Analysis of the Change Feature and Action Mechanism of Coal Pore Structure under the Action of Supercritical CO 2

In this study, supercritical CO 2 - (ScCO 2 ) induced permeability enhancement tests were performed on two di ﬀ erent-rank coal samples (i


Introduction
With the development of the clean coal technology, the collection, storage, and utilization of greenhouse gases have attracted widespread attention.In recent years, supercritical CO 2 , a new technology for anhydrous coalbed methane (CBM) extraction has been extensively researched and applied [1].CBM extraction in deep coal seams is susceptible to high temperature and high-ground stress [2].When tem-perature equals 31.19 °C and pressure equals 7.38 MPa, CO 2 becomes supercritical [3].As supercritical CO 2 (ScCO 2 ) features high permeability and low viscosity, it enters coal pores more easily than CO 2 .Once ScCO 2 enters the coal body, it will trigger a series of complex changes in the pore structure and improve coal seam permeability, thus enhancing CBM extraction [4].Compared to permeability enhancement by hydraulic fracturing [5], ScCO 2 -induced permeability enhancement, as an anhydrous CBM extraction technology, can save considerable water resources in practical engineering applications.However, if CO 2 sequestration and collection is not performed to coalbeds treated with this technology, a large amount of CO 2 will be emitted to the atmosphere, which is harmful to the environment.
In recent years, scholars all over the world have performed many experiments on ScCO 2 -activated coal bodies for the purpose of exploring changes in the pore structure of coal under the ScCO 2 condition.Their experimental results provide some theoretical guidance for the application of the ScCO 2 CBM extraction technology in practical engineering.In China, scholars carried out a CO 2 adsorption isotherm test on the coal body and found that the adsorption amount of the coal body reached its maximum value when the pressure of CO 2 was about 7.38 MPa, indicating that ScCO 2 was more easily adsorbed by the coal body and imposed the greatest influence on the structure of the coal body [6].Liu et al. [7] studied the influence of ScCO 2 on the pore structure and connectivity of different-rank coal samples by mercury intrusion porosimetry (MIP) and nitrogen adsorption test and concluded that ScCO 2 expanded the pore volume and specific surface area of different pores to different degrees; besides, the coal of a higher rank corresponds to more significant increases in the pore volume and the specific surface area and greater changes in the connectivity of macropores and mesopores.Scholars carried out an experimental study on CH 4 displacement in coal with ScCO 2 , and the results showed that injecting ScCO 2 into coal seams could improve the connectivity between coal pores, thus enhancing the permeability of coal seams and increasing the CH 4 extraction rate of coal seams [8][9][10].Yang et al. [11] investigated the seepage law of coal under the action of ScCO 2 and discovered that ScCO 2 triggered the appearance of honeycomb-shaped pores and improved the permeability of the coal body.The coal sample treated with ScCO 2 -H 2 O mainly experienced changes in micropores, while the low-rank coal sample mainly experienced changes in macropores and mesopores [12].In the experimental study on fracture initiation and extension of ScCO 2 -H 2 O fractured coal, it was observed that ScCO 2 -H 2 O led to the formation of numerous complex laminar bifurcation fractures in the coal body [13].In the study on the change features of the seepage pore structure of coal under the action of ScCO 2 , it was concluded that chemical dissolution-migration, water precipitation, and adsorption swelling of minerals by ScCO 2 were responsible for the changes in the seepage pore structure [14].The changes in functional groups and pore structure before and after the ScCO 2 action were analyzed by infrared spectroscopy and MIP, and the results revealed that the content of weak polar functional compounds decreased and the seepage pores changed [15].
In CO 2 adsorption experiments on coal, scholars found that oxygen-containing functional groups such as carboxyl and hydroxyl on the coal surface are the main adsorption sites for CO 2 molecules [16,17].In the experimental study on CBM displacement by injecting CO 2 (CO 2 -ECBM), it was concluded that the adsorption capacity of CO 2 in coal varies under different conditions, and the adsorption capacity of CO 2 is the strongest when CO 2 becomes supercritical; in addition, the ability of ScCO 2 to enhance coal seam permeability is independent of coal rank; whereas, the volumetric strain generated by ScCO 2 has an effect on coal seam permeability [18].Besides, CO 2 can extract polycyclic aromatic hydrocarbons (PAHs) from the coal matrix, induce matrix swelling, and present an inverted U-shaped change with coal rank.Compared with gaseous and liquid CO 2 , ScCO 2 is more capable of dissolving coal, can extract more hydrocarbons from the coal matrix, and boasts a better adsorption-swelling effect [19].
The above studies focus on changes in the pore structure induced by ScCO 2 displacement, adsorption, and chemical action on coal seams, as well as the resultant enhancement of coal seam permeability.This study is aimed at better clarifying the changes in the pore structure under the action of ScCO 2 .First, ScCO 2 -induced permeability enhancement treatment was performed on the coal samples using a selfdeveloped test device.Furthermore, the different-rank coal samples before and after the treatment were subjected to MIP and scanning electron microscopy (SEM) tests for comparatively analyzing their pore structure change features.Meanwhile, the corresponding mechanism was analyzed from physical and chemical perspectives.

Sample Preparation
Two types of coal with different metamorphic degrees, i.e., coking coal from Zhujiao Coal Mine in Anyang City and anthracite coal from Jiulishan Coal Mine in Jiaozuo City, Henan Province, China, were selected as coal samples.The samples were taken from underground working faces and transported to the laboratory in sealed packages.To minimize the influence of coal fractures on porosity, the samples were crushed and sieved into particles with a size of 3-6 mm [20].To ensure reliability of the experimental results, control tests were performed on raw coal samples for each group of experiments.The physical property parameters of coal samples before and after ScCO 2 treatment are listed in Table 1.

Test Device and Procedure
3.1.Test Device.Permeability enhancement tests were performed on coal samples using a self-developed ScCO 2induced permeability enhancement test.The test device consists of four parts, namely, a permeability processing module, an electrical control module, a temperature and pressure control module, and a data acquisition module.The schematic diagram of the test system is depicted in Figure 1.To determine the feature parameters related to the pore structure changes of coal samples before and after ScCO 2 treatment and obtain the microscopic views of the pore structure, MIP and SEM tests were conducted using a mercury porosimeter and a scanning electron microscope.
The test device mainly comprises a booster pump, a constant-temperature water tank, a buffer bottle, an air compressor, and a reactor.The functions or parameters of these instruments are described as follows: the booster pump provides sufficient critical pressure for CO 2 .The constanttemperature water tank is used for heating CO 2 at a constant The buffer bottle whose working pressure limit is 40 MPa serves to temporarily store pressurized and warmed ScCO 2 .
The air compressor (NCS50, the USA) is responsible for extracting air from the test setup.The instrument for the MIP test is a fully automatic mercury porosimeter (AutoPore IV 9500, Micromeritics, USA) with a pore size measurement range of 5-360,000 nm, a minimum working pressure of 0.1 Psi and a maximum working pressure of 60,000 Psi.The instrument for the SEM test is a tungsten filament scanning electron microscope (JSM-6390LV, Electronics Co., Ltd., Japan).

ScCO 2 Treatment of Coal Samples
Step 1.A coal sample was put into the reactor, and the air compressor was activated to clear the air and residues in the pipeline and device so that the test was carried out in vacuum.
Step 2. The CO 2 storage cylinder was opened, and the booster pump and constant-temperature water tank were turned on.To ensure that the CO 2 was in a supercritical state, the temperature was set to be constant at 32 °C, and the pressure was adjusted to 8 MPa and increased gradually by 2 MPa.As soon as the temperature and pressure had stabilized, ScCO 2 was put into the buffer bottle.
Step 3. ScCO 2 in the buffer bottle was injected into the reactor to allow the coal sample to be treated in the reactor for 2 d until the reaction reached equilibrium.
Step 4. The treated coal sample was removed from the reactor and cooled to room temperature naturally before the next test.

International Journal of Energy Research
In the test, CO 2 always maintained the supercritical stable state with the temperature being 32 °C and the pressure being 8 MPa, and the influences of temperature and pressure changes beyond the critical temperature and pressure of ScCO 2 were not considered.Such a setting ensured that the test results were not affected by temperature and pressure changes.

MIP and SEM.
MIP is based on the capillary flow of nonwetting liquid mercury in the pore.Different pressures are needed to inject mercury into different-sized pores in the solid, and the higher the applied external pressure, the smaller the radius of pores that mercury can enter.Hence, feature parameters related to the pore structure can be calculated based on the external pressure and the amount of injected mercury.In the MIP test, pores are assumed to be cylindrical, and the interfacial tension that resists mercury injection into pores is acting along the circumference of the pore wall.The interfacial tension is −2πrγ cos θ, and the external force overcoming the interfacial tension, which is πr 2 P, acts on the entire pore cross-section.According to Newton's third law, Equation ( 1) is obtained Based on Equation ( 1), the Washburn equation for the external pressure and pore size required for mercury intrusion into the pore is deduced as follows [21]: where r is the pore size (nm); γ is the surface tension of mercury (0.485 N/m); θ is the contact angle between mercury and the coal sample (130 °); P is the mercury injection pressure (10 MPa).
The relationship between the surface area of the pore and the external pressure required to fill the corresponding pore space with mercury is given below: then, where ΔV is the volume change of the injected mercury.Then, the entire pore volume is integrated to obtain the formula for calculating the pore specific surface area: The porosity is calculated as [22].
where V 1 is the volume of mercury at any pressure; V 2 is the volume of injected mercury in the steady state; V 3 is the volume of mercury at the maximum pressure.
The treated coal samples are fixed, dehydrated, and sprayed with heavy metal particles.Subsequently, the electron beam emitted by the SEM is used to bombard metal particles on their surfaces.In this way, the emitted secondary electron signals are converted into optical signals and displayed as images on the screen for further analysis on the morphological features of pore development at different scales.Figures 2 (a1, a2) and Figure 3 (a1, a2) shows the SEM images of the coal samples before and after treatment.To better analyze the pore structure distribution by MIP, the pore size of the coal body was classified into micropores (0.1-10 nm), small pores (10-10 2 nm), mesopores (10 2 -10 3 nm), macropores (10 3 -10 5 nm), and fractures (> 10 5 nm) according to the classification standard proposed by Hodot [23].

Test Results and Analysis
Based on the MIP test, the cumulative mercury intrusion curves of coking coal and anthracite before and after ScCO 2 treatment with pressure were plotted (Figure 4).As revealed by Figure 4, the cumulative mercury intrusion amounts of the two types of ScCO 2 -treated coal samples both show a three-stage increase, i.e., they surge first, then rise slowly and finally soar.Such a result can be explained with reference to the inverse relationship between pressure and pore size in Washburn's Equation (1) as follows: in stage 1, the slope of the curve is large; i.e., the cumulative mercury intrusion increases rapidly.The reason is that, at the initial low pressure, the mercury first enters some macropores and fractures where the resistance is slight, indicating remarkable connectivity between macropores and fractures.In stage 2, the slope of the curve is smaller; i.e., the cumulative mercury intrusion rises slowly.The reason is that the mercury enters into mesopores and some of the small pores with high resistance, which suggests that the connectivity between mesopores and small pores weakens and the pore channels between pores are blocked.In stage 3, as the pressure continues to rise, the cumulative mercury intrusion soars, demonstrating that the ScCO 2 action promotes the connectivity between small pores and micropores and increases the number of pore channels between small pores and micropores.
The mercury withdrawal and entry curves of the two types of coal samples before and after the treatment form a "hysteresis loop" (Figure 4), which indicates that the mercury entering the pores does not completely withdraw under the same pressure conditions.This phenomenon is related to the shape of the pores.The volume of the "hysteresis loop" after ScCO 2 treatment is smaller than that before the treatment, suggesting that ScCO 2 has transformed some of the dead or isolated pores into well-connected effective pores.That is, ScCO 2 plays a certain role in dredging and expanding the pores, thus increasing the number of effective pores, 4 International Journal of Energy Research improving the connectivity between pores, and facilitating the flow of CBM in the coal body.
4.1.2.Pore Volume Changes.As can be seen in Figure 4, the cumulative mercury intrusion curves of the two types of ScCO 2 -treated coal samples lie above those of the raw coal samples, and the two samples both experience mercury increment changes under the same pressure.This demonstrates that the ScCO 2 action expands the total pore volume of the coal body.Besides, the increment change is greater for anthracite and smaller for coking coal.As shown in Tables 2  and 3, the pore volumes of coking coal and anthracite increase by 0.0106 ml/g and 0.0142 ml/g, respectively, an increase of 52.51% and 79.32%, the latter being 47.72%  5 International Journal of Energy Research greater than the former, which reveals that ScCO 2 is more likely to change the pore size of higher-rank anthracite.In addition, the total porosities of coking coal and anthracite also increase by 1.80% and 3.44%, respectively, indicating that the ScCO 2 -induced increase in pore volume is attributed to the increase in the number of different-sized pores in the coal samples.According to Figure 5, after ScCO 2 treatment, for coking coal, the proportion of macropores and mesopores increases by 19.90% while that of small pores and micropores decreases by 19.90%; in contrast, for anthracite, the proportion of macropores rises by 52.63% while that of small pores, mesopores, and micropores falls by 60.44%.It can be concluded that the ScCO 2 -induced increase in total pore volume of coal samples mainly results from the increase in macropores and mesopores (pore size > 100 nm), and this increase is obviously related to the coal rank.Concretely, the higher the coal rank, the more significant the increase in pore volume.In the study on surface area variation of coal pore structure, scholars have investigated the roughness of the pore surface by combining the surface fractal dimension method and MIP and concluded that the more irregular the pore surface, the more inhomogeneous the pore structure and the larger the pore surface area [24].After ScCO 2 treatment, the specific surface area of macropores, mesopores, and small pores of coking coal increases by 5.71%, that of macropores and mesopores of anthracite increases by 4.41% and those of micropores of both types of coal samples decrease (Figure 5).This phenomenon signifies that ScCO 2 influences the pore surface structure in two ways: For one thing, ScCO 2 promotes the surface roughness of macropores and mesopores, resulting in an increase in pore specific surface area (Figure 5), which is related to the entry of ScCO 2 into the pores and the dissolution with minerals on the pore surface.For another, ScCO 2 reduces the surface roughness of micropores and the pore specific surface area (Figure 5) for the following reason: as the sizes of the micropores are small, the entered ScCO 2 reacts with minerals on the surface of micropores, making the surface uniform, and the reaction product blocks the channels between micropores, reducing the amount of mercury entry into micropores.As a result, the specific surface area calculated using Equation ( 5) is smaller.
Here is an analysis from the perspective of the coupling changes in different pore sizes and pore specific surface areas of coal samples before and after treatment.ScCO 2 treatment brings about an increase in the pore volume and pore specific surface area of macropores, mesopores, and small pores of coking coal; whereas for the anthracite, it causes an increase in the pore volumes and pore specific surface areas of macropores and barely alters those of small pores and micropores (Figure 5).This phenomenon further demonstrates that for low-rank coal, ScCO 2 action tends to change the pore shape and connectivity between macropores, meso-pores, and small pores; for high-rank coal, it tends to alter the channel and pore volume between macropores.and Figure 3(a2)).Since ScCO 2 is more permeable in the coal body than gaseous CO 2 , liquid CO 2 , and CH 4 , it is more likely to seep into fractures and matrix pores [25] and be adsorbed by pores, resulting in volumetric swelling and coal cracking.This phenomenon can be explained in the light of the potential theory adsorption isotherms D-R model proposed by Dubinin-Polanyi [26] and the coal surface-free energy principle [27] as follows.The D-R model considers that an adsorption potential energy field exists on the solid surface, through which gas molecules will be immediately adsorbed on the sold surface to form an adsorption layer; moreover, gas molecules are adsorbed in the pores in a filling manner and fail to form discrete single-molecule layers in the pores.The adsorption potential and the volume of the adsorbed phase in the D-R model are expressed as  7) and ( 8),

Analysis on the Action
where V and V 0 are the adsorption phase volume and maximum adsorption volume of coal body (ml/g); ε is the adsorption potential (J/mol); β is the relative affinity coefficient; P and P 0 are the adsorption equilibrium pressure and room pressure, respectively, (Pa); k is a parameter related to adsorption heat; R is the gas constant (J/mol K); T is the temperature (K).The amount of free energy change on the coal surface, i.e., π s , refers to the change in free energy of the system caused by raising the adsorbent content per unit area at constant temperature and pressure.When adsorbed gases such as CO 2 are present in coal pores, gas molecules are adsorbed on the surface of coal pores, so that the surface will reach an equilibrium state with the minimum energy.The value of π s when the adsorption equilibrium is reached under certain temperature and pressure conditions can be calculated based on Equation (10) [28] where π s is the amount of free energy change on the coal surface (J/m 2 ); σ is the amount of free energy change on the coal surface after gas adsorption (J/m 2 ); σ 0 is the amount of free energy on the solid surface under vacuum (J/m 2 ); E is the surface overload [29] (mol/m 2 ).

Macropores Mesopores Small pores
According to both the D-R model and the coal surfacefree energy principle, the amount of free energy change of two types of ScCO 2 -treated coal samples at the adsorption equilibrium of pores are calculated by simplifying V 0 and S 0 in Equation ( 12) into the volume and specific surface area of pores after ScCO 2 treatment, representing the critical pressure and temperature of ScCO 2 by P and T, and taking 0.1 MPa as the value of P 0 (Figure 6).
According to the surface-free energy principle, greater surface-free energy of pores corresponds to stronger adsorption for gas; more gas absorbed on the pore surface, a faster decrease in energy.Ultimately, the surface approaches an equilibrium state with the minimum energy.In other words, the amount of surface-free energy change at adsorption equilibrium is inversely proportional to the adsorption capacity of pores.When the two ScCO 2 -treated coal samples reach adsorption equilibrium, their variations in surface-free energy drop stepwise from macropores to micropores, and the surface-free energy is close to 0 in small pores and micropores (Figure 6).This indicates that for both types of coal samples, small pores and micropores contribute the most to ScCO 2 adsorption and boast the strongest adsorption capacity.The reduced surface-free energy in small pores and micropores of anthracite is lower than that of coking coal (Figure 6), demonstrating that the adsorption capacity of small pores and micropores (pore size < 100 nm) of anthracite is stronger than that of coking coal.As a result, they are capable of absorbing more ScCO 2 , producing a larger volume expansion effect, and thus causing the phenomenon that the opening of the newly generated fracture network in anthracite is larger than that in coking coal (Figures 2(a2) and Figure 3(a2)).Before the treatment, small pores and micropores account for 71.6% for coking coal and 81.56% for anthracite (Figures 5(a) and 5(b)), which further verifies that the stronger adsorption capacity of fine pores of anthracite than coking coal is attributed to the more developed fine pores of anthracite.Therefore, the coal of a higher rank has more developed fine pores with greater surface-free energy, stronger adsorption capacity, and a more remarkable adsorption expansion effect; accordingly, the coal body gets cracked more easily.This finding is consistent with the findings by Cai et al. and Wei et al. that the gas adsorption capacity of coal pores strengthen as the coal rank rises and the coal containing more developed micropores has stronger adsorption capacity [24,30].Due to the error of pore volume measurement by MIP, the micropore volume of coking coal before treatment in Table 2 is a little higher than that of anthracite in Table 3, and the actual situation is the opposite.This error can be explained as follows: compared with anthracite, coking coal is softer and thus its pore structure is more likely to be changed by mercury pressure, which results in an increase in its micropores before treatment.
In summary, the development degree of fine pores in the coal body determines the adsorption capacity of coal pores.After absorbing ScCO 2 , small pores and micropores in the coal body get swelled and converted into macropores and mesopores due to their strong adsorption capacity, resulting in an increase in the number of macropores and mesopores as well as a decrease in the number of small pores and micropores.Besides, after the adsorption swelling among pores, the 9 International Journal of Energy Research fracture network develops so that the connectivity between pores improves.Figure 7 shows a schematic diagram of the adsorption-swelling effect of the coal matrix under the action of ScCO 2 .
The swelling effect of ScCO 2 adsorption on coal pores is jointly affected by thermal expansion caused by critical temperature and matrix deformation caused by critical pressure.According to Sun and Li [31], the volumetric strain of the coal body varies in the form of an "Sshaped" logistic function during ScCO 2 treatment; i.e., as the critical temperature and pressure increase, the coal body experiences an increase in its volumetric strain which allows it to adsorb more ScCO 2 , but the adsorption does not grow infinitely and will finally converge to a stable state.Therefore, CO 2 always maintains the supercritical stable state with the temperature being 32 °C and the pressure being 8 MPa in this study, which ensures that the pore adsorption expansion keeps a steady state during ScCO 2 treatment.Otherwise, the adsorption expansion will become excessively large and cause coal pores to squeeze each other, which is inconducive for permeability enhancement.Furthermore, in actual engineering, when ScCO 2 is injected into the coal body, ScCO 2 and CBM will compete for adsorption, displacing CBM at the original position in the coal matrix and on pore surface.And some of the displaced CBM transforms from a free state to a gaseous state and experiences volumetric swelling, causing matrix and pore deformation that will further produces new fractures.
It is noteworthy that the effect of competition between ScCO 2 and CBM for adsorption on the pore structure is not considered for ScCO 2 -treated coal samples here.

Analysis on Chemical Changes.
Coal is composed of a large amount of organic matter and a small amount of inorganic matter.Organic matter is a complex natural organic polymer compounds, while inorganic matter is mainly salt minerals.Once exposed to the ScCO 2 fluid, inorganic matter will have a dissolution reaction with ScCO 2 [32].Resultantly, some minerals in coal, such as calcite and dolomite, will dissolve and precipitate out of coal, further changing the pore structure of coal.Unevenly scattered micron-sized flocculent particles with different degrees of surface erosion exist on the surface of the coal samples after the test treatment, with the gray-white part being mineral particles and the gray-black part being coal matrix (Figures 2(a2) and Figure 3(a2)).On top of this, it is also found that the surfaces of the two raw coal samples are relatively flat, without obvious pores, fractures, or minerals.However, after the ScCO 2 treatment, the mineral particles on the surface of coking coal are distributed along the fractures, large in particle size and small in number, while those on the surface of anthracite are densely distributed, small in particle size.In addition, the coal matrix area is wrapped with more mineral particles.The above phenomena primarily result from the following chemical changes.Besides, as ScCO 2 acts on the coal body for a longer period of time, more mineral components in the coal matrix (1) Dissolution-Migration of Minerals.The mechanisms of dissolution of different minerals in coal by ScCO 2 differ, and the interaction of minerals in the coal body with carbonic acid generated by CO 2 affects the pore structure of the coal seam [33].CO 2 molecules react with water to form carbonic acid that can further react with strong acid minerals such as carbonate, silicate, and sulfate in a soluble reaction.Saline minerals such as calcite and dolomite in coal generate less stable carbonate under the reaction with carbonic acid.Furthermore, carbonate will migrate in the pore channels between the coal matrix (Figure 8) and adsorb outside the pores and around the coal matrix, thus promoting the pore connectivity and pore volume.
(2) Dissolution-Precipitation of Minerals.Under the condition of carbonic acid generated by ScCO 2 , saline substances such as potassium feldspar and sodium feldspar in the coal matrix are corroded to produce more stable acids such as kaolinite, montmorillonite, illite, and pyrite crystal.These reaction products are easily deposited to block the pores and channels when migrating in the pore space owing to their strong stability (Figure 8), thus lowering the connectivity between the pores.The changes in the chemical reactions of carbonic acids generated by ScCO 2 with major saline minerals in coal are illustrated in Table 4, and the schematic diagram of mineral dissolution under the action of ScCO 2 is depicted in Figure 8.
The formation of carbonic acid from ScCO 2 and water is a reversible chemical reaction.When the temperature is too high (above the critical temperature), it accelerates the reverse reaction, increasing the amount of carbonic acid decomposed into CO 2 and reducing the carbonate ions in the coal pores.Consequently, saline minerals in the coal body can merely react with fewer carbonate ions, which drastically weakens the ability of ScCO 2 to dissolve minerals.In short, the dissolution-migration and dissolutionprecipitation stability of ScCO 2 occurring in the coal body varies with the change in the critical temperature, and an excessively high temperature will weaken the dissolution reaction, which is not favorable for ScCO 2 to unblock the pore channels.Keeping the critical temperature and pressure constant during the test can ensure the steady state of the S C CO 2 dissolution reaction.

Conclusions and Prospects
By discussing the pore structure change features and the corresponding action mechanisms of two different-rank coal  (1) Regarding the changes in pore connectivity and pore volume, the injection of ScCO 2 brings the following changes: the connectivity between macropores and fractures increases; the connectivity between macropores and mesopores is reduced as the pore channels between them are blocked; the connectivity between small pores and micropores increases as the number of pore channels grow.The total pore volumes of both coking coal and anthracite increase by 0.0106 ml/g and 0.0142 ml/g, an increment of 52.51% and 79.32%, respectively.The ScCO 2induced increase in total pore volume of coal samples mainly results from the increase in macropores and mesopores (2) As for the change of pore shape, under the action of ScCO 2 , some ineffective pores such as isolated pores and dead pores are transformed into effective pores that are connected at least at two ends, which helps to dredge and enlarge pores, improve the pore connectivity, and change the original shape distribution of pores.ScCO 2 primarily induces an increase in the surface roughness of macropores and mesopores and a decrease in the roughness of small pores and micropores, thus triggering changes in the pore surface area (3) ScCO 2 -induced changes in the coal pore structure are jointly caused by pore adsorption swelling, dissolution-migration, and dissolution-precipitation in the coal body.To be specific, the surface-free energy of fine pores (pore size < 100 nm) is the internal cause of pore adsorption-swelling effect.The coal of a higher rank contains more fine pores with greater surface-free energy and stronger adsorption swelling.Resultantly, it experiences greater changes in its pore structure.Carbonic acid generated by ScCO 2 and strong acid minerals in the coal matrix and pore space undergo dissolution-migration, which stands to dredge and expand the pore space and meanwhile promote the pore connectivity and volume.Dissolution-precipitation leads to the blockage of pore space and pore channels, hence reducing the pore connectivity and pore volume In this study, the ambient temperature and pressure of the coal samples were always room temperature and atmospheric pressure, for the purpose of ensuring that ScCO 2 always maintained the critical temperature and pressure.Underground deep coal seams are affected by a combination of ground stress and ground temperature.If ScCO 2 is adopted in the actual project to increase the penetration of coal seams for gas extraction, it will be subject to the coupling effect of ground stress and geothermal heat.As a result, its temperature and pressure are unstable, which affects its effect on the coal body.Therefore, in the future simulation test design, it is necessary to enhance the set-ting of environmental conditions of coal samples and to attach greater importance to the coupling effect during ScCO 2 treatment.

2
International Journal of Energy Research temperature in the range from room temperature to 200 °C.

Figure 1 :
Figure 1: Schematic diagram of the test system.

Figure 4 :
Figure 4: Cumulative mercury intrusion of coking coal and anthracite with pressure before and after ScCO 2 treatment.
Mechanism of the Pore Structure 4.2.1.Analysis on Physical Changes.The SEM images of coal samples before and after treatment are presented in Figures 2(a1, a2) and Figure 3(a1, a2).The samples were sprayed with metal without being polished.As demonstrated in Figures 2(a2 ) and Figure3(a2), both coking coal and anthracite produce a new fracture network after ScCO 2 treatment.The fracture network of coking coal is larger in number and smaller in opening, while that of anthracite is smaller in number and larger in opening (Figures 2(a2 )

Figure 5 :
Figure 5: Changes in pore volume and pore specific surface area percentage of coal samples before and after ScCO 2 treatment.

Figure 6 :
Figure 6: Amounts of free energy change of two types of ScCO 2 -treated coal samples at the adsorption equilibrium of pores.

Figure 7 :
Figure 7: Schematic diagram of the adsorption-swelling effect of coal matrix under the action of ScCO 2 .

Figure 8 :
Figure 8: Schematic diagram of mineral dissolution under the action of ScCO 2 .

Table 1 :
Basic parameters of coal samples before and after ScCO 2 treatment.

Table 2 :
Pore parameters of coking coal before and after ScCO 2 treatment in the MIP test.

Table 3 :
Pore parameters of anthracite before and after ScCO 2 treatment in the MIP test.
Note: V g : pore volume; S b : pore specific surface area; η: total porosity.