. More poorly-ordered kaolinite, however, appears to be present in the LTA of the lower coal plies. SEM analysis shows that the kaolinite in the coals primarily occurs as thin bands intimately associated with vitrinite, and as vermicular crystals, cell cavity and cleat infillings (Figures 6.12A, B, C). Together with the well- ordered structure, this indicates that the kaolinite in the coals, especially in the upper part of the seam, was formed mainly by authigenic precipitation in the original peat swamp (Ward, 1989). The kaolinite-rich coal plies at Newvale (samples 21381, 21383 and 21385, Table 6.3) also have higher inertinite percentages than those in the lower part of the seam. These higher inertinite percentages are in turn associated with a higher macroporosity, which would have allowed more kaolinite precipitation in the cavities of the maceral components.

Some of the kaolinite in the lower part of the seam appears to represent pseudomorphs after biotite (Figure 6.12D). Although rare in comparison to the authigenic kaolinite, this may represent altered remnants of detrital sediment, or possibly volcanic ash, introduced to the coal swamp in the early stages of peat accumulation.

The non-kaolinite clay minerals, illite and I/S, mostly occur in the coal as thin bands and laminae, and are interpreted to be essentially of detrital origin. Although present in the upper and lower coal plies at Newvale (in the two coal plies near the floor they make up of roughly half of the clay fraction), these minerals are virtually absent in the middle part of the coal bed. This further suggests that the original peat included more admixed detrital minerals derived from the sediment source region supplied to the basin in the early and late stages of seam formation, before and after the peat swamp was fully established.

Oriented-aggregate XRD data indicate that the clay minerals in the roof and floor comprise roughly equal proportions of kaolinite, illite and expandable clay minerals (Figure 6.5). However, within the coals themselves, kaolinite is the dominant clay mineral, and only very small proportions of illite and expandable clay minerals are indicated by oriented-aggregate XRD techniques.

173

Oriented-aggregate XRD of the Newvale samples (Table 6.5) further shows that the nature of the expandable clay minerals varies through the seam section (Figure 6.5). The expandable clays consist mainly of smectite in the lower part of the coal bed, smectite together with regularly interstratified I/S in the upper part of the seam, and regular I/S only the uppermost layers. Figure 6.13 shows XRD traces of two coal LTAs from the upper and lower parts of the seam, after glycol saturation and heat treatment.

Figure 6.12 Kaolinite in coal sample 21388 from Newvale. (A) Vermicular kaolinite. (B) Cell infillings of kaolinite. (C) Cleat infillings of kaolinite. (D) Probable kaolinite pseudomorphs after biotite.

Figure 6.13 XRD traces obtained from <2 μm fractions of (A) coal sample 21383 and (B) 21388 after glycol saturation and heating at 400°C for 2h.

174

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

6.3.3.2 Quartz

Quartz is more abundant as a fraction of the LTA for the Catherine Hill Bay samples than for the Newvale samples. This may reflect a lateral change across the field, but may also partly reflect dilution of the mineral matter in the Newvale samples by epigenetic dolomite and/or ankerite; such carbonates are not present in the LTA of the coals from Catherine Hill Bay.

Quartz is particularly abundant in the lowermost ply of the Catherine Hill Bay section (sample GN6), and is more abundant in the lower part of the seam at both locations than in the upper part. For example, quartz makes up 15-20% of the LTA in samples 21388 to 21385 from Newvale, but <10% of the LTA of the coal plies above the parting represented by sample 23184. Davis et al. (1984) noted that quartz appears to be most abundant in the basal parts and near the margins of peat beds in the Okefenokee swamp-marsh complex, and suggested that it represents detrital quartz introduced to the peat by mixing with the sediment of the swamp floor, due to either bioturbation or contemporaneous clastic deposition. The LTA in the topmost ply at Catherine Hill Bay (sample GN27), however, has a relatively high quartz content (52%). If the mineralogy is calculated on a dolomite-free basis, quartz also appears to be very abundant in the topmost coal ply (sample 21380) of the Newvale seam section.

Although some detrital input may be involved, the mode of quartz occurrence favours a dominantly authigenic origin for the high quartz percentage in the lowermost coal (sample 21390) of the Newvale section. The quartz mostly occurs as cell and micropore infillings within the macerals (Figures 6.14A, B), and occasionally as detrital fragments in collodetrinite (Figure 6.14C). Given the similar occurrence of quartz and K-feldspar, the lowermost intra-seam band at Newvale has characteristics that more closely resemble the floor strata. Ward (1991) also noted a hard siliceous siltstone floor of a Tertiary coal seam of the Mae Moh Basin in Thailand, and suggested that an accumulation of authigenic or biogenic silica may partly be developed. Sykes and Lindqvist (1993) observed diagenetic quartz and amorphous silica of different forms in Tertiary coals from a number of New Zealand coalfields. They suggested that sub-horizontal silicification of some coals was due to the infiltration of the peat bed by silica-saturated groundwater and crystallization of quartz at greater depth; the origin of the silica may have been from leaching of the basement rocks or from siliceous phytoliths within the coal-forming plant material. In addition, the silica could have been derived from the alteration of volcanic glass, which may also lead to the formation of porcellanite or chert-like rocks (cf. Loughnan and Ray,

175

1978). The mineral matter in coals from Xuanwei, Yunnan, southwestern China also contains a high proportion of fine-grained quartz (Dai et al., 2008c; Large et al., 2009), which was deposited from silica-bearing solutions that originated from weathering of basaltic rocks in the hinterland (e.g. Ren, 1996).

Since the clay mineralogy of the coals in the upper part of the Newvale section (discussed below) is dominated by authigenic kaolinite, and the relative abundance of quartz in the middle part of coal seam is relatively low, authigenic processes are thought to have been dominant in the middle part of the original peat accumulation. Ruppert et al. (1991) also found authigenic quartz to be much more abundant in the laterally interior part than the marginal parts of the Upper Freeport coal of the Appalachian Basin of the eastern USA.

A quartz-rich rock fragment about 50 μm in diameter was observed under the SEM in the uppermost coal sample (21380) at Newvale (Figures 6.14D, E). EDS studies indicate that the rind of this fragment is composed of K-bearing aluminosilicate, possibly illite, while the interior is essentially composed of even-grained quartz crystals with some disseminated chlorite, the latter possibly being an alteration product of biotite. This rock fragment is thought to be a devitrified glass spherule, and was probably derived from the same source as the abundant acid volcanic rock fragments that make-up the overlying conglomerate bed.

6.3.3.3 K-feldspar

K-feldspar occurs in the lower part of the coal seam in both sections. It is especially abundant in the lowermost ply at Newvale (sample 21390) and the associated non-coal parting, but also occurs in the LTA of the coal plies in the lower part of both seam sections. XRD studies suggest that an albite-rich plagioclase, as well as K-feldspar, also occurs in the LTA of the lower plies at Catherine Hill Bay.

SEM studies indicate that the K-feldspar in both the lowermost coal ply and the adjacent non-coal parting at Newvale is intimately associated with laminae of mixed-layer illite- smectite (I/S), the distribution of which is discussed more fully below. The feldspar in some cases occurs as irregular to elongated aggregates of interlocking euhedral crystals (Figure 6.14F), partly cross-cutting the layering of the maceral components. This mode of feldspar occurrence is even more strongly developed in the lowermost intra-seam claystone band, and is discussed more fully below. Given the occurrence of similar veins in the floor sample as well, at least some of the feldspar in the coal is thus possibly of

176

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam hydrothermal origin. However, this does not exclude additional feldspar input from re- working of penecontemporaneous volcanic material, such as the debris that made up the underlying Awaba Tuff.

Figure 6.14 Minerals in coal samples from Newvale. (A) Quartz (Q) along with kaolinite (K) in cell lumens in coal 21388. (B) Quartz in maceral micropores in coal sample 21388. (C) Detrital quartz grains in collodetrinite in coal sample 21388. (D) Quartz-rich rock fragment in coal sample 21380. (E) Enlargement of (D) showing chlorite (C) and quartz (Q) crystals. (F) K-feldspar in coal sample 21390.

The abundances of both quartz and K-feldspar in the lowermost plies of the coal seam are comparable to those in the floor samples from Catherine Hill Bay. This suggests that the

177 basal part of the original peat bed was made-up of organic matter admixed with the same detrital sediment as was supplied to the basin before the swamp was established.

6.3.3.4 Siderite

Siderite is present in most of the coal plies in the seam section at Newvale (Table 6.3), but not in the intra-seam non-coal partings. It mainly occurs as syngenetic nodules, as chemically-zoned euhedral crystals (Figure 6.15A) and as cell infillings. EDS data indicate that some of the cell infillings have overgrowths (Figure 6.15B) with higher concentrations of Mn than the earlier-formed material.

The Great Northern seam was deposited under terrestrial conditions, with no indication of marine influence. Since the pore waters of the peat would thus have contained little if any dissolved SO4, the nodular siderite was probably formed by interaction of dissolved Fe with CO2 generated by fermentation of the organic matter shortly after peat accumulation (cf. Botz and Hart, 1983).

6.3.3.5 Dolomite/ankerite

Dolomite and/or ankerite (ferroan dolomite) occur in most of the coal plies of the Newvale section, mostly as epigenetic cleat infillings in the bright coal layers. Abundant veins of these minerals, oriented generally parallel to bedding, also occur in the topmost ply of the coal seam, immediately below the roof material.

Both the cleat infillings and the abundant veins are clearly post-depositional. As discussed by Ward (2002), the veins near the top of the seam, in particular, may have been derived from expulsion of organically-associated Ca and Mg from the macerals during rank advance. Studies of other coals using electron microprobe techniques (e.g. Ward et al., 2005; Li et al., 2007b; Ward et al., 2007) have shown that such organically-bound inorganic elements are liberated from coal macerals, especially vitrinite, as part of the molecular changes associated with the sub-bituminous to bituminous transition in the rank advance process.

No carbonate minerals have been observed in the LTA of the Catherine Hill Bay coal samples. This contrasts with their presence in most of the Newvale samples, and may be a result of weathering associated with exposure of the coal seam.

178

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

6.3.3.6 Other minerals

Apatite and alumino-phosphate minerals of the crandallite group are noted in the XRD data for sample 21386 from the Newvale seam section. The alumino-phosphate occurs with kaolinite as cleat or crack infillings, and also co-exists with kaolinite as cell infillings (Figure 6.15C). The EDS spectrum (Figure 6.15D) indicates that the material is a member of the crandallite series containing P, Sr, Ca and a trace of Ba, suggesting either a goyazite or a Sr-bearing crandallite (cf. Ward et al., 1996). Alumino-phosphate minerals would be expected from intra-seam precipitation if Al was also available in reactive form at the site of phosphate deposition, and apatite if Al was not available to react with the precipitated phosphatic material (Ward, 2002).

Minor proportions of bassanite are also present in some of the Newvale LTAs, especially those from plies with low ash yields and therefore abundant organic matter. As discussed by Frazer and Belcher (1973), the bassanite is most likely an artefact of the low- temperature ashing process, derived from the interaction of organic sulphur and organically associated calcium released by oxidation of the maceral components.

Bassanite was not found in the LTA of the coals from Catherine Hill Bay, and this may also assist to rule out the possibility of bassanite in the Newvale coals as an oxidation product during storage. Alunogen was frequently detected in the Catherine Hill Bay LTA samples. The alunogen may also have been produced by interaction of inorganic elements and organic sulphur during low-temperature ashing (cf. Ward et al., 2001), or it may have been formed in the outcropping coals by processes associated with weathering.

179

Figure 6.15 (A) Zoned euhedral siderite crystals in coal sample 21388. (B) Cell-infilling siderite overgrown by a later siderite in coal sample 21390. (C) Goyazite-crandallite (G) and kaolinite (K) cell infillings in coal sample 21386. (D) EDS spectrum of point 1 from (C).

6.4 Geochemistry of the Great Northern seam

6.4.1 Mineralogical and chemical analysis data

Major element chemical data for the high-temperature (815 qC) ashes (HTA) of the Great Northern samples in the Newvale section are given in Table 6.7, and those of the HTAs for the samples from Catherine Hill Bay are given in Table 6.8. In both cases the results have been expressed to an LOI- and SO3-free basis, to allow better comparison to the mineralogical data.

The relation between ash chemistry and mineralogy for the Newvale samples was studied to check the reliability of the quantitative XRD data, following the procedure described by Ward et al. (1999a). The inferred chemical composition of the mineral assemblages determined by Siroquant was calculated and compared with the actual chemical composition of the same samples as determined by XRF. The inferred chemical

180

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

+ composition was adjusted for each LTA sample by deducting the CO2 and H2O to derive an equivalent to a coal ash analysis. The actual chemical composition determined by XRF was also normalized to an SO3-free basis, to allow for differences in SO3 retention by the low- and high-temperature ashing processes.

Table 6.7 Major element analyses of Great Northern coal ash and non-coal samples from Newvale No.1 Colliery (%), as determined by XRF analysis (n.d. = not detected).

Sample HTA SiO2 Al2O3 TiO2 Fe2O3 MgO CaO Na2O K2O MnO P2O5 SO3 LOI 21379 - 71.82 12.73 0.58 3.17 0.96 0.30 2.36 2.35 0.059 0.08 0.04 4.38 21380 40.82 6.34 2.39 0.21 18.52 25.06 38.80 0.05 0.22 0.307 0.03 1.14 10.16 21381 11.95 48.45 34.34 1.08 2.81 2.71 4.35 0.20 0.86 0.028 0.02 4.48 n.d. 21382 - 27.43 20.73 0.76 0.50 0.18 0.14 0.12 0.90 0.007 0.02 0.66 48.84 21383 8.74 45.24 32.09 0.89 5.13 3.22 5.60 0.16 0.60 0.081 0.15 6.47 n.d. 21384 - 37.28 25.40 1.66 0.63 0.24 0.19 0.13 1.59 0.010 0.04 0.14 32.22 21385 9.90 53.17 32.48 1.15 5.06 1.56 2.03 0.16 0.74 0.075 0.07 2.91 n.d. 21386 6.54 42.24 28.43 1.65 7.46 3.38 5.72 0.17 0.69 0.122 0.55 7.27 3.83 21387 11.02 54.87 30.77 1.85 3.93 1.41 1.83 0.34 2.31 0.068 0.05 2.16 n.d. 21388 8.97 57.63 24.84 1.29 8.70 1.41 0.94 0.64 2.70 0.111 0.02 0.75 n.d. 21389 - 36.94 15.72 1.25 0.72 0.24 0.08 0.28 4.09 0.007 0.02 0.18 38.91 21390 27.41 78.84 11.69 0.49 1.52 0.30 0.13 0.28 5.32 0.025 0.03 0.08 n.d. 21391 - 44.88 9.97 0.26 0.60 0.25 0.04 0.18 2.68 0.006 0.01 0.24 42.69

Table 6.8 Major element analyses of Great Northern non-coal samples from Catherine Hill Bay (%), as determined by XRF analysis.

Sample SiO2 Al2O3 TiO2 Fe2O3 MgO CaO Na2O K2O MnO P2O5 SO3 LOI Mudstone roof 53.14 21.58 0.26 3.57 2.94 0.15 0.20 2.14 0.02 0.02 -0.04 15.59 Claystone 1 39.12 29.56 1.10 0.50 0.27 0.04 0.11 1.19 0.01 0.03 0.09 27.24 Claystone 2 46.32 33.60 1.83 0.73 0.20 0.02 0.01 1.11 0.01 0.03 -0.01 15.44 Hard floor 69.35 17.45 0.45 0.90 0.56 0.11 0.46 1.32 0.01 0.02 0.00 8.99 Mudstone floor 61.74 18.07 0.29 2.23 1.75 0.36 0.20 1.37 0.00 0.01 0.11 14.44

The percentages of each element indicated by both sets of data were plotted against each other (Figure 6.16), to provide a basis for comparing the XRD results to the chemical analysis data for the same coal or parting samples. As discussed for other materials by Ward et al. (1999), the respective data sets are presented as X-Y plots, with a diagonal line on each plot indicating where the points would fall if the estimates from the two different techniques were equal.

The plots for SiO2, Al2O3, CaO and MgO in Figure 6.16 show that all points fall very close to the diagonal equality line, indicating that the Siroquant results are consistent with the ash chemistry. With one exception, the plot for K2O is also consistent for those samples with K2O greater than 1%. This confirms the identification of K-feldspar in the XRD

181 analysis. The remaining samples, however, have low proportions of K-feldspar, and the

K2O inferred from Siroquant is based mainly on the illite or I/S content. The proportion of

K2O inferred from the XRD data is based on the assumption that the illite is fully saturated with K+ ions; if this is not the case (as would be expected for sedimentary materials), the proportion of K2O indicated by Siroquant will be less than that determined directly by XRF analysis.

The proportion of Na2O is low (<1%) and under-estimated by Siroquant for most samples. This may in part be explained by the presence of minor Na in the K-feldspar based on EDS data (see below), which was not allowed for in calculating the inferred chemical composition. A sample of the conglomerate roof material, however, which is indicated in Table 6.3 as having a significant proportion of albite (Na-feldspar), is the exception to this trend, and plots very close to the equality line.

182

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

Figure 6.16 Comparison between proportions of major element oxides in coal LTAs and non-coal strata from Newvale, inferred from Siroquant and determined by XRF. The diagonal line represents equality in each plot. Relevant trendlines and squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case.

Although there is a broad positive correlation between the XRD and XRF results for Fe2O3, most points plot below the equality line. This may be because the dolomite identified by the XRD data in most samples also contains some Fe, which was not allowed for in the

183 calculation process. Alternatively, some of the iron in the rocks or the LTAs may occur in non-crystalline oxy-hydroxide form. The proportion of TiO2–bearing minerals also appears to be under-estimated by the XRD analysis. However, SEM analysis shows that the Ti- bearing minerals occur largely as fine-grained crystallites (<0.5μm) in the kaolinite matrix or vermicular aggregates, and because of this fine grainsize, possibly poor crystallinity and probably small percentages, such materials are difficult to evaluate by XRD analysis.

6.4.2 Major element oxides in coals

The major oxides in the coal ashes are dominated by SiO2 and Al2O3. The concentrations of MgO, CaO, Fe2O3, K2O and TiO2 in most coal plies are within 5%, and those of the rest major oxides are within 1%. Abundant CaO, MgO and Fe2O3 in the uppermost coal plies in both seam sections represent ankerite, dolomite and siderite. High proportions of K2O reflect the greater abundance of smectite and especially K-feldspar in the relevant coal and non-coal materials. Mn is as high as 0.12% in the topmost coal ply in the Great Northern seam. SEM-EDS results indicate that Mn mainly occurs in carbonates (dolomite, ankerite and siderite). The TiO2 concentrations in the partings are higher than in the roof and floors (normal sediments), which is consistent with the occurrence of anatase in the partings. One coal ply (21386) contains the highest proportion of P2O5 (0.04% of the HTA) among all the coal samples, which is mainly due to the presence of goyazite and fluorapatite.

6.4.3 Associations of element components and minerals

Trace element data for the Great Northern coal and non-coal samples are given in Table 6.9. The concentrations of most trace elements in the Great Northern coals are lower than that of average worldwide coals (Ketris and Yudovich, 2009). Only Sc, Y, Sn, Cs, and Hf are generally slightly higher in the Great Northern coals than the averages in the worldwide coal dataset.

Boron in the Great Northern coals varies from 19.9 to 52.5 ppm, indicating that the seam was mainly influenced by fresh water, according to the classification by Goodarzi and Swaine (1994). Chlorine is generally low in the Great Northern coals (60-260 ppm). The highest concentration of Cl occurs in uppermost coal ply of the Great Northern seam, the mineral assemblage of which is dominated by dolomite, although Cl is not correlated with carbonates. No obvious correlation between Cl and ash yield is observed.

184

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

To evaluate the relationship between individual elements and minerals in the Great Northern coals, correlation coefficients were calculated between the proportions of individual elements and minerals on the whole coal basis. The major oxide percentages determined by XRF analysis of the coal ashes were recalculated to give the percentages of those oxides in the whole coal, which are also given in Table 6.9.

6.4.3.1 Elements associated with kaolinite

The lithophile element Li shows a strong positive correlation with kaolinite in the Great Northern coals, with a coefficient (R) of 0.97 (Figure 6.17A). The correlation trend line intersects the origin of the graph, which indicates the occurrence of Li almost entirely associated with kaolinite. However, lithium is not correlated with ash yield in the Great Northern coals (R=-0.65). This may be due to the lack of correlation between kaolinite and ash yield. Dai et al. (2012a) noted that, apart from kaolinite and possibly illite, a large proportion of Li occurs in cookeite (Li-bearing chlorite) in coals from the Guanbanwusu Mine, Inner Mongolia, China, where the individual coal samples contain Li concentration up to 505 ppm. Tourtelot and Brenner-Tourtelot (1978) reported kaolinite clayrocks from some states of the US that contain >2000 ppm Li, and the Li mainly occurs in cookeite in those samples.

Like Li, lithophile elements, such as Th and U, are positively correlated with kaolinite in the coals, having correlation coefficients of 0.51 and 0.58 respectively. Th and U also show no apparent correlation with ash yield (R = 0.01 and -0.12, respectively).

Besides the lithophile elements above, chalcophile elements Se, Pb and Cu also appear to be correlated with kaolinite, having relatively strong correlation coefficients of 0.74, 0.7 and 0.65, respectively (Figures 6.17B, C, D). These chalcophile elements may be associated with kaolinite in the Great Northern coals. Alternatively, the correlations between chalcophile elements and the kaolinite probably indicate a common source.

185

Table 6.9 Major oxide and trace element analyses of Great Northern samples in the Newvale section (Major oxides in %, recalculated from the data by XRF analysis; trace elements in ppm; all data on whole-coal basis).

21379 21380 21381 21382 21383 21384 21385 21386 21387 21388 21389 21390 21391

SiO2 72.68 2.51 5.83 27.35 3.97 37.46 5.29 2.72 6.07 5.22 37.53 21.89 44.09

TiO2 0.58 0.08 0.13 0.76 0.08 1.67 0.11 0.11 0.20 0.12 1.26 0.14 0.25

Al2O3 12.89 0.94 4.13 20.66 2.82 25.52 3.23 1.83 3.40 2.25 15.97 3.25 9.80

Fe2O3 3.21 7.32 0.34 0.50 0.45 0.63 0.50 0.48 0.44 0.79 0.73 0.42 0.59 MgO 0.97 9.91 0.33 0.18 0.28 0.24 0.16 0.22 0.16 0.13 0.25 0.08 0.24 CaO 0.30 15.34 0.52 0.14 0.49 0.19 0.20 0.37 0.20 0.08 0.08 0.04 0.03

Na2O 2.38 0.02 0.02 0.12 0.01 0.13 0.02 0.01 0.04 0.06 0.28 0.08 0.18

K2O 2.38 0.09 0.10 0.90 0.05 1.60 0.07 0.04 0.26 0.24 4.15 1.48 2.63 MnO 0.060 0.121 0.003 0.007 0.007 0.010 0.007 0.008 0.008 0.010 0.007 0.007 0.006

P2O5 0.078 0.013 0.002 0.019 0.013 0.035 0.007 0.036 0.006 0.001 0.018 0.008 0.010 Li 17.4 3.45 19.2 176 15.4 111 17.1 10.4 13.6 5.30 38.2 5.36 11.2 Be 1.19 4.88 6.26 2.10 0.86 0.65 0.81 0.33 0.63 0.66 1.54 2.47 3.01 B <10 19.9 30.0 26.5 31.6 21.2 34.2 52.5 44.4 40.7 24.4 36.3 15.4 %Cl 0.034 0.026 0.011 0.021 0.009 0.177 0.023 0.020 0.006 0.020 0.090 0.019 0.031 Sc 82.8 8.39 6.78 35.9 5.23 44.3 7.85 3.87 6.77 6.40 47.6 22.3 57.1 V 64.8 79.6 15.1 20.7 18.3 79.8 54.1 15.1 13.2 17.0 32.2 40.4 56.0 Cr 43.1 52.5 7.01 11.0 3.30 13.0 4.99 2.45 2.71 5.13 18.9 27.0 44.6 Co 64.8 9.33 3.19 9.48 14.5 19.3 3.36 3.04 3.40 5.46 29.2 7.04 6.23 Ni 58.3 7.66 2.09 20.0 11.10 7.17 4.14 0.82 1.12 1.71 25.9 0.40 1.82 Cu 28.0 3.66 6.42 17.5 5.63 101 6.19 7.02 4.85 4.05 31.9 4.65 18.2 Zn 83.5 69.1 8.68 30.9 9.89 49.0 20.9 11.2 10.2 12.2 45.2 20.7 21.4 Ga 31.3 15.0 7.98 38.3 4.95 40.4 4.76 6.73 7.14 5.49 38.4 11.1 21.8 Ge 1.65 6.44 7.17 2.11 1.78 0.86 0.68 0.45 0.38 0.48 0.77 2.02 6.35 As 4.27 0.27 0.32 0.58 0.39 2.16 0.43 0.43 0.47 0.36 1.62 0.60 1.02 Se 0.12 0.18 0.44 0.84 0.40 0.62 0.73 0.43 0.52 0.26 0.16 0.25 0.22 Rb 80.4 4.0 4.17 29.7 2.16 50.8 2.76 2.77 11.1 18.0 157 56.2 126 Y 14.3 22.0 18.0 12.1 8.25 5.68 7.8 5.01 11.2 9.60 20.8 22.1 27.8 Zr 198 70.1 57.1 358 46.3 150 36.4 29.4 83.8 38.4 190 67.2 107 Nb 5.06 0.88 2.79 15.2 1.48 7.55 1.40 1.28 4.46 1.26 9.77 2.29 4.64 Mo 0.02 <0.1 0.40 0.79 0.72 1.20 0.45 0.60 0.71 1.32 1.38 2.88 1.42 Ag 0.14 0.02 0.07 0.47 0.04 0.21 0.04 0.03 0.09 0.03 0.22 0.06 0.11 Cd 0.18 0.08 0.07 0.45 0.08 0.31 0.09 0.05 0.11 0.07 0.23 0.09 0.22 Sn 3.00 0.80 10.6 8.93 0.67 4.54 1.17 1.02 1.68 0.81 4.12 37.7 3.09 Sb 0.45 5.02 2.05 0.80 0.30 0.07 0.16 0.11 0.12 0.11 0.12 0.44 3.49 Te 0.31 <0.1 <0.1 0.17 <0.1 0.21 <0.1 <0.1 <0.1 <0.1 0.12 <0.1 0.09 Cs 2.50 0.47 0.52 2.57 0.28 3.47 0.25 0.55 1.14 3.22 7.26 3.97 8.32 Hf 8.74 2.19 2.44 15.9 1.94 6.77 1.46 1.20 3.14 1.50 8.34 3.00 4.75 Ta 0.13 <0.2 <0.2 1.22 <0.2 0.49 <0.2 <0.2 <0.2 <0.2 0.73 <0.2 0.14 W 1.47 0.11 1.09 2.55 0.63 2.41 0.24 0.41 0.57 0.72 2.83 3.45 10.4 Hg 0.06 0.04 0.02 0.10 0.01 0.11 0.03 0.02 0.03 0.01 0.11 0.02 0.05 Tl 0.57 0.10 0.17 0.43 0.30 1.29 0.40 0.69 0.53 0.38 1.24 0.51 0.87 Pb 10.7 2.82 11.9 36.4 6.83 9.98 7.12 3.72 4.09 3.03 6.67 6.42 13.0 Bi 0.16 0.20 0.43 1.51 0.33 0.43 0.49 0.32 0.34 0.26 0.28 0.21 0.34 Th 8.42 2.14 5.35 43.5 3.96 10.6 2.73 1.59 2.38 1.49 3.85 3.86 10.6 U 2.51 0.48 1.75 7.81 1.26 3.60 0.65 0.46 0.70 0.42 1.43 1.04 3.01

186

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

Figure 6.17 Correlation of elements (Li, Se, Pb and Cu) with kaolinite in the Great Northern coals in the Newvale section. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case.

6.4.3.2 Elements associated with ash yield

Vanadium and Cr show significant correlations with each other, and with the ash yield (R=0.72 and 0.92), although no obvious correlation was observed between them and any specific minerals except kaolinte. Vanadium and Cr have poor correlations with the clay minerals, though both clay-associated and organic-associated V has been reported in other coals. Vanadium tends to be organic-associated in lower-rank coals and clay- associated in higher-rank coals (Ren et al., 2006).

6.4.3.3 Potassium, Rubidium and Sodium

A certain Na2O/K2O ratio is indicated in Figure 6.18A when the coal samples do not contain orthoclase (K22 Figure 6.18A). Sodium may substitute for K in orthoclase, which results in higher concentrations of Na2O in the relevant samples. Rubidium shows a very strong correlation with potassium (Figure 6.18B), having a correlation coefficient of 0.97. Rubidium does not form minerals of its own, but is always camouflaged in K- minerals, because of its similarity with K (Heier and Adams, 1964). This is consistent with results noted by Ward et al. (1999).

187

6.4.3.4 Silver and niobium, scandium and silicon

Sc exhibits a quite strong positive correlation with silica (Figure 6.18C), with a coefficient of 0.93. This may be due to the spectral interference of Si on Sc in ICP-MS analysis. Relatively high concentrations of Si in the coal samples may result in a spectral interference of Si on Sc. Silver exhibits a strong correlation (R=0.92) with niobium (Figure 6.18D). Very similar correlation between these elements also exists in coals from the Songzao coalfield (Chapter 8). This is also probably because the value for Ag concentration was subjected to spectral interference caused by Nb in ICP-MS analysis (e.g. Guo et al., 2011).

6.4.4 Selected elements in non-coal samples

The assemblage of the relatively immobile elements such as Y, Zr, Th, Ga and Ti, along with the REE, can be used to determine the composition of the original ash and thus the type of the source magma (Burger et al., 2002). In the study of Spears and Kanaris-

Sotiriou (1979), tonsteins with TiO2/Al2O3 values of <0.02 and >0.07 are grouped to indicate parent magmas of acid and mafic composition, respectively; those with values in

188

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

between are thought to represent intermediate ash materials. The comparison of TiO2 and

Al2O3 in the Great Northern non-coal samples was potted in Figure 6.19A. All the partings fall between the two lines indicating TiO2/Al2O3 values of 0.02 and 0.07, and are thus suggested to be mainly derived from intermediate volcanic ash. The key parameters for the partings were also plotted in the magma source discrimination diagram of Winchester and Floyd (1977) (Figure 6.19B). The three claystones fall in the andesite, trachyandesite and alkali basalt fields, respectively. However, this is not exactly consistent with the results indicated by Figure 6.19A, which may partly be due to possible contamination of the samples from the zirconia grinding mill during sample preparation.

The roof and floor samples are plotted between lines indicating TiO2/Al2O3 values of 0.07 and 0.02 in the Figure 7.7.A, and in the rhyodacit/dacite fields in Figure 7.7B. These materials are suggested to be derived probably from source material of silicic to intermediate compositions.

Nb and Ta are lower in Great Northern tonsteins (up to 29.3 and 2.35 μg/g, organic-free basis) which is comparable to the silicic tonsteins from the Songzao Coalfield reported by Dai et al. (2011). Nb and Ta often substituted in Ti minerals or enriched in clay minerals and remain immobile to weathering and alteration (Zhou et al., 2000). The concentrations of Nb and Ta are much higher in the alkali tonsteins from the Songzao Coalfield (Chapter 8) and other alkali tonsteins from SW China (Zhou et al., 2000).

Trace elements, including Li, Se, Nb, Th, and U, are generally enriched in the tonstein samples, relative to these elements in the roof and floor samples. The influence of the tonstein bands on the geochemistry of the adjacent coals is not as significant as in the Songzao coal seams (Chapter 8). Trace elements, such as Li, Th, and U, are relatively high in most of the coal plies adjacent the acid tonsteins, relative to those away from the tonsteins.

189

(B) Plot of Zr/TiO2 against Nb/Y ratios. Magma source discrimination diagram of Winchester and Floyd (1977).

6.4.5 Distribution and affinity of REE and Y

7KH WRWDO UDUH HDUWK HOHPHQW FRQWHQWV 5((< LQ WKH *UHDW Northern coal samples typically range between 25.9 and 90 ppm (Table 6.10). The REE concentrations were normalised against the Upper Continental Crust (UCC) (Taylor and McLennan, 1985) for each coal and associated non-coal sample, in order to obtain a more clear indication of the distribution patterns (Figure 6.20).

The normalized REE distributions in all the Great Northern coal samples have LaN/LuN values less than 1 (Table 6.10), indicating heavy REE enrichment (H-type) compared to the UCC (Seredin, 2001). Eskenazy (1999) ascribed the enrichment of HREE relative to LREE to the formation by HREE of complexes with organic matter in the coal, which would also increase the stability of the HREE relative to the LREE. As discussed by Seredin (2001), H-type distribution patterns may be attributed to the circulation of water which was enriched in HREE through coal basins. This is consistent with the occurrence of epigenetic minerals (mainly carbonates) in the Great Northern coals.

As indicated in Figure6.21 the correlation coefficient between REE and the ash yield appears to increase with the increasing atomic number of REE, especially for LREE and MREE. This indicates that the HREE in the Great Northern coals generally have a much greater mineral affinity than the LREE and MREE.

190

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

Table 6.10 Rare earth elements in coal samples and associated strata from the Great Northern coal and associated non-coal samples in the Newvale section (REE concentrations in ppm, on whole-coal basis).

21379 21380 21381 21382 21383 21384 21385 21386 21387 21388 21389 21390 21391

La 15.6 2.6 10.6 14.0 7.8 7.4 3.5 4.0 5.9 3.7 17.5 10.1 13.8 Ce 31 5.9 20.6 29 15.6 16 10.3 8.0 13.6 9.0 40 24.5 29 Pr 3.6 0.8 2.6 3.4 2.0 1.9 1.5 1.0 1.7 1.2 4.9 3.2 3.6 Nd 14.2 4.1 10.3 13.2 8.4 7.5 6.6 3.9 7.1 5.2 20.1 13.9 16.1 Sm 2.9 1.4 2.3 2.6 1.7 1.7 1.6 0.9 1.8 1.3 4.6 3.4 3.6 Eu 0.79 0.47 0.46 0.54 0.36 0.48 0.42 0.20 0.39 0.32 1.01 0.72 0.76 Gd 1.7 2.0 2.2 2.3 1.6 1.2 1.3 0.8 1.5 1.2 3.8 3.1 3.5 Tb 0.41 0.42 0.40 0.40 0.24 0.22 0.23 0.13 0.31 0.24 0.63 0.54 0.55 Dy 2.4 2.9 2.3 2.1 1.4 1.15 1.3 0.8 1.8 1.4 3.6 3.2 3.4 Y 14 22 18.0 12 8.3 6 7.8 5.0 11.2 9.6 21 22 28 Ho 0.56 0.69 0.53 0.47 0.29 0.25 0.28 0.16 0.41 0.33 0.71 0.71 0.78 Er 1.7 1.9 1.4 1.2 0.8 0.73 0.8 0.5 1.1 1.0 2.1 2.1 2.4 Tm 0.26 0.29 0.20 0.18 0.12 0.11 0.12 0.07 0.17 0.15 0.31 0.30 0.35 Yb 1.7 1.8 1.2 1.2 0.8 0.70 0.8 0.5 1.1 1.0 2.1 2.0 2.2 Lu 0.31 0.28 0.19 0.18 0.13 0.13 0.12 0.07 0.17 0.16 0.33 0.33 0.36 REE 91.2 47.5 73.3 82.6 49.5 45.6 36.6 25.9 48.3 35.9 122 90.0 108

(La/Lu)N 0.54 0.10 0.60 0.85 0.65 0.61 0.30 0.58 0.37 0.25 0.57 0.32 0.40

(La/Sm)N 0.82 0.28 0.70 0.82 0.68 0.66 0.32 0.68 0.50 0.42 0.57 0.44 0.58

(Gd/Lu)N 0.47 0.61 0.99 1.11 1.05 0.77 0.91 0.94 0.76 0.64 0.97 0.78 0.81 Eu 1.64 1.29 0.96 1.04 1.03 1.59 1.35 1.14 1.12 1.19 1.14 1.05 1.01 anomaly Ce 0.93 0.91 0.89 0.96 0.90 0.99 0.99 0.92 0.97 0.95 0.98 0.97 0.94 anomaly Y 0.94 1.18 1.24 0.93 0.98 0.80 0.96 1.05 0.98 1.07 0.99 1.12 1.30 anomaly Enrichment H H H H&M H&M H H H H H H H H type

191

Figure 6.20 Distribution patterns of REE in the Great Northern seam (Newvale). REE are normalized to Upper Continental Crust (UCC) data from Taylor and McLennan (1985). (A) Coal samples 21380 to 21390; (B) Non- coal samples including roof sample 21379, claystones 21382, 21384, and 21389, and floor 21391.

Figure 6.21 Correlation coefficients between mean individual REE and Y with LTA% in the Great Northern coal samples (Newvale).

6.5 Summary

Mineralogical and chemical variations have been identified through the vertical profile of the Great Northern coal seam, involving contrasts between the mineral matter in the individual coal plies as well as between the coal plies, the floor strata, and the intra-seam non-coal bands. There are also differences between the mineral matter in the coal plies of the outcrop section at Catherine Hill Bay and the sample from the mine exposure at Newvale, which may be related, at least in part, to weathering of the coal at the outcrop site.

192

Chapter 6 Mineralogy and Geochemistry of the Great Northern Seam

The floor strata below the Great Northern seam are made-up of abundant quartz, with minor and equal proportions of kaolinite, illite and expandable clay minerals. The floor material also contains abundant detrital K-feldspar. Coals from the middle parts of the seam are dominated by well-ordered kaolinite with minor quartz, and in some cases contain minor phosphates and anatase. Carbonate minerals, siderite, and dolomite or ankerite, occur in the fresh coal at Newvale but not in the weathered coal at Catherine Hill Bay. However, non-kaolinite clay minerals and detrital feldspar are abundant in the lowermost ply of the coal seams, suggesting that the immediate base of the peat bed was made-up of organic matter admixed with the same detrital sediment as supplied to the basin before the swamp was established. The presence of disordered kaolinite in the lower coal plies, similar to that in the floor samples, further supports such a conclusion.

Authigenic K-feldspar occurs in the lower part of the coal seam, especially in the Newvale section, including the siltstone floor, the lowermost claystone parting, and the coal plies between. The K-feldspar shows a variety of modes of occurrence, including cell infillings, irregular veins, and thin laminae parallel to the bedding of organic matter or detrital clay bands. The texture and chemical composition of the K-feldspar suggest that it originated from low-temperature hydrothermal fluids in the late syngenetic stage of peat development, probably associated with contemporaneous volcanic activity. The hydrothermal activity may have occurred in a relatively localized area, given that similar vein minerals are rare in the Catherine Hill Bay siltstone floor, and that the feldspar in the floor at that location largely occurs as detrital grains.

Although the lowermost intra-seam band at Newvale has characteristics more closely resembling the floor strata, the uppermost two bands in that section, and also the bands at a similar level in the Catherine Hill Bay exposure, consist almost entirely of well-ordered kaolinite. The kaolinite in these bands displays graupen to vermicular textures, and the bands appear to represent tonsteins within the coal seam. The nature of the non-kaolinite clay minerals in the coal also changes from smectite in the lowermost plies, similar to the underlying floor materials, to a regularly interstratified illite/smectite component in the section that includes the tonstein layers. Smectite is again developed, however, in the mudstone overlying the main part of the coal seam in the Catherine Hill Bay exposure.

The coals above and below the tonsteins are also richer in inertinite than the other parts of the seam section. This may indicate a different peat-forming environment, possibly one that allowed more widespread deposition of volcanic ash without interruption by

193 contemporaneously growing trees (cf. Creech, 1998). While volcanic ash may have been admixed with the organic debris at different times during peat accumulation, it may have only been able to form thin but persistent beds at times when the peat-forming environment was not interrupted by major upright tree growth.

Compared with worldwide coal, trace element concentrations for the Great Northern samples are generally low. Lithium occurs essentially in kaolinite in the coals. Rubidium and K are strongly correlated, although Rb does not form any mineral of its own. Chalcophile elements including Se, Pb and Cu in the coals are probably associated with kaolinite, rather than any other mineral phases. The correlation coefficient between REE and the ash yield appears to increase with increasing atomic number of REE, especially for LREE and MREE. This indicates that the HREE in the Great Northern coals generally have a much greater mineral affinity than the LREE and MREE.

194

CHAPTER 7 MINERALOGY AND GEOCHEMISTRY OF THE BULLI SEAM

This part of the study aims to investigate the variation in mineral matter in Late Permian bituminous coals and associated strata from the Bulli seam of the southern Sydney Basin. Such a study provides an opportunity to assess the depositional environment according to coal mineralogical characteristics. It also provides a basis for comparison to the other seam sections included in the research program, especially the non-marine Great Northern seam and the marine-influenced Greta coal seam of the northern Sydney Basin.

7.1 Coal characteristics

The proximate analysis data, as well as total sulphur, calorific value and random vitrinite reflectance of the Bulli coal samples are given in Table 7.1. All the Bulli coal samples are low in total sulphur (<0.4%), with the sulphur probably mainly being organic. Based on the fixed carbon and volatile matter values, the Bulli coal is mainly classified as a medium volatile bituminous coal under the ASTM classification (ASTM, 2007).

Table 7.1 Proximate analysis and vitrinite reflectance data of Bulli coal and associated strata. Proximate analysis data from CSIRO reports.

Thickness M Ash VM FC CV Total S Rv, ran Sample (Btu/lb, (m) (%, ad) (%, ad) (%, ad) (%, daf) (%, ad) (%, daf) (%, ad) (%) daf)

8117 - 1.08 42.02 - - - - 0.22 - 8118 0.08 0.64 26.04 20.25 27.6 53.07 72.4 14820 0.32 1.29 8119 0.03 0.88 79.17 - - - - 0.06 - 8120 0.61 0.64 12.31 19.11 22 67.94 78 15520 0.34 1.26 8121 0.66 0.62 10.32 19.2 21.6 69.86 78.4 15490 0.34 1.31 8122 0.25 0.62 6.8 21.82 23.6 70.76 76.4 15680 0.38 1.3 8123 0.48 0.6 10.08 20.36 22.8 68.96 77.2 15470 0.36 1.31 8124 0.41 0.6 8.99 21.38 23.6 69.03 76.4 15640 0.38 1.29 8125 - 0.74 76.76 - - - - 0.1 - ad = air-dried basis; daf = dry ash-free basis; M = inherent moisture; FC = fixed carbon; CV = calorific value; Total S = total sulphur; - = no data.

The Bulli coal samples are mainly low in ash yield, in the range of 6.8% to 12.3%, except for one high ash coal sample (8118) immediately below the roof material. The moisture content of the coal samples is constantly around 0.62% (air-dried basis) and. the volatile

195 matter value ranges between 21.6% and 27.6% (dry, ash-free basis). The highest volatile matter value occurs in the uppermost coal sample (8118), which contains a high proportion of carbonates (Table 7.3). The high volatile yield may thus be associated with the loss of carbon dioxide from the carbonates during the analysis process.

7.2 Coal petrology

An evaluation of the macerals in the Bulli coal samples was carried-out on the crushed grain mount samples, in accordance with the Australian Standard (Standards Australia, 1998b). The maceral composition is listed in Table 7.2. On a mineral-free basis, the individual coal plies have 37-71% vitrinite, 29-63% inertinite; liptinite was not detected.

Table 7.2 Maceral analysis (vol., %, mineral-free basis) of the Bulli coal samples (bdl = below detection limit)

Maceral 8118 8120 8121 8122 8213 8124 Telinite 0 0.7 0 0 0 0.2 Collotelinite 10.8 8.7 5.5 11.9 8.8 7.4 Collodetrinite 34.2 35.4 31.9 52.7 38.2 63.1 Vitrodetrinite 0 0.7 0 0 0 0.2 Corpogelinite 0 0 0 0 0 0 Gelinite 0 0.4 0 0 0 0 Total vitrinites 45.0 45.9 37.4 64.6 47.0 71.0

Fusinite 1.4 1.3 2.4 1.8 2.4 2.6 Semifusinite 24.9 24.9 31.7 13.9 27.5 14.4 Funginite 0.5 0.2 0.2 0 0 0.2 Inertodetrinite 8.9 10.7 13.4 10.4 13.6 6.0 Macrinite 7.7 12.9 13.4 7.3 7.7 5.6 Micrinite 11.7 4.1 1.3 2.0 1.8 0.2 Secretinite 0 0 bdl 0 0 0 Total inertinites 55.0 54.1 62.6 35.4 53.0 29.0

Sporinite bdl 0 0 0 0 0 Cutinite bdl 0 0 0 0 0 Resinite 0 0 0 0 0 0 Liptodetrinite 0 0 0 0 0 0 Suberinite 0 0 0 0 0 0 Fluorinite 0 0 0 0 0 0 Bituminite 0 0 0 0 0 0 Total liptinites 0 0 0 0 0 0

Collotelinite is the most abundant vitrinite maceral. Typical collotelinite in the Bulli coals occurs as bands, sometimes with cracks (Figures 7.1A, B). Other vitrinite macerals are rare. Inertinite macerals in the Bulli coals are shown in Figure 7.1. The inertinite macerals in the Bulli coals are mainly semifusinite (Figures 7.1A, E, F), inertodetrinite (Figures 7.1A,

196

Chapter 7 Mineralogy and Geochemistry of the Bulli Seam

B), macrinite (Figure 7.1F), with minor fusinite (Figures 7.1C, F), micrinite (Figure 7.1E). Funginite (Figure 7.1B) and secretinite (Figure 7.1D) also occur.

Figure 7.1 Photomicrographs showing typical macerals in the Bulli coals. Oil immersion, reflected light. (A) Collotelinite (ct) and semifusinite (sf) bands, associated with collodetrinite (cd) and inertodetrinite (id). Sample 8121. (B) Collotelinite (ct) band and funginite (fg). Sample 8121. (C) Fusinite exhibiting bogen structure. Sample 8122. (D) Secretinite (se) associated with semifusinite. Sample 8121. (E) Clustered micrinite (mi) in collodetrinite matrix. Sample 8118. (F) Association of macrinite (ma), semifusinite (sf) and fusinite (f). Sample 8118.

197

7.3 Mineralogy of the coals and associated strata

Quantitative mineralogical results from powder XRD analysis and Siroquant interpretation for the Bulli coal LTA and non-coal rock samples are given in Table 7.2. Clay mineralogy IRU WKH OHVV WKDQ ȝP IUDFWLRQV RI WKH FRDO /7$V DQG DVVRFLDWHG URFNV E\ RULHQWHG aggregate XRD is given in Table 7.4. An indication of clay mineralogy of the coal seam by oriented aggregate XRD is also provided by the profile in Figure 7.2.

Table 7.3 Mineralogy of the Bulli coal LTAs and non-coal rock samples by XRD and Siroquant (wt. %)

Sample LTA% Qtz Kao I I/S Ank Sid Ana Alb Goy Fap Bass 8117 47.6 11.7 43.5 17.1 24.6 0.7 1.4 1 8118 30.8 8.8 54.6 11.2 9.6 13.7 1.4 0.7 8119 - 1.8 97.8 0.4 8120 13.9 7.7 77 7.1 2.5 3.5 0.8 0.9 0.4 8121 11.6 3.3 80.5 5 1.4 5.2 1.1 1.6 1.9 8122 7.5 11.7 78.8 4.8 2.5 0.5 0.7 1 8123 11.6 6.4 84.3 0.4 3.3 0.5 1.9 3.2 8124 9.8 27.6 58.4 9.6 3.6 0.8 8125 - 51.3 17.2 21.1 8.4 0.7 1.3 Qtz=quartz; Kao=kaolinite; I=illite; I/S=illite/smectite; Ank=ankerite; Sid=siderite; Ana=anatase; Alb=albite; Goy=goyazite; Fap=fluorapatite; Bass=bassanite.

Table 7.4 Mineralogy of <2 μm fraction of coal LTAs and non-coal strata using oriented aggregate XRD techniques (wt. %) Sample Kaolinite Illite Expandable clay (mainly I/S) 8117 61.1 25.5 13.4 8118 83.5 13.5 3.0 8119 100 0 0 8120 92.0 4.9 3.1 8121 100 0 0 8122 100 0 0 8123 100 0 0 8124 90.4 9.6 0 8125 38.8 37.6 23.5

7.3.1 Mineralogy of the non-coal strata

The floor of the Bulli seam in the section studied is a carbonaceous claystone, consisting of abundant quartz (51.3%) and illite (21.1%), relatively minor kaolinite (17.2%) and I/S (8.4%), and trace amounts of anatase and siderite. Although under the detection limit for bulk (whole rock) XRD analysis, a small proportion of I/S was indicated by the oriented

198

Chapter 7 Mineralogy and Geochemistry of the Bulli Seam

XRD analysis (Table 7.4). The roof of the Bulli seam is a highly carbonaceous shale (42% ash yield), which also comprises relatively abundant kaolinite (43.5%), illite (17.1%), I/S (24.6%) and quartz (11.7%), and small proportions of anatase and albite. The kaolinite in both the roof and floor samples is poorly-ordered.

The claystone sample (8119) is almost entirely made-up of kaolinite (97.8%), with trace proportions of quartz and anatase. Kaolinite as the only clay mineral is also confirmed by the oriented-aggregate XRD data. In contrast to that in the roof and floor samples, XRD analysis indicates that the kaolinite in the claystone band has a well-ordered crystal structure. The occurrence of negligible amounts of quartz and other (detrital) clay minerals in the claystone is also in contrast to that in the adjacent coal plies and the roof and floor strata. This may indicate that the claystone was derived from sources other than normal clastic sediments. Except for the absence of volcanogenic minerals, the claystone sample in the Bulli seam has characteristics that resemble those in the Great Northern seam (Chapter 6).

K-feldspar, which occurs in the floor and intra-seam claystones of the Great Northern seam, and also in the coal plies overlying the floor (Chapter 6), is absent in the Bulli coal seam. This may reflect deposition at a greater distance from the sediment source, which was in the New England Fold Belt to the north of the Sydney Basin.

Figure 7.2 Vertical column sections showing variations in clay mineral compositions of the Bulli seam.

199

7.3.2 Mineralogy of the coal samples

The mineral assemblage in most of the Bulli coal plies consists of abundant kaolinite, minor quartz and carbonates (dolomite, ankerite and siderite), and, in some cases, minor anatase, goyazite and fluorapatite are present.

7.3.2.1 Kaolinite and quartz

Kaolinite is the dominant component of the mineral matter in the Bulli seam, although minor proportions of illite and expandable clay minerals are also noted in the oriented- aggregate XRD data for the lowermost and two uppermost coal plies of the profile. The expandable clay mineral in the Bulli seam is mainly I/S. Although not as abundant as in the floor strata, quartz is still relatively abundant in the lowermost coal ply of the Bulli seam, comprising 27.6% of the LTA assemblage. Above this level it decreases to < 10%.

With the exception of the immediate basal and uppermost coal plies, the clay mineralogy is very different in the coal samples compared to the roof and floor strata of the Bulli seam. Significant proportions of illite and expandable clay minerals occur in the clay mineral assemblage of both the roof and floor samples (Table 7.4, Figure 7.2). However, within most of the coal samples themselves, well-ordered kaolinite is dominant with only trace amounts of illite and expandable clay minerals. Like that in the Great Northern coals, the kaolinite in the Bulli coals primarily occurs in thin bands intimately associated with vitrinite and as cell-cavity fillings. This indicates that the kaolinite in the coals was formed mainly by authigenic precipitation in the original peat swamp (cf. Ward, 1989). Non-kaolinite clay minerals, mainly present in the lowermost and uppermost coal plies, are interpreted to be essentially of detrital origin. Ward and Christie (1994) found that the uppermost and lowermost parts of some Late Permian coal beds from the Bowen Basin, regardless of rank, also contain relatively abundant illite and/or interstratified illite/smectite.

7.3.2.2 Carbonates

Ankerite (ferroan dolomite) in the Bulli coal plies largely occurs as cleat infillings in the bright coal layers. Abundant veins made-up of these minerals also occur in the topmost ply of the coal bed, immediately below the roof material. Ward et al. (1996) and Ward et al. (1999a) also reported abundant carbonates as cleat or fracture fillings in the coal at the very top of the Bulli seam at another location, and coal from the upper part of the Black Jack Group, Gunnedah Basin, respectively. Along with the cleat infillings the carbonates

200

Chapter 7 Mineralogy and Geochemistry of the Bulli Seam are clearly post-depositional; the veins near the top of each seam may have been derived from expulsion of organically-associated Ca and Mg from the macerals. As discussed in previous studies (Ward et al., 2005; Li et al., 2007b; Ward et al., 2007), organically-bound inorganic elements may be liberated from coal macerals, especially vitrinite, as part of the molecular changes associated with the rank advance process.

Siderite is present in most of the coal plies in the Bulli seam, typically comprising < 5% of LTA assemblages. Like that in the Great Northern coals, siderite in the Bulli coals also occurs as cleat/fracture infillings. Siderite concretions also occur, some of which contain various concentrations of Ca, Mg, and Mn (Figure 7.3A).

7.3.2.3 Phosphates

XRD analysis indicates that small proportions of goyazite and fluorapatite occur in the four thick coal plies (8120 to 8123) of the Bulli seam. SEM examination shows that the fluorapatite commonly co-exists with kaolinite in cell cavities or pore spaces (Figure 7.3B). Although referred to as goyazite (the Sr end-member of the aluminophosphate series), the XRD patterns do not match exactly with that mineral, or with the other end-members (Ba or Ca end members). Apart from Al, P and Sr, EDS spectra indicate that significant concentrations of Ca and Ba also occur in most cases (Figure 7.3C), and hence the mineral is most likely a solid solution of the Sr, Ca, and Ba end-member phases. Goyazite and fluorapatite are common in Australian coals (cf. Ward et al., 1996). The formation of these two phosphates in coal has been discussed in Chapter 5.

7.3.2.4 Other minerals

Anatase occurs persistently in all the Bulli coal samples in trace proportions (0.5-1.5%, LTA basis). Like that in the other coals of the present study, the anatase may occur as microcrystalline particles in a matrix of clay minerals, especially kaolinite in the Bulli coals, and was probably co-precipitated with the clay minerals in the early diagenetic stage.

201

Figure 7.3 SEM images of minerals in coal sample 8123. (A) Siderite concretion, containing minor Ca, Mg and Mn. (B) Fluorapatite (A) co-existing with kaolinite (K) in pore space. Weak Al and Si peaks are from the intimate mixture with the kaolinite. (C) Ca, Ba-bearing goyazite in pore spaces of kaolinite (K) aggregates.

202

Chapter 7 Mineralogy and Geochemistry of the Bulli Seam

7.4 Geochemistry of the Bulli seam

7.4.1 Mineralogical and chemical analysis data

Major element chemical data for the high-temperature (815 qC) ashes (HTA) of the Bulli coal and non-coal samples are given in Table 7.5. The results have been expressed to an

LOI- and SO3-free basis to allow better comparison to the mineralogical data.

The relation between ash chemistry and mineralogy was studied to check on the reliability of the quantitative XRD data determined by Siroquant. The chemical composition of the mineral assemblage in the LTAs and rock samples determined by Siroquant was calculated and compared with the actual ash composition of the 815 °C coal ash and rocks as determined by XRF, following the calculation procedure described by Ward et al. (1999) and in other chapters of the present thesis. The chemical composition was + modified for each sample by deducting the CO2 and H2O to derive an equivalent to an ash analysis. The actual chemical composition determined by XRF was also normalized to an SO3-free basis. The correlations of major oxides are shown in Figure7.4.

SiO2 and Al2O3 plots show strong correlations of these oxides from the Siroquant data and those indicated by the ash chemistry. This indicates that Siroquant gives consistent results for those major minerals which contribute to the major element oxides in the analysed coal ashes.

The plot for Fe2O3 shows a broad scatter, with most of the points falling below the equality line. The coal sample containing the highest observed Fe2O3 concentration from XRF analysis falls farthest from the equality line. This may reflect an underestimation of siderite or ankerite by XRD and Siroquant analysis for that sample, but may also indicate that some of the Fe in the coal occurs in a non-crystalline form. Both the plots for CaO and MgO show somewhat scattered correlations, with CaO and MgO data inferred from Siroquant often being less than those observed ash chemistry. As noted above, significant concentrations of Ca and Mg may also occur in the siderite of the coals, which was not allowed for in the stoichiometric calculations.

203

Table 7.5 Major element analyses of Bulli coal ash and non-coal samples (%), as determined by XRF analysis (bdl = below detection limits; n.d. = no data).

Sample HTA SiO2 Al2O3 TiO2 Fe2O3 MgO CaO Na2O K2O MnO P2O5 SO3 LOI

8117 41.99 57.66 34.78 1.06 1.78 0.70 0.17 0.24 2.41 0.007 0.01 0.11 n.d. 8118 26.70 48.31 32.70 1.45 2.76 3.27 5.26 0.10 1.18 0.017 0.09 3.28 n.d. 8119 - 41.66 35.14 0.48 0.44 0.10 0.04 bdl 0.08 0.002 0.04 0.09 21.34 8120 12.15 51.48 37.88 1.20 2.70 1.00 1.88 0.15 0.51 0.013 0.68 2.01 n.d. 8121 10.02 48.04 39.41 1.52 4.40 0.87 1.93 0.16 0.14 0.014 1.58 0.69 n.d. 8122 6.60 51.29 34.03 2.07 5.14 1.49 2.39 0.13 0.29 0.030 0.33 2.90 n.d. 8123 10.13 44.97 34.84 1.24 8.91 1.01 1.71 0.08 0.11 0.125 1.60 1.39 4.12 8124 9.09 60.80 28.42 1.51 3.42 0.52 0.30 0.04 0.85 0.035 0.12 0.66 3.96 8125 - 56.15 12.63 0.89 2.68 0.48 0.08 0.01 2.11 0.047 0.03 0.09 23.59 8126 10.72 51.00 35.33 1.38 4.53 1.11 1.95 0.12 0.44 0.046 0.90 1.95 n.d.

The K2O plot (Figure 2F) also shows a good level of agreement, with all points close to the 1:1 diagonal line. This indicates that consistent results for non-kaolinite clay minerals (illite and I/S) are given by Siroquant. The plot for Na2O shows a broad scatter, which may due to the small concentrations involved; the poorly-defined correlation that does exist may reflect traces of Na in the I/S or illite. The scatter shown in the TiO2 plot may again reflect the low overall concentrations and associated errors in XRD determination, but this may also be due to the microcrystalline nature of the anatase particles, which would make them difficult to detect by XRD and Siroquant analysis.

7.4.2 Major element oxides in coals

The major oxides in the coal ashes are dominated by SiO2 and Al2O3. The concentrations of MgO, CaO, Fe2O3, K2O and TiO2 in coal ashes are mostly less than 5%, and those of the rest of the major oxides are less than 1%. The ashes of the uppermost coal plies are rich in CaO and MgO, which reflect the abundant ankerite. High proportions of K2O occur in the coal samples near the floor and roof strata, reflecting the greater abundance of illite and I/S in the relevant coals. Relatively high proportions of P2O5 occur in the ashes of the coal plies from the main part of the seam (8120 to 8123), mainly due to the presence of goyazite and fluorapatite. The TiO2 concentrations are higher in the coal ashes than in the non-coal rock samples, which may reflect greater abundance of anatase in the mineral matter of the coals than in the non-coal rock samples. However, this is not indicated by the XRD analysis, due to the difficulty in detecting the microcrystalline particles of anatase in the coal samples.

204

Chapter 7 Mineralogy and Geochemistry of the Bulli Seam

Figure 7.4 Comparison between proportions of major element oxides in the Bulli coal ash and non-coal samples inferred from Siroquant and determined by XRF. The diagonal line represents equality in each plot. Relevant trendlines and the squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case.

205

7.4.3 Selected trace elements

Trace element data for the Bulli coal and non-coal samples are given in Table 7.6. With the exception of the high-ash, uppermost coal ply (8118), the concentrations of most trace elements in the coals are lower than that of average worldwide coals (Ketris and Yudovich, 2009). Only Li, Sc, Ga, Th and U are generally slightly higher in the coals than the averages in the worldwide coal dataset. Boron in the Bulli coals varies from 5 to 35 ppm, indicating that the seam was influenced by fresh water, according to the classification by Goodarzi and Swaine (1994). Chlorine is generally low in the Bulli coals (<40-90 ppm).

To evaluate the relationship between individual elements and minerals in the Bulli samples, correlation coefficients were calculated between the proportions of individual elements and minerals, with both calculated on a whole coal basis. The major oxide percentages determined by XRF analysis of the coal ashes were recalculated to give the percentages of those oxides in the whole coal, which are also given in Table 7.6.

Lithium in the Bulli coals is strongly correlated with kaolinite (Figure 7.5A), having a correlation coefficient of 0.97. The association of lithium with kaolinite is also noted in coals in other chapters. Beryllium appears to be correlated with ash yield (Figure 7.5B), having a correlation coefficient (R=0.74) higher that that with kaolinite (R=0.52), indicating an inorganic affinity of the element in coal. Elevated concentrations of many elements, such as Co, Cu, Ni, As, and Tl, are present in a high-ash coal sample (8118). However, overall correlation does not exist between these elements and the ash yield, and the examples are shown in Figures 7.5C, D.

Most trace elements, however, do not show any significant correlations with the abundance of any mineral, major oxides or the ash yield. This may partly be due to the small number of coal samples. Comparison of elements for both coal and non-coal samples, however, shows some correlations between lithophile elements Li, Be and Ga and the major oxide Al2O3 (Figures 7.6A, B, C) (R=0.95, 0.83 and 0.85, respectively). The chalcophile element As also exhibits a relatively strong correlation with Al2O3 (Figure 7.6D), having a correlation coefficient of 0.80. The correlations between chalcophile elements and alumina probably indicates a common source. Like the other coals in the present study, Rb and K are also strongly correlated (Figure 7.6E) (R=1.0), although Rb is not associated with any mineral phase in the coals.

206

Chapter 7 Mineralogy and Geochemistry of the Bulli Seam

Sc exhibits a strong positive correlation with silica (Figure 7.6F), with a coefficient of 1.0. Again, this may be due to the spectral interference of Si on Sc in the ICP-MS analysis. Relatively high concentrations of Si in the coal samples may result in a spectral interference of Si on Sc.

Figure 7.5 Comparison of elements to kaolinite and ash yield of the Bulli coal samples. (A) Li against kaolinite. (B) Be against the ash yield. (C) As against the ash yield. (D) Tl against the ash yield. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case.

7.4.4 Selected elements in non-coal samples

In the study of Spears and Kanaris-Sotiriou (1979), tonsteins with TiO2/Al2O3 values of <0.02 and >0.07 are grouped to indicate parent magmas of acid and basic composition, respectively; those with values in between are thought to represent intermediate ash materials. The comparison of TiO2 and Al2O3 in the Bulli non-coal samples is plotted in

Figure 7.7A. The intra-seam claystone falls below the line indicating TiO2/Al2O3 value of 0.02, and are thus suggested to be mainly derived from acid volcanic ash. The key parameters for the claystone were also plotted in the magma source discrimination diagram of Winchester and Floyd (1977) (Figure 7.7B). The results indicate that, if the claystone was originally derived from volcanic ash, the magma source would have been trachyandesitic in composition. However, this is not exactly consistent with the results

207 indicated by Figure 6.19A, which may partly be due to possible contamination of the samples from the zirconia grinding mill during sample preparation.

Figure 7.6 Comparison of elements in the Bulli coal and non-coal samples. (A) Li against Al2O3. (B) Be against Al2O3. (C) Ga against Al2O3. (D) As against Al2O3. (E) Rb against K2O. (F) Sc against SiO2. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case.

208

Chapter 7 Mineralogy and Geochemistry of the Bulli Seam

Table 7.6 Major oxide and trace element analyses of samples from the Bulli seam (Major oxides in %, recalculated from the data by XRF analysis, on whole-coal basis; Trace elements in ppm, unless otherwise indicated, on whole-coal basis; bdl = below detection limit).

8117 8118 8119 8120 8121 8122 8123 8124 8125

SiO2 24.47 13.10 41.90 6.29 4.87 3.38 4.55 5.49 56.84

TiO2 0.45 0.39 0.48 0.15 0.15 0.14 0.13 0.14 0.90

Al2O3 14.76 8.87 35.35 4.63 4.00 2.24 3.53 2.57 12.79

Fe2O3 0.75 0.75 0.44 0.33 0.45 0.34 0.90 0.31 2.72 MgO 0.295 0.886 0.103 0.122 0.089 0.098 0.102 0.047 0.481 CaO 0.071 1.426 0.043 0.229 0.195 0.157 0.173 0.027 0.078

Na2O 0.103 0.027 bdl 0.018 0.016 0.009 0.008 0.003 0.007

K2O 1.024 0.320 0.081 0.062 0.014 0.019 0.011 0.077 2.140 MnO 0.003 0.005 0.002 0.002 0.001 0.002 0.013 0.003 0.048

P2O5 0.01 0.02 0.04 0.08 0.16 0.02 0.16 0.01 0.03 Li 130 98.7 515 55.9 62.8 26.2 46.5 22.2 39.7 Be 2.85 2.53 3.79 0.94 0.83 0.44 0.83 2.08 2.77 B 45.36 7.49 16.59 33.92 4.71 19.95 34.95 29.69 25.27 Cl (%) <0.004 <0.004 0.07 <0.004 0.01 0.01 0.01 <0.004 0.10 Sc 35.14 17.29 46.78 8.26 6.42 4.34 6.55 8.17 70.94 V 68.11 29.40 11.38 14.83 14.20 10.20 9.81 17.35 76.18 Cr 62.66 16.91 11.76 4.99 8.38 4.22 5.87 11.92 81.75 Co 12.95 83.52 9.25 10.13 3.31 4.76 3.21 12.83 170 Ni 61.00 88.84 98.23 17.58 4.16 6.15 13.20 31.14 126 Cu 49.47 26.75 8.73 9.36 8.27 8.37 8.12 13.20 44.64 Zn 36.44 88.63 15.09 13.24 11.96 6.03 11.96 19.15 62.12 Ga 25.01 17.51 48.65 10.85 19.00 9.27 20.97 7.62 43.25 Ge 1.06 2.97 5.86 5.51 3.11 3.13 3.06 6.31 1.56 As 0.79 1.14 1.23 0.27 0.21 0.15 0.14 0.19 0.52 Se 2.63 1.51 1.62 0.48 0.68 0.66 0.55 0.59 0.23 Rb 48.70 15.69 3.59 1.92 0.58 0.78 0.50 3.85 117 Y 31.42 22.14 8.56 13.04 7.44 4.24 6.15 9.72 21.43 Zr 130 155 132 49.1 2.31 47.8 2.73 46.6 261 Nb 4.99 7.17 8.59 1.49 1.61 2.46 1.73 2.63 9.97 Mo 0.29 0.37 0.86 0.40 0.62 0.54 0.53 0.31 0.03 Ag 0.22 0.20 0.16 0.05 0.05 0.07 0.05 0.08 0.27 Cd 0.23 0.27 0.11 0.08 0.05 0.07 0.08 0.10 0.22 Sn 2.42 2.88 4.88 0.85 1.45 1.28 1.30 1.49 8.73 Sb 1.48 2.95 1.61 2.11 0.51 0.34 0.37 4.44 0.67 Te 0.13 0.14 0.17 <0.1 <0.1 <0.1 <0.1 <0.1 0.25 Cs 6.21 2.12 0.59 0.18 0.08 0.09 0.05 0.53 28.47 Hf 5.80 6.40 8.47 2.16 0.05 2.00 0.07 2.06 11.21 Ta <0.2 0.32 1.16 <0.2 <0.2 <0.2 <0.2 <0.2 0.75 W 1.43 0.51 0.88 0.25 0.27 0.35 0.27 0.29 3.99 Hg (ppb) 69.0 93.9 41.6 24.5 130.7 26.5 32.6 31.9 254.5 Tl 0.39 0.11 0.03 0.03 0.03 0.03 0.03 0.05 0.72 Pb 25.07 19.97 8.88 5.15 9.96 5.49 8.36 14.72 34.95 Bi 0.62 0.37 0.26 0.15 0.42 0.21 0.25 0.41 1.08 Th 11.44 6.72 8.60 2.77 8.07 2.58 6.00 3.35 21.27 U 1.67 4.45 1.89 0.95 3.21 0.75 1.63 2.16 7.42

209

The roof and floor samples are plotted between lines indicating TiO2/Al2O3 values of 0.07 and 0.02 in the Figure 7.7.A, and in the andesite and rhyodacit/dacite fields in Figure 7.7B. These materials are suggested to be derived probably from source material of mafic to intermediate compositions.

Figure 7.7 Plots of elements for the Bulli non-coal samples. (A) Comparison of TiO2 and Al2O3 concentrations.

The upper and lower diagonal lines represent TiO2/Al2O3 values of 0.07 and 0.02, respectively. (B) Plot of

Zr/TiO2 against Nb/Y ratios. Magma source discrimination diagram of Winchester and Floyd (1977).

7.4.5 Distribution and affinity of REE and Y

7KH WRWDO UDUH HDUWK HOHPHQW DQG \WWULXP FRQWHQWV 5((< in the Bulli coal samples typically range between 18.8 and 66 ppm (Table 7.7). The REE concentrations were normalised against the Upper Continental Crust (UCC) (Taylor and McLennan, 1985) for each coal and associated non-coal sample, in order to obtain a clearer indication of the distribution patterns (Figure 7.8).

Like the Great Northern coals, the normalized REE distributions in all of the Bulli coal samples have LaN/LuN values less than 1 (Table 7.7), indicating heavy REE enrichment (H-type) compared to the UCC (Seredin, 2001). As mentioned in Chapter 6, the H-type enrichment may be attributed to the circulation of water which was enriched in HREE through coal basins (Seredin, 2001). Eskenazy (1999) ascribed the enrichment of HREE relative to LREE to the formation by HREE of complexes with organic matter in the coal, which would also increase the stability of the HREE relative to the LREE. Coal sample 8118 which is immediately underlying the roof shows the greatest fractionation between the normalised LREE and HREE (Figure 7.8A). Likewise, greatest fractionation of the normalized LREE and HREE is shown for the roof sample 8117 (Figure 7.8B). These two adjacent samples may have a similar source of REE.

210

Chapter 7 Mineralogy and Geochemistry of the Bulli Seam

As indicated in Figure 7.9, the LREE generally have higher correlation coefficients with the ash yield than the MREE and HREE. The correlation coefficient of the LREE and ash yield also increases from 0.1 (La) to 0.74 (Sm), with increasing REE atomic number. This indicates a general greater mineral affinity of the LREE than the MREE and HREE in the Bulli coals.

Table 7.7 Rare earth elements in coal samples and associated strata from the Bulli coal and associated non- coal samples (REE concentrations in ppm, on whole-coal basis)

Sample 8117 8118 8119 8120 8121 8122 8123 8124 8125 La 7.02 5.29 1.50 9.93 7.16 2.12 5.61 6.34 19.56 Ce 13.52 13.32 4.49 19.61 15.68 5.34 11.56 14.44 39.28 Pr 1.76 1.97 0.57 2.53 1.74 0.69 1.24 1.81 4.61 Nd 7.12 8.18 2.78 10.20 6.45 2.72 4.53 7.13 16.84 Sm 2.06 2.30 0.78 2.27 1.39 0.69 1.08 1.53 3.28 Eu 0.53 0.57 0.28 0.48 0.35 0.15 0.28 0.29 0.62 Gd 2.97 2.81 1.13 2.10 1.11 0.61 1.00 1.38 2.81 Tb 0.63 0.51 0.23 0.34 0.21 0.12 0.20 0.25 0.50 Dy 4.78 3.29 1.31 1.97 1.25 0.82 1.17 1.46 3.21 Y 31.42 22.14 8.56 13.04 7.44 4.24 6.15 9.72 21.43 Ho 1.16 0.74 0.25 0.44 0.26 0.18 0.24 0.34 0.70 Er 3.42 1.89 0.74 1.32 0.72 0.51 0.62 1.02 2.26 Tm 0.53 0.27 0.10 0.20 0.11 0.08 0.09 0.16 0.34 Yb 3.36 1.71 0.63 1.40 0.73 0.50 0.52 1.01 2.39 Lu 0.55 0.25 0.11 0.22 0.12 0.08 0.08 0.15 0.37 REE 80.83 65.23 23.46 66.04 44.72 18.84 34.37 47.04 118.2 (La/Lu)N 0.14 0.22 0.14 0.48 0.65 0.28 0.75 0.44 0.56 (La/Sm)N 0.51 0.35 0.29 0.66 0.78 0.46 0.78 0.62 0.89 (Gd/Lu)N 0.46 0.93 0.85 0.81 0.80 0.63 1.06 0.76 0.63 Eu anomaly 0.97 1.03 1.38 1.03 1.33 1.06 1.27 0.95 0.96 Ce anomaly 0.88 0.92 1.08 0.89 1.01 1.00 1.00 0.97 0.94 Y anomaly 1.02 1.08 1.14 1.06 0.99 0.84 0.89 1.05 1.08 Enrichment type H H H H H H H&M H H

Figure 7.8 Distribution patterns of REE in the Bulli seam. REE are normalized to Upper Continental Crust (UCC) data from Taylor and McLennan (1985). (A) Coal samples 8118 to 8124; (B) Non-coal samples including roof sample 8117, claystone sample 8119, and floor sample 8125.

211

Figure 7.9 Correlation coefficients between mean individual REE and Y with LTA% in the Bulli coal samples.

7.5 Summary

The roof and floor materials of the Bulli seam contain abundant quartz, and minor but significant proportions of kaolinite, illite and expandable clay minerals. The mineral matter of the coals from the main part of the seam is dominated by well-ordered kaolinite, with minor quartz, carbonates (ankerite and siderite), goyazite and fluorapatite, and in some cases anatase. Both quartz and non-kaolinite clay minerals are also abundant in the lowermost ply of the coal seam, suggesting that the immediate base of the peat bed was made-up of organic matter admixed with the same detrital sediment as supplied to the basin before the swamp was established.

K-feldspar, which is present in the coals and non-coal bands in the lower metre of the Great Northern seam section and restricted to the non-coal bands in the upper part of that seam, is not present, even in the intra-seam non-coal bands of the Bulli seam. This may reflect deposition of the Bulli seam at a greater distance from the sediment source in the New England Fold Belt, to the north of the Sydney Basin.

With the exception of the high-ash, uppermost coal ply, the concentrations of most trace elements in the coals are lower than the respective values for average worldwide coals. Only Li, Sc, Ga, Th and U are generally slightly higher in the Bulli coals than the averages in the worldwide coal dataset. Most trace elements do not show any obvious correlations with the abundance of particular minerals or major oxides in the coal samples. This may be partly due to the small number of coal samples studied for the present chapter. Lithium in the Bulli coals is mainly associated with kaolinite. Beryllium in the coals appears to have an inorganic affinity. Elevated concentrations of many elements, such as Be, Co, Cu, Ni, As, and Tl, are present in a high-ash coal sample. However, overall correlation does not

212

Chapter 7 Mineralogy and Geochemistry of the Bulli Seam exist between these elements and the ash yield. The normalized REE distributions in all the Bulli coal samples have a heavy REE enrichment (H-type) compared to the UCC and the LREE have in general a greater mineral affinity than the MREE and HREE in the Bulli coals.

213

214

CHAPTER 8 MINERALOGY AND GEOCHEMISTRY OF THE SONGZAO COAL SEAMS

This chapter discusses the modes of occurrence of the mineral matter and trace elements in the coal and associated non-coal strata from three relatively thin seam sections (Tonghua, Datong and Yuyang sections) in the Songzao Coalfield. The purposes of this study were to investigate more fully the mineral assemblages, and to evaluate the associations between the different groups of minerals and trace elements within the coals. The study was also expected to provide an opportunity to evaluate the geological factors responsible for the mineralogical and geochemical characteristics of the coal seams, and provide a basis for comparison to the other seam sections included in the research program.

8.1 Coal quality and chemistry

Proximate analysis and total sulphur data for the individual coal plies, and forms of sulphur of selected coals from the Tonghua, Datong and Yuyang sections, are listed in Table 8.1. The variations in proximate analysis results are further shown in Figure 8.1. Overall, the Datong and Yuyang coals have medium to high ash and high sulphur percentages, while the Tonghua coals have high ash and varying sulphur percentages. Based on the volatile matter value and fixed carbon percentage, the Songzao coal is mainly classified as a semi-anthracite under the ASTM classification (ASTM, 2007).

The moisture content of the coal samples appears to be relatively constant within individual seam sections. In general, the ash yield is highest in the Tonghua section, varying from 32.3% to 43.7%, and lowest in the Datong section, varying from 18.2 to 35.4%. The volatile matter is similar in the three sections, in the range of 6.9% to 11.1% (dry, ash-free basis). High volatile values occur in the Tonghua (TH) samples (th-k2b-4, th-k2b-6). As shown below, these contain relatively high proportions of carbonate minerals, and the high volatile yields may be associated with the loss of carbon dioxide from the carbonates during the analysis process. The fixed carbon value shows the opposite trend through the seam sections to the volatile matter value.

215

Table 8.1 Proximate analysis and forms of sulphur (selected samples) and mean maximum vitrinite reflectance value of the Songzao coal samples (%, air-dried basis, unless indicated)

Thickness VM FC Sample M Ash TS SS PS OS Rv,max (cm) ad daf ad daf

dt-7-1 15 2.7 20.6 11 10 65.7 90 6.73 0.78 5 0.95 2.22

dt-7-2 22 2.1 18.2 11 10.3 68.8 89.7 4.89 - - - 2.17

dt-7-3 12 1.9 19.4 10.3 9.4 68.4 90.6 4.62 0.16 4.1 0.36 2.36

dt-7-4 24 2.1 23.2 9.1 8 65.6 92 3.73 - - - 2.33

dt-7-5 20 1.9 35.4 11.7 9.5 51.1 90.5 3.21 - - - 2.29 th-k2b-1 30 2 32.3 9.9 8.1 55.9 91.9 2.97 - - - 2.4 th-k2b-2 7 2.1 43.7 9.3 6.9 44.9 93.1 0.78 - - - 2.42 th-k2b-4 11 1.6 36.6 13.7 11.1 48.2 88.1 10.1 - - - 2.3 th-k2b-6 6 1.7 36.6 13.2 10.7 48.7 89.3 1.22 - - - 2.42 yy-11-1 10 1.4 35.9 8.1 7 53.8 93 8.109 - - - 2.09 yy-11-2 7 1.4 24.1 10 8.9 64.5 91.2 8.13 - - - 2.25 yy-11-3 7 1.4 24.2 11.3 10 63.1 90 13.39 0.51 11 1.88 2.31 yy-11-4 14 1.4 23.6 10.7 9.5 64.3 90.5 10.08 0.6 8.5 0.98 2.4 yy-11-5 8 2 41.3 10.1 7.7 46.6 92.3 3.03 0.21 2.4 0.42 2.28 yy-11-7 11 1.6 31.5 10.2 8.4 56.7 91.6 1.64 - - - 2.25 ad=air-dried basis; daf=dry ash-free basis; M = inherent moisture; FC = fixed carbon; TC = total sulphur; SS = sulphate sulphur; PS = pyritic sulphur; OS = organic sulphur; - = no data.

The Songzao coals are high in total sulphur, with the sulphur being mainly pyritic, consistent with results from other studies (Dai et al., 2007b, 2010a). The total sulphur contents in both the Datong and Yuyang sections increase upwards. It varies considerably, from 1.64% to 13.39%, and is generally higher in the Yuyang section. With the exception of one coal ply (th-k2b-4), total sulphur in the Tonghua section is relatively low.

216

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Figure 8.1 Variation of proximate analysis and vitrinite reflectance data and total sulphur through three seam sections of the Songzao Coalfield.

8.2 Mineralogy of the coal seams

Although the coals are high in pyrite, no obvious oxidation of the pyrite was observed in the Songzao coals from hand specimen examination. The proportion of minerals in each coal LTA and each non-coal sample from each seam section, as well as the LTA percentages of the coals, are given in Table 8.2. The relative proportions of the minerals were also recalculated and normalised to a carbonate-free basis. This was done in order to facilitate a better investigation of the mineralogical distribution and variation within each seam section, without dilution from the carbonates, which are mainly epigenetic. The vertical trends in abundance of the different minerals in the coal LTA and associated non- coal strata in each seam section on a carbonate-free basis are illustrated in Figure 8.2.

217

Data on clay mineralogy was obtained from the < 2 μm fractions of all the coal LTA and non-coal samples (Table 8.3), and the results are expressed in graphic profiles in Figure 8.3.

Figure 8.2 Plots showing vertical variation of abundance of major minerals (normalised to carbonate-free) in the Songzao seam sections.

218

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Table 8.2 Mineralogy of the Songzao LTA and associated non-coal samples by XRD and Siroquant (wt. %)

Sample LTA% Qtz Kao I/S I Alb Cha Py Mar Cal Dol Ank Sid Ana Rut Jar Bas Gp Sz dt-7-0 95.7 4.5 34.0 19.9 16.7 3.2 18.3 0.8 2.6

dt-7-1 28.5 6.4 37.2 10.2 40.4 0.7 1.7 2.9 0.7

dt-7-2 24.2 1.9 54.9 21.3 10.3 1.8 1.4 2.9 5.4

dt-7-3 25.4 2.1 46.9 13.8 7.8 0.8 20.9 4.2 0.6 0.6 0.4 2.1

dt-7-4 31.6 5.3 50.8 24.2 14.4 1.8 1.6 0.1 1.8

dt-7-5 47.6 19.9 35.1 4.5 16.2 8.4 10.4 1.0 1.2 1.8 1.4

dt-7-6 98.4 1.0 22.5 33.1 31.7 6.4 5.4 th-k2b- 91.8 18.3 2.1 25.8 27.3 2.5 23.0 1.0 0 th-k2b- 45.0 15.8 24.2 33.9 5.0 5.0 6.0 2.1 3.5 2.0 2.4 1 th-k2b- 54.6 8.2 48.5 20.4 11.9 2.1 2.9 1.8 3.3 0.8 2 th-k2b- 95.0 0.5 8.0 62.7 17.7 4.8 2.4 4.0 3 th-k2b- 47.3 5.7 16.9 21.9 3.0 2.8 3.0 24.9 0.5 17.4 0.9 0.9 1.2 1.0 4 th-k2b- 95.7 2.7 45.1 32.1 13.0 4.4 0.3 2.5 5 th-k2b- 47.6 14.8 31.7 14.9 4.6 12.0 2.4 14.2 0.5 0.7 0.6 3.7 6 th-k2b- 98.0 0 14.9 56.7 20.0 4.2 4.2 7 yy-11-0 95.0 24.9 5.4 39.4 22.0 5.7 1.1 0.7 0.8

yy-11-1 43.6 56.9 8.4 1.6 30.1 0.5 0.6 1.9

yy-11-2 31.5 37.2 9.6 5.7 1.1 39.3 1.9 0.7 0.8 1.9 1.9

yy-11-3 35.3 15.0 22.7 53.0 1.9 0.3 0.7 2.0 2.9 1.7

yy-11-4 32.8 11.4 23.7 1.9 2.2 51.5 1.3 1.6 0.4 1.6 2.9 1.5

yy-11-5 49.8 10.7 62.3 12.5 2.1 4.3 4.4 1.1 0.5 0.9 1.1

yy-11-6 90.4 1.2 78.4 15.6 3.2 1.5

yy-11-7 37.7 22.3 48.10 6.9 12.0 4.2 3.4 0.3 2.1

yy-11-8 96.5 5.7 18.0 40.8 26.2 2.6 3.3 0.4 2.8

Qtz = quartz; Kao = kaolinite; I = illite; I/S = mixed-layer illite/smectite; Alb= albite; Cha = chamosite; Py = pyrite; Mar = marcasite; Cal = calcite; Dol = dolomite; Ank = ankerite; Sid = siderite; Ana= anatase; Rut= rutile; Gp = gypsum; Bas = bassanite; Jar = jarosite; Sz = szomolnokite

219

8.2.1 Minerals in roof and floor strata

The roof and floor strata of the seam sections are mainly carbonaceous shale. The dominant minerals in the roof and floor strata are generally I/S, illite and kaolinite, with quartz, albite and anatase as minor components. Pyrite is present as a minor mineral in most of these clastic materials, but is relatively abundant in the roof samples of the Datong and Tonghua sections (Table 8.2, Figure 8.2).

The clay mineralogy of the roof and floor samples in the Tonghua and Yuyang sections is dominated by illite and expandable clay (Table 8.3, Figure 8.3). The roof and floor samples in the Datong section contain slightly lesser proportions of kaolinite than the LTA of the coal samples. XRD patterns indicate that the kaolinite in all the roof and floor samples has a poorly-ordered structure.

The expandable clays in the roof and floor samples are mainly I/S, which appears to be regularly interstratified. This is indicated by peaks at around 28 Å in the powder XRD traces, and at around 30 Å in the glycol saturated material. An example is the Tonghua floor sample, th-k2b-7, as shown in Figure 8.4.

Figure 8.3 Column section showing vertical variations in clay mineralogy for the three seam sections, Datong (left), Tonghua (centre) and Yuyang (right), in the Songzao Coalfield.

220

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Figure 8.4 Powder (lower trace) and oriented-aggregate XRD traces of Tonghua floor sample (th-k2b-7), showing peaks due to poorly-ordered kaolinite, regularly interstratified I/S, quartz, albite and anatase, plus aluminum from the sample holder; the upper two traces show more detailed characteristics of the clay minerals.

8.2.2 Minerals in Partings

XRD analysis indicates that minerals in the three non-coal (claystone) partings of the Tonghua and Yuyang sections are essentially clay minerals. Non-clay minerals pyrite, quartz, albite and anatase are present only in trace proportions (Table 8.2). The thick claystone band, th-k2b-3, shows elongated pellets under the optical microscope, mainly consisting of cryptocrystalline to microcrystalline clay minerals (Figure 8.5A). However, the thin claystones, th-k2b-5 and yy-11-6, show a more dense texture, with the clay minerals largely being fine-grained (Figure 8.5B). Quartz grains in the samples mainly exhibit an angular, resorbed structure, which is characteristic of volcanic quartz (Figure 8.5C). The presence of volcanic quartz and chloritised biotite (Figure 8.5D) indicate a volcanic origin for these claystones.

The difference between the microscopic appearance of the thick and thin claystones may be related to the leaching efficiency of the claystone-forming environment. Dense tonsteins (or bentonites) may result from more complete in situ decomposition (Diessel, 1965). The thin claystones, which may have been more efficiently leached, would therefore tend to have a relatively dense appearance.

221

Table 8.3 0LQHUDORJ\ RI ȝP IUDFWLRQ RI FRDO /7$V DQG QRQ-coal strata using oriented-aggregate XRD techniques (wt. %) dt-7-0 dt-7-1 dt-7-2 dt-7-3 dt-7-4 dt-7-5 dt-7-6 Kaolinite 71 86 100 86 78 80 69 Illite 7 0 0 0 4 7 26 Expandable clays 23 14 0 14 18 12 5

Sample th-k2b-0 th-k2b-1 th-k2b-2 th-k2b-3 th-k2b-4 th-k2b-5 th-k2b-6 th-k2b-7 Kaolinite(+Chlorite) 11 76 89 22 78 66 90 40 Illite 25 1 0 28 0 10 0 35 Expandable clays 64 22 10 50 22 25 10 26

Sample yy-11-0 yy-11-1 yy-11-2 yy-11-3 yy-11-4 yy-11-5 yy-11-6 yy-11-7 yy-11-8 Kaolinite 27 100 100 100 100 93 88 93 41 Illite 45 0 0 0 0 0 0 0 26 Expandable clays 28 0 0 0 0 7 12 7 33

Figure 8.5 Thin section photomicrographs of claystone samples. (A) Elongated pellets in th-k2b-3, PPL. (B) A homogeneous texture shown in yy-11-6. (C) Volcanic quartz in lower part of image, yy-11-6, PPL. Note the resorbed angular texture. (D) Chloritised biotite in th-k2b-3, XPL.

222

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Figure 8.6 X-ray powder diffractograms showing minerals in claystone samples (th-k2b-3, th-k2b-5 and yy-11- 6), including poorly-ordered kaolinite (K), I/S, anatase (An), albite (Alb) and pyrite (py). Aluminum (Al) was derived from the sample holder.

8.2.2.1 Clay minerals

Kaolinite Kaolinite in the claystones makes up 21.8% to 88.1% of the clay mineral assemblage. Kaolinite appears to be more abundant in the thin claystones than in the thicker claystones. X-ray diffraction patterns show that kaolinite in the claystones has a poorly- ordered structure or decreased crystallinity (Figure 8.6). Poorly-ordered kaolinite was also noted in the coals from the Bukit Asam deposit, Indonesia, as described by Susilawati and Ward (2006). Kaolinite of vermicular (Figure 8.7A) and tabular (Figure 8.7B) forms is rarely present. When present, these forms usually indicate in situ crystallization.

I/S and illite XRD analysis of the <2 μm fractions of the claystone samples shows relatively more abundant I/S and illite in the clay mineral assemblage than in that of the adjacent coal LTA samples (Figure 8.3). The thick claystone, th-k2b-3, appears to contain more I/S and less kaolinite than the thin claystones. According to Spears (2012), volcanic ash layers that have I/S exceeding 50% of the clay mineral assemblage are more appropriately referred to as K-bentonites rather than tonsteins. The claystone th-k2b-3 is therefore referred to as

223 a K-bentonite in the present study, while the other two claystones, which have kaolinite making up > 50% of the clay mineral assemblage, are regarded as tonsteins.

The XRD patterns also show that the I/S in these claystones is regularly interstratified, indicated by the peaks at around 30, 13 and 9.2 Å in the glycol saturated oriented- aggregate XRD patterns (Figures 8A, B). EDS data show that the I/S in all the Songzao claystones have Na as the predominant cation over K (Figures 8.9, 8.11). Such Na-rich I/S resembles the Na-rich rectorite-like clay in the claystones associated with coal seams at Bukit Asam, Indonesia, described by Susilawati and Ward (2006).

Although intra-seam altered volcanic ashes are typically tonsteins, K-bentonites and metatonsteins (illite-rich altered volcanic ashes) have also been reported in coal seams (e.g. Burger et al., 1990). Bentonites, smectite-rich altered volcanic ashes, are frequently present in Mesozoic marine sediments, which formed when the relevant ions were sufficiently available. Altaner et al. (1984) suggested that the I/S in K-bentonite was produced by the reaction of smectite in the original bentonite with K in the pore fluids, with the K being derived from breakdown of K-bearing minerals (e.g. micas and K-feldspar) in the host rock. In such a case, the I/S in the K-bentonites could be an intermediate product of the conversion of smectite to illite during burial diagenesis (e.g. Spears, 2012).

The I/S in K-bentonite is typically regularly interstratified, with various percentages of illite and smectite layers (e.g. Spears, 1971; Pevear et al., 1980; Altaner et al., 1984), although randomly interstratified I/S has also been reported (e.g. Huff et al., 1998). The I/S in typical K-bentonite which formed in a marine environment, however, is mainly K-I/S. This is probably dependant on the availability of the cations for the I/S formation. Relatively high contents of K in K-bentonite probably reflect both seawater and parent material composition at the time of formation (Huff and Tuerkmenoglu, 1981). The Na in the Na- rich I/S of the Songzao claystones may have been derived from a relevant source during burial diagenesis. Na was probably released as a non-mineral inorganic component from the organic matter during the coal rank advance, especially during anthracitization.

Another possible mechanism for the formation of illite involves the illitisation of a kaolinite precursor, which may take place in high rank coal seams. Burger et al. (1990) described a relationship between the clay mineralogy of metatonsteins (which they described as illite tonsteins) and the rank of the adjacent coals. They found that the proportion of illite increases in tonsteins or metatonsteins that are associated with coals having lower

224

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams volatile matter (VM) percentages (i.e. higher rank), accompanied by chloritization when VM is less than 8%. Susilawati and Ward (2006) described rectorite-like minerals in claystones in coal seams at Bukit Asam, Indonesia, with a vitrinite reflectance ൒1.3%, but the rectorite is absent from the lower rank parts of the Bukit Asam deposit. Susilawati and Ward (2006) also suggested that the rectorite-like clay may have been altered from a kaolinite precursor.

Although the Songzao coals are high in rank, illitisation of kaolinite does not appear to have taken place. This is especially notable in the thin claystones, in which kaolinite is the dominant clay mineral. Given the presence of significant proportions of smectite in the associated coal samples, it is most likely that the volcanic ash may have been originally (at least partly) converted to smectite in the marine-influenced coal-forming environment. The smectite was in turn converted to I/S and illite during diagenesis, assuming that the necessary ions (e.g. K, Na and Mg) were available from the marine water percolating through the peat deposit.

Figure 8.7 Kaolinite in claystone sample th-k2b-5. (A) Thin section photomicrograph of vermicular kaolinite, PPL. (B) SEM image of kaolinite of tabular structure in an I/S matrix. Bright area is Ti-rich material.

8.2.2.2 Accessory minerals

Chlorite XRD patterns of the clay fractions of the tonstein/K-bentonites show the presence of a small proportion of chlorite. As noted above, biotite pseudomorphs with chlorite laminae have also been observed (Figure 8.5D). Chlorite may also form diagenetically from kaolinite in metatonsteins, which have been described as illite tonsteins, in high rank coal seams (e.g. Burger et al., 1990).

225

Figure 8.8 XRD traces obtained from clay fractions of claystones. (A) Claystone th-k2b-3, showing regularly interstratified I/S, kaolinite (K) and chlorite (C). (B) Claystone th-k2b-5, showing regularly interstratified I/S and kaolinite (K). (C) Claystone yy-11-6, showing kaolinite (K) and a trace proportion of I/S.

Quartz Quartz is a minor constituent in all the claystones, making up 0.5% to 2.7% of the mineral assemblage. The proportions of quartz are also much lower than in the LTA of the adjacent coals. The difference is even greater if allowance is made for dilution by the abundant pyrite in the coals. Under the optical and electron microscopes the quartz is mostly angular, and elongated in some cases. Some quartz has fluid inclusions (Figure 8.5C), which indicates that the quartz crystallized from a fluid source. Small proportions of quartz in the claystones indicate that the input of epiclastic sediment into the peat swamp during the accumulation of the volcanic ash layers was rare.

Anatase Small proportions of anatase (1.5% to 4%) are present in all the claystones, at concentrations that are generally higher than those in the LTA residues of the adjacent coal samples. Under the SEM, different modes of anatase occurrence were observed in the claystones (Figures 8.9, 8.10, 8.11). The mineral largely occurs as discrete particles (with a grain size commonly less than 1 μm) disseminated in the I/S matrix (Figure 8.9A), and probably as a replacement of glass shards (Figure 8.11B). Anatase also appears to

226

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams be a replacement of pumice or other volcanogenic components; however, this is not yet clearly understood and needs further investigation. Coal tonsteins throughout the world frequently contain anatase (Price and Duff, 1969). Ruppert and Moore (1993) described anatase occurring as replacement of glass shards and probably glass gas bubbles in an Indonesian tonstein. Anatase is a common secondary mineral in tonsteins, where it may be derived from the break-down of Ti-rich volcanic glass, ilmenite, magnetite or rutile (Ruppert and Moore, 1993).

Phosphates La-Ce-phosphates, probably gorceixite, commonly having a grain size < 2 μm, were observed in claystone sample th-k2b-5 (Figure 8.12), although gorceixite is below the detection limit of the XRD and Siroquant analyses. Such fine-grained material was probably precipitated from the REE-rich leachate of the original volcanic components. Triplehorn and Bohor (1983) reported Ce-bearing goyazite in a kaolinised tuff from Colorado, and suggested that the goyazite was probably precipitated from solution during early diagenesis. REE-rich Ca phosphate, which may be replacing pumice, was also reported in an Indonesian tonstein by Ruppert and Moore (1993).

227

Figure 8.9 SEM images showing the modes of occurrence of anatase in claystone sample th-k2b-3. (A), (B), (C) and (D): Anatase appears to be replacement of pumice or other volcanogenic components.

228

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Figure 8.10 SEM image showing the modes of occurrence of anatase in claystone sample th-k2b-5.

Figure 8.11 SEM images showing the modes of occurrence of anatase in claystone sample yy-11-6. (A), (B) and (C): Anatase appears to be replacement of glass shards and other volcanogenic components. 229

Figure 8.12 SEM image showing REE- and Ba- phosphates, probably gorceixite, in claystone sample th-k2b-5.

8.2.3 Minerals in coal samples

The overall percentage of LTA of the Tonghua coals, which is in the range of 45% to 54.6% (Table 8.2), is the highest among the three Songzao seam sections. The LTA percentage of the Datong coals is relatively low, and is in the range of 24.2% to 47.6%. Minerals in the LTA residues of the coal samples are mainly kaolinite, pyrite (and marcasite in a few cases), I/S, quartz, and minor proportions of carbonates (calcite, dolomite, and ankerite), feldspar (albite), anatase, and secondary sulphate minerals.

8.2.3.1 Clay minerals

Kaolinite In contrast to most of the roof and floor samples, the clay mineralogy of all the coals is dominated by kaolinite (Table 8.3, Figure 8.3), especially in the Yuyang section. With the exception of coal th-2b-1, the XRD patterns show a well-ordered structure of the kaolinite in all the LTAs, including those from coals near the top and bottom of each section. Under the SEM, both detrital and authigenic kaolinite can be recognised. The former occurs as laminae and bands, and the later as infills of cells or crack cavities (Figure 8.23B), and as cleat/fracture infillings (Figure 8.14A). Multiple stages of kaolinite veining is evident (Figure 8.14A), suggesting a number of late diagenetic stages. Vermicular kaolinite also occurs (Figure 8.14B), a feature which indicates in situ precipitation.

As discussed by Ward (1989), the well-ordered kaolinite in the coal appears to be the result of in situ leaching and reprecipitation processes. Although not entirely diagenetic, an increase in the proportion of diagenetic kaolinite leads to an increase in the overall degree of order in the kaolinite as a whole (e.g. Spears, 2012).

230

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

The presence of poorly-ordered kaolinite in an Australian bituminous coal affected by igneous intrusions was described by Ward et al. (1989). Poorly-ordered kaolinite occurs in heat-affected coal from the Bukit Asam deposit, Indonesia, with Rvmax >1.0% (Susilawati and Ward, 2006). However, among the coal samples of the present study, coal th-k2b-1 is the only coal that has poorly-ordered kaolinite. This may indicate that the well-ordered structure of the kaolinite is probably more persistent in coals that have been subjected to burial metamorphism, such as the Songzao seams, than in those subjected to thermal metamorphism.

Figure 8.13 XRD traces obtained from clay fractions of coal LTAs. (A) dt-7-5, showing kaolinite (K), illite (I), and a trace of I/S. (B) th-k2b-1, showing kaolinite (K) and smectite (S). (C) th-k2b-4, showing kaolinite (K), smectite (S) and chlorite (C). (D) yy-11-5, showing kaolinite (K) and probably a trace of I/S.

Expandable clay minerals The Songzao coals have significant proportions of expandable clays in the Tonghua and Datong sections, and the lower few samples in the Yuyang section. However, oriented- aggregate XRD analysis indicates that the proportion of I/S does not exceed that of kaolinite in any coal sample. The upper few coals in the Yuyang section have kaolinite as the only clay mineral in the clay fractions of the LTA residues (Table 8.3, Figure 8.3).

Glycol saturated oriented-aggregate XRD patterns of the Tonghua coals show the 001 reflection at 16.97 to 17.26Å (Figures 8. 13B, C), which indicates the presence of smectite. 231

However, under the SEM, I/S is more frequently observed. The I/S in the Tonghua coals occurs as thin bands and laminae (Figure 8.24B). Such material is probably a diagenetic product of dispersed volcanic ash, which was incorporated in the original peat swamp. I/S and smectite could then have been formed from the alteration of the volcanic ash, due to availability of necessary ions (e.g. K, Na, Mg) in the marine-influenced coal swamp.

Under the SEM, I/S also commonly occurs in cell cavities or pore spaces (Figure 8. 14B), and occurs in cracks within the macerals (Figure 8.24A) in a Tonghua coal (th-k2b-4). The EDS spectrum shows the presence of K, Na and Mg, as well as Al and Si. This suggests that at least some of the I/S in the Tonghua coal is of authigenic origin. As noted below, this coal is overlain by a mafic bentonite (th-k2b-3) and underlain by an alkali tonstein (th- k2b-5). The original alkali volcanic ash was enriched in K and Na, and the mafic volcanic ash was also relatively enriched in Mg. Such ions may have been leached from the ash during diagenesis, which may have then led to the precipitation of I/S in cracks, cell cavities and pore spaces of the maceral components. Although the marine water was also a possible supplier of the alkali elements, authigenic I/S is rare in other coals that are further away from the altered volcanic layers.

Under the SEM, the expandable clay in the Datong coals occurs as irregular flakes, with the presence of K, Na, Mg and Fe indicated by the EDS spectrum (Figure 8. 15A). Such material resembles the I/S in the claystone partings, and was probably diagenetically altered from the original volcanic ash. The ordering of the I/S in the Datong coals, however, cannot be accurately determined, due to the small proportion of I/S present and the poor resolution of the XRD patterns.

An increase in the proportion of I/S, accompanied by a decreasing proportion of kaolinite, has been reported in sediments due to thermal metamorphism or burial metamorphism. A significant decrease in kaolinite was noted in the thermally metamorphosed coal from the Bukit Asam deposit, Indonesia, with a reflectance > 1.41%, and kaolinite disappeared in coals with vitrinite reflectance > 2.2%, associated with the appearance of I/S (Susilawati and Ward, 2006). No significant change in the proportions of kaolinite, however, is observed in most of the Songzao coals, although all the coals are of high rank levels. The most likely precursor for the formation of I/S in coals is smectite, which was mainly pyroclastic. The I/S is probably an intermediate product of the conversion of smectite to illite during burial diagensis, with the Na and K being released from the organic matter

232

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams during the coal rank advance, especially the anthracitization. The formation of I/S from a kaolinite precursor may be less important, if not totally ruled out.

Figure 8.14 SEM images of clay minerals in coal samples. (A) Detrital kaolinite (K) laminae and cleat- and fracture-filling kaolinite (K). Also shown are syngenetic pyrite (Py) and detrital quartz (Q). yy-11-1. (B) Smectite or I/S in pore spaces. Also shown are vermicular kaolinite (K), pyrite (py) and detrital quartz (Q). th-k2b-4.

Illite XRD analysis indicates that small but significant proportions of illite occur in the lower two coal plies of the Datong section (Figure 8.13A). Other coals, including those from the two other sections, have no or trace amounts of illite. SEM data indicate the common presence of Na-rich illite or paragonite, within regular flakes I/S matrix in the Datong coals (Figure 8.15B). Paragonite, if present, however, is below the detection limit of the XRD analysis for both the LTA residues and the clay fractions.

Daniels and Altaner (1990) suggested smectite as a precursor for Na-bearing illite and paragonite formation in high rank anthracites of eastern Pennsylvania, with the Na being provided from metasomatic hydrothermal fluids. Paragonite is also present in the heat- affected coals at Bukit Asam, Indonesia, where it was suggested to have formed from the

233 reaction of kaolinite with inorganic Na released from the coal’s organic matter by igneous intrusions (Susilawati and Ward, 2006).

As indicated in the clay mineral profiles (Figure 8.3), illite generally coexists with expandable clay minerals when present in the Songzao coals. This may suggest the formation of Na-rich illite in the coals from the alteration of smectite. The Na for the formation of Na-rich illite was probably also derived from organically bound Na which was expelled from the organic matter with coal rank advance.

Figure 8.15 SEM images showing clay minerals in coal sample dt-7-5. (A) Na-rich I/S, or paragonite in an I/S matrix. The bright area is probably phosphate. (B) Na-rich I/S, or paragonite, showing regular flakes.

234

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

8.2.3.2 Chamosite

XRD analysis shows small proportions of chlorite (<5%) present in some of the Tonghua coal LTA residues. The EDS spectrum of the chlorite shows peaks of Fe and Mg, as well as Al and Si. EDS data indicate that all the chlorite in the Tonghua coals is Fe-rich, and suggest that it is chamosite. Under the SEM, the chamosite was commonly observed as cell and cleat/fracture infillings, in most cases coexisting with kaolinite (Figure 8.16).

Chlorite having an intergrowth texture with kaolinite has been reported in coals of the South Walker Creek area, Bowen Basin, Australia (Permana, 2011). Dai and Chou (2007) reported chamosite replacement of kaolinite in cell cavities in a semianthracite from the Zhaotong Coalfield, SW China, and suggested that the chamosite was derived from the reaction between kaolinite and Fe-Mg-rich fluids during early diagenesis. Chamosite with a similar mode of occurrence in the No. 12 coal of the Songzao coalfield was also described by Dai et al. (2010). Chlorite in the thermally metamorphosed Bukit Asam coal was also suggested to have formed from reactions between kaolinite and Fe and Mg ions, with the Fe and Mg ions probably driven from the organic matter with rank advance (Susilawati and Ward, 2006).

The intergrowth texture of chamosite and kaolinite (Figures 8.16B, D) in the present study indicates that kaolinite was probably precipitated earlier in the fractures, and is thus the precursor of the chamosite. The formation of the chamosite in the Songzao coal may be the result of reactions between the earlier-precipitated kaolinite and Fe–Mg-bearing fluids during late diagenesis.

Coexistence of chamosite, ankerite and quartz in fractures was also observed in the Songzao coals (Figure 8.17). Well-defined contacts among these minerals indicate that they formed from different fluid injection events. Both the chamosite and the ankerite contain fragments of quartz, indicating that quartz is the earliest-formed mineral, followed by ankerite and chamosite. Such chamosite therefore has a different origin from that in the intergrowths with kaolinite, and was probably formed epigenetically from fluid reactions at a late diagenetic stage. Chlorites formed from epigenetic processes have been reported in several coals, most of which are of high rank levels (Faraj et al., 1996; Dai et al., 2008c, 2012a).

235

Figure 8.16 SEM images of chamosite in coal sample th-k2b-4. (A) Kaolinite and chamosite occurring in fractures probably formed in different stages. The earlier formed fractures are parallel to each other, and are confined in probably a vitrinite band. Note the displacement of the vitrinite band later formed during tectonic deformation. (B) Enlargement of (A), showing chamosite intergrown with kaolinite. (C) Enlargement of (A), showing fracture-filling chamosite and kaolinite. (D) Enlargement of (C), showing chamosite intergrown with kaolinite.

236

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Figure 8.17 SEM images showing modes of occurrence of chlorite in coal sample th-k2b-4. (A) Chlorite (C), ankerite (A) and quartz (Q) in fracture. (B) Enlargement of (A) showing chlorite intergrown with ankerite. (C) Chlorite (C), ankerite (A) and quartz (Q) in fracture. (D) Enlargement of (C) showing chlorite intergrown with quartz.

8.2.3.3 Quartz

Quartz is abundant in most of the coal samples, making up to 57% of the LTA residues. It is less abundant when the coal samples contain abundant pyrite, due to dilution effects. The quartz occurs largely as detrital grains (Figures 8.14A, 8.18A). High proportions of detrital quartz in the coals throughout all the seam sections indicate that detrital material introduced by epiclastic processes was important during peat accumulation. On the other hand, poorly-ordered kaolinite, which also represents detrital material introduced by epiclastic processes, is almost absent in most of the coals. Ward (2002) suggested that there may be greater opportunities for alteration of the detrital input, if any, in dominant and widespread peat-forming conditions. This may be the case for the Songzao coals, and detrital material other than quartz was not preserved.

Authigenic quartz is relatively minor, occurring as euhedral crystals (Figure 8.18B), and cell cavity and cleat infillings. Such authigenic quartz was syngenetically precipitated from silica-bearing solutions. Euhedral quartz also appears to be enclosed by ankerite and

237 chamosite in fractures (Figure 8.17). Quartz in the fractures was precipitated from fluids during the epigenetic stage.

Figure 8.18 Photomicrographs of quartz in Songzao coal samples. In air, reflected light. (A) Abundant detrital quartz grains in dt-7-1. (B) Euhedral quartz in th-k2b-4.

8.2.3.4 Pyrite/marcasite

Pyrite occurs in all the Songzao coal samples in various proportions (3% to 53% of the LTA residues). It appears to be more abundant in the coals from the upper part of each section. Minor marcasite also occurs in some samples, and one coal ply (th-k2b-4) has a particularly high concentration of marcasite (13.9%).

Pyrite has a variety of modes of occurrence in the coals, including isolated or clustered framboids (Figure 8.19A), subhedral to euhedral crystals (Figure 8.19B), cell cavity infillings (Figure 8.19C), and to a lesser extent, fracture/cleat infillings (Figure 8.19D). Pyrite also tends to be disseminated in vitrinite as aggregates in bands parallel to stratification. Under the SEM, pyrite framboids appear to have a different form, and can be seen to consist of pyrite microcrystals, generally < 0.5 μm in diameter, under high magnification (Figure 8.20). Pyrite, especially the microcrystals, also has a tendency to be embedded in a clay matrix (Figure 8.19E). Later-formed pyrite includes overgrowths on earlier-formed subhedral pyrite and massive pyrite on framboids (Figure 8.19F).

Marcasite has tabular and bladed crystal habits under the optical microscope. Marcasite occurs as massive bodies (Figure 8.21A) and as a replacement of pyrite (Figure 8.21B). Although below the detection limit for XRD analysis, marcasite was observed in coal dt-7- 3 under the optical microscope and SEM. The marcasite occurs as radiating crystals growing on pyrite framboids and coated with a layer of massive pyrite (Figures 8.21C, D). Three stages of early diagenesis are shown: Pyrite framboids formed in the early

238

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams syngenetic stage; marcasite precipitated in the middle syngenetic stage; and massive pyrite precipitated in the late syngenetic stage. Spherical grains composed of marcasite without any pyrite nuclei were also observed (Figure 8.21C). Similar occurrences and associations between pyrite and marcasite have been reported in coals from elsewhere (Wiese and Fyfe, 1986; Querol et al., 1989). Epigenetic pyrite, however, is not common in the Songzao coals.

Figure 8.19 Modes of occurrence of pyrite in Songzao coal samples. (A) Clustered and isolated framboidal pyrite. yy-11-4. In air, reflected light. (B) Euhedral pyrite crystals. yy-11-1. In air, reflected light. (C) Cell-filling pyrite. th-k2b-4. In air, reflected light. (D) Cleat-filling pyrite, dt-7-1. Oil immersion, reflected light. (E) SEM image of framboidal pyrite and isolated pyrite crystals in a clay matrix. dt-7-2. (F) SEM image showing framboidal cemented by later-formed massive pyrite. dt-7-2.

239

Figure 8.20 SEM images showing the association of pyrite and marcasite in coal sample dt-7-3. (A) Multistage aggregates of pyrite and marcasite. (B) Enlargement of (A) showing pyrite framboids with different density. (C) Enlargement of (B) showing that pyrite framboids consisting of pyrite microcrystals.

Figure 8.21 Photomicrographs showing modes of occurrence of pyrite and maracasite in coal samples. In air, reflected light. (A) Massive marcasite. dt-7-3 (B) Marcasite with bladed morphology. th-k2b-4. (C) Different associations of pyrite and marcasite. dt-7-3. (D) Association of pyrite (Py) and marcasite (Ma). dt-7-3. 240

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

8.2.3.5 Carbonates

Calcite, ankerite, and dolomite XRD analysis indicates that minor proportions of calcite, dolomite, and ankerite occur in the Songzao coal LTAs. Carbonates in coals from the Datong section mainly occur as cleat and fracture infillings, but the carbonates in coals of the Tonghua section are more commonly present as cell infillings (Figure 8.22A). The mineral matter of the Tonghua seam contains more abundant Mg-bearing carbonates, dolomite and ankerite, than the other two seams. In some cases the carbonates show evidence of formation postdating that of euhedral pyrite in inertinite (Figure 8.22B). Such carbonates are probably epigenetic, precipitated from hydrothermal fluids.

Figure 8.22 SEM images showing modes of occurrence of carbonates in coal sample th-k2b-2. (A) I/S and dolomite coexisting in cell cavities. Cell-filling calcite (Ca) is also indicated. (B) Ankerite (A) enclosing earlier- formed pyrite (Py).

8.2.3.6 Phosphates

Although occurring at concentrations below the detection limit for XRD analysis and Siroquant interpretation, gorceixite was frequently observed under the SEM in coal sample th-k2b-4 from the Tonghua section. The gorceixite particles are up to 2 μm in diameter, and tend to occur particularly in the matrix of I/S-rich bands or pore-filling kaolinite (Figures 8.23A, B). The EDS spectrum shows the presence of the peaks of Ba 241 and P, as well as Al and Si. In some cases a P-bearing material is seen to occur between illite flakes in the Datong coal (Figure 8.15A). Phosphate in that coal also occurs as small particles, generally <0.5 μm in diameter, disseminated in the I/S matrix (Figure 8.15A). As noted above, Sr-, Ba-, Ca-, and REE-aluminophosphate minerals have been reported in tonsteins and or altered tuff layers (e.g. Triplehorn and Bohor, 1983). The presence of such phosphates has also been frequently recognized in coals, most of which are associated with tonsteins, or may have been affected by volcanic ash (Crowley et al., 1989; Crowley et al., 1993; Hower et al., 1999a).

Fine-grained (particle size <2 μm) REE-rich phosphates, tentatively identified as rhabdophane, were also identified under SEM in coal sample th-k2b-4. The phosphate occurs in both Na-rich I/S in cracks of the organic matter (Figure 8.24A) and in thicker I/S bands (Figure 8.24B). Apart from the REE, some of these phosphate particles contain EDS-detectable Ba (Figure 8.24A). The authigenic rhabdophane particles in clay bands or clay-filled cracks in the Tonghua coals may be crystallisation products of REE-rich leachates (probably also Ba-bearing) derived from the overlying tonstein layers.

Figure 8.23 SEM images showing gorceixite in coal sample th-k2b-4. (A) Gorceixite in the matrix of band rich in I/S. (B) Gorceixite (G) in a matrix of kaolinite within pores of the organic matter.

242

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Gorceixite in the Datong coals that does not contain REE was probably crystallised from Ba-bearing fluids. The Sr, Ba and REE in the phosphates found in coal are generally thought to have been leached from overlying tonstein bands or from volcanic components incorporated into the peat (Crowley et al., 1989; Hower et al., 1999). As mentioned in other chapters, the P in the phosphates may have multiple possible sources, for example leaching of the volcanic components or decomposition of plant material (Ward et al., 1996; Rao and Walsh, 1999).

Figure 8.24 SEM images showing REE-phosphates in coal sample th-k2b-4. (A) Fine-grained REE- phosphates, some of which probably contain Ba, in the matrix of crack-filling I/S. (B) Fine-grained REE- phosphates in the matrix of I/S bands.

243

8.2.3.7 Anatase

XRD analysis indicates that small proportions (up to 3.3% of the LTAs) of anatase occur in coals throughout the three sections. Multiple modes of anatase occurrence were observed under the SEM, particularly in coal sample th-k2b-4.

Euhedral anatase crystals are associated with kaolinite, which also contains Ti and appears to be parallel to bedding (Figure 8.25A). The Ti-bearing kaolinite is not chemically homogeneous, as indicated by the back scattered electron SEM image. The association of anatase and kaolinite indicates that the euhedral anatase postdates the kaolinite. As suggested by Ward et al. (1999a), Ti was possibly precipitated in conjunction with kaolinite as a separate phase or incorporated within the kaolinite structure. There were probably multiple stages of Ti-mineral formation, as the euhedral anatase appears to be later-formed. Anhedral anatase containing small proportions of Al and Si also occurs (Figure 8.25B). Coal sample th-k2b-4 contains anatase with unusual morphologies. Although the origin is not very clear, some of the anatase appears to be replacing glass spherules (Figure 8.25C, D), and other anatase probably replaces shell or wood fragments (Figures 8.25 E, F). The bulk of the fragments are also parallel to bedding. EDS data indicate that some of the anatase is Nb-bearing.

As noted below, the coal ply represented by sample th-k2b-4 is overlain by a mafic bentonite (sample th-k2b-3), which is relatively high in TiO2. The underlying alkali tonstein is rich in high field strength elements including Nb. These high concentrations of TiO2 and Nb were probably leached from the original Ti-rich mafic and alkali volcanic ashes, respectively.

Anatase has been generally found in coals that are adjacent to tonstein layers (Dewison, 1989; Ruppert and Moore, 1993), and anatase has been observed replacing glass shard material in an Indonesian tonstein (Ruppert and Moore, 1993). Anatase replacement of maceral components has also been reported in coals (Querol et al., 1989). Dai et al (2007) noted that some fine-grained anatase is distributed in I/S, and also occurs as a cementing material for pyrite particles in coals from the Songzao coalfield.

8.2.3.8 REE-bearing minerals

Veins or fracture fillings comprising what appear to be REE-bearing minerals were identified under the SEM in coal sample th-k2b-4 (Figure 8.26). EDS data indicate the

244

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams presence of two distinct rare earth minerals in these veins. Phase I in Figure 8.26B is bright in the backscattered electron image, and shows a pattern of lamellar crystallisation; the EDS spectrum of this phase shows only the peak of Ce. Phase II, the grey area in the centre of Phase I, contains both EDS-detectable Ce and Ca. A similar association between these two phases is common in all of the REE mineral-filled fractures in this particular coal sample.

The concentrations of lanthanide oxides in the veins of these samples were also determined by an electron microprobe equipped with wavelength-dispersive X-ray spectrometry (WDS), due to the lower detection limit and more accurate quantitative analyses than available using SEM and EDS techniques. The results showed that, except for different calcium contents, both of the phases consist mainly of Nd and Ce with small amounts of La and Y (Table 8.4). Based on the chemical compositions, two REE- carbonates were considered possible: lanthanite-(Nd) [(Nd,Ce,La)2(CO3)3+2O] and kimuraite [Ca(Y,Nd)2(CO3)4+2O].

245

Figure 8.25 SEM images showing modes of occurrence of anatase in coal sample th-k2b-4. (A) Euhedral anatase crystals associated with Ti-bearing kaolinite. (B) Probably an intimate mixture of anatase and kaolinite. (C) Flattened circular bodies of anatase possibly replacing glass spherules. (D) Enlargement of (C) showing detail of the possible replacement texture; the cavities are filled with I/S. (E) Anatase possibly replacing shell or wood fragments. (F) Enlargement of (E). The fine-grained, bright particles are probably Nb-bearing zircon (Zr). 246

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Figure 8.26 SEM images of REE-bearing minerals in coal sample th-k2b-4. (A) Fracture-filling REE minerals. (B) Two different REE minerals in fracture (see EDS images). Grey area contains Ca, and bright areas are without Ca.

Table 8.4 Averages and ranges of concentrations of REE and Ca determined by electron microprobe analyses for REE mineral veins in coal sample th-k2b-4 (wt. %)

La2O3 Ce2O3 Nd2O3 Y2O3 CaO Total

Ca-poor mineral 6.4 24.9 27.1 0.3 0.7 59.4

(18 points) 0-8.8 0-29.5 0-34.3 0-1.2 0-2.4 30.5-66.7

Ca-rich mineral 5.1 19.4 23.4 2.4 13.5 63.8

(10 points) 4.4-7.6 16.2-23.9 19.7-28 0-15.9 9.7-16 56.1-71.6

Laser Raman spectroscopy was further undertaken to evaluate the anionic and water molecules in these minerals, as a guide to better identification. No stretching carbonate ion vibration at around 1080 cmí1, however, was observed in the Raman spectra obtained from the minerals, and thus the presence of carbonate ions was not confirmed. Although accurate mineral species are not certain, this result ruled-out the identification of the materials as REE-carbonates, at least for Phase I. The same conclusions probably also apply to Phase II, although this could not be exactly located in the laser Raman analysis

247 due to the small grainsize. Based on the above chemical and laser Raman analyses, Phase I is thus thought to represent a REE-hydroxide or oxyhydroxide mineral.

Naturally-occurring REE-hydroxides or oxyhydroxides, however, have not been reported in the literature. REE-carbonates, such as kimuraite, are also rare in coal. Only one reference has been found (Seredin, 1998), that mentions lanthanite in Russian coals, but the identification was based solely on the morphology and EDS data, without confirmation from any other analyses.

The REE-minerals in the Tonghua sample were probably crystallized from ascending hydrothermal fluids carrying high REE concentrations. The REE-minerals appear to have formed earlier than the fracture-filling kaolinite, both of which were precipitated from epigenetic fluids during a late stage of the diagenesis. The REE-bearing fluid in the Tonghua coals may be associated with contemporaneous volcanic activity.

8.2.3.9 Sulphates

The iron sulphate minerals jarosite and szmolnokite, found in the samples, are secondary minerals derived from oxidation of pyrite during the storage of the coal. Bassanite and gypsum are also secondary, and were most likely formed from the reaction of organic sulfur with calcium released from organic matter during the low-temperature ashing process. Although non-mineral inorganic matter (e.g. Ca, Mg) is usually more prominent in lower-rank coals (e.g. Ward, 2002; Li et al., 2010), the presence of bassanite and gypsum in the Songzao coal LTAs indicates the occurrence of organically bound Ca in the Songzao coals despite its higher rank level. Other mechanisms for the formation of gypsum and/or bassanite include interaction of sulphuric acid released by pyrite oxidation with Ca-bearing carbonates (Rao and Gluskoter, 1973), but the gypsum or bassanite is not always associated with jarosite and other products of pyrite oxidation (Table 8.2).

8.3 Geochemistry of the Songzao coal seams

8.3.1 Relation between mineralogical and ash chemical data

The relation between ash chemistry and mineralogy for the Songzao coal and non-coal samples was studied to check the reliability of the quantitative XRD data, following the procedure described by Ward et al. (1999a). The inferred chemical composition of the 248

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams mineral assemblages determined by Siroquant was calculated based on the mineral percentages and the chemical stoichiometry or expected chemical composition of the individual minerals, and compared with the actual chemical composition of the high- temperature ash (HTA) from the same samples as determined by XRF (Table 8.5). The inferred chemical composition was adjusted for each LTA sample by deducting the CO2 + and H2O to derive an equivalent to a coal ash analysis. The actual chemical composition determined by XRF was also normalized to an SO3-free basis, to allow for differences in

SO3 retention by the low- and high-temperature ashing processes.

The percentages of each major oxide indicated by both sets of data were plotted against each other (Figure 8.27), to provide a basis for comparing the XRD results to the chemical analysis data for the same coal or parting samples. As discussed for other materials by Ward et al. (1999), the respective data sets are presented as X-Y plots, with a diagonal line on each plot indicating where the points would fall if the estimates from the two different techniques were equal. The comparison is presented as a graphic plot for each major oxide.

Al2O3 and SiO2

The plot for Al2O3 shows a high degree of correlation between the Al2O3 proportions from the Siroquant data and those indicated by the ash chemistry. The SiO2 plot, however, has the majority of the points falling above the diagonal equality line, indicating an overestimation of SiO2 inferred from Siroquant. The overestimation of SiO2 appears to be accompanied by an underestimation of Fe2O3, as discussed below, which is also a dominant oxide in most of the Songzao coal samples.

Fe2O3, CaO and MgO

A relatively strong correlation is observed for the Fe2O3 plot, but with most points plotting slightly below the equality line, indicating an underestimation of Fe2O3 inferred from Siroquant, relative to the observed values. MgO values in the majority of samples, all of which have observed MgO values lower than 1.5%, appear to be underestimated by Siroquant. A fair degree of scatter is observed for the CaO plot, especially where high percentages of CaO are indicated. As discussed above, EDS data indicates that calcite, dolomite and ankerite, in most cases, show chemical variation due to element substitution, e.g. Fe and/or Mn for Mg in the dolomite, which was not allowed for in the stoichiometric calculations used in this part of the study.

249

Table 8.5 Major element analyses of Songzao coal ash and non-coal samples from three seam sections (%), as determined by XRF analysis.

sample HTA SiO2 Al2O3 TiO2 Fe2O3 MgO CaO Na2O K2O MnO P2O5 SO3 LOI

dt-7-0 81.59 45.273 30.467 4.681 11.081 0.803 0.613 1.066 1.277 0.181 0.277 0.475 3.12 dt-7-1 20.93 33.165 22.752 1.448 35.106 0.303 2.086 0.377 0.349 0.028 0.23 2.366 1.84 dt-7-2 19.45 27.165 20.708 1.004 28.924 0.437 9.093 0.206 0.287 0.056 0.147 9.913 3.08 dt-7-3 20.18 35.26 24.967 1.273 25.327 0.569 5.48 0.263 0.584 0.038 0.054 5.037 2.34 dt-7-4 23.49 43.429 29.115 2.893 16.996 0.549 2.216 0.486 1.086 0.018 0.063 2.124 1.75 dt-7-5 37.23 45.675 20.494 1.575 11.7 0.859 9.058 0.291 0.942 0.04 0.073 5.66 2.68 dt-7-6 87.3 48.185 36.03 6.86 1.617 0.409 0.243 1.944 1.378 0.011 0.077 0.215 3.54 th-k2b-0 77.59 53.266 20.789 2.312 14.425 0.924 0.825 1.169 2.532 0.064 0.147 0.709 2.19 th-k2b-1 33.35 46.722 22.927 1.526 12.583 2.153 5.419 0.544 1.029 0.061 0.111 4.25 1.9 th-k2b-2 44.35 51.968 34.261 4.033 2.249 0.568 1.694 0.653 0.89 0.014 0.127 1.352 1.64 th-k2b-3 84.26 46.935 33.886 4.855 4.947 0.633 0.295 2.393 1.543 0.03 0.061 0.299 4.19 th-k2b-4 38.42 21.615 12.624 0.836 35.926 3.811 8.747 0.377 0.46 0.085 0.103 10.656 3.18 th-k2b-5 82.3 51.6 37.654 1.748 1.073 0.664 0.263 1.464 1.671 0.008 0.038 0.062 4.43 th-k2b-6 38.32 35.651 18.772 1.14 18.906 4.474 9.662 0.208 0.453 0.105 0.176 6.604 1.93 th-k2b-7 88.8 47.502 35.39 5.95 2.696 0.735 0.27 2.464 1.53 0.016 0.056 0.22 3.17 yy-11-0 85.14 59.156 23.844 2.158 6.18 1.041 0.333 1.48 2.888 0.034 0.132 0.198 2.29 yy-11-1 35.83 60.272 6.433 1.037 29.467 0.067 0.345 0.039 0.113 0.03 0.034 0.364 0.55 yy-11-2 24.21 45.266 8.637 1.168 37.888 0.238 2.532 0.092 0.17 0.019 0.028 2.666 1.01 yy-11-3 23.14 21.813 8.86 0.728 58.282 0.39 3.961 0.08 0.103 0.027 0.041 4.24 nd yy-11-4 28.63 26.908 13.689 0.945 48.583 0.527 3.358 0.051 0.206 0.033 0.065 3.588 nd yy-11-5 41.77 50.306 32.448 1.424 8.457 0.707 2.284 0.445 0.8 0.025 0.064 1.197 2.04 yy-11-6 78.12 53.278 40.723 1.818 0.471 0.46 0.254 0.736 1.035 0.01 0.023 0.037 2.77 yy-11-7 32.14 52.207 24.92 0.754 7.045 1.306 6.75 0.29 0.494 0.064 0.156 3.015 2.49 yy-11-8 85.27 50.465 30.007 3.444 5.644 1.123 0.365 0.867 3.459 0.028 0.086 0.206 4.26

250

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Figure 8.27 Comparison between proportions of major element oxides in coal ashes and non-coal strata from three seam sections in the Songzao Coalfield, inferred from Siroquant and determined by XRF. The diagonal line represents equality in each plot. Circled points are discussed in the text. Relevant trendlines and squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case.

251

TiO2

Comparison of TiO2 data from the two sources shows a relatively good level of agreement, but with all the points falling below the equality line to a certain degree. It appears, therefore, that the proportions of anatase in most samples were slightly underestimated. Fine grained Ti-phases (mostly <0.5 μm in diameter), which are associated with clay minerals in the Greta (Chapter 5) and Great Northern seams (Chapter 6), were also observed in the Songzao claystones. Fine Ti-phases, with possibly poor crystallinity, may possibly not have been detected by the XRD and Siroquant techniques. Ti was also detected by EDS in kaolinite in some of the coal samples, and this was not allowed for in the stoichiometric calculations.

K2O and Na2O

The plots for K2O and Na2O show relatively scattered correlations. However, with the exception of a few points (circled) that are farthest away from their respective equality lines, both plots are generally parallel to the equality lines. It appears an overestimation of

K2O in most samples is coupled with an underestimation of Na2O by Siroquant. The three points in the circles represent three non-coal samples showing a predominance of sodium over potassium (Table 8.5).

As noted above, EDS data show the common presence of Na-rich illite and Na-I/S, rather than regular K-illite and K-I/S. This indicates an overall low saturation of K and higher saturation of Na than allowed for in the stoichiometric calculations. The calculated K2O and Na2O values from the Siroquant data, however, are based only on the regular illite and I/S. Lesser degrees of K saturation in naturally-occurring I/S was also suggested to explain similar differences in the studies of Ruan and Ward (2002) and Permana et al. (2010).

8.3.2 Geochemical associations in coal samples

Trace element data for the Songzao coal and non-coal samples are given in Table 8.6. In general, the Songzao coals have relatively high concentrations of most trace elements to that of average worldwide coals (Ketris and Yudovich, 2009). This is especially prominent for the lithophile elements, which are usually associated with the ash yield of coal (e.g. Finkelman, 1995; Ren et al., 2006).

252

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

The geochemical results from the Songzao coal samples were analysed using cluster analysis, to identify groups of associated trace elements and major element oxides. The major element oxide percentages determined by XRF analysis of the coal ashes were recalculated to give the percentages of those oxides in the whole coal, before the cluster analysis was undertaken. The major oxide percentages on a whole coal basis are also given in Table 8.6.

Hierarchical clustering was performed using the Pearson correlation coefficients. The likely organic/mineral affinity of the elements in the Songzao coals is indicated by the statistical correlation of the different trace element concentrations with the ash yield. Elements with a strong inorganic affinity would be expected to show a positive correlation to the ash percentage, and those with a strong organic affinity would show a negative correlation. Elements which are the most correlated are firstly linked, and then elements or element groups with decreasing correlation are further linked, until a dendrogram is achieved.

253

Table 8.6 Major oxide and trace element analyses of the Songzao coal and associated non-coal samples from three seam sections (Major oxides in %, trace elements in ppm, unless otherwise indicated. All data on a whole-coal basis. Major element oxides recalculated from the data by XRF analysis. Trace elements determined by ICP- MS/OES analysis)

Sample SiO2 Al2O3 TiO2 Fe2O3 MgO CaO Na2O K2O P2O5 MnO Li Be F Sc V Cr Co Ni Cu Zn Ga Ge As

dt-7-0 36.94 24.86 3.82 9.04 0.66 0.5 0.87 1.042 0.226 0.148 197 4.9 354 62 359.6 167.7 65.4 100.9 191.6 213.4 44 2.21 9.26 dt-7-1 6.94 4.76 0.3 7.35 0.06 0.44 0.079 0.073 0.048 0.006 57 2.1 71 10 46.5 22.3 11.6 19.4 34.5 12.1 8 1.71 3.91 dt-7-2 5.28 4.03 0.2 5.63 0.08 1.77 0.04 0.056 0.029 0.011 57 1.6 64 11.4 108.2 23.1 15.8 15.3 59.5 16 6.1 1.99 2.96 dt-7-3 7.12 5.04 0.26 5.11 0.11 1.11 0.053 0.118 0.011 0.008 69 1.4 41 11.5 96.8 23.8 20.4 22.9 56.4 49.4 8.6 1.65 2.87 dt-7-4 10.2 6.84 0.68 3.99 0.13 0.52 0.114 0.255 0.015 0.004 70 6 76 18.9 156.7 56.3 23.1 41.2 98.3 251.3 10.1 1.12 4.39 dt-7-5 17 7.63 0.59 4.36 0.32 3.37 0.108 0.351 0.027 0.015 107 1.4 96 25.2 154 49.7 35.5 212.9 71.2 33.5 13 1.05 2.46 dt-7-6 42.07 31.45 5.99 1.41 0.36 0.21 1.697 1.203 0.067 0.01 235 4.6 309 65.5 432.9 296.9 22.9 55.1 151.6 84.8 51.4 2.7 <1 th-k2b-0 41.33 16.13 1.79 11.19 0.72 0.64 0.907 1.965 0.114 0.05 53 5.6 1697 47.6 279.2 136.6 51.5 125.1 88.2 81.1 31.2 1.75 36.68 th-k2b-1 15.58 7.65 0.51 4.2 0.72 1.81 0.181 0.343 0.037 0.02 80 3.5 533 33.7 335.4 64.9 43.3 91 334.1 24.3 14.6 1.68 2.85 th-k2b-2 23.05 15.19 1.79 1 0.25 0.75 0.29 0.395 0.056 0.006 185 4.2 536 48.4 456.9 102.6 9.6 38.9 304.5 42.9 24.5 2.45 0.8 th-k2b-3 39.55 28.55 4.09 4.17 0.53 0.25 2.016 1.3 0.051 0.025 186 12.8 1014 65.8 384.5 200.9 32.8 139.9 260.2 109.7 53.3 4.46 1.01 th-k2b-4 8.3 4.85 0.32 13.8 1.46 3.36 0.145 0.177 0.04 0.033 50 2.2 326 19.5 170 31.8 18.7 76.2 81.4 41.7 9 1.12 5.38 th-k2b-5 42.47 30.99 1.44 0.88 0.55 0.22 1.205 1.375 0.031 0.007 211 9.4 1288 33.6 26.9 8.1 3.5 34.3 <2 44.6 75.3 3.12 <1 th-k2b-6 13.66 7.19 0.44 7.24 1.71 3.7 0.08 0.174 0.067 0.04 97 1.6 365 26 159.6 40.6 25.7 70.2 68.6 48.3 12.7 1.45 5.19 th-k2b-7 42.18 31.43 5.28 2.39 0.65 0.24 2.188 1.359 0.05 0.014 233 6.8 1184 60.9 501.7 233.8 21.7 110.7 124.3 86.1 59.1 4.39 <1 yy-11-0 50.37 20.3 1.84 5.26 0.89 0.28 1.26 2.459 0.112 0.029 39 5.8 1635 45.1 182.3 38.6 25.6 42.2 143.6 119.3 39.8 1.98 2.34 yy-11-1 21.6 2.3 0.37 10.56 0.02 0.12 0.014 0.04 0.012 0.011 18 17.4 58 16.6 32.9 15.2 15.4 28.3 18.1 12.5 5.6 3.77 4.81 yy-11-2 10.96 2.09 0.28 9.17 0.06 0.61 0.022 0.041 0.007 0.005 15 6.8 70 10.2 23.5 15 17.2 21.9 16.6 10 6.2 3.56 5.32 yy-11-3 5.05 2.05 0.17 13.49 0.09 0.92 0.019 0.024 0.009 0.006 14 7.8 69 6.1 18.3 10.8 13.1 6 23.4 10.7 6.7 3.4 6.93 yy-11-4 7.7 3.92 0.27 13.91 0.15 0.96 0.015 0.059 0.019 0.009 30 5.9 98 6.7 22.3 15 107 67.7 91.6 11.3 8.5 3.72 6.87 yy-11-5 21.01 13.55 0.59 3.53 0.3 0.95 0.186 0.334 0.027 0.01 110 4.9 530 16.5 25.6 14.3 4.2 9.5 6.6 41.7 26.5 3.84 2.33 yy-11-6 41.62 31.81 1.42 0.37 0.36 0.2 0.575 0.809 0.018 0.008 251 14 1225 27.7 10.7 4.9 9.8 6.1 <2 19.5 62.4 7.85 <1 yy-11-7 16.78 8.01 0.24 2.26 0.42 2.17 0.093 0.159 0.05 0.021 59 7.3 343 15.1 22.7 13.2 3.7 16.1 2.6 8.9 12.2 1.95 1.55 yy-11-8 43.03 25.59 2.94 4.81 0.96 0.31 0.739 2.949 0.073 0.024 75 9.3 2587 52.8 299.6 89.3 30.6 64.5 123.1 182.4 55 3.35 16.6

254

Table 8.6 (continued)

Sample Se Rb Cs Sr Y Zr Nb Mo Ag Cd Sn Sb Te Ba Hf Ta W Hg (ppb) Tl Pb Bi Th U

dt-7-0 3.43 24.05 1.94 613 56.2 536 69 1.26 1.34 0.22 4.65 0.7 1.14 216.9 26.68 6.48 1.76 108 <0.02 16.58 0.18 20.64 6.11

dt-7-1 8.63 1.67 0.18 206 12.8 56 5.6 1.67 0.12 0.09 1.06 0.47 0.16 18.2 3.1 0.5 0.37 459 0.06 16.92 0.35 6.29 2.5

dt-7-2 9.17 1.51 0.18 213 17.6 49 3.8 1.01 0.1 0.1 1 0.53 0.23 14.6 3.18 0.48 0.21 450 0.03 13.79 0.33 6.7 1.85

dt-7-3 12.32 3.16 0.44 133 16.9 67 10 2.37 0.16 0.46 1.79 1.05 0.29 17.6 4.3 0.64 1.09 852 0.02 16.46 0.48 10.32 2.34

dt-7-4 15.3 6.06 0.73 120 23.3 99 10.4 1.66 0.22 0.58 1.65 0.86 0.11 36.6 6.2 0.96 0.54 1102 0.04 46.62 0.32 10.82 2.55

dt-7-5 13.05 8.48 1.07 292 32.4 221 12.4 1.54 0.35 0.22 3.31 1.14 0.2 45.1 9.12 1.34 0.49 679 0.06 10.07 0.48 11.81 3.65

dt-7-6 6.46 21.32 1.17 841 49 611 107.5 1.47 1.69 0.22 5.85 0.78 1.06 266.2 30.88 8.25 2.24 855 <0.02 6.78 0.13 25.13 5.61

th-k2b-0 8.21 45.79 2.94 719 38.3 521 73 20.43 1.47 0.34 5.96 1.27 0.8 173.3 28.52 7.64 0.97 414 0.03 21.68 0.3 25.17 10.18

th-k2b-1 10.81 7.22 1 285 77.1 162 20.2 2.04 0.81 0.25 4.77 0.5 0.36 35 9.9 1.83 0.58 906 0.07 22.81 0.38 17.63 4.16

th-k2b-2 5.28 8.54 1.09 301 42.3 283 51 1.2 0.92 0.15 4.12 0.39 0.18 57.3 17.15 4.01 1.56 196 0.03 10.01 0.42 23.02 4.64

th-k2b-3 10.64 22.92 1.44 1134 36.8 581 82.8 2.19 1.6 0.31 6.18 1.15 0.58 214.5 30.37 7.35 3.17 369 0.05 30.2 0.17 21.5 4.66

th-k2b-4 15.62 3.68 0.53 311 70.3 334 41.6 3.53 0.76 0.39 4.94 0.86 0.29 20.9 15.12 1.42 0.8 781 0.03 23.32 0.41 8.29 4.27

th-k2b-5 3.29 23.1 1.65 825 36.5 857 355.6 1.05 5.16 0.34 13.16 0.75 0.7 183.8 66.78 31.71 1.87 457 <0.02 9.19 0.63 76.84 9.28

th-k2b-6 11.93 3.97 0.59 432 101.4 665 38.1 5.98 0.76 0.34 5.69 0.78 0.41 23.7 26.63 2.04 1.16 361 0.03 11.66 0.39 13.72 4.96

th-k2b-7 3.07 23.97 1.6 971 52.7 739 106.6 2.73 1.67 0.29 6.93 0.72 0.27 204.1 36.06 7.79 3.07 148 0.03 10.62 0.19 24.66 5.04

yy-11-0 9.7 43.5 2.55 729 41.3 531 88.7 1.47 1.47 0.22 5.71 0.92 0.28 167.5 30.5 7 2.89 196 0.03 22.84 0.07 25.76 4.4

yy-11-1 7.88 0.73 0.14 72 12.8 106 7.3 3.2 0.17 0.09 1.44 0.43 0.15 9.1 5.56 0.91 0.7 917 0.5 8.85 0.24 4.89 2.91

yy-11-2 12.05 0.76 0.14 99 14.2 85 6.4 3.49 0.14 0.06 1.25 0.39 <0.1 9 4.62 0.77 0.2 764 0.46 7.8 0.18 5.82 1.81

yy-11-3 14.01 0.51 0.1 112 15.6 68 4.9 5.81 0.12 0.12 0.87 0.43 <0.1 5.2 3.73 0.64 0.38 599 0.37 19.98 0.09 4.04 1.27

yy-11-4 12.57 1.11 0.2 95 21.8 123 11.9 4.05 0.21 0.11 1.26 0.41 <0.1 7.4 6.6 1.29 0.13 861 0.57 16.6 0.13 7.11 2.57

yy-11-5 7.06 6.04 0.99 239 55.6 537 134.2 2.46 2.58 0.28 8.55 0.55 0.18 37.5 32.16 12.69 1.55 443 0.11 21.03 0.3 27.22 6.5

yy-11-6 0.24 12.2 1.98 418 40.3 529 403.8 2.47 6.28 0.13 14.48 0.78 0.43 84.4 31.65 33.27 4.22 93 <0.02 1.2 0.17 29.17 4.79

yy-11-7 6.28 3.72 0.65 204 75.1 773 57.4 1.77 1.12 0.16 6.12 0.36 0.23 19.6 37.79 3.82 0.24 203 0.02 8.19 0.11 23.37 7.79

yy-11-8 10.85 57.98 4.01 794 68.6 1068 155.7 3.4 2.46 0.4 8.6 1.01 0.47 174.5 50.25 12.22 7.57 175 0.09 29.95 0.24 37.73 10.25

255

Associations of elements in the Songzao coals are broadly indicated by the resulting dendrogram (Figure 8.28). Some six groups, and also the statistical correlation coefficients between selected elements and ash yield, are shown in Table 8.7. Apart from the inter-correlation among elements in the same group, each group may also include elements of different sub-groups that have different correlations with Al2O3, ash yield, or the abundance of particular minerals. The possible modes of occurrence of the different elements can be inferred based on the correlation of their concentrations with particular mineralogical abundances, with both of the element and mineral contents recalculated to a whole-coal basis prior to the correlation analysis.

Figure 8.28 Dendrogram developed from cluster analysis on the geochemical data of the coals from three seam sections in the Songzao Coalfield (cluster method, centroid clustering; interval, Pearson correlation; transform values, maximum magnitude of 1).

Group A includes TiO2, V, Cr, Sc, Cu, Rb, Cs, Li, Ba, Na2O, K2O, and Al2O3. With the exception of Sc, Cr, V and Cu, elements in Group A are all strongly correlated with Al2O3, with high correlation coefficients (R>0.79). On the other hand, slightly lower correlation coefficients generally exist between these elements and the ash yield. The elements in this group probably have a common source.

Group B includes Zr, Hf, Sn, U, F, W, Ga, Th, Nb, Ag, Ta. With the exception of Ga and

Th, these elements have relatively strong correlations with Al2O3, with correlation coefficients in the range of 0.5-0.79. Ga and Th stand out in this group, as they have greater affinity with Al2O3 than other elements (R=0.97 and 0.91, respectively). The comparison of Ga against Al2O3 is shown in Figure 8.29F. Ga and Th are clustered in this

256

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams group for their close association with other elements in this group, for example Th with U (R=0.87), and Ga with Nb (R=0.83). The elements in this group also probably have a common source.

Table 8.7 Broad classification of elements according to the results from cluster analysis and correlation coefficients (R) between the content of individual elements (E) and Al2O3 in coal (Correlation coefficients for all inter-element and ash relationships are presented in Appendix 2).

Al2O3 (1), TiO2 (0.79), Rb (0.82), Cs (0.86), Li (0.92), Ba (0.87), Na2O (0.9), Group A R(E-Al2O3) K2O (0.86), Sc (0.73), Cr (0.63), V(0.59), Cu (0.47)

Hf (0.63), Sn (0.72), U (0.69), F (0.8), W (0.74), Ga (0.97), Th (0.91), Nb (0.75), Group B R(E-Al2O3) Ag (0.76), Ta (0.73), Zr (0.5)

Tm (0.44), Yb (0.47), Tb (0.52), Dy (0.5), Ho (0.47), Er (0.49), Y (0.46), Lu (0.43), Group C R(E-Al2O3) La (0.59), Ce (0.57), Pr (0.55), Nd (0.53), Sm (0.51), Gd (0.59)

CaO (1), MgO (0.8), MnO (0.86), Sr (0.77), Eu (0.42), Te (0.66), P2O5 (0.52), Group D R(E-CaO) Mo (0.19)

Fe2O3 (0.95), As (0.89), Hg (0.32), Zn (-0.32), Pb (-0.06), Cd (-0.37), Co (0.35), Group E R(E-(Py+Mar)) Ni (-0.17), Se (0.43), Sb (-0.3), Bi (-0.58)

Group F Be (-0.28), Tl (-0.16), SiO2 (0.84) R(E-Al2O3)

Group C includes all the REE except Eu. The correlation coefficients between REE in this group and the Al2O3 concentration in the coals are in the range of 0.44 to 0.59. Higher correlation coefficients generally exist between these elements and the ash yield (R=0.6- 0.68). This may indicate that REE are mainly moderately associated with ash yield in the Songzao coals.

Group D includes MgO, MnO, Mo, Sr, Eu, CaO, Te, and P2O5. With the exception of Mo, all the elements in this group are strongly or relatively strongly correlated with CaO. MgO,

MnO and CaO, and thus have a carbonate affinity. P2O5 and Sr are probably associated with aluminophosphate minerals, which were detected by EDS analysis in the coals. Molybdenum, although clustered in this group, also has an affinity with the sum of the abundances of pyrite and marcasite (correlation coefficient of 0.66), expressed on a whole-coal basis.

Group E includes Fe2O3, As, Zn, Pb, Cd, Hg, Co, Ni, Se, Sb, and Bi. These elements have no or negative correlation with Al2O3 (R in the range of -0.73 to 0.39) or with ash yield (R in the range of -0.36 to 0.3). The elements in this group are mainly chalcophile elements. However, only As is significantly correlated with pyrite, having a correlation coefficient of

257

0.89. Hg and Se are weakly correlated with the sum of the abundances of pyrite and marcasite in the coals, with the correlation coefficients of 0.32 and 0.43, respectively.

Group F includes Be, Tl, and SiO2, which do not have obvious correlation with each other.

Overall correlation does not exist between Be and quartz (or SiO2), although an elevated Be concentration is present in a quartz-rich coal sample (yy-11-1).

8.3.3 Associations of major elements in the coals

The major element components in the Songzao coals are dominated by SiO2, Al2O3 and

Fe2O3 (Table 8.4). The main carriers of these elements are quartz, clay minerals and pyrite. TiO2 is relatively high in the Songzao coals. EDS analysis indicates that TiO2 occurs in the clay minerals, as well as anatase and fine Ti-bearing phases. A positive correlation exists between TiO2 and Al2O3 (R=0.79) in the ash chemistry (Figure 8.29A).

The TiO2/Al2O3 ratio of the Songzao coal ashes ranges from 0.03 to 0.16. As discussed by

Ward et al. (1999), part of the TiO2/Al2O3 ratio might be attributed to the incorporation of Ti in the aluminosilicate (e.g. kaolinite) structure. The intimate association of anatase (and fine Ti-bearing phases) and clay minerals in the Songzao coals in some cases indicates that kaolinite and TiO2 may be co-precipitated. More TiO2 in some coals also occurs as separate masses of anatase replacing probable volcanic components. The high proportions of TiO2 in the Songzao coals may partly reflect high TiO2 contents in the sediment input to the original peat swamp, derived from the mafic basaltic rocks of the Kangdian Oldland on the western edge of the coal basin (Dai et al., 2011).

Na2O shows great variability in the HTAs of the Songzao coals. Several coal and non-coal samples in the Datong and Tonghua sections are high in Na2O, which is mainly contributed by the presence of albite. High proportions of Na2O in I/S and illite in a few of the coal samples also result in elevated Na2O contents.

Dai et al. (2007) noted the presence of fine grained alabandLWH 0Q6 DERXW ȝP RI hydrothermal origin in the Songzao No. 11 coal, which is probably the most important carrier of manganese. However, alabandite was not observed in the Songzao coals in the present study. Manganese, if present in siderite, is also below the SEM-EDX detection limit. High correlations of MnO-CaO (R=0.86) and MnO-MgO (0.96) indicate that the MnO in the present Songzao coals may be closely associated with carbonates (calcite, dolomite and ankerite).

258

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

P2O5 shows a positive correlation with MnO (Figure 8.29B). As discussed further below,

P2O5 also shows significant correlation with Sr (Figure 8.29C) and Ca (Figure 8.29D), but no correlation with Ba (Figure 8.29C). This indicates that P2O5 mainly occur in aluminophosphates of the goyazite and probably the crandallite groups. Although no correlation exists between P2O5 and Ba, the presence of gorceixite was indicated by the SEM study. Ba in the Songzao coals may have additional sources other than gorceixite (e.g. barite), which were not identified by either XRD or microscope studies.

Figure 8.29 Correlations between selected elements in the Songzao coal samples. (A) TiO2 against Al2O3. (B)

P2O5 against MnO. (C) Sr and Ba against P2O5. (D) P2O5 against CaO. (E) Ba against Al2O3. (F) Ga against

Al2O3. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case.

259

8.3.4 Selected trace elements in coal

As indicated by cluster analysis (Section 8.3.2), many elements are associated with Al2O3 or ash yield of the coals with different degrees of affinity. Apart from the ash yield, other factors may also control the concentrations of different trace elements in the different coal seams.

Sr and Ba The concentrations of Sr and Ba are relatively high in the Tonghua and Datong coals. As noted above, Sr and Ba mainly occur in the aluminophosphates (goyazite and gorceixite), although such minerals were below the detection limit of XRD analysis for many of the coal LTA samples. Some Ba also occurs in the authigenic monazite, as indicated by EDS analysis. As mentioned above, poor correlation between Ba and P2O5 (Figure 8.29C) may indicate that some of the Ba also has sources other than gorceixite. A significant positive correlation, however, was shown in the plot of Ba against Al2O3 (Figure 8.29E). This indicates that a large proportion of the Ba in the Songzao coals is associated with aluminosilicates, probably clay minerals, with only minor Ba occurring in the aluminophosphates (mainly gorceixite and monazite). Alternatively, the correlation between Ba and Al2O3 may indicate a common source.

V, Cr, and Cu The concentrations of V, Cr and Cu are generally high in the Datong and Tonghua coals, especially the latter, with the highest values in the coals being 457 ppm, 103 ppm, and 334 ppm, respectively. The correlations among these elements are strong, with R being 0.96 for V and Cr, and 0.92 for V and Cu (Figure 8.30A). Significant correlation between V and Cr was also observed in the coals from Gunnedah Basin studied by Ward et al. (1999), who suggested that a common magmatic source may be reflected. Glick and Davis (1987) suggested that Cr may have an association with illite in a large number of US coals.

Copper in the Songzao coals has poor correlation with Al2O3 (Figure 8.30C). However, along with V and Cr, Cu shows significant positive correlation with the sum of illite and I/S (Figure 8.30B) (R=0.88, 0.84 and 0.76 respectively). The correlations are apparent in samples with the proportion of illite +I/S >5% (whole coal basis). Siroquant may have difficulties in quantification of illite and I/S when in small proportions (<5%, whole coal basis). This may account for the poor correlations between illite+I/S and these elements.

260

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Cu has been reported to be associated with with sulphides (e.g. Spears and Martinez- Tarazona, 1993; Querol et al., 1996; Kolker, 2012), carbonates (Dai et al., 2005a), and occassionally as Cu sulphides and oxides (Finkelman, 1980). Organically bound Cu has also been suggested in some coals (e.g. Miller and Given, 1986; Querol et al., 2001). However, neither of these associations is indicated in the Songzao coals. Cu is not only associated with illite and I/S, but also appears to have a similar pattern of variation to V and Cr (Figure 8.30).

The correlations between V, Cr and Cu and the sum of illite and I/S may indicate a common source of clastic material supplied to the coal basin, which was in turn probably derived from the mafic basaltic rocks of the Kangdian Oldland. Vanadium, Cr, and Cu are especially concentrated in the Tonghua coals, with the coals of the upper section being more enriched in these elements than those in the lower section. This is mostly likely related to the underlying mafic bentonite. Leaching of the original mafic ash may have led in higher concentrations of these elements in the adjacent coals than in coals without such influence or affected by alkaline ashes.

Figure 8.30 Correlations between selected elements in the Songzao coal samples. (A) Cr and Cu against V. (B) V, Cr and Cu against the sum of illite and I/S, on a whole-coal basis. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case.

Chalcophile elements The plot of As against the proportion of iron sulphides (the sum of pyrite and marcasite) shows a relatively strong and consistent correlation (Figure 8.31A). The correlation trend indicates the presence of approximately 0.3 ppm of As per 1% of iron sulphides, or 30 ppm of As present in the iron sulphides themselves. This ratio is comparable with that in the Greta coals (10 ppm in pyrite) noted in Chapter 5, but much lower than that in the coals from the Gunnedah Basin, Australia (1000 ppm in pyrite) discussed by Ward et al., (1999). Further evaluation of Figure 8.31A indicates that the correlation line intersects the

261 y (As) axis, indicating a value of 1.4 ppm As when the pyrite concentration is zero. This may be due to the presence of small proportions of organically-associated As in the coal 3 samples. This may also be due to the presence of arsenate (AsO4 -), formed during pyrite oxidation with storage. The presence of arsenate has been indicated by XAFS studies in a range of US bituminous coals (e.g. Huggins et al., 1993; Kolker et al., 2000).

Figure 8.31 Correlation of selected elements (As, Mo, Hg, Se, Tl and Ge) with the sum of pyrite and marcasite in the Songzao coal samples, on a whole-coal basis. (A) As against iron sulphides (the sum of pyrite and marcasite). (B) Mo against iron sulphides. (C) Hg against iron sulphides. (D) Se against iron sulphides. (E) Tl against iron sulphides. (F) Ge against iron sulphides. Relevant correlation coefficients (R), obtained from linear regression analysis, are also shown in each case.

As indicated in Figure 8.31B, a positive correlation also exists between Mo and the total iron sulphides in the coals, with correlation coefficient of 0.66. Figure 8.31B shows the presence of approximately 0.2 ppm of Mo with 1% of iron sulphdes, or 20 ppm of Mo

262

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams present in the iron sulphides themselves. An association of Mo with iron sulfides is observed in some coal deposits (Lindahl and Finkelman, 1986; Spears and Zheng, 1999) but not in others (e.g. Ward, 1999). LA-ICP-MS studies of coals from the Black Warrior Basin, USA, by Diehl et al. (2004), indicate that varying but significant concentrations of Mo (<10-582 ppm) and As (<100-27400 ppm) are present in the pyrite of that deposit. The results of the present study show an overall consistent relationship of As and Mo to the proportion of iron sulphides in the Songzao coals. However, there appears to be some variation in Mo concentration in these sulphides, expressed by the scatter of the individual points on the graph.

Mercury and Se are no more than broadly correlated with the iron sulphide contents (Figures 8.31C, D). Both of these elements have been reported to be commonly associated with pyrite in other coals, but the degree of scatter found in the present study is relatively high. The concentration of Hg in the individual coal plies of the Songzao Coalfield is mostly in the range of 0.3 to 0.9 ppm, with the highest value being up to 1.1 ppm. This is much higher than the average Hg concentration of Chinese coals, which is 0.163 ppm (Dai et al., 2012b). A relatively high correlation coefficient exists between Tl and iron sulphides (Figure 8.31E). However, this relationship is dominated by the presence of several samples with very high iron sulphide proportions (>12.4%, on a whole coal basis). The same iron sulphide-rich coals contain the highest concentrations of Ge (Figure 8.31F), despite poor correlation between Ge and the iron sulphides in other coals of the sample suite.

Other chalcophile elements, such as Sb, Pb, Co, Ni, Cu and Zn, show either poor or negative correlations with total iron sulphides, although different degrees of positive correlation have also been reported in other coal deposits. This may reflect the variation of the concentrations of these elements in individual pyrite/marcasites (including Tl and Ge), or simply poor retention of those elements in the pyrite/marcasite of the Songzao coals.

Nb, Ta, Hf, Ga, Th, and U Niobium, Ta, Hf, Ga, Th, and U, also referred to as high field strength elements, are enriched in all the coals in the Tonghua section and two coals near the alkali tonstein band in the Yuyang section (Table 8.6). A similar distribution pattern is observed for Zr, although the variation in Zr may also be derived from contamination of the coals from the zirconia grinding mill used during sample preparation.

263

The concentrations of Nb, Ta and Hf are notably higher in the coal ply (yy-11-5) overlying the tonstein in the Yuyang section. The concentrations of Zr, Hf and REE (including Y) are especially high in the coal plies (th-k2b-6 and yy-11-7) under the alkali tonsteins in the Tonghua and Yuyang sections (Figure 8.32). It is also worth noting that the REE in both the overlying and underlying coal plies are higher that that in the tonsteins in the Tonghua and Yuyang sections.

Elevated concentrations of trace elements in coals near tonsteins are relatively common, and volcanic minerals or volcanic glass are probably the sources for the elevated element concentrations in such coals (Crowley et al., 1989). For example, the enrichment of Zr, Nb, Th and Ce in coals directly above and below tonsteins in the C coal bed of the Emery Coal Field, Utah was reported by Crowley et al. (1989). Crowley et al. (1989) suggested that the mechanism of enrichment for some elements in the coal was leaching of volcanic ash by groundwater and subsequent incorporation in organic matter or authigenic minerals, or, alternatively, the incorporation of volcanic ash in the original peat material. Leaching of the volcanic ash by ground water was used to explain the high concentrations of Zr, Y and REE in the coal directly underlying a tonstein in the Fire Clay coal bed, Kentucky, by Hower et al. (1999a). Similar observations suggesting enrichment of elements due to leaching of volcanic ash beds in coals were also made by Wang (2009).

Zielinski (1985) ascribed the significant mobility of Nb, Ta and REE, which are relatively immobile in most low-temperature environments, during the alteration of volcanic ash to tonsteins to the high leaching efficiency in acid coal-forming swamp. Zielinski (1985) indicated that Zr and Hf were the best resist mobilization in his study. In the present study, although Nb and Ta are significantly enriched in the alkali tonsteins, no significant elevation of them was observed in the adjacent coal samples (Figure 8.32).

The positive correlations between the high field strength elements and Al2O3 in coal most likely indicate a common source, namely the original volcanic ash. In the present study, Nb was also detected in anatase by EDS in the Tonghua coals, and in fine Zr-phases (<0.5 μm), probably zircon, in the tonsteins. The fine Zr-phases in the tonsteins are probably authigenic, and similar phases may also occur in the coal samples. If such material is also present in the coals, it may possibly be overlooked during SEM examination due to the fine particle size. Primary minerals of volcanogenic origin (e.g. zircon) mainly made-up of these elements, however, were not observed in the coals in the present study, and thus are probably not the main carrier of the elements in question.

264

Figure 8.32 Plots showing vertical variation of HTA percentage and selected trace elements in the Songzao seam sections

265

Although the REE mineral veins also lead to high REE concentrations in a Tonghua coal (th-k2b-4), the localisation of the minerals may not account for the elevated REE in all the coal samples adjacent the tonsteins. The main carrier of REE in coals is fine-grained authigenic REE-phosphates, probably rhabdophane, which is also indicated by the SEM study.

8.3.5 Selected elements in non-coal samples

The concentration of TiO2 is as high as 4.86% in the K-bentonite band of the Tonghua section (sample th-k2b-3), which is much higher than that in the tonsteins (around 1.8%).

The ratio TiO2/Al2O3 has been compared with that found in volcanic rocks to identify sediments with a possible volcanic component in coal-bearing sequences, or to indicate the possible composition of the parent magma in many studies (Price and Duff, 1969; Spears and Rice, 1973; Spears and Kanaris-Sotiriou, 1976; Addison et al., 1983; Burger et al., 2002; Dai et al., 2011). In the study of Spears and Kanaris-Sotiriou (1979), tonsteins with TiO2/Al2O3 values of <0.02 and >0.07 are grouped to indicate parent magmas of acid and mafic composition, respectively; those with values in between are thought to represent intermediate ash materials.

The comparison of TiO2 and Al2O3 in the Songzao non-coal samples was potted in Figure

8.33A. The TiO2–rich bentonite, sample th-k2b-3, is plotted in the mafic field. The other two tonsteins are plotted between the two lines indicating TiO2/Al2O3 values of 0.02 and 0.07. The key parameters for the claystones were also plotted in the magma source discrimination diagram of Winchester and Floyd (1977) (Figure 8.33B). The bentonite (th- k2b-3) falls in the alkali basalt field, while the two tonsteins fall in the trachyte and basanite nephelinite field. To some extent, the true ratios of Zr/TiO2 of the claystones are probably less than those shown in the plot, due to possible contamination of the samples from the zirconia grinding mill during sample preparation. However, this may not significantly affect the broad field that the samples are plotted in. It can thus be tentatively concluded that the bentonite and two tonsteins were derived from mafic and alkali ashes, respectively.

As indicated in Figure 8.33, the roof and floor samples generally have high TiO2 contents.

This reflects a high TiO2 content in the detrital sediment input from the source region, probably derived from mafic basalts in the Kangdian Oldland on the western edge of the coal basin.

266

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Figure 8.33 Plots of elements for the Songzao non-coal samples. (A) Comparison of TiO2 and Al2O3 concentrations. The upper and lower diagonal lines represent TiO2/Al2O3 values of 0.07 and 0.02, respectively. (B) Plot of Zr/TiO2 against Nb/Y ratios using the magma source discrimination diagram of Winchester and Floyd (1977).

Two groups of elements were enriched in the mafic K-bentonite and alkali tonsteins, respectively, not only relatively to the coal plies, but to the roof and floor samples in the respective seam section (Figure 8.32). The high field strength elements, including Nb, Ta, Hg, and REE, are enriched in the alkali tonsteins. Another group of elements (V, Cr, Co, Cu, and Ni), most of which are transition elememts, are concentrated in the mafic K- bentonite band (Figure 8.32).

The relatively immobile elements in tonsteins are considered to be present in resistate minerals, such as ilmenite (Ti) and zircon (Zr and Hf), and diagenetic minerals such as kaolinite (Al) and anatase (Ti) (Zielinski, 1985). Spears and Rice (1973) suggested that Ga and Th can be accommodated in kaolinite, and U and probably Y in zircon. Zircon in coal may also contain Nb, Ta and Th (Crowley et al., 1989). These minerals, if incorporated in coal with a volcanogenic origin, may be responsible for enrichment of the immobile elements.

High concentrations of Nb, Ta, Zr, Hf, and REE have been reported in alkali tonsteins, relative to silicic and mafic tonsteins (Zhou et al., 2000; Dai et al., 2011). Dai et al. (2007) suggested that alkaline volcanic ash is responsible for the enrichment of elements such as Nb, Zr, Ga, Hf and REE in the No.11 coal of the Songzao Coalfield, where only clayey micro-sized bands were observed.

267

8.3.6 Distribution and affinity of REE and Y

7KH WRWDO UDUH HDUWK HOHPHQW FRQWHQW 5((< RI WKH 6RQJ]DR FRDOV IURP WKH WKUHH sections ranges between 56 and 772 ppm (Table 8.8). The maximum REE concentration in each section appears to occur in the coal immediately above the floor strata (Table 8.8, Figure 8.29A, C, E). The highest REE concentration in the Datong section is in the lowermost coal ply, which may be related to its high ash yield. In both the Tonghua and Yuyang sections, the coal plies near alkali tonsteins have elevated REE contents. However, all the three altered volcanic ash bands have REE concentrations lower than those in the respective overlying and underlying coal samples.

As noted above, the REE in the Tonghua coals occur in authigenic fine rhabdophane and fracture-filling REE minerals. The low REE concentrations in the partings are probably due to leaching by groundwater during parting formation (Dai et al., 2006a, 2008a). As mentioned above, the rhabdophane in coal th-k2b-4 was probably recrystallized from the leachate of the overlying bentonite. The abundant REE in that coal are also attributed to the REE-bearing vein minerals.

268

Table 8.8 Rare earth elements in the Songzao coal samples and associated strata (REE concentrations in ppm, on whole-coal basis)

dt-7-0 dt-7-1 dt-7-2 dt-7-3 dt-7-4 dt-7-5 dt-7-6 th-k2b-0 th-k2b-1 th-k2b-2 th-k2b-3 th-k2b-4 th-k2b-5 th-k2b-6 th-k2b-7

La 85.56 20.33 19.49 18.38 32.03 61.68 85.92 88.63 92.56 121.1 59.46 100 58.17 151.5 68.12 Ce 187.1 43.26 36.89 38.05 61.1 125.4 172.2 194.2 203.3 239.2 115.1 206.6 117.2 316.3 131.9 Pr 24.32 5.35 4.53 4.68 6.82 14.65 21.57 22.78 24.27 27.36 16.45 24.54 13.88 39 17.82 Nd 91.54 19.59 17.81 18.4 25.51 57.94 77.68 81.41 91.4 99.71 63.58 90.02 49.82 153.4 69.01 Sm 19.43 3.59 3.73 3.43 4.4 10.24 12.66 10.67 18.37 19.69 10.7 16.85 10.38 32.14 16.1 Eu 4.96 0.88 1.05 0.68 0.82 2.08 2.59 1.44 3.73 3.26 2.38 3.75 2.65 5.81 2.68 Gd 23.72 3.05 4 3.37 5.08 9.95 20.11 16.8 19.17 20.19 16.37 15.99 19.48 25.83 22.52 Tb 2.72 0.46 0.63 0.52 0.74 1.29 2.29 1.73 2.78 2.2 2.4 2.64 2.9 3.5 2.98 Dy 14.52 2.58 3.37 3.02 4.25 6.6 13.08 10.29 14.99 10.45 13.74 14.32 17.82 18.98 15.28 Y 56.21 12.85 17.64 16.87 23.29 32.44 48.98 38.27 77.11 42.29 36.78 70.26 36.54 101.4 52.67 Ho 2.79 0.51 0.7 0.61 0.92 1.33 2.48 1.98 3.17 1.92 2.51 3.19 3.44 3.92 2.66 Er 7.36 1.41 1.8 1.7 2.59 3.87 6.96 5.72 8.36 5.49 6.48 8.71 9.26 10.4 6.78 Tm 0.99 0.2 0.24 0.25 0.35 0.51 0.96 0.77 1.23 0.73 0.76 1.36 1.21 1.45 0.88 Yb 6.45 1.27 1.48 1.53 2.26 3.42 5.75 4.69 7.27 4.76 4.71 8.19 7.34 8.79 5.43 Lu 0.96 0.19 0.21 0.24 0.32 0.52 0.88 0.67 1.07 0.67 0.64 1.29 1.05 1.36 0.77 REE 472 103 96 95 147 299 425 442 492 557 315 497 315 772 363 Eu/ Eu* 1.07 1.24 1.27 0.93 0.8 0.96 0.73 0.48 0.93 0.76 0.81 1.07 0.81 0.95 0.64 Ce/ Ce* 0.93 0.94 0.9 0.93 0.94 0.95 0.91 0.98 0.98 0.95 0.84 0.95 0.94 0.94 0.86 Y/Y* 0.67 0.85 0.87 0.94 0.9 0.83 0.65 0.64 0.85 0.71 0.47 0.79 0.35 0.89 0.62

(La/Lu)N 0.95 1.13 0.98 0.81 1.07 1.27 1.05 1.41 0.93 1.92 0.997 0.83 0.59 1.19 0.95

(La/Sm)N 0.66 0.85 0.78 0.8 1.09 0.9 1.02 1.25 0.76 0.92 0.83 0.89 0.84 0.71 0.63

(Gd/Lu)N 2.08 1.34 1.58 1.17 1.34 1.62 1.93 2.11 1.51 2.52 2.17 1.04 1.56 1.6 2.47 Type H-M L-M H-M H-M L L-M L L H-M L-M M M M L-M H-M

269

Table 8.8 (Continued)

Sample yy-11-0 yy-11-1 yy-11-2 yy-11-3 yy-11-4 yy-11-5 yy-11-6 yy-11-7 yy-11-8

La 98.11 24.25 11.69 20.84 30.15 69.42 44.3 160.9 122.1 Ce 215.3 47.44 23.26 41 60.41 128.6 83.98 304.5 244.6 Pr 25.64 4.75 2.56 4.5 6.91 13.99 8.33 32.81 24.22 Nd 94.73 15.57 9.08 15 24.36 47.5 26.09 114 70.8 Sm 17.4 2.69 1.81 2.75 4.28 8.35 5.03 19.46 12.26 Eu 3.15 0.4 0.28 0.37 0.56 1.09 0.81 2.18 2.69 Gd 20.98 2.78 1.54 2.38 4.53 8.92 9.93 17.58 26.77 Tb 2.02 0.36 0.32 0.46 0.67 1.59 1.22 2.84 3.11 Dy 11.04 2.16 2.13 2.72 4.05 9.54 7.45 15.31 18.37 Y 41.25 12.77 14.24 15.65 21.85 55.61 40.32 75.1 68.58 Ho 2.09 0.47 0.48 0.58 0.85 2.09 1.48 3.03 3.39 Er 6.11 1.43 1.37 1.64 2.48 5.88 4.19 8.32 9.75 Tm 0.85 0.21 0.21 0.23 0.35 0.82 0.54 1.2 1.42 Yb 5.33 1.47 1.3 1.4 2.2 5 3.25 7.42 9.52 Lu 0.78 0.22 0.21 0.22 0.33 0.74 0.44 1.13 1.3 REE 503 104 56 94 142 304 197 691 550 Eu/ Eu* 0.76 0.68 0.78 0.67 0.6 0.59 0.49 0.55 0.63 Ce/ Ce* 0.98 1 0.97 0.96 0.95 0.94 0.99 0.95 1.02 Y/Y* 0.65 0.96 1.07 0.95 0.9 0.95 0.92 0.84 0.66

(La/Lu)N 1.35 1.19 0.61 1.01 0.98 0.998 1.07 1.52 0.999

(La/Sm)N 0.85 1.35 0.97 1.13 1.06 1.25 1.32 1.24 1.49

(Gd/Lu)N 2.28 1.08 0.63 0.91 1.16 1.01 1.89 1.31 1.73 Type L-M L H L H H L L H

270

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

The correlation coefficient between individual REE and the ash yield is mainly in the range of 0.6 to 0.7, except for Eu, which has a correlation coefficient with ash of 0.58 (Figure

8.34). The REE exhibit less significant correlations with Al2O3 (R=0.4-0.6). This indicates that the REE generally have similar mineral affinities. Ce may have slightly greater organic affnity than the other REE. A stronger organic affinity for heavy REE has been observed in some coal deposits (e.g. Eskenazy, 1999; Dai et al., 2012d), as well as in the Greta coals of the present study (Chapter 5), while the light REE exhibit greater organic affinity in other coals (e.g. Dai et al., 2008a).

Figure 8.34 Correlation coefficients between mean individual REE and Y with ash yield in the Songzao coal samples

The REE concentrations in each coal and non coal sample were also normalised against the Upper Continental Crust (UCC) (Taylor and McLennan, 1985), in order to obtain a more clear indication of the distribution patterns (Figure 8.35). Seredin and Dai (2012) classified the distribution of REE into three enrichment types, namely L-type (light REE type enrichment), M-type (medium REE enrichment type) and H-type (heavy REE enrichment type). All the three types, as well as a mixed type, of REE enrichment based on this classification occur in the Songzao coal and non-coal samples (Table 8.8).

The REE enrichment in most of the Datong and Tonghua coals is dominated by a mixed type, either L-M or H-M type. The Yuyang coals have either L-type or H-type enrichment. Nevertheless, the dominance of M-type for the Datong and Tonghua coals indicate a different REE source from that for the Yuyang coals. As discussed by Seredin and Dai (2012), an M-type of REE plot, normalized to the UCC, may be due to the circulation of acid natural water, including acid hydrothermal solutions with high REE concentrations, in the coal basin.

271

Despite the diversity in the enrichment types, relatively flat REE distribution patterns tend to exist for the coals in the upper part of the Datong (Figure 8.35A) and Yuyang (Figure 8.35E) sections. The non-coal samples, including the claystone partings, show no obvious Ce anomalies, pronounced negative Y anomalies, and either no obvious or negative Eu anomalies (Figures 8.34B, D, F).

Figure 8.35 Distribution patterns of REE in the three seam sections. REE are normalized to Upper Continental Crust (UCC) data from Taylor and McLennan (1985). (A) Coal samples in the Datong section. (B) Rock samples in the Datong section. (C) Coal samples in the Tonghua section. (D) Rock samples in the Tonghua section. (E) Coal samples in the Yuyang section. (F) Rock samples in the Yuyang section.

272

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

8.4 Summary

The coals from the three seam sections in the Songzao Coalfield are mainly high ash, high sulphur semianthracites. XRD analysis indicates that minerals within the Songzao coals are mainly kaolinite, pyrite (or marcasite in some cases), and quartz, with minor proportions of carbonates (calcite, dolomite, and ankerite), feldspar (albite), anatase and sulphate minerals. Separate oriented-aggregate XRD study indicates that significant proportions of illite and expandable clays also present, especially in some of the Datong and Tonghua coals.

The expandable clay in the Tonghua coals is mainly I/S and smectite. The lower two coals of the Datong section contain significant proportions of illite and I/S, and at least some of the illite and I/S is Na-rich. A significant change in the proportions of kaolinite is not observed in other Songzao coals, although all the coals are of high rank levels. This may indicate that the precursor for the formation of I/S and illite in coals is probably smectite, which was mainly pyroclastic. The I/S in the Songzao coal and claystones is an alteration product of the original dispersed volcanic ash, due to the availability of necessary ions (e.g. K, Na, Mg) in the marine-influenced coal swamp. Organically-bound Na, which was expelled from the organic matter with coal rank advance, especially with anthracitization, probably supplied additional Na for the formation of Na-rich illite and I/S.

Cell or pore-filling I/S also occurs in a Tonghua coal ply that is overlain by a mafic bentonite and underlain by an alkali tonstein. K, Na and Mg for the formation of such authigenic I/S were probably derived from the leaching of the adjacent alkali tonstein and mafic bentonite. Such ions may be leached from the volcanic ash during diagenesis. Although the marine water was also a possible supplier of the alkali elements, authigenic I/S is rare in other coals that are away from the altered volcanic layers.

The minerals in the two thin intra-seam claystones are essentially poorly-ordered kaolinite, regularly interstratified I/S, and illite, with pyrite, quartz, albite and anatase present in mior proportions. Kaolinite appears to be more abundant in the thin claystones than in the thicker claystone. The relatively thick claystone (th-k2b-3), which contains 50% I/S in the clay mineral assemblage, is referred to as a K-bentonite. The other two thin claystones, which have > 50% kaolinite, are referred to as tonsteins.

273

The volcanic ash layers in the peat swamp may have been originally converted to smectite under the marine-influenced coal-forming environment. The smectite was in turn altered to I/S and illite during diagenesis and/or rank advance, assuming that necessary ions (e.g. K, Na and Mg) were available from the marine water percolating in the peat swamp. Na-rich I/S may have been formed in the claystones, with the additional Na probably being released from the organic matter during the coal’s rank advance. Thin tonstein bands in the seams contain relatively high proportions of kaolinite. This may be due to a high leaching efficiency derived from contact with the organic matter.

A coal underlain by the bentonite in the Tonghua section contains abundant fine-grained authigenic rhabdophane, which was probably precipitated from leachates derived from the overlying bentonite. REE minerals also occur as fracture infillings in that coal, which were probably crystallized from ascending hydrothermal fluids carrying high REE concentrations. The REE-bearing fluid in the Tonghua coals may be associated with contemporaneous volcanic activity. Two species of REE minerals, probably REE- hydroxides or oxyhydroxides, and REE-carbonate were tentatively identified. Naturally occurring REE-hydroxides or oxyhydroxides, however, have not previously been reported in the literature. REE-carbonates are also rare in coal.

The geochemistry of the Songzao coals has also been affected by the adjacent tonstein/bentonite bands. Coals near the alkali tonsteins in the Tonghua and Yuyang sections are high in Nb, Ta, Hf, Ga, Th, U, and REE. Coal samples overlying the mafic bentonite in the Tonghua section are high in TiO2, V, Cr, Zn and Cu. Leaching of the original mafic ash may have led to higher concentrations of these elements in the adjacent coals than in coals without such influence, or in coals affected by alkaline ashes. The coals in the Datong section, which has neither visible tonstein layers nor obvious volcanogenic minerals, have concentrations of TiO2, V, Cr, Ni, Cu, and Zn in intervals between coals affected by mafic and alkaline volcanic ashes. This is consistent with the suggestion that a common source material was supplied to the coal basin, derived from the mafic basaltic rocks of the Kangdian Oldland.

In the Songzao coals, only As and Mo show positive correlations with iron sulphides. No definitive correlations have been found between other chalcophile trace elements (e.g. Sb, Pb, Co, Ni, Cu and Zn) and iron sulphides. This may reflect wide variations of the concentrations of these elements in individual pyrite/marcasites (including Tl and Ge), or simply poor retention of those elements in the pyrite/marcasite of the Songzao coals. Lead,

274

Chapter 8 Mineralogy and Geochemistry of the Songzao Coal Seams

Cu, Sn and Sb are positively correlated with Al2O3, rather than pyrite, probably indicating a common source.

The normalized REE distributions (compared to the UCC) of the Great Northern and Bulli coals show heavy REE enrichment (H-type distribution). H-type distribution patterns may be attributed to the circulation of water which was enriched in HREE through coal basins (Seredin, 2001). This is consistent with the occurrence of epigenetic minerals (mainly carbonates) in the Great Northern and Bullli coals. The REE enrichment in most of the Datong and Tonghua coals is dominated by a mixed type, either L-M or H-M type. The Yuyang coals have either L-type or H-type enrichment. The dominance of M-type enrichment for the Datong and Tonghua coals may indicate a different REE source from that for the Yuyang coals. An M-type of REE plot may reflect the circulation of acid natural water, including acid hydrothermal solutions with high REE concentrations, in the coal basin (Seredin, 2001). However, further investigation of the mechanism of enrichment and distribution of the REE in the Songzao coals is beyond the scope of the present study.

275

276

CHAPTER 9 COMPARISON AND INTERGRATION

In the previous chapters, the abundance, modes of occurrence, and origins of the mineral matter in the coal and associated non-coal strata from coal seams of the Sydney Basin and Songzao Coalfield have been discussed. Also included are discussions on the relationships between trace elements and mineral matter components within the coals.

This chapter is intended to integrate the findings from the individual studies, and to draw some more general conclusion. More specifically, it aims to compare the results obtained from the different analytical techniques used in this study, to evaluate the role of the different inorganic processes and sedimentary inputs that were associated with coal formation, and to evaluate different depositional conditions and geological factors that may have controlled the abundance and distribution of minerals and elements, including trace elements, within the various coal seams.

9.1 Cross-checking of different analytical techniques

As noted in Chapter 4, a variety of analytical techniques were applied to the coal and non- coal samples involved in this study, not only to assess the mineralogical and geochemical characteristics, but also to provide a basis for cross-checking of the results obtained from the different analytical techniques.

9.1.1 XRF and XRD

XRD is generally regarded as a tool for mineral identification and assessment on a semi- quantitative basis. In this study, the reliability of the quantitative XRD data obtained from the coal LTA and the non-coal samples was checked by comparison with the observed ash chemistry determined by XRF. Before the comparison, data from the two methods

277

+ were both normalised to allow for difference in LOI, CO2 and H2O percentages from the calculated data, and SO3 retention by the ashing processes.

The results show that most of the major element components from the two methods were relatively consistent. For all the coal seams studied, SiO2 and Al2O3 generally show a high degree of consistency, due to the presence of relatively abundant clay minerals (mainly kaolinite) and quartz in the coal seams studied. The correlations of other major oxides, such as Fe2O3, CaO, MgO, K2O, and Na2O, from the two methods, depend on the abundance of relevant minerals and the mode of occurrence of these elements. The correlations of these elements thus vary for the different case studies. For example, pyrite-rich and carbonate-rich coal seams generally show good correlations for Fe2O3, and CaO and MgO, respectively. The occurrence of organically-associated elements (e.g. Ca and Mg), mainly in lower rank coals, can result in underestimation of these elements by XRD and Siroquant, relative to the actual ash chemistry as determined by XRF.

The calculated major oxide values from the Siroquant data, however, are based only on the chemical compositions of the relevant minerals published in the literature. Variation in the actual chemical composition of minerals, such as Ca, Mg and Fe in carbonates, may lead to different degrees of consistency for these elements as inferred or determined by the two methods. Na is underestimated in most of the samples from the all the coal seams studied, due to lesser degrees of K saturation in naturally-occurring I/S. The degree of underestimation of Na varies for the different cases, depending on the degree of Na saturation in the illite and I/S in the relevant samples.

The crystallinity of minerals may also affect the correlations. Anatase, when present in the samples analysed, is generally underestimated by XRD and Siroquant. This is probably due to the poor crystallinity, small particle size and small proportions in the mineral assemblages, which may result in at least some of the components not being detected by the XRD and Siroquant techniques.

9.1.2 XRD and SEM-EDS

The SEM examination of coal and associated non-coal samples, provides direct evidence of the modes of occurrence of the minerals in the coals, including those in concentrations below the detection limit of the XRD analysis. For example, the fine-grained rhabdophane and fracture-filling REE minerals in the Songzao coals, which were not detected by XRD

278

Chapter 9 Comparison and Intergration analysis, were identified using SEM-EDS. Minerals occurring in small proportions may be overlooked by XRD analysis, due to their small or poorly-resolved peaks. Anatase is indicated as being in proportions of less than 1% in the most of the LTAs of the Sydney Basin and Songzao coal samples by XRD analysis. The occurrence of fine-grained anatase crystallites in kaolinite and in the matrix of other clay minerals is further confirmed by SEM-EDS.

XRD analysis indicates the occurrence of Sr-, Ba-, or Ca- aluminophosphate minerals in one or more coal samples from each coal seam in the study, as well as the intra-seam claystones samples from the Greta seam. During the quantitative XRD analysis, only one end-member of the aluminophosphate series is included, and the Sr-, Ba-, and Ca- aluminophosphates are referred to as goyazite, gorceixite, and crandallite, respectively. However, in each case where these minerals occur, the XRD pattern of the sample does not match exactly with that mineral, or with the other end-members. Thus it is most likely a solid solution of the end-member phases. Further analysis using SEM-EDS provides more information about the composition of the minerals. For example, apart from Al and P, EDS data for the aluminophosphates in the Greta claystones show that Sr is the dominant cation; Ba and in some cases, trace amounts of Ca also occur, and hence it is more precisely identified as a barian goyazite.

9.1.3 ICP-MS/OES and XRF

Trace element concentrations in the coal and rock samples were determined by ICP-MS or ICP-OES. These techniques were used because of their low detection limits, which are typically in the ppb to ppt range for most trace elements. XRF techniques are ideal for the determination of major elements, including Si, Al, Ti, Mg, Ca, Fe, K, Na, Mn, and P in coal ash and rock samples, and have also been extended to determinations of trace elements in coal (Huggins, 2002). In addition to ICP-MS/OES, the concentrations of many trace elements in the Songzao coals were also analysed using “regular” XRF, and XRF with PANalytical Pro-Trace software. This was intended to evaluate the consistency of trace element results determined by ICP and XRF techniques. It was also intended to check whether the ICP results may have been affected by difficulties with mineral dissolution. The concentrations of selected trace elements are presented as X-Y plots, with a diagonal line on each plot indicating where the points would fall if the results from the two different techniques were equal (Figures 9.1, 9.2).

279

The plots for a number of trace elements, including Sr, Y, Cu, Nb, Th, Zn and Zr, show relatively strong correlations (Figures 9.1), although there is somewhat scatter of the individual points on the graphs. This indicates that XRF techniques generally give consistent results for these elements with ICP techniques. Comparison between results by “regular” XRF and XRF Pro-Trace shows that the concentrations of most elements from the former are somewhat higher than that from the latter. In most cases, results from XRF and XRF Pro-Trace are comparable, and sometimes, Pro-Trace gives results that are closer to that from ICP analysis (e.g. Figures 9.1B, H).

There is, however, a slight but significant difference between the concentrations of Zr indicated by the two XRF techniques and that indicated by the ICP analysis (Figure 9.1H). In this case the Pro-Trace results are significantly higher than the ICP concentrations, although slightly higher Zr concentrations are also indicated by the more basic XRF data compared to those from the ICP techniques. A possible explanation for the difference is some degree of incomplete dissolution of the Zr-bearing phase(s) during the digestion used to prepare the solutions for ICP analysis, However, the consistency in the differences between the ICP and XRF (especially the Pro-Trace) results suggests that calibration factors associated with the XRF analysis are probably responsible for the differences observed in the results for Zr obtained by the two techniques.

For other trace elements, a fair degree of scatter is observed, such as in the examples (V, Cr, Ga, Th, La, Ce, Nd, and Sm) shown in Figure 9.2. However, a lesser degree of scatter is observed for most elements, e.g. V (Figure 9.2A) and Cr (Figure 9.2B), determined by XRF-ProTrace than by the “regular” XRF processing system. Exceptions are Ga (Figure 9.2C) and Th (Figure 9.2D), for which the concentrations from the “regular” XRF are closer to the ICP results.

280

Chapter 9 Comparison and Intergration

Figure 9.1 of Sr (A), Y (B), Cu (C), Nb (D), Th (E), Pb (F), Zn (G), and Zr (H), determined by ICP and XRF, XRF Pro-trace techniques. All data in ppm. Relevant trendlines and the squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case.

281

Figure 9.2 Comparisons of V (A), Cr (B), Ga (C), Th (D), La (E), Ce (F), Nd (G), and Sm (H), determined by ICP and XRF, XRF Pro-trace techniques. All data in ppm. Relevant trendlines and the squared correlation coefficients (R2), obtained from linear regression analysis, are also shown in each case.

The comparison above indicates that, at least for the trace elements discussed, XRF techniques are able to provide comparable results to the data derived from ICP analysis.

282

Chapter 9 Comparison and Intergration

In most cases, but not always, the Pro-Trace calibration algorithm enables a better correlation of trace element concentration in the Songzao coals to the ICP data. Satisfactory results for trace elements at sub-ppm levels in coal, however, would still be provided by ICP or other techniques (e.g. AFS), because of their low detection limits or high sensitivity characteristics.

9.2 Processes of mineral formation

As discussed in Chapter 2, mineral matter may be formed by different processes during the different stages of coal formation. The origin and precipitation processes involved in mineral formation in coals are usually indicated by the modes of mineral occurrence, and also the associations of the minerals with the organic matter.

9.2.1 Detrital processes

Detrital processes mainly include input of sediment into the original peat swamp by epiclastic and pyroclastic processes. Materials from both of these sources may then be subjected to leaching in the coal-forming environment.

9.2.1.1 Epiclastic materials

The typical detrial minerals in the coal seams studied are quartz, disordered kaolinite, non-kaolinite clay minerals (mainly illite and I/S) and feldspars (albite and orthoclase). Minerals of epiclastic origin are particularly abundant in the coal plies near the roof and floor strata, or in epiclastic horizons within the seam. This suggests that the original peat included more admixed epiclastic minerals derived from the sediment source region supplied to the basin in the early and late stages of seam formation, before and after the peat swamp was fully established.

Such detrital minerals are only a minor part of the mineral matter in the coal samples away from the roof and floor strata and any epiclastic intra-seam horizons. As discussed by Ward (2002), the input of epiclastic sediment into the main part of the peat may have been reduced, because of the filtering effect of the vegetation, flocculation of minerals caused by the acid or saline swamp waters, or peat formation in raised mire environments. Ward (2002) also suggested that there may be greater opportunities for alteration of the detrital input, if any, in such dominant and widespread peat-forming conditions. This may

283 be the case for the Songzao, Bulli and Great Northern coals, where detrtial material other than quartz was not preserved. The presence of detrital K-feldspar in the Great Northern coal samples may reflect deposition close to the sediment source area in the New England Fold Belt. Such material, however, is absent from the mineral matter of the more distal but otherwise similar Bulli seam.

Although also located close to the New England Fold Belt, no detrital minerals, including quartz, are present in significant proportions in the coal samples from the main part of the Greta seam. This may be due to almost complete sediment by-passing of the depositional system, if peat formation took place in front of the slowly encroaching sea and under conditions of continued subsidence (Diessel, 1992).

9.2.1.2 Pyroclastic materials

Pyroclastic materials blown into the peat swamp may either be incorporated into the peat or become inherent constituents of the coal itself, or form intra-seam altered volcanic ash layers (tonstein or K-bentonite bands) within the coal seams. Minerals clearly of pyroclastic origin were not identified in any of the coal samples in this study. This may be due to intensive leaching in the swamp environment, and thus the dispersed volcanic components, if admixed with the peat, have not been preserved. Visible tonstein/K- bentonite bands occur in the Great Northern seam and the coal seams of the Songzao Coalfield, with mineralogical and geochemical indicators of a volcanic origin. These are discussed more fully in section 9.3 below.

9.2.2 Syngenetic processes

Minerals formed by syngenetic processes are those authigenically precipitated in place, during or shortly after peat accumulation. The syngenetic minerals in the coals studied are mainly clay minerals, quartz, pyrite and siderite, and occur largely as infillings of cell cavities or pore spaces, as well as euhedral crystals. Syngenetic pyrite and siderite also typically occur as microcrystalline framboids and nodules, respectively.

Authigenic processes are thought to have been dominant in the middle parts of the original peat accumulations. Alternatively, the bulk of the detrital minerals in those parts of the seams, if they were originally present, may have been leached by diagenetic or post- diagenetic processes. This may have resulted in the formation of soluble ions that were

284

Chapter 9 Comparison and Intergration removed from the system, or in the deposition of authigenic minerals elsewhere within the peat bed.

9.2.2.1 Clay minerals

Syngenetic kaolinite is the most abundant clay mineral in the coal samples. It characteristically has a well-ordered structure, probably as the result of in situ leaching and reprecipitation processes within the peat swamp (Ward, 1989). However, kaolinite in the marine-influenced coal seams (Greta and Songzao) is not as dominant a part of the mineral matter as that in the less-marine-influenced coal seams (Great Northern and Bulli). The marine water may have provided less favourable conditions for kaolinite formation, due to higher concentrations of relevant ions for precipitation of non-kaolinite clay minerals.

The presence of a Na-rich I/S in the Greta coal seam, occurring as microcrystalline infillings of cell cavities and pore spaces within the macerals, also indicates an authigenic origin. Cell or pore-filling I/S is particularly abundant in the Tonghua coal ply that is overlain by a mafic bentonite and underlain by an alkali tonstein. K, Na and Mg for the formation of such authigenic I/S were probably derived from the leaching of the volcanic ash that formed the adjacent alkali tonstein and mafic bentonite during diagenesis. Although the marine water was also a possible supplier of the alkali elements, authigenic I/S is less common in the Songzao coals that are away from such altered volcanic layers.

9.2.2.2 Pyrite

Syngenetic pyrite is abundant in the Songzao coals and the coals in the upper split of the Greta seam, mainly occurring as euhedral crystals, framboids and pore infillings, and as a massive form with a porous structure. Pyrite in the Greta coals sometimes occurs as a replacement of wood structures. All of these forms indicate that the pyrite is largely a syngenetic precipitate, formed during or shortly after peat accumulation.

The formation of syngenetic pyrite requires the availability of dissolved ferrous iron, and also of H2S from bacterial reduction of sulphate in the peat swamp. The sulphate may have been supplied by the swamp water, or introduced from waters that penetrated into the peat after its accumulation (Diessel, 1992; Ward, 2002). The Greta coal seam, as the lower coal seam in the Greta Coal Measures, is thought to have been influenced by

285 marine water after the peat was buried. The seam sections of the Songzao Coalfield were probably subjected to marine influence during the peat accumulation.

9.2.2.3 Siderite

Syngenetic siderite nodules or concretions are relatively common in coals from the Great Northern and Bulli seams, and sometimes the lower split of the Greta seam. The relatively common presence of syngenetic siderite, instead of sulphide minerals, indicates deposition of the coal mainly under non-marine conditions, or at least under the influence of swamp or formation waters with a low sulphate content (Ward, 2002).

9.2.2.4 K-feldspar

As discussed further in Chapter 6, and by Zhao et al. (2012), vein and cell-infilling K- feldspar occurs in the lower part of the Great Northern seam in the Newvale section, including the siltstone floor, the lowermost claystone parting, and the coal plies in between. Feldspar with such a mode of occurrence has not been reported in other coal deposits. The apparent precipitation of K-feldspar in shrinkage fissures of vitrinite appears to indicate hydrothermal activity in the late syngenetic stage, before the peat underwent diagenesis. The K-feldspar is tentatively identified as adularia, a low-temperature K- feldspar most commonly associated with hydrothermal mineralization in epithermal deposits. The hydrothermal activity may have occurred in a relatively localized area, given that similar vein minerals are rare in the Catherine Hill Bay siltstone floor, and that the feldspar in the floor at that location largely occurs as detrital grains.

9.2.2.5 Other minerals

Minor albite coexists with K-feldspar in the cleat of some Great Northern coal samples, and was probably co-precipitated with adularia from the same hydrothermal fluids.

9.2.3 Epigenetic processes

Minerals formed by epigenetic processes are those have been precipitated from evolving pore fluids during burial diagenesis (Spears and Caswell, 1986), and are typically deposited in post-depositional cleats, fractures or other shrinkage fissures. A number of epigenetic minerals occur in the studied coal seams, including kaolinite, quartz, carbonates, and other accessory minerals.

286

Chapter 9 Comparison and Intergration

9.2.3.1 Chlorite

Fracture-filling chamosite and kaolinite, probably formed at different stages, occurs in one of the Tonghua coal samples from the Songzao Coalfield. The chamosite also shows an intergrowth texture with kaolinite, indicating that kaolinite probably precipitated earlier in the fractures, and is thus the precursor of the chamosite. The formation of chamosite in the Songzao coal may be the result of reactions between the earlier-precipitated kaolinite and Fe–Mg-bearing fluids during late diagenesis.

Coexistence of chamosite, ankerite and quartz in fractures was also observed in the same Tonghua coal sample. Well-defined contacts among these minerals indicate that they formed from different fluid injection events. Both the chamosite and the ankerite contain fragments of quartz, indicating that quartz is the earliest-formed mineral, followed by ankerite and chamosite. Such chamosite therefore has a different origin from that in the intergrowths with kaolinite, and was probably formed epigenetically from fluid reactions at a late diagenetic stage.

9.2.3.2 Carbonates

The most common epigenetic carbonates in the coal samples from the Sydney Basin and the Songzao Coalfield are calcite, dolomite and ankerite, These largely occur as epigenetic cleat/fracture infillings in the bright coal layers, as well as in cell cavities and pore spaces. Abundant veins of these minerals, generally oriented parallel to bedding, also occur in the topmost ply of the Great Northern and Bulli seams, immediately below the roof material. Along with the cleat infillings, the bedding-parallel carbonates are clearly post-depositional; the veins near the top of each seam may have been derived from expulsion of organically-associated Ca and Mg from the macerals.

Abundant dawsonite veins commonly cut through coal plies made up of different megascopic lithotypes mainly in the coals from the lower split of the Greta seam. Dawsonite also occurs as thin bands, within inertinite macerals, and, in some cases, shows intergrowth with kaolinite in cleats. Such forms of dawsonite indicate that the mineral was probably precipitated epigenetically from fluids at one of the latest stages of the coal’s burial history. The intergrowth of kaolinite and dawsonite in the cleat appears to suggest that the kaolinite was a precursor of the dawsonite. The Greta coal had not been subjected to substantial rank advance, and the Na released from organic matter, if there is any, may not have supplied enough material for the formation of such abundant dawsonite.

287

Thus, it is speculated that the fluid that the dawsonite was precipitated from was probably introduced at a relatively high coal rank, after the fracture and cleat patterns had been developed.

9.2.3.3 Sulphides

Late stage cleat-filling pyrite sometimes occurs in the Greta and the Songzao coals. However, the abundance of pyrite in this form is relatively minor. In both coals, and, as discussed above, the sulphides occur largely as syngenetic precipitates within the seams.

9.2.3.4 Other minerals

Minor albite occurs in some cell or clearly broken cavities in the Greta coal samples; this material may also be epigenetic. The coexistence of dawsonite and albite in the Greta coals may indicate that both of these minerals were formed epigenetically at later stages of the coal’s burial history, and the albite was probably derived from the same Na-rich hydrothermal fluid as the dawsonite. Active Al and Si ions may have been released from dissolution of detrital minerals (e.g. quartz and feldspars) or biogenic material (e.g. siliceous sponge spicules or diatoms). The formation of dawsonite, as noted above, may have required a kaolinite precursor.

REE minerals occur as fracture infillings in one of the Tonghua coal samples. Two species of REE minerals, probably REE-hydroxides or oxyhydroxides on the one hand, and REE- carbonate on the other, were tentatively identified. The REE-minerals appear to have formed earlier than the fracture-filling kaolinite, although both were precipitated from epigenetic fluids during a late stage of diagenesis. The REE-bearing fluid in the Tonghua coals may have been associated with contemporaneous volcanic activity.

9.2.4 Other minerals formed by diagenetic processes

Some diagenetic minerals in the coal samples studied are probably precipitated from leachate of the intra-seam claystone bands, possibly due to the passage of groundwaters. These minerals tend to be precipitated as fine-grained particles, disseminated in pore or crack spaces of clay minerals.

288

Chapter 9 Comparison and Intergration

La-Ce-phosphates, probably rhabdophane (grain size commonly < 2 μm), occur in the coal sample underlying the tonstein sample, although the rhabdophane is below the detection limit of the XRD and Siroquant analyses.

Sr-Ba-aluminophosphates (Barian goyazite) in the Greta claystones occur as subhedral to euhedral crystals incorporated in the kaolinite or I/S matrix, indicating an authigenic origin. Although there is lack of clear indications of volcanic origin of the claystones, the barian goyazite may also be precipitated from the leachate during the formation of the claystones. The occurrence of gorceixite was also observed in the claystone sample itself (e.g. K- bentonite sample th-k2b-5). Such fine-grained material was probably precipitated from the REE-rich leachate of the original volcanic components.

9.3 Mineralogy and geochemistry of tonstein/K-bentonite bands

The intra-seam claystone bands in the Songzao coals contain altered biotite and altered glass shards, which indicate a volcanic origin. Anorthoclase-sanidine and graupen to vermicular kaolinite occur in the upper two intra-seam claystones of the Great Northern seam. According to the classification described by Spears (2012), claystone sample th- k2b-5 from the Tonghua section of the Songzao Coalfield, which has I/S exceeding 50% of the clay mineral assemblage, is referred to as a K-bentonite; claystone sample th-k2b-3 from the Tonghua section and the upper two claystones of the Great Northern seam, which consist essentially of kaolinite, are referred to as tonsteins.

EDS data for the K-feldspar crystals in the Great Northern seam indicate the presence of an anorthoclase-sanidine series or a sodic sanidine material, which suggests an silicic to intermediate volcanic ash input to form the upper two partings in this coal seam. Bentonite sample th-k2b-3 and tonstein sample th-k2b-5, from the Songzao Coalfield, are inferred from the geochemical data to have been derived from mafic and alkali volcanic ashes, respectively.

Apart from the clay minerals and K-feldspar, all the tonstein/K-bentonite bands from the two locations typically contain trace proportions of quartz and anatase, as well as albite in some cases. Although a lesser proportion of quartz than in normal mudrocks is not necessarily a characteristic specific to tonsteins/K-bentonites, the low quartz content serves in conjunction with other evidence to indicate a volcanic origin.

289

Anatase is common in the Great Northern and Songzao tonstein/K-bentonite bands. The anatase in the Great Northern tonsteins is inferred to represent reprecipitation products of chemically leached components of the original volcanic ash material. Anatase in the Songzao tonstein/K-bentonite samples appears to be a replacement of glass shards, pumice or other volcanogenic components, as well as discrete particles (size <1 μm) disseminated in the clay mineral matrix. Anatase in the Songzao tonstein/K-bentonites is thus also diagenetic, derived from the break-down of Ti-rich volcanic components.

9.4 Factors influencing the geochemistry of the coal seams

9.4.1 Ash yield

The concentrations of most trace elements in the Great Northern, Bulli and low-ash Greta coals are lower than the averages for worldwide coals, while trace element concentrations are generally high in the Songzao coals. As indicated by cluster analysis, many of the elements in the high-ash Songzao and Greta coals are associated with Al2O3 or with the ash yield of the coals with different degrees of affinity. Apart from the ash yield, however, other factors may also have controlled the concentrations of different trace elements in the different coal seams.

9.4.2 Nature of the sediment source

The nature of sediment source region on the margin is generally the dominant factor for the background values of trace elements in coal basins (Dai et al., 2012b). High TiO2 contents generally occur in the roof and floor samples of the Songzao coal seams. This reflects a high TiO2 content in the detrial sediment input from the source region, probably derived from mafic basalts in the Kangdian Oldland on the western edge of the coal basin. For the coals in the Datong section, which has neither visible tonstein layers nor obvious volcanogenic minerals, the concentrations of TiO2, V, Cr, Ni, Cu, and Zn in the intervals between the coal plies are affected by mafic and alkaline volcanic ashes. This is consistent with the suggestion that a common source material was supplied to the coal basin, derived from the mafic basaltic rocks of the Kangdian Oldland.

290

Chapter 9 Comparison and Intergration

9.4.3 Influence of volcanic ashes

The intra-seam claystones in the coal seams present in this study were derived from volcanic ashes with different geochemical compositions. A large number of elements have been leached during the alteration of volcanic ashes to the claystones. This has also affected the geochemical and mineralogical compositions of the adjacent underlying and, sometimes overlying coal plies.

The volcanic claystones in the Songzao coal seams were derived from mafic and alkaline volcanic ashes. Two groups of elements were enriched in the mafic K-bentonite and alkali tonsteins, respectively. The high field strength elements, including Nb, Ta, Zr, Hf, Ga and REE (including Y), are enriched in the alkali tonsteins. The other group of transition elements, including V, Cr, Co, Cu, and Ni, are concentrated in the mafic K-bentonite band.

The relatively immobile elements enriched in the altered volcanic ashes also tend to be enriched in the adjacent coal plies, possibly due to leaching by groundwaters. The coals near the alkali tonstein bands in the Tonghua and Yuyang sections are high in Nb, Ta, Hf, Ga, Th, U, and REE. Coal samples overlying the mafic bentonite in the Tonghua section are high in TiO2, V, Cr, Zn and Cu. Leaching of the original alkali and mafic ash may have led to higher concentrations of relevant elements in the adjacent coals than in coals without such influence. Alkali and alkali earth elements (e.g. K, Na and Mg) that were expected to have been leached from the volcanic materials were not found to be particularly abundant in coal samples near the alkali tonsteins. Most of these elements liberated during the formation of tonsteins may have been removed due to their sifinificant mobility in the acid coal-forming environment. A small proportion of K, Na and Mg, however, may have been fixed in the nearly coal plies in the forms of non-kaolinite clay minerals.

The tonsteins of the Great Northern coal seam were derived from silicic to intermediate volcanic ashes based on the geochemical and mineralogical compositions. Trace elements, such as Li, Th, and U, are relatively high in most of the coal plies adjacent the tonsteins, relative to those away from the tonsteins. These elements are also generally enriched in the tonstein samples, relative to the concentrations in the roof and floor samples. The influence of the tonstein materials on the geochemistry of the adjacent coals, however, is not as significant as that in the Songzao coal seams. This may be partly because that the source materials of the Great Northern coal seam are also of silicic to

291 intermediate compositions, indicated by the trace-element associations of the roof and floor samples of the coal seam. The similar sources of tonsteins and mineral matter in coal may make the chemical difference much less. However, the sediment source region of the Songzao coals is the basalt in the Kangdian Oldland, and the source of the tonsteins is alkaline; such different sources can result in greater difference in elemental compositions. On another matter, elements that were originally enriched in the volcanic ashes (e.g. SiO2) in the the Great Northern coal seam may have largely removed from the deposit during the alteration from volcanic ashes to claystones or during the diagenetic stage, rather than being fixed in the adjacent coal plies; however, the fate of the SiO2 is still unknown (Zhou and Ren, 1992; Spears, 2012).

9.4.4 Modes of occurrence of trace elements

In the sulphur-rich Songzao and Greta coals, most of the chalcophile trace elements show either poor or negative correlations with total iron sulphide contents. In the Songzao coals, only Hg and Se show positive correlations with iron sulphides; only Hg, Tl and As are positively correlated with pyrite in the Greta coals. This may be because the pyrite in the Songzao and Greta coals is mostly of syngenetic origin. The absence of traditional pyrite- metal associations may reflect wide variations in the concentrations of these elements in individual pyrite/marcasites, or simply poor retention of those elements in the pyrite/marcasite of the relevant coals. In both the Songzao and Greta coals, some chalcophile elements are correlated with Al2O3, may indicating an association with aluminosilicates, probably clay minerals, rather than iron sulphides. Or more likely, this indicates a common source of chalcophile elements and alumina.

In addition to the lithophile elements, chalchophile elements in the Great Northern coals, including Se, Pb and Cu, also appear to be associated with kaolinite, rather than any other mineral phases. This may indicate that more elements tend to be associated with clay minerals in coals that are poor in sulphur. Most of the trace elements in the Bulli coals, however, do not show any significant correlations with the abundance of any mineral. This may partly be due to the small number of coal samples, and the relatively low concentrations of the elements in question.

In the Songzao coals, Cu, V and Cr show significant positive correlations with the sum of illite and I/S, with all the elements and minerals being expressed on a whole coal basis. Similar patterns of variation in Cu, V and Cr in different coal samples also appear to exist.

292

Chapter 9 Comparison and Intergration

These correlations may indicate a common source of clastic material supplied to the coal basin, which was in turn probably derived from the mafic basaltic rocks of the Kangdian Oldland, as indicated in section 9.4.2. Vanadium, Cr, and Cu are especially concentrated in the Tonghua coals, with the coals of the upper section being more enriched in these elements than those in the lower section. This is mostly likely related to the underlying mafic bentonite. As indicated in section 9.4.3, leaching of the original mafic ash may have led in higher concentrations of these elements in the adjacent coals than in coals without such influence or affected by alkaline ashes.

9.4.5 REE distributions

The rare earth elements in the coals have various affinities with the mineral matter of the different coal samples. The correlation coefficient between individual REE in the Songzao coals and the ash yield is mainly in the range of 0.58 to 0.7, indicating that the REE generally have similar mineral affinities. In the Great Northern and Bulli coals, however, the LREE generally have higher correlation coefficients with the ash yield than the MREE and HREE, which indicates a greater mineral affinity of the LREE than the MREE and HREE. The LREE and MREE in the Greta coals have generally higher correlation coefficients with the ash yield than the HREE, which indicates a general greater affinity of the LREE and MREE to the mineral matter of the coal than that of the HREE.

293

294

CHAPTER 10 CONCLUSIONS

The mineralogy and geochemistry of the individual coal plies, roof and floor strata, and intra-seam non-coal bands in the Greta, Great Northern, and Bulli seams of the Sydney Basin, and three seam sections in the Songzao Coalfield has been investigated. A range of analytical techniques have been used to obtain relevant data, including optical microscopy, electron microscopy/microprobe analyses, quantitative X-ray diffraction, chemical techniques (ICP-MS/OES, CV-AFS, HG-AFS and Eschka method), and also Laser Raman spectroscopy.

The mineral matter in the coals from the Sydney Basin and the Songzao Coalfield was formed by a range of geological processes, including syngenetic, detrital and epigenetic processes, during different stages of coal formation. Various factors, such as the depositional environment, distance from the source area, and incorporation of volcanic ash or influence of volcanic ash layers, have each had an impact on the mineralogical and geochemical characteristics of the individual layers within the coal seams.

10.1 Mineralogy of the coal seams

The Greta coal is a high-volatile bituminous coal and typically contains a high proportion of vitrinite and liptinite. Although the Greta seam is not immediately overlain by the marine strata, the percolation of marine water is indicated by the petrological, mineralogical and geochemical characteristics. The upper section of the Greta seam has several different indicators of marine influence, such as anomalously low vitrinite reflectance and abundant syngenetic pyrite. The coals in the upper and lower sections of the seam have contrasting mineralogy. Pyrite typically comprises 40 to 56% of the mineral assemblage in the coals from the marine-influenced upper part of the seam section, especially the section sampled from the Austar borehole. The lower section of the seam, however, contains only minor pyrite, indicating a lesser marine influence. It also contains relatively abundant dawsonite, which may be the result of reactions between earlier-precipitated kaolinite and Na2CO3- or

NaHCO3-bearing fluids.

The minerals, including most of the clay minerals, pyrite, siderite and quartz, that occur within most of the Greta coal plies are largely of authigenic origin. Authigenic Na-rich I/S is relatively common in the Greta coals. However, as there seems to be no indication of any

295 external heat sources associated with the Greta coals, which might have produced such Na-rich phases, the Na-rich I/S is thought to have been syngenetically precipitated, after the peat had accumulated, with abundant Na and relatively minor K ions required to form the I/S being supplied by the marine water. This is different from the situation in a more normal acid swamp environment, where the formation of kaolinite as the only syngenetic clay mineral is favoured, due in part to the low availability of other cations from the pore water.

Detrital minerals, such as quartz grains, poorly ordered kaolinite, illite and I/S, mainly occur in the Greta coal plies near the epiclastic horizons of the seam. This indicates a greater amount of clastic influx into those parts of the original peat-forming swamp, when the coal-forming environment was becoming established or when it was in the process of being overwhelmed by other sedimentary deposits. Detrital minerals are rare in the coals away from the epiclastic horizons. This probably reflects almost complete sediment by- passing of the depositional system after the swamp had been established. Alternatively, the bulk of the detrital minerals, if they did infiltrate into the peat-forming environment, may have been leached by early diagenetic processes.

The beds of the Great Northern coal seam are mainly high volatile A bituminous in rank. The mineral matter of the Great Northern coal beds, especially those in the middle and upper parts of the seam, is dominated by authigenic kaolinite with a very low abundance of quartz. Apart from the intra-seam tonsteins, authigenic processes therefore appear to be the dominant mechanism of mineral matter formation. Authigenic K-feldspar also occurs in the lower few metres of the seam, with a variety of unusual modes of occurrence including cell and cleat infillings, cross-cutting veins, and thin laminae parallel to the organic matter and the detrital clay bands. A late syngenetic low-temperature hydrothermal fluid injection process is suggested for formation of these feldspar veins. The origin of the fluid is uncertain, but it was most likely associated with contemporaneous volcanic activity.

The uppermost two claystone bands of the Great Northern seam are tonsteins, consisting mainly of well-ordered kaolinite with graupen to vermicular textures. Idiomorphic crystals of K-feldspar within these tonsteins may represent members of the anorthoclase-sanidine series or a sodic sanidine, and indicate an acid to intermediate volcanic ash input. In contrast, the lowermost parting was largely derived from epiclastic sediment, admixed with

296

Chapter 10 Conclusions minor volcanic material such as high-temperature quartz and a different type of K-feldspar component.

The coal plies of the Bulli seam are mainly medium volatile bituminous in rank, and typically contain high proportions of inertinite. The mineral matter of the coals from the main part of the Bulli seam is dominated by well-ordered kaolinite, with minor quartz, carbonates (ankerite and siderite), goyazite and fluorapatite, and in some cases anatase. Both quartz and non-kaolinite clay minerals are also abundant in the lowermost ply of the coal seam, suggesting that the immediate base of the peat bed was made-up of organic matter admixed with the same detrital sediment as supplied to the basin before the swamp was established. K-feldspar, which is present in the coals and non-coal bands in the lower metre of the Great Northern seam section and also occurs in the non-coal bands in the upper part of that seam, is not present, even in the intra-seam non-coal bands of the Bulli seam. This may reflect deposition of the Bulli seam at a greater distance from the sediment source in the New England Fold Belt, to the north of the Sydney Basin.

Some of the illite and I/S in the lower two coals of the Datong section is Na-rich. No significant change is observed in the proportions of kaolinite in the mineral matter of the other Songzao coals, although all the coals are of high rank levels. Unlike some other coals of semianthracite rank, there is no indication that the kaolinite has been converted to (ammonian) illite.

The precursor for the formation of I/S and illite in the coals was probably smectite, which was mainly of pyroclastic origin. The I/S in the Songzao coal and claystones is thought to be an alteration product of contemporaneous dispersed volcanic ash, and was due to the availability of necessary ions (e.g. K, Na, Mg) in the marine-influenced coal swamp. Organically-bound Na, which was expelled from the organic matter with coal rank advance,

297 especially with anthracitization, may have supplied additional Na for the formation of Na- rich illite and I/S.

Cell or pore-filling I/S also occurs in one of the plies of the Tonghua seam which is overlain by a mafic bentonite and underlain by an alkali tonstein. K, Na and Mg for the formation of this authigenic I/S were probably derived from the leaching of the adjacent alkali tonstein and mafic bentonite during diagenesis. Although the marine water was also a possible supplier of the alkali elements, authigenic I/S is rare in other coals of that seam that are not immediately adjacent to such altered volcanic layers.

The minerals in the two thin intra-seam claystones, which are referred to as tonsteins, are essentially well-ordered kaolinite, regularly interstratified I/S, and illite, with pyrite, quartz, albite and anatase present in mior proportions. The relatively thick claystone, which contains 50% I/S in the clay mineral assemblage, is referred to as a K-bentonite. The volcanic ash layers in the peat swamp may have been originally converted to smectite under the marine-influenced coal-forming conditions. The smectite was in turn altered to I/S and illite during diagenesis and/or rank advance, assuming that the necessary ions (e.g. K, Na and Mg) were available from the marine water percolating through the peat bed. Na-rich I/S may also have been formed in the claystones, with the additional Na probably being released from the organic matter during the coal’s rank advance.

10.2 Geochemistry of the coal seams

The concentrations of most trace elements in the relatively low-ash Greta, Great Northern, and Bulli coals are lower than the average values for the corresponding elements in worldwide coals as given by Ketris and Yudovich (2009). By contrast, the high-ash Songzao coals have relatively high concentrations of most trace elements relative to the average published values for worldwide coal seams.

Despite the relatively high pyrite contents, most of the chalcophile trace elements in the sulphur-rich Songzao and Greta coals show either poor or negative correlations with total iron sulphide percentages. In the Songzao coals, only As and Mo show positive correlations with iron sulphides; only Hg, Tl and As are positively correlated with pyrite in the Greta coals. This may be because the pyrite in the Songzao and Greta coals is mostly of syngenetic origin.

298

Chapter 10 Conclusions

Instead of being correlated with iron sulphides, some of the chalcophile elements are correlated with Al2O3, probably indicating a common source rather than association with iron sulphides. The absence of traditional pyrite-metal associations may reflect wide variations in the concentrations of these elements in individual pyrite/marcasites, or simply poor retention of those elements in the pyrite/marcasite of the relevant coal beds. In addition to the lithophile elements, the chalchophile elements in the Great Northern coals, including Se, Pb and Cu, also appear to be associated with kaolinite. Most of the trace elements in the Bulli coal samples, however, do not show a significant correlation with the abundance of any mineral. This may partly be due to the small number of coal samples, and the relatively low concentrations of the elements in question.

In the Songzao coals, Cu, V and Cr show significant positive correlations with the sum of illite and I/S, when the elements and the minerals are both expressed on a whole coal basis. Similar patterns of variation also appear to exist for Cu, V and Cr in the different coal samples. These correlations may indicate a common source of clastic material supplied to the coal basin, which was probably derived from the mafic basaltic rocks of the Kangdian Oldland.

Since the Songzao coals are mainly used as fuel for thermal power plants, trace elements that may affect environment and human health are of great concern. The poor correlation of Hg and Se with iron sulphides in the Songzao coals may indicate multiple modes of occurrence of these elements. Organically associated Hg and Se can hardly be removed by cleaning processes in coal preparation. Arsenic and Mo are mainly associated with iron sulphides in the Songzao coals. However, the pyrite/marcasite in the Songzao coals is mainly syngenetic rather than epigenetic, in the form of cell infillings, framboids, and euhedral crystals, indicating an intimate association with the organic matter. Arsenic and Mo thus cannot be readily and effectively removed in coal preparation processes.

10.3 Significance of intra-seam volcanic claystones

Elements liberated during the alteration from the original volcanic ash layers to the claystones have been either largely removed from the system or partly fixed in the adjacent coals. This has also affected the geochemistry and mineralogy of the coal bed. Additionally, Intra-seam claystones may be incorporated with mined coal products, and if not removed in the preparation plant, become part of the coal when it is used.

299

The coals near the alkali tonstein bands in the Tonghua and Yuyang sections have high concentrations of Nb, Ta, Hf, Ga, Th, U, and REE. Coal plies overlying the mafic bentonite in the Tonghua section are high in TiO2, V, Cr, Zn and Cu. This may be because of either the incorporation of the volcanic ash material in the adjacent peat, or the liberation of these elements from the ash layers. Leaching of the original alkali and mafic ash may have led to higher concentrations of relevant elements in the adjacent coals than in coals without such influence. The tonsteins of the Great Northern coal seam were derived from acid to intermediate volcanic ashes. Trace elements, such as Li, Th, and U, are relatively high in most of the coal plies adjacent the acid tonsteins, relative to those away from the tonstein layers. However, the influence of the tonstein materials on the trace element geochemistry of the adjacent coals is not as significant as in the Songzao coal seams.

The occurrence of alkali tonsteins may be of significance and an indicator in finding rare mental ore deposit in coal or associated strata (Seredin and Dai, 2012). Dai et al. (2010c) recently described Nb-, Zr-, REE-, and Ga- bearing tuffecous horizons in the Late Permian coal-bearing strata of Yunnan Province, SW China, with thickness up to 10 m that were derived from alkali pyroclasitic rocks. The concentrations of Nb2O5, ZrO2, REE oxides and Ga in those ore beds are up to 0.06%, 0.85%, 0.14% and 0.08%, respectively (Dai et al., 2010c). As concluded by Seredin and Dai (2012), the recovery of REE and other rare metals from coal-bearing deposits is promsing for industrial utilisation.

300

REFERENCES

Addison, R., Harrison, R.K., Land, D.H., Young, B.R., Davis, A.E. and Smith, T.K., 1983. Volcanogenic tonsteins from Tertiary coal measures, East Kalimantan, Indonesia. International Journal of Coal Geology 3, 1-30.

Admakin, L.A., 2002. Accumulation and Post-Sedimentary Diagenesis of Tonsteins. Lithology and Mineral Resources 37, 60-67.

Agnew, D., Bocking, M., Brown, K., Ives, M., Johnson, D., Howes, M., Preston, B., Rigby, R., Warbrooke, P. and Weber, C.R., 1995. Sydney Basin, Newcastle Coalfield. In: C.R. Ward, H.J. Harrington, C.W. Mallett and J.W. Beeston (Eds.), Geology of Australian Coal Basins. Geological Society of Australia Coal Geology Group Special Publication, pp. 197-212.

Alder, J., Hawley, S., Mullard, B. and Shaw, R., 1998. Origin of the Sydney Basin: A new structural model, Advances in the Study of the Sydney Basin, 32th Newcastle Symposium proceedings, Newcastle, pp. 1-11.

Altaner, S.P., Hower, J., Whitney, G. and Aronson, J.L., 1984. Model for K-bentonite formation: Evidence from zoned K-bentonites in the disturbed belt, Montana. Geology 12, 412-415.

Armstrong, M., Bamberry, W.J., Hutton, A.C. and Jones, B.G., 1995. Sydney Basin, Southern Coalfield. In: C.R. Ward, H.J. Harrington, C.W. Mallett and J.W. Beeston (Eds.), Geology of Australian Coal Basins. Geological Society of Australia Coal Geology Group Special Publication, pp. 213-230.

ASTM, 2007. Annual Book of ASTM Standards. Gaseous Fuels; Coal and Coke, vol. 05. 06, Classification of Coals by Rank. American Society for Testing and Materials, West Conshohocken, PA, pp. 244-250.

Baker, J.C., Bai, G.P., Hamilton, P.J., Golding, S.D. and Keene, J.B., 1995. Continental-scale magmatic carbon dioxide seepage recorded by dawsonite in the Bowen-Gunnedah-Sydney Basin system, eastern Australia. Journal of Sedimentary Research 65, 522-530.

Banerjee, I. and Goodarzi, F., 1990. Paleoenvironment and sulfur-boron contents of the Mannville (Lower Cretaceous) coals of southern Alberta, Canada. Sedimentary Geology 67, 297-310.

Batchelor, R.A. and Jeppsson, L., 1994. Late Llandovery bentonites from Gotland, Sweden, as chemostratigraphic markers. Journal - Geological Society (London) 151, 741-746.

Bayan, M.R. and Hower, J.C., Illite crystallinity and coal metamorphism for selected central Appalachian coals and shales. International Journal of Coal Geology.

Beaton, A.P., Goodarzi, F. and Potter, J., 1991. The petrography, mineralogy and geochemistry of a paleocene lignite from southern Saskatchewan, Canada. International Journal of Coal Geology 17, 117-148.

Bembrick, C.S., Herbert, C., Scheibner, E. and Stuntz, J., 1973. Structural subdivision of the New South Wales portion of the Sydney-Bowen Basin.

301

Benning, L.G., Wilkin, R.T. and Barnes, H.L., 2000. Reaction pathways in the Fe–S system below 100°C. Chemical Geology 167, 25-51.

Benson, S.A. and Holm, P.L., 1985. Comparison of inorganic constituents in three low-rank coals. Industrial and Engineering Chemistry Product Research and Development 24, 145-149.

Besly, B.M. and Fielding, C.R., 1989. Palaeosols in Westphalian coal-bearing and red-bed sequences, central and northern England. Palaeogeography, Palaeoclimatology, Palaeoecology 70, 303-330.

Bieg, G. and Burger, K., 1992. Preliminary study of tonsteins of the Pastora formation (stephanian B) of the Ciñera-Matallana coalfield, northwestern Spain. International Journal of Coal Geology 21, 139-160.

Bocking, M., Howes, M. and Weber, C.R., 1988. Palaeochannel development in the Moon Island Beach Sub-Group of the Newcastle Coal Measures., Proceedings of 22nd Symposium on Advances in the Study of the Sydney Basin, Department of Geology, University of Newcastle, NSW, pp. 37-45.

Bohor, B.F. and Gluskoter, H.J., 1973. Boron in illite as an indicator of paleosalinity of Illinois coals. Journal of Sedimentary Petrology 43, 945-956.

Bohor, B.F. and Triplehorn, D.M., 1993. Tonsteins: altered volcanic-ash layers in coal-bearing sequences. Geological Society of America, Special paper 285.

Botz, R.W. and Hart, G.H., 1983. Mineralogical, petrographical and geochemical investigations of outbursting in Australian coal mines. Proceedings of the Australian Institute of Mining and Metallurgy 286, 41-49.

Bouška, V., 1981. Geochemistry of Coal. Elsevier, Amsterdam, 284 pp.

Bouška, V., Pešek, J. and Sýkorová, I., 2000. Probable modes of occurrence of chemical elements in coal. Acta Montana. Serie B, Fuel, Carbon, Mineral Processing, Praha 10 (117), 53-90.

Bowman, H.N., 1980. Southern Coalfield, Upper Shoalhaven Group and Illawarra Coal Measures. In: C. Herbert and R.E. Helby (Eds.), A guide to the Sydney Basin. Geological Survey of New South Wales, Bulletin 26, pp. 117-132.

Boyd, R.J., 2002. The partitioning behaviour of boron from tourmaline during ashing of coal. International Journal of Coal Geology 53, 43-54.

Branagan, D.F. and Johnson, M.W., 1970. Permian sedimentation in the Newcastle Coalfield. Proceedings of the Australian Institute of Mining and Metallurgy 135, 1-36.

Brockway, D.J. and Borsaru, R.M., 1985. Ion concentration profiles in Victorian brown coals, International Conference on Coal Science. Pergamon Press, Sydney, Australia, pp. 593-596.

Brown, G. and Weir, A.H., 1963. The identity of rectorite and allevardite, the Intenational Clay Conference, Stockholm, 1, 27-35 and 2, 87-90.

302

References

Brown, H.R. and Swaine, D.J., 1964. Inorganic constituents of Australian coals--Part l: nature and mode of occurrence. Journal of the Institute of Fuel 37, 422-440.

Brownfield, M.E., Affolter, R.H., Stricker, G.D. and Hildebrand, R.T., 1995. High chromium contents in Tertiary coal deposits of northwestern Washington — A key to their depositional history. International Journal of Coal Geology 27, 153-169.

Brownfield, M.E., Affolter, R.H., Cathcart, J.D., Johnson, S.Y., Brownfield, I.K. and Rice, C.A., 2005. Geologic setting and characterization of coals and the modes of occurrence of selected elements from the Franklin coal zone, Puget Group, John Henry No. 1 mine, King County, Washington, USA. International Journal of Coal Geology 63, 247-275.

Brownlow, J.W., 1981. A thermal model for the origin of the Sydney and Gunnedah Basins, Proceedings of the 15st Newcastle Symposium, "Advances in the study of the Sydney Basin". Department of Geology, the University of Newcastle, New South Wales, pp. 10-12.

Burger, K., Zhou, Y. and Tang, D., 1990. Synsedimentary volcanic-ash-derived illite tonsteins in Late Permian coal-bearing formations of southwestern China. International Journal of Coal Geology 15, 341-356.

Burger, K., Bandelow, F.K. and Bieg, G., 2000. Pyroclastic kaolin coal-tonsteins of the Upper Carboniferous of Zonguldak and Amasra, Turkey. International Journal of Coal Geology 45, 39-53.

Burger, K., Zhou, Y. and Ren, Y., 2002. Petrography and geochemistry of tonsteins from the 4th Member of the Upper Triassic Xujiahe formation in southern Sichuan Province, China. International Journal of Coal Geology 49, 1-17.

Caballero, E., Jiménez De Cisneros, C., Huertas, F.J., Huertas, F., Pozzuoli, A. and Linares, J., 2005. Bentonites from Cabo de Gata, Almería, Spain: a mineralogical and geochemical overview. Clay Minerals 40, 463-480.

Carroll, D., 1970. Clay minerals: A Guide to Their X-ray Identification. The Geology Society of America, Boulder, Colorado, 75 pp.

Caswell, S.A., Holmes, L.F. and Spears, D.A., 1984a. Total chlorine in coal seam profiles from the South Staffordshire (Cannock) coalfield. Fuel 63, 782-787.

Caswell, S.A., Holmes, L.F. and Spears, D.A., 1984b. Water-soluble chlorine and associated major cations from the coal and mudrocks of the Cannock and North Staffordshire coalfields. Fuel 63, 774-781.

Cecil, C.B., Stanton, R.W., Dulong, F.T. and Renton, J.J., 1982. Geological factors that control mineral matter in coal. In: R.H. Filby, B.S. Carpenter and R.C. Ragaini (Eds.), Atomic and Nuclear Methods in Fossil Energy Research. Plenum, New York, pp. 323-335.

Chagué-Goff, C., Goodarzi, F. and Fyfe, W.S., 1996. Elemental distribution and pyrite occurrence in a freshwater Peatland, Alberta. Journal of Geology 104, 649-663.

Chou, C.L., 1984. Relationship between geochemistry of coal and nature of strata overlying Herrin coal in the Illinois Basin, United States. Geological Society of China, Mem. 6, 269-280.

303

Chou, C.L., 1991. Distribution and forms of chlorine in Illinois basin coals. In: J. Stringer and D.D. Banerjee (Eds.), Chlorine in Coal. Coal Science and Technology, 17. Elsevier, Amsterdam.

Chung, S.-L. and Jahn, B.-m., 1995. Plume-lithosphere interaction in generation of the Emeishan flood basalts at the Permian-Triassic boundary. Geology 23, 889-892.

Coleman, L., Bragg, L. and Finkelman, R., 1993. Distribution and mode of occurrence of selenium in US coals. Environmental Geochemistry and Health 15, 215-227.

Coombs, D.S., 1961. Some recent work on the lower grades of metamorphism. Australian Journal of Science 24, 203-215.

Cooper, J.R., Crelling, J.C., Rimmer, S.M. and Whittington, A.G., 2007. Coal metamorphism by igneous intrusion in the Raton Basin, CO and NM: Implications for generation of volatiles. International Journal of Coal Geology 71, 15-27.

Craw, D., 1997. Fluid inclusion evidence for geothermal structure beneath the Southern Alps, New Zealand. New Zealand Journal of Geology and Geophysics 40, 43-52.

Creech, M., 1998. So where are all the trees? Implication for the origin of coal, Proceedings of the 32nd Newcastle Symposium on Advances in the Study of the Sydney Basin, University of Newcastle, New South Wales, pp. 29-43.

Creech, M., 2002. Tuffaceous deposition in the Newcastle Coal Measures: challenging existing concepts of peat formation in the Sydney Basin, New South Wales, Australia. International Journal of Coal Geology 51, 185-214.

Cressey, B.A. and Cressey, G., 1988. Preliminary mineralogical investigation of Leicestershire low- rank coal. International Journal of Coal Geology 10, 177-191.

Crowley, S.S., Stanton, R.W. and Ryer, T.A., 1989. The effects of volcanic ash on the maceral and chemical composition of the C coal bed, Emery Coal Field, Utah. Organic Geochemistry 14, 315-331.

Crowley, S.S., Ruppert, L.F., Belkin, H.E., Stanton, R.W. and Moore, T.A., 1993. Factors affecting the geochemistry of a thick, subbituminous coal bed in the Powder River Basin: volcanic, detrital, and peat-forming processes. Organic Geochemistry 20, 843-853.

Dai, S., Ren, D. and Ma, S., 2004. The cause of endemic fluorosis in western Guizhou Province, Southwest China. Fuel 83, 2095-2098.

Dai, S., Chou, C.-L., Yue, M., Luo, K. and Ren, D., 2005a. Mineralogy and geochemistry of a Late Permian coal in the Dafang Coalfield, Guizhou, China: influence from siliceous and iron-rich calcic hydrothermal fluids. International Journal of Coal Geology 61, 241-258.

Dai, S., Ren, D., Tang, Y., Yue, M. and Hao, L., 2005b. Concentration and distribution of elements in Late Permian coals from western Guizhou Province, China. International Journal of Coal Geology 61, 119-137.

304

References

Dai, S., Ren, D., Chou, C.-L., Li, S. and Jiang, Y., 2006a. Mineralogy and geochemistry of the No. 6 Coal (Pennsylvanian) in the Junger Coalfield, Ordos Basin, China. International Journal of Coal Geology 66, 253-270.

Dai, S., Zeng, R. and Sun, Y., 2006b. Enrichment of arsenic, antimony, mercury, and thallium in a Late Permian anthracite from Xingren, Guizhou, Southwest China. International Journal of Coal Geology 66, 217-226.

Dai, S. and Chou, C.L., 2007. Occurrence and origin of minerals in a chamosite-bearing coal of Late Permian age, Zhaotong, Yunnan, China. American Mineralogist 92, 1253-1261.

Dai, S., Li, W., Tang, Y., Zhang, Y. and Feng, P., 2007a. The sources, pathway, and preventive measures for fluorosis in Zhijin County, Guizhou, China. Applied Geochemistry 22, 1017-1024.

Dai, S. and Ren, D., 2007. Effects of magmatic intrusion on mineralogy and geochemistry of coals from the Fengfeng-Handan coalfield, Hebei, China. Energy and Fuels 21, 1663-1673.

Dai, S., Li, D., Chou, C.-L., Zhao, L., Zhang, Y., Ren, D., Ma, Y. and Sun, Y., 2008a. Mineralogy and geochemistry of boehmite-rich coals: New insights from the Haerwusu Surface Mine, Jungar Coalfield, Inner Mongolia, China. International Journal of Coal Geology 74, 185-202.

Dai, S., Ren, D., Zhou, Y., Chou, C.-L., Wang, X., Zhao, L. and Zhu, X., 2008b. Mineralogy and geochemistry of a superhigh-organic-sulfur coal, Yanshan Coalfield, Yunnan, China: Evidence for a volcanic ash component and influence by submarine exhalation. Chemical Geology 255, 182-194.

Dai, S., Tian, L., Chou, C.-L., Zhou, Y., Zhang, M., Zhao, L., Wang, J., Yang, Z., Cao, H. and Ren, D., 2008c. Mineralogical and compositional characteristics of Late Permian coals from an area of high lung cancer rate in Xuan Wei, Yunnan, China: Occurrence and origin of quartz and chamosite. International Journal of Coal Geology 76, 318-327.

Dai, S., Wang, X., Chen, W., Li, D., Chou, C.-L., Zhou, Y., Zhu, C., Li, H., Zhu, X., Xing, Y., Zhang, W. and Zou, J., 2010a. A high-pyrite semianthracite of Late Permian age in the Songzao Coalfield, southwestern China: Mineralogical and geochemical relations with underlying mafic tuffs. International Journal of Coal Geology 83, 430-445.

Dai, S., Zhao, L., Peng, S., Chou, C.-L., Wang, X., Zhang, Y., Li, D. and Sun, Y., 2010b. Abundances and distribution of minerals and elements in high-alumina coal fly ash from the Jungar Power Plant, Inner Mongolia, China. International Journal of Coal Geology 81, 320-332.

Dai, S., Zhou, Y., Zhang, M., Wang, X., Wang, J., Song, X., Jiang, Y., Luo, Y., Song, Z., Yang, Z. and Ren, D., 2010c. A new type of Nb (Ta)–Zr(Hf)–REE–Ga polymetallic deposit in the late Permian coal-bearing strata, eastern Yunnan, southwestern China: Possible economic significance and genetic implications. International Journal of Coal Geology 83, 55-63.

Dai, S., Wang, X., Zhou, Y., Hower, J.C., Li, D., Chen, W., Zhu, X. and Zou, J., 2011. Chemical and mineralogical compositions of silicic, mafic, and alkali tonsteins in the late Permian coals from the Songzao Coalfield, Chongqing, Southwest China. Chemical Geology 282, 29-44.

Dai, S., Jiang, Y., Ward, C.R., Gu, L., Seredin, V.V., Liu, H., Zhou, D., Wang, X., Sun, Y., Zou, J. and Ren, D., 2012a. Mineralogical and geochemical compositions of the coal in the

305

Guanbanwusu Mine, Inner Mongolia, China: Further evidence for the existence of an Al (Ga and REE) ore deposit in the Jungar Coalfield. International Journal of Coal Geology 98, 10-40.

Dai, S., Ren, D., Chou, C.-L., Finkelman, R.B., Seredin, V.V. and Zhou, Y., 2012b. Geochemistry of trace elements in Chinese coals: A review of abundances, genetic types, impacts on human health, and industrial utilization. International Journal of Coal Geology 94, 3-21.

Dai, S., Wang, X., Seredin, V.V., Hower, J.C., Ward, C.R., O'Keefe, J.M.K., Huang, W., Li, T., Li, X., Liu, H., Xue, W. and Zhao, L., 2012c. Petrology, mineralogy, and geochemistry of the Ge-rich coal from the Wulantuga Ge ore deposit, Inner Mongolia, China: New data and genetic implications. International Journal of Coal Geology 90–91, 72-99.

Dai, S., Zou, J., Jiang, Y., Ward, C.R., Wang, X., Li, T., Xue, W., Liu, S., Tian, H., Sun, X. and Zhou, D., 2012d. Mineralogical and geochemical compositions of the Pennsylvanian coal in the Adaohai Mine, Daqingshan Coalfield, Inner Mongolia, China: Modes of occurrence and origin of diaspore, gorceixite, and ammonian illite. International Journal of Coal Geology 94, 250-270.

Dai, S.F., Ren, D.Y. and Tang, Y., 2005c. Modes of occurrence of major elements in coal and their study signif icance. Coal Geology and Exploration 33, 1-4 (In Chinese with English abstract).

Dai, S.F., Zhou, Y.P., Ren, D.Y., Wang, X.B., Li, D. and Zhao, L., 2007b. Geochemistry and mineralogy of the Late Permian coals from the Songzao Coalfield, Chongqing, southwestern China. Science in China, Series D: Earth Sciences 50, 678-688.

Damberger, H.H., Harver, R.D., Ruch, R.R. and Thomas, J.J., 1984. Coal Characterization. In: B.R. Cooper and W.A. Ellingson (Eds.), The Science and Technology of Coal and Coal Utilization. Plenum Press, New York, pp. 7-45.

Daniels, E.J. and Altaner, S.P., 1990. Clay mineral authigenesis in coal and shale from the Anthracite region, Pennsylvania. American Mineralogist 75, 825-839.

Daniels, E.J. and Altaner, S.P., 1993. Inorganic nitrogen in anthracite from eastern Pennsylvania, USA. International Journal of Coal Geology 22, 21-35.

Davis, A., Russell, S.J., Rimmer, S.M. and Yeakel, J.D., 1984. Some genetic implications of silica and aluminosilicates in peat and coal. International Journal of Coal Geology 3, 293-314.

Deer, W.A., Howie, R.A. and Zussman, J., 1992. An introduction to the rock-forming minerals. Longman Scientific and Technical, London, 696 pp.

Delano, J.W., Tice, S.J., Mitchell, C.E. and Goldman, D., 1994. Rhyolitic glass in Ordovician K- bentonites: a new stratigraphic tool. Geology 22, 115-118.

Deul, M., 1955. The origin of ash-forming ingredients in coal. Economic Geology 50, 103-104.

Deul, M., 1958. Biochemical and geochemical origins of ash-forming ingredients in coal. Journal Name: Am. Chem. Soc., Div. Gas Fuel Chem., Prepr.; (United States), Medium: X; Size: Pages: vp.

Dewison, M.G., 1989. Dispersed kaolinite in the Barnsley Seam coal (U.K.): evidence for a volcanic origin. International Journal of Coal Geology 11, 291-304.

306

References

Diehl, S.F., Goldhaber, M.B. and Hatch, J.R., 2004. Modes of occurrence of mercury and other trace elements in coals from the warrior field, Black Warrior Basin, Northwestern Alabama. International Journal of Coal Geology 59, 193-208.

Diehl, S.F., Goldhaber, M.B., Koenig, A.E., Lowers, H.A. and Ruppert, L.F., 2012. Distribution of arsenic, selenium, and other trace elements in high pyrite Appalachian coals: Evidence for multiple episodes of pyrite formation. International Journal of Coal Geology 94, 238-249.

Diessel, C.F.K., 1965. On the petrography of some Australian tonsteins. In: P. Schmidt-Thomé and R. Schönenberg (Eds.), Max Richter-Festschrift. University of Clausthal-Zellerfeld, pp. 149-166.

Diessel, C.F.K., 1983. Tuffs and tonsteins in the coal measures of New South Wales, Australia, Proceedings of the 10th International Congress on Carboniferous Stratigraphy and Geology, Madrid, Madrid, pp. 197-210.

Diessel, C.F.K., 1992. Coal-Bearing Depositional Systems. Springer-Verlag, Berlin, 721 pp.

Diessel, C.F.K. and Gammidge, L., 1998. Isometamorphic variations in the reflectance and fluorescence of vitrinite—a key to depositional environment. International Journal of Coal Geology 36, 167-222.

Diessel, C.F.K. and Hutton, A.C., 2004. Field trip B- An overview of the Newcalstle Coal Measures. 21st Annual Meeting of the Society for Organic Petrology. , Sydney.

Ding, Z., Zheng, B., Long, J., Belkin, H.E., Finkelman, R.B., Chen, C., Zhou, D. and Zhou, Y., 2001. Geological and geochemical characteristics of high arsenic coals from endemic arsenosis areas in southwestern Guizhou Province, China. Applied Geochemistry 16, 1353-1360.

Dominik, J. and Stanley, D.J., 1993. Boron, beryllium and sulfur in Holocene sediments and peats of the Nile delta, Egypt: Their use as indicators of salinity and climate. Chemical Geology 104, 203-216.

Drakonaki, S., Diamadopoulos, E., Vamvouka, D. and Lahaniatis, M., 1998. Leaching behavior of lignite fly ash. Journal of Environmental Science and Health - Part A Toxic/Hazardous Substances and Environmental Engineering 33, 237-248.

Dreher, G.B. and Finkelman, R.B., 1992. Selenium mobilization in a surface coal mine, Powder River Basin, Wyoming, U.S.A. Environmental Geology and Water Sciences 19, 155-167.

Drits, V.A., 2003. Structural and chemical heterogeneity of layer silicates and clay minerals. Clay Minerals 38, 403-432.

Du, G., Zhuang, X., Querol, X., Izquierdo, M., Alastuey, A., Moreno, T. and Font, O., 2009. Ge distribution in the Wulantuga high-germanium coal deposit in the Shengli coalfield, Inner Mongolia, northeastern China. International Journal of Coal Geology 78, 16-26.

Duan, L., Meng, Q.-R., Zhang, C.-L. and Liu, X.-M., 2011. Tracing the position of the South China block in Gondwana: U–Pb ages and Hf isotopes of Devonian detrital zircons. Gondwana Research 19, 141-149.

307

Dutta, B.K., Khanra, S. and Mallick, D., 2009. Leaching of elements from coal fly ash: Assessment of its potential for use in filling abandoned coal mines. Fuel 88, 1314-1323.

Eskenazy, G., Delibaltova, D. and Mincheva, E., 1994. Geochemistry of boron in Bulgarian coals. International Journal of Coal Geology 25, 93-110.

Eskenazy, G.M., 1995. Geochemistry of arsenic and antimony in Bulgarian coals. Chemical Geology 119, 239-254.

Eskenazy, G.M., 1999. Aspects of the geochemistry of rare earth elements in coal: an experimental approach. International Journal of Coal Geology 38, 285-295.

Eskenazy, G.M. and Valceva, S.P., 2003. Geochemistry of beryllium in the Mariza-east lignite deposit (Bulgaria). International Journal of Coal Geology 55, 47-58.

Eskenazy, G.M., 2006. Geochemistry of beryllium in Bulgarian coals. International Journal of Coal Geology 66, 305-315.

Eskenazy, G.M. and Stefanova, Y.S., 2007. Trace elements in the Goze Delchev coal deposit, Bulgaria. International Journal of Coal Geology 72, 257-267.

Eskenazy, G.M., 2009. Trace elements geochemistry of the Dobrudza coal basin, Bulgaria. International Journal of Coal Geology 78, 192-200.

Eugster, H.P. and Munoz, J., 1966. Ammonium Micas: Possible Sources of Atmospheric Ammonia and Nitrogen. Science 151, 683-686.

Faraj, B.S.M. and Mackinnon, I.D.R., 1993. Micrinite in southern hemisphere sub-bituminous and bituminous coals: redefined as fine grained kaolinite. Organic Geochemistry 20, 823-841.

Faraj, B.S.M., Fielding, C.R. and Mackinnon, D.R., 1996. Cleat Mineralisation of Upper Permian Baralaba/Rangal Coal Measures, Bowen Basin, Australia. In: R. Gayer and I. Harris (Eds.), Coalbed methane and Coal Geology. Geological society, London, Special Publication 109, pp. 151-164.

Fellowes, P.M., 1979. An investigation of the water-soluble elements in coal and coal-bearing strata. PhD Thesis, University of Sheffield.

Feng, X. and Hong, Y., 1999. Modes of occurrence of mercury in coals from Guizhou, People's Republic of China. Fuel 78, 1181-1188.

Finkelman, R.B. and Stanton, R.W., 1978. Identification and significance of accessory minerals from a bituminous coal. Fuel 57, 763-768.

Finkelman, R.B., 1980. Mode of occurrence of trace elements in coal. Ph.D Thesis, University of Maryland, College Park.

Finkelman, R.B., 1981. Recognition of authigenic and detrital minerals in coal: Geological Society of America, Abstracts with Program 13, 451.

308

References

Finkelman, R.B., 1994. Modes of occurrence of potentially hazardous elements in coal: levels of confidence. Fuel Processing Technology 39, 21-34.

Finkelman, R.B., 1995. Modes of occurrence of environmentally-sensitive trace elements in coal. In: D.J. Swaine and F. Goodarzi (Eds.), Environmental Aspect of Trace Elements in Coal. Kluwer Academic Publishers, Amsterdam.

Finkelman, R.B., Bostick, N.H., Dulong, F.T., Senftle, F.E. and Thorpe, A.N., 1998. Influence of an igneous intrusion on the inorganic geochemistry of a bituminous coal from Pitkin County, Colorado. International Journal of Coal Geology 36, 223-241.

Finkelman, R.B., Belkin, H.E. and Zheng, B., 1999. Health impacts of domestic coal use in China. Proceedings of the National Academy of Sciences 96, 3427-3431.

Folk, R.L., 1974. Petrology of Sedimentary rocks. Hemphill Austin, 182 pp.

Francis, W., 1961. Coal: its formation and composition. Edward Arnold, London, 806 pp.

Frazer, F.W. and Belcher, C.B., 1973. Quantitative determination of the mineral-matter content of coal by a radiofrequency -oxidation technique. Fuel 52, 41-46.

Fu, F.F., Akagi, T., Suzuki, Y., Watanabe, K. and Yabuki, S., 2004. Rare earth element distribution in the acetic acid soluble fraction of combusted coals: Its implication as a proxy for the original coal-forming plants. Geochemical Journal 38, 333-343.

Fynes, G., Herod, A.A., Hodges, N.J., Stokes, B.J. and Ladner, W.R., 1988. The thermal emission of chlorine-containing compounds from coal following brine and chlorine gas treatments. Fuel 67, 822-830.

Galushkin, Y.I., 1997. Thermal effects of igneous intrusions on maturity of organic matter: A possible mechanism of intrusion. Organic Geochemistry 26, 645-658.

Gayer, R.A., Rose, M., Dehmer, J. and Shao, L.Y., 1999. Impact of sulphur and trace element geochemistry on the utilization of a marine-influenced coal—case study from the South Wales Variscan foreland basin. International Journal of Coal Geology 40, 151-174.

GB/T 482-2008 (National Standard of P.R. China), 2008. Sampling of coal in seam GB 482-1995, pp. 6 (in Chinese).

Gibbs, R.J., 1971. Preparation of X-ray diffraction mounts. In: R.E. Carver (Ed.), Procedures in Sedimentary Petrology. Wiley, New York, pp. 531-540.

Glen, R.A. and Beckett, J., 1989. Geological note thinϋskinned tectonics in the hunter coalfield of New South Wales. Australian Journal of Earth Sciences 36, 589-593.

Glick, D.C. and Davis, A., 1987. Variability in the inorganic element content of U.S. coals including results of cluster analysis. Organic Geochemistry 11, 331-342.

Gluskoter, H.J., 1975. Mineral Matter and Trace Elements in Coal. In: S.P. Babu (Ed.), Trace Elements in Fuel. Advances in Chemistry. AMERICAN CHEMICAL SOCIETY, pp. 1-22.

309

Godbeer, W.C. and Swaine, D.J., 1987. Fluorine in Australian coals. Fuel 66, 794-798.

Golab, A.N. and Carr, P.F., 2004. Changes in geochemistry and mineralogy of thermally altered coal, Upper Hunter Valley, Australia. International Journal of Coal Geology 57, 197-210.

Golab, A.N., Carr, P.F. and Palamara, D.R., 2006. Influence of localised igneous activity on cleat dawsonite formation in Late Permian coal measures, Upper Hunter Valley, Australia. International Journal of Coal Geology 66, 296-304.

Golab, A.N., Hutton, A.C. and French, D., 2007. Petrography, carbonate mineralogy and geochemistry of thermally altered coal in Permian coal measures, Hunter Valley, Australia. International Journal of Coal Geology 70, 150-165.

Goldbery, R. and Loughnan, F.C., 1970. Dawsonite and nordstrandite in Permian Berry Formation of Sydney Basin, New South Wales. American Mineralogist 55, 477-490.

Goldbery, R. and Loughnan, F.C., 1977. Dawsonite, alumohydrocalcite, nordstrandite and gorceixite in Permian marine strata of the Sydney Basin, Australia. Sedimentology 24, 565-579.

Goodarzi, F., 1987a. Concentration of elements in lacustrine coals from zone A hat creek deposit no. 1, British Columbia, Canada. International Journal of Coal Geology 8, 247-268.

Goodarzi, F., 1987b. Comparison of elemental distribution in fresh and weathered samples of selected coals in the Jurassic-Cretaceous Kootenay Group, British Columbia, Canada. Chemical Geology 63, 21-28.

Goodarzi, F. and Van Der Flier-Keller, E., 1988. Distribution of major, minor and trace elements in Hat Creek Deposit No. 2, British Columbia, Canada. Chemical Geology 70, 313-333.

Goodarzi, F. and Cameron, A.R., 1990. Organic petrology and elemental distribution in thermally altered coals from Telkwa, British Columbia. Energy Sources 12, 315-343.

Goodarzi, F. and Swaine, D.J., 1994. The influence of geological factors on the concentration of boron in Australian and Canadian coals. Chemical Geology 118, 301-318.

Goodarzi, F., 1995. Geology of Trace elements in Coal. In: D.J. Swaine and F. Goodarzi (Eds.), Environmental Aspects of Trace elements in Coal. Kluwer Academic Publishers, Amsterdam, pp. 51-75.

Goodarzi, F., 2002. Mineralogy, elemental composition and modes of occurrence of elements in Canadian feed-coals. Fuel 81, 1199-1213.

Goodarzi, F. and Huggins, F.E., 2005. Speciation of chromium in feed coals and ash byproducts from Canadian power plants burning subbituminous and bituminous coals. Energy and Fuels 19, 2500-2508.

Greb, S.F., Eble, C.F. and Hower, J.C., 1999. Depositional history of the Fire Clay coal bed (Late Duckmantian), Eastern Kentucky, USA. International Journal of Coal Geology 40, 255-280.

Grevenitz, P., 2003. Origin, alteration and geochemical correlation of Late Permian airfall tuffs in coal measures, Sydney Basin, Australia. International Journal of Coal Geology 55, 27-46.

310

References

Grieve, D.A. and Goodarzi, F., 1993. Trace elements in coal samples from active mines in the Foreland Belt, British Columbia, Canada. International Journal of Coal Geology 24, 259-280.

Griffin, G.M., 1971. Interpretation of X-ray diffraction data. In: R.E. Carver (Ed.), Procedures in Sedimentary Petrology. Wiley, New York, pp. 541-569.

Grim, R.E. and Güven, N., 1978. Bentonites : geology, mineralogy, properties and uses Developments in sedimentology. Elsevier Scientific Publishing Co., Amsterdam, 256 pp.

Guerra-Sommer, M., Cazzulo-Klepzig, M., Santos, J.O.S., Hartmann, L.A., Ketzer, J.M. and Formoso, M.L.L., 2008. Radiometric age determination of tonsteins and stratigraphic constraints for the Lower Permian coal succession in southern Paraná Basin, Brazil. International Journal of Coal Geology 74, 13-27.

Guo, W., Hu, S., Zhang, J. and Zhang, H., 2011. Elimination of oxide interferences and determination of ultra-trace silver in soils by ICP-MS with ion–molecule reactions. Science of The Total Environment 409, 2981-2986.

Gupta, R., Wall, T.F. and Baxter, L.A. (Eds.), 1999. The impact of mineral impurities in solid fuel combustion. Plenum, New York, 768 pp.

Gurba, L.W. and Ward, C.R., 1998. Vitrinite reflectance anomalies in the high-volatile bituminous coals of the Gunnedah Basin, New South Wales, Australia. International Journal of Coal Geology 36, 111-140.

Hagen, E., Kelley, S.P., Dypvik, H., Nilsen, O. and Kjolhamar, B., 2001. Direct dating of authigenic K-feldspar overgrowths from the Kilombero Rift of Tanzania. Journal of the Geological Society 158, 801-807.

Hatch, J.R., Gluskoter, H.J. and Lindahl, P.C., 1976. Sphalerite in coals from the Illinois Basin. Economic Geology 71, 613-624.

Haynes, J.T., Melson, W.G. and Kunk, M.J., 1995. Composition of biotite phenocrysts in Ordovician tephras casts doubt on the proposed trans-Atlantic correlation of the Millbrig K- bentonite (United States) and the Kinnekulle K-bentonite (Sweden). Geology 23, 847-850.

He, B., Xu, Y.-G., Chung, S.-L., Xiao, L. and Wang, Y., 2003. Sedimentary evidence for a rapid, kilometer-scale crustal doming prior to the eruption of the Emeishan flood basalts. Earth and Planetary Science Letters 213, 391-405.

Heier, K.S. and Adams, J.A.S., 1964. The geochemistry of the alkali metals. In: L.H. Ahrens, F. Press and S.K. Runcorn (Eds.), Physics and Chemistry of the Earth, Vol. 5. Pergamon Press Ltd., Oxford, pp. 253-381.

Hemingway, J.E., 1968. Sedimentology of coal-bearing strata. In: D. Murchison and T.S. Westoll (Eds.), Coal and Coal-bearing Strata. Oliver and Boyd, Edinburgh, pp. 43-69.

Herbert, C., 1980. Depositional development of the Sydney Basin In: C. Herbert and R.E. Helby (Eds.), A guide to the Sydney Basin. Geological Survey of New South Wales, Bulletin 26, pp. 11-52.

311

Herbert, C., 1994. Cyclic sedimentation in the lower Newcastle Coal Measures, the 28th Newcastle Symposium, "Advances in the Study of the Sydney Basin". Department of Geology, University of Newcastle, Newcastle, New South Wales, pp. 134-141.

Hickmott, D.D. and Baldridge, W.S., 1995. Application of PIXE microanalysis to macerals and sulfides from the Lower Kittanning Coal of western Pennsylvania. Economic Geology 90, 246- 254.

Hill, P.A., 1988. Tonsteins of Hat Creek, British Columbia: A preliminary study. International Journal of Coal Geology 10, 155-175.

Hillier, S. and Clayton, T., 1989. Illite/smectite diagenesis in Devonian lacustrine mudrocks from northern Scotland and its relationship to organic maturity indicators. Clay Minerals 24, 181-196.

Hillier, S., Matyas, J., Matter, A. and Vasseur, G., 1995. Illite/smectite diagenesis and its variable correlation with vitrinite reflectance in the Pannonian Basin. Clays & Clay Minerals 43, 174-183.

Holz, M., Kalkreuth, W. and Banerjee, I., 2002. Sequence stratigraphy of paralic coal-bearing strata: an overview. International Journal of Coal Geology 48, 147-179.

Howard, K.J. and Fisk, M.R., 1988. Hydrothermal alumina-rich clays and boehmite on the Gorda Ridge. Geochimica et Cosmochimica Acta 52, 2269-2279.

Hower, J., Eslinger, E.V., Hower, M.E. and Perry, E.A., 1976. Mechanism of burial metamorphism of argillaceous sediment: 1. Mineralogical and chemical evidence. Geological Society of America Bulletin 87, 725-737.

Hower, J.C., Rimmer, S.M. and Bland, A.E., 1991. Geochemistry of the blue gem coal bed, Knox county, Kentucky. International Journal of Coal Geology 18, 211-231.

Hower, J.C., Robertson, J.D., Thomas, G.A., Wong, A.S., Schram, W.H., Graham, U.M., Rathbone, R.F. and Robl, T.L., 1996. Characterization of fly ash from Kentucky power plants. Fuel 75, 403-411.

Hower, J.C., Graham, U.M., Wong, A.S., Robertson, J.D., Haeberlin, B.O., Thomas, G.A. and Schram, W.H., 1997a. Influence of flue-gas desulfurization systems on coal combustion by- product quality at kentucky power stations burning high-sulfur coal. Waste Management 17, 523-533.

Hower, J.C., Robertson, J.D., Wong, A.S., Eble, C.F. and Ruppert, L.F., 1997b. Arsenic and lead concentrations in the Pond Creek and Fire Clay coal beds, eastern Kentucky coal field. Applied Geochemistry 12, 281-289.

Hower, J.C., Ruppert, L.F. and Eble, C.F., 1999a. Lanthanide, yttrium, and zirconium anomalies in the Fire Clay coal bed, Eastern Kentucky. International Journal of Coal Geology 39, 141-153.

Hower, J.C., Trimble, A.S., Eble, C.F., Palmer, C.A. and Kolker, A., 1999b. Characterization of fly ash from low-sulfur and high-sulfur coal sources: Partitioning of carbon and trace elements with particle size. Energy Sources 21, 511-525.

312

References

Hower, J.C., Calder, J.H., Eble, C.F., Scott, A.C., Robertson, J.D. and Blanchard, L.J., 2000. Metalliferous coals of the Westphalian A Joggins Formation, Cumberland Basin, Nova Scotia, Canada: petrology, geochemistry, and palynology. International Journal of Coal Geology 42, 185-206.

Hower, J.C., Williams, D.A., Eble, C.F., Sakulpitakphon, T. and Moecher, D.P., 2001. Brecciated and mineralized coals in Union County, Western Kentucky coal field. International Journal of Coal Geology 47, 223-234.

Hower, J.C., Ruppert, L.F. and Williams, D.A., 2002. Controls on boron and germanium distribution in the low-sulfur Amos coal bed, Western Kentucky coalfield, USA. International Journal of Coal Geology 53, 27-42.

Hower, J.C. and Robertson, J.D., 2003. Clausthalite in coal. International Journal of Coal Geology 53, 219-225.

Hower, J.C., Mastalerz, M., Drobniak, A., Quick, J.C., Eble, C.F. and Zimmerer, M.J., 2005. Mercury content of the Springfield coal, Indiana and Kentucky. International Journal of Coal Geology 63, 205-227.

Hower, J.C., Campbell, J.L., Teesdale, W.J., Nejedly, Z. and Robertson, J.D., 2008. Scanning proton microprobe analysis of mercury and other trace elements in Fe-sulfides from a Kentucky coal. International Journal of Coal Geology 75, 88-92.

Huang, W., Zhou, J. and Zheng, R., 2007. Magmatic influences on coal bed and coal quality in the Jining Coalfield, China China Coal Geology 19, 16-17 (In Chinese).

Huang, W., Wan, H., Du, G., Sun, L., Ma, Y., Tang, X., Wu, W. and Qin, S., 2008. Research on Element Geochemical Characteristics of Coal-Ge Deposit in Shengli Coalfield, Inner Mongolia, China. Earth Science Frontiers 15, 56-64.

Huang, X., Gordon, T., Rom, W.N. and Finkelman, R.B., 2006. Interaction of iron and calcium minerals in coals and their roles in coal dust-induced health and environmental problems. Reviews in Mineralogy and Geochemistry 64, 153-178.

Huff, W.D. and Tuerkmenoglu, A.G., 1981. Chemical characteristics and origin of Ordovician K- bentonites along the Cincinnati Arch. Clays and Clay Minerals 29, 113-123.

Huff, W.D., 1983. Misuse of the term “bentonite” for ash beds of Devonian age in the Appalachian basin: Discussion and reply. Geological Society of America Bulletin 94, 681-682.

Huff, W.D. and Kolata, D.R., 1990. Correlation of the Ordovician Deicke and Millbrig K-bentonites between the Mississippi Valley and the southern Appalachians. American Association of Petroleum Geologists Bulletin 74, 1736-1747.

Huff, W.D., Bergström, S.M. and Kolata, D.R., 1992. Gigantic Ordovician volcanic ash fall in North America and Europe: Biological, tectonomagmatic, and event-stratigraphic significance. Geology 20, 875-878.

Huff, W.D., Merriman, R.J., Morgan, D.J. and Roberts, B., 1993. Distribution and tectonic setting of Ordovician K-bentonites in the United Kingdom. Geological Magazine 130, 93-100.

313

Huff, W.D., Bergstrom, S.M., Kolata, D.R. and Sun, H., 1998. The Lower Silurian Osmundsberg K- bentonite. Part II: mineralogy, geochemistry, chemostratigraphy and tectonomagmatic significance. Geological Magazine 135, 15-26.

Huffman, G.P., Huggins, F.E., Shah, N. and Zhao, J., 1994. Speciation of arsenic and chromium in coal and combustion ash by XAFS spectroscopy. Fuel Processing Technology 39, 47-62.

Huggins, F.E., Shah, N., Zhao, J., Lu, F. and Huffman, G.P., 1993. Nondestructive determination of trace element speciation in coal and coal ash by XAFS spectroscopy. Energy & Fuels 7, 482- 489.

Huggins, F.E. and Huffman, G.P., 1995. Chlorine in coal: an XAFS spectroscopic investigation. Fuel 74, 556-569.

Huggins, F.E., Goodarzi, F. and Lafferty, C.J., 1996. Mode of Occurrence of Arsenic in Subbituminous Coals. Energy & Fuels 10, 1001-1004.

Huggins, F.E. and Huffman, G.P., 1996. Modes of occurrence of trace elements in coal from XAFS spectroscopy. International Journal of Coal Geology 32, 31-53.

Huggins, F.E., 2002. Overview of analytical methods for inorganic constituents in coal. International Journal of Coal Geology 50, 169-214.

Huggins, F.E. and Huffman, G.P., 2004. How do lithophile elements occur in organic association in bituminous coals? International Journal of Coal Geology 58, 193-204.

Huggins, F.E., Seidu, L.B.A., Shah, N., Huffman, G.P., Honaker, R.Q., Kyger, J.R., Higgins, B.L., Robertson, J.D., Pal, S. and Seehra, M.S., 2009. Elemental modes of occurrence in an Illinois #6 coal and fractions prepared by physical separation techniques at a coal preparation plant. International Journal of Coal Geology 78, 65-76.

Hutton, A.C., 2009. Geological Setting of Australasian Coal Deposits. In: R. Kininmonth and E. Baafi (Eds.), Australasian Coal Mining Practice. The Australasian Institute of Mining and Metallurgy, Carlton, Victoria, pp. 40-84.

International Committee for Coal and Organic Petrology, 1998. The new vitrinite classification (ICCP System 1994). Fuel 77, 349-358.

International Committee for Coal and Organic Petrology, 2001. The new inertinite classification (ICCP System 1994). Fuel 80, 459-471.

Juster, T.C., Brown, P.E. and Bailey, S.W., 1987. NH4-bearing illite in very low grade metamorphic rocks associated with coal, northeastern Pennsylvania. American Mineralogist 72, 555-565.

Karayigit, A.I. and Whateley, M.K.G., 1997. Properties of a lacustrine subbituminous (k1) seam, with special reference to the contact metamorphism, Soma-Turkey. International Journal of Coal Geology 34, 131-155.

Karayigit, A.I., Spears, D.A. and Booth, C.A., 2000. Antimony and arsenic anomalies in the coal seams from the Gokler coalfield, Gediz, Turkey. International Journal of Coal Geology 44, 1-17.

314

References

Karuppiah, M. and Gupta, G., 1997. Toxicity of and metals in coal combustion ash leachate. Journal of Hazardous Materials 56, 53-58.

Kemezys, M. and Taylor, G.H., 1964. Occurrence and distribution of minerals in some Australian coals. Journal of the Institute of Fuel 37, 389-397.

Ketris, M.P. and Yudovich, Y.E., 2009. Estimations of Clarkes for Carbonaceous biolithes: World averages for trace element contents in black shales and coals. International Journal of Coal Geology 78, 135-148.

Kimura, T., Kawashima, H. and Saito, I., 1994. Smectite and illite/smectite mixed-layer clay minerals in the Ashibetsu coals. International Journal of Coal Geology 26, 215-231.

Kisch, H.J., 1966. Chlorite-illite tonstein in high-rank coals from Queensland, Australia; notes on regional epigenetic grade and coal rank. American Journal of Science 264, 386-397.

Kisch, H.J. and Taylor, G.H., 1966. Metamorphism and alteration near an intrusive-coal contact. Economic Geology 61, 343-361.

Klein, C. and Dutrow, B., 2007. Manual of Mineral Science. John Wiley and Sons, 675 pp.

Knight, J.A., Burger, K. and Bieg, G., 2000. The pyroclastic tonsteins of the Sabero Coalfield, north-western Spain, and their relationship to the stratigraphy and structural geology. International Journal of Coal Geology 44, 187-226.

Knight, W.C., 1898. Bentonite. Engineering and Mining Journal 66, 491-495.

Kolata, D.R., Frost, J.K. and Huff, W.D., 1987. Chemical correlation of K-bentonite beds in the Middle Ordovician Decorah Subgroup, upper Mississippi Valley. Geology 15, 208-211.

Kolker, A. and Chou, C.-L., 1994. Cleat-filling calcite in Illinois Basin Coals: Trace-element evidence for meteoric fluid. Journal of Geology 102, 111.

Kolker, A. and Finkelman, R.B., 1998. Potentially Hazardous Elements in Coal: Modes of Occurrence and Summary of Concentration Data for Coal Components. Coal Preparation 19, 133-157.

Kolker, A., Huggins, F.E., Palmer, C.A., Shah, N., Crowley, S.S., Huffman, G.P. and Finkelman, R.B., 2000. Mode of occurrence of arsenic in four US coals. Fuel Processing Technology 63, 167-178.

Kolker, A., Panov, B.S., Panov, Y.B., Landa, E.R., Conko, K.M., Korchemagin, V.A., Shendrik, T. and McCord, J.D., 2009. Mercury and trace element contents of Donbas coals and associated mine water in the vicinity of Donetsk, Ukraine. International Journal of Coal Geology 79, 83-91.

Kolker, A., 2012. Minor element distribution in iron disulfides in coal: A geochemical review. International Journal of Coal Geology 94, 32-43.

Kortenski, J. and Kostova, I., 1996. Occurrence and morphology of pyrite in Bulgarian coals. International Journal of Coal Geology 29, 273-290.

315

Kostova, I., Petrov, O. and Kortenski, J., 1996. Mineralogy, geochemisty and pyrite content of Bulgarian subbituminous coals, Pernik Basin. In: R. Gayer and I. Harris (Eds.), Coalbed Methane and Coal Geology. Geological society, London, Special Publication 109, London, pp. 301-314.

Koukouzas, N., Ward, C.R. and Li, Z., 2010. Mineralogy of lignites and associated strata in the Mavropigi field of the Ptolemais Basin, northern Greece. International Journal of Coal Geology 81, 182-190.

Kowallis, B.J. and Christiansen, E.J., 1989. Applications of zircon morphology: Correlation of pyroclastic rocks and petrogenic inferences. Geological Society of America Abstracts with Programs 21, A-244.

Kramer, W., Weatherall, G. and Offler, R., 2001. Origin and correlation of tuffs in the Permian Newcastle and Wollombi Coal Measures, NSW, Australia, using chemical fingerprinting. International Journal of Coal Geology 47, 115-135.

Large, D.J., Kelly, S., Spiro, B., Tian, L., Shao, L., Finkelman, R., Zhang, M., Somerfield, C., Plint, 6$OL<DQG=KRX<6LOLFDíYRODWLOHLQWHUDFWLRQDQGWKHJHRORJLFDOFDXVHRIWKH;XDQ Wei lung cancer epidemic. Environmental Science & Technology 43, 9016-9021.

Larssen, T. and Carmichael, G.R., 2000. Acid rain and acidification in China: the importance of base cation deposition. Environmental Pollution 110, 89-102.

Lawrence, L.J., Warne, S.S.J. and Booker, M., 1960. Millerite in the Bulli coal seam. Australian Journal of Science 23, 87-88.

Li, J., Zhuang, X. and Querol, X., 2011. Trace element affinities in two high-Ge coals from China. Fuel 90, 240-247.

Li, S., Kusky, T.M., Wang, L., Zhang, G., Lai, S., Liu, X., Dong, S. and Zhao, G., 2007a. Collision leading to multiple-stage large-scale extrusion in the Qinling orogen: Insights from the Mianlue suture. Gondwana Research 12, 121-143.

Li, Z., Clemens, A.H., Moore, T.A., Gong, D., Weaver, S.D. and Eby, N., 2005. Partitioning behaviour of trace elements in a stoker-fired combustion unit: An example using bituminous coals from the Greymouth coalfield (Cretaceous), New Zealand. International Journal of Coal Geology 63, 98-116.

Li, Z., Ward, C.R. and Gurba, L.W., 2007b. Occurrence of non-mineral inorganic elements in low- rank coal macerals as shown by electron microprobe element mapping techniques. International Journal of Coal Geology 70, 137-149.

Li, Z., Ward, C.R. and Gurba, L.W., 2010. Occurrence of non-mineral inorganic elements in macerals of low-rank coals. International Journal of Coal Geology 81, 242-250.

Lindahl, P.C. and Finkelman, R.B., 1986. Factors Influencing Major, Minor, and Trace Element Variations in U.S. Coals. In: K.S. Vorres (Ed.), Mineral Matter and Ash in Coal. ACS Symposium Series. American Chemical Society, pp. 61-69.

316

References

Liu, D., Yang, Q., Tang, D., Kang, X. and Huang, W., 2001. Geochemistry of sulfur and elements in coals from the Antaibao surface mine, Pingshuo, Shanxi Province, China. International Journal of Coal Geology 46, 51-64.

Liu, J., Hay, R.L., Deino, A. and Kyser, T.K., 2003. Age and origin of authigenic K-feldspar in uppermost Precambrian rocks in the North American Midcontinent. Bulletin of the Geological Society of America 115, 422-433.

Liu, Y., 2006. The influence of structural change of energe sources of Chongqing on the quality of atmosphere. Journal of Chongqing Technology and Business University (Natural Science Edition) 23, 35-39 (in Chinese with English abstract).

Lohe, E.M., 1992a. Regional Structure and Tectonic Setting of the Sydney Basin. In: E.M. Lohe, T.P.T. Mclennan, T.D. Sullivan, K.P. Soole and C.W. Mallett (Eds.), Sydney Basin-Geological Structure and Mining Conditions, Assessment for Mine Planning. CISRO Division of Geomechanics, Institute of Minerals, Energy and Construction, Kenmore, Queensland.

Lohe, E.M., 1992b. Southern Coalfield. In: E.M. Lohe, T.P.T. Mclennan, T.D. Sullivan, K.P. Soole and C.W. Mallett (Eds.), Sydney Basin-Geological Structure and Mining Conditions, Assessment for Mine Planning. CISRO Division of Geomechanics, Institute of Minerals, Energy and Construction, Kenmore, Queensland.

López-Buendía, A.M., Whateley, M.K.G., Bastida, J. and Urquiola, M.M., 2007. Origins of mineral matter in peat marsh and peat bog deposits, Spain. International Journal of Coal Geology 71, 246-262.

López, I.C. and Ward, C.R., 2008. Composition and mode of occurrence of mineral matter in some Colombian coals. International Journal of Coal Geology 73, 3-18.

Loughnan, F.C., 1966. Analcite in the Newcastle Coal Measure sediments of the Sydney Basin, Australia. American Mineralogist 51, 486-494.

Loughnan, F.C. and See, G.T., 1967. Dawsonite in the Greta Coal Measures at Muswellbrook, New South Wales. American Mineralogist 52, 1216-1219.

Loughnan, F.C., 1969. Chemical Weathering of the Silica Minerals. Elsevier, Amsterdam, 153 pp.

Loughnan, F.C., 1970. Flint clay in the coal-barren Triassic of the Sydney Basin, Australia. Journal of Sedimentary Research 40, 822-828.

Loughnan, F.C., 1971a. kaolinite claystones associated with the wongawilli seam in the southern part of the Sydney Basin Journal of the Geological Society of Australia 18, 293-302.

Loughnan, F.C., 1971b. Refractory flint clays of the Sydney Basin. Journal of the Australian Ceramic society 7, 34-43.

Loughnan, F.C. and Ward, C.R., 1971. Pyrophyllite-bearing flint clay from the Cambewarra area, New South Wales. Clay Minerals 9, 83-95.

Loughnan, F.C. and Goldbery, R., 1972. Dawsonite and analcite in the Singleton Coal Measures of the Sydney Basin. American Mineralogist 57, 1437-1447.

317

Loughnan, F.C. and Corkery, R.W., 1975. Oriented-kaolinite aggregates in flint clays and kaolin tonsteins of the Sydney Basin, New South Wales. Clay Minerals 10, 471-473.

Loughnan, F.C. and See, G.T., 1976. Dawsonite in the Greta Coal Measures at Muswellbrook, New South Wales. American Mineralogist 52, 1216-1219.

Loughnan, F.C., 1978. Flint clays, tonsteins and the kaolinite clayrock facies. Clay Minerals 13, 387-400.

Loughnan, F.C. and Ray, A.S., 1978. The Reid's Mistake Formation at Swansea Head, New South Wales. Journal of the Geological Society of Australia 25, 473 - 481.

Lowers, H.A., Breit, G.N., Foster, A.L., Whitney, J., Yount, J., Uddin, M.N. and Muneem, A.A., 2007. Arsenic incorporation into authigenic pyrite, Bengal Basin sediment, Bangladesh. Geochimica et Cosmochimica Acta 71, 2699-2717.

Luo, Z., 1981. The influence of taphrogenesis from late Paleozoic Era in southern China on petroleum and other deposits. Sichuan Geological Journal 2, 1-22 (in Chinese).

Luo, Z., Jin, Y. and Zhao, X., 1990. The Emei Taphrogenesis of the upper Yangtze Platform in south China. Geological Magazine 127, 393-405.

Lyons, P.C., Palmer, C.A., Bostick, N.H., Fletcher, J.D., Dulong, F.T., Brown, F.W., Brown, Z.A., Krasnow, M.R. and Romankiw, L.A., 1989. Chemistry and origin of minor and trace elements in vitrinite concentrates from a rank series from the eastern United States, England, and Australia. International Journal of Coal Geology 13, 481-527.

Lyons, P.C., Outerbridge, W.F., Triplehorn, D.M., Evans, H.T., JR., Congdon, R.D., Capiro, M., Hess, J.C. and Nash, W.P., 1992. An Appalachian isochron: A kaolinized Carboniferous air-fall volcanic-ash deposit (tonstein). Geological Society of America Bulletin 104, 1515-1527.

Lyons, P.C., Spears, D.A., Outerbridge, W.F., Congdon, R.D. and Evans, H.T., 1994. Euramerican tonsteins: overview, magmatic origin, and depositional-tectonic implications. Palaeogeography, Palaeoclimatology, Palaeoecology 106, 113-134.

Lyons, P.C., Krogh, T.E., Kwok, Y.Y., Davis, D.W., Outerbridge, W.F. and Evans Jr, H.T., 2006. Radiometric ages of the Fire Clay tonstein [Pennsylvanian (Upper Carboniferous), Westphalian, Duckmantian]: A comparison of U–Pb zircon single-crystal ages and 40Ar/39Ar sanidine single-crystal plateau ages. International Journal of Coal Geology 67, 259-266.

Mackowsky, M.-T., 1982. Minerals and trace elements occuring in coal. In: E. Stach, M.-T. Mackowsky, M. Teichmuller, G.H. Taylor, D. Chandra and R. Teichmuller (Eds.), Textbook of Coal Petrology. Gebruder Borntraeger, Berlin, pp. 153-170.

Martinez-Tarazona, M.R., Suarez-Fernandez, G.P. and Cardin, J.M., 1994. Fluorine in Asturian coals. Fuel 73, 1209-1213.

Mastalerz, M. and Drobniak, A., 2005. Vertical variations of mercury in Pennsylvanian coal beds from Indiana. International Journal of Coal Geology 63, 36-57.

318

References

Mastalerz, M. and Drobniak, A., 2007. Arsenic, cadmium, lead, and zinc in the Danville and Springfield coal members (Pennsylvanian) from Indiana. International Journal of Coal Geology 71, 37-53.

Mastalerz, M., Drobniak, A. and Schimmelmann, A., 2009. Changes in optical properties, chemistry, and micropore and mesopore characteristics of bituminous coal at the contact with dikes in the Illinois Basin. International Journal of Coal Geology 77, 310-319.

Mastalerz, M. and Drobniak, A., 2012. Gallium and germanium in selected Indiana coals. International Journal of Coal Geology 94, 302-313.

Maynard, J.B., Elswick, E.R. and Hower, J.C., 2001. Reflectance of dispersed vitrinite in shales hosting Pb–Zn–Cu ore deposits in western Cuba: comparison with clay crystallinity. International Journal of Coal Geology 47, 161-170.

Merritt, R.D., 1990. Thermal alteration and rank variation of coals in the Matanuska field, south- central Alaska. International Journal of Coal Geology 14, 255-276.

Meyers, P.A. and Simoneit, B.R.T., 1999. Effects of extreme heating on the elemental and isotopic compositions of an Upper Cretaceous coal. Organic Geochemistry 30, 299-305.

Miller, R.N. and Given, P.H., 1986. The association of major, minor and trace inorganic elements with lignites. I. Experimental approach and study of a North Dakota lignite. Geochimica et Cosmochimica Acta 50, 2033-2043.

Minkin, J.A., Finkelman, R.B., Thompson, C.L., Chao, E.C.T., Ruppert, L.F., Blank, H. and Cecil, C.B., 1984. MICROCHARACTERIZATION OF ARSENIC- AND SELENIUM-BEARING PYRITE IN UPPER FREEPORT COAL, INDIANA COUNTY, PENNSYLVANIA. Scanning Electron Microscopy, 1515-1529.

Moore, D.M. and Reynolds, R.C., 1997. X-ray Diffraction and the Identification and Analysis of Clay Minerals. Oxford University Press, Oxford, UK.

Murowchick, J.B. and Barnes, H.L., 1986. Marcasite precipitation from hydrothermal solutions. Geochimica et Cosmochimica Acta 50, 2615-2629.

Murray, C.G., 1990. Tectonic evolution and metallogeny of the Bowen Basin, Bowen Basin Symposium. Geological Society of Australia, Mackay, Queensland, pp. 201-212.

Mutakyahwa, M.K.D., 2002. Mineralogy and chemistry of bentonite (?) deposits at Minjingu, Lake Manyara, North Tanzania. Journal of African Earth Sciences 34, 213-221.

Nicholls, G.D., 1968. the geochemistry of coal-bearing strata. In: D.G. Murchison and T.S. Westoll (Eds.), Coal and Coal-bearing Strata. Oliver and Boyd, Edinburgh, pp. 267-307.

Nieto, F., 2002. Characterization of coexisting NH4- and K-micas in very low-grade metapelites. American Mineralogist 87, 205-216.

Nissenbaum, A. and Swaine, D.J., 1976. Organic matter-metal interactions in Recent sediments: the role of humic substances. Geochimica et Cosmochimica Acta 40, 809-816.

319

Oliveira, M.L.S., Ward, C.R., French, D., Hower, J.C., Querol, X. and Silva, L.F.O., 2012. Mineralogy and leaching characteristics of beneficiated coal products from Santa Catarina, Brazil. International Journal of Coal Geology 94, 314-325.

Palmer, C.A. and Lyons, P.C., 1990. Chemistry and origin of minor and trace elements in selected vitrinite concentrates from bituminous and anthracitic coals. International Journal of Coal Geology 16, 189-192.

Palmer, C.A. and Lyons, P.C., 1996. Selected elements in major minerals from bituminous coal as determined by INAA: implications for removing environmentally sensitive elements from coal. International Journal of Coal Geology 32, 151-166.

Palmer, C.A., Finkelman, R.B., Krasnow, M.R., Sellers, G.A. and Aruscavage, P.J., 1989. Behavior of selected elements during sequential leaching of 10 lignite and bituminous coals, International Conference on Trace Elements in Coal, Western Kentucky University, pp. 21.

Patterson, J.H., Corcoran, J.F. and Kinealy, K.M., 1995. Carbonate mineralogy of the Hoskissons seam and equivalents in the Sydney and Gunnedah Basins. In: R.L. Boyd and G.A. Mckenzie (Eds.), the 29th Newcastle Symposium, "Advances in the Study of the Sydney Basin". Department of Geology, University of Newcastle, Newcastle, New South Wales, pp. 94-101.

Pearson, D.E. and Kwong, J., 1979. Mineral matter as a measure of oxidation of a coking coal. Fuel 58, 63-66.

Percival, C.J., 1983. A definition of the term ganister. Geological Magazine 120, 187-190.

Permana, A., 2011. Mineralogical varation and changes in the South Walker Creek Coals, Bowen Basin, Queensland, Australia. Master Thesis, University of New South Wales, Sydney, 276 pp.

Permana, A.K., Ward, C.R., Li, Z., Gurba, L.W. and Davison, S., 2010. Mineral matter in the high rank coals of the South Walker Creek area, Northern Bowen Basin. In: J.W. Beeston (Ed.), Bowen Basin Symposium 2010—Back in (the) Black Geological Society of Australia Inc. Coal Geology Group and the Bowen Basin Geologists Group, Mackay, Queensland, pp. 27-34.

Perry, E. and Hower, J., 1970. Burial diagenesis in Gulf Coast pelitic sediments. Clays & Clay Minerals 18, 165-177.

Pevear, D.R., Williams, V.E. and Mustoe, G.E., 1980. Kaolinite, smectite, and K- rectorite in bentonites: relation to coal rank at Tulameen, British Columbia. Clays & Clay Minerals 28, 241- 254.

Price, N.B. and Duff, P.M.D., 1969. Mineralogy and chemistry of tonsteins from Carboniferous sequences in Great Britain. Sedimentology 13, 45-69.

Qi, H., Hu, R. and Zhang, Q., 2007a. Concentration and distribution of trace elements in lignite from the Shengli Coalfield, Inner Mongolia, China: Implications on origin of the associated Wulantuga Germanium Deposit. International Journal of Coal Geology 71, 129-152.

Qi, H., Hu, R. and Zhang, Q., 2007b. REE Geochemistry of the Cretaceous lignite from Wulantuga Germanium Deposit, Inner Mongolia, Northeastern China. International Journal of Coal Geology 71, 329-344.

320

References

Querol, X., Chinchon, S. and Lopez-Soler, A., 1989. Iron sulfide precipitation sequence in Albian coals from the Maestrazgo Basin, southeastern Iberian Range, northeastern Spain. International Journal of Coal Geology 11, 171-189.

Querol, X., Fernandez Turiel, J.L., Lopez Soler, A. and Duran, M.E., 1992. Trace elements in high- S subbituminous coals from the teruel Mining District, northeast Spain. Applied Geochemistry 7, 547-561.

Querol, X. and Chenery, S., 1995. Determination of trace element affinities in coal by laser ablation microprobe-inductively coupled plasma spectrometry. In: M.K.G. Whateley and D.A. Spears (Eds.), European Coal Geology. Geological Society Special Publication No. 82, pp. 147-155.

Querol, X., Fernández-Turiel, J. and López-Soler, A., 1995. Trace elements in coal and their behaviour during combustion in a large power station. Fuel 74, 331-343.

Querol, X., Cabrera, L., Pickel, W., López-Soler, A., Hagemann, H.W. and Fernández-Turiel, J.L., 1996. Geological controls on the coal quality of the Mequinenza subbituminous coal deposit, northeast Spain. International Journal of Coal Geology 29, 67-91.

Querol, X., Alastuey, A., Lopez-Soler, A., Plana, F., Fernandez-Turiel, J.L., Zeng, R., Xu, W., Zhuang, X. and Spiro, B., 1997a. Geological controls on the mineral matter and trace elements of coals from the Fuxin basin, Liaoning Province, northeast China. International Journal of Coal Geology 34, 89-109.

Querol, X., Whateley, M.K.G., Fernández-Turiel, J.L. and Tuncali, E., 1997b. Geological controls on the mineralogy and geochemistry of the Beypazari lignite, central Anatolia, Turkey. International Journal of Coal Geology 33, 255-271.

Querol, X., Klika, Z., Weiss, Z., Finkelman, R.B., Alastuey, A., Juan, R., López-Soler, A., Plana, F., Kolker, A. and Chenery, S.R.N., 2001. Determination of element affinities by density fractionation of bulk coal samples. Fuel 80, 83-96.

Rao, C.P. and Gluskoter, H.J., 1973. Occurrence and distribution of minerals in Illinois coals 476, 56 pp.

Rao, P.D. and Walsh, D.E., 1997. Nature and distribution of phosphorus minerals in Cook Inlet coals, Alaska. International Journal of Coal Geology 33, 19-42.

Rao, P.D. and Walsh, D.E., 1999. Influence of environments of coal deposition on phosphorous accumulation in a high latitude, northern Alaska, coal seam. International Journal of Coal Geology 38, 261-284.

Ren, D., 1996. Mineral matters in coal. In: D. Han (Ed.), Coal petrology of China. Publishing House of China University of Mining and Technology, Xuzhou, pp. 67-77 (in Chinese with English abstract).

Ren, D., Zhao, F., Wang, Y. and Yang, S., 1999. Distributions of minor and trace elements in Chinese coals. International Journal of Coal Geology 40, 109-118.

Ren, D., Zhao, F., Dai, S., Zhang, J. and Luo, K., 2006. Geochemistry of Trace Elements in Coal. Science Press, Beijing, 556 pp (in Chinese with English abstract).

321

Renton, J.J., 1982. Mineral matter in coal. In: R.A. Meyers (Ed.), Coal Structure. Academic Press, New York, pp. 283-325.

Renton, J.J. and Bird, D.S., 1991. Association of coal macerals, sulfur, sulfur species and the iron disulphide minerals in three columns of the Pittsburgh coal. International Journal of Coal Geology 17, 21-50.

Retallack, G.J., 1977. Triassic palaeosols in the upper Narrabeen Group of New South Wales. Part I: Features of the palaeosols. Journal of the Geological Society of Australia 23, 383-399.

Riley, K.W., French, D.H., Lambropoulos, N.A., Farrell, O.P., Wood, R.A. and Huggins, F.E., 2007. Origin and occurrence of selenium in some Australian coals. International Journal of Coal Geology 72, 72-80.

Riley, K.W., French, D.H., Farrell, O.P., Wood, R.A. and Huggins, F.E., 2012. Modes of occurrence of trace and minor elements in some Australian coals. International Journal of Coal Geology 94, 214-224.

Roberts, B. and Merriman, R.J., 1990. Cambrian and Ordovician metabentonites and their relevance to the origins of associated mudrocks in the northern sector of the Lower Palaeozoic Welsh marginal basin. Geological Magazine 127, 31-43.

Ruan, C.-D. and Ward, C.R., 2002. Quantitative X-ray powder diffraction analysis of clay minerals in Australian coals using Rietveld methods. Applied Clay Science 21, 227-240.

Ruppert, L., Finkelman, R., Boti, E., Milosavljevic, M., Tewalt, S., Simon, N. and Dulong, F., 1996. Origin and significance of high nickel and chromium concentrations in Pliocene lignite of the Kosovo Basin, Serbia. International Journal of Coal Geology 29, 235-258.

Ruppert, L.F., Cecil, C.B., Stanton, R.W. and Christian, R.P., 1985. Authigenic quartz in the Upper Freeport coal bed, west-central Pennsylvania. Journal of Sedimentary Petrology 55, 334-339.

Ruppert, L.F., Stanton, R.W., Blaine Cecil, C., Eble, C.F. and Dulong, F.T., 1991. Effects of detrital influx in the Pennsylvanian Upper Freeport peat swamp. International Journal of Coal Geology 17, 95-116.

Ruppert, L.F., Minkin, J.A., McGee, J.J. and Cecil, C.B., 1992. An unusual occurrence of arsenic- bearing pyrite in the Upper Freeport coal bed, west-central Pennsylvania. Energy & Fuels 6, 120-125.

Ruppert, L.F. and Moore, T.A., 1993. Differentiation of volcanic ash-fall and water-borne detrital layers in the Eocene Senakin coal bed, Tanjung Formation, Indonesia. Organic Geochemistry 20, 233-247.

Ruppert, L.F., Neuzil, S.G., Cecil, C.B. and Kane, J.S., 1993. Inoragnic constituents from samples of a domed and lacustrine peat, Sumatra, Indonesia. Geological Society of America, Special paper 286, 83-96.

Ruppert, L.F., Hower, J.C. and Eble, C.F., 2005. Arsenic-bearing pyrite and marcasite in the Fire Clay coal bed, Middle Pennsylvanian Breathitt Formation, eastern Kentucky. International Journal of Coal Geology 63, 27-35.

322

References

Sandler, A., Harlavan, Y. and Steinitz, G., 2004. Early formation of K-feldspar in shallow-marine sediments at near-surface temperatures (southern Israel): Evidence from K-Ar dating. Sedimentology 51, 323-338.

Scheibner, E. and Basden, H., 1998. Geology of New South Wales-Synthesis, Volume 2 Geological Evolution, 13(2). Geological Survey of New South Wales, Memoir Geology, 666 pp.

Schoonen, M.A.A. and Barnes, H.L., 1991. Reactions forming pyrite and marcasite from solution: I. Nucleation of FeS2 below 100°C. Geochimica et Cosmochimica Acta 55, 1495-1504.

Schopf, J.M., 1956. A definition of coal. Economic Geology 51, 521-527.

Senior, C.L., Zeng, T., Che, J., Ames, M.R., Sarofim, A.F., Olmez, I., Huggins, F.E., Shah, N., Huffman, G.P., Kolker, A., Mroczkowski, S., Palmer, C. and Finkelman, R., 2000. Distribution of trace elements in selected pulverized coals as a function of particle size and density. Fuel Processing Technology 63, 215-241.

Senkayi, A.L., Dixon, J.B., Hossner, L.R., Abder-Ruhman, M. and Fanning, D.S., 1984. Mineralogy and genetic relationships of tonstein, bentonite, and lignitic strata in the Eocene Yegua Formation of east-central Texas Clay and Clay minerals 32, 259-271.

Seredin, V.V., 1996. Rare earth element-bearing coals from the Russian Far East deposits. International Journal of Coal Geology 30, 101-129.

Seredin, V.V., 1998. Rare earth mineralization in Late Cenozoic explosion structures (Khankai Massif, Primorskii Krai, Russia). Geology of Ore Deposits 40, 357-371.

Seredin, V.V. and Shpirt, M.Y., 1999. Rare earth elements in the humic substance of metalliferous coals. Lithology and Mineral Resources 34, 281-286.

Seredin, V.V., 2001. Major regularities of the REE distribution in coal. Doklady Earth Sciences 377, 250-253.

Seredin, V.V., 2004. The Au-PGE mineralization at the Pavlovsk brown coal deposit, primorye. Geology of Ore Deposits 46, 36-63.

Seredin, V.V. and Finkelman, R.B., 2008. Metalliferous coals: A review of the main genetic and geochemical types. International Journal of Coal Geology 76, 253-289.

Seredin, V.V. and Dai, S., 2012. Coal deposits as potential alternative sources for lanthanides and yttrium. International Journal of Coal Geology 94, 67-93.

Shi, G.R. and McLoughlin, S., 1997. Permian stratigraphy, sedimentology and palaeontology of the southern Sydney Basin, eastern Australia - A field excursion guide, School of Aquatic Science and Natural Resources Management, Deakin University Clayton, Victoria.

Simmons, S.F. and Browne, P.R.L., 1997. Saline fluid inclusions in sphalerite from the Broadlands- Ohaaki geothermal system; a coincidental trapping of fluids being boiled toward dryness. Economic Geology 92, 485-489.

323

Simpson, M.P., Mauk, J.L. and Simmons, S.F., 2001. Hydrothermal Alteration and Hydrologic Evolution of the Golden Cross Epithermal Au-Ag Deposit, New Zealand. Economic Geology 96, 773-796.

Skipsey, E., 1974. Distribution of alkali chlorides in British coal seams. Fuel 53, 258-267.

Skipsey, E., 1975. Relations between chlorine in coal and the salinity of strata waters. Fuel 54, 121-125.

Smyth, M., 1966. A siderite-pyrite association in Australian coals. Fuel 45, 221-231.

Solari, J.A., Fiedler, H. and Schneider, C.L., 1989. Modelling of the distribution of trace elements in coal. Fuel 68, 536-539.

Song, C., 1989. A brief description of the Yutangba sedimentary type Se mineralized area in southwestern Hubei. Mineral Deposits 8, 83–88 (in Chinese with English abstract).

Sowa, W.A., Hedman, P.O., Smoot, L.D. and Blackham, A.U., 1989. The behaviour of chlorine in Kentucky and Illinois coals during combustion and its effects on ash deposits. Fuel 68, 573- 580.

Spears, Kanaris, S., Riley and Krause, 1999a. Namurian bentonites in the Pennine Basin, UK – origin and magmatic affinities. Sedimentology 46, 385-401.

Spears, D.A., 1970. A kaolinite mudstone (tonstein) in the British coal measures. Journal of Sedimentary Petrology 40, 386-394.

Spears, D.A., 1971. The mineralogy of the Stafford tonstein, the Yorkshire Geological Society, pp. 497-516.

Spears, D.A. and Rice, C.M., 1973. An Upper Carboniferous tonstein of volcanic origin. Sedimentology 20, 281-294.

Spears, D.A. and Kanaris-Sotiriou, R., 1976. Titanium in some Carboniferous sediments from Great Britain. Geochimica et Cosmochimica Acta 40, 345-351.

Spears, D.A. and Kanaris-Sotiriou, R., 1979. A geochemical and mineralogical investigation of some British and other European tonsteins. Sedimentology 26, 407-425.

Spears, D.A. and Duff, P.M.D., 1984. Kaolinite and mixed-layer illite–smectite in Lower Cretaceous bentonites from the Peace River coalfield, British Columbia. Canadian Journal of Earth Sciences 21, 465-476.

Spears, D.A. and Caswell, S.A., 1986. Mineral matter in coals: cleat minerals and their origin in some coals from the English midlands. International Journal of Coal Geology 6, 107-125.

Spears, D.A., 1987. Mineral matter in coal, with special reference to the Pennine Coalfields. In: A.C. Scott (Ed.), Coal and Coal-Bearing Strata-Recent Advances. Geological Society Special Publication 32, pp. 171-185.

324

References

Spears, D.A., Duff, P.M.D. and Caine, P.M., 1988. The West Waterberg tonstein, South Africa. International Journal of Coal Geology 9, 221-233.

Spears, D.A. and Martinez-Tarazona, M.R., 1993. Geochemical and mineralogical characteristics of a power station feed-coal, Eggborough, England. International Journal of Coal Geology 22, 1-20.

Spears, D.A. and Lyons, P.C., 1995. An update on British Tonsteins. In: M.K.G. Whateley and D.A. Spears (Eds.), European Coal Geology. Geological Society Special Publication No. 82, pp. 137-146.

Spears, D.A., Manzanares-Papayanopoulos, L.I. and Booth, C.A., 1999b. The distribution and origin of trace elements in a UK coal; the importance of pyrite. Fuel 78, 1671-1677.

Spears, D.A. and Zheng, Y., 1999. Geochemistry and origin of elements in some UK coals. International Journal of Coal Geology 38, 161-179.

Spears, D.A., 2000. Role of clay minerals in UK coal combustion. Applied Clay Science 16, 87-95.

Spears, D.A. and Martinez-Tarrazona, M.R., 2004. Trace elements in combustion residues from a UK power station. Fuel 83, 2265-2270.

Spears, D.A., 2005. A review of chlorine and bromine in some United Kingdom coals. International Journal of Coal Geology 64, 257-265.

Spears, D.A., 2006. Clay mineralogy of onshore UK Carboniferous mudrocks. Clay Minerals 41, 395-416.

Spears, D.A., Borrego, A.G., Cox, A. and Martinez-Tarazona, R.M., 2007. Use of laser ablation ICP-MS to determine trace element distributions in coals, with special reference to V, Ge and Al. International Journal of Coal Geology 72, 165-176.

Spears, D.A. and Tewalt, S.J., 2009. The geochemistry of environmentally important trace elements in UK coals, with special reference to the Parkgate coal in the Yorkshire– Nottinghamshire Coalfield, UK. International Journal of Coal Geology 80, 157-166.

Spears, D.A., 2011. The origin of tonsteins, an overview, and links with seatearths, fireclays and fragmental clay rocks. International Journal of Coal Geology 94, 22-31.

Speight, J.G., 1983. the chemistry and technology of coal. Marcel Dekker, New York, 528 pp.

Standards Australia, 1993. Guide to the technical evaluation of higher rank coal deposits. Australian Standard 1038.10.5.2, pp. 120.

Standards Australia, 1995. Coal and coke – glossary of terms. Australian Standard 2418, pp. 180.

Standards Australia, 1998a. Coal and coke – Analysis and testing. Part 10.3: Determination of trace elements – Coal and coke – Determination of boron content – ICP-AES method. Australian Standard 1038.10.3, pp. 7.

325

Standards Australia, 1998b. Coal petrography – maceral analysis. Australian Standard 2856-2, pp. 32.

Standards Australia, 1998c. Coal and coke – Analysis and testing. Part 10.2: Determination of trace elements-Coal and coke-Determination of arsenic, antimony and selenium – Hydride generation method. Australian Standard 1038.10.2, pp. 9.

Standards Australia, 1999. Coal and coke – Analysis and testing. Part 8.1: Coal and coke – Chlorine – Eschka method. Australian Standard 1038.8.1, pp. 6.

Standards Australia, 2000a. Coal petrography – methods for microscopical determinations of the reflectance of coal macerals. Australian Standard 2856-3, pp. 16.

Standards Australia, 2000b. Coal and coke – analysis and testing: Higher rank coal – mineral matter and water of constitution. Australian Standard 1038.22, pp. 20.

Standards Australia, 2007a. Coal and coke—Analysis and testing. Part 10.5.2: Coal and fly-ash— Trace elements—Determination of mercury content—Acid extraction method. Australian Standard 1038.10.5.2, pp. 14.

Standards Australia, 2007b. Symbols for graphic representation of coal seams and associated strata. Australian Standard 2916, pp. 21.

Stewart, A.K., Massey, M., Padgett, P.L., Rimmer, S.M. and Hower, J.C., 2005. Influence of a basic intrusion on the vitrinite reflectance and chemistry of the Springfield (No. 5) coal, Harrisburg, Illinois. International Journal of Coal Geology 63, 58-67.

Stewart, J.R. and Alder, J.D. (Eds.), 1995. New South Wales Petroleum Potential. New South Wales Department of Mineral Resources, Sydney, 188 pp.

Susilawati, R. and Ward, C.R., 2006. Metamorphism of mineral matter in coal from the Bukit Asam deposit, south Sumatra, Indonesia. International Journal of Coal Geology 68, 171-195.

Swaine, D.J., 1983. Geological aspects of trace elements in coal. In: S.S. Augustithis (Ed.), The Significance of Trace Elements in Solving Petrogenetic Problems and Controversies. Theophrastus Publications, Athens, pp. 521-532.

Swaine, D.J., 1990. Trace elements in coal. Butterworths, London.

Swaine, D.J., 1992. The organic association of elements in coals. Organic Geochemistry 18, 259- 261.

Sykes, R. and Lindqvist, J.K., 1993. Diagenetic quartz and amorphous silica in New Zealand coals. Organic Geochemistry 20, 855-866.

Tadros, N.Z., 1995. Sydney-Gunnedah Basin overiew. In: C.R. Ward, H.J. Harrington, C.W. Mallett and J.W. Beeston (Eds.), Geology of Australian Coal Basins. Geological Society of Australia Coal Geology Group Special Publication, pp. 163-175.

326

References

Tang, X., Luo, L. and Zhang, H., 2009. The study on the space-time distribution characteristics of acid rain in Southwest. Plateau and Mountain Meteorology Research 29, 33-36 (in Chinese with English abstract).

Taylor, G.H., 1991a. Reflections on reflectance, Advances in the Study of the Sydney Basin, 25th Newcastle Symposium proceedings, Newcastle, pp. 222-226.

Taylor, G.H., Teichmuller, M., Davis, A., Diessel, C.F.K., Littke, R. and Robert, P., 1998. Organic Petrology. Gebruder Borntraeger, Berlin, 704 pp.

Taylor, J.C., 1991b. Computer programs for standardless quantitative analysis of minerals using the full powder diffraction profile. Powder Diffraction 6, 2-9.

Taylor, S.R. and McLennan, S.M., 1985. The Continental Crust: Its Composition and Evolution. Blackwell, Oxford, 312 pp.

Teichmuller, M. and Teichmuller, R., 1979. Diagenesis of Coal (Coalification). In: G. Larsen and C. G.V. (Eds.), Deveolopments in Sedimentology 25A. Elsevier Scientific Publishing Co., New York, pp. 207-246.

Tian, L., Dai, S., Wang, J., Huang, Y., Ho, S., Zhou, Y., Lucas, D. and Koshland, C., 2008. Nanoquartz in Late Permian C1 coal and the high incidence of female lung cancer in the Pearl River Origin area: a retrospective cohort study. BMC Public Health 8, 398.

Tourtelot, H.A. and Brenner-Tourtelot, E.F., 1978. Lithium, a preliminary survey of its mineral occurrence in flint clay and related rock types in the United States. Energy 3, 263-272.

Triplehorn, D. and Bohor, B., 1986. Volcanic Ash Layers in Coal: Origin, Distribution, Composition, and Significance, Mineral Matter and Ash in Coal. ACS Symposium Series. American Chemical Society, pp. 90-98.

Triplehorn, D.M. and Bohor, B.F., 1983. Goyazite in kaolinitic altered tuff beds of Cretaceous age near Denver, Colorado. Clays & Clay Minerals 31, 299-304.

Triplehorn, D.M., Stanton, R.W., Ruppert, L.F. and Crowley, S.S., 1991. Volcanic ash dispersed in the Wyodak-Anderson coal bed, Powder River Basin, Wyoming. Organic Geochemistry 17, 567-575.

Turner, D.L., Frizzell, V.A., Triplehorn, D.M. and Naeser, C.W., 1983. Radiometric dating of ash partings in coal of the Eocene Puget Group, Washington: Implications for paleobotanical stages. Geology 11, 527-531.

Uysal, I.T., Glikson, M., Golding, S.D. and Audsley, F., 2000. The thermal history of the Bowen Basin, Queensland, Australia: vitrinite reflectance and clay mineralogy of Late Permian coal measures. Tectonophysics 323, 105-129. van Der Flier-Keller, E. and Fyfe, W.S., 1988. Relationships between inorganic constituents and organic matter in a northern Ontario lignite. Fuel 67, 1048-1052.

327

Van Heeswijck, A., 2001. Sequence stratigraphy of coal-bearing, high-energy clastic units: The maitland-cessnock-greta coalfield and cranky corner basin. Australian Journal of Earth Sciences 48, 417-426.

Vassilev, S.V., Yossifova, M.G. and Vassileva, C.G., 1994. Mineralogy and geochemistry of Bobov Dol coals, Bulgaria. International Journal of Coal Geology 26, 185-213.

Vassilev, S.V. and Vassileva, C.G., 1996. Occurrence, abundance and origin of minerals in coals and coal ashes. Fuel Processing Technology 48, 85-106.

Vassilev, S.V. and Vassileva, C.G., 1997. Geochemistry of coals, coal ashes and combustion wastes from coal-fired power stations. Fuel Processing Technology 51, 19-45.

Vassilev, S.V., Eskenazy, G.M. and Vassileva, C.G., 2000. Contents, modes of occurrence and behaviour of chlorine and bromine in combustion wastes from coal-fired power stations. Fuel 79, 923-938.

Vassilev, S.V., Vassileva, C.G., Karayigit, A.I., Bulut, Y., Alastuey, A. and Querol, X., 2005. Phase– mineral and chemical composition of composite samples from feed coals, bottom ashes and fly ashes at the Soma power station, Turkey. International Journal of Coal Geology 61, 35-63.

Wang, H., Shao, L., Hao, L., Zhang, P., Glasspool, I.J., Wheeley, J.R., Wignall, P.B., Yi, T., Zhang, M. and Hilton, J., 2011. Sedimentology and sequence stratigraphy of the Lopingian (Late Permian) coal measures in southwestern China. International Journal of Coal Geology 85, 168-183.

Wang, J., Bao, C., Lou, Z. and Guo, Z., 1989. Formation and development of the Sichuan Basin. In: X. Zhu (Ed.), Chinese Sedimentary Basins. Elesvier, Amsterdam, pp. 147-163.

Wang, W., Qin, Y., Sang, S., Jiang, B., Zhu, Y. and Guo, Y., 2007. Sulfur variability and element geochemistry of the No. 11 coal seam from the Antaibao mining district, China. Fuel 86, 777- 784.

Wang, x. (Ed.), 1996. Sedimentary Environments and Coal Accumulation of Late Permian Coal Formation in Western Guizhou, Southern Sichuan, and Eastern Yunnan, China. Chongqing University Press, Chongqing, 242 pp (in Chinese with English abstract).

Wang, X., 2009. Geochemistry of Late Triassic coals in the Changhe Mine, Sichuan Basin, southwestern China: Evidence for authigenic lanthanide enrichment. International Journal of Coal Geology 80, 167-174.

Wang, Y., Mo, J. and Ren, D., 1999a. Distribution of minor and trace elements in magmatic hydrothermal metamorphic coal of Meitian Coal Mine, Hunan Province. Geochimica 28, 289- 296.

Wang, Y., Ren, D. and Zhao, F., 1999b. Comparative leaching experiments for trace elements in raw coal, laboratory ash, fly ash and bottom ash. International Journal of Coal Geology 40, 103-108.

Ward, C.R., 1972. Sedimentation in the Narrabeen Group, southern Sydney Basin, New South Wales. Journal of the Geological Society of Australia 19, 393-409.

328

References

Ward, C.R., 1978. Mineral matter in Australian bituminous coals. Proceedings of the Australian Institute of Mining and Metallurgy 267, 7-25.

Ward, C.R. (Ed.), 1984. Coal geology and coal technology. Blackwell, Melbourne, 345 pp.

Ward, C.R., 1989. Minerals in bituminous coals of the Sydney Basin (Australia) and the Illinois Basin (U.S.A.). International Journal of Coal Geology 13, 455-479.

Ward, C.R., Warbrooke, P.R. and Ivor Roberts, F., 1989. Geochemical and mineralogical changes in a coal seam due to contact metamorphism, Sydney Basin, New South Wales, Australia. International Journal of Coal Geology 11, 105-125.

Ward, C.R. and Roberts, F.I., 1990. Occurrence of spherical halloysite in bituminous coals of the Sydney Basin, Australia. Clays and Clay Minerals 38, 501-506.

Ward, C.R., 1991. Mineral matter in low-rank coals and associated strata of the Mae Moh basin, northern Thailand. International Journal of Coal Geology 17, 69-93.

Ward, C.R., 1992. Mineral matter in triassic and tertiary low-rank coals from South Australia. International Journal of Coal Geology 20, 185-208.

Ward, C.R. and Christie, P.J., 1994. Clays and other minerals in coal seams of the Moura-Baralaba area, Bowen Basin, Australia. International Journal of Coal Geology 25, 287-309.

Ward, C.R., Corcoran, J.F., Saxby, J.D. and Read, H.W., 1996. Occurrence of phosphorus minerals in Australian coal seams. International Journal of Coal Geology 30, 185-210.

Ward, C.R., Spears, D.A., Booth, C.A., Staton, I. and Gurba, L.W., 1999a. Mineral matter and trace elements in coals of the Gunnedah Basin, New South Wales, Australia. International Journal of Coal Geology 40, 281-308.

Ward, C.R., Taylor, J.C. and Cohen, D.R., 1999b. Quantitative mineralogy of sandstones by X-ray diffractometry and normative analysis. Journal of Sedimentary Research 69, 1050-1062.

Ward, C.R., Taylor, J.C., Matulis, C.E. and Dale, L.S., 2001. Quantification of mineral matter in the Argonne Premium Coals using interactive Rietveld-based X-ray diffraction. International Journal of Coal Geology 46, 67-82.

Ward, C.R., 2002. Analysis and significance of mineral matter in coal seams. International Journal of Coal Geology 50, 135-168.

Ward, C.R., Li, Z. and Gurba, L.W., 2005. Variations in coal maceral chemistry with rank advance in the German Creek and Moranbah Coal Measures of the Bowen Basin, Australia, using electron microprobe techniques. International Journal of Coal Geology 63, 117-129.

Ward, C.R., Li, Z. and Gurba, L.W., 2007. Variations in elemental composition of macerals with vitrinite reflectance and organic sulphur in the Greta Coal Measures, New South Wales, Australia. International Journal of Coal Geology 69, 205-219.

329

Ward, C.R., French, D., Jankowski, J., Dubikova, M., Li, Z. and Riley, K.W., 2009. Element mobility from fresh and long-stored acidic fly ashes associated with an Australian power station. International Journal of Coal Geology 80, 224-236.

Weaver, C.E., 1953. Mineralogy and petrology of some Ordovician K-bentonite and related limestones. Geological Society of America Bulletin 64, 921-944.

Wedepohl, K.H., 1978. Handbook of geochemistry 2, 2. Springer, Heidelberg.

West, P.H. and Bradley, G.M., 1986. The Gunnedah and Sydney Basins - a re-appraisal, the 21st Newcastle Symposium, "Advances in the study of the Sydney Basin", the University of Newcastle, pp. 1-9.

White, R.N., Smith, J.V., Spears, D.A., Rivers, M.L. and Sutton, S.R., 1989. Analysis of iron sulphides from UK coal by synchrotron radiation X-ray fluorescence. Fuel 68, 1480-1486.

Wiese, R.G. and Fyfe, W.S., 1986. Occurrences of iron sulfides in Ohio coals. International Journal of Coal Geology 6, 251-276.

Winchester, J.A. and Floyd, P.A., 1977. Geochemical discrimination of different magma series and their differentiation products using immobile elements. Chemical Geology 20, 325-343.

Wright, P.C., 1968. The meandu creek bentonite—a reply. Journal of the Geological Society of Australia 15, 347-350.

Xiao, L., Xu, Y.G., Mei, H.J., Zheng, Y.F., He, B. and Pirajno, F., 2004. Distinct mantle sources of low-Ti and high-Ti basalts from the western Emeishan large igneous province, SW China: implications for plume–lithosphere interaction. Earth and Planetary Science Letters 228, 525- 546.

Xu, Y.-G., He, B., Chung, S.-L., Menzies, M.A. and Frey, F.A., 2004. Geologic, geochemical, and geophysical consequences of plume involvement in the Emeishan flood-basalt province. Geology 32, 917-920.

Xu, Y., He, B., Huang, X., Luo, Z., Zhu, D., Ma, J. and Shao, H., 2007. Late Permian Emeishan Flood Basalts in Southwestern China. Earth Science Frontiers 14, 1-9.

Yang, G., Wang, S., Zhou, R., Sun, S., 1983. Endemic selenium intoxication of humans in China. The American Journal of Clinical Nutrition 37, 872-881.

Yudovich, Y.E. and Ketris, M.P., 2005a. Mercury in coal: a review: Part 1. Geochemistry. International Journal of Coal Geology 62, 107-134.

Yudovich, Y.E. and Ketris, M.P., 2005b. Mercury in coal: a review Part 2. Coal use and environmental problems. International Journal of Coal Geology 62, 135-165.

Yudovich, Y.E. and Ketris, M.P., 2006a. Selenium in coal: A review. International Journal of Coal Geology 67, 112-126.

Yudovich, Y.E. and Ketris, M.P., 2006b. Chlorine in coal: A review. International Journal of Coal Geology 67, 127-144.

330

References

Zhang, J., Ren, D., Zheng, C., Zeng, R., Chou, C.-L. and Liu, J., 2002. Trace element abundances in major minerals of Late Permian coals from southwestern Guizhou province, China. International Journal of Coal Geology 53, 55-64.

Zhang, Y., 2007. The characteristics of acid rain pollution and the causes for its formation during the Tenth Five-year Plan in Chongqing Journal of Southwest University (Natural Science Edition) 29, 164-168 (in Chinese with English abstract).

Zhang, Z., Mao, J., Wang, Y., Pirajno, F., Liu, J. and Zhao, Z., 2010. Geochemistry and geochronology of the volcanic rocks associated with the Dong'an adularia-sericite epithermal gold deposit, Lesser Hinggan Range, Heilongjiang province, NE China: Constraints on the metallogenesis. Ore Geology Reviews 37, 158-174.

Zhao, L., Ward, C.R., French, D. and Graham, I.T., 2012. Mineralogy of the volcanic-influenced Great Northern coal seam in the Sydney Basin, Australia. International Journal of Coal Geology 94, 94-110.

Zheng, L., Liu, G., Chou, C.-L., Qi, C. and Zhang, Y., 2007. Geochemistry of rare earth elements in Permian coals from the Huaibei Coalfield, China. Journal of Asian Earth Sciences 31, 167-176.

Zheng, L., Liu, G. and Chou, C.-L., 2008. Abundance and modes of occurrence of mercury in some low-sulfur coals from China. International Journal of Coal Geology 73, 19-26.

Zhou, Y., Ren, Y. and Bohor, B.F., 1982. Origin and distribution of tonsteins in late permian coal seams of Southwestern China. International Journal of Coal Geology 2, 49-77.

Zhou, Y., Tang, D. and Ren, Y., 1992. Characteristics of Zircons from volcanic ash-derivd tonsteins in late Permian coalfields of eastern Yunnan, China. Acta Sedimentologica Sinica 10, 28-38 (In Chinese with English abstract).

Zhou, Y. and Ren, Y., 1994. Element geochemistry of volcanic ash derived tonsteins in Late Permian Coal-bearing formation of eastern Yunnan and western Guizhou, China. Acta Sedimentologica Sinica 12, 123-132 (In Chinese with English abstract).

Zhou, Y., Ren, Y., Tang, D. and Bohor, B., 1994. Characteristics of zircons from volcanic ash- derived tonsteins in Late Permian coal fields of eastern Yunnan, China. International Journal of Coal Geology 25, 243-264.

Zhou, Y., Bohor, B.F. and Ren, Y., 2000. Trace element geochemistry of altered volcanic ash layers (tonsteins) in Late Permian coal-bearing formations of eastern Yunnan and western Guizhou Provinces, China. International Journal of Coal Geology 44, 305-324.

Zhou, Y.P., 1999. The alkali pyroclastic tonsteins of early stage of late Permian age in southwestern China. Coal Geology and Exploration 27, 5-9 (in Chinese with English abstract).

Zhuang, X., Querol, X., Zeng, R., Xu, W., Alastuey, A., Lopez-Soler, A. and Plana, F., 2000. Mineralogy and geochemistry of coal from the Liupanshui mining district, Guizhou, south China. International Journal of Coal Geology 45, 21-37.

Zhuang, X., Querol, X., Plana, F., Alastuey, A., Lopez-Soler, A. and Wang, H., 2003. Determination of elemental affinities by density fractionation of bulk coal samples from the Chongqing coal district, Southwestern China. International Journal of Coal Geology 55, 103-115.

331

Zhuang, X., Querol, X., Alastuey, A., Juan, R., Plana, F., Lopez-Soler, A., Du, G. and Martynov, V.V., 2006. Geochemistry and mineralogy of the Cretaceous Wulantuga high-germanium coal deposit in Shengli coal field, Inner Mongolia, Northeastern China. International Journal of Coal Geology 66, 119-136.

Zielinski, R.A., 1982. The mobility of uranium and other elements during alteration of rhyolite ash to montmorillonite: A case study in the Troublesome Formation, Colorado, U.S.A. Chemical Geology 35, 185-204.

Zielinski, R.A., 1985. Element mobility during alteration of silicic ash to kaolinite-a study of tonstein. Sedimentology 32, 567-579.

Zielinski, R.A. and Budahn, J.R., 1998. Radionuclides in fly ash and bottom ash: Improved characterization based on radiography and low energy gamma-ray spectrometry. Fuel 77, 259- 267.

Zubovic, P., 1966. Physicochemical Properties of Certain Minor Elements as Controlling Factors in their Distribution in Coal. In: R.F. Gould (Ed.), Coal Science. Advances in Chemistry. American Chemical Society, Washington, DC, pp. 221-231.

332

Appendix 1 Coefficients among ash yield, minerals and elements in the Greta coals

LTA Kao Py SiO2 TiO2 Al2O3 Fe2O3 MgO CaO Na2OK2O MnO P2O5 Li Be LTA 1.00 Qtz 0.86 Kao 0.31 1.00 Py 0.31 -0.20 1.00

SiO2 0.87 0.33 -0.17 1.00

TiO2 0.67 0.78 -0.15 0.73 1.00

Al2O3 0.74 0.74 -0.29 0.87 0.90 1.00

Fe2O3 0.32 -0.19 0.97 -0.18 -0.16 -0.29 1.00 MgO 0.71 0.09 0.19 0.62 0.25 0.47 0.21 1.00 CaO 0.49 -0.15 0.70 0.15 -0.05 -0.01 0.67 0.70 1.00

Na2O 0.82 0.26 -0.23 0.97 0.66 0.81 -0.21 0.67 0.15 1.00

K2O 0.81 0.49 -0.26 0.96 0.78 0.93 -0.26 0.62 0.11 0.93 1.00 MnO 0.16 -0.14 0.38 -0.06 -0.15 -0.07 0.42 0.20 0.23 -0.06 -0.04 1.00

P2O5 0.28 -0.03 0.74 -0.11 -0.07 -0.16 0.77 0.19 0.72 -0.19 -0.15 0.24 1.00 Li 0.52 0.77 -0.48 0.75 0.84 0.93 -0.48 0.26 -0.29 0.68 0.82 -0.18 -0.36 1.00 Be 0.16 0.21 -0.37 0.32 0.11 0.34 -0.36 0.51 0.23 0.41 0.43 -0.33 -0.17 0.27 1.00 F 0.43 -0.13 0.67 0.13 0.07 0.00 0.59 0.42 0.84 0.04 0.09 0.22 0.72 -0.24 -0.03 Sc 0.82 0.45 -0.22 0.95 0.72 0.89 -0.23 0.73 0.24 0.94 0.97 -0.12 -0.11 0.77 0.54 V 0.45 0.71 -0.31 0.56 0.75 0.80 -0.33 0.43 0.15 0.53 0.71 -0.16 -0.10 0.71 0.50 Cr 0.27 0.34 -0.24 0.38 0.22 0.42 -0.26 0.56 0.17 0.43 0.49 -0.20 -0.22 0.41 0.81 Co 0.10 -0.27 0.07 0.11 0.01 0.00 -0.01 0.32 0.40 0.14 0.11 0.05 -0.07 -0.13 0.14 Ni 0.40 0.26 -0.36 0.59 0.30 0.52 -0.35 0.54 0.13 0.65 0.64 -0.35 -0.17 0.48 0.88 Cu 0.61 0.62 -0.16 0.70 0.79 0.78 -0.20 0.33 -0.03 0.62 0.80 -0.15 -0.09 0.74 0.27 Zn 0.68 0.38 -0.25 0.82 0.67 0.79 -0.25 0.53 0.20 0.78 0.87 -0.14 0.02 0.63 0.43 Ga 0.64 0.58 -0.14 0.68 0.77 0.79 -0.16 0.51 0.26 0.65 0.81 0.03 0.11 0.62 0.38 Ge -0.34 -0.09 -0.42 -0.16 -0.28 -0.12 -0.38 -0.06 -0.24 -0.01 -0.08 -0.19 -0.44 -0.06 0.68 As 0.28 -0.18 0.87 -0.11 -0.11 -0.22 0.80 0.16 0.67 -0.19 -0.19 0.31 0.75 -0.38 -0.25 Se -0.20 0.01 0.15 -0.27 -0.07 -0.18 0.12 -0.15 -0.14 -0.31 -0.27 0.09 -0.08 -0.01 -0.42 Rb 0.81 0.47 -0.26 0.96 0.76 0.91 -0.27 0.64 0.13 0.92 0.99 -0.05 -0.14 0.81 0.45 Y 0.42 0.36 -0.22 0.51 0.48 0.59 -0.26 0.61 0.43 0.52 0.61 -0.07 0.02 0.44 0.61 Zr 0.63 0.37 -0.35 0.82 0.73 0.83 -0.37 0.45 -0.05 0.82 0.86 -0.02 -0.41 0.75 0.28 Nb 0.63 0.52 -0.05 0.70 0.67 0.73 -0.14 0.46 0.22 0.60 0.75 -0.08 0.06 0.67 0.39 Mo 0.53 0.19 0.27 0.42 0.57 0.39 0.23 0.24 0.32 0.38 0.42 0.07 0.26 0.24 -0.09 Ag 0.71 0.32 -0.06 0.75 0.65 0.75 -0.09 0.56 0.35 0.69 0.78 0.12 0.08 0.56 0.23 Cd 0.85 0.27 0.06 0.84 0.62 0.75 0.06 0.56 0.24 0.80 0.82 0.28 0.11 0.59 0.06 Sn 0.71 0.44 -0.11 0.82 0.81 0.78 -0.18 0.47 0.11 0.76 0.83 -0.11 -0.21 0.70 0.24 Sb 0.38 0.74 -0.36 0.50 0.73 0.76 -0.35 0.28 -0.01 0.45 0.69 0.06 -0.02 0.70 0.39 Cs 0.70 0.56 -0.20 0.80 0.70 0.83 -0.22 0.72 0.29 0.77 0.88 -0.12 -0.08 0.73 0.63 Ba 0.34 0.05 0.23 0.22 0.24 0.20 0.20 0.38 0.55 0.19 0.28 0.16 0.46 -0.05 0.08 W -0.18 -0.20 -0.13 -0.09 0.05 -0.09 -0.20 -0.11 0.02 -0.06 -0.08 0.10 -0.23 -0.12 -0.08 Tl 0.36 -0.17 0.91 -0.11 -0.10 -0.24 0.95 0.25 0.65 -0.14 -0.21 0.30 0.75 -0.39 -0.35 Hg 0.19 -0.38 0.89 -0.24 -0.25 -0.39 0.86 0.24 0.72 -0.25 -0.34 0.32 0.60 -0.59 -0.35 Pb 0.46 0.67 -0.42 0.66 0.74 0.81 -0.42 0.28 -0.24 0.58 0.77 0.03 -0.33 0.87 0.21 Bi 0.21 0.02 0.16 0.25 0.31 0.11 -0.01 0.05 0.09 0.14 0.19 -0.41 0.04 0.14 -0.14 Th 0.51 0.54 -0.32 0.68 0.63 0.71 -0.35 0.47 -0.01 0.66 0.73 -0.21 -0.28 0.73 0.50 U 0.70 0.48 -0.04 0.75 0.66 0.71 -0.08 0.60 0.27 0.71 0.77 -0.09 0.05 0.64 0.43 La 0.62 0.37 -0.33 0.82 0.73 0.77 -0.37 0.43 -0.01 0.83 0.87 -0.12 -0.23 0.67 0.38

Appendix 1 (Continued)

F Sc V Cr Co Ni Cu Zn Ga Ge As Se Rb Y F 1.00 Sc 0.14 1.00 V 0.22 0.71 1.00 Cr -0.11 0.61 0.45 1.00 Co 0.57 0.15 0.38 -0.04 1.00 Ni -0.10 0.74 0.44 0.86 -0.09 1.00 Cu 0.10 0.73 0.62 0.44 0.02 0.49 1.00 Zn 0.29 0.82 0.74 0.24 0.32 0.51 0.69 1.00 Ga 0.37 0.76 0.85 0.31 0.33 0.42 0.79 0.89 1.00 Ge -0.47 -0.02 -0.01 0.52 -0.13 0.54 -0.19 -0.16 -0.21 1.00 As 0.82 -0.15 -0.16 -0.23 0.22 -0.25 -0.06 -0.10 0.01 -0.41 1.00 Se 0.07 -0.30 -0.14 -0.26 0.11 -0.42 -0.03 -0.29 -0.19 -0.35 0.29 1.00 Rb 0.09 0.98 0.69 0.54 0.10 0.68 0.81 0.85 0.79 -0.07 -0.19 -0.28 1.00 Y 0.45 0.68 0.87 0.45 0.62 0.48 0.43 0.71 0.78 0.02 0.00 -0.19 0.61 1.00 Zr 0.06 0.80 0.73 0.27 0.45 0.38 0.66 0.79 0.72 -0.08 -0.25 -0.09 0.83 0.64 Nb 0.30 0.78 0.64 0.62 0.06 0.61 0.77 0.54 0.64 -0.09 0.14 -0.14 0.78 0.58 Mo 0.61 0.35 0.44 -0.17 0.48 -0.04 0.49 0.53 0.64 -0.42 0.46 0.26 0.39 0.46 Ag 0.51 0.74 0.78 0.12 0.57 0.25 0.55 0.86 0.83 -0.32 0.11 -0.14 0.75 0.81 Cd 0.34 0.75 0.47 0.07 0.20 0.28 0.62 0.77 0.70 -0.38 0.20 -0.03 0.79 0.45 Sn 0.22 0.82 0.68 0.42 0.34 0.43 0.80 0.67 0.73 -0.21 -0.06 -0.08 0.83 0.59 Sb 0.09 0.62 0.84 0.42 0.07 0.43 0.73 0.66 0.87 -0.02 -0.21 -0.17 0.69 0.69 Cs 0.18 0.93 0.80 0.77 0.19 0.75 0.74 0.71 0.76 0.05 -0.13 -0.25 0.90 0.76 Ba 0.73 0.25 0.44 -0.17 0.60 -0.05 0.31 0.63 0.70 -0.43 0.36 -0.08 0.26 0.60 W 0.29 -0.09 0.28 -0.21 0.78 -0.28 -0.12 0.03 0.16 -0.02 0.03 0.14 -0.10 0.44 Tl 0.60 -0.17 -0.30 -0.25 0.01 -0.31 -0.14 -0.20 -0.13 -0.45 0.80 0.30 -0.21 -0.22 Hg 0.73 -0.28 -0.25 -0.35 0.43 -0.48 -0.31 -0.21 -0.14 -0.44 0.77 0.26 -0.34 -0.06 Pb -0.18 0.71 0.62 0.53 -0.14 0.46 0.80 0.49 0.62 -0.08 -0.31 0.04 0.78 0.39 Bi 0.33 0.22 0.10 0.09 0.23 0.05 0.49 0.16 0.22 -0.46 0.20 0.27 0.24 0.12 Th -0.12 0.79 0.56 0.79 -0.11 0.73 0.70 0.41 0.50 0.14 -0.26 -0.16 0.77 0.45 U 0.17 0.84 0.55 0.71 -0.05 0.69 0.72 0.50 0.60 -0.05 0.00 -0.23 0.82 0.52 La 0.08 0.83 0.63 0.39 0.21 0.58 0.82 0.80 0.82 -0.04 -0.19 -0.22 0.87 0.59 Appendix 1 (Continued) Zr Nb Mo Ag Cd Sn Sb Cs Ba W Tl Hg Pb Bi Th U La Zr 1.00

Nb 0.54 1.00

Mo 0.50 0.37 1.00

Ag 0.84 0.58 0.65 1.00

Cd 0.75 0.55 0.62 0.80 1.00

Sn 0.83 0.77 0.55 0.72 0.64 1.00

Sb 0.57 0.67 0.40 0.62 0.46 0.60 1.00 Cs 0.71 0.84 0.31 0.68 0.57 0.82 0.73 1.00 Ba 0.30 0.15 0.63 0.65 0.43 0.29 0.40 0.24 1.00 W 0.31 -0.02 0.44 0.35 -0.03 0.26 0.14 -0.01 0.32 1.00 Tl -0.32 -0.10 0.32 -0.06 0.12 -0.13 -0.34 -0.17 0.19 -0.20 1.00 Hg -0.24 -0.25 0.33 0.04 0.00 -0.11 -0.44 -0.25 0.39 0.18 0.84 1.00 Pb 0.63 0.77 0.23 0.45 0.52 0.72 0.77 0.76 -0.03 -0.07 -0.37 -0.55 1.00 Bi 0.16 0.45 0.43 0.16 0.10 0.54 0.09 0.27 0.22 0.18 0.12 0.14 0.29 1.00 Th 0.53 0.84 0.10 0.33 0.36 0.77 0.60 0.86 -0.13 -0.11 -0.29 -0.45 0.83 0.38 1.00 U 0.49 0.90 0.27 0.47 0.50 0.79 0.61 0.89 0.08 -0.13 -0.02 -0.22 0.76 0.43 0.93 1.00 La 0.79 0.67 0.56 0.67 0.72 0.84 0.69 0.75 0.39 0.15 -0.31 -0.36 0.70 0.36 0.68 0.70 1.00

Appendix 2 Coefficients among ash yield, minerals and elements in the Songzao coals Py

Ash Kao SiO2 TiO2 Al2O3 Fe2O3 MgO CaO Na2OK2O MnO P2O5 Li Be +Mar

Ash 1 Kao 0.52 1 Py+Mar -0.35 -0.77 1

SiO2 0.84 0.53 -0.52 1

TiO2 0.6 0.61 -0.53 0.62 1

Al2O3 0.68 0.93 -0.77 0.67 0.79 1

Fe2O3 -0.25 -0.74 0.95 -0.52 -0.54 -0.73 1 MgO 0.46 0.05 -0.07 0.05 -0.01 0.15 0.13 1 CaO 0.32 0.1 -0.2 -0.01 -0.12 0.11 -0.01 0.8 1

Na2O 0.68 0.74 -0.66 0.58 0.82 0.9 -0.6 0.26 0.14 1

K2O 0.67 0.74 -0.75 0.63 0.72 0.86 -0.68 0.24 0.28 0.89 1 MnO 0.43 0 -0.07 0.09 -0.15 0.07 0.11 0.96 0.86 0.13 0.15 1

P2O5 0.47 0.44 -0.39 0.29 0.38 0.53 -0.37 0.63 0.52 0.54 0.4 0.62 1 Li 0.6 0.83 -0.79 0.57 0.85 0.92 -0.73 0.21 0.23 0.86 0.85 0.12 0.6 1

Be 0.14 -0.32 0.38 0.37 -0.07 -0.28 0.27 -0.37 -0.53 -0.32 -0.34 -0.27 -0.42 -0.44 1.00 F 0.76 0.64 -0.51 0.61 0.54 0.8 -0.44 0.5 0.28 0.83 0.72 0.44 0.63 0.66 -0.19 Sc 0.71 0.5 -0.6 0.66 0.86 0.73 -0.56 0.34 0.25 0.85 0.81 0.26 0.57 0.83 -0.18 V 0.46 0.36 -0.51 0.37 0.79 0.59 -0.46 0.3 0.2 0.8 0.72 0.17 0.48 0.75 -0.33 Cr 0.44 0.42 -0.55 0.42 0.87 0.63 -0.53 0.19 0.13 0.79 0.76 0.06 0.44 0.79 -0.29 Co -0.2 -0.31 0.35 -0.23 -0.14 -0.22 0.4 0.01 0.05 -0.25 -0.09 0.02 -0.17 -0.19 -0.07 Ni 0.3 0.01 -0.17 0.19 0.14 0.11 -0.06 0.31 0.62 0.17 0.48 0.34 0.14 0.28 -0.32 Cu 0.31 0.21 -0.35 0.29 0.66 0.47 -0.34 0.17 0.04 0.68 0.62 0.06 0.34 0.57 -0.24 Zn -0.09 0.22 -0.32 -0.04 0.26 0.18 -0.27 0 -0.11 0.21 0.32 -0.13 -0.13 0.18 -0.06 Ga 0.73 0.9 -0.66 0.69 0.72 0.97 -0.61 0.14 0.06 0.86 0.83 0.06 0.44 0.84 -0.21 Ge 0.06 -0.11 0.43 0.19 -0.07 -0.12 0.35 -0.46 -0.6 -0.28 -0.36 -0.43 -0.45 -0.33 0.64 As -0.36 -0.71 0.89 -0.57 -0.54 -0.73 0.91 0.12 -0.07 -0.66 -0.66 0.07 -0.4 -0.73 0.26 Se -0.36 -0.49 0.43 -0.62 -0.39 -0.52 0.54 0.27 0.3 -0.35 -0.19 0.17 -0.4 -0.41 -0.23 Hg -0.34 -0.56 0.32 -0.27 -0.3 -0.5 0.39 -0.16 -0.23 -0.33 -0.15 -0.2 -0.7 -0.48 0.25 Rb 0.6 0.7 -0.77 0.58 0.71 0.82 -0.7 0.23 0.33 0.85 0.99 0.15 0.4 0.85 -0.38 Cs 0.67 0.76 -0.78 0.64 0.66 0.86 -0.7 0.27 0.35 0.85 0.99 0.22 0.42 0.83 -0.34 Sr 0.57 0.46 -0.48 0.27 0.35 0.54 -0.33 0.79 0.77 0.59 0.58 0.75 0.84 0.66 -0.61 Y 0.6 0.36 -0.36 0.34 0.13 0.46 -0.24 0.86 0.7 0.47 0.48 0.86 0.73 0.39 -0.29 Zr 0.6 0.54 -0.39 0.47 0.11 0.52 -0.33 0.59 0.54 0.34 0.35 0.65 0.65 0.36 -0.09 Nb 0.66 0.78 -0.46 0.56 0.31 0.75 -0.37 0.25 0.13 0.58 0.53 0.23 0.34 0.49 -0.07 Mo 0.04 -0.39 0.66 -0.25 -0.36 -0.4 0.7 0.43 0.19 -0.43 -0.43 0.39 -0.07 -0.42 0.19 Ag 0.68 0.77 -0.5 0.59 0.3 0.76 -0.41 0.28 0.16 0.62 0.59 0.26 0.35 0.5 -0.08 Cd 0.04 0.24 -0.37 -0.09 0.09 0.22 -0.21 0.38 0.26 0.27 0.38 0.26 0 0.24 -0.33 Sn 0.75 0.69 -0.54 0.59 0.26 0.72 -0.42 0.56 0.47 0.63 0.64 0.56 0.55 0.54 -0.19 Sb 0 0.09 -0.3 -0.14 -0.04 0.03 -0.12 0.31 0.52 0.05 0.31 0.27 -0.07 0.22 -0.49 Ba 0.61 0.76 -0.79 0.61 0.83 0.87 -0.74 0.14 0.2 0.89 0.96 0.04 0.43 0.92 -0.35 Hf 0.63 0.65 -0.46 0.54 0.19 0.63 -0.41 0.47 0.4 0.44 0.42 0.52 0.59 0.42 -0.05 Ta 0.58 0.79 -0.46 0.57 0.3 0.73 -0.39 0.05 -0.04 0.53 0.51 0.03 0.18 0.46 0.00 W 0.68 0.66 -0.51 0.56 0.63 0.74 -0.39 0.31 0.07 0.68 0.59 0.19 0.34 0.73 -0.14 Tl -0.16 -0.58 0.74 -0.06 -0.28 -0.54 0.67 -0.38 -0.47 -0.58 -0.56 -0.34 -0.59 -0.63 0.68 Pb -0.22 0.09 -0.06 -0.29 0.02 0.04 0 0.02 -0.16 0.16 0.21 -0.13 -0.23 -0.03 -0.12 Bi 0.24 0.3 -0.58 0.14 0.39 0.39 -0.42 0.33 0.39 0.5 0.56 0.24 0.32 0.63 -0.60 Th 0.64 0.87 -0.75 0.68 0.54 0.91 -0.73 0.18 0.15 0.77 0.76 0.18 0.5 0.73 -0.17 U 0.69 0.65 -0.54 0.64 0.26 0.69 -0.5 0.43 0.39 0.55 0.55 0.5 0.62 0.49 -0.05 La 0.68 0.48 -0.44 0.49 0.36 0.59 -0.38 0.7 0.62 0.57 0.52 0.72 0.82 0.54 -0.20

Appendix2(Continued)

F Sc V Cr Co Ni Cu Zn Ga Ge As Se Hg Rb Cs Sr Y

F 1.00

Sc 0.71 1.00

V 0.59 0.93 1.00

Cr 0.50 0.92 0.96 1.00

Co -0.16 -0.13 -0.03 -0.02 1.00

Ni 0.04 0.37 0.32 0.37 0.39 1.00

Cu 0.56 0.81 0.92 0.86 0.22 0.28 1.00

Zn -0.07 0.15 0.20 0.36 -0.05 0.03 0.11 1.00

Ga 0.83 0.66 0.50 0.52 -0.19 0.08 0.42 0.08 1.00

Ge -0.05 -0.33 -0.43 -0.44 0.14 -0.47 -0.26 -0.40 0.06 1.00

As -0.49 -0.58 -0.48 -0.49 0.44 -0.09 -0.35 -0.02 -0.62 0.31 1.00

Se -0.43 -0.31 -0.13 -0.11 0.35 0.35 -0.11 0.40 -0.50 -0.32 0.66 1.00

Hg -0.44 -0.26 -0.12 -0.09 0.43 0.20 0.04 0.42 -0.46 -0.01 0.47 0.66 1.00

Rb 0.63 0.81 0.73 0.79 -0.09 0.55 0.62 0.36 0.75 -0.47 -0.67 -0.13 -0.13 1.00

Cs 0.72 0.78 0.66 0.71 -0.10 0.50 0.56 0.30 0.82 -0.39 -0.69 -0.19 -0.19 0.98 1.00

Sr 0.66 0.64 0.57 0.50 -0.16 0.43 0.38 -0.06 0.50 -0.55 -0.36 -0.09 -0.50 0.58 0.60 1.00

Y 0.78 0.49 0.36 0.27 -0.07 0.23 0.29 -0.03 0.45 -0.36 -0.23 -0.08 -0.35 0.45 0.54 0.80 1.00

Zr 0.64 0.27 0.02 0.01 -0.24 0.03 -0.11 -0.08 0.51 -0.12 -0.33 -0.34 -0.60 0.31 0.44 0.58 0.82

Nb 0.75 0.26 0.05 0.03 -0.30 -0.15 -0.03 -0.02 0.83 0.20 -0.42 -0.45 -0.46 0.41 0.55 0.38 0.53

Mo -0.09 -0.30 -0.33 -0.37 0.21 -0.09 -0.28 -0.19 -0.26 0.32 0.78 0.43 0.08 -0.46 -0.41 0.01 0.20

Ag 0.80 0.31 0.11 0.08 -0.27 -0.09 0.06 -0.03 0.84 0.17 -0.45 -0.44 -0.42 0.48 0.62 0.42 0.58

Cd 0.12 0.17 0.23 0.27 -0.07 0.14 0.11 0.78 0.15 -0.58 -0.07 0.50 0.39 0.42 0.40 0.22 0.30

Sn 0.86 0.44 0.23 0.17 -0.25 0.12 0.13 -0.04 0.76 -0.10 -0.46 -0.33 -0.44 0.56 0.70 0.65 0.83

Sb -0.19 0.09 0.12 0.17 0.02 0.59 -0.06 0.42 -0.04 -0.67 -0.09 0.57 0.33 0.40 0.34 0.30 0.09

Ba 0.61 0.84 0.76 0.84 -0.18 0.43 0.61 0.37 0.80 -0.40 -0.71 -0.25 -0.23 0.97 0.93 0.56 0.34

Hf 0.71 0.29 0.04 0.03 -0.28 -0.06 -0.07 -0.07 0.63 -0.02 -0.42 -0.44 -0.61 0.37 0.51 0.51 0.77

Ta 0.66 0.20 -0.01 -0.01 -0.25 -0.17 -0.05 -0.02 0.83 0.34 -0.42 -0.49 -0.39 0.39 0.53 0.24 0.36

W 0.63 0.59 0.47 0.45 -0.34 -0.06 0.29 0.15 0.76 -0.01 -0.42 -0.31 -0.28 0.53 0.57 0.48 0.37

Tl -0.40 -0.45 -0.48 -0.46 0.47 -0.16 -0.26 -0.30 -0.39 0.81 0.67 0.11 0.37 -0.61 -0.59 -0.67 -0.47

Pb -0.03 -0.05 0.10 0.17 0.07 -0.06 0.14 0.84 0.01 -0.30 0.19 0.52 0.54 0.21 0.16 -0.13 -0.04

Bi 0.21 0.55 0.60 0.58 -0.14 0.48 0.40 0.22 0.30 -0.66 -0.47 0.08 0.04 0.61 0.53 0.59 0.20

Th 0.84 0.57 0.40 0.40 -0.25 0.00 0.33 0.06 0.90 -0.05 -0.74 -0.57 -0.51 0.70 0.80 0.47 0.60

U 0.76 0.41 0.17 0.14 -0.24 0.07 0.08 -0.07 0.67 -0.12 -0.56 -0.50 -0.53 0.50 0.63 0.53 0.77

La 0.77 0.61 0.44 0.39 -0.16 0.20 0.31 -0.06 0.53 -0.34 -0.41 -0.30 -0.57 0.51 0.59 0.75 0.91

Appendix2(Continued)

Zr Nb Mo Ag Cd Sn Sb Ba Hf Ta W Tl Pb Bi Th U La

Zr 1.00

Nb 0.71 1.00

Mo 0.15 -0.07 1.00

Ag 0.70 0.99 -0.09 1.00

Cd 0.14 0.15 -0.03 0.15 1.00

Sn 0.86 0.89 -0.02 0.92 0.26 1.00

Sb -0.03 -0.10 -0.07 -0.10 0.72 0.06 1.00

Ba 0.26 0.43 -0.52 0.47 0.34 0.50 0.30 1.00

Hf 0.98 0.83 0.04 0.82 0.13 0.90 -0.11 0.33 1.00

Ta 0.57 0.97 -0.12 0.96 0.07 0.79 -0.15 0.42 0.71 1.00

W 0.37 0.65 -0.01 0.63 0.39 0.60 0.23 0.60 0.44 0.61 1.00

Tl -0.37 -0.29 0.48 -0.31 -0.52 -0.46 -0.47 -0.60 -0.36 -0.18 -0.38 1.00

Pb -0.22 -0.01 -0.09 0.01 0.71 -0.05 0.29 0.18 -0.18 0.01 0.01 -0.22 1.00

Bi -0.06 0.04 -0.38 0.06 0.50 0.19 0.72 0.62 -0.08 -0.01 0.53 -0.67 0.09 1.00

Th 0.71 0.83 -0.34 0.86 0.18 0.86 -0.08 0.69 0.82 0.80 0.60 -0.49 -0.04 0.17 1.00

U 0.92 0.80 -0.16 0.81 0.13 0.92 -0.07 0.45 0.96 0.68 0.43 -0.43 -0.17 0.04 0.87 1.00

La 0.88 0.52 0.05 0.54 0.16 0.78 -0.01 0.45 0.84 0.34 0.39 -0.49 -0.20 0.14 0.70 0.85 1.00

Appendix 3

List of publications

The following peer reviewed journal paper, covering the scope of this dissertation has been published:

The following papers or abstracts relevant to this dissertation have also been presented at conferences during my research candidature:

Zhao, L. 2009. Mineralogy and geochemistry of coals from the Songzao coalfield, Southwestern China (Oral). Proceedings of 26th Annual Meeting of the Society for Organic Petrology. Gramado/Porto Alegre, Brazil

Zhao, L. 2010. Mineralogy and geochemistry of individual coal seam subsections in the Sydney Basin, Australia (Oral). Proceedings of 27th Annual Meeting of the Society for Organic Petrology. Denver, Colorado, USA.

Zhao, L. 2010. Comparative study of the mineralogy and geochemistry of the Bulli and Great Northern coal seams, Sydney Basin (Oral). Proceedings of 37th symposium on the Geology of the Sydney Basin, Hunter Valley, NSW, Australia.

Zhao, L. 2011. Comparative Mineralogy of Volcanic Influenced Coal Seams in the Songzao Coalfield, China and the Sydney Basin, Australia (Poster). XVII International Congress on the Carboniferous and Permian, Perth, Australia.

Zhao, L. 2012. Mineralogy and geochemistry of the marine-influenced Greta coal seam, Sydney Basin (Oral). Proceedings of 38th symposium on the Geology of the Sydney Basin, Hunter Valley, NSW. Australia.

Zhao L., Ward, C., French D. and Graham, I. 2012. Mineralogy and geochemistry of contrasting sulphur-rich coals from the Songzao Coalfield, SW China and the Greta Coal Measures, Hunter Valley, Australia (Poster). 34th International Geological Congress, Brisbane, Australia.

Zhao L., Ward, C., French D. and Graham, I. 2012. Mineralogy and Geochemistry of Sulphur-rich Coals from the Songzao Coalfield, SW China (Oral). Proceedings of 29th Annual Meeting of the Society for Organic Petrology. Beijing, China. APPENDIX 4:

Reprint of “Mineralogy of the volcanic-influenced Great Northern coal seam in the Sydney Basin, Australia” – a paper published on

International Journal of Coal Geology, Vol. 113, 2012. International Journal of Coal Geology 113 (2012) 94–110

Contents lists available at SciVerse ScienceDirect

International Journal of Coal Geology

journal homepage: www.elsevier.com/locate/ijcoalgeo

Mineralogy of the volcanic-inﬂuenced Great Northern coal seam in the Sydney Basin, Australia

Lei Zhao a,⁎, Colin R. Ward a, David French b, Ian T. Graham a a School of Biological, Earth and Environmental Sciences, University of New South Wales, Sydney 2052, Australia b CSIRO Energy Technology, PO Box 52, North Ryde 1670, Australia article info abstract

Article history: The mineralogy of the individual coal plies and intra-seam claystone bands in the Great Northern seam of the Received 26 July 2011 Sydney Basin has been evaluated using optical and scanning electron microscopy, and quantitative X-ray Received in revised form 20 September 2011 diffraction techniques. The uppermost two claystone bands are tonsteins, consisting mainly of well-ordered Accepted 20 September 2011 kaolinite with graupen to vermicular textures. Idiomorphic crystals of K-feldspar within these tonsteins may Available online 2 October 2011 represent members of the anorthoclase–sanidine series or a sodic sanidine, and indicate an acid to intermediate volcanic ash input. In contrast, the lowermost parting was largely derived from epiclastic sediment, Keywords: Mineral matter admixed with minor volcanic material such as high-temperature quartz and a different type of K-feldspar Inorganic geochemistry component. Tonsteins The mineral fraction of the coals, especially in the middle and upper parts of the seam, is dominated by authi- Permian genic kaolinite with a very low abundance of quartz. Apart from the tonsteins, authigenic processes therefore Australia appear to be the dominant mechanism of mineral matter formation. Authigenic K-feldspar also occurs in the lower few metres of the seam, with a variety of modes of occurrence including cell and cleat infillings, cross- cutting veins, and thin laminae parallel to the organic matter and detrital clay bands. A late syngenetic low- temperature hydrothermal fluid injection process is suggested for formation of the feldspar veins. The origin of the fluid is uncertain, but is most likely associated with contemporaneous volcanic activity. © 2011 Elsevier B.V. All rights reserved.

1. Introduction 1995; Bocking et al., 1988; Branagan and Johnson, 1970; Diessel and Hutton, 2004; Herbert, 1980). However, the upper part of the New- Minerals in coal provide information about the depositional condi- castle Coal Measures (including the Moon Island Beach Subgroup) tions and thus the geologic history of coal-bearing sequences, along also contains a number of tuffaceous units (Fig. 2), representing prod- with the regional sedimentary and tectonic history (Ren, 1996; ucts of contemporaneous volcanism associated with the New England Ward, 2002). Knowledge of the mineral matter in coal is important Orogen, transported into the basin by ash fall and ash flow processes in understanding both the inorganic processes associated with coal (Agnew et al., 1995; Diessel, 1965, 1985; Grevenitz, 2003; Kramer et formation (Ward et al., 2001) and aspects such as materials handling, al., 2001; Loughnan and Ray, 1978). The coal seams of the upper New- boiler erosion, ash formation, and slagging in coal processing or utili- castle Coal Measures were thus deposited under conditions that were sation (Gupta et al., 1999; Ward, 1984). influenced by two different types of non-coal sediment input, and The Late Permian Newcastle Coal Measures is one of several strat- these, along with authigenic processes associated with peat accumula- igraphic units in the Sydney Basin of eastern Australia (Figs. 1, 2) from tion, may both have impacted on the mineral matter in the individual which significant resources of bituminous coals are currently coal beds. extracted. As discussed more fully by Agnew et al. (1995), the upper- The Great Northern seam is one of the principal economic coal most part of the sequence, the Moon Island Beach Subgroup (Fig. 2), seams in the Moon Island Beach Subgroup. It is exposed in coastal includes a number of coarse pebble conglomerates, with associated outcrops at Catherine Hill Bay, 30 km SSW of Newcastle (Fig. 1), and sandstones and shales, derived from erosion of older strata immedi- is mined from a number of underground collieries in the area to the ately to the north, in what was then the tectonically active orogen north and west. The seam is underlain by the Awaba Tuff, an extensive of the New England Fold Belt (Fig. 1), and deposited by alluvial fan tuffaceous unit which commonly shows cross-stratification (Kramer et and braided river systems in a foreland basin setting (Agnew et al., al., 2001) and may in part represent ash flow or reworked ash fall material. It is partly interbedded with alluvial channel deposits, and overlain mainly by the Teralba Conglomerate, a sequence of sandstone and con- ⁎ Corresponding author. Tel.: +61 430166280; fax: +61 93851558. glomerate also formed by fluvial channel processes (Agnew et al., E-mail address: [email protected] (L. Zhao). 1995). Another volcanic unit, the Booragul Tuff, also overlies the Great

Fig. 1. Locations of Great Northern seam sections at Newvale (A) and Catherine Hill Bay (B) in the northern Sydney Basin.

Northern seam in the eastern part of the coalfield, occurring in places illustrated in Fig. 3, which also shows the variation in abundance of between the coal and the Teralba Conglomerate interval. dull and bright lithotypes in the individual coal plies. The contact be- Like other seams in the upper Newcastle Coal Measures tween the conglomerate and the underlying coal seam at Catherine (Warbrooke, 1987), the coal of the Great Northern seam is relatively Hill Bay is disconformable, and in one part of the outcrop a thick layer rich in inertinite. Agnew et al. (1995) indicated around 43% vitrinite, of mudstone overlain by a thin bed of coal occurs between the conglom- 53% inertinite and 4% liptinite for typical Great Northern seam products erate roof and the main part of the coal seam. This section is illustrated on a mineral-free basis. Coal from the seam is used as a fuel source in separately (Fig. 3). A tentative correlation between the two sections is both domestic and international markets. also indicated in Fig. 3. The Great Northern seam at Newvale is thinner This study aims to investigate the characteristics and vertical than that exposed at Catherine Hill Bay; correlation of the two sections variation in the mineral matter of the Great Northern seam and asso- suggests that this is probably due to contemporaneous erosion of the ciated strata, including the immediate roof and floor rocks and a num- upper part of the coal bed by processes associated with deposition of ber of thin intra-seam claystone bands, as well as the different the Teralba Conglomerate. Although the organic matter of the coal is individual layers of coal (coal plies) within the seam section. Such a affected by weathering, the exposure at Catherine Hill Bay allows a study provides an opportunity to assess the mineralogical characteris- more extensive study of the sediments deposited in the area before tics of both the coal and non-coal strata in a major and economically- and after the peat swamp was established. important coal seam, with an emphasis on evaluating the role of the The samples were ground to less than 200 mesh using either a different inorganic processes and sedimentary inputs that may be asso- zirconia mill (Newvale) or a ceramic mortar and pestle (Catherine ciated with coal formation. Hill Bay), and split into representative portions for analysis. Use of the zirconia mill was intended to facilitate study of the trace element 2. Sampling and methodology components, but discussion of these is outside the scope of the present paper. All the coal samples were subjected to low-temperature A series of 13 coal and associated non-coal rock samples from the oxygen-plasma ashing, using techniques described by Standards Great Northern seam at Newvale No. 1 Colliery, located approximately Australia (2000), and the percentage of low-temperature ash (LTA) 35 km SSW of Newcastle (Fig. 1), were supplied for this study from a was determined in each case. Other samples of coal from Newvale sample bank maintained by CSIRO Energy Technology. An additional were ashed at 815 °C; these ashes, and also samples of the powdered set of 18 samples, including coal, claystone partings and roof and floor non-coal rocks from both locations, were fused into borosilicate disks materials, was taken from an outcrop of the Great Northern seam at (Norrish and Hutton, 1969) and analysed by X-ray fluorescence (XRF) Catherine Hill Bay (Fig. 1). The seam section at each location is spectrometry, using a Philips PW2400 XRF system, to determine the 96 L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110

Hawkesbury Sandstone Triassic Narrabeen Group

Moon Island Beach Subgroup

Newcastle Boolaroo Coal Subgroup Measures Wallarah seam Adamstown Subgroup Mannering Park Tuff

Lambton Subgroup Teralba

Permian Tomago Coal Measures Booragul Tuff Mulbring Siltstone Conglomerate

Maitland Muree Great Northern seam Group Sandstone Awaba Branxton Tuff Formation Bolton Point Greta Coal Measures Conglomerate

Dalwood Group Fassifern seam

Fig. 2. Stratigraphic units in the Newcastle Coalfield of the Sydney Basin (left) and the Moon Island Beach Subgroup of the Newcastle Coal Measures (right) (after Agnew et al., 1995). concentrations of major and minor elements present. Because of 21380 and 21390) immediately below the roof and above the floor of changes expected with weathering of the coal in the outcrop, only the the seam. All the coals are low in total sulphur (b0.4%), with the sulphur non-coal strata from Catherine Hill Bay were subjected to XRF analysis. being mainly organic in form. The mean random vitrinite reflectance is X-ray diffraction analysis was carried-out on both the coal LTAs typically around 0.82%. and the non-coal samples using a Philips PW 1830 diffractometer sys- Table 2 lists the maceral analysis results for the individual coal tem with Cu-Kα radiation and a graphite monochrometer. The XRD samples, except for the high-ash coal of sample 21380. On a miner- pattern was recorded over a 2θ interval of 2–60°, with a step size of al-free basis, the individual plies have 44–88% vitrinite, 11–55% iner- 0.04° and a count time of 2 s per step. The XRD patterns of each ash tinite (mainly semifusinite) and up to 5% liptinite. The coal plies in were analysed using the Rietveld-based Siroquant™ software pack- the upper part of the seam, especially samples 21383 and 21385, age, described by Taylor (1991), to obtain their quantitative mineral have higher inertinite percentages than those in the lower part, pos- proportions (cf. Ward et al., 2001). Oriented aggregates of the clay sibly reflecting greater exposure and oxidation of the peat during (b2 μm) fraction of the coal LTAs and the non-coal rocks were also seam formation. analysed by XRD, as described by Ruan and Ward (2002), to obtain a better understanding of the clay mineralogy, especially the nature 3.2. Mineralogical and chemical analysis data of the expandable clay minerals. Polished sections of selected coal and parting samples were exam- Quantitative mineralogical results from XRD analysis and Siroquant ined using a Zeiss Axioplan microscope to provide data on maceral interpretation for the coal LTAs and for the non-coal rocks in the abundance and vitrinite reflectance. The sections were also studied Newvale and Catherine Hill Bay sections are given in Tables 3 and 4, using a Hitachi 3400 scanning electron microscope equipped with respectively. Major element geochemical data, based on XRF analysis, an energy-dispersive X-ray spectrometer (SEM-EDS), to identify the are given in Tables 5 and 6, and oriented-aggregate XRD data in Table 7. modes of occurrence of the different mineral components within The relation between ash chemistry and mineralogy for the the coal and non-coal strata, as well as any variations in mineral com- Newvale samples was studied to check the reliability of the quantita- position. Thin sections of some non-coal samples were examined tive XRD data, following the procedure described by Ward et al. using a Leica DM 2500P optical microscope, mainly to investigate (1999). The inferred chemical composition of the mineral assem- the texture of the floor strata from the Catherine Hill Bay exposure. blages determined by Siroquant was calculated and compared with the actual chemical composition of the same samples as determined 3. Results and discussion by XRF. The inferred chemical composition was adjusted for each + LTA sample by deducting the CO2 and H2O to derive an equivalent 3.1. Coal characteristics to a coal ash analysis. The actual chemical composition determined

by XRF was also normalised to an SO3-free basis, to allow for differ- A summary of the properties of each sample from Newvale Colliery ences in SO3 retention by the low- and high-temperature ashing is given in Table 1.AccordingtoASTMD388-05(ASTM, 2007), the coals processes. are high volatile A bituminous in rank. The individual coal plies mainly The percentages of each element indicated by both sets of data have low ash yields (b12%), except for two high-ash plies (samples were plotted against each other (Fig. 4), to provide a basis for L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110 97

Fig. 3. Lithologic column sections of the Great Northern seam at Newvale colliery (A) and Catherine Hill Bay (B). Samples were taken from the CSIRO sample bank and the outcrop, respectively.

comparing the XRD results to the chemical analysis data for the same and the K2O inferred from Siroquant is based mainly on the illite coal or parting samples. As discussed for other materials by Ward et and/or I/S content. The proportion of K2O inferred from the XRD data al. (1999), the respective data sets are presented as X-Y plots, with is based on the assumption that the illite is fully saturated with K+ a diagonal line on each plot indicating where the points would fall if ions; if this is not the case (as would be expected for sedimentary mate- the estimates from the two different techniques were equal. rials), the proportion of K2O indicated by Siroquant will be less than that The plots for SiO2,Al2O3,CaOandMgOinFig. 4 show that all points determined directly by XRF analysis. fall very close to the diagonal equality line, indicating that the Siroquant The proportion of Na2Oislow(b1%) and under-estimated by results are consistent with the ash chemistry. With one exception, the Siroquant for most samples. This may in part be explained by the pres- plot for K2O is also consistent for those samples with K2Ogreaterthan ence of minor Na in the K-feldspar based on EDS data (see below), 1%. This conﬁrms the identiﬁcation of K-feldspar in the XRD analysis. which was not allowed for in calculating the inferred chemical compo- The remaining samples, however, have low proportions of K-feldspar, sition. A sample of the conglomerate roof material, however, which is

Table 1 Proximate analysis and vitrinite reﬂectance of Great Northern coal and associated strata from Newvale No. 1 Colliery. Proximate analysis data from CSIRO reports.

Sample Thickness Moisture Ash Fixed carbon Volatile matter CV Total S Rv ran (m) (%, ad) (%, ad) (%, ad) (% daf) (Btu/1b, daf) (%, ad) (%)

21379-roof 0.1 0.58 95.85 –– –0.06 – 21380 0.02 1.02 40.62 –– –0.16 – 21381 0.2 2.59 12.06 51.16 40.1 14,550 0.36 0.81 21382-p 0.03 2.57 51.93 24.10 ––0.21 – 21383 0.37 2.68 8.84 55.96 36.8 14,570 0.35 0.85 21384-p 0.01 2.40 68.92 5.38 ––0.11 – 21385 0.14 2.70 10.06 58.52 32.9 14,570 0.33 0.83 21386 0.6 2.65 6.67 53.82 40.6 14,750 0.39 1.07 21387 0.5 2.82 11.11 53.80 37.5 14,620 0.36 – 21388 0.52 2.86 9.57 53.30 39.1 14,520 0.39 – 21389-p 0.01 2.69 62.45 –– –0.17 – 21390 0.14 2.44 28.50 42.56 38.4 14,700 0.41 – 21391-floor 0.08 1.98 57.56 –– –0.2 – ad = air-dried basis; daf = dry ash-free basis; CV = calorific value; Rv ran = mean random reflectance of vitrinite; – = no data. 98 L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110

Table 2 quartz, a significant proportion of K-feldspar, and a lesser proportion Maceral analysis (vol. %, mineral-free basis) of Great Northern coal samples from of illite (Table 3). A similar siltstone is present below the seam in the Newvale No. 1 Colliery. exposure at Catherine Hill Bay, although the material at that location Maceral 21381 21383 21385 21386 21388 21390 contains less quartz, higher proportions of kaolinite and smectite, and Telinite 10.6 6.0 12.0 16.1 30.3 4.2 minor proportions of both K-feldspar and a plagioclase component Collotelinite 21.4 23.5 22.4 31.2 32.4 10.2 (Table 4). Collodetrinite 32.6 18.3 8.2 23.1 17.3 71.1 The siltstone at the base of the seam at Catherine Hill Bay ranges Vitrodetrinite 0.4 0.2 0.4 0.4 0.2 0 from 0.06 to 0.2 m in thickness (Fig. 3). Thin-section examination Corpogelinite 0 0.2 0 1.0 2.2 2.1 Gelinite 0.2 0.8 1.0 bdl bdl 0 shows a framework consisting mainly of detrital quartz, K-feldspar Total vitrinites 65.3 49.0 44.1 71.8 82.5 87.5 and kaolinized biotite, supported by abundant silica cement. Fusinite 3.1 3.2 6.4 3.1 4.1 2.1 Dispersed plant fragments with silica-impregnated tissues are also Semifusinite 21.8 36.3 40.5 17.1 7.7 4.2 present. Minor veins of K-feldspar cut across the sedimentary fabric, Funginite 0 0 0 bdl 0.2 0 including the plant fragments (Fig. 5A). SEM examination shows Inertodetrinite 3.5 2.6 5.2 1.8 1.4 3.9 Macrinite 1.9 2.2 1.8 1.0 0.6 0.7 that the K-feldspar occurs in laminae nearly parallel to bedding in Micrinite 1.9 2.0 1.2 1.4 0.2 0.2 the siltstone floor of the Newvale section (Fig. 5B). Minor detrital Secretinite 0 0 0.2 0.2 0 0 microperthite and albite are also present, consistent with the XRD Total inertinites 32.2 46.2 55.3 24.7 14.3 11.1 data. Sporinite 0.8 1.2 0.2 1.4 0.6 0.5 fi Cutinite 0.8 1.2 bdl 1.2 2 0.7 Quartz also occurs as veins (Fig. 5C) and an in lling of crushed fusi- Resinite 0 0.4 0.2 bdl bdl 0 nite-like material in this sample (Fig. 5D), as well as in more massive Liptodetrinite 0.8 2 0 0.8 bdl 0.2 forms. The siltstone floors in both localities resemble ganisters, which Suberinite 0 0 0 0 0.4 0 are hard, compact very fine to medium-grained quartz arenites (cf. Fluorinite 0 0 0.2 0 0 0 Folk, 1974), cemented by silica, that have gone through pedogenesis Bituminite 0 0 0 0 0.2 0 Total liptinites 2.5 4.8 0.6 3.5 3.3 1.4 (Retallack, 1977). They occur commonly below coal seams and contain carbonaceous traces (e.g. Besly and Fielding, 1989; Hemingway, 1968). However, quartz in the siltstone floors of both sections is not as abun- indicated in Table 3 as having a significant proportion of albite (Na- dant as that within true ganisters, which as Percival (1983) has sug- feldspar), is the exception to this trend, and plots very close to the gested, should be N95%. Correspondingly, clay minerals are less equality line. depleted in the floor samples of the present study than in ganisters Although there is a broad positive correlation between the XRD that have gone through leaching processes in a palaeosol profile. and XRF results for Fe2O3, most points plot below the equality line. The siltstone at Catherine Hill Bay is underlain by a mudstone This may be because the dolomite identified by the XRD data in containing abundant smectite and significant proportions of kaolinite most samples also contains some Fe, which was not allowed for in and quartz, along with a minor proportion of feldspar (probably an the calculation process. Alternatively, some of the iron in the rocks albite-rich plagioclase) and muscovite. Oriented-aggregate XRD or the LTAs may occur in non-crystalline oxy-hydroxide form. The analysis ( Table 7)confirms that smectite is more abundant than in proportion of TiO2-bearing minerals also appears to be under- the overlying siltstone, and is the main expandable clay mineral pre- estimated by the XRD analysis. However, SEM analysis shows that sent. Similar material may also occur below the siltstone in the Newvale the Ti-bearing minerals occur largely as fine-grained crystallites section, but was not available for sampling from the mine workings. (b0.5 μm) in the kaolinite matrix or vermicular aggregates, and The roof of the seam in both sections is a framework-supported because of this fine grainsize, possibly poor crystallinity and probably polymictic conglomerate, with a variety of different lithic fragments small percentages, such materials are difficult to evaluate by XRD representing chert and acid volcanic materials (cf. Ward et al., analysis. 1986) derived from the older strata in the New England Fold Belt. At the Catherine Hill Bay exposure this unit is around 25 m in thick- 3.3. Minerals in roof and floor samples ness (Diessel and Hutton, 2004). Mineralogically, the material forming the immediate roof of the Newvale section contains abundant The Great Northern seam in the Newvale Colliery section rests on quartz and plagioclase feldspar (albite), together with some musco- a floor of silicified carbonaceous siltstone, containing abundant vite and small percentages of chlorite and siderite (Table 3). This

Table 3 Mineralogy of Great Northern coal LTAs and non-coal rock samples from Newvale No. 1 Colliery by XRD and Siroquant (wt.%).

Sample LTA Qtz Kao I I/S M Chl Dol /Ank Sid K-feld Alb Ana Goy Fap Bass

21379 – 48.0 10.8 4.3 7.9 2.8 1.3 24.9 -roof 21380 68.2 3.2 2.3 1.4 83.5 9.2 0.4 21381 14.4 6.3 83.6 6.2 1.8 2.0 21382-p – 2.5 87.1 0.5 9.1 0.7 21383 10.6 8.8 77.3 8.2 3.3 2.4 21384-p – 4.9 83.8 9.7 1.6 21385 11.6 15.0 74.6 2.9 3.3 2.9 1.3 21386 8.0 11.2 59.8 12.0 5.7 1.2 2.0 2.8 5.4 21387 12.8 14.0 61.9 8.5 1.3 1.7 10.4 0.7 1.5 21388 10.2 19.8 28.4 30.6 1.6 12.1 6.8 0.7 21389-p – 13.1 44.8 6.9 34.3 0.9 21390 28.1 53.9 1.9 16.7 1.3 26.3 21391-ﬂoor – 65.9 6.1 10.9 17.2

Qtz = quartz; Kao = kaolinite; I = illite; I/S = mixed-layer illite/smectite; S = smectite; M = muscovite; Chl = chlorite; Dol = dolomite; Ank = ankerite; Sid = siderite; Ana = anatase; Alb = albite; K-feld = K-feldspar; Goy = goyazite; Fap = ﬂuorapatite; Bass = bassanite; Alun = Alunogen. L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110 99

Table 4 Mineralogy of Great Northern coal LTAs and non-coal rock samples from Catherine Hill Bay by XRD and Siroquant (wt.%).

Sample LTA Qtz Kao I I/S S M Chl K-feld Alb Ana Alun

Mudstone roof – 1.3 7.7 90.3 0.7 GN27 16.2 51.7 35.3 11.6 1.3 GN26 16.6 28.0 59.1 13.0 GN25 14.1 23.3 61.9 2.4 10.0 2.4 Claystone 1 – 2.0 89.0 8.6 0.4 GN23 19.2 14.0 72.1 11.9 2.0 GN22 11.1 15.0 70.4 12.1 2.6 Claystone 2 – 2.3 77.2 18.7 1.8 GN19 14.2 22.3 62.3 2.7 9.9 2.9 GN17 13.9 23.1 62.8 11.4 2.7 GN16 18.6 24.9 57.2 8.4 9.4 GN11 12.2 34.0 39.4 20.7 5.9 GN10 15.4 66.5 21.7 8.1 1.8 2.0 GN9 23.4 57.2 19.9 10.6 5.8 6.5 GN7 24.2 58.2 28.9 12.8 GN6 53.3 86.9 9.3 3.8 Siltstone ﬂoor – 34.9 33.6 1.4 19.8 4.8 5.4 Mudstone ﬂoor – 20.0 24.2 46.1 5.1 4.6

Abbreviations same as Table 3.

Table 5 Major element analyses of Great Northern coal ash and non-coal samples from Newvale No. 1 Colliery (%, on LOI-free basis).

Sample HTA SiO2 Al2O3 TiO2 Fe2O3 MgO CaO Na2OK2O MnO P2O5 21379-roof – 76.08 13.49 0.61 3.36 1.01 0.32 2.50 2.49 0.06 0.08 21380 40.82 6.90 2.60 0.23 20.15 27.26 42.20 0.06 0.24 0.33 0.04 21381 11.95 51.09 36.21 1.14 2.97 2.85 4.58 0.21 0.90 0.03 0.02 21382-p – 54.01 40.80 1.50 0.99 0.36 0.28 0.24 1.77 0.01 0.04 21383 8.74 48.57 34.45 0.95 5.50 3.45 6.02 0.17 0.65 0.09 0.16 21384-p – 55.51 37.82 2.47 0.93 0.35 0.28 0.20 2.37 0.01 0.05 21385 9.90 55.11 33.66 1.19 5.24 1.62 2.11 0.17 0.76 0.08 0.07 21386 6.54 46.72 31.44 1.82 8.25 3.74 6.33 0.19 0.77 0.13 0.61 21387 11.02 56.31 31.58 1.89 4.04 1.45 1.88 0.35 2.37 0.07 0.06 21388 8.97 58.64 25.28 1.31 8.86 1.43 0.95 0.65 2.75 0.11 0.02 21389-p – 62.26 26.49 2.10 1.21 0.41 0.14 0.47 6.88 0.01 0.03 21390 27.41 79.94 11.85 0.49 1.54 0.30 0.13 0.29 5.40 0.03 0.03 21391-floor – 76.24 16.94 0.43 1.03 0.42 0.06 0.31 4.54 0.01 0.02 represents a combination of the minerals in both the framework par- the organic matter in the outcropping coal seam. Similar quartz and ticles and the finer-grained matrix material. As discussed above, the clay mineral assemblages, however, are present in the LTAs of the identification of albite as the dominant feldspar is confirmed by the individual coal plies at both locations (Tables 3, 4). high proportion of Na2O indicated by the XRF analysis. The mudstone exposed between the conglomerate and the main Table 7 part of the coal bed in one part of the Catherine Hill Bay exposure Mineralogy of b2 μm fraction of coal LTAs and non-coal strata using oriented aggregate (Fig. 3) was also evaluated. This mudstone consists almost entirely XRD techniques. of smectite (Table 4), with minor proportions of kaolinite, quartz Sample K(+Chl) Illite Expandable Nature of expandable and possibly a trace of quartz-feldspar. Oriented aggregate XRD data (%) (%) clay minerals (%) clay minerals shows that the b2 μm fraction consists almost entirely of smectite, Great Northern seam in Newvale No. 1 Colliery along with a trace of kaolinite. 21379-roof 54.4 30 15.6 Regular I/S 21380 74.6 11.5 13.9 Regular I/S 3.4. Minerals in coal plies 21381 93.3 6.7 Regular I/S 21382-p 100 – The coal plies at Catherine Hill Bay have higher LTA percentages 21383 93.5 6.5 Regular I/S+smectite (minor) than those sampled at Newvale, possibly due in part to oxidation of 21384-p 96.7 3.3 21385 94.9 5.1 Smectite+regular I/S (minor) Table 6 21386 100 – Major element analyses of Great Northern non-coal samples from Catherine Hill Bay (%, on 21387 87.5 4.9 7.7 Smectite LOI-free basis). 21388 56.9 9.6 33.5 Smectite 21389-p 84.9 3.7 11.3 Smectite Sample SiO Al O TiO Fe O MgO CaO Na OKO MnO P O 2 2 3 2 2 3 2 2 2 5 21390 31.8 11 57.2 Smectite Mudstone 63.25 25.68 0.31 4.25 3.50 0.17 0.24 2.55 0.02 0.02 21391-floor 28.9 38.5 32.6 Irregular I/S roof Claystone 1 54.40 41.11 1.52 0.69 0.38 0.05 0.15 1.65 0.01 0.04 Great Northern seam at Catherine Hill Bay Claystone 2 55.25 40.08 2.18 0.86 0.24 0.02 0.01 1.32 0.01 0.04 Mudstone roof 1.0 99.0 Smectite Siltstone 76.51 19.25 0.50 1.00 0.62 0.12 0.51 1.46 0.01 0.02 Claystone1 100 floor Claystone 2 100 Mudstone 71.77 21.01 0.34 2.59 2.04 0.42 0.23 1.59 0 0.01 Siltstone floor 72.6 4.5 22.9 Smectite floor Mudstone floor 41.7 58.3 Smectite 100 L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110

100 50 80 40

60 30

40 20 from Siroquant (%) 20 10 3 from Siroquant (%) 2 O 0 2 0 Al SiO 0 20 40 60 80 100 0 1020304050

SiO2 by XRF (%) Al2O3 by XRF (%)

50 40 40 30 30 20 20 10 10

0 0 MgO from Siroquant (%) CaO from Siroquatnt (%) 0 2040600 10203040 CaO by XRF (%) MgO by XRF (%)

8 4

6 3

4 2

2 1 O from Siroquant (%) 2 O from Siroquant (%) 2 0 0 K 02468Na 01234

K2O by XRF (%) Na2O by XRF (%)

25 3 20 2 15 10 1 5 from Siroquant (%)

from Siroquant (%) 3 2 O

2 0 0 0 5 10 15 20 25 TiO 0123 Fe Fe2O3 by XRF (%) TiO2 by XRF (%)

Fig. 4. Comparison between proportions of major element oxides in coal LTAs and non-coal strata from Newvale, inferred from Siroquant and determined by XRF. The diagonal line represents equality in each plot.

The LTA percentage of the lowermost coal plies in each section is the parting represented by sample 23,184. Davis et al. (1984) have very high, with a little over 50% at the immediate base of the seam noted that quartz appears to be most abundant in the basal parts at Catherine Hill Bay and 25–30% in the overlying parts of both and near the margins of peat beds in the Okefenokee swamp-marsh seam sections. The coal plies in the middle and upper parts of the complex, and suggested that it represents detrital quartz introduced seam in both sections have significantly lower LTA percentages. An to the peat by mixing with the sediment of the swamp floor, due to exception is the topmost ply in the Newvale section, which has an either bioturbation or contemporaneous clastic deposition. The LTA LTA yield of almost 70%. However, as discussed further below, the in the topmost ply at Catherine Hill Bay (sample GN27), however, LTA of this ply consists almost entirely of authigenic carbonate has a relatively high quartz content (52%). If the mineralogy is calcu- minerals (dolomite/ankerite), in contrast to the clay-dominated lated on a dolomite-free basis, quartz also appears to be very abun- assemblages of the other ply samples. dant in the topmost coal ply (sample 21380) of the Newvale seam section. 3.4.1. Quartz Although some detrital input may have been involved, the mode Quartz is more abundant as a fraction of the LTA for the Catherine of quartz occurrence favours a dominantly authigenic origin for the Hill Bay samples than for the Newvale samples. This may reflect a high quartz percentage in the lowermost coal (sample 21390) from lateral change across the field, but may also partly reflect dilution of the Newvale section. The quartz mostly occurs as cell and micropore the mineral matter in the Newvale samples by epigenetic dolomite infillings within the macerals (Fig. 6A, B), and occasionally as detrital and/or ankerite; such carbonates are not present in the LTA of the fragments in collodetrinite (Fig. 6C). Given the similar occurrence of coals from Catherine Hill Bay. quartz and K-feldspar, the lowermost intra-seam band at Newvale Quartz is particularly abundant in the lowermost ply of the has characteristics that more closely resemble the floor strata. Ward Catherine Hill Bay section (sample GN6), and is more abundant in (1991) also noted a hard siliceous siltstone floor of a Tertiary coal the lower part of the seam at both locations than in the upper part. seam in the Mae Moh Basin of Thailand, and suggested that an accu- For example, quartz makes up 15–20% of the LTA in samples 21388 mulation of authigenic or biogenic silica may partly be developed. to 21385 from Newvale, but b10% of the LTA of the coal plies above Sykes and Lindqvist (1993) observed diagenetic quartz and L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110 101

A B

C D

Fig. 5. (A) Vein probably made up of K-feldspar cutting well-preserved plant tissue in the siltstone floor of the Catherine Hill Bay section; thin section under crossed polars. (B) SEM image of K-feldspar veins in the siltstone floor of the Newvale section. (C) SEM image of quartz veins in the siltstone floor of the Newvale section. (D) SEM image of quartz with crushed fusinite in the siltstone floor of the Newvale section. amorphous silica of different forms in Tertiary coals from a number of in the LTA of the coal plies in the lower part of both seam sections. New Zealand coalfields. They suggested that sub-horizontal silicifica- XRD studies suggest that an albite-rich plagioclase, as well as K- tion of some coals was due to the infiltration of the peat bed by silica- feldspar, also occurs in the LTA of the lower plies at Catherine Hill Bay. saturated groundwater and crystallisation of quartz at greater depth; SEM studies indicate that the K-feldspar in both the lowermost the silica may have been derived from leaching of the basement rocks coal ply and the adjacent non-coal parting at Newvale is intimately or from siliceous phytoliths within the coal-forming plant material. In associated with laminae of mixed-layer illite–smectite (I/S), the addition, the silica could have been derived from the alteration of distribution of which is discussed more fully below. The feldspar in volcanic glass, which may also lead to the formation of porcellanite some cases occurs as irregular to elongated aggregates of interlocking or chert-like rocks (cf. Loughnan and Ray, 1978). The mineral matter euhedral crystals (Fig. 6F), partly cross-cutting the layering of the in coals from Xuanwei, Yunnan, southwestern China also contains a maceral components. This mode of feldspar occurrence is even more high proportion of quartz (Dai et al., 2008b), which was deposited strongly developed in the lowermost intra-seam claystone band, from silica-bearing solutions that originated from weathering of and is discussed more fully below. Given the occurrence of similar basaltic rocks in the hinterland (e.g. Ren, 1996). veins in the floor sample as well, at least some of the feldspar in the Since the clay mineralogy of the coals in the upper part of the coal is thus possibly of hydrothermal origin. However, this does not Newvale section (discussed below) is dominated by authigenic exclude additional feldspar input from re-working of penecontem- kaolinite, and the relative abundance of quartz in the middle part of poraneous volcanic material, such as the debris that made up the the coal seam is relatively low, authigenic processes are thought to underlying Awaba Tuff. have been dominant in the middle part of the original peat accumula- The abundance of both quartz and K-feldspar in the lowermost tion. Ruppert et al. (1991) also found authigenic quartz to be much plies of the coal seam is comparable to those in the floor samples more abundant in the laterally interior part than the marginal parts from Catherine Hill Bay. This suggests that the basal part of the orig- of the Upper Freeport coal of the Appalachian Basin of the eastern inal peat bed was made-up of organic matter admixed with the same USA. detrital sediment that was supplied to the basin before the swamp A quartz-rich rock fragment about 50 μm in diameter was was established. observed under the SEM in the uppermost coal sample (21380) at Newvale (Fig. 6D, E). EDS studies indicate that the rind of this frag- 3.4.3. Clay minerals ment is composed of K-bearing aluminosilicate, possibly illite, whilst As indicated in Table 3, the clay minerals in the coal plies in both the interior is essentially composed of even-grained quartz crystals sections are represented by a combination of kaolinite and mixed- with some disseminated chlorite, the latter possibly being an alter- layer illite-smectite (I/S). With the exception of the lowermost coal ation product of biotite. This rock fragment is thought to be a devitri- ply at Newvale (sample 21390), kaolinite is the most abundant of fied glass spherule, and was probably derived from the same source these components. Kaolinite is the only clay mineral identified by as the abundant acid volcanic rock fragments that make-up the over- powder XRD techniques above the ply represented by sample 21387 lying conglomerate bed. in the Newvale section, and is the dominant component (N60% of the LTA) in the upper plies of the Catherine Hill Bay seam section. 3.4.2. K-feldspar As noted for other coals in the Sydney Basin (Ward, 1989), the K-feldspar occurs in the lower part of the coal seam in both seam kaolinite in the coal plies, especially those in the upper part of the sections. It is especially abundant in the lowermost ply at Newvale seam, has a powder XRD pattern indicating a well-ordered crystal (sample 21390) and the associated non-coal parting, but also occurs structure. More poorly-ordered kaolinite, however, appears to be 102 L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110

A B

Q K

C D

E F Q

Fig. 6. Minerals in coal samples from Newvale: (A) Quartz (Q) along with kaolinite (K) in cell lumens in coal 21388. (B) Quartz in maceral micropores in coal 21388. (C) Detrital quartz grains in collodetrinite in coal 21388. (D) Quartz-rich rock fragment in coal 21380. (E) Enlargement of (D) showing chlorite (C) and quartz (Q) crystals. (F) K-feldspar in coal 21390. present in the LTA of the lower coal plies. SEM analysis shows that the A graphical profile of the clay mineralogy at Newvale, based on kaolinite in the coals primarily occurs as thin bands intimately associ- XRD data of oriented-aggregates prepared from the LTAs and non- ated with vitrinite, and as vermicular crystals, cell cavity and cleat coal rocks, is given in Fig. 8. The clay minerals in the roof and floor infillings (Fig. 7A, B, C). Together with the well-ordered structure, comprise roughly equal proportions of kaolinite, illite and expandable this indicates that the kaolinite in the coals, especially in the upper clay minerals. However, within the coals themselves, kaolinite is the part of the seam, was formed mainly by authigenic precipitation in dominant clay mineral, and only very small proportions of illite and the original peat swamp (Ward, 1989). The kaolinite-rich coal plies expandable clay minerals are indicated by oriented-aggregate XRD at Newvale (samples 21381, 21383 and 21385, Table 2) also have techniques. higher inertinite percentages than those in the lower part of the Oriented-aggregate XRD of the Newvale samples (Table 7) further seam. These higher inertinite percentages are in turn associated shows that the nature of the expandable clay minerals varies through with a higher macroporosity, which would have allowed more kaolinite the seam section (Fig. 8). The expandable clays consist mainly of precipitation in the cavities of the maceral components. smectite in the lower part of the coal bed, smectite together with reg- Some of the kaolinite in the lower part of the seam appears to ularly interstratified I/S in the upper part of the seam, and regular I/S represent pseudomorphs after biotite (Fig. 7D). Although rare in com- only in the uppermost layers. Fig. 9 shows XRD traces of two coal LTAs parison to the authigenic kaolinite, this may represent altered rem- from the upper and lower parts of the seam, after glycol saturation nants of detrital sediment, or possibly volcanic ash, introduced to and heat treatment. the coal swamp in the early stages of peat accumulation. The non-kaolinite clay minerals, illite and I/S, mostly occur in the 3.4.4. Siderite coal as thin bands and laminae, and are interpreted to be essentially Siderite is present in most of the coal plies in the seam section at of detrital origin. Although present in the upper and lower coal plies Newvale (Table 3), but not in the intra-seam non-coal partings. It at Newvale (in the two coal plies near the floor, they make up roughly mainly occurs as syngenetic nodules, as chemically-zoned euhedral half of the clay fraction), these minerals are virtually absent in the crystals (Fig. 10A), and as cell infillings. EDS data indicate that some middle part of the coal bed. This further suggests that the original of the cell infillings have overgrowths (Fig. 10B) with higher concen- peat included more admixed detrital minerals derived from the sedi- trations of Mn than the earlier-formed material. ment source region supplied to the basin before and after the peat The Great Northern seam was deposited under terrestrial condi- swamp was established. tions, with no indication of marine influence. Since the pore waters L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110 103

A B

C D

Fig. 7. Kaolinite in coal 21388 from Newvale: (A) Vermicular kaolinite. (B) Cell inﬁllings of kaolinite. (C) Cleat inﬁllings of kaolinite. (D) Probable kaolinite pseudomorphs after biotite.

of the peat would thus have contained little if any dissolved SO4, the Both the cleat inﬁllings and the abundant veins are clearly post- nodular siderite was probably formed by interaction of dissolved Fe depositional. As discussed by Ward (2002), the veins near the top of with CO2 generated by fermentation of the organic matter shortly the seam, in particular, may have been derived from expulsion of after peat accumulation (cf. Botz and Hart, 1983).

3.4.5. Dolomite/ankerite Dolomite and/or ankerite (ferroan dolomite) occur in most of the coal plies of the Newvale section, mostly as epigenetic cleat inﬁllings in the bright coal layers. Abundant veins of these minerals, oriented generally parallel to bedding, also occur in the topmost ply of the coal seam, immediately below the roof material.

Fig. 8. Column section showing vertical variations in clay mineralogy for the Great Fig. 9. XRD traces obtained from b2 μm fractions of (A) coal 21383 and (B) coal 21388 Northern seam at Newvale. after glycol saturation and after heating at 400 °C for 2 h. 104 L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110

A B

C D O Al Sr K P

1 Ca Ba G

Fig. 10. (A) Zoned euhedral siderite crystals in coal 21388. (B) Cell-infilling siderite overgrown by a later siderite in coal 21390. (C) Goyazite–crandallite (G) and kaolinite (K) cell infillings in coal 21386. (D) EDS spectrum of point 1 from (C). organically-associated Ca and Mg from the macerals during rank ashing (cf. Ward et al., 2001), or it may have been formed in the out- advance. Studies of other coals using electron microprobe techniques cropping coals by processes associated with weathering. (e.g. Li et al., 2007, 2010; Ward et al., 2005, 2007) have shown that such organically-bound inorganic elements are liberated from coal 3.5. Minerals in non-coal partings macerals, especially vitrinite, as part of the molecular changes associated with the sub-bituminous to bituminous transition in the rank The intra-seam partings in the Newvale section are represented by advance process. carbonaceous mudrocks with ash yields of 52–69% (Table 1). XRD No carbonate minerals have been observed in the LTA of the analysis (Table 3) indicates that the upper two partings (samples Catherine Hill Bay coal samples. This contrasts with their presence 21382-p and 21384-p) both have similar mineral assemblages, with in most of the Newvale samples, and may be a result of weathering dominant kaolinite, minor quartz and K-feldspar, and traces of ana- associated with exposure of the coal seam. tase and possibly siderite. The two claystone partings analysed in the section at Catherine Hill Bay, correlating with these two horizons 3.4.6. Other minerals (Fig. 3), also have similar mineralogy. The mineral assemblage in the Apatite and alumino-phosphate minerals of the crandallite group lowermost parting at Newvale, however (sample 21389-p), has a are noted in the XRD data for sample 21386 from the Newvale seam lesser proportion of kaolinite and more quartz, as well as abundant section. The alumino-phosphate occurs with kaolinite as cleat or K-feldspar and some mixed-layer I/S. crack infillings, and also co-exists with kaolinite as cell infillings (Fig. 10C). The EDS spectrum (Fig. 10D) indicates that the material 3.5.1. Kaolinite is a member of the crandallite series containing P, Sr, Ca and a trace As with the LTA of the coal plies, the XRD patterns show that the of Ba, suggesting either a goyazite or a Sr-bearing crandallite (cf. upper two partings at Newvale (21382-p and 21384-p) are dominat- Ward et al., 1996). Alumino-phosphate minerals would be expected ed by well-ordered kaolinite. However, the kaolinite in the lowermost from intra-seam precipitation if Al was also available in reactive parting (21389-p) is mostly disordered in nature, and is quite similar form at the site of phosphate deposition, and apatite if Al was not to that of the LTA from the coal samples in the lower part of the seam. available to react with the precipitated phosphatic material (Ward, SEM observations show that the kaolinite in samples 21382-p and 2002). 21384-p is represented mostly by graupen to vermicular kaolinite Minor proportions of bassanite are also present in some of the forms (Fig. 11A, B). A vermicular texture in matrix kaolinite is often Newvale LTAs, especially those from plies with low ash yields used as evidence that the sediment is a tonstein (Ruppert and and therefore abundant organic matter. As discussed by Frazer and Moore, 1993; Spears, 1971). As defined by Bohor and Triplehorn Belcher (1973), the bassanite is most likely an artefact of the low- (1993) among others, such materials generally represent altered temperature ashing process, derived from the interaction of organic air-fall volcanic ash layers deposited in a non-marine, commonly sulphur and organically associated calcium released by oxidation of coal-forming environment. the maceral components. Numerous papers have been published on tonsteins interbedded Bassanite was not found in the LTA of the coals from Catherine Hill with coal seams (e.g. Addison et al., 1983; Burger et al., 1990, 2000, Bay, and this may also assist to rule out the possiblity that the bassa- 2002; Dai et al., 2011; Hill, 1988; Knight et al., 2000), and the miner- nite in the Newvale coals is an oxidation product formed during stor- alogical and geochemical relationships between them and the associ- age. Alunogen was frequently detected in the Catherine Hill Bay LTA ated coals (Brownfield et al., 2005; Crowley et al., 1989, 1993; Hower samples. The alunogen may also have been produced by interaction et al., 1999). In some cases volcanic ash was not abundant enough to of inorganic elements and organic sulphur during low-temperature form visible ash layers in the coal beds (Dai et al., 2008a; Dewison, L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110 105

A B

C D

Fig. 11. Kaolinite in claystone partings from Newvale: (A) Graupen to vermicular kaolinite in 21382-p. (B) Graupen kaolinite in 21384-p. (C) Typical vermicular kaolinite with inclusions of anatase in 21384-p. (D) Platy kaolinite with ﬁne anatase inclusions in 21389-p.

1989), or only clayey micro-sized bands (around 100 μm in thick- however, matches all of these peaks, and this may indicate a mixture ness) are developed (Dai et al., 2007). Nevertheless the volcanic of two or more K-feldspar components, or possibly a range of feldspar material may still have an important impact on the coal geochemistry. structures (sanidine-anorthoclase series) in the parting samples. The common presence of vermicular kaolinite in the partings of Although comparison with the XRF data (see discussion above) the Great Northern seam provides evidence of in-situ alteration confirms that the material is essentially a K-feldspar, EDS analysis of (Zhou et al., 1982), and the partings may thus have formed by exten- individual particles within the parting samples (Table 8) indicates sive leaching of easily degraded volcanic ash material. The vermicular that the feldspars also contain a small proportion of Na. Deer et al. kaolinite in the tonstein partings appears to be rich in titanium, prob- (1992) note that similar proportions of Na may be found in K- ably in the form of anatase inclusions (Fig. 11C). Kaolinite with feldspars generally. The Na may exsolve and be represented by crystallites of anatase has also been identified in Indonesian tonsteins micro-perthite in feldspars of plutonic origin, but may remain incorpo- by Ruppert and Moore (1993). The kaolinite in sample 21389-p is rated in the lattice of more quickly-cooled volcanic feldspars such as mostly in the form of clasts, with rare platy crystals, but nevertheless sanidine. also contains anatase inclusions (Fig. 11D). SEM examination shows that the feldspar in the two upper partings Tabular or vermicular kaolinite phenocrysts have been regarded at Newvale (21382-p and 21384-p) mostly occurs as idiomorphic crys- as kaolinized biotite by some researchers (e.g. Knight et al., 2000). tals (Fig. 12A, B). As discussed above, EDS data indicates the presence of However, the diagenetic textures of the vermicular kaolinite crystals, a range in compositions from nearly pure potassium feldspar to sodi- and their relationship, if any, with the original volcanic texture, still um-bearing K-feldspar, with Na being up to 3.5% in single K-feldspar needs to be established (Spears, 2006). crystals (Table 8). This range may represent an anorthoclase–sanidine Tonsteins, perhaps more precisely described as kaolinite– series or a sodic sanidine material, indicating an acid to intermediate tonsteins, are also reported in other Late Permian coal seams of the volcanic ash input to form the upper two partings in the coal seam. Sydney Basin (Creech, 2002; Loughnan, 1971; Loughnan and Corkery, The SEM study also shows that some K-feldspar appears to have 1975; Loughnan and Ward, 1971), and several intervals of ash-fall been etched (Fig. 12C), especially in sample 21384-p. Based on EDS and ash-flow tuff are known to occur in other parts of the coal-bear- data, the etched K-feldspar appears to be free of Na, but some un- ing Sydney Basin sequence (Agnew et al., 1995; Diessel, 1965, 1985; etched K-feldspar particles (not all of them) contain both K and Na. Grevenitz, 2003; Kramer et al., 2001; Loughnan and Ray, 1978). How- Etched feldspar and quartz grains in pelitic layers of pyroclastic origin ever, although tonsteins of apparent volcanic origin have been have also been observed by Ruppert and Moore (1993), who sug- reported to occur in coal seams around the world, other processes gested that they had been leached in the acid-rich mires that formed may also be responsible for the thin beds of non-coal sediment com- the associated coal seams. Etched K-feldspar was also observed in the monly found within coal seams (Ward, 2002). Evidence of a volcanic high-sulphur coals from Yanshan, Yunnan, southwestern China (Dai origin, such as the textural features described above, are therefore et al., 2008a). significant in determining the origin of the kaolinite-dominated A small amount of kaolinite is also present as fragmented aggre- bands in the present study. gates in the upper two partings (Fig. 12D). Hence, although the common presence of anorthoclase–sanidine series material is suggestive 3.5.2. K-feldspar of a dominant volcanic origin, the presence of detrital K-feldspar indi- K-feldspar is present in all partings sampled from at Newvale and cates that the upper two partings could be mixed with minor Catherine Hill Bay. The XRD patterns of these samples display peaks epigenetic clasts. Mixing of mineral sources could have occurred if the at d-spacings of 3.29 Å, 3.25 Å, 2.90 Å and 2.60 Å that may be attribut- relevant ash-fall had temporarily slowed or halted peat accumulation ed to sanidine, anorthoclase or orthoclase. No single mineral, (Ruppert and Moore, 1993). 106 L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110

Table 8 EDS micro analyses of K-feldspar in Great Northern non-coal samples from the Newvale No. 1 Colliery (%, O by difference).

Element Parting 21382 Parting 21384 Parting 21389 Siltstone ﬂoor

(17 points) (15 points) (5 points) (18 points)

Mean Max Min Mean Max Min Mean Max Min Mean Max Min

Al 10.6 11.3 9.0 7.9 9.5 7.0 9.8 10.5 9.3 9.9 10.5 9.0 Si 28.4 30.1 26.1 24.1 30.0 20.9 32.5 33.0 32.1 32.3 34.3 30.9 O 48.5 53.1 43.1 60.8 65.6 48.7 44.4 45.2 42.5 44.4 45.7 43.5 K 9.6 14.7 8.4 5.4 11.3 3.2 12.9 14.7 11.8 13.3 15.2 11.7 Na 2.9 3.5 1.4 1.8 3.1 0.3 0.3 0.7 0 0.1 0.3 0 K/(K +Na) 0.8 0.9 0.7 0.7 1.0 0.5 1.0 1.0 0.9 1.0 1.0 1.0

Unlike that in the upper two partings, K-feldspar in the form of Veins consisting mainly of K-feldspar also occur in thin bands or detrital grains is dominant in the lowermost parting at Newvale laminae that are parallel to the bedding of the organic matter (sample 21389-p), along with poorly-ordered kaolinite in banded or (Fig. 13E) in sample 21389-p. This texture is similar to that of the massive aggregates (Fig. 13A). This may indicate K-feldspar formation K-feldspar laminae in the siltstone floor of the Newvale section. mainly by epiclastic input from a sediment source close to the peat Fig. 13F shows an intergrowth of K-feldspar with lesser quartz. This deposit. As indicated in Table 8, Na is much lower in the K-feldspar intergrowth indicates that both minerals were precipitated contem- of this bed than in the upper two partings. poraneously from the same fluid. Some K-feldspar in the lower parting at Newvale, however, occurs Both Simmons and Browne (1997) and Craw (1997) suggest as veins cross-cutting organic stringers (Fig. 13B) and cell-infillings temperatures of 220 °C–300 °C for adularia crystallisation from hy- (Fig. 13C, D). EDS elemental maps of Al, K and Na (Fig. 14) confirm drothermal fluids. Since no conversion has been observed in the that the veins consist essentially of pure K-feldspar with a negligible clay minerals surrounding the veins in the floor samples (Fig. 5B), amount of Na, if any. EDS also indicates very rare albite co-existing the veins most likely originated from low-temperature hydrothermal with K-feldspar in the cleat. No feldspar was detected with a compo- fluids in the late syngenetic stage, before the peat underwent diagene- sition falling between those of albite and K-feldspar. Although authi- sis. The origin of the hydrothermal fluid is uncertain, but was probably genic K-feldspar has been reported in sedimentary formations as associated with contemporaneous volcanic activity. overgrowths on detrital grains, as cementing materials and as veins (e.g. Hagen et al., 2001; Liu et al., 2003; Sandler et al., 2004), vein 3.5.3. Quartz and cell-infilling K-feldspars have not been reported in coal deposits. Despite the relative abundance of K-feldspar, only very small pro- The apparent precipitation of K-feldspar in shrinkage fissures of vitri- portions of quartz (1.3% and 4.9%) are present in the upper two part- nite appears to indicate hydrothermal activity in the late syngenetic ings at both locations (samples 21382-p and 21384-p from Newvale stage of peat development. Adularia, a low-temperature K-feldspar and claystone-1 and claystone-2 from Catherine Hill Bay). The quartz and a polymorph of orthoclase (Klein and Dutrow, 2007), has been in the upper two partings at Newvale occurs mostly as euhedral crys- found to be associated with hydrothermal mineralization, although tals (Fig. 12B). These may be beta-form quartz, but would require most commonly in epithermal deposits (e.g. Simpson et al., 2001; confirmation by cathodoluminescent emission analysis. Quartz is Zhang et al., 2010). more abundant (13.1%) in the lowermost parting at Newvale (sample

A B

F F

C D F

Fig. 12. K-feldspar in two uppermost partings from Newvale: (A) Euhedral K-feldspar (F; ? sanidine) in parting 21382-p. (B) Euhedral K-feldspar (F) and quartz (Q) in 21384-p. (C) Etched K-feldspar in 21382-p. Fine white grains are anatase. (D) K-feldspar (F) fragments in clastic kaolinite (K) matrix in 21384-p. L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110 107

A B

F F Q K

C D F Q

F K Q

F Q

Fig. 13. K-feldspar in lowermost claystone parting 21389-p from Newvale: (A) K-feldspar (F) and kaolinite (K) in banded or massive aggregates. (B) K-feldspar (F) veins cross- cutting organic stringers. Note a volcanic quartz (Q) in the upper left of the image. (C) Fusinite with mineral inﬁllings. (D) Enlargement of an area in (C) showing K-feldspar (F), quartz (Q) and kaolinite (K). (E) K-feldspar and quartz in thin bands. (F) Enlargement of an area in (E) showing an intergrowth of K-feldspar (K) and quartz (Q).

21389-p). Although mainly made-up of epiclastic material, this as fine grains in kaolinite aggregates (Fig. 11C, D). Discrete crystals sample contains occasional volcanic quartz grains (Fig. 13B). of anatase (Fig. 15A, B) also occur. Sample 21384-p contains anatase The proportion of quartz appears to be lower in the partings com- that appears to have replaced coal maceral components (Fig. 15C, D). pared to the LTA of the adjacent coal plies, especially if allowance is Anatase is a common secondary mineral in tonsteins (Spears and made for dilution of the silicates in the coal plies by authigenic Kanaris-Sotiriou, 1979), and is inferred to represent reprecipitation carbonate components. Although a lesser proportion of quartz than products of chemically leached labile components in the original ash in normal mudrocks is not necessarily a characteristic specificto material (Triplehorn et al., 1991). The titanium may have been tonsteins, it serves in conjunction with other evidence to identify a derived from the breakdown of Ti-rich volcanic glass, ilmenite, volcanic origin (Bohor and Triplehorn, 1993). Together with the magnetite or rutile (Ruppert and Moore, 1993). modes of quartz occurrence as cell infillings in the coal, this may indicate that the volcanic ash that formed the partings, especially the upper two partings, contained little if any free quartz. Such material 3.5.5. Other minerals may also have been admixed with the peat itself as the seam devel- Although not in sufficient concentration to be detected by XRD oped, but if so additional quartz was possibly added to the peat by analysis, apatite has been identified by SEM studies in the kaolinite precipitation in the pores of the maceral components. Some of this matrix of sample 21384-p (Fig. 15B). Idiomorphic apatite observed silica may have been released into the peat waters by break-down by other authors (Knight et al., 2000), however, was not identified. of volcanic glass in the coal-forming mire environment. A phosphate particle containing Fe, Mn, Mg and Ca was detected in sample 21384-p (Fig. 16A, B), but not in any of the other partings 3.5.4. Anatase studied. Small proportions of anatase occur in all partings from both the No other phases reported elsewhere to be diagnostic indicators of Newvale (0.7–0.9%) and Catherine Hill Bay sections (0.4–1.8%). The volcanic origin (e.g. biotite, zircon, pyroxene and glass shards) were anatase appears to be embedded between the kaolinite platelets; it observed in the partings of the seam section studied, presumably also commonly occurs as crack infillings in the kaolinite matrix and because of the extensive alteration involved. 108 L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110

Fig. 14. Elemental maps of Al, Si, K and Na in K-feldspar veins cross-cutting organic matter in claystone parting 21389-p.

4. Conclusions The floor strata below the Great Northern seam are made-up of abundant quartz, with minor and equal proportions of kaolinite, illite Mineralogical and chemical variations have been identified and expandable clay minerals. The floor material also contains abun- through the vertical profile of the Great Northern coal seam, involving dant detrital K-feldspar. Coals from the middle parts of the seam are contrasts between the mineral matter in the individual coal plies as dominated by well-ordered kaolinite with minor quartz, and in well as between the coal plies, the floor strata, and the intra-seam some cases contain minor phosphates and anatase. The carbonate non-coal bands. There are also differences between the mineral minerals, siderite, and dolomite or ankerite, occur in the fresh coal matter in the coal plies of the outcrop section at Catherine Hill Bay at Newvale but not in the weathered coal at Catherine Hill Bay. How- and the samples from the mine exposure at Newvale; these may be ever, non-kaolinite clay minerals and detrital feldspar are abundant in related, at least in part, to weathering of the coal at the outcrop site. the lowermost ply of the coal seams, suggesting that the immediate

A B A A F

C D

Fig. 15. Anatase in claystone parting 21384-p: (A) Anatase grains (A) and K-feldspar (F). (B) Apatite (Ap) and anatase (A) in a kaolinite matrix. (C) Anatase (A) replacing macerals. (D) Enlargement of an area in (C). L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110 109

A B

1 2

Fig. 16. (A) SEM image of Fe, Mn, Mg, Ca phosphate (point 1) and K-feldspar (point 2) in kaolinite matrix of claystone parting 21384-p. (B) EDS spectrum of point 1.

base of the peat bed was made-up of organic matter admixed with also like to thank Don Triplehorn and Shifeng Dai for their construc- the same detrital sediment as supplied to the basin before the tive comments on the manuscript. swamp was established. The presence of disordered kaolinite in the lower coal plies, similar to that in the floor samples, further supports such a conclusion. References Authigenic K-feldspar occurs in the lower part of the coal seam, fl Addison, R., Harrison, R.K., Land, D.H., Young, B.R., Davis, A.E., Smith, T.K., 1983. especially in the Newvale section, including the siltstone oor, the Volcanogenic tonsteins from Tertiary coal measures, East Kalimantan, Indonesia. lowermost claystone parting, and the coal plies between. The K- International Journal of Coal Geology 3, 1–30. feldspar shows a variety of modes of occurrence, including cell infill- Agnew, D., Bocking, M., Brown, K., Ives, M., Johnson, D., Howes, M., Preston, B., Rigby, R., Warbrooke, P., Weber, C.R., 1995. Sydney Basin, Newcastle Coalfield. In: Ward, C.R., ings, irregular veins, and thin laminae parallel to the bedding of or- Harrington, H.J., Mallett, C.W., Beeston, J.W. (Eds.), Geology of Australian Coal Basins: ganic matter or detrital clay bands. The texture and chemical Geological Society of Australia Coal Geology Group Special Publication, vol. 1, composition of the K-feldspar suggest that it originated from low- pp. 197–212. temperature hydrothermal fluids in the late syngenetic stage of peat ASTM, 2007. Annual book of ASTM standards. Gaseous Fuels; Coal and Coke, vol. 05. 06. American Society for Testing and Materials, West Conshohocken, PA, pp. 244–250. development, probably associated with contemporaneous volcanic Besly, B.M., Fielding, C.R., 1989. Palaeosols in Westphalian coal-bearing and red-bed activity. The hydrothermal activity may have occurred in a relatively sequences, central and northern England. Palaeogeography, Palaeoclimatology, – localised area, given that similar vein minerals are rare in the Catherine Palaeoecology 70, 303 330. fl fl Bocking, M., Howes, M., Weber, C.R., 1988. Palaeochannel development in the Moon Hill Bay siltstone oor, and that the feldspar in the oor largely occurs as Island Beach Sub-Group of the Newcastle Coal Measures. Proceedings of 22nd detrital grains. Symposium on Advances in the Study of the Sydney Basin, Department of Geology, Although the lowermost intra-seam band at Newvale has character- University of Newcastle, NSW, pp. 37–45. fl Bohor, B.F., Triplehorn, D.M., 1993. Tonsteins: altered volcanic-ash layers in coal- istics more closely resembling the oor strata, the uppermost two bands bearing sequences. Geological Society of America, Special Paper, 285. 40 pp. in that section, and also the bands at a similar level in the Catherine Hill Botz, R.W., Hart, G.H., 1983. Mineralogical, petrographical and geochemical investiga- Bay exposure, consist almost entirely of well-ordered kaolinite. The tions of outbursting in Australian coal mines. Proceedings of the Australasian Insti- tute of Mining and Metallurgy 286, 41–49. kaolinite in these bands displays graupen to vermicular textures, and Branagan, D.F., Johnson, M.W., 1970. Permian sedimentation in the Newcastle Coal- the bands appear to represent tonsteins within the coal seam. The field. Proceedings of the Australasian Institute of Mining and Metallurgy 135, 1–36. nature of the non-kaolinite clay minerals in the coal also changes from Brownfield, M.E., Affolter, R.H., Cathcart, J.D., Johnson, S.Y., Brownfield, I.K., Rice, C.A., 2005. Geologic setting and characterization of coals and the modes of occurrence smectite in the lowermost plies, similar to the underlying floor mate- of selected elements from the Franklin coal zone, Puget Group, John Henry No. 1 rials, to a regularly interstratified illite/smectite component in the sec- mine, King County, Washington, USA. International Journal of Coal Geology 63, tion that includes the tonstein layers. Smectite is again developed, 247–275. however, in the mudstone overlying the main part of the coal seam in Burger, K., Zhou, Y., Tang, D., 1990. Synsedimentary volcanic-ash-derived illite tonsteins in Late Permian coal-bearing formations of southwestern China. Interna- the Catherine Hill Bay exposure. tional Journal of Coal Geology 15, 341–356. The coals above and below the tonsteins are also richer in inerti- Burger, K., Bandelow, F.K., Bieg, G., 2000. Pyroclastic kaolin coal-tonsteins of the Upper nite than in other parts of the seam section. This may indicate a differ- Carboniferous of Zonguldak and Amasra, Turkey. International Journal of Coal Geology 45, 39–53. ent peat-forming environment, possibly one that allowed more Burger, K., Zhou, Y., Ren, Y., 2002. Petrography and geochemistry of tonsteins from the widespread deposition of volcanic ash without interruption by con- 4th Member of the Upper Triassic Xujiahe formation in southern Sichuan Province, temporaneously growing trees (cf. Creech, 1998). Whilst volcanic China. International Journal of Coal Geology 49, 1–17. Craw, D., 1997. Fluid inclusion evidence for geothermal structure beneath the Southern ash may have been admixed with the organic debris at different Alps, New Zealand. New Zealand Journal of Geology and Geophysics 40, 43–52. times during peat accumulation, it may have only been able to form Creech, M., 1998. So where are all the trees? Implication for the origin of coal. Proceed- thin but persistent beds at times when the peat-forming environment ings of the 32nd Newcastle Symposium on Advances in the Study of the Sydney Basin, University of Newcastle, New South Wales, pp. 29–43. was not interrupted by major upright tree growth. Creech, M., 2002. Tuffaceous deposition in the Newcastle Coal Measures: challenging existing concepts of peat formation in the Sydney Basin, New South Wales, Australia. International Journal of Coal Geology 51, 185–214. Acknowledgements Crowley, S.S., Stanton, R.W., Ryer, T.A., 1989. The effects of volcanic ash on the maceral and chemical composition of the C coal bed, Emery Coal Field, Utah. Organic Geo- chemistry 14, 315–331. Thanks are expressed to Irene Wainwright of the UNSW Analytical Crowley, S.S., Ruppert, L.F., Belkin, H.E., Stanton, R.W., Moore, T.A., 1993. Factors affect- Centre for provision of the XRF analyses, to Eugene White of the ing the geochemistry of a thick, subbituminous coal bed in the Powder River Basin: UNSW Electron Microscopy Unit for technical assistance, and to volcanic, detrital, and peat-forming processes. Organic Geochemistry 20, 843–853. Dai, S.F., Zhou, Y.P., Ren, D.Y., Wang, X.B., Li, D., Zhao, L., 2007. Geochemistry and Joanne Wilde of the UNSW School of Biological, Earth and Environ- mineralogy of the Late Permian coals from the Songzao Coalfield, Chongqing, mental Sciences, for thin-section preparation. The authors would southwestern China. Science in China Series D: Earth Sciences 50, 678–688. 110 L. Zhao et al. / International Journal of Coal Geology 113 (2012) 94–110

Dai, S., Ren, D., Zhou, Y., Chou, C.-L., Wang, X., Zhao, L., Zhu, X., 2008a. Mineralogy and Percival, C.J., 1983. A definition of the term ganister. Geological Magazine 120, geochemistry of a superhigh-organic-sulfur coal, Yanshan Coalfield, Yunnan, 187–190. China: evidence for a volcanic ash component and influence by submarine exhala- Ren, D., 1996. Mineral matters in coal. In: Han, D. (Ed.), Coal Petrology of China. tion. Chemical Geology 255, 182–194. Publishing House of China University of Mining and Technology, Xuzhou, pp. 67–77 Dai, S., Tian, L., Chou, C.-L., Zhou, Y., Zhang, M., Zhao, L., Wang, J., Yang, Z., Cao, H., Ren, (in Chinese with English abstract). D., 2008b. Mineralogical and compositional characteristics of Late Permian coals Retallack, G.J., 1977. Triassic palaeosols in the upper Narrabeen Group of New South from an area of high lung cancer rate in Xuan Wei, Yunnan, China: occurrence Wales. Part I: features of the palaeosols. Journal of the Geological Society of Australia and origin of quartz and chamosite. International Journal of Coal Geology 76, 23, 383–399. 318–327. Ruan, C.-D., Ward, C.R., 2002. Quantitative X-ray powder diffraction analysis of clay Dai, S., Wang, X., Zhou, Y., Hower, J.C., Li, D., Chen, W., Zhu, X., Zou, J., 2011. Chemical minerals in Australian coals using Rietveld methods. Applied Clay Science 21, and mineralogical compositions of silicic, mafic, and alkali tonsteins in the late 227–240. Permian coals from the Songzao Coalfield, Chongqing, Southwest China. Chemical Ruppert, L.F., Moore, T.A., 1993. Differentiation of volcanic ash-fall and water-borne Geology 282, 29–44. detrital layers in the Eocene Senakin coal bed, Tanjung Formation, Indonesia. Davis, A., Russell, S.J., Rimmer, S.M., Yeakel, J.D., 1984. Some genetic implications of Organic Geochemistry 20, 233–247. silica and aluminosilicates in peat and coal. International Journal of Coal Geology Ruppert, L.F., Stanton, R.W., Cecil, C.B., Eble, C.F., Dulong, F.T., 1991. Effects of detrital 3, 293–314. influx in the Pennsylvanian Upper Freeport peat swamp. International Journal of Deer, W.A., Howie, R.A., Zussman, J., 1992. An Introduction to the Rock-Forming Coal Geology 17, 95–116. Minerals, Second edition. Longman Scientific and Technical, London. 696 pp. Sandler, A., Harlavan, Y., Steinitz, G., 2004. Early formation of K-feldspar in shallow- Dewison, M.G., 1989. Dispersed kaolinite in the Barnsley Seam coal (U.K.): evidence for marine sediments at near-surface temperatures (southern Israel): evidence from a volcanic origin. International Journal of Coal Geology 11, 291–304. K–Ar dating. Sedimentology 51, 323–338. Diessel, C.F.K., 1965. On the Petrography of Some Australian Tonsteins. In: Schmidt- Simmons, S.F., Browne, P.R.L., 1997. Saline fluid inclusions in sphalerite from the Thomé, P., Schönenberg, R. (Eds.), Max Richter-Festschrift. University of Broadlands-Ohaaki geothermal system; a coincidental trapping of fluids being Clausthal-Zellerfeld, pp. 149–166. boiled toward dryness. Economic Geology 92, 485–489. Diessel, C.F.K., 1985. Tuffs and tonsteins in the coal measures of New South Wales, Australia. Simpson, M.P., Mauk, J.L., Simmons, S.F., 2001. Hydrothermal alteration and hydrologic Proceedings of the 10th International Congress on Carboniferous Stratigraphy and Geol- evolution of the Golden Cross epithermal Au–Ag deposit, New Zealand. Economic ogy, Madrid, Madrid, pp. 197–210. Geology 96, 773–796. Diessel, C.F.K., Hutton, A.C., 2004. Field trip B — an overview of the Newcalstle Coal Spears, D.A., 1971. The mineralogy of the Stafford tonstein. Proceedings of the York- Measures. 21st Annual Meeting of the Society for Organic Petrology, 27 September shire Geological Society 38, 497–516. – 1 October, 2004, Sydney, Australia. 43 pp. Spears, D.A., 2006. Clay mineralogy of onshore UK Carboniferous mudrocks. Clay Minerals Folk, R.L., 1974. Petrology of Sedimentary Rocks. Hemphill, Austin, Texas. 182 pp. 41, 395–416. Frazer, F.W., Belcher, C.B., 1973. Quantitative determination of the mineral-matter Spears, D.A., Kanaris-Sotiriou, R., 1979. A geochemical and mineralogical investigation content of coal by a radiofrequency–oxidation technique. Fuel 52, 41–46. of some British and other European tonsteins. Sedimentology 26, 407–425. Grevenitz, P., 2003. Origin, alteration and geochemical correlation of Late Permian Standards Australia, 2000. Coal and coke—analysis and testing: higher rank coal– airfall tuffs in coal measures, Sydney Basin, Australia. International Journal of mineral matter and water of constitution. Australian Standard 1038.22. Standards Coal Geology 55, 27–46. Australia, Homebush. 20 pp. Gupta, R., Wall, T.F., Baxter, L.A. (Eds.), 1999. The Impact of Mineral Impurities in Solid Sykes, R., Lindqvist, J.K., 1993. Diagenetic quartz and amorphous silica in New Zealand Fuel Combustion. Plenum, New York. 768 pp. coals. Organic Geochemistry 20, 855–866. Hagen, E., Kelley, S.P., Dypvik, H., Nilsen, O., Kjolhamar, B., 2001. Direct dating of authi- Taylor, J.C., 1991. Computer programs for standardless quantitative analysis of genic K-feldspar overgrowths from the Kilombero Rift of Tanzania. Journal of the minerals using the full powder diffraction profile. Powder Diffraction 6, 2–9. Geological Society 158, 801–807. Triplehorn, D.M., Stanton, R.W., Ruppert, L.F., Crowley, S.S., 1991. Volcanic ash Hemingway, J.E., 1968. Sedimentology of coal-bearing strata. In: Murchison, D., Westoll, dispersed in the Wyodak-Anderson coal bed, Powder River Basin, Wyoming. T.S. (Eds.), Coal and Coal-bearing Strata. Oliver and Boyd, Edinburgh, pp. 43 –69. Organic Geochemistry 17, 567–575. Herbert, C., 1980. Depositional development of the Sydney Basin. In: Herbert, C., Helby, Warbrooke, P.R., 1987. Depositional and chemical environments in Permian coal form- R.E. (Eds.), A Guide to the Sydney Basin: Geological Survey of New South Wales, ing swamps from the Newcastle area. In: Engel, B.A. (Ed.), Proceedings of the 21st Bulletin, 26, pp. 11–52. Newcastle Symposium, “Advances in the study of the Sydney Basin”. Department Hill, P.A., 1988. Tonsteins of Hat Creek, British Columbia: a preliminary study. Interna- of Geology, University of Newcastle, Callaghan, NSW, Australia, pp. 1–9. tional Journal of Coal Geology 10, 155–175. Ward, C.R. (Ed.), 1984. Coal Geology and Coal Technology. Blackwell, Melbourne. 345 Hower, J.C., Ruppert, L.F., Eble, C.F., 1999. Lanthanide, yttrium, and zirconium anomalies in pp. the Fire Clay coal bed, Eastern Kentucky. International Journal of Coal Geology 39, Ward, C.R., 1989. Minerals in bituminous coals of the Sydney Basin (Australia) and the 141–153. Illinois Basin (U.S.A.). International Journal of Coal Geology 13, 455–479. Klein, C., Dutrow, B., 2007. Manual of Mineral Science, Twenty third ed. John Wiley and Ward, C.R., 1991. Mineral matter in low-rank coals and associated strata of the Mae Sons. 675 pp. Moh basin, northern Thailand. International Journal of Coal Geology 17, 69–93. Knight, J.A., Burger, K., Bieg, G., 2000. The pyroclastic tonsteins of the Sabero Coalfield, Ward, C.R., 2002. Analysis and significance of mineral matter in coal seams. Interna- north-western Spain, and their relationship to the stratigraphy and structural ge- tional Journal of Coal Geology 50, 135–168. ology. International Journal of Coal Geology 44, 187–226. Ward, C.R., Bradley, G.M., Skilbeck, C.G., Stutchbury, R., 1986. A Guide to Cored Rocks in Kramer, W., Weatherall, G., Offler, R., 2001. Origin and correlation of tuffs in the the Sydney Basin. Earth Resources Foundation, University of Sydney. 150 pp. Permian Newcastle and Wollombi Coal Measures, NSW, Australia, using chemical Ward, C.R., Corcoran, J.F., Saxby, J.D., Read, H.W., 1996. Occurrence of phosphorus fingerprinting. International Journal of Coal Geology 47, 115–135. minerals in Australian coal seams. International Journal of Coal Geology 30, 185–210. Li, Z., Ward, C.R., Gurba, L.W., 2007. Occurrence of non-mineral inorganic elements in Ward, C.R., Spears, D.A., Booth, C.A., Staton, I., Gurba, L.W., 1999. Mineral matter and low-rank coal macerals as shown by electron microprobe element mapping tech- trace elements in coals of the Gunnedah Basin, New South Wales, Australia. Inter- niques. International Journal of Coal Geology 70, 137–149. national Journal of Coal Geology 40, 281–308. Li, Z., Ward, C.R., Gurba, L.W., 2010. Occurrence of non-mineral inorganic elements in Ward, C.R., Taylor, J.C., Matulis, C.E., Dale, L.S., 2001. Quantification of mineral matter in macerals of low-rank coals. International Journal of Coal Geology 81, 242–250. the Argonne Premium Coals using interactive Rietveld-based X-ray diffraction. Liu, J., Hay, R.L., Deino, A., Kyser, T.K., 2003. Age and origin of authigenic K-feldspar in International Journal of Coal Geology 46, 67–82. uppermost Precambrian rocks in the North American Midcontinent. Bulletin of Ward, C.R., Li, Z., Gurba, L.W., 2005. Variations in coal maceral chemistry with rank ad- the Geological Society of America 115, 422–433. vance in the German Creek and Moranbah Coal Measures of the Bowen Basin, Aus- Loughnan, F.C., 1971. Kaolinite claystones associated with the Wongawilli seam in the tralia, using electron microprobe techniques. International Journal of Coal Geology southern part of the Sydney Basin. Journal of the Geological Society of Australia 18, 63, 117–129. 293–302. Ward, C.R., Li, Z., Gurba, L.W., 2007. Variations in elemental composition of macerals Loughnan, F.C., Corkery, R.W., 1975. Oriented-kaolinite aggregates in flint clays and with vitrinite reflectance and organic sulphur in the Greta Coal Measures, New kaolin tonsteins of the Sydney Basin, New South Wales. Clay Minerals 10, 471–473. South Wales, Australia. International Journal of Coal Geology 69, 205–219. Loughnan, F.C., Ray, A.S., 1978. The Reid's mistake formation at Swansea Head, New Zhang, Z., Mao, J., Wang, Y., Pirajno, F., Liu, J., Zhao, Z., 2010. Geochemistry and geochro- South Wales. Journal of the Geological Society of Australia 25, 473–481. nology of the volcanic rocks associated with the Dong'an adularia-sericite epither- Loughnan, F.C., Ward, C.R., 1971. Pyrophyllite-bearing flint clay from the Cambewarra mal gold deposit, Lesser Hinggan Range, Heilongjiang province, NE China: area, New South Wales. Clay Minerals 9, 83–95. Constraints on the metallogenesis. Ore Geology Reviews 37, 158–174. Norrish, K., Hutton, J.T., 1969. An accurate X-ray spectrographic method for the analysis of Zhou, Y., Ren, Y., Bohor, B.F., 1982. Origin and distribution of tonsteins in Late Permian a wide range of geological samples. Geochimica et Cosmochimica Acta 33, 431–453. coal seams of Southwestern China. International Journal of Coal Geology 2, 49–77.