Copyright © the author(s). This work is licensed under a Creative Commons Attribution 4.0 International License. DOI:10.14800/IOGR.1189 Received April 12, 2021; revised July 2, 2021; accepted August 9, 2021. *Corresponding author: liushiqi@cumt.edu.cn 1 Effects of CO2 on the Micron-Scale Pore- Fracture Structure and Connectivity in Coals from the Qinshui Basin Shiqi Liu*, Shuxun Sang, Tian Wang, Yi Du, and Huihuang Fang, China University of Mining and Technology, Xuzhou, China Abstract Changes in the micron-scale pore and fracture structure in coal caused by CO2 are critical for CO2 injection and CH4 production in coal seams. To investigate the effects of CO2 on the characteristics and connectivity of micrometer-scale pores and fractures in coal, four coal samples from the Qinshui basin were selected. These samples were exposed to CO2 and water for 240 hours at 80 °C and 20 MPa using a CO2 geochemical reactor. X-ray micro-CT (computed tomography), field emission scanning electron microscopy (FESEM), and energy dispersive spectroscopy (EDS) were used to identify the characteristics and connectivity of micron-scale pores and fractures in the coal samples before and after CO2 treatment. Then, the influence of mineral dissolution on micrometer-scale pores and fractures was discussed. After CO2 treatment, the massive dissolution of carbonate minerals significantly increased the pore contents and volumes of the coal samples. For this reason, the grain size and volume of carbonate minerals determined the increase in pore number and volume after CO2 treatment. The dissolution of calcite, dolomite, and other carbonate minerals in the coal matrix formed a large number of pores created by dissolution <10 μm in diameter, which affected the number of pores in the coal after CO2 treatment, but contributed little to the connectivity of the coal at the micro-scale. The carbonate minerals that filled the microfractures were heavily dissolved in CO2, increasing the aperture and connectivity of the microfractures. The dissolution of carbonate minerals in microfractures was the major contributor to the increase in the volume of pores >50 μm in diameter and the main reason for the increase in coal connectivity at the micrometer scale. Introduction The adsorption of CO2 is superior to that of CH4 in coal seams (Day et al. 2008; White et al. 2005). By injecting and storing CO2 in coal seams, CH4 can be displaced by CO2 and expelled from the coal seam, thereby improving CH4 recovery and reducing CO2 emissions (White et al. 2005; Hol et al. 2014; Fujioka et al. 2010). This technique is called CO2 geological storage-enhanced coalbed methane recovery (CO2- ECBM). CO2-ECBM has environmental and energy benefits and has quickly become one of the hot spots in research on coalbed methane and emissions reduction (Hol et al. 2014; Fujioka et al. 2010). The United States, Canada, the Netherlands, Japan, and China have conducted pilot tests of CO2-ECBM, and the results are satisfactory (Fujioka et al. 2010; Faiz et al. 2007; Pan et al. 2018; Wong et al. 2007). Mixing of CO2 and water forms an acid fluid containing H2CO3, which can dissolve calcite, dolomite, magnesite, and other minerals and promote Ca and Mg migration (Bertier et al. 2006; Dawson et al. 2015; Du et al. 2018; Hayashi et al. 1991; Kolak and Burruss 2014; Liu et al. 2018). The CO2-induced migration, dissolution and precipitation of inorganic minerals in coal change the structure of the coal, eg, opening some closed and semi-closed pores in the coal, changing the distribution of pore size in the coal, and increasing the porosity and permeability of the coal (Anggara et al. 2013; Kutchko et al. 2013; Liu et 2 al. 2015; Liu et al. 2010 ;Massarotto et al. 2010; Liu et al. 2019). CO2 can also dissolve minerals filling coal fractures, thus increasing the aperture and connectivity of coal fractures and changing the mechanical properties of the coal (Anggara et al. 2013; Massarotto et al. 2010; Perera et al. 2011; Ranjith and Perera 2012). Other scholars have suggested that the reaction between the CO2-H2O system and coal is a long- term process in which dissolved mineral components can migrate and precipitate in fractures that have not been filled with minerals or dissolved with CO2, thus reducing their connectivity (Du et al. 2018; Kutchko et al. 2013; Xu et al. 2016; Zerai et al. 2006). The characteristics and connectivity of pores and fractures in coal determine the storage, diffusion, and migration of CH4 and CO2 in coal (Liu et al. 2015; Wang et al. 2017; Zhou et al. 2018). It is generally believed that the micron-scale pores in coal are mainly secondary gas pores and dissolution-created pores, while the fractures are micro-fractures and some small-scale cleats (Liu et al. 2015; Liu et al. 2016; Liu et al. 2017). These pores and fractures are the main seepage channels of CH4 and CO2 and connect the microscopic structure (e.g., adsorption pores and diffusion pores) and the macroscopic structure (e.g., macroscopic fractures) of the coals (Liu et al. 2015; Liu et al. 2016; Liu et al. 2017). Therefore, changes in the characteristics and connectivity of micron-scale pores and fractures in coal determine the injectivity and storage capacity of CO2 and the production of CH4, which are crucial factors in CO2-ECBM. Research on changes in coal structures caused by CO2 reactions was mainly focused on the nano- to sub- micrometer scale and the macroscale. The changes in the characteristics and connectivity of pores and fractures at the micron-scale are still unclear. In this paper, using typical low-volatile bituminous coal and anthracite coal as examples, X-ray microCT (computed tomography), field emission scanning electron microscopy (FESEM), and energy disperse spectroscopy (EDS) were used to study the effects and mechanisms of CO2 on the characteristics and connectivity of micrometer-scale pores and fractures in coal. This study aims to provide a better understanding of the effectiveness of CO2 injection and CH4 production. Samples and Methodology Samples. Four groups of coal samples were collected from the Qinshui basin, China, including low- volatile bituminous coal from the Xinyuan Mine, semi-anthracite coal from the Yuwu Mine and the Xinjing Mine, and anthracite coal from the Sihe Mine. These samples were named Coal #1 to Coal #4 (Table 1). The coal samples were systematically collected from the working faces of the coal mines. The collection, retention, and preparation of the coal samples were conducted in line with the relevant standard GB/T 19222-2003 in China and the international standard ISO 7404-2:1985. To prevent further oxidization, coal samples were wrapped in absorbent paper, hermetically sealed in plastic bags and stored at 5 °C after sample collection. The key properties of these samples are shown in Table 1. Table 1—Properties of the coals used. Samples Sampling location Ro, max (%) Proximate (wt. %) Ultimate (wt. %) Mad Aad Vdaf FCad Odaf Cdaf Hdaf Ndaf #1 Xinyuan Mine 1.81 0.81 5.35 15.26 80.20 9.30 80.32 4.43 1.14 #2 Yuwu Mine 2.19 1.10 11.98 13.44 76.19 2.44 91.73 4.12 2.44 #3 Xinjing Mine 2.64 1.66 10.02 10.10 80.89 3.05 91.52 3.96 1.06 #4 Sihe Mine 3.33 1.48 13.12 6.32 81.39 2.98 93.45 2.15 1.00 Note: Ro, max, the mean maximum reflectance values of vitrinite; wt. %, weight percent; Mad, moisture; Aad, ash yield; Vdaf, volatile matter; FCad, fixed carbon content; Odaf, oxygen content; Cad, carbon content; Had, hydrogen content; Nad, nitrogen content; “ad” means air-dried basis; “daf” means dry ash-free basis. CO2 Treatment. CO2 treatments were performed to replicate a burial depth of 2000 m. The temperature and pressure at this burial depth (80 °C and 20 MPa, respectively) were calculated from the temperature and depth of the sub-surface constant temperature zone, the average geothermal gradient, and the average pressure gradient at the sampling location. The coal samples chosen for CO2 treatment consisted of small coal pillars for X-ray CT and bulk coal for scanning electron microscopy (SEM) analysis. Details of the high-pressure reactor and the experimental duration used in CO2 treatment can be found in our previous 3 studies (Liu et al. 2018; Liu et al. 2019). After the CO2 treatment, the coal samples were vacuum dried at 50 °C for 24 hours for the X-ray CT scan and SEM analysis. Pore-fracture Network Modelling. X-ray CT scan. X-ray CT scanning was performed with an Xradia 520 Versa X-ray CT scanner produced by the Carl Zeiss Foundation Group. Samples for X-ray CT scanning were small coal pillars approximately 2 mm in diameter and 2 mm in height. These were drilled from bulk coal samples using a mechanical sampler. The scanning area of the X-ray CT scan was 1 mm in diameter and 1 mm in height. The total scan number was 1000 and the voxel resolution was 1.0 μm. After the X-ray CT scan, the small coal pillars were loaded into 800 mesh nylon bags that are resistant to high temperatures and corrosion for CO2 treatment. After CO2 treatment, X-ray CT scans were taken again of the small coal pillars. To compare the X-ray CT results, the scanning range, total scan number, voxel resolution, and scanning position of the small coal pillars before and after CO2 treatment were the same. To ensure the same scanning areas before and after CO2 treatment, the central point of each small coal pillar was identified as the center of the scanning area. Due to the manually set scanning area, slight errors may exist. However, the X-ray CT results show that these errors have a weak impact on the research and can be ignored. Establishing the pore-fracture network model. Three-dimensional (3D) digital models of coal were established using Avizo 9, which is professional software for 3D digital cores based on X-ray CT images. The process of establishing the 3D digital model of the coal includes several steps, such as 3D imaging reconstruction, image denoising, image binarization, and model construction. Threshold selection is the key to identifying pores and fractures during the binarization process. In this paper, the X-ray CT images were first converted into 8-bit TIFF bitmaps (Tag image file format), and then their greyscales were normalized to the range of 0-255. Thus, all bitmaps had 256 greyscales and the same grey range. Threshold segmentation of the X-ray CT images revealed that the grey ranges for pore fractures, organic matter and minerals are 0-110, 110-180 and 180-255, respectively. Based on the 3D digital model of coal, the characteristics of pores, fractures and minerals, including porosity, pore size distribution, pore volumes, mineral grain size distribution, and mineral volumes, were further extracted. The maximum inscribed ball method was used to extract the pore size and grain size of the mineral, and the equivalent diameters (EqDiameters) of the pores and minerals were obtained. Furthermore, the equivalent pore-fracture network models which are ball-and-stick models, and interconnected pore-fracture models, were established, and the coordination numbers of pores and fractures and throat lengths were extracted. Due to the large number of calculations and limited calculation capacity of the workstation, cubic ball-and-stick models and interconnected pore-fracture network models of coal samples that were 500 μm on each side were established. According to the maximum inscribed ball method, the fractures are filled with a number of balls and cut into a number of pores. Therefore, the contents, volumes, and numbers of pores extracted from the 3D digital model contain the contents, volumes, and numbers of fractures. When establishing the ball-and- stick model, a series of balls filling the fractures are identified as throats. For this reason, fractures are generally considered throats in the ball-and-stick model. Scanning Electron Microscopy Analysis. Pores, micro-fractures, and minerals in coal before and after CO2 treatment were investigated using a Sigma 300 FESEM instrument produced by the Carl Zeiss Foundation Group, Germany, with a QUANTAX 200 EDS produced by Bruker Company, USA, with amplification from 103 to 105. Samples used for FESEM were bulk coal. The coal samples were polished into small samples approximately 10-30 mm across and 4-5 mm high using a polishing and burnishing machine. Then, the small samples of coal were polished using a cross section polisher. After FESEM analysis, coal samples were first loaded into 800 mesh nylon bags for CO2 treatment and then investigated again using FESEM to observe changes in pores, fracture, and minerals after CO2 treatment. Coal is known as a non-conducting substance. Therefore, to achieve better experimental results, a thin gold coating is commonly applied to coal samples by sputtering. In this study, the gold coating can hamper reactions between coal samples and CO2. Therefore, instead of a gold coating, before CO2 treatment, the sub-face and side faces of each coal sample were wrapped in conductive tape. After CO2 treatment, to achieve better experimental results, a thin gold coating was applied to the coal samples by sputtering. The grain size of the mineral is small in coal and the minerals are difficult to locate. A 4 photograph of the minerals was first taken using the back-scattering mode with low amplification. With the help of the Advanced Mineral Identification and Characterization System (AMICS), the mineral compositions were initially identified, and the typical minerals were marked in the photographs. Then, using the secondary electron mode and increasing the amplification step by step, details on the surface of typical minerals were observed. AMICS is the latest software package for automated identification and quantification of minerals and synthetic phases (Du et al. 2018). Combined with EDS, AMICS can be used to synthesize SEM images and automatically determine the compositions of minerals with grain sizes greater than 4 μm via an amplification of 200 (Du et al. 2018). Results and Discussion Changes in Pore Structure. Pore Content. Before CO2 treatment, the pore contents of the coal samples are relatively low, ranging from 0.84-2.18 % (Table 2). After CO2 treatment, the pore contents (1.22- 5.93 %) of the coal samples increase significantly, with an average increase of 82.61 % (Table 2). Changes in pore volumes show the same trend as changes in pore contents (Table 2). Table 2—Volumes and contents of pores and minerals in coal samples based on X-ray CT. Samples Pore content, % Mineral content, % Pore volume, μm3 Mineral volume, μm3 Before After Before After Before After Before After #1 2.18 3.33 5.43 3.33 1.62E+07 2.46E+07 4.01E+07 2.46E+07 #2 1.20 1.54 3.77 2.83 8.86E+06 1.14E+07 2.78E+07 2.09E+07 #3 1.95 5.93 8.18 3.73 1.48E+07 4.50E+07 6.22E+07 2.83E+07 #4 0.84 1.22 3.19 2.40 2.03E+07 2.98E+07 7.72E+07 5.86E+07 Pore Number. Before CO2 treatment, the pores in the coal samples are primarily <5 μm in eqDiameter, and the pore number decreases rapidly with increasing pore eqDiameter (Figure 1). The number of pores >10 μm in eqDiameter is small (Figure 1). After CO2 treatment, the number of pores <2 μm of Coal Sample #2 changes slightly, while the number in the other coal samples decreases (Figure 1). The number of pores ranging from 2-10 μm in eqDiameter of Coal Sample #3 decreases, while the number of the other coal samples increases slightly (Figure 1). Moreover, the number of pores >10 μm in eqDiameter changes slightly (Figure 1). In general, changes in pore number are not significant, indicating that changes in pore content and volume are not determined by changes in pore number. Pore Volume. Before CO2 treatment, the pore volumes of the coal samples are primarily associated with pores <10 μm in eqDiameter, with pores >10 μm in eqDiameter accounting for a small volume (Figure 2). The volumes of pores <50 μm in eqDiameter follow normal distribution, and the peak of pore volume ranges from 3 to 4 μm (Figure 2). With the increase and decrease in the pore eqDiameter, the pore volumes of the coal samples decrease rapidly (Figure 2). Coal Sample #3 and #4 have larger pores >50 μm in eqDiameter, which may result from the presence of micro-fractures (Figure 2). Combined with the pore number distribution, pores >10 μm in eqDiameter contribute greatly to the pore volume. After CO2 treatment, the pore volume distribution of the coal samples changes significantly, which is mainly caused by the significant increase in the volumes of pores >50 μm in eqDiameter (Figure 2). In addition, the volumes of pores <50 μm in eqDiameter of Coal Sample #2 exhibit no marked changes, while the volumes of pores <4 μm in eqDiameter in Coal Sample #1, #3, and #4 decrease slightly (Figure 2). 5 Figure 1—Pore number distribution of coal samples based on X-ray CT. Figure 2—Pore volume distributions of coal samples based on X-ray CT. Changes in Connectivity. Interconnected Pore Models of Coal Samples. Before CO2 treatment, there are a large number of pores and a certain number of micro-fractures in the 3D digital models of the coal samples (Figure 3A). The 3D digital models and the ball-and-stick models show that micro-fractures are throats in the ball-and-stick models (Figure 3B). Although some pores and microfractures are connected 6 (Figure 3B), the connectivity is weak, resulting in poor connectivity in the coal samples. Therefore, the interconnected pore models of the coal samples cannot be extracted (Figure 3C). Notes: A, 3D digital models of coal; in cubes, pores and microfractures are red, minerals are blue, and organic matter is grey; B, ball-and-stick models of pores and throats; in ball-and-stick models, the pores are red, and the throats are green; and C, interconnected pore models. Figure 3—3D digital models, ball-and-stick models, and interconnected pore models of coal samples based on X-ray CT before CO2 treatment. After CO2 treatment, some minerals filled the pores and microfractures, increasing the connectivity of the pores and microfractures (Figure 4A). Ball-and-stick models show that the number of pores and throats interconnected in the coal samples increases (Figure 4B). However, the increase in connectivity of pores and micro-fractures caused by CO2 does not obviously improve the connectivity of the coal samples on the macro-scale (Figure 4C), and the connectivity of pores and micro-fractures is still limited to local areas in the coal samples and does not extend throughout the whole coal samples (Figure 4C). After CO2 treatment, the interconnected pore models of Coal Sample #1 and Coal Sample #2 were successfully extracted (Figure 4C). By comparing the interconnected pore models and 3D digital models of Coal Samples #1 and #2, the connectivity of coal samples is improved by a micro-fracture (Figure 4A and 4C), which means that the connectivity of coal samples on the micron-scale is mainly contributed by micro-fractures, while the contribution of pores is weak. Figure 4—3D digital models, ball-and-stick models, and interconnected pore models of coal samples based on X-ray CT after CO2 treatment. Coordination Numbers. The coal samples are dominated by isolated pores (pores with a coordination number of 0), and the number of pores with coordination numbers > 0 only accounts for 0.56-7.83 % (Table 3). The coordination numbers of the interconnected pores (pores with coordination numbers >0) are low, mainly 1-2 (Figure 5). The number of pores with a coordination number >2 decreases rapidly 7 (Figure 5), indicating that the connectivity of pores and fractures on the macro-scale is weak and that pores are only connected with 1-2 adjacent pores and throats. Table 3—Numbers and contents of pores with different coordination numbers. Samples Pore number Pore content, % Before After Before After 0 >0 Total 0 >0 Total 0 >0 0 >0 #1 1029934 16289 1046223 1005892 34226 1040118 98.44 1.56 96.71 3.29 #2 961181 81686 1042867 948262 94502 1042764 92.17 7.83 90.94 9.06 #3 1021998 5804 1027802 343714 69641 413355 99.44 0.56 83.15 16.85 #4 982578 60727 1043305 968959 73782 1042741 94.18 5.82 92.92 7.08 Notes: “0”, pores with a coordination number of 0; ”>0”, pores with coordination numbers >0. Figure 5—Coordination numbers of coal samples based on X-ray CT. After CO2 treatment, the total number of pores in coal samples remained unchanged, except for Coal Sample #3 (Table 3). The number of pores with coordination numbers >0 increases, while the number of pores with a coordination number of 0 decreases correspondingly (Table 3), indicating that CO2 improves the connectivity of the pores in coal and has little influence on the number of pores. Although the total number of pores and the number of pores with a coordination number of 0 in Coal Sample #3 decrease, the number of pores with coordination numbers >0 increases significantly (Table 3). This is because a large number of pores in Coal Sample #3 become connected to form larger pores. The number of pores with coordination numbers >1 in coal samples all increases (Figure 5), indicating that CO2 improves the connectivity of pores and fractures in coal to a certain extent. Throat Lengths. A throat is the connecting channel between pores and is representative of the connectivity of pores and fractures (Song et al. 2018). Before CO2 treatment, throats <20 μm in length and >100 μm in length are the most common (Figure 6). Throats >100 μm in length are mainly composed of microfractures. 8 Figure 6—Throat lengths of coal samples based on X-ray CT. After CO2 treatment, the number of throats >100 μm in length obviously increases (Figure 6). The number of throats <100 μm in length in Coal Samples #2 and #4 changes weakly (Figure 6). The number of throats <20 μm in length in Coal Sample #1 increases obviously, while the number in Coal Sample #3 decreases (Figure 6). The number of 20-100 m long throats in Coal Sample #1 does not show obvious changes, while the number in Coal Sample #3 obviously increases (Figure 6). Changes in Mineral. Similarly to the distribution of the pores, the mineral grain size is mainly <5 μm, and as the grain size increases, the mineral number decreases rapidly (Figure 7). Minerals with a grain size <50 μm are normally distributed, and the peaks of mineral volume are distributed at 3-6 μm (Figure 8). With the increase and decrease in the grain size, the mineral volumes of coal samples decrease rapidly (Figure 8). The volumes of minerals with grain sizes ranging from 10-50 μm are small, while those with grain sizes >50 μm are large (Figure 8), indicating that there is a large amount of minerals with grain sizes >50 μm or a large amount of minerals filling the micro-fractures. After CO2 treatment, mineral numbers show a significant decrease trend, while the number of minerals with grain sizes <6 μm in Coal Samples #1, #3 and #4 increases significantly, indicating that minerals with grain sizes >6 μm are partially dissolved, resulting in a decrease in grain size (Figure 7). After CO2 treatment, the volume of minerals with grain sizes >50 μm largely decreases greatly (Figure 8). Moreover, the 3D digital models of the coal samples (Figure 5A, Figure 4A) show that the minerals that fill the micro-fractures are largely dissolved by CO2, which improves the connectivity of the micro- fractures and increases the pore volume in the coal samples. Therefore, the dissolution of minerals with grain sizes >50 μm and filling microfractures is the main contributor to the decrease in mineral volume in the coal samples and the increase in the volume of pores >50 μm. Some minerals with grain sizes <50 μm are not completely dissolved, resulting in a significant decrease in the volume of minerals with grain sizes of 6-50 μm and a slight increase in the volume of minerals with grain sizes <6 μm in some coal samples (Figure 8). 9 Figure 7—Mineral number distributions in the coal samples based on X-ray CT. Figure 8—Mineral volume distributions of coal samples based on X-ray CT. Effects of Mineral Dissolution on Pore Structure. The Relationship Between Pore Content and Mineral Dissolution. The mineral contents of the coal samples range from 3.19 % to 8.18 %, and these values decrease significantly (2.40-3.73 %) after CO2 treatment, with an average decrease of 35.69 % (Table 2). After CO2 treatment, there are significant positive correlations between increased pore content and decreased mineral content and between pore volume and mineral volume, with R2=0.8851 and 10 R2=0.9624, respectively (Figure 9). These relationships show that the change in the pore content of the coal samples is closely related to the mineral dissolution caused by CO2. Figure 9—Relationships between pores and minerals before and after CO2 treatment. The Effect of Mineral Dissolution on Pore Structure. The minerals with grain sizes >1 μm in the coal samples are mainly clay minerals and carbonate minerals. Clay minerals are mainly kaolinite (46.01 %) and muscovite (13.52 %), and carbonate minerals include calcite (15.98 %), ankerite (1.60%), and dolomite (0.15 %). In addition, there is a certain amount of gibbsite (2.48 %) and organic-clay complex (mainly organic-kaolinite complex) (5.28 %) (Figure 10, Table 4). Clay minerals are mainly distributed in the coal matrix, and minerals filling the microfractures and pores are mainly carbonate minerals and gibbsite, as well as some organic-clay complexes (Figure 10). Table 4—Area percentages of minerals in Coal Sample #2 before and after CO2 treatment. Minerals Kaolinite Calcite Muscovite Ankerite Gibbsite Dolomite Organic-clay complexes Other minerals Area percentage, % Before 46.01 15.98 13.52 1.60 2.48 0.15 5.28 14.98 After 69.17 0.20 6.14 0.62 3.50 0.20 4.54 15.63 Figure 10—The FESEM and EDS images of Coal Sample #2 before and after CO2 treatment. After CO2 treatment, carbonate minerals are largely dissolved and disappear (Figure 10 AA’, BB’). Among them, calcite has the highest degree of dissolution and disappears almost completely (0.20%) (Table 4, Figure 10 AA’), followed by microfractures filled with ankerite (0.62%) (Table 4, Figure 10 BB’). The dolomite in the coal matrix has a relatively low degree of dissolution (0.20 %) (Table 4, Figure 10 AA’), while the dolomite filling micro-fractures is largely dissolved (Figure 10 CC’). Therefore, the dissolution of carbonate minerals is the main reason for the decrease in the mineral content in the coal. Kaolinite reacts weakly with CO2. Therefore, kaolinite shows no significant change after CO2 treatment 11 (69.17 %) (Table 4, Figure 10 AA’). The gibbsite in the coal matrix is difficult to dissolve with CO2 (Figure 10 DD’), while the gibbsite filling the micro-fractures is partially dissolved (Figure 10 CC’). Muscovite mainly exists in the coal matrix and can be partially dissolved by CO2 (Figure 10 AA’). Organic-clay complexes mainly fill microfractures and are partially dissolved by CO2 (Figure 10 BB’). After CO2 treatment, calcite, dolomite, and other carbonate minerals in the coal matrix (including those that fill pores) dissolve or partially dissolve, forming a large number of pores created by dissolution (Figure 11a and 11c). The shape of these dissolution-created pores is irregular, and the pore diameter is generally <10 μm. Residual carbonate minerals can be found in these pores. These dissolution-created pores are the primary cause of the changes in the number and volume of pores <10 μm in diameter, and they control the changes in the pore number distribution in the coal samples (Figure 11a and 11c). Some of the pores created by dissolution become connected to each other (Figure 11a) and with microfractures (Figure 11c); however, most of them are associated with local connectivity in a small part of coal samples, which are characterized by overall poor connectivity. Therefore, these pores make a weak contribution to the connectivity of the coal samples. Note: a and c are coal samples from the Xinjing Mine; c is a coal sample from the Sihe Mine; d is a coal sample from the Yuwu Mine. Figure 11—Pores, microfractures, and minerals in coal samples before and after CO2 treatment. After CO2 treatment, the carbonate minerals, such as calcite and dolomite, with grain sizes >50 μm in the coal matrix (including those filling pores) are dissolved or partially dissolved, forming dissolution- created pores with diameters >50 μm (Figure 11b). Residual carbonate minerals can also be found in the pores created by dissolution (Figure 11b). These dissolution-created pores have a relatively large contribution to the number of pores >50 μm in diameter after CO2 treatment, while they have a relatively small contribution to the pore volume. Similar to the dissolution-created pores <10 μm in diameter, the connectivity of the dissolution-created pores with diameters >50 μm is weak. A large number of carbonate minerals filling the micro-fractures are dissolved by CO2, resulting in a partial or complete opening of the micro-fractures and significantly increasing the apertures of the micro-fractures (Figure 11c). Although the number of micro-fractures is less than the number of dissolution-created pores >50 μm in diameter caused by CO2 and has a much smaller contribution to the pore number, the volume of micro- fractures is much larger than that of dissolution-created pores >50 μm in diameter, and micro-fractures are the main contributor to the increase in the volume of pores >50 μm in diameter. This increase is also the reason why the pore volumes of the coal samples increase significantly after CO2 treatment, while the pore numbers change only slightly. Furthermore, according to Section 3.2.3, the connectivity of coal samples on the micrometer scale is mainly contributed by microfractures. Therefore, increasing the microfracture aperture significantly improves the connectivity of the coal samples and throat lengths. Moreover, some micro-fractures are filled with organic-clay complexes. After CO2 treatment, several organic-clay complexes that fill the microfractures are removed, which also increase the apertures and connectivity of the microfractures. In general, microfractures in coal samples are mainly filled with carbonate minerals (Figure 10), and the removal of organic-clay complexes from microfractures has a 12 relatively weak effect on microfractures. Furthermore, neither the clay minerals in the coal matrix nor the clay minerals that fill the microfractures exhibit obvious changes after CO2 treatment, and their influence on the structure of the pore fracture and the connectivity of the coal is weak (Figure 11d). Conclusions In this paper, taking low-volatile bituminous coals and anthracite coals collected from the Qinshui basin as examples, changes in pore fracture structure and connectivity on the micron scale after CO2 treatments were studied using X-ray micro-CT, FESEM and EDS. The following conclusions can be drawn from this study. 1. After CO2 treatment, the pore contents and volumes of low-volatile bituminous and anthracite coal increase significantly and the changes in pore numbers are small. The increase in pore volume is mainly caused by pores > 50 μm in diameter, while the change in pore number is related to pores <10 μm in diameter. The connectivity of low-volatile bituminous coal and anthracite coal on the micron-scale is mainly contributed by micro-fractures. 2. After CO2 treatment, the number of pores with coordination numbers >1 and the number of throats >100 μm in length increase significantly, confirming that CO2 improves the connectivity of pores and fractures; however, the improvement in pore-fracture connectivity is not enough to improve the connectivity of the coal on the micron-scale. 3. After CO2 treatment, the changes in pore numbers and volumes are mainly caused by the dissolution of carbonate minerals. Calcite, dolomite and other carbonate minerals in the coal matrix dissolve to form a large number of dissolution-created pores <10 μm in diameter. These dissolution-created pores are the main causes of the increase in the number and volume of pores <10 μm in diameter and determine the changes in pore numbers in coal samples. The carbonate minerals that fill the microfractures are greatly dissolved by CO2, which increases the apertures and connectivity of the microfractures. These microfractures are the main contributor to the increase in the volume of pores >50 μm in diameter and improve the connectivity of coal on the micrometer scale. Acknowledgments This study was supported by the National Key Research and Development Plan (No. 2018YFB0605601), the China National Natural Science Foundation (No. 41972168), and the Jiangsu Key Laboratory of Coal- Based Greenhouse Gas Control and Use (No. 2019A001). We would like to thank engineers from the Shanxi CBM Branch of Huabei Oilfield Company and Lu’an Group and a number of research students from China University of Mining and Technology for their assistance in the coal sampling and some experiments. Conflicts of Interest The author(s) declare that they have no conflicting interests. References Anggara, F., Sasaki, K., and Sugai, Y. 2013. Mineral Dissolution/Precipitation During CO2 Injection into Coal Reservoir: A Laboratory Study. Energy Procedia 37:6722-6729. Bertier, P., Swennen, R., Laenen, B., et al. 2006. Experimental Identification of CO2-water-rock Interactions Caused by Sequestration of CO2 in Westphalian and Buntsandstein Sandstones of the Campine Basin (NE- Belgium). Journal of Geochemical Exploration 89(1): 10-14. Dawson, G.K.W., Golding, S.D., Biddle, D., et al. 2015. Mobilization of Elements from Coal Due to Batch Reactor Experiments with CO2 and Water at 40 °C and 9.5 MPa. International Journal of Coal Geology 140: 63-70. Day, S., Duffy, G., Sakurovs, R., et al. 2008. Effect of Coal Properties on CO2 Sorption Capacity under Supercritical Conditions. International Journal of Greenhouse Gas Control 2(3): 342-352. 13 Du, Y., Sang, S., Wang, W., et al. 2018. Experimental Study of the Reactions of Supercritical CO2 and Minerals in High-Rank Coal under Formation Conditions. Energy & Fuels 32(2): 1115-1125. Fujioka, M., Yamaguchi, S., and Nako, M. 2010. CO2-ECBM Field Tests in the Ishikari Coal Basin of Japan. International Journal of Coal Geology 82(3): 287-298. Faiz, M.M., Saghafi, A., Barclay, S.A., et al. 2007. Evaluating Geological Sequestration of CO2 in Bituminous Coals: The Southern Sydney Basin, Australia as A Natural Analogue. International Journal of Greenhouse Gas Control 1(2): 223-235. Hayashi, J., Takeuchi, K., Kusakabe, K., et al.1991. Removal of Calcium From Low Rank Coals by Treatment with CO2 Dissolved in Water. Fuel 70(10): 1181-1186. Hol, S., Gensterblum, Y., and Massarotto, P. 2014. Sorption and Changes in Bulk Modulus of Coal -Experimental Evidence and Governing Mechanisms for CBM and ECBM Applications. International Journal of Coal Geology 128(12):119-133. Kutchko, B.G., Goodman, A.L., Rosenbaum, E., et al. 2013. Characterization of Coal Before and After Supercritical CO2 Exposure Via Feature Relocation Using Field-Emission Scanning Electron Microscopy. Fuel 107(2): 777-78. Kolak, J.J. and Burruss, R.C. 2014. The Use of Solvent Extractions and Solubility Theory to Discern Hydrocarbon Associations in Coal, with Application to The Coal–Supercritical CO2 System. Organic Geochemistry 73(2): 56-69. Liu, S.Q., Ma, J.S., Sang, S.X., et al. 2018. The Effects of Supercritical CO2 on Mesopore and Macropore Structure in Bituminous and Anthracite Coal. Fuel 223: 32-43. Liu, C.J., Wang, G.X., Sang, S.X., et al. 2015. Fractal Analysis in Pore Structure of Coal Under Conditions of CO2 Sequestration Process. Fuel 139: 125-132. Liu, C.J., Wang, G.X., Sang, S.X., et al. 2010. Changes in Pore Structure of Anthracite Coal Associated with CO2 Sequestration Process. Fuel 89: 2665-2672. Liu, S.Q., Sang, S.X., Ma, J.S., et al. 2019. Effects of Supercritical CO2 on Micropores in Bituminous and Anthracite Coal. Fuel 242: 96-108. Liu, S.Q., Sang, S.X., Liu, H. ., et al. 2015. Growth Characteristics and Genetic Types of Pores and Fractures in A High-Rank Coal Reservoir of The Southern Qinshui Basin. Ore Geology Reviews 64: 140-151. Liu, S.Q., Sang, S.X., Pan, Z.J., et al. 2016. Study of Characteristics and Formation Stages of Macroscopic Natural Fractures in Coal Seam #3 for CBM Development in The East Qinnan Block, Southern Qinshui Basin, China. Journal of Natural Gas Science and Engineering 34(2):1321-1332. Liu, S.Q., Sang, S.X., Wang, G.X., et al. 2017. FIB-SEM and X-ray CT Characterization of Interconnected Pores in High-Rank Coal Formed from Regional Metamorphism. Journal of Petroleum Science and Engineering 148(1): 21-32. Massarotto, P., Golding, S.D., Bae, J.S., et al. 2010. Changes in Reservoir Properties from Injection of Supercritical CO2 into Coal Seams-A Laboratory Study. International Journal of Coal Geology 82(34):269-279. Pan, Z., Ye, J., Zhou, F., et al. 2018. CO2 Storage in Coal to Enhance Coalbed Methane Recovery: A Review of Field Experiments in China. International Geology Review 60(5):754-776. Perera, M.S. A., Ranjith, P.G., Choi, S.K., et al. 2011. The Effects of Sub-Critical and Super-Critical Carbon Dioxide Adsorption-Induced Coal Matrix Swelling on the Permeability of Naturally Fractured Black Coal. Energy 36(11): 6442-6450. Ranjith, P.G. and Perera, M.S A. 2012. Effects of Cleat Performance on Strength Reduction of Coal in CO2 Sequestration. Energy 45(1):1069-1075. Song, W., Yao, J., Ma, J., et al. 2018. Pore-Scale Numerical Investigation into The Impacts of The Spatial and Pore-Size Distributions of Organic Matter on Shale Gas Flow and Their Implications on Multiscale Characterization. Fuel 216:707-721. Wang, A., Wei, Y., Yuan, Y., et al. 2017. Coalbed Methane Reservoirs’ Pore-Structure Characterization of Different Macrolithotypes in the Southern Junggar Basin of Northwest China. Marine and Petroleum Geology 86: 675-688. White, C.M., Smith, D.H., Jones, K.L., et al. 2005. Sequestration of Carbon Dioxide in Coal with Enhanced Coalbed Methane Recovery-A Review. Energy & Fuels 19(3): 659-724. Wong, S., Law, D., Deng, X.H., et al. 2007. Enhanced Coalbed Methane and CO2 Storage in Anthracitic Coals- Micro-Pilot Test at South Qinshui, Shanxi, China. International Journal of Coal Geology 1(2): 215-222. Xu, Y., Liu, X., Zhang, P., et al. 2016. Role of Chlorine in Ultrafine Particulate Matter Formation during the Combustion of a Blend of High-Cl Coal and Low-Cl Coal. Fuel 184: 185-191. Zerai, B., Saylor, B.Z., and Matisoff, G. 2006. Computer Simulation of CO2 Trapped Through Mineral Precipitation in the Rose Run Sandstone, Ohio. Applied Geochemistry 21(2): 223-240. 14 Zhou, H.W., Zhong, J.C., Ren, W.G., et al. 2018. Characterization of Pore-Fracture Networks and Their Evolution at Various Measurement Scales in Coal Samples Using X-Ray ΜCT and a Fractal Method. International Journal of Coal Geology 189:35-49. Shiqi Liu is a Professor of geological resources and geological engineering at China University of Mining and Technology, where he has worked as faculty for 9 years. His research interests are in the exploration and development of unconventional gas (coalbed methane, shale gas, coal measure gas), and CO2 geological storage and utilization. He holds B.S. from the China University Of Petroleum (East China) in information and computer science, M.S. from the China University of Petroleum (East China) in oil and gas field development engineering, and Ph.D. from China University of Mining and Technology in geological resources and geological engineering. Shuxun Sang is a Professor of geology at China University of Mining and Technology, where he has worked as faculty for 28 years. His research interests are in the exploration and development of unconventional gas (coalbed methane, shale gas, coal measure gas) and in geological storage and utilization of CO2. He holds a B.S., M.S., and Ph.D. from China University of Mining and Technology, all in geology. Tian Wang is a doctoral candidate at China University of Mining and Technology. Her research interests are in CO2 geological storage and utilization. She holds a B.S. from China University of Mining and Technology in geology. Yi Du is a doctoral candidate at China University of Mining and Technology. Her research interests are in CO2 geological storage and utilization. She holds a B.S. from China University of Mining and Technology in geology. Huihuang Fang is a doctoral candidate at China University of Mining and Technology. His research interests are in CO2 geological storage and utilization. He holds M. S. from China University of Mining and Technology in geological resources and geological engineering. Abstract Introduction Samples and Methodology Results and Discussion Conclusions Acknowledgments Conflicts of Interest References