Mechanisms and Genesis of Acidic Goaf Water in Abandoned Coal Mines: Insights from Mine Water–Surrounding Rock Interaction

Abstract

The formation of acidic goaf water in abandoned coal mines poses significant environmental threats, especially in karst regions where the risk of groundwater contamination is heightened. This study investigates the geochemical processes responsible for the generation of acidic water through batch and column leaching experiments using coal mine surrounding rocks (CMSR) from Yangquan, China. The coal-bearing strata, primarily composed of sandstone, mudstone, shale, and limestone, contain high concentrations of pyrite (up to 12.26 wt%), which oxidizes to produce sulfuric acid, leading to a drastic reduction in pH (approximately 2.5) and the mobilization of toxic elements. The CMSR samples exhibit elevated levels of arsenic (11.0 mg/kg to 18.1 mg/kg), lead (69.5 mg/kg to 113.5 mg/kg), and cadmium (0.6 mg/kg to 2.6 mg/kg), all of which exceed natural crustal averages and present significant contamination risks. The fluorine content varies widely (106.1 mg/kg to 1885 mg/kg), with the highest concentrations found in sandstone. Sequential extraction analyses indicate that over 80% of fluorine is bound in residual phases, which limits its immediate release but poses long-term leaching hazards. The leaching experiments reveal a three-stage release mechanism: first, the initial oxidation of sulfides rapidly lowers the pH (to between 2.35 and 2.80), dissolving heavy metals and fluorides; second, slower weathering of aluminosilicates and adsorption by iron and aluminum hydroxides reduce the concentrations of dissolved elements; and third, concentrations stabilize as adsorption and slow silicate weathering regulate the long-term release of contaminants. The resulting acidic goaf water contains extremely high levels of metals (with aluminum at 191.4 mg/L and iron at 412.0 mg/L), which severely threaten groundwater, particularly in karst areas where rapid cross-layer contamination can occur. These findings provide crucial insights into the processes that drive the acidity of goaf water and the release of contaminants, which can aid in the development of effective mitigation strategies for abandoned mines. Targeted management is essential to safeguard water resources and ecological health in regions affected by mining activities.

1. Introduction

Coal continues to serve as a fundamental component of global energy systems, playing a vital role in ensuring energy security, meeting industrial needs, and fostering economic development, particularly in emerging economies. Global proven coal reserves exceed 1.07 trillion metric tons, geopolitically concentrated in five nations controlling over 75% of total deposits: the United States (23%), Russia (15%), China (13%), Australia (14%), and India (10%) [1,2]. In China, coal resources are a leading source of national energy supply [3,4] and have significantly contributed to the country’s economic growth. However, the improper exploitation of coal resources has led to serious environmental issues [5,6,7]. Mining and processing of coal cause environmental harm by releasing toxic substances, such as heavy metals (e.g., arsenic and mercury), fluoride, hydrocarbons, and acidic byproducts [6,8,9,10,11]. These pollutants contaminate aquatic systems through multiple pathways, including drainage into rivers, lakes, reservoirs, and groundwater [12,13]. The persistent nature of these contaminants raises concerns about long-term ecological impacts and sustainable resource management.

The exploitation of coal resources significantly disrupts groundwater systems through physical, chemical, and hydrological alterations, as evidenced by studies across global mining regions [14,15,16,17,18]. The fracturing and subsidence induced by longwall mining create hydraulic pathways that connect the aquifer with shallower, contaminated water layers or surface pollutants, exacerbating water quality degradation. The karst aquifer, a critical water resource in China, is characterized by high permeability and karst development, making it particularly vulnerable to anthropogenic disturbances [19,20,21]. In China’s coal-rich regions, coal mining activities intersecting with the Ordovician (Majiagou Formation, Fengfeng Formation) or Permian (Maokou Formation, Qixia Formation) limestone aquifers have raised significant concerns about groundwater contamination and hydrological disruption [22,23]. A primary concern is the infiltration of coal mining-derived contaminants into the karst aquifer. Studies in the Shanxi and Henan provinces have documented elevated concentrations of sulfates, total dissolved solids (TDS), and trace metals (such as Fe, Mn, Pb and Cd) in groundwater samples from mining-affected areas [18]. These pollutants mainly originate from the oxidation of pyrite (FeS₂) in coal seams and overlying strata, which generates acidic mine drainage (AMD) or goaf water. The acidic conditions enhance the solubility of heavy metals, allowing them to migrate into the aquifer. Huang et al. [24] demonstrated that AMD from coal mines in the Ordos Basin has led to the altering of the hydrochemical composition of the karst aquifer. This geochemical interaction degrades the water quality of the aquifers, posing long-term risks to water security. Furthermore, organic pollutants, including polycyclic aromatic hydrocarbons (PAHs) and volatile organic compounds (VOCs), have been detected in karst aquifers adjacent to active mines, likely originating from coal-derived organic matter [25,26,27].

Coal mining activities significantly alter hydrogeological conditions in mining areas, often disrupting groundwater systems and causing ecological degradation [28,29,30]. In China, where coal is predominantly extracted via underground mining [31], the closure of abandoned mines without proper procedures has led to the formation of acidic goaf water in multi-mined-out areas. This occurs as groundwater accumulates and reacts with sulfide minerals (e.g., pyrite [FeS₂] and sphalerite [ZnS]) in sulfur-rich coal or CMSR [32,33], producing sulfuric acid through oxidation [34,35,36]. Goaf water is characterized by elevated sulfate concentrations, high TDS, heavy metal contamination, and low pH [37,38,39,40,41,42], posing severe environmental risks, particularly in karst regions [43].

The scale of the problem is alarming: China’s coal mines decreased from 100,000 in the 1990s to fewer than 4300 by 2024 (The China Coal Industry Association, 2025), with improper mine closures exacerbating goaf water accumulation. This contaminated water can discharge into rivers, threatening aquatic ecosystems and public health [44,45,46,47], or seep into karst aquifers, causing cross-layer pollution [48]. Regions such as Shanxi [49,50,51,52], Zibo in Shandong [53], Kaili in Guizhou [54], the Huainan region [55], and the Yunnan-Guizhou Plateau and Xinjiang mining areas [56] are already heavily impacted. Furthermore, many coal reserves overlap with ecologically fragile karst zones [57], which are highly sensitive to pollution. As mining depths and production volumes increase, risks of karst water resource deterioration, water inrushes, and other ecological disasters linked to goaf water [27] underscore the urgent need for comprehensive research and mitigation strategies.

Despite growing recognition of these risks, gaps persist in understanding the geochemical mechanisms governing contaminant release from abandoned coal mines under dynamic leaching conditions. The main purpose of this paper is to investigate the occurrence of heavy metals (Al, Fe, Mn, Zn), and fluorine (F) in CMSR, elucidate the genesis mechanisms of goaf water, and quantify contaminant mobilization under dynamic leaching. By integrating geochemical analysis and kinetic leaching experiments, we aim to provide a theoretical foundation for mitigating goaf water contamination. Our research would provide theoretical and practical foundations for addressing the environmental challenges posed by acidic goaf water, contributing to the sustainable development of coal mining regions, especially coal mining in karst areas. The study highlights the need for targeted management strategies to mitigate contamination from abandoned mines. The study is also relevant for the protection of soil and water systems from pollution in coal mining areas.

2. Geologic Setting

Yangquan is a major coal-producing city in Shanxi province (China), historically known for its high-quality hard coal production. The Yangquan coal mining district is located to the west of the city of Yangquan (Figure 1a,b). The total known reserve of coal at Yangquan is estimated to be 118.2 Gt, covering an area of 1218 km². The major periods of coal formation are Carboniferous (Taiyuan Formation) and Early Permian (Shanxi Formation) (Figure 1c). The formation mainly consists of sandstone, mudstone, limestone and coal seams. The coal-bearing strata are made up of 19 coal seams, with a total average depth of 175 m. T15, T12 and S3 are the major coal seams in the area (Figure 1c). The major coal types include anthracite, meager coal, and little charred coal.

3. Materials and Methods

3.1. Sample Collection and Characterization

The overflowed goaf water samples, named MG, FMT, LG, were collected from the abandoned coal mines: Shandi, Miaogou and Hongtuyan coal mines (Figure 1b), where the goaf water overflowed from the entrance of the abandoned coal mines. The water samples were collected in three HDPE bottles: one for anion analysis and one with extra acid for cation and trace element analysis. The in situ physicochemical parameters (including temperature, pH, and EC) were measured at the sampling sites using a previously calibrated HANNA portable pH meter (HI8424, pH: ±0.01, T: ±0.4 °C). Major cations (K⁺, Na⁺, Ca²⁺ and Mg²⁺) and trace elements were determined by inductively coupled plasma-atomic emission spectrometry (ICP-AES) (IRIS Intrepid XSP, Thermo Elemental, Madison, WI, USA).

The CMSR samples were collected from the main coal mines (numbers I to V of Yangquan coal mines) and abandoned coal mines (Shandi, Hongtuyan and Miaogou) (Figure 1b). All collected samples were stored in plastic bags after being packaged by silver paper to prevent contamination and weathering.

The mineral composition of the surrounding rock samples was performed by means of X-ray diffraction (XRD). Semi-quantitative XRD analysis was undertaken by using the Reference Internal Standard Method [59].

The samples were digested by Aqua Regia (a mixture of concentrated nitric acid (HNO₃) and hydrochloric acid (HCl) in a 1:3 volume ratio). The content of heavy metals, such as As, Cr, Co, Ni, Cd, Hg and Pb, were analyzed using an Inductively Coupled Plasma-Mass Spectrometer (ICP-MS, Elan (R) DRC II, PerkinElmer SCIEX, Shelton, CT, USA). The accuracy of the analyses was confirmed using the certified reference material TMDA-54.4. The analysis limitation is 0.02 μg/L with the standard solution. The total fluorine content in the samples was analyzed by the alkali fusion/fluorine ion-selective electrode method (ISE, Leici, PF-1/PF-1-01, Shanghai INESA Scientific Instrument Co., Ltd., Shanghai, China), as described by Gao et al. (2016) [60]. The detection limit of the ISE method for F⁻ is 0.02 mg/L in neutral solutions.

3.2. Sequential Extraction

A sequential extraction was performed to study the concurrence of fluorine (F) and heavy metal in CMSR: (1) water soluble fraction (F1), 50 mL of H₂O (20 °C, 1:25 coal gangue to deionised water ratio, 30 min); (2) exchangeable fraction (F2), 1 M MgCl₂ (pH 7.0, 20 °C, 1:25 coal gangue to solution ratio, 30 min); (3) fraction bound to amorphous Fe-Al oxides (F3) and crystalline Fe/Al oxides (F4), 0.04 M NH₂OH·HCl + 25% HOAc (pH 2.0, 90 °C in a water bath, 1:25 coal gangue to solution ratio, 3 h); (4) fraction bound to organic matter (F5), 50 mL of 3.2 M NH₄Ac (1:25 coal gangue to solution ratio) after the mixture of 20 mL of 0.02 M HNO₃ (pH 2.0) and 30 mL of 30% H₂O₂ was added and heated to 85 °C (in a water bath) for 2 h; and (5) residual fraction (F6), determined by subtracting the other four fractions from the total content.

3.3. Batch Test

A batch test was applied to the surrounding rock samples in order to simulate the water–rock interaction under batch leaching conditions in the abandoned underground mines. It is a single batch water–rock interaction test performed at a liquid to solid ratio of 10 mL/g, with an agitation time of 24 h and deionised water as leachant. Additionally, another batch test of surrounding rock samples in solution with different pH was also performed following the method mentioned above (the pH of the solution was adjusted by HCl or NaOH).

3.4. Column Tests

To simulate the water–rock interaction under dynamic leaching conditions in the abandoned underground mines, a set of continuous flow through leaching experiments was conducted in column reactors. Most of the surrounding rock samples belong to sandstone, according to the field investigation. The predominant sandstone sample was therefore chosen for the column test. The surrounding rock sample with a particle size of 2 mm (previously smashed with a hammer) was packed uniformly in the side glass columns with an inner diameter of 2.0 cm and a length of 10.0 cm. Micropore filters (0.45 μm pore diameter) were used at the top and bottom of each column. A constant upward flow rate of ca. 0.2 mL/min was maintained using a peristaltic pump to minimize depositional effects due to gravity and potential preferential flow along the column walls.

The column test was performed with a total coal spoil weight of 12.5 g packed into a 10 cm height Omega column, using 1 mM Ca(NO₃)₂ as the electrolyte.

Most of the coal spoils belong to sandstone, according to the field investigation. The predominant sandstone coal spoils sample (CS4) was therefore chosen for the column leaching test. The laboratory column test was employed with an up-flow of 1 mM Ca(NO₃)₂ (electrolyte, pH 6.7), operated at a flow rate of about four column pore volumes per hour for a total volume of 500 mL. The column was connected to a peristaltic pump, and the effluent solution was collected with an automatic fraction collector. An increasing pumping speed method (from four column pore volumes per hour to five column pore volumes per hour) was applied to simulate the possible velocity change of the leaching solution. To represent the pulsed input of precipitation, a stop-flow method (stopped for two hours) was applied, followed by another consecutive leaching. The effluent solution was filtrated through a 0.2 μm filter and analyzed with an ICP-MS for major cations and heavy metals.

The initial pH value of the leaching solution was adjusted with diluted HCl or NaOH to 7.5 ± 0.1 to simulate the neutral to weakly alkaline natural water. The effluent was collected every 5 mL. Frequent pH measurements were taken in all column experiments. The pH was determined by immersing a combined pH microelectrode (Hach H160g model, Hach Company, Loveland, CO, USA) in column effluent. Part of the effluent was used for the determination of aqueous fluorine and heavy metal concentrations.

4. Results

4.1. Hydrochemistry of Goaf Water

The hydrochemical analysis of goaf water samples MG, FMT, and LG reveals distinct geochemical characteristics shaped by varying environmental conditions. The sample MG is highly acidic (pH 3.26) with elevated concentrations of heavy metals, such as Al (191.4 mg/L), Fe (272.0 mg/L) and Cu (152.0 μg/L), indicative of acid mine drainage processes likely driven by sulfide mineral oxidation (

Table 1. Concentration of heavy metals in the batch leaching solution of CMSR and overflow goaf water. The concentrations that do not have between parentheses are in μg/L. MG, FMT and LG are overflowed goaf water, collected from the Shandi, Miaogou and Hongtuyan abandoned coal mines.

Heavy Metals	Shale	Sandstone	Limestone	Mudstone	MG	FMT	LG
pH	2.80–6.04	2.95–5.37	4.98–6.42	5.50–6.13	3.26	6.82	4.71
TDS (mg/L)	1115	2309	571	515	2528	3026	3672
Al (mg/L)	8.6	2.8	2.0	3.3	191.4	0.2	141.0
Fe (mg/L)	3.5	4.4	3.2	69.3	272.0	0.4	412.0
Mn (mg/L)	2.5	4.6	1.2	3.2	60.2	10.7	30.7
Zn (mg/L)	0.8	0.2	<0.1	1.9	10.3	0.9	2.5
F (mg/L)	2.58	8.76	0.75	2.51	0.14	0.86	0.89
Cu	3.30	1.11	0.43	2.30	152.0	2.78	10.60
Sr	31.6	9.6	123.0	34.8	4450	7160	3950
As	0.02	1.25	0.07	10.65	2.87	3.12	3.54
Pb	1.07	0.52	0.12	0.03	2.70	2.13	1.45
Cr	0.29	0.42	0.01	0.29	3.21	2.58	0.98
Cd	1.36	2.60	0.30	3.03	13.70	1.80	1.97
Co	6.70	0.47	0.33	7.0	1.95	0.21	0.38
Ni	5.24	2.17	1.07	7.85	29.90	9.44	12.0
Mo	0.05	1.33	0.56	0.66	4.52	3.28	4.29
V	0.57	1.79	0.10	0.56	3.20	1.82	3.10

). In contrast, FMT exhibits a near-neutral pH (6.82) and markedly lower metal levels, suggesting a weak oxidation of sulfide mineral or carbonate buffering. The higher Sr (7160 μg/L) and moderate fluorine (0.86 mg/L) in FMT potentially reflect the interactions with Sr-rich lithologies (like calcite) or mixing with groundwater (Table 1). LG, moderately acidic (pH 4.71), combines high TDS (3672 mg/L) with elevated Fe (412.0 mg/L) and Al (141.0 mg/L) (Table 1), implying a moderately intense acidic leaching. The progressive increase in toxic element concentrations in goaf water was likely influenced by pH-dependent solubility (e.g., enhanced metal mobility under acidic conditions) or distinct mineralogical sources. The variations of elements in goaf water samples underscore heterogeneous contamination mechanisms, highlighting diverse water–rock interaction pathways.

4.2. Mineralogy of CMSR

Field and laboratory investigations reveal that the coal-bearing strata are dominated by medium- to coarse-grained quartz sandstone, fine-grained mudstone/shale, and bioclastic/argillaceous limestone, interbedded with high-quality anthracite coal. These lithologies exhibit distinct mineral assemblages, as detailed below. The mineral assemblage of the sandstone samples is mainly quartz, kaolinite, and illite with minor proportions of calcite and pyrite (

Table 2. The mineralogy of CMSR samples (unit: %, BDL—below detection limit).

Sample ID	Lithology	Quartz	Kaolinite	Illite	Calcite	Pyrite	Gypsum	Others
YKA	Mudstone	22.63	59.39	14.80	BDL	0.20	0	3.00
YKC	Shale	0.31	1.67	95.63	0.98	0.35	0	1.05
YKD	Sandstone	48.48	21.80	17.82	0	11.90	0	0
YKE	Limestone	20.15	5.47	BDL	35.26	0.24	39.12	0
YKF	Shale	42.32	21.77	31.92	0.49	3.51	0	0
YKG	Shale	26.21	31.06	34.43	0	4.36	0	3.93
YKH	Sandstone	59.87	20.78	11.85	0	7.12	0	0.38
YKI	Sandstone	46.27	24.15	16.84	0.48	12.26	0	0
YKJ	Mudstone	7.30	89.17	BDL	0	3.54	0	0
EKA	Sandstone	50.00	6.70	37.00	4.00	0.10	0	2.20
EKB	Mudstone	29.22	61.42	8.82	0.53	0.14	0	0
EKC	Shale	1.17	19.13	77.78	0	0.16	0	1.92
EKE	Sandstone	86.96	BDL	3.30	1.04	8.70	0	0
EKF	Sandstone	96.07	1.79	1.09	0	0.08	1.04	0
EKG	Mudstone	23.73	70.91	2.70	0	2.56	0	0
EKH	Mudstone	14.08	69.73	11.81	0.49	2.07	0	1.82
SDA	Limestone	27.49	3.70	1.49	44.38	0.17	12.10	9.17
XGA	Limestone	1.35	0.65	4.94	48.67	0.22	43.24	1.58

, Figure 2). The minerals of the mudstone/shale samples are mainly kaolinite, or illite, with medium content of quartz and minor proportions of calcite and pyrite. The limestone sample is mainly composed of minerals and the mudstone/shale samples are mainly kaolinite, or illite, with calcite and gypsum, with medium content of quartz and minor proportions of clay minerals and pyrite. Pyrite is notably enriched i sandstone surrounding rocks with the highest value of 12.26% and a mean value of 6.69% by weight. The pyrite abundance in other CMSR decreases systematically across lithologies in the order of shale (2.1%), mudstone (1.7%) and limestone (0.21%). This trend may reflect the redox conditions during deposition, with sandstone facies favoring sulfate reduction in organic-rich, permeable environments. A high content of gypsum was only found in limestone samples, reaching up to 39.12%.

4.3. Heavy Metals and Trace Elements in CMSR

The heavy metal contents in CMSR samples from Yangquan coal mines are summarized in

Table 3. Summary of heavy metal contents in CMSR samples from Yangquan coal mines (mg/kg).

Heavy Metals	Sandstone	Limestone	Mudstone	Shale	Abundance in Continental Crust
Al (g/kg)	116.1	20.2	108.9	137.4
Fe (g/kg)	63.9	30.4	76.9	101.7	51
Mn	149.9	40.8	204	214.1	100
Cu	37.9	19.8	45.3	46.7	54
Zn	198.6	73.7	159.8	120.9	85
As	18.1	11.0	14.5	11.6	2.2
Pb	78.3	113.5	69.5	89.4	13
Cr	39.2	12.3	33.4	25.4	90
Cd	1.7	0.6	1.9	2.6	0.14
Sr	104.9	85.0	121.3	147.1	480
Ba	26.3	31.4	48.9	24.0
Co	7.1	4.2	8.0	9.1	20
Ni	36.0	6.8	44.5	11.8	71
Mo	11.7	1.9	1.5	1.5	1.2
V	149.8	151.9	134.6	138.3

. Aluminum (Al) and iron (Fe) exhibit significant enrichment across all lithologies, with Al concentrations ranging from 108.9 to 137.4 g/kg and Fe from 30.4 to 101.7 g/kg, both far exceeding their average continental crustal abundances (Al: ~8%, Fe: ~5.1%). Manganese (Mn), zinc (Zn), strontium (Sr), and vanadium (V) are moderately enriched, while cobalt (Co), and nickel (Ni) align closely with typical crustal values (Table 3). Notably, arsenic (As), lead (Pb), and cadmium (Cd) show pronounced enrichment, with As (11.0–18.1 mg/kg) and Pb (69.5–113.5 mg/kg) exceeding crustal averages over five times, respectively.

Elements such as V, Cr and As display wide confidence intervals in CMSR samples, reflecting their heterogeneous distribution. Most elements (e.g., Mn, Zn, Sr) increase progressively from limestone → sandstone → mudstone → shale, suggesting lithology-driven geochemical controls. For example, shale hosts the highest Mn (214.1 mg/kg) and Cd (2.6 mg/kg), while limestone shows the lowest Cr (12.3 mg/kg) and Ni (6.8 mg/kg). Chromium (Cr: 12.3–39.2 mg/kg) and nickel (Ni: 6.8–44.5 mg/kg) are consistently below crustal averages, likely due to leaching or dilution by silicate minerals. The extremely high concentrations of As, Pb and Cd pose high risks of groundwater contamination in the study area.

4.4. Occurrence of Fluorine in CMSR

4.4.1. Total Fluorine

The total fluorine content in the surrounding rock samples varies significantly, ranging from 106.1 mg/kg to 1885 mg/kg, with a mean value of 423.4 mg/kg (

Table 4. Statistical summary of total fluorine and fluorine speciation in CMSR samples.

Sample ID	Total Fluorine (mg/kg)	Water Soluble Fluorine (F1)		Exchangeable Fluorine Fraction (F2)		Amorphous Fe/Al Oxide Bound Fluorine (F3)		Crystalline Fe/Al Oxide Bound Fluorine (F4)		Fluorine Fraction Bound to Organics (F5)		Fluorine Fraction in Recalcitrant Phases (F6)
		Content (mg/kg)	%	Content (mg/kg)	%	Content (mg/kg)	%	Content (mg/kg)	%	Content (mg/kg)	%	Content (mg/kg)	%
YKA	416.7	16.5	3.96	7.9	1.90	14.3	3.42	6.0	1.43	6.4	1.52	365.7	87.77
YKC	273.5	7.5	2.74	9.1	3.31	16.1	5.87	5.0	1.81	6.8	2.51	229.1	83.76
SDA	106.1	2.3	2.12	8.1	7.59	10.2	9.60	3.8	3.61	3.8	3.61	77.9	73.47
YKD	539.2	28.8	5.34	12.2	2.27	22.4	4.15	11.2	2.08	11.6	2.15	453.1	84.01
YKE	190.0	3.5	1.82	6.7	3.51	20.0	10.53	9.0	4.74	3.8	1.98	147.2	77.42
YKF	453.8	16.7	3.68	7.5	1.66	5.0	1.09	9.2	2.03	8.7	1.92	406.7	89.62
YKG	616.0	22.4	3.64	17.7	2.87	5.7	0.92	4.6	0.75	8.6	1.40	557.1	90.42
YKH	494.9	18.1	3.67	7.5	1.52	6.1	1.22	11.2	2.27	7.3	1.46	444.7	89.86
YKI	539.7	6.2	1.14	9.8	1.81	18.5	3.43	4.2	0.78	10.3	1.90	490.8	90.94
YKJ	155.9	3.7	2.38	23.2	14.89	16.8	10.74	7.4	4.76	15.0	9.60	89.8	57.63
XGA	208.1	5.5	2.65	6.1	2.95	10.7	5.16	4.3	2.04	6.6	3.15	174.9	84.05
EKA	175.3	7.1	4.04	9.9	5.66	6.3	3.56	13.0	7.39	11.7	6.65	127.5	72.70
EKB	296.9	16.7	5.61	8.4	2.83	14.2	4.78	3.2	1.09	12.8	4.31	241.6	81.38
EKC	309.8	5.0	1.61	14.1	4.55	7.3	2.37	6.8	2.20	14.5	4.69	262.0	84.58
EKE	1885	200.0	10.60	9.8	0.52	18.8	1.00	86.6	4.60	39.5	2.10	1530	81.18
EKF	232.4	2.8	1.21	6.5	2.78	29.6	12.71	33.9	14.60	23.1	9.96	136.5	58.74
EKG	273.5	11.6	4.23	9.3	3.39	6.7	2.46	7.1	2.61	27.8	10.16	211.0	77.15
EKH	454.0	14.4	3.18	8.5	1.87	16.2	3.56	5.9	1.31	13.3	2.93	395.7	87.15
Mean value	423.4	21.6	3.53	10.1	3.66	13.6	4.81	12.9	3.34	12.9	4.00	352.3	80.66

). This concentration exceeds typical values reported for Shanxi coal (Datong coalfield, 37 mg/kg to 164 mg/kg; Gujiao coalfield, 42 mg/kg to 163 mg/kg; Ningwu coalfield, 255 mg/kg to 380 mg/kg; Shuozhou coalfield, 74 mg/kg to 1050 mg/kg; Pingshuo coalfield, 53 mg/kg to 226 mg/kg) [61,62,63], suggesting additional geological or environmental influences on fluorine enrichment.

Lithology and mineralogy may control the fluorine content in the surrounding rock. The highest fluorine concentration (1885 mg/kg) occurs in sandstone, which is dominated by quartz and contains minor clay minerals and pyrite. Fluorine enrichment in sandstone may be associated with clay mineral impurities (e.g., illite) or fluoride adsorption onto secondary mineral coating quartz surfaces. However, the sample EKF (96.07% quartz) shows lower fluorine (232.4 mg/kg), suggesting that pure quartz-rich sandstone does not necessarily retain high fluorine. Mudstone and shale showed a moderate fluorine content, with mean values of 319.4 mg/kg and 413.3 mg/kg, respectively. These lithologies contain abundant clay minerals (kaolinite, illite), which are known to strongly adsorb fluorine due to their high surface area and ion-exchange capacity. For example, YKA (mudstone) and YKG (shale) contain over 60% of clay minerals (kaolinite and illite) and correlate with elevated fluorine (416.7 mg/kg and 616.0 mg/kg). However, YKC (shale) has 95.63% illite but a relatively low fluorine content (273.5 mg/kg), indicating that fluorine retention is not solely dependent on clay mineral abundance but also on geochemical conditions (e.g., pH, competing ions). Limestone samples, dominated by calcite (35%–49%) and gypsum (12%–43%), with minimal clay minerals, exhibit the lowest fluorine content (mean value of 166.4 mg/kg). The low fluorine levels suggest that carbonate minerals (calcite, gypsum) do not efficiently retain fluoride, possibly due to low adsorption capacity or leaching under alkaline conditions.

4.4.2. Speciation of Fluorine

The distribution of fluorine fractions shows distinct patterns of the samples (Table 4).

The F1 (1.14%–10.6%) is a bioavailable fraction that shows the strongest positive correlation with quartz content (r = 0.72, p < 0.01), particularly in sandstone samples. For instance, the quartz-rich sandstone EKE (86.96% quartz) contains the highest F1 percentage (10.6%). In contrast, clay-dominated samples, such as YKC shale (95.63% illite), show lower F1 (2.74%). A negative correlation between F1 and clay mineral content (r = −0.65, p < 0.05) was observed. This suggests that clay minerals may inhibit fluorine mobility through adsorption mechanisms.

The F2 (0.52%–14.89%) demonstrates a strong positive correlation with kaolinite content (r = 0.81, p < 0.01). Kaolinite-rich mudstone YKJ (89.17% kaolinite) shows the highest F2 percentage (14.89%). The data suggest that the cation exchange capacity of kaolinite plays a crucial role in retaining exchangeable fluorine. Quartz-dominated samples, such as EKF (96.07% quartz), have a minimal F2 value (2.78%).

The sequential extraction steps 3 and 4 employ strong reducing agents (NH₂OH·HCl and HOAc) designed to selectively dissolve iron and aluminum oxides. These reagents effectively break down both amorphous (step 3) and crystalline (step 4) Fe/Al oxide phases, releasing the associated fluorine into solution. The fluorine content in F3/F4 is 0.75%–14.60% total fluorine. This value is higher in mudstone and sandstone, according to the extraction experiment results (Table 4). A positive correlation of F3 + F4 with pyrite content (r = 0.68, p < 0.05) was found. Samples with higher pyrite showed elevated F3 + F4 values as seen in sandstone YKD (11.9% pyrite, 6.23% F3 + F4) and YKI (12.3% pyrite, 4.21% F3 + F4). It is deduced that oxidation of pyrite may produce secondary Fe oxides, such as ferrihydrite and goethite, providing abundant sorption sites for fluoride. A high value of F3 + F4 was found in mudstone samples containing 50%–70% clay minerals (kaolinite, illite). This may be due to the abundant edge sites for Fe oxides, structural Fe generated crystalline oxides and pyrite.

The extraction agent for Step 5 is an oxidant, with a mixture of hydrogen peroxide and nitric acid used as an extractant for the fluorine fraction bound to organics. Mudstones consistently exhibit the highest proportion of F5 (1.52%–10.16%, mean value 5.7%) compared to other lithologies, sandstones (1.46%–9.96%, mean 4.0%), shales (1.40%–4.69%, mean 2.6%) and limestone (1.98%–3.61%, mean 2.9%). This trend can be explained by organic matter content and geochemical conditions. Mudstones typically contain more organic matter (OM) than sandstones or limestones. The mechanism of F retention includes fluorine (F⁻) binding to functional groups in organic matter (e.g., carboxyl [-COOH], phenolic [-OH] and amine [-NH₂] groups) and organo-fluorine complexes forming under mildly reducing conditions, where F⁻ replaces -OH or -Cl in organic structures. The low-medium value of fluorine percentages in fraction 5 indicates that the organic-bound fluorine is not one of the major occurrences in rocks (CMSR).

The F6 represents fluorine incorporated in recalcitrant mineral structures. This fraction was calculated by subtracting the other five fractions from the total fluorine and shows systematic variation according to lithology and mineral composition. F6 dominates all lithologies with 57.63%–90.94% of total fluorine (Table 4) and the widest range reflecting mineral heterogeneity. The shale samples have the highest residual proportions (F6, 83.76%–90.42%, mean 87.1%), which strongly correlated with illite content (r = 0.89, p < 0.05). The exceptionally high F6 in shales reflects the tight structural binding in illite (structural incorporation in octahedral sheets), low accessibility to extraction reagents and limited exchangeable sites. The sandstones (58.74%–90.94%) and mudstones (57.63%–87.77%) showed a moderately higher F6 percentage value.

5. Discussion

5.1. Water–Surrounding Rock Interaction

The water–rock interaction between coal mine water and surrounding rocks plays a critical role in mobilizing heavy metals and other elements, driven by oxidation, weathering and acid leaching processes. When exposed to air and mine water, sulfide minerals (e.g., pyrite) within the rocks oxidize, generating sulfuric acid and lowering the pH of the system. This acidic environment accelerates weathering rates and promotes the release of metals and metalloids into solution [64]. Batch leaching tests of shale, sandstone, limestone, and mudstone reveal that leachate pH ranges from acidic to neutral (2.80–6.42), with lower pH values correlating to higher concentrations of dissolved elements (Table 1). This suggests that water-soluble, exchangeable, and acid-soluble fractions dominate the release of heavy metals in the system [51]. The key trends in element mobility were discussed based on the rock types and geochemical conditions. Shale and sandstone yield acidic leachates with elevated Al, Fe, and Mn content. Sandstone releases higher TDS (2309 mg/L), reflecting a greater ionic load. Limestone produced a solution with near-neutral pH (5.85), reducing metal solubility. The notable Sr (123.0 µg/L) release in the batch solution suggests carbonate dissolution dominates during the coal mine water–surrounding limestone interaction. Sulfide-rich rocks (CMSR; e.g., shale, sandstone) exhibit the lowest pH values (2.80–2.95), creating conditions conducive to elements leaching. The acidic conditions enhance the dissolution of minerals, releasing Al, Fe, Mn, and Zn at high concentrations. It is deduced that the key factors driving element release may be sulfide oxidation and pH control.

The goaf water (MG, FMT, LG) exhibits extreme TDS (2528 mg/L to 3672 mg/L) and metal concentrations (e.g., Al, 191.4 mg/L in MG; Fe, 412.0 mg/L in LG), with a lower pH value of 3.26–4.71. In the batch test, shale leachate contains the highest Al concentration of 8.6 mg/L, but this surges to 191.4 mg/L in goaf water (MG), indicating a prolonged acid leaching in mine environments. Similarly, the highest Iron (Fe) release from mudstone is 69.3 mg/L, most likely due to pyrite oxidation. However, the highest Fe concentration in goaf water (LG) reaches 412.0 mg/L, reflecting the cumulative weathering effects in underground mines.

Except for Al and Fe, heavy metals with high mobility and release potential include Mn and Zn. These elements dominate leachates due to their solubility in acidic conditions. For instance, Mn in goaf water (MG: 60.2 mg/L) exceeds most rock leachates, suggesting enhanced mobilization in mine systems. Copper, As and Mo are heavy metals with moderate mobility. Copper (Cu) in shale leachate (3.3 µg/L) increases to 152.0 µg/L in goaf water (MG), while arsenic peaks in mudstone (10.65 µg/L) due to its association with sulfides or iron oxides. Lead, Cr and Cd showed limited release in rock leachates, but accumulate in goaf water, likely due to prolonged leaching or acidic enhanced dissolution. Strontium (Sr) is particularly enriched in goaf water (up to 7160 µg/L), likely sourced from primary and secondary carbonate dissolution under acidic conditions.

5.2. Variation of pH, TDS and Fluorine in Leachate Solution

The column leaching test investigated heavy metal mobilization from CMSR under dynamic flow conditions. As illustrated in Figure 3, the effluent pH exhibited distinct trends linked to geochemical processes. The pH dropped sharply from 6.7 (input solution) to about 2.5 within the first 20 mL of effluent. The sharp decrease in the pH value may be driven by the rapid oxidation of sulfide minerals (e.g., pyrite, FeS₂) in CMSR, which generates sulfuric acid in the following equation:

FeS₂ + 7/2O₂ + H₂O → Fe²⁺ + 2SO₄²⁻ + 2H⁺

(1)

The release of Fe²⁺ and the generation of sulfuric acid sharply lower the solution pH. Fe²⁺ may further oxidize to Fe³⁺, which hydrolyzes and releases more H⁺, intensifying acidity by the following equation:

4Fe²⁺ + O₂ + 4H⁺ → 4Fe³⁺ + 2H₂O

(2)

Fe³⁺ + 3H₂O → Fe(OH)₃ + 3H⁺

(3)

This acidic environment dissolves soluble minerals, such as sulfide (pyrite, sphalerite, chalcopyrite), clay minerals (kaolinite, chlorides, illite) and even trace content of metallic oxide or metal hydroxide, releasing cations (Fe²⁺, Al³⁺) and further lowering pH.

The solution pH slowly rose from 2.35 to 3.28 as leaching progressed, which reflects the buffering capacity of CMSR. The buffering mechanisms include the depletion of reactive sulfides, which reduces acid generation, ion exchange (e.g., H⁺ replacing Ca²⁺/Mg²⁺ on clay surfaces) moderates acidity, and the dissolution of carbonate-like minerals (e.g., calcite, dolomite) consumes H⁺ ions. The pH of the leaching solution stabilized near 3.3–3.5 after 300–440 volumes, reflecting the residual acid production from slower sulfide oxidation and limited buffering capacity during the leaching process.

The leaching experiment demonstrated significant variations in TDS that reflect the dynamic geochemical processes occurring within the CMSR system. TDS, representing the sum of all dissolved inorganic and organic constituents, exhibited a characteristic multi-phase release pattern that correlated strongly with pH changes and metal mobilization trends. The initial leaching stage (0–20 volume) showed an explosive increase in TDS, peaking at 1093 mg/L before slightly decreasing to 875 mg/L at 20 volume. This rapid early release indicates the flushing of highly soluble salts, including sulfates, chlorides, and carbonates, which are readily dissolved upon contact with percolating water. The concurrent pH drop from 6.7 to 2.35 suggests that acidic conditions generated by sulfide oxidation (e.g., pyrite weathering) enhanced mineral dissolution, contributing to the high TDS load. The peak TDS coincided with maximum concentrations of major cations and major heavy metals (Al, Fe, Mn, Cu and Zn), reinforcing that the initial flush represents the most aggressive phase of geochemical weathering in the system (Figure 3 and Figure 4).

Following the initial peak, TDS concentrations entered a prolonged decline phase, decreasing from 875 mg/L to 530 mg/L, with a relatively steady attenuation rate. This decline reflects the progressive depletion of the most soluble mineral phases and the establishment of slower, diffusion-controlled dissolution processes. However, the rate of TDS decrease was not linear, showing minor fluctuations that may correspond to localized pockets of more soluble materials being intermittently encountered, formation of new solid phases (e.g., gypsum, iron hydroxides) temporarily reducing dissolved loads and the periodic release of fine particulates that may contribute to TDS measurements.

5.3. Major Cations in Leaching Solution

The leaching behavior of major cations, such as Na⁺, K⁺, Ca²⁺, Mg²⁺, from CMSR is governed by distinct mineral weathering sequences [65,66]. These sequences control both the temporal release patterns and aqueous concentrations, as observed in the experimental data. The complex mobilization patterns of these elements directly reflect their diverse mineralogical sources and weathering susceptibilities. The dataset, which tracks cation concentrations across varying leaching volumes and pH conditions (initial pH 6.7 dropping to ~3.2–3.4), reveals distinct patterns that mirror the weathering sequences of primary minerals in acidic environments. This provides critical insights into the geochemical processes driving elemental mobilization, especially under conditions influenced by pyrite oxidation and sulfuric acid generation. The dynamic interaction of mineral dissolution, pH evolution, and secondary mineral formation governs the leaching behavior [67,68,69].

The first and most rapid weathering stage involves the immediate dissolution phase (water-soluble salts): the dissolution of highly soluble evaporite minerals, occurring within the initial leaching volumes (0–20 volume). This phase accounts for the dramatic concentration spikes observed for Na⁺ (10.73 mg/L) and K⁺ (1.48 mg/L). The dominant minerals undergoing dissolution include halite (NaCl), sylvite (KCl) and mirabilite/thenardite (Na₂SO₄). The extremely rapid Na⁺ release indicates an abundant halite presence, with its high solubility (360 g/L at 25 °C) enabling complete dissolution within the first leaching intervals. The subsequent steep decline in Na⁺ concentrations reflects the exhaustion of this readily soluble phase. Similar to halite, sylvite has high solubility (340 g/L at 25 °C), which explains the parallel K⁺ release pattern, although the greater retention of K⁺ in later stages suggests additional sources. The minor content of thenardite/mirabilite (Na₂SO₄) contributes to both the Na⁺ and sulfate loads in early leaching.

Following the initial flush of soluble salts, the system transitions to the acid-promoted weathering stage. The mineral dissolution is dominated by calcite, dolomite, aragonite (CaCO₃ polymorph), gypsum/anhydrite, and pyrite. The pyrite oxidation drives the system’s acidity, creating the low pH conditions that sustain carbonate dissolution. The continuous Ca²⁺ increase (reaching 47.34 mg/L at 100 volume) demonstrates sustained calcite dissolution (Equation (4)), enhanced by the acidic conditions (pH 2.35–3.04). The increased Mg²⁺ peak includes contributions from dolomite dissolution, although its slower kinetics result in a more extended release profile. The acidic condition also enhances gypsum/anhydrite (CaSO₄) dissolution. These sulfate minerals provide a continuous source of both Ca²⁺ and sulfate, with their moderate solubility (~2.4 g/L at 25 °C) maintaining a long-term release.

CaCO₃ + H⁺ → Ca²⁺ + HCO₃⁻

(4)

As the system progresses, weathering of more resistant minerals, such as silicates, feldspars, biotite/phlogopite micas and clay minerals becomes increasingly important. The persistent lower concentrations of sodium in these stages may derive from albite weathering (NaAlSi₃O₈), ion exchange from smectite clays, and dissolution of sodium-bearing micas. The subsequent Na⁺ release diminishes due to (1) depletion of readily soluble Na sources and (2) increased competition from H⁺ ions for mineral surfaces, slowing albite dissolution [70]. The K⁺ release is tied to the dissolution of K-bearing minerals, such as K-feldspar (orthoclase/microcline, KAlSi₃O₈) and micas (biotite, muscovite). However, K⁺ is more strongly retained in weathered residues compared to Na⁺ due to its incorporation into secondary clays (e.g., illite) or fixation in interlayer sites [71,72,73], which explains its significant decline. Except for carbonate sources, calcium/magnesium release shifts to silicate mineral breakdown at this stage. The calcium/magnesium-rich endmember (plagioclase feldspars, anorthite, CaAl₂Si₂O₈) weathers, contributing to the sustained Ca²⁺ and Mg²⁺ release. Meanwhile, as exchangeable cations in smectites/vermiculites, Ca²⁺ and Mg²⁺ may be released, governed by ion exchange equilibrium. The lower Mg²⁺ retention compared to Ca²⁺ is attributed to its weaker incorporation into secondary clays and greater mobility in acidic solutions. Clay minerals may slowly release incorporated cations through desorption and structural breakdown.

5.4. Dynamic Leaching of Major Heavy Metals and Fluorine

The heavy metals (Al, Fe, Mn, Cu, Zn and Sr), and fluorine are of increasing concern due to their potential ecological risks. They exhibited a complex and dynamic leaching behavior characterized by multiple concentration peaks and significant fluctuations throughout the leaching experiment, reflecting its sensitivity to both geochemical conditions and mineralogical controls within the CMSR system. The leaching profile of heavy metals can be divided into several distinct phases, each revealing important aspects of its mobilization and attenuation mechanisms (Figure 4). The initial leaching phase (0–20 volumes) showed a rapid increase in their concentrations (Al—11.3 mg/L, Fe—181.4 mg/L (30 volumes), Mn—2.5 mg/L, Cu 20.0 µg/L, Zn 1501 µg/L), followed by a rapid decrease despite continued leaching. This early peak suggests the presence of readily soluble metal-bearing phases, possibly including adsorbed heavy metals on mineral surfaces, metal sulfides (e.g., pyrite, chalcocite, covellite, sphalerite, marcasite, pyrrhotite), and amorphous or poorly crystalline metal phases.

The subsequent decline at 20–30 volumes may indicate the onset of secondary processes, such as adsorption onto newly formed Fe/Mn oxyhydroxides.

Following the initial phase, major heavy metal concentrations showed a general declining trend from 30 to 60 volumes. This dramatic decrease suggests strong attenuation mechanisms became dominant, likely including (1) the depletion of these easily mobilized heavy metal pools, (2) adsorption onto newly formed Fe/Al oxyhydroxides and clay minerals, and (3) precipitation as secondary minerals. However, this declining trend was interrupted by several notable resurgences, which demonstrate the non-linear and potentially reversible nature of heavy metals release processes. These secondary peaks may represent localized dissolution of heavy metal-bearing mineral grains, pH-dependent desorption from mineral surfaces, or redox-driven release from sulfide minerals [74,75].

The long-term leaching behavior (>200 volumes) revealed the persistent heavy metal release at significant concentrations (Al—0.08 mg/L to 0.4 mg/L, Fe—5.7 mg/L to 13.5 mg/L, Mn—0.06 mg/L to 0.25 mg/L, Cu—0.7 µg/L to 8.5 µg/L, Zn—54.5 µg/L to 255.5 µg/L). This enduring presence of Cu in the solution suggests continuous, slow-release mechanisms are operative, potentially including (1) oxidative dissolution of primary metal sulfides, (2) dissolution of metal-containing secondary minerals, or (3) desorption from organic matter or clay mineral surfaces. Notably, heavy metal concentrations never returned to the initial background levels, indicating that CMSR represents a long-term source of potentially heavy metal pollution in the environment.

Several key factors appear to control heavy metals’ mobility in this system. First, pH plays a crucial role, as their solubility is typically highest under acidic conditions (pH < 5) due to decreased adsorption and greater mineral dissolution rates. Second, the presence of competing cations (e.g., K⁺, Na⁺, Ca²⁺, Mg²⁺ and H⁺) can influence heavy metals adsorption/desorption equilibrium on mineral surfaces [76]. Third, redox conditions are particularly important for heavy metals, as they can exist in multiple states with differing solubility characteristics [77,78]. The intermittent nature of heavy metals release suggests that redox fluctuations within the column may have periodically created conditions favorable for their mobilization. Fourth, the formation of aqueous complexes with inorganic ligands (e.g., Cl⁻, SO₄²⁻, CO₃²⁻) or organic matter can significantly enhance heavy metals’ solubility and mobility [79].

The mobilization of strontium (Sr²⁺) from calcite (CaCO₃) in CMSR represents a particularly important but often overlooked geochemical process, as calcite serves as both a direct and indirect source of aqueous strontium through multiple mechanisms. While calcite is predominantly a calcium carbonate mineral, it frequently incorporates trace amounts of strontium through isomorphic substitution in its crystal lattice, where Sr²⁺ (ionic radius 1.18 Å) replaces Ca²⁺ (1.00 Å) during mineral formation [80]. Strontium (Sr²⁺) concentrations peak at 4.55 mg/L (Volume 30) before declining to about 0.2 mg/L. This behavior mirrors Ca²⁺ trends but with earlier attenuation, highlighting Strontium’s geochemical affinity with calcium. During initial acid-driven calcite dissolution (pH 2.35–2.80), incorporated Sr²⁺ is released congruently with Ca²⁺ according to the following Equation (5):

(Sr_1−x, Ca_x)CO₃ + 2H⁺ → (1 − x)Sr²⁺ + xCa²⁺ + H₂O + CO₂

(5)

As carbonates deplete, Sr²⁺ is increasingly adsorbed onto clay minerals (e.g., montmorillonite) or incorporated into sulfate minerals (celestite, SrSO₄). The decline in Sr²⁺ after volume 30 suggests saturation with respect to celestite (Ksp ≈ 10^−6.6), limiting its mobility.

The elevated fluorine (F⁻) concentrations observed in the leaching solutions likely originate from the sequential weathering of fluorine-bearing minerals under acidic conditions, with dissolution kinetics and mineral stability governed by pH, redox potential, and interactions with coexisting ions. The initial surge in F⁻ concentrations (0.02 mg/L at Volume 0 to 29.30 mg/L at Volume 10) coincides with the most aggressive acidic conditions (pH 2.35), suggesting rapid dissolution of highly reactive fluorite (CaF₂), Cryolite (Na₃AlF₆) or fluorapatite (Ca₅(PO₄)₃F), which are primary F⁻ reservoirs in coal bearing geological systems in Yangquan city. The dissolution of these fluoride-bearing minerals (CaF₂ + 2H⁺ → Ca²⁺ + 2HF) is strongly pH-dependent, with proton-promoted breakdown dominating at pH < 3.

Coal mine surrounding rocks often contain F⁻ substituted into aluminosilicate minerals (e.g., biotite, muscovite) and clays (e.g., illite, smectite), where F⁻ replaces hydroxyl (OH⁻) groups [81]. Acidic conditions protonate mineral surfaces, accelerating Al³⁺ and Fe³⁺ release while freeing structural F⁻ (Equation (6)). The early co-release of F⁻ with Al³⁺ and Fe³⁺ supports this mechanism. However, aluminosilicate weathering is slower than fluorite dissolution, resulting in a prolonged but diminishing F⁻ flux. Late-stage F⁻ stabilization (beyond volume 200) aligns with pH-neutralizing reactions (3.22–3.45) and secondary Al/Fe hydroxide precipitation. As leaching progresses, slower aluminosilicate weathering releases residual F⁻, while secondary Ca²⁺-mediated fluorite precipitation (when local [Ca²⁺][F⁻]² > K_sp) and Al/Fe hydroxide adsorption reduce dissolved F⁻. Rising pH promotes the formation of amorphous Al(OH)₃ and Fe(OH)₃, which strongly adsorb F⁻ via ligand exchange (Equation (7)) [82,83].

KAl₂(AlSi₃O₁₀)(OH)F + 10H⁺ +H₂O → K⁺ + 3Al³⁺ + 3H₄SiO₄ + HF

(6)

Al(OH)₃ + F⁻ → Al(OH)₂F + OH⁻

(7)

5.5. Trace Content Elements During Leaching Process

The leaching behavior of trace content elements (As, Pb, Cr, Cd, Co, Ni, Mo, and V) in CMSR (Figure 5) is governed by the dissolution of primary sulfides, oxides, and organic matter, modulated by pH, redox conditions, and competitive adsorption [84]. The dataset reveals distinct concentration patterns for these elements, reflecting their mineralogical hosts, weathering sequences, and sensitivities under evolving pH (2.35–3.45) and ionic strength conditions. Below, we dissect their sources, mobilization mechanisms, and interrelationships within the context of acid-driven weathering.

The mobilization of trace content elements in the leaching solutions is governed by a hierarchical mineral weathering sequence driven by pH, redox potential, and ligand availability. These elements exhibit distinct release patterns tied to their host mineral phases, with their interactions mediated by competitive adsorption, coprecipitation, and pH-dependent speciation. Below is a detailed analysis of their sources and interrelationships across weathering stages. The trace elements are predominantly hosted in sulfides, such as arsenopyrite (FeAsS), galena (PbS), sphalerite (ZnS), pentlandite ((Fe,Ni)₉S₈), violarite (FeNi₂S₄), and molybdenite (MoS₂) or as a trace impurity in pyrite (FeS₂). Pyrite oxidation drives initial acidity (pH 2.35–2.68) in the leaching system. Fe³⁺ from pyrite oxidation accelerates arsenopyrite (FeAsS), galena (PbS), sphalerite (ZnS), pentlandite ((Fe,Ni)₉S₈), violarite (FeNi₂S₄) and molybdenite (MoS₂) dissolution via proton/oxidant synergy. Under strong acidic conditions, the dramatic release of trace elements is observed at 10–20 volumes, revealing the key weathering mechanisms of sulfide mineral oxidation. The primary sources of trace elements derive from sulfide mineral dissolution, particularly the host sulfides. Acid generation from sulfide oxidation creates self-accelerating dissolution conditions. This explains the trace elements concentration increase within the first pH drop to 2.35.

During the leaching process, newly formed Fe/Al oxyhydroxides may further affect the adsorption–desorption behaviour of trace elements in leaching solution. As leaching progresses, dissolved Fe²⁺ oxidizes to Fe³⁺, precipitating as ferrihydrite (FeOOH) or schwertmannite (Fe₈O₈(OH)₆SO₄), while Al³⁺ hydrolysis forms gibbsite (Al(OH)₃). These phases strongly adsorb trace elements via ligand exchange. Under low pH (<3.0), however, Fe/Al oxyhydroxides remain unstable, limiting their adsorption ability and allowing high levels of dissolved trace elements. As pH rises to >3.0, forming stable Al³⁺/Fe³⁺ oxyhydroxides that sequester trace elements through surface complexation, explaining the gradual decline in trace elements concentrations post 30 volumes.

Therefore, the leaching of trace content elements from CMSR is a hierarchical process dictated by sulfide weathering, pH-dependent dissolution and adsorption, and redox transformations. Early acid generation liberates sulfide-hosted metals, while later silicate weathering and secondary mineralization modulate their mobility. Understanding these sequences is critical for predicting contaminant fluxes and designing targeted remediation in coal-impacted landscapes.

6. Conclusions

This study thoroughly examined the mechanisms and origins of acidic goaf water in abandoned coal mines, particularly focusing on how mine water interacts with surrounding rocks in the Yangquan coal mining area. The research revealed that the coal-bearing strata primarily consist of sandstone, mudstone, shale, and limestone, with a significant presence of pyrite, a sulfide mineral that plays a crucial role in generating acid mine drainage (AMD) through oxidative weathering. The goaf water samples collected from the field showed alarmingly high concentrations of iron (Fe) at 412.0 mg/L and aluminum (Al) at 191.4 mg/L, along with elevated levels of heavy metals, such as arsenic (As), lead (Pb), and cadmium (Cd), which were found to be significantly enriched in the surrounding rocks, surpassing crustal averages and presenting serious risks for environmental and ecological contamination. The fluorine content in the samples varied widely, ranging from 155.9 to 1885 mg/kg, with the highest concentrations found in sandstone. Sequential extraction analysis indicated that more than 80% of the fluorine is retained in residual phases, which limits its immediate mobility but raises concerns about long-term leaching risks. The oxidation of pyrite leads to the production of sulfuric acid, which drastically lowers the pH to approximately 2.5, facilitating the mobilization of heavy metals, such as aluminum, iron, manganese, and zinc, as well as trace elements such as arsenic, lead, and cadmium. Both batch and column leaching tests confirmed that acidic conditions significantly enhance the release of these toxic elements. Throughout the dynamic leaching processes, major cations including sodium (Na⁺), potassium (K⁺), calcium (Ca²⁺), and magnesium (Mg²⁺), along with heavy metals, such as copper (Cu), zinc (Zn), and strontium (Sr), exhibited distinct release patterns that were closely linked to the sequences of mineral weathering. Notably, strontium (Sr²⁺) and fluorine (F⁻) displayed pH-dependent mobility, with the formation of secondary minerals, such as iron and aluminum oxyhydroxides, influencing their long-term release dynamics. The presence of acidic goaf water presents considerable threats to groundwater quality, especially in ecologically sensitive karst regions where cross-layer pollution is a possibility. This research lays both a theoretical and practical groundwork for tackling the environmental issues associated with acidic goaf water, thereby aiding in the sustainable development of coal mining areas. The findings underscore the necessity for specific management strategies aimed at reducing contamination stemming from abandoned mines.