The coal samples in this study are composed of clay minerals (including kaolinite, illite, ammonium illite, and zeolite), oxide minerals (quartz and rutile), sulfide minerals (pyrite and sphalerite), carbonate minerals (calcite, dolomite, and siderite), and sulfate minerals (gypsum, pyrite) (Figure 3). Taking the coal sample from the CLB Mine as an example, pyrite (2.4%–38%) is the main mineral component, followed by ammonium illite (0%–5.8%) and a small amount of kaolinite (0%–4.4%) (Table 2).

The mineral content of the tonstein sample differs from that of the coal sample. In the TDY Mine, the main mineral type of tonstein is kaolinite (60%), followed by pyrite (1.2%). Kaolinite is the main mineral (51%) in the YD Mine, and contains trace amounts of pyrite (0.88%). In the CL mine, the main mineral of tonstein is kaolinite (with an average value of 51%), followed by pyrite (with an average value of 2.6%) and a small amount of calcite (with an average value of 1.2%). Tonstein in the JQ Mine is mainly composed of kaolinite (65%), followed by illite (4.4%). Only kaolinite (75%) and a small amount of rutile (4.8%) were detected in tonstein samples from the K1 coalbed of XX Mine.

Kaolinite is the main clay mineral in coal samples, occurring in strip or lens-shaped (Figures 4A, B). Kaolinite can also serve as a matrix for pyrite, rutile, and quartz particles (Figures 4C, D). A small amount of kaolinite is a fracture-filling cement (Figure 4A), formed by precipitation of aluminum and silicon-containing solutions. A small amount of kaolinite appears in the organic matter cell lumens (Figure 4C), indicating syndiagenetic-early diagenetic origin. Worm-like kaolinite may be of volcanic ash origin (Figure 4E). Illite has a low content and is often present as a cell-filling cement for microscopic components (Figures 4F, I), or dispersed in granular form in organic matter (Figure 4G).

In some samples, quartz and illite are co-filled in the cell lumens and micro-fractures of organic matter (Figure 4F), indicating a authigenic origin. In other samples, quartz occurs in granular form in organic matter matrix or kaolinite matrix (Figure 4G). The long axis of this type of quartz is parallel to the bedding plane and has a certain degree of roundness, belonging to terrigenous detrital origin. Calcite occurs in fractures or vein forms in organic matter (Figure 4H). It is suggested that calcite precipitates directly from Ca-rich hydrothermal fluids. In floor samples, calcite occurs in granular form in the kaolinite matrix (Figure 4M). Siderite is only observed in coal samples from the YD Mine and JQ Mine, filling the organic pores in an irregular shape (Figure 4J), suggesting a hydrothermal precipitation origin.

The occurrence forms of pyrite include 1) subeuhedral to euhedral crystal aggregates (Figures 4D, E); 2) Cell lumen-filling cement (Figure 4K); 3) Spherical polycrystalline framboidal (Figure 4I); 4) Other granular forms in organic matter matrix or kaolinite matrix (Figure 4P); 5) The fracture-filling cement (Figure 4L) is of hydrothermal origin.

The forms of gypsum occurrence include euhedral crystals (Figure 4N) and fracture-filling cements (Figure 4K). Gypsum clusters are concentrated near pyrite or iron sulfate, indicating that they are likely formed by the interaction between sulfuric acid formed by the oxidation of pyrite and calcite (Figure 4L). A small amount of rutile is distributed in small dispersed particles in the organic matter matrix (Figure 4P), indicating that rutile originates from the reprecipitation of Ti-rich solution. Anatase occurs as cylindrical form in the kaolinite matrix. Low roundness of the anatase indicates its terrigenous detrital origin. There are dissolved spaces in the self-shaped crystal rutile, which are filled with kaolinite (Figure 4O), indicating the influence of hydrothermal solutions.

The average content and proportion of major element oxides are shown in Table 3 and Figure 5A. The distribution of major elements in different coal mines exhibits strong heterogeneity. Taking CLB Mine as an example, its major elemental composition is Fe₂O₃ (7.05%), followed by SiO₂ (5.0%) and Al₂O₃ (2.96%). Compared with the average value of Chinese coal (Figure 5B), Si, Al, K, and Ti are more enriched in coal samples in this study, especially in the K1 Coalbed of XX Mine and the 4th Coalbed of Well ZK802. Fe, Na, and other minerals are relatively depleted, especially TY and YD Mines. Furthermore, Ca is depleted in the 4th coal seam of JQ Mine, K3 Coalbed of XX Mine, and the 4th Coalbed of ZK802 well, but is more enriched in TYD Mine and CL Mine. K is only depleted in the K3 Coalbed of XX Mine.

According to the enrichment coefficient proposed by Dai et al. (2012) (CC = element content/average trace element content in coal worldwide), the enrichment levels of critical metal elements are divided into six categories: CC<0.5 (depleted), 0.5<CC<2 (normal), 2<CC<5 (slightly enriched), 5<CC<10 (enriched), and 10<CC<100 (highly enriched) (Figure 6). The K1 Coalbed of XX Mine is mainly enriched in Se-Nb-Ta-Zr-Hf, while the K3 Coalbed of XX Mine shows Se enrichment. TDY Mine is enriched in both Zr-Nb-Be-Se-Hf-Ta and REY. CL Mine exhibits enrichment characteristics of Zr-Nb-Be-Se-REY-Hf-Ta. The YD Mine is characterized by enrichment of Se-Zr-Hf-Nb-Ta-REY-Be-Th, while the JQ Mine is characterized by enrichment of Se-Zr-Hf-Nb-Ta-REY-Mo-U. CLB Mine is characterized by Se-Mo-U enrichment. The coal samples from Well ZK802 have Se-Nb-Ta-Zr-Hf-REY-Mo-U-V enrichment, with a higher degree of enrichment of critical elements compared to samples from coal mines.

The Late Permian coalfields in northeastern Guizhou have enriched elements such as Nb, Ta, Zr, Hf, REY, Se, Th, and U to varying degrees, these enrichment element combinations were also found in adjacent coalfields/coal mines, such as MXP Mine, HYS Mine, SZ Mine, XD Mine, Northern Guizhou and Western Guizhou coalfields, etc. In addition, there were certain differences in the combination and content of enrichment elements among different mines in northeastern Guizhou coalfields (Table 4).

For example, the Zr content ranges from 8.9 to 3,083 μg/g, with an average value of 393 μg/g, far higher than the world hard coal (average value of 36 μg/g). It is also higher than the average value of Zr in XD, LPS, ZN, XY Mine and Northern Guizhou, Western Guizhou coalfields (<200 μg/g), slightly lower than the content of Zr in MXP, GX and SZ Mine (>400 μg/g), and lower than that in HYS Mine (695 μg/g). In the three coal mines of CL, YD, and JQ, the average value of Zr is relatively high, at 987 μg/g, 820 μg/g, and 886 μg/g, respectively. In the coal samples collected from the CLB Mine, the degree of Zr enrichment is the lowest (average value of 67 μg/g), but still higher than the Zr content in the world hard coal.

Figure 7 shows that critical metal elements such as Li, Nb, Ga, and REY are most enriched in or near tonstein, while their enrichment in roof (or floor) samples is relatively weak. The variation trends of Al2O3, ash content, kaolinite, and critical metal elements are basically consistent (Figure 7), indicating that detrital input is the main factor controlling the enrichment of trace elements.

The average REY content of coal samples from JQ and YD coal mines was the highest, at 660 μg/g and 621 μg/g, respectively (Table 5), followed by CL Mine. Compared with the other coalfields in Guizhou, the coal in the northeastern Guizhou coalfield is more enriched in REY. Compared with the coal samples from adjacent Mines, the content of REY is significantly lower in HYS, lower than that of MXP, SZ and XD (423 μg/g, 460 μg/g and 465 μg/g), and higher than the average content of REY in GX (276 μg/g). From the continental environment in the western study area to the marine environment in the east, there is a trend of increasing REY content in coal, which is related to the influence of terrestrial input on peat deposition (Zhuang et al., 2000). The classification and enrichment types of REY were based on the classification scheme proposed by Seredin and Dai (2012). Taking the K1 Coalbed of XX Mine as an example, LREY ranges from 157 to 328 μg/g (average 263 μg/g), and MREY ranges from 34 to 68 μg/g (average 52.0 μg/g). HREY ranges from 12 to 6.3 μg/g (average 9.4 μg/g), and LREY content is significantly higher than MREY and HREY. The δEu value ranges from 0.42 to 0.37, indicating a significant negative Eu anomaly. δCe ranges from 0.77 to 1.01, showing a weak negative Ce anomaly. Except for one sample characterized by weak negative La anomaly (<1), all other samples exhibit positive La anomalies (1.08–1.15). All samples showed varying degrees of positive anomalies in Gd (1.11–2.83).

As shown in the REY curves of K1 Coalbed in XX Mine (Figure 8A), the REY content of tonstein (Sample No. XXK1-2) is significantly higher than that of other samples. XXK1-2, XXK1-4, and XXK1-5 exhibit LREY enrichment with La_N/Yb_N>1. XXK1-1 and XXK1-3 exhibit MERY enrichment, with La_N/Sm_N<1 and Gd_N/Yb_N>1. Tonstein (Sample No. XXK1-2) showed significant negative Ce anomalies, while the rest samples had normal or weak negative Ce anomalies. All samples exhibit Eu negative anomaly. The REY characteristics of coal samples from other coal mines/drilling wells are shown in Figure 8. The coals of Well ZK802 occur weakly negative Eu anomalies only (Figure 8H), this may be explained by the influence of high-temperature hydrothermal solutions (Hower et al., 2015; Zhou et al., 2023).

Terrigenous detrital input has a significant controlling effect on the abnormal enrichment of minerals and trace elements in coal in southwest China (Ren, 2006; Dai et al., 2012; Shen et al., 2023). In order to determine whether terrigenous detrital input has an effect on the enrichment of key elements in coal in this area, principal component analysis (PCA) was used to explore the relationship between sulfur content, ash content, mineral composition and geochemical elements in coal (Figure 9). Based on the ordination diagram Figure 9A, we noted that the loadings angle of the variables in red dotted box such as Zr, Al₂O₃, REY, Ga, Hf, Ta, etc., is very small, and the projection value on the X-axis (PC1) is significantly larger and closer, while Mo, U, Fe₂O₃, Se has a larger projection value on the Y-axis (PC2). This indicates that the sources and genesis of Zr, Al₂O₃, REY, Ga, Hf, Ta, etc., Are similar, and they are obviously affected by terrestrial sources, but the enrichment of U, Mo, Se and Fe elements was of non-terrigenous origin. Similarly, we find that the kaolinite and ash content in Figure 9B are projected on PC1 with high values, which also proves that the enrichment of these elements in the red dotted box is obviously influenced by terrigenous detritus. In addition, we find that the loadings of sulfur content and pyrite content in Figure 9B are very close, and the projection value on PC2 is higher, which indicates that most of the sulfur in coal originates from pyrite, and also proves that the enrichment of U, Mo, Se is related to pyrite and sulfur.

The weathering products of ELIP in Kangdian Upland provide the main provenance of terrestrial input for the Late Permian coalfields in Yunnan and Guizhou provinces in SW China, and are an important basis for the enrichment of REY and trace elements in coals (Yang et al., 2011; Shen et al., 2021). In this study, Al₂O₃-TiO₂, Zr/TiO₂-Nb/Y, and Zr-Ti cross-plots were applied to discuss the influence of basaltic provenance in the Kangdian Upland on the Late Permian coalfields (Figure 10). The Al₂O₃/TiO₂ ratio can be used to determine the magma source of altered volcanic ash in coal-bearing strata. Hayashi et al. (1997) proposed that Al₂O₃/TiO₂ ratios ranging from 3–8, 8–21, and 21–70 respectively indicate the presence of ferromagnesian, neutral rock, and felsic provenance. Winchester and Floyd (1977) found the abundance and distribution of selected minor, trace elements (Ti, Zr, Y, Nb, Ce, Ga and Sc) and their cross-plot may be used to distinguish common volcanic rock types. Although elements including Al, Ti, Nb, and Y, can be relatively mobile in a few specific supergene environments (Ward, 2016; Dai et al., 2015). The Al₂O₃/TiO₂-Nb/Y, Zr-Ti and Zr/TiO₂-Nb/Y discrimination diagrams are used in this study as these elements appear immobile in our study (Liu et al., 2020).

The coal samples and clay rocks from the Well ZK802, as well as tonstein samples from the K1 Coalbed of XX Mine, all exhibit low Al₂O₃/TiO₂ ratio (3–8) (Figure 10A). In Figure 10B, the data points corresponding to these samples fall within the range of alkaline basalt and are close to Emeishan basalt. In Figure 10C, data points fall within the range of intraplate volcanic rocks, while in Figure 10D, they are close to the low titanium basalt of the Kangdian Upland, suggesting that the inorganic components in this sample come from the low titanium basalt of the Kangdian Upland. These samples have the characteristics of low Ti/Y (427–452) and low Nb/La (0.75–0.82), similar to the low titanium basalt of the Kangdian Upland (Ti/Y<500, and Nb/La ranging between 0.6 and 1.4). It is suggested that the inorganic components in coal samples and overlying clay rocks come from low titanium basalts in the Kangdian Upland. In study area, only coal samples from Well ZK802 showed varying degrees of enrichment in Sc, V, Cr, Co., Ni, Cu, Zn, and Se. This enrichment feature was discovered in coal samples from the Late Permian period, which were sourced from Kangdian Upland basalt (Xiao et al., 2004), indirectly confirming the view that the inorganic components in the Well ZK802 samples come from Kangdian Upland basalt.

Except for coal samples and clay rocks from CLB Mine and Well ZK802, almost all data points of other coal samples fall within the range of neutral rocks, indicating that the provenance belongs to neutral rocks (Figures 10A, B). Most studies propose that the characteristics of acidic rocks and intermediate acidic volcanic ash in the Upper Permian strata in Southwest China and the weathering debris input of acidic rocks at the top of the Emeishan basalt are related (Li et al., 2020; Zhao et al., 2019; Liu et al., 2020). Figure 9C shows that most of the coal samples and tonstein samples from TDY, CL, YD, and JQ Mines fell into the rough rock and alkali flow rock areas, and overlapped with the Emeishan acidic volcanic rock area, indicating that the main source of inorganic components in these coal mines is the input of intermediate acidic volcanic ash. In addition, the data points of these samples are concentrated within the range of Permian-Triassic sedimentary rocks with unknown volcanic rock provenance (Figure 10D). Moreover, in Figure 10E, these data points are concentrated in the alkaline rock range of Emeishan, which also implies the presence of volcanic rock provenance. Comparison of REY distribution patterns shows that these coal samples and tonstein samples differ significantly from ELIP felsic intrusive rocks (such as granite and diorite) (Figure 11), but are more similar to alkaline rhyolite, indicating the input of volcanic ash from eruptions rather than the weathering products of felsic intrusive rocks.

The mineral content and morphological characteristics in coal samples also indicate that the inorganic components come from the input of intermediate acidic volcanic ash. The quartz content is low (Table 2), and it has not even been detected in most samples. Under SEM, quartz occurs as extremely fine particles (Figures 4F, G), belonging to authigenic origin rather than terrestrial detrital origin. It has been reported that the intermediate acidic rocks at the top of the Emeishan basalt have a high quartz content (Shellnutt and Jahn, 2010). Quartz deposition in peat bogs can lead to an increase in quartz content in coal seams and surrounding rocks (Liu et al., 2019). In addition, worm-like kaolinite was observed in coal samples (Figure 4E), which also confirms that the provenance belongs to volcanic rocks (Dai et al., 2017). Vertically, the Al₂O₃/TiO₂ of tonstein is significantly higher than that in coal sample, belonging to the range of felsic provenance. This indicates that the increased input of intermediate acidic volcanic ash led to the termination of peat accumulation and the formation of interbedded tonstein (Figures 12D–G). The variation of Al₂O₃/TiO₂ in coal samples is small. However, Al₂O₃/TiO₂ showed high values (>21) in the samples near the tonstein and at the bottom of the coalbed (Figures 12D–G), reflecting input from intermediate acidic volcanic materials.

Most of the sample data points from XX Mine fall within the range of neutral provenance (Figures 10A, B), while a small number of coal samples and tonstein samples fall within the range of ferromagnesian provenance. The data points corresponding to the K1 Coalbed are distributed at the common boundary of coarse-grained rock and coarse-grained andesite, as well as at the common boundary of coarse-grained andesite and rhyolite (Figures 10C–E). The tonstein sample is close to the range of Emeishan basalt. The data points corresponding to the K3 Coalbed are distributed within the range of coarse andesite and alkaline basalt. Compared to samples influenced by intermediate acidic volcanic rocks, the samples from XX Mine are closer to the alkaline basalt range. Liu et al. (2019) proposed that the main factor contributing to the enrichment of REY and critical elements in coalfields of western Guizhou Province is the input of alkaline volcanic material from the top of ELIP. It can be inferred that the plagioclase neutral rocks at the top of ELIP are the main provenance of XX Mine (Dai et al., 2016; Xie et al., 2017). Similar to the coal samples from southern Sichuan Province, there is a negative Ce anomaly in the samples from XX Mine, indicating a provenance dominated by felsic or felsic neutral rocks (Dai et al., 2015).

The coal samples from XX Mine have a high quartz content. Quartz particles are large and have a certain degree of roundness, indicating terrestrial debris origin. Meanwhile, some small self-generated quartz particles have also been observed, which are the weathering products of basalt in the Kangdian Upland or siliceous hydrothermal precipitation products (Dai et al., 2012). Vertically, some samples have low Al₂O₃/TiO₂, corresponding data points falling within the range of mafic provenance (Figures 11A, 12B), suggesting terrigenous detrital input from Kangdian Upland basalt.

Most of the clay rocks and coal samples from Well ZK802 belong to alkaline basalt (Figure 12E). The low titanium Emeishan basalt exhibits sub-alkaline characteristics, indicating that these samples have other provenance regions. The deposition of alkaline volcanic ash in coal-accumulating environments is an important factor in the enrichment of critical metal elements in Late Permian coal in the study area (Dai et al., 2016; 2017). Some studies have suggested that the felsic inorganic components in coalfields in western Guizhou come from remote volcanic arcs or Paleozoic orogenic belts, rather than ELIP (Liu et al., 2020). The coal samples from CLB Mine have the highest Al₂O₃/TiO₂ content, indicating that the inorganic components come from the input of felsic substances. The Nb/Y-Zr/TiO2 cross-plot indicates the input of alkaline and sub-alkaline intermediate acidic volcanic ash (Figure 10C). The extrusive rocks induced by the Emeishan mantle plume belong to the alkaline - peralkaline category (Hei et al., 2018), providing an alkaline-peralkaline provenance for the study area. The acidic volcanic ash at the Permian-Triassic boundary in southern China provides sub-alkaline volcanic debris (Liao et al., 2016). In Figure 9D, the coal sample from CLB Mine is closer to the arc volcano range, while most of the data points fall within the Late Permian-Early Triassic volcanic rock range (Figure 10E), showing the contribution of more distant volcanic arcs to inorganic components. In the coal samples from CLB Mine, the element enrichment combination is Se-Mo-Ge-Nb-U, which differs significantly from the enrichment combinations of Kangdian Upland basalt, felsic volcanic rocks at the top of basalt, and intermediate acidic volcanic ash, but is similar to two reported coalbeds in western Guizhou Province (Liu et al., 2020). The inorganic components of these two coalbeds are considered to originate from unknown volcanic arcs and/or distal orogenic belts. Vertically, Al₂O₃/TiO₂ shows a decreasing trend, indicating a continuously decreasing input intensity of volcanic arcs and/or distal orogenic belt debris (Figure 12C).

The presence of ammonium illite is direct evidence of high-temperature fluid activity, and it is commonly observed in samples from CLB Mine. Ammonium illite occurs in two ways in coal: 1) illite is formed by alteration through the substitution of NH₄ ⁺ and K⁺ (Williams et al., 1992); 2) Kaolinite and nitrogen interact with coalification to form (Dai et al., 2012). Illite was not detected in the CLB Mine samples, therefore the formation of ammonium illite is related to the second way. The nitrogen involved in the formation of ammonium illite may come from bacterial decay and thermal decomposition of organic matter, reacting with H+ to form NH4⁺. Under reducing and acidic conditions, NH4⁺ easily enters the pore fluid. Ammonium illite occurs in pores or cell lumen, coexisting with kaolinite, indicating that ammonium illite may originate from the kaolinite precursor produced by the interaction between kaolinite and NH4⁺ during coalification process (Williams et al., 1989; 1992).

Kaolinite was detected in coal and tonstein samples from TDY, YD, CL, and XX mines, while other types of clay minerals were not detected. However, various types of clay minerals were detected in the samples of CLB, JQ Mine, and Well ZK802. The clay minerals in the coal samples of study area are mainly kaolinite, which is induced by the decomposition of debris in the highly acidic solution of peat swamp, promoting the transformation of other aluminosilicates into kaolinite. In the coal sample from Well ZK802, zeolite is also evidence of low-temperature hydrothermal fluid activity. Letoshite is mainly formed in some low-temperature hydrothermal alteration zones and widely replaces potassium feldspar along this vein, or it may have been formed during low-grade metamorphism (Susilawati and Ward, 2006). Kaolinite, montmorillonite, zeolite, and plagioclase can all form zeolite through weathering. The process of kaolinite deformation into zeolite requires additional material supply, including SiO₂, Na⁺, and K⁺, and the most likely provider of these substances is hydrothermal fluid (Hower and Gayer, 2002). Due to the fact that the coal sample mainly contains kaolinite and quartz, this indicates that zeolite is mainly formed through low-temperature hydrothermal alteration.

Carbonate minerals and sulfate minerals can also confirm hydrothermal fluid activity. Calcite is widely distributed in various coalbeds in the study area. Except for CL Mine, no carbonate minerals were detected in the tonstein and floor samples of other mines. Calcite mainly appears in the form of fracture cement in coal (Figure 4H). The fractures formed after coalbed rupture provide permeable channels for Ca²⁺ rich hydrothermal fluids to pass through the coalbed (rather than the tonstein). The gypsum content that occurs as fracture cement is low and coexists with pyrite (Figures 4I–K). Iron sulfides near fractures are easily oxidized, and fluids rich in oxygen and calcium may be more likely to promote the production of iron containing sulfate minerals along fractures (Figure 4L) (Dai et al., 2013). In addition, the authigenic small quartz (Figure 4F), fracture-cemented pyrite (Figure 4L), and authigenic siderite (Figure 4J) all suggest the input of hydrothermal fluids.

As a high field strength element, Nb-Ta-Zr-Hf-REY is believed to not migrate under low-temperature hydrothermal conditions (Arbuzov et al., 2019), migration occurs only under the action of strong acidic hydrothermal solution (Spears and Arbuzov, 2019). In Figure 13, all data points of the tonstein samples fall within the CHARAC range. Some data points of coal samples fall outside the CHARAC range, indicating that the fluid has fractionated Zr-Hf and Y-Ho (Bau M.,1996). For samples from different coal mines, Nb/Ta and Zr/Hf vary greatly, especially in CL, YD, and JQ coal mines (Table 6). Under the action of acidic fluid, the two elements with strong migration ability, Nb and Zr, undergo migration, resulting in fractionation of the combination of Nb-Ta and Zr-Hf elements. The Nb/Ta and Zr/Hf ratios in the coal samples are higher than those in the tonstein sample, indicating that Nb and Zr are relatively more actively leached from the tonstein and then deposited in the underlying organic matter. The re-deposition of elements is the reason for the enrichment of Nb and Zr in multiple coalfields in southwestern China (Dai et al., 2013). Meanwhile, few samples in the present study are characterized by Gd-maximum as well as MREY enrichment (Figures 8A, B), which is indicative of acid solutions (e.g., high P_CO2-waters) (Shand et al., 2005).

Compared to Nb, Ta, Zr, and Hf, REY has a higher mobility in aqueous solutions (Liu et al., 2019). Therefore, the vertical variation of REY content is inconsistent with Nb, Ta, Zr, and Hf. In XX, CL, YD, and JQ coal mines, the REY/Zr ratio of the interbedded tonstein is lower than that of the underlying coalbed (Figure 14), indicating that some REY leaches out of the interbedded tonstein and subsequently re-enriches in the underlying coalbed. Meanwhile, in the tonstein of TDY and CL coal mines, the REY/Zr ratio is higher than that of the underlying coalbed (Figures 14B, E), indicating that the overlying coalbed provides a source of REY. Under the action of gravity, fluids rich in REY enter the underlying tonstein. However, due to the barrier effect or changes in acidity or alkalinity, REY precipitates in the tonstein and no longer penetrates downwards. Acidic water can cause leaching and migration of REY, while alkaline water can cause re-precipitation of REY. The variation of U/Th can be used to reflect the migration and re deposition of acidic aqueous solutions between coalbeds and tonstein. In all coalbeds, the U/Th of the interbedded tonstein is lower than that of the underlying coal sample, indicating that U leaches out of the interbedded tonstein and then re precipitates in the underlying coalbed. It is worth noting that in the samples from the top of the coalbed, U/Th is higher than in other positions. This reflects the leaching of U elements from the overlying roof, which entered the top coalbed and led to an increase in U/Th. In addition, the roof composed of carbonate rock is easily influenced by seawater and may contain a high content of U, which may also lead to a high U content in the coalbed.