Petrography, Mineralogy, and Geochemistry of Thermally Altered Coal in the Tashan Coal Mine, Datong Coalfield, China

Abstract

A suite of coal samples near a diabase dike were collected to investigate the petrographic, mineralogical, and geochemical characteristics of thermally altered coal in Datong Coalfield, China. Proximate analysis, vitrinite reflectance measurement, and petrographic analysis were applied to identify and characterize the alteration halo; optical microscope observation, qualitative X-ray diffractometry, and SEM-EDS were applied to study the phases, occurrence, and composition of minerals; XRF, ICP-MS, and AFS were applied to determine concentrations of major and trace elements; and the occurrence modes of elements were studied by correlation and hierarchical cluster analysis as well as SEM-EDS. The results demonstrated that the 3.6 m dike has caused an alteration halo of approximately 2 m in diameter. In addition, the thermally altered coals were characterized by high vitrinite reflectance, low volatile matter, and the occurrence of thermally altered organic particles. Dolomite and ankerite in the thermally altered coal may be derived from hydrothermal fluids, while muscovite and tobelite may be transformed from a kaolinite precursor. The average concentration of Sr in the Tashan thermally altered coal reached 1714 μg/g, which is over 12 times that of the Chinese coal; the phosphate minerals and Sr-bearing kaolinite account for this significant enrichment. The cluster analysis classified elements with geochemical associations into four groups: group 1 and 2 were associated with aluminosilicates, clays, and carbonates and exhibited enrichment in the coal/rock contact zone, indicating that the dike may be the source of the elements; group 3 included P₂O₅, Sr, Ba, and Be, which fluctuate in coals, suggesting that their concentrations were influenced by multiple-factors; group 4 did not manifest obvious variations in coals, implying that the coal itself was the source.

1. Introduction

Igneous intrusions in coal-bearing sedimentary basins have been reported in many countries, such as the USA [1,2,3], South Africa [4,5], Australia [6,7,8], Spain [9], India [10,11,12,13], Turkey [14,15,16], and China [17,18,19,20,21]. Igneous intrusions in coal-bearing strata greatly affect the safety of the mining process and have impacts on coal quality, petrography, mineralogy, and geochemistry.

The heating and expansion of igneous intrusions and the cooling and contraction of the surrounding rocks result in massive joints and fractures in coals [22], inducing enhanced permeability and declined stability of coal seams, which may lead to gas outbursts and roof falls [23]. Igneous intrusions reduce the economic viability of coal mines because of the difficult mining conditions they create. In addition, thermally altered coal, which is not marketable, may be abandoned in the practical mining process since it generally has high ash yields, low calorific values, and difficulties with combustion.

Coal is a type of sedimentary rock that is sensitive to changing temperatures and stress that may cause physical and chemical variations. Heating from igneous intrusions petrographically alters the coal. Thermally altered coal is macroscopically dull and dense [10,14]; microtextures such as mesophase spheres, mosaics, flow structures, pyrolytic carbon, and graphitic sphaeroliths occur because of contact metamorphism [12,24,25,26,27]. Thermally altered coal exhibits an increase in reflectance and thus coal rank [28,29]; the coal may be altered to natural coke and in some cases even undergo natural graphitization in the immediate area of the igneous body [30,31,32]. However, coals thermally altered by igneous intrusions may follow a different maturation pathway and exhibit subtle changes compared with unaltered coal [33,34,35].

Paleoenvironment and geological evolution are important factors affecting coal mineralogy, and igneous intrusions may substantially change the mineralogy of thermally altered coals in two ways. First, the original minerals may be altered [36]; for example, Chen et al. [37] suggested that the clinochlore and halloysite in the thermally altered coals may be alteration products of aluminosilicates. Moreover, Li et al. [38] stated that chlorite, muscovite, and illite in anthracite had a precursor of kaolinite and that was possibly formed by alteration of pre-existing kaolinite by hydrothermal solutions. Second, some minerals may be deposited directly from the hydrothermal fluids; for instance, Dai et al. [27] reported hydrothermally derived alabandite (MnS) and chalcopyrite in thermally altered coals.

Coal may be one of the most complex geological materials and includes almost all of the elements in the periodic table [39]. Although Finkelman et al. [2] stated that studies on the effects of igneous intrusions on inorganic constituents of coal are scarce, many studies have been conducted in recent years [9,19,40,41,42]. Elements may enrich, deplete, or exhibit no variations in coal thermally altered by contact metamorphism [43,44]. The enriched elements may be introduced into coal by hydrothermal fluids or groundwater [37], whereas the depleted elements may be vaporized or leached out [2]. However, Dai et al. [45] stated that the migration mechanisms of elements between intrusions and coals are still not clear.

As suggested by Finkelman et al. [2], “intrusion of magma into coal offers the opportunity to study the mobilization of potentially hazardous elements under high temperature, reducing conditions that simulate in situ gasification conditions.” A suite of coal samples near a diabase dike were systematically collected in this study, and an integrated approach comprising petrographic, mineralogical, and geochemical data were used to achieve: (i) the identification and characterization of the alteration halo; (ii) the identification of crystalline phase, occurrence modes of minerals, and mineralogical composition; (iii) the evaluation of concentrations and distributions and the assessment of affinities and modes of occurrence of elements, as along with the potential sources and geological factors that may have caused their enrichment or depletion.

2. Geological Setting

Datong Coalfield is located in the north of Shanxi Province [46], China, bordering Inner Mongolia to the north. The northeast portion of the coalfield is bounded by the Qingciyao fault, while the west, south, and southeast borders contain coal seam outcrops. The Cambrian–Ordovician strata are the basement of the Datong coalfield and overlain by Carboniferous–Permian and Jurassic coal-bearing strata and Cretaceous, Neogene, and Quaternary strata (Figure 1). Structurally, the coalfield is a large asymmetrical syncline; its southeast flank steeper is than its northwest. The Tashan Coal Mine is located 30 km southwest of Datong City in the middle portion of the east flank of the Datong syncline. The coal mine is 20 km long and 12.5 km wide, covering an area of 170.9-km² that extends from 39°53’23” N to 40°00’09” N and 112°49’23” E to 113°04’03” E. Structurally, the region displays single oblique faults with a strike of 10–50 NE and a dip to the NW; some small-scale anticlines and synclines develop locally.

2.1. Coal-Bearing Strata

The coal-bearing strata of the Datong coalfield include the Jurassic Datong, Lower Permian Shanxi, and Upper Carboniferous Taiyuan formations (Figure 2). At present, the Jurassic coal seam has been exhausted, and exploitation has turned to the underlying Carboniferous–Permian coal seams [21].

The Taiyuan formation has a thickness ranging from 33.2 to 138.25 m, with an average of 96.27 m. It contains 10 coal seams (nos. 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10), of which nos. 2, 3, 5, and 8, with an average accumulation thickness of 22 m, are mineable. The no. 3 coal seam splits into 3⁻¹ and 3⁻² coal seams in the west of the Tashan Coal Mine, whereas the no. 3 and 5 coal seams merge in the east of the region and are referred to as the 3–5 coal seam.

The Shanxi formation conformably overlies the Taiyuan formation. Its thickness ranges up to 96.1 m, with an average of 67.2 m. The coal seams include Shan₁, Shan₂, Shan₃, and Shan₄, of which only the Shan₄ coal seam is mineable, with an average thickness of 2.63 m.

2.2. Igneous Intrusion

Datong Coalfield was intruded by igneous intrusions in both the Indosinian (Late Permian to the end of the Triassic) and the Yanshan (Late Jurassic to the Early Cretaceous) periods [47,48,49,50,51]. Intrusions dating back to the Indosinian period are dominated by mafic to ultramafic lamprophyre and carbonatite [49,51], whereas rocks that intruded in the Yanshan period are mainly diabase and basaltic andesite [52]. The detailed lithological characteristics of the igneous intrusions in the Datong coalfield were described in Chen et al. [21]. In the practical mining process, sills and dikes were encountered in the Shan₄, 2, 3⁻¹, 3⁻², 5(3–5), and 8 coal seams in the Tashan Coal Mine. The large-scale igneous intrusions in coal seam affect the safety of coal mines by causing roof instability. For example, two severe roof falling incidents induced by igneous intrusion emplacement were recorded in the Tashan Coal Mine in 2005 and 2020. Moreover, they play a deleterious role in economic efficiency by reducing the mineable thickness and altering coal into cinder or natural coke, which are not marketable or combustible.

3. Materials and Methods

3.1. Sampling

The 8222 and 5222 tunnels of the Tashan Coal Mine parallel to each other and encountered the same diabase dike. The detailed characteristics of the dike were described in Song et al. [53]. Although we did not obtain photographs of the dike in the 5222 tunnel because of poor lighting conditions, the overlying dike in the 8222 tunnel exhibited the same characteristics (Figure 3). The 5222 dike is 3.6 m in thickness, comprising a 2.6 m dark green core and two pale chill zones on each side. A suite of 16 coal samples was collected along the 5222 tunnel following the Chinese Standard Method GB/T 482-2008 [54]. The coal samples were collected at essentially the same height along the transect at closely spaced intervals in the vicinity of the dike and larger intervals farther away from the dike, up to a distance of approximately 20 m. All collected samples were immediately sealed in plastic bags to minimize contamination and oxidation.

3.2. Methods

The coal samples were subjected to proximate analysis according to ASTM Standard for moisture (ASTM D3173/D3173M-17a [55]), ash yield (ASTM D3174-12(2018)e1 [56]), and volatile matter (ASTM D3175-20 [57]). Particulate blocks for petrographic analysis were prepared according to the standard procedures described in ISO 7404-2 [58]. The mean maximum and random reflectance of vitrinite were measured using a Leica DM4500P microscope equipped with a DETA V4000 Spectrometer according to ISO 7404-5 [59]. Point counts were conducted under incident white light, immersion oil, at a 500× magnification using 500 counts per block and two blocks per sample to quantify the contents of the different macerals and evaluate the extent of thermal alteration. The identification of macerals followed the ICCP system 1994 proposed by the International Committee for Coal and Organic Petrology for vitrinite [60], inertinite [61], and liptinite macerals [62]. The classification of organic particles in thermally altered coals was performed using the ICCP classification system [63].

Optical microscopic observation, powder X-ray diffractometry (XRD), and scanning electron microscope analysis (SEM) were applied to study the mineralogical characteristics of the Tashan coal. Qualitative XRD analysis was applied to confirm the main mineralogical phases. Powdered coal samples were examined on a Bruker D8 advance powder X-ray Cu-Kα radiation diffractometer. The XRD pattern was recorded over a 2θ interval of 5–70° with a step size of 0.02°. The resulting diffractogram peaks were qualitatively analyzed using HighScore Plus software. Scanning electron microscopes (Hitachi SU8010 and Bruker Quanta FEG 450, 1kV and 20 kV accelerating voltage) in conjunction with an energy-dispersive X-ray spectrometer (SEM-EDS) were used to study the morphology, distribution, and composition of the minerals in the coal.

Samples were crushed and ground to pass 200 mesh for geochemical analysis. Concentrations of major and trace elements in the Tashan coal were determined at the Analytical Laboratory Beijing Research Institute of Uranium Geology following the methods described in Chinese Standards GB/T 14506-2010 [64]. The contents of the major element oxides, including SiO₂, Al₂O₃, Fe₂O₃, MgO, CaO, Na₂O, K₂O, MnO, TiO₂, and P₂O₅, were determined by X-ray fluorescence spectrometry (PANalytical Axios-max); high resolution inductively coupled plasma mass spectrometry (Thermo Fisher ELEMENT XR) was used to determine the concentrations of elements, such as Li, Be, Sc, V, Cr, Co, Ni, Cu, Zn, Ga, Rb, Sr, Mo, Cd, In, Sb, Cs, Ba, W, Tl, Pb, Bi, Th, U, Nb, Ta, Zr, and Hf; the concentrations of Se and As in the Tashan coal were determined using liquid chromatography-atomic fluorescence spectrometry (LC-AFS 6500).

4. Results and Discussion

4.1. Identification and Characterization of the Alteration Halo

An alteration halo was induced in the ambient area of the dike because of thermal effects, in which the samples exhibited high vitrinite reflectance, abnormal coal composition (ash and volatile matter), and changes in petrographic characteristics. Identification and characterization of the alteration halo were performed following the standard coal classification methods described in ISO 11760-2018 [65] and ASTM D 388-19a [66].

4.1.1. Vitrinite Reflectance

Vitrinite reflectance increased sharply in the immediate area of the intrusive body (

Table 1. The results of the proximate analysis and vitrinite reflectance measurements of the Tashan coal (in %).

Sample	Distance (m)	Rr	Rmax	Mad	Ad	VMdaf	FCdaf
T1	Contact zone	3.45	3.67	4.1	27.5	11.9	88.1
T2	0.1	3.29	3.50	4.8	25.5	7.4	92.6
T3	0.6	2.69	2.86	2.0	24.4	8.3	91.7
T4	1.1	1.75	1.87	1.1	11.8	19.7	80.3
T5	1.6	1.47	1.57	1.2	11.7	21.6	78.4
T6	2.5	0.74	0.79	1.8	8.1	28.9	71.2
T7	2.7	0.88	0.94	1.4	9.7	27.2	72.8
T8	4.4	0.70	0.75	1.8	6.6	33.2	66.8
T9	4.5	0.71	0.75	1.9	7.6	34.6	65.5
T10	6.3	0.68	0.72	1.9	8.6	49.9	50.1
T11	7.1	0.68	0.73	1.9	8.3	29.7	70.3
T12	9.9	0.63	0.67	1.6	9.3	38.0	62.0
T13	10	0.75	0.79	1.7	9.6	33.0	67.0
T14	13.6	0.68	0.73	1.8	10.3	33.6	66.4
T15	16.6	0.82	0.87	1.8	14.2	29.1	70.9
T16	17.3	0.73	0.78	1.2	11.8	35.8	64.2

R_r—vitrinite random reflectance; R_max—vitrinite maximum reflectance; M_ad—moisture, on air-dry basis; A_d—ash yield, on dry basis; VM_daf—volatile matter, on dry and ash-free basis; FC_daf—fixed carbon, on dry and ash-free basis.

, Figure 4), indicating an upgrading coal rank approaching the dike [3,67]. According to ISO 11760-2018 [65], T1–T3 are high-rank coals (or anthracite), whereas T1 and T2 are high-rank B coal (or anthracite B coal with a mean vitrinite random reflectance of ≥3.0% and <4.0 %), T3 is high-rank C coal (or anthracite C coal with a mean vitrinite random reflectance of ≥2.0% and <3.0%); T4 and T5 are medium rank B coal (or bituminous B coal with a mean vitrinite random reflectance of ≥1.0% and <1.4%); T6–T16 are medium rank C coals (or bituminous C coal with a mean vitrinite random reflectance of ≥0.6% and <1.0%).

The vitrinite maximum reflectance (R_max) is also indicative of coal rank upgrading. The coal within 1.6 m of the dike (T1–T5) had a R_max higher than 1%, whereas the R_max fluctuated between 0.67% and 0.94% in the more distant areas.

4.1.2. Proximate Analysis

The ash yields (dry basis) of the Tashan coal ranged from 8.06% to 26.39% and exhibited an increasing trend when approaching the dike (Table 1, Figure 4). According to ISO 11760-2018 [65], T1–T3 are moderately high-ash coal with ash yields of ≥20% and <30%, T4–T5 and T14–T16 are medium-ash coal with ash yields of ≥10% and <20%, and the others are low-ash coal with ash yields of ≥5% and <10%.

The volatile matter yield (dry, ash-free basis) of the Tashan coal ranged from 7.36% to 49.92% and increased with increasing distance from the intrusive body (Table 1, Figure 4). As suggested by the American standard ASTM D 388-19a [66], samples T1 and T3 are semianthracite, T2 is anthracite, T4 and T5 are low volatile bituminous coal, and T6–T16 are medium to high volatile bituminous coal.

4.1.3. Petrographic Characteristics

Intruding magma caused morphological and optical changes in the organic particles of thermally altered coals; for example, liptinite becomes difficult to distinguish [28,29,33,34,35], and porous and massive altered organic particles as well as newly formed components occur [63]. These altered organic particles indicate that coals experienced plastic phases at high temperatures during the intrusive event [36], differentiating thermally altered coals from unaltered coals. Macerals of the Tashan thermally altered coal exhibited various melting and plasticity characteristics (Figure 5); for instance, vitrinite was converted into porous coke in the adjacent area of the dike; the edges of the macerals became rounded, which may have been caused by melting; and fissures and cracks became prevalent in the coal, inducing a ‘fish bone’ structure in some isolated carbonized vitrinite [68].

Quantitative analysis was carried out to evaluate the extent of thermal alteration (Figure 6,

Table 2. Quantification of organic particles in thermally altered coal (vol. %).

Sample	Non-Altered Organic Particles			Altered Organic Particles		Newly Formed Organic Particles
	Vitrinite	Inertinite	Liptinite	Porous	Massive	Pyrolytic Carbon
T1	23.7	19.7	-	24.0	32.1	0.5
T2	21.4	16.2	-	18.6	42.6	1.2
T3	15.5	18.3	-	15.1	49.7	1.4
T4	19.8	12.3	0.3	4.4	62.8	0.4
T5	60.5	14.5	0.4	1.3	22.5	0.8

). For example, the contents of non-altered organic particles ranged from 32.4% to 43.4% in T1–T4 and reached a peak at 75.4% in T5; the contents of altered organic particles, however, exhibited a reverse trend compared with non-altered particles, ranging between 56.1% and 67.2% in T1–T4, followed by a sharp decrease to 23.8% in T5; and newly formed organic particles only accounted for a small proportion of particles, and no obvious trend was observed in T1–T5. The quantification of the organic particles in thermally altered coal demonstrated that thermal alteration dissipated at a further distance; fewer altered organic particles in T5 could be a sign of the heat pulse running out.

Macerals in T6–T16 showed prominent isotropy, indicating little or no thermal alteration.

Table 3. Quantification of macerals in unaltered coal (vol. %).

Sample No.	T	CT	CD	CG	G	VD	TV	SF	F	Mac	Mic	FG	ID	TI	SP
T6	15.4	9.2	36.1	2.0	3.6	1.6	67.9	7.9	8.9	1.3	4.3	-	2.3	24.6	7.5
T7	18.8	18.8	30.7	-	1.0		69.3	9.4	10.4	2.0	5.9	0.5	0.5	28.7	2.0
T8	10.8	12.6	33.9	-	0.7	0.7	58.8	12.6	7.2	2.5	7.9	-	2.2	32.5	8.7
T9	15.0	12.3	33.3	0.3	0.6	2.4	64.0	15.3	13.2	1.5	2.1	0.3	0.9	33.3	2.7
T10	17.0	13.8	36.4	0.4	1.8	2.5	71.7	11.3	8.5	3.2	1.4	-	0.7	25.1	3.2
T11	11.7	10.8	28.6	-	1.5	1.5	54.2	15.4	10.8	1.5	5.5	-	3.4	36.6	9.2
T12	21.0	8.2	22.7	2.7	3.1	1.0	58.8	17.5	10.0	1.0	4.8	-	2.4	35.7	5.5
T13	19.8	5.1	36.3	0.4	1.1	4.0	66.7	9.9	7.3	1.8	3.7	0.4	1.5	24.5	8.8
T14	25.7	13.8	23.8	1.1	1.1	1.1	66.7	11.9	11.1	1.1	1.9	-	2.7	28.7	4.6
T15	10.8	2.4	50.0	2.0	3.0	2.0	70.3	8.1	7.4	4.4		-	6.1	26.0	3.7
T16	28.1	11.7	32.4	0.3	-	1.9	74.4	13.9	5.9	0.9	2.2	-	0.6	23.5	2.2
Av.	17.6	10.8	33.1	1.2	1.8	1.9	65.7	12.1	9.1	1.9	4.0	-	2.1	29.0	5.3

T—Telinite; CT—Collotellinite; CD—Collodetrinite; CG—Corpogelinite; G—Gelinite; VD—Vitrodetrinite; TV—Total Vitrinite; SF—Semifusinite; F—Fusinite; Mac—Macrinite; Mic—Micrinite; FG—funginite; ID—Inertodetrinite; TI—Total Inertinite; SP—Sporinite; MM—Mineral matter; bdl—below detection limit; Trace—Trace amount; Av.—Average content.

summarizes the point counting analysis data of T6–T16, and the results indicate that vitrinite, inertinite, and liptinite in the unaltered coal accounted for an average of 65.7%, 29%, and 5.3%, respectively. The vitrinite group mainly comprised collodetrinite, telinite, and collotellinite, with average contents of 33.1%, 17.6%, and 10.8%, respectively; the inertinite group was mainly composed of semifusinite and fusinite, with average contents of 12.1% and 9.1%, respectively, whereas macrinite and micrinite only accounted for a small proportion of the contents; and liptinite was represented by sporinite. According to the classification by petrographic composition described in ISO 11760-2018 [65], most of the Tashan coals are of moderately high vitrinite, except T8, T11, and T12, which are medium vitrinite coals.

4.1.4. Characterization of the Alteration Halo

Both the ISO and ASTM classification methods indicated an elevated coal rank in the vicinity of the dike; however, the results were not very consistent. It was suggested that the T1–T3 samples were in the anthracite stage on the basis of ISO 11760-2018 [65], while ASTM D388-19a [66] classified T1 and T3 as semianthracite. This is likely because T1 and T3 have considerable quantities of carbonate minerals that can decompose during the volatile matter determination process, affecting the results to some extent [69,70]. In the present study, two methods as well as thermally altered coal petrographic analysis were combined to identify and characterize the alteration halo (

Table 4. Classification of the Tashan coal.

Sample	Description Term		Nature
	ISO 11760-2018 a	ASTM D388-19a b
T1	Anthracite B, moderately high ash	Semianthracite	Thermally-altered
T2	Anthracite B, moderately high ash	Anthracite	Thermally-altered
T3	Anthracite C, moderately high ash	Semianthracite	Thermally-altered
T4	Bituminous B, medium ash	Low volatile bituminous	Thermally-altered
T5	Bituminous B, medium ash	Low volatile bituminous	Thermally-altered
T6	Bituminous C, low ash, moderately high vitrinite	Medium volatile bituminous	Unaltered
T7	Bituminous C, low ash, moderately high vitrinite	Medium volatile bituminous	Unaltered
T8	Bituminous C, low ash, medium vitrinite	High volatile bituminous	Unaltered
T9	Bituminous C, low ash, moderately high vitrinite	High volatile bituminous	Unaltered
T10	Bituminous C, low ash, moderately high vitrinite	High volatile bituminous	Unaltered
T11	Bituminous C, low ash, medium vitrinite	High volatile bituminous	Unaltered
T12	Bituminous C, low ash, medium vitrinite	High volatile bituminous	Unaltered
T13	Bituminous C, low ash, moderately high vitrinite	High volatile bituminous	Unaltered
T14	Bituminous C, medium ash, moderately high vitrinite	High volatile bituminous	Unaltered
T15	Bituminous C, medium ash, moderately high vitrinite	Medium volatile bituminous	Unaltered
T16	Bituminous C, medium ash, moderately high vitrinite	High volatile bituminous	Unaltered

^a based on ISO 11760-2018 [65]; ^b based on ASTM D388-19a [66].

). Accordingly, thermally altered coals had an R_max of >1%; coals adjacent to the intrusion (T1–T3, <0.6 m) were altered to anthracite or semianthracite with low volatile matter and moderately high ash; T4 and T5, which were slightly farther away from the dike, were altered to bituminous B coal with low volatile matter; and the unaltered coal was bituminous C coal with medium-high volatile matter, low-medium ash, and an R_max lower than 1%. Thermally altered organic particles were only observed in T1–T5, whereas macerals in T6–T16 did not manifest thermal alteration. Overall, proximate analysis, vitrinite reflectance, and petrographic characteristics consistently suggested that the edge of the alteration halo is located between T5 and T6.

4.2. Mineralogical Characteristics

4.2.1. Mineral Composition

Table 5. Minerals identified in the Tashan coal.

Sample	Distance (m)	Nature	Minerals
T1	Contact	Thermally-altered	Dolomite, ankerite, muscovite, kaolinite, quartz, calcite
T2	0.1	Thermally-altered	Kaolinite, dolomite, ankerite, anhydrite, muscovite, quartz
T3	0.6	Thermally-altered	Dolomite, muscovite, kaolinite, quartz, ankerite
T4	1.1	Thermally-altered	Kaolinite, muscovite, calcite, tobelite
T5	1.6	Thermally-altered	Kaolinite, dolomite, ankerite, tobelite
T6	2.5	Unaltered	Kaolinite, quartz
T7	2.7	Unaltered	Kaolinite, quartz
T8	4.4	Unaltered	Kaolinite, calcite, dolomite
T9	4.5	Unaltered	Kaolinite, dolomite, calcite, quartz
T10	6.3	Unaltered	Kaolinite, calcite, quartz, dolomite
T11	7.1	Unaltered	Kaolinite, quartz, analcime
T12	9.9	Unaltered	Kaolinite, calcite
T13	10	Unaltered	Kaolinite, calcite, quartz, diopside, ankerite
T14	13.6	Unaltered	Quartz, kaolinite, calcite, dolomite
T15	16.6	Unaltered	Kaolinite, dolomite, calcite
T16	17.3	Unaltered	Calcite, Kaolinite, boehmite, dolomite, quartz

Contact-rock/coal contact zone.

shows the main crystalline phases identified from the diffractograms of the Tashan coal. Minerals found in the unaltered coal were dominated by kaolinite, quartz, calcite, and dolomite and were less abundant in ankerite, analcime, diopside, and boehmite. Minerals identified in the thermally altered coal included kaolinite, dolomite, ankerite, quartz, muscovite, tobelite (NH₄-illite), and anhydrite.

4.2.2. Occurrence and Origin of Mineral Matter

1.

Kaolinite, Muscovite, and Tobelite

Clay minerals in the Tashan coal are dominated by kaolinite, which was identified in all samples. Authigenic kaolinite occurring as cell fillings or replacing the organic constituent was found in both thermally altered and unaltered coals, for example, in the form of detrital pellets as well as book-like and vermicular aggregates (Figure 7a–c). This occurrence may indicate that authigenic kaolinite formed during peat accumulation or during early diagenetic processes [71,72]. However, kaolinite filling in the cleats and fractures as well as in devolatilization vacuoles was also observed in thermally altered coals (Figure 7d–e), indicating its epigenetic origin.

Kaolinite can be transformed into chlorite, muscovite, and illite with elevated coal ranks [38,73,74,75,76]. According to the XRD results, muscovite occurred in the thermally altered coals T1–T4, which is consistent with the conclusion drawn by Golab and Carr [44] that muscovite is indicative of thermal alteration. In the present study, muscovite was intimately intergrown with kaolinite (Figure 7f), and the identification of muscovite was supported by SEM-EDS observations. The occurrence modes of muscovite indicate its epigenetic formation by the transformation of a kaolinite precursor.

The XRD pattern indicates the occurrence of tobelite (NH₄-Illite) in T4 and T5, which was further demonstrated by SEM analysis. NH₄-Illite has been found in some high-rank coals [76,77,78]. It was suggested by Dai et al. [77] that ammonian illite may form because of the interaction between kaolinite and nitrogen released from the organic matter at high temperatures during an intrusive event. In the present study, cleat-filling tobelite with flake morphology was observed through SEM analysis (Figure 7g). The identification of tobelite was supported by the SEM-EDS spectra in Figure 7h, and similar results were reported by Dai et al. [79]. The occurrence modes of tobelite differentiate it from other clays and may suggest an epigenetic origin.

2.

Quartz

According to the XRD results, quartz occurred in most samples and was lacking in some samples (e.g., T4, T5, T8, T12, and T15), which may be because the quartz content was below the detection limit. However, quartz was rarely observed by optical analysis except in the T1 sampled directly adjacent to the intrusion, where, for example, epigenetic quartz occurring as cell- and fracture-fillings was observed (Figure 8). Chen et al. [37] reported quartz veins in thermally altered coals intruded by syenite porphyry and lamprophyre sills from the Huainan Coalfield, where the epigenetic quartz was derived from Si-rich hydrothermal fluids. In the present study, the restricted epigenetic mineralization of quartz reflects that the mafic diabase dike was unlikely to provide massive silicon-rich hydrothermal fluids during the intrusive event.

3.

Carbonate minerals

Dolomite and ankerite were more prevalent than calcite in the thermally altered coal, while carbonates in the unaltered coal were dominated by calcite (Table 5). Because of a lack of epigenetic mineralization, carbonates in unaltered coal were mainly of syn-depositional origins, such as the calcite filling in the fusinite (Figure 9). However, most dolomite and ankerite in thermally altered coal formed epigenetically, and this view has been supported by many studies [2,16,22,33,76,80,81,82]. It is generally suggested that epigenetic carbonate minerals form through the reaction of magma or hydrothermal solution with CO and CO₂ released from coal at high temperatures during an intrusive event [2,37,71,72]. Dai et al. [77] suggested that during coal metamorphism, calcium released from the organic matter may interact with CO₂ in the pore water, forming carbonates. In the Tashan coals, epigenetic dolomite and ankerite included considerable quantities of Fe, Mg, and Ca, indicating that these minerals may be deposited by hydrothermal fluids.

Epigenetic carbonate filling in the devolatilization vacuoles and cleats of macerals was observed in the Tashan coals, providing the following insights into the interactions between coal and rocks during the emplacement event. (i) The occurrence of thermal alteration microtextures demonstrates temperatures of more than 400 °C [83,84]. Volatile matter was released at this temperature (see Table 1, reduced V_daf value of thermally altered coal), and devolatilization vacuoles formed and finally served as accommodation for deposited minerals (Figure 10a,b). (ii) Dikes intruded vertically and may serve as relatively long-lived magma conduits [26], and heat expansion caused by the supplement of magmas followed by the shrinkage of host rocks eventually caused fractures in coal, for example, the development of prismatic fracturing perpendicular to the heating surface [24]. Joints and fractures in the coal acted as paths for magmas and hydrothermal fluids and were finally filled by carbonates (Figure 10c,d). (iii) Organic inclusions in carbonate minerals may reflect the alteration of magma on coal seams during emplacement (Figure 10e,f), that is, when a large amount of the magma invades quickly, organic debris may be carried and transported and eventually mixed in epigenetic minerals.

4.3. Geochemical Characteristics

4.3.1. Major Elements in the Tashan Coal

The contents of major element oxides in the Tashan coal are shown in

Table 6. Contents of major element oxides in the Tashan coal (wt.%, whole coal basis).

Sample No.	SiO2	Al2O3	Fe2O3	MgO	CaO	Na2O	K2O	MnO	TiO2	P2O5
T1	9.61	6.56	1.27	2.41	5.86	0.14	0.13	0.022	0.15	0.332
T2	6.73	6.40	0.48	0.63	1.63	0.07	0.16	0.009	0.11	0.297
T3	6.44	6.07	0.33	0.43	1.35	0.07	0.07	0.006	0.13	0.425
T4	6.86	7.17	0.03	0.05	0.18	0.05	0.02	bdl	0.07	0.484
T5	6.66	6.13	0.53	0.24	0.61	0.03	0.05	0.008	0.12	0.027
T6	4.59	4.86	0.04	0.05	0.21	0.03	0.02	bdl	0.03	0.335
T7	5.43	5.26	0.04	0.05	0.27	0.04	0.04	bdl	0.04	0.430
T8	3.47	3.22	0.13	0.10	0.58	0.02	0.03	0.004	0.10	0.030
T9	2.80	3.09	0.49	0.29	1.22	0.02	0.02	0.010	0.04	0.191
T10	2.85	2.74	1.84	0.27	1.74	0.01	0.02	0.020	0.06	0.009
T11	4.72	4.83	bdl	0.04	0.22	0.04	0.02	bdl	0.04	0.373
T12	5.11	4.53	0.04	0.07	1.15	0.01	0.03	bdl	0.03	0.016
T13	4.16	3.85	2.08	0.19	1.25	0.02	0.04	0.015	0.07	0.012
T14	6.97	6.20	0.06	0.06	1.00	0.02	0.03	0.004	0.06	0.013
T15	8.20	8.24	0.08	0.07	0.29	0.03	0.03	bdl	0.34	0.455
T16	3.44	4.31	0.59	0.13	2.48	0.02	0.03	0.007	0.04	0.080
Chinese coal	8.47	5.98	4.85	0.22	1.23	0.16	0.19	0.015	0.33	0.092
Av. thermally altered	7.26	6.47	0.53	0.75	1.93	0.07	0.09	0.011	0.12	0.313
Av. unaltered	4.70	4.65	0.54	0.12	0.95	0.02	0.03	0.010	0.08	0.18
CC. thermally altered	0.86	1.08	0.11	3.41	1.57	0.44	0.45	0.75	0.35	3.40
CC. unaltered	0.56	0.78	0.11	0.55	0.77	0.15	0.15	0.67	0.23	1.92

bdl, below detection limit; Av., average value; Chinese coal, data are from Ren et al. [85] and Dai et al. [45]; Av. thermally altered, arithmetic mean of T1–T5; Av. Unaltered, arithmetic mean of T6–T16; CC., concentration coefficient, calculated on the basis of the average element concentration of the thermally altered and unaltered coal samples versus the value established for Chinese coal.

. Compared with unaltered coals, the thermally altered coals had enhanced average contents of major elements because of the devolatilization and interaction with magma (Figure 11). SiO₂ and Al₂O₃ were major components of the Tashan coal and averaged 7.26% and 6.47% in the thermally altered coals, respectively, slightly higher than the averages of 4.7% and 4.65% in the unaltered coals (Table 6). The average contents of MgO and CaO in the thermally altered coals were 0.75% and 1.93%, respectively, 6.22 and 2.04 times the averages for the unaltered coals (Table 6). P₂O₅ accounted for 0.31% on average in the thermally altered coal, slightly higher than the average of 0.18% in the unaltered coals. Na₂O and K₂O were enriched in the samples immediately adjacent to the intrusion, thus inducing increased average values in thermally altered coal (Table 6). The contents of Fe₂O₃ and MnO exhibited similar trends at different distances from the dike (Figure 11); although they tended to be enriched in the coal/rock contact zone, their high values in samples T10 and T13 led to virtually equal average contents in the thermally altered and unaltered coal (Table 6). The average contents of TiO₂ did not exhibit obvious differences in the thermally altered and unaltered coals (Table 6).

The average contents of MgO, CaO, and P₂O₅ in the thermally altered coal were significantly higher than those in average Chinese coals [45,85]—3.41, 1.57, and 3.40 times higher, respectively—while the Al₂O₃ content was 1.08 times that of Chinese coals. However, the average contents of SiO₂, Fe₂O₃, Na₂O, K₂O, MnO, and TiO₂ were lower than those of Chinese coals. The average P₂O₅ content in the unaltered coal was 1.92 times higher than that in the Chinese coals, whereas the contents of other elements were lower.

4.3.2. Trace Elements in the Tashan Coal

Table 7. Trace elements in the Tashan coal (μg/g, whole coal basis).

Sample No.	Li	Be	V	Cr	Co	Ni	Cu	Zn	Ga	Rb	Sr	Mo	Cd	In	Sb
T1	33	1.1	22	38	1.7	6.2	37	25	14	3.4	1649	1.2	0.04	0.05	0.24
T2	35	1.8	24	14	2.4	4.5	37	17	24	5.1	1452	1.8	0.03	0.04	0.45
T3	46	1.8	16	13	1.7	3.2	18	14	13	1.6	2318	1.4	0.04	0.03	0.24
T4	59	1.8	14	5.7	1.0	2.0	33	11	9.2	0.41	3067	0.62	0.01	0.03	0.46
T5	121	1.3	21	5.6	1.6	2.4	35	18	21	0.97	85	2.2	0.04	0.05	0.92
T6	37	1.5	17	5.1	1.2	2.2	31	13	10	0.29	1839	1.1	0.02	0.02	0.33
T7	39	1.6	12	4.8	1.4	2.6	14	9.3	11	0.38	2618	0.72	0.01	0.02	0.27
T8	27	1.6	19	7.4	2.2	7.1	31	16	25	0.66	108	1.8	0.08	0.10	0.76
T9	23	1.5	11	5.4	2.5	3.8	30	18	21	0.41	57	2.7	0.03	0.02	0.42
T10	27	0.98	19	6.4	2.8	4.0	33	20	33	0.37	24	2.2	0.06	0.07	1.2
T11	37	1.8	8.7	4.8	1.6	3.6	35	13	12	0.60	2110	0.77	0.02	0.02	0.38
T12	34	1.7	11	4.1	3.7	7.4	31	13	27	0.62	36	3.1	0.03	0.03	0.88
T13	43	1.9	11	6.1	1.4	2.5	32	24	12	0.89	36	1.1	0.08	0.09	0.33
T14	62	1.6	13	4.5	6.9	13	31	15	21	0.49	30	1.6	0.03	0.05	0.56
T15	60	2.8	23	8.2	1.6	10	32	17	12	0.55	2729	1.0	0.08	0.14	0.37
T16	49	0.81	14	4.7	2.3	4.8	29	15	24	0.46	485	2.7	0.06	0.02	0.51
Av. thermally altered	59	1.5	19	15	1.7	3.7	32	17	16	2.3	1714	1.4	0.03	0.04	0.46
Av. unaltered	40	1.6	15	5.6	2.5	5.6	30	16	19	0.52	916	1.7	0.05	0.05	0.54
Ratio	1.5	0.94	1.3	2.7	0.67	0.66	1.1	1.1	0.8	4.4	1.9	0.83	0.6	0.8	0.85
Sample No.	Cs	Ba	W	Tl	Pb	Bi	Th	U	Nb	Ta	Zr	Hf	Se	As	Sc
T1	0.36	184	1.0	0.35	32	0.16	7.4	2.5	5.2	0.40	63	2.2	1.3	0.94	4.2
T2	0.99	138	0.59	0.28	28	0.19	6.3	3.8	3.6	0.23	68	2.4	2.1	0.66	4.6
T3	0.23	173	0.47	0.05	22	0.21	6.1	1.6	4.0	0.32	54	1.5	1.4	0.40	3.9
T4	0.04	171	0.21	0.16	20	0.16	6.2	1.5	1.6	0.24	49	1.9	2.4	bdl	3.1
T5	0.12	28	1.03	0.40	44	0.39	13	6.5	6.4	0.42	110	3.7	4.6	0.51	4.6
T6	0.01	93	0.16	0.10	19	0.08	2.0	1.1	1.4	0.08	26	1.0	2.5	bdl	1.6
T7	0.01	134	0.17	0.17	13	0.10	2.5	1.1	1.1	0.11	29	1.1	1.3	bdl	1.9
T8	0.04	42	0.71	0.22	49	0.39	9.2	6.1	3.9	0.40	148	4.2	2.8	bdl	5.1
T9	0.02	56	0.24	0.50	17	0.10	3.3	0.82	1.0	0.12	43	1.4	1.7	0.43	2.1
T10	0.02	13	0.17	0.24	20	0.16	6.4	6.3	2.7	0.15	147	4.1	2.4	0.32	6.5
T11	0.01	133	0.23	0.03	15	0.13	3.1	1.6	1.2	0.11	37	1.1	1.6	bdl	2.8
T12	0.01	13	0.41	0.12	34	0.12	2.3	1.8	2.2	0.10	49	1.6	2.5	bdl	2.1
T13	0.03	12	0.18	0.25	25	0.43	6.5	4.5	4.2	0.23	122	3.2	3.1	0.30	6.3
T14	0.03	17	0.27	0.11	34	0.13	2.2	2.2	3.1	0.17	57	1.9	3.9	bdl	3.7
T15	0.03	156	0.97	0.13	45	0.83	8.9	4.5	8.6	0.75	146	4.1	3.8	bdl	10
T16	0.03	34	0.95	0.28	21	0.12	2.6	0.86	1.2	0.12	44	1.5	2.5	0.25	2.2
Av. thermally altered	0.35	139	0.66	0.25	29	0.22	7.8	3.2	4.2	0.32	69	2.3	2.3	0.63	4.1
Av. unaltered	0.02	64	0.41	0.20	27	0.24	4.5	2.8	2.8	0.21	77	2.3	2.6	0.32	4.0
Ratio	17.5	2.17	1.6	1.25	1.1	0.92	1.73	1.1	1.5	1.52	0.89	1	0.88	2.0	1.0

bdl, below detection limit; Av. thermally altered, arithmetic mean of T1–T5; Av. unaltered, arithmetic mean of T6–T16; Ratio, average element concentration in thermally altered coal versus unaltered coal.

summarizes the concentrations of trace elements in the Tashan coal. Lithium, V, Cr, Rb, Sr, Cs, Ba, W, Tl, Th, Nb, Ta, and As were more enriched in the thermally altered than in the unaltered coal, whereas Co, Ni, Ga, Mo, Cd, In, and Sb exhibited depletions in the thermally altered coal. The average concentrations of Be, Cu, Zn, Pb, Bi, U, Zr, Hf, Se, and Sc were virtually equal in both coals.

In the current study, the average element concentrations in the North China Carboniferous–Permian coal [85,86], Chinese coal [45], and average world coal [87] were collected to calculate concentration coefficients (CC) of elements in the Tashan coal; for example, CC_i and CC_i’ were used to denote the corresponding concentration coefficients in the thermally altered and unaltered coal, respectively (

Table 8. Concentration coefficients of trace elements in the Tashan coal.

Element	C–P coal in North China a	Chinese Coal b	World Coal c	Thermally Altered Coal d			Unaltered Coal e
				CC1	CC2	CC3	CC1’	CC2’	CC3’
Li	18	31.8	12	3.3	1.9	4.9	2.2	1.3	3.3
Be	1.92	2.11	1.6	0.8	0.7	1.0	0.9	0.8	1.0
V	39.79	35.1	25	0.5	0.6	0.8	0.4	0.4	0.6
Cr	15.81	15.4	16	1.0	1.0	1.0	0.4	0.4	0.4
Co	4.24	7.08	5.1	0.4	0.2	0.3	0.6	0.4	0.5
Ni	11.75	13.7	13	0.3	0.3	0.3	0.5	0.4	0.4
Cu	21.97	17.5	16	1.5	1.8	2.0	1.4	1.7	1.9
Zn	48.99	41.4	23	0.3	0.4	0.7	0.3	0.4	0.7
Ga	9.88	6.55	5.8	1.7	2.5	2.8	1.9	2.9	3.3
Rb	6	9.25	14	0.4	0.3	0.2	0.1	0.06	0.04
Sr	154	140	110	11.1	12.2	15.6	6.0	6.5	8.3
Mo	3.47	3.08	2.2	0.4	0.5	0.7	0.5	0.6	0.8
Cd	0.3	0.25	0.22	0.1	0.1	0.1	0.2	0.2	0.2
In	Nd	0.047	0.031	Nd	0.8	1.3	Nd	1.1	1.7
Sb	0.68	0.84	0.92	0.7	0.6	0.5	0.8	0.6	0.6
Cs	1.2	1.13	1	0.3	0.3	0.4	0.02	0.02	0.02
Ba	110	159	150	1.3	0.9	0.9	0.6	0.4	0.4
W	1.5	1.08	1.1	0.4	0.6	0.6	0.3	0.4	0.4
Tl	0.35	0.47	0.63	0.7	0.5	0.4	0.6	0.4	0.3
Pb	20.28	15.1	7.8	1.4	1.9	3.7	1.3	1.8	3.4
Bi	1.6*	0.79	0.97	0.1	0.3	0.2	0.2	0.3	0.2
Th	8.7	5.84	3.3	0.9	1.3	2.4	0.5	0.8	1.4
U	2.6	2.43	2.4	1.2	1.3	1.3	1.1	1.2	1.2
Nb	28	9.44	3.7	0.2	0.4	1.1	0.1	0.3	0.8
Ta	0.5	0.62	0.28	0.6	0.5	1.2	0.4	0.3	0.8
Zr	188.28	89.5	36	0.4	0.8	1.9	0.4	0.9	2.1
Hf	3	3.71	1.2	0.8	0.6	1.9	0.8	0.6	1.9
Se	4.84*	2.47	1.3	0.5	1.0	1.8	0.5	1.0	2.0
As	2.59*	3.79	8.3	0.2	0.2	0.1	0.1	0.1	0.04
Sc	6*	4.38	3.9	0.7	0.9	1.1	0.7	0.9	1.0

Nd, no data; ^a data with * are from Tang and Huang [86], the others are from Ren et al. [85]; ^b data are from Dai et al. [45]; ^c data are from Ketris and Yudovich [87]; ^d average concentration of trace elements in thermally altered coal; ^e average concentration of trace elements in unaltered coal; CC_i and CC_i’ respectively denote the concentration coefficient of the corresponding element in thermally altered and unaltered coal calculated on the basis of the average element concentration of the thermally altered and unaltered coal samples versus the respective value established for the C–P coals in North China, Chinese coal, and world coal.

, Figure 12). The enrichment of elements in the Tashan coal was classified according to the method proposed by Dai et al. [88].

Compared with the average North China Carboniferous–Permian coals (Table 8, Figure 12), the Tashan thermally altered coal was significantly enriched in Sr (CC₁ = 11.1) and slightly enriched in Li (CC₁ = 3.3). In the unaltered coal, Sr (CC₁’ = 6.0) was enriched, and Li (CC₁’ = 2.2) was slightly enriched.

Compared with average Chinese coals, strontium was significantly enriched in both thermally altered (CC₂ = 12.2) and unaltered coals (CC₂’ = 6.5). Ga was slightly enriched in both coals, with concentration coefficients of 2.5 and 2.9 in the thermally altered and unaltered coals, respectively.

Compared with average world coals, the Tashan thermally altered coal was significantly enriched in Sr (CC₃ = 15.6) and slightly enriched in Li (CC₃ = 4.9), Cu (CC₃ = 2.0), Ga (CC₃ = 2.8), Pb (CC₃ = 3.7), and Th (CC₃ = 2.4). The unaltered coal was enriched in Sr (CC₃’ = 8.3) and slightly enriched in Li (CC₃’ = 3.3), Ga (CC₃’ = 3.3), Pb (CC₃’ = 3.4), and Zr (CC₃’ = 2.1).

4.3.3. Modes of Occurrence of Elements

The modes of occurrence of elements in coal are important, as this parameter dictates the behavior of the element during coal combustion, beneficiation, conversion, weathering, leaching, or any other chemical reaction that the coal undergoes [89]. Studies on the modes of occurrence of elements in thermally altered coal may provide information on the sources of elements and elemental migration during the intrusive event. In this study, correlation and hierarchical cluster analysis, as well as SEM-EDS, were used to study the affinities and geochemical associations of elements in the thermally altered coals.

1.

Correlation and hierarchical cluster analysis

Correlation between element concentrations and ash yields may provide preliminary insights into the elemental inorganic-organic affinities [77,90,91]. Hierarchical cluster analysis based on Pearson’s correlation, which categorized elements in the Tashan thermally altered coal into four groups, was useful to analyze the geochemical associations and potential sources of the elements in the coal (Figure 13). Pearson’s correlation coefficients between elements and ash yields, as well as inter-elements, are summarized in

Table 9. Pearson’s correlations between elements and ash yields and selected elements.

Correlation with Ash Yield
Group 1	rash=	MgO (0.594), CaO (0.623), Cr (0.697), Na2O (0.835), SiO2 (0.457), MnO (0.449), Fe2O3 (0.348), Zn (0.239), Al2O3 (0.592), As (0.481), Ni (0.790), TiO2 (0.284), V (0.002), Cu (-0.087), In (−0.103), W (−0.453), Tl (−0.423)
Group 2	rash=	K2O (0.751), Rb (0.676), Co (0.379), Cs (0.516)
Group 3	rash=	P2O5 (0.907), Sr (0.867), Ba (0.967), Be (0.200)
Group 4	rash=	Ga (−0.394), Sc (−0.479), Cd (−0.725), Nb (−0.763), Ta (−0.516), Mo (−0.876), Pb (−0.810), Hf (−0.840), Sb (−0.948), Se (−0.965), Zr (−0.944), U (−0.882), Li (−0.996), Bi (−1.00), Th (−0.945)
Correlation with Aluminosilicate
rSi-Al > 0.7		MnO, Fe2O3, MgO, CaO, Zn, Cr, As, TiO2, Na2O, Ni, In, K2O, Rb, Cu, V
rSi-Al 0.5–0.69		W, Cs
rSi-Al 0.35–0.49		Ba, Ta, Tl
Correlation with Carbonate
rCa > 0.7		MgO (0.999), Cr (0.995), SiO2 (0.980), MnO (0.971), Na2O (0.949), Fe2O3 (0.945), Ni (0.920), As (0.909), Zn (0.887), TiO2 (0.852), Al2O3 (0.835)
rCa 0.5–0.69		In (0.668), Ba (0.628), K2O (0.557)
Correlation with Phosphate
rp > 0.5		Sr (0.996), Ba (0.958), Na2O (0.646)
Correlation Coefficient between Selected Elements
SiO2-Al2O3 = 0.820, Na2O-K2O = 0.659, MnO-Fe2O3 = 0.994, TiO2-SiO2 = 0.881, Co-K2O = 0.709, Rb- K2O = 0.977, Sr-Ba = 0.936, Zr-Hf = 0.972, Li-Se = 0.947, V-Cu = 0.966, Co-Cs = 0.972, Cd-Nb = 0.969, Pb-Hf = 0.970, Li-Bi = 0.995

.

Group 1 included MgO, CaO, Cr, Na₂O, SiO₂, MnO, Fe₂O₃, Zn, Al₂O₃, As, Ni, TiO₂, V, Cu, In, W, and Tl (Figure 13). In this group, CaO, Na₂O, Cr, and Ni were strongly correlated with the ash yield (r_ash ≥ 0.6), suggesting their inorganic affinity (Table 9); MgO, SiO₂, Al₂O₃, MnO, Fe₂O₃, and As had lower but still positive correlation coefficients with the ash yield (0.6 > r_ash ≥ 0.3), suggesting that these elements are mainly associated with minerals in coal; TiO₂, V, and Zn were weakly correlated with the ash yield (0.3 > r_ash ≥ 0), suggesting that the modes of occurrence of these elements are complex. Cu, In, W, and Tl had negative correlation coefficients with the ash yield; however, these elements are not necessarily associated with organics. All of the group 1 elements had significant correlation coefficients with the aluminosilicates, and most of them were strongly correlated with carbonates, except V, Cu, W, and Tl. Therefore, it is deduced that the group 1 elements are mainly associated with aluminosilicates and carbonates in coal.

Group 2 included K₂O, Rb, Co, and Cs (Figure 13). K₂O and Rb were strongly correlated with the ash yield (rash ≥ 0.6, Table 9), while Cs and Co had lower but still positive correlation coefficients (0.6 > r_ash ≥ 0.3). The close correlation between K₂O and Rb (r = 0.977) suggests that they both from illite or muscovite in the coal, which is consistent with Finkelman [92], who stated that substantial amounts of rubidium may be present in illite and mixed-layer clays, as rubidium can readily substitute for potassium. Finkelman et al. [93] suggested that Cs is geochemically associated with K and Rb. In the present study, the association of Cs and Co (r = 0.972) suggests that they are both from the same potassium-bearing minerals, which may be muscovite and illite.

Group 3 included P₂O₅, Sr, Ba, and Be (Figure 13). Elements in this group were strongly correlated with the ash yield (r_ash = 0.867–0.967, Table 9), except Be (r_ash = 0.200). The high correlation coefficients between P₂O₅, Sr, and Ba suggest that these elements may be associated with phosphates in coal, such as goyazite, gorceixite, crandallite-group minerals, and apatite [39,93]. As for Be, it only exhibited significant correlations with Co (r = 0.616) and was weakly correlated with the ash yield. Previous studies have suggested organic and aluminosilicate associations of Be in coal [93,94]. It is thus inferred that beryllium in the Tashan coal may be associated with clays and organics; however, further studies are needed.

Group 4 included Ga, Sc, Cd, Nb, Ta, Mo, Pb, Hf, Sb, Se, Zr, U, Li, Bi, and Th (Figure 13), which were negatively correlated with the ash yield (ranging from −0.394 to −1, Table 9). However, these elements are not necessarily associated with organic matter. Lead, Bi, Se, and Cd may be associated with sulfides and selenides; for example, Pb and Se can exist together as clausthalite [95]. These minerals may occur as fine-grained particles within or shielded by organics [93] and only account for a small proportion, which is below the XRD detection limit. Zirconium, Ta, Th, Hf, and U may be associated with accessory zircon. Lithium, however, may be associated with clays and accumulated at the edge of the alteration halo; therefore, it exhibits a reverse trend with the ash yields.

2.

SEM-EDS Analysis

While statistical analysis provides useful information on the modes of occurrence in coal, it should be combined with other methods for reliable interpretations [39,45,91,96,97]. In the present study, the direct determination of some mineral compositions in the Tashan coals was accomplished by SEM-EDS.

The SEM-EDS analysis confirmed the modes of occurrence of some elements in the Tashan coals (

Table 10. SEM-EDS semi-quantitative analysis of some minerals in the Tashan coal.

Mineral		O	Si	Al	Fe	Ca	Mg	K	P	Sr	Sb	Zr	Nb	Se	W	Pb	Mo
Kaolinite (n = 19)	Min	15.36	4.04	2.60	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl
	Max	58.04	38.45	28.74	bdl	9.30	bdl	2.24	bdl	23.46	3.54	12.80	bdl	bdl	bdl	bdl	bdl
	Average	44.43	18.33	18.87	bdl	1.11	bdl	0.12	bdl	3.13	0.19	0.67	bdl	bdl	bdl	bdl	bdl
Illite (n = 3)	Min	44.52	19.47	20.79	bdl	bdl	bdl	2.01	bdl	6.34	bdl	bdl	bdl	bdl	bdl	bdl	bdl
	Max	50.24	24.44	21.94	bdl	bdl	bdl	2.56	bdl	6.87	bdl	bdl	bdl	bdl	bdl	bdl	bdl
	Average	47.02	22.60	21.45	bdl	bdl	bdl	2.33	bdl	6.60	bdl	bdl	bdl	bdl	bdl	bdl	bdl
Dolomite (n = 3)	Min	46.75	bdl	bdl	bdl	18.47	11.76	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl
	Max	57.62	bdl	bdl	bdl	27.20	15.38	bdl	bdl	bdl	10.38	bdl	bdl	bdl	bdl	bdl	bdl
	Average	51.33	bdl	bdl	bdl	21.50	13.23	bdl	bdl	bdl	5.85	bdl	bdl	bdl	bdl	bdl	bdl
Ankerite (n = 14)	Min	18.16	bdl	bdl	3.10	3.51	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl
	Max	59.37	bdl	9.69	29.54	28.00	15.86	bdl	bdl	bdl	14.88	1.94	2.60	5.32	4.33	0.30	42.02
	Average	45.82	bdl	0.94	8.85	19.02	9.83	bdl	bdl	bdl	2.39	0.14	0.37	0.38	0.31	0.02	6.47
Quartz (n = 5)	Min	53.13	43.95	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl
	Max	56.05	46.87	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl
	Average	54.65	45.35	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl	bdl
Goyazite (n = 2)	Min	27.80	3.57	3.82	bdl	bdl	bdl	bdl	3.34	2.73	bdl	bdl	bdl	bdl	bdl	bdl	bdl
	Max	30.25	4.10	5.87	bdl	bdl	bdl	bdl	4.60	3.86	bdl	bdl	bdl	bdl	bdl	bdl	bdl
	Average	29.03	3.84	4.85	bdl	bdl	bdl	bdl	3.97	3.30	bdl	bdl	bdl	bdl	bdl	bdl	bdl

N, number of detection spots; Min, minimum; Max, maximum; Av, average; bdl, below detection limit (0.01%).

, Figure 14). For example, Si and Al were mainly associated with clays, while quartz was another source of Si in coal. Iron, Ca, and Mg were mainly associated with ubiquitous carbonates in the thermally altered coals; however, a substantial amount of Fe was also detected in pyrite. In addition, Ca and Mg were also detected in anhydrite, and some Ca occurred in apatite; P was detected in goyazite and apatite; and Zr, Nb, Hf, and Ta were associated with zircon.

4.3.4. Strontium in the Tashan Coal

Concentrations of strontium in the Tashan coal were extremely high (peak at 3067 μg/g in sample T4) when compared with other coals. Despite the relatively low concentration in T5, the average concentration of Sr in the thermally altered coal was 1714 μg/g, which is more than 12 times that of Chinese coal (140 μg/g on average). Finkelman [92] suggested that the modes of occurrence of strontium can be both organic and inorganic; inorganic Sr may be associated with phosphates such as crandallite-group minerals and goyazite, barite, celestite, carbonates, clay minerals, and zeolites [17,92,96,98]. The quantification of the modes of occurrence of strontium by the sequential leaching procedure demonstrated that in bituminous coals, 50% of the Sr was associated with phosphates and carbonates, 25% was associated with silicates, and 25% was associated with organics [93]. In the present study, the high correlation coefficients between the elements in group 3 reflect that Sr is associated with phosphates, which was confirmed by the presence of goyazite by SEM (Figure 7g). Moreover, SEM-EDS analysis demonstrated that Sr is also associated with clays, such as kaolinite (Figure 15). A total of 22 clay minerals were analyzed by SEM-EDS, and Sr was detected in 13 of them, with a maximum concentration of 23.46 wt.% in kaolinite (Table 10). Although SEM-EDS analysis only provides semi-quantitative elemental compositions of the minerals in coal, it indicates that the abundant Sr-bearing clays resulted in high strontium values in the Tashan coal.

4.3.5. Element Sources

As discussed above, the interaction between rocks and coals has caused changes in the concentrations of elements, which may enrich, deplete, or have no impact in the contact zone. The enriched elements in the thermally altered coal have two potential sources: the elements either migrated out from the intrusion and enriched in coals and/or were enriched through residual accumulation during pyrolysis of the coal [99], and the contribution of each source differs depending on the mineral affinity. Elements that do not exhibit significant variations in the altered and unaltered zone may primarily have a coal origin; otherwise, they would be derived from the intrusion or driven off from the alteration halo [43]. Hierarchical cluster analysis categorizes elements that exhibit similar trends into the same group; these elements may have the same source (Figure 16).

The group 1 elements fluctuated in the altered and unaltered zone and increased sharply in the contact zone (Figure 16a), which indicates that intrusion is the source of these elements. It is generally suggested that minerals are the major hosts of the vast majority of elements present in coal [39], and thus the input of these elements from the intrusive body can be represented by epigenetic mineralization derived from hydrothermal fluids.

The intrusive rock itself may be altered by coal [37,44,53]. It is evident that the dike in the area adjacent to the coal became pale, whereas the color was dark green in the area far away from coal (Figure 3a). This phenomenon may provide more evidence for element migration. Song et al. [53] reported that mafic and ferrous minerals, such as pyroxene ((Ca,Na,Li) (Mg,Fe²⁺,Fe³⁺,Mn,CrAl) Si₂O₆), olivine ((Mg,Fe)₂SiO₄), and magnetite (FeFe₂O₄), interact with CO₂ during the intrusive event. It is assumed that alteration of these minerals and subsequent elemental migration may cause the pale margins of the dike and the enrichment of group 1 elements in the thermally altered coal.

Concentrations of the group 2 elements remained relatively stable in the unaltered coal but increased in the contact zone (Figure 16 b), which may be due to an input of these elements from the intrusion.

The group 3 elements fluctuated greatly in the altered and unaltered coal. However, these elements were enriched in the contact zone (Figure 16c). The alteration of minerals in the intrusion, such as apatite (Ca₅(PO₄)₃(F,Cl,OH)), and the migration of these elements may play a part in the elevated group 3 concentrations; however, other factors may have also affected the concentrations of these elements. For example, it is possible that some of the group 3 elements contained in the roof rock or the intrusion were leached out by groundwater and introduced into the adjacent coal [37].

The trend of group 4 element exhibits a different pattern from the above elements; since these elements were not enriched or depleted with distance (Figure 16d), the concentration pattern may indicate that the coal itself is the source of these elements and that they responded neutrally to the intrusive event.

5. Conclusions

Petrographic, mineralogical, and geochemical compositions of the Tashan coal were investigated in this paper, and the following conclusions were drawn:

(1) The 3.6 m diabase dike has caused an alteration halo of approximately 2 m. Within the area, coals exhibit an elevated vitrinite reflectance and ash yield; volatile matter is driven off; and thermally altered organic particles occur.

(2) More crystalline phases were identified in the thermally altered coal than in the unaltered coal. Epigenetic carbonates such as dolomite and ankerite may have been deposited by hydrothermal fluids. Muscovite and tobelite may have been transformed from a kaolinite precursor.

(3) Significantly enriched strontium in the Tashan coal is associated with phosphate minerals such as goyazite as well as the Sr-bearing clays.

(4) Elements associated with aluminosilicates, clays, and carbonates are enriched in the coal/rock contact zone; the enrichment was caused by epigenetic mineralization and element migration. On the other hand, P₂O₅, Sr, Ba, and Be may be controlled by multiple factors, such as hydrothermal fluids and groundwater, while elements exhibiting no variations in thermally altered and unaltered coals indicate that the coal itself is the source.