The coal sample used in the present study was from the Sijiazhuang Coalmine, Shanxi, China. Table 1 shows proximate and ultimate analyses of raw coal. Raw coal was dried, crushed and sieved to particle sizes of 250 μm–420 μm and 420 μm–840 μm, respectively. The as-prepared coal particles were dried in an oven overnight at 383 K and stored in sealed plastic containers. The resulting samples were named SJZ-x-y, where x refers to the mass proportion of the coal sample with 420 μm–840 μm and y represents the mass ratio for the 250 μm–420 μm coal sample. Therefore, the samples were denoted as SJZ-0-100, SJZ-20-80, SJZ-40-60, SJZ-60-40, SJZ-80-20, and SJZ-100-0, respectively. The samples of various particle sizes were considered in evaluating the effects of particle size on the coal pore structure and CH₄ and CO₂ adsorption/desorption characteristics.

Table 1.

Table 1.

Table 1. Chemical analyses of raw coal (air-dried basis, wt.%). . Coal Proximate analysis Ultimate analysis Fixed carbon Ash Volatile matter Moisture C H N S O* Value 61.95 29.31 8.74 0.49 66.54 3.22 <0.5 2.13 >27.61 * : by difference.

Table 1. Chemical analyses of raw coal (air-dried basis, wt.%). .

The proximate analysis of raw coal was investigated by a GF-A6 automatic proximate analyzer (Hebi Celestica Instruments Co., Ltd).^[22,30]

The ultimate analysis of raw coal was performed in a CARLO ERBA 1106 element analyzer (Italy).^[30]

X-ray diffraction pattern of raw coal was recorded in DX-1000 powder diffractometer equipped with Cu Kα x-ray source and an internal standard of silicon power.^[24,32]

FT-IR analysis of raw coal was performed on a Nicolet DXC20 FT-IR spectrometer.^[22]

Thermogravimetric analysis (TG) of raw coal was carried out on a thermo gravimetric analyzer (TGA Q500). The samples were heated at a rate of 10 °C/min under N₂ atmosphere.^[32]

Surface morphology of raw coal was investigated by scanning electron microscopy (SEM) (JEOL/EO JSM-5900 microscope).^[1]

The textural properties of the samples with varying particle sizes were obtained by N₂ adsorption/desorption isotherms, determined at 77 K with a NOVA1000e surface area and pore size analyzer (Quantachrome Company).^[1,34,35]

Gas adsorption and desorption measurements were performed by a volumetric method similar to that previously described.^{[22,24,30,32]} Approximately 5-g coal was dried at 110 °C and evacuated for 3 h, prior to the determination of void volume in the adsorption setup by helium calibration. Subsequent to helium calibration, the samples were again evacuated for 3 h. CH₄ or CO₂ was introduced into the adsorption setup to produce an adsorption isotherm under pressures up to 5.0 MPa at 308 K. The test was repeated three times and the mean value of the equilibrium data was used to calculate the adsorption isotherm to ensure the validity of the experiment.

Dynamic gas adsorption test was carried out with about 20.0-g coal at 298 K under 0.5 MPa and 0.1 MPa, respectively. Prior to measurement, the coal sample was dried at 110 °C overnight. The content of He/CO₂/CH₄ used in the experiment was 55.08:6.07:38.85 (volume ratio, %).

The structural properties of raw coal are characterized by XRD, FT-IR, TG and SEM techniques, respectively. Figure 1(a) shows the XRD pattern of raw coal. There exists a broad peak at around 25°, attributed to the (002) characteristic plane of graphite. FT-IR profile for raw coal is shown in Fig. 1(b). An absorption band appears at around 3400 cm⁻¹, assigned to −OH stretching. The adsorption peaks observed at 2900 cm⁻¹–3000 cm⁻¹ and 1670 cm⁻¹–1760 cm⁻¹ are attributed to −COOH group vibrations. The peak appearing at 1000 cm⁻¹–1200 cm⁻¹ can be attributed to esters, such as those in ethers, phenols and hydroxyl groups. Shoulder bonds at low wavenumbers (< 1000 cm⁻¹) may be related to out-of-plane bending modes. Figure 1(c) displays the normalized weight change of raw coal due to the heat treatment from 100 °C to 750 °C under N₂ atmosphere. Dehydration mainly happens below 350 °C, together with small evolutions of other gases. In a temperature range of 400 °C–750 °C, extensive evolution of hydrocarbons that correspond to volatile organic components takes place. As confirmed in Fig. 1(c), the weight loss for raw coal SJZ due to heat treatment is 8.43%, which is consistent with the proximate analysis results. Figure 2 shows the SEM images of raw coal, which exhibits a dense structure with an irregular, heterogeneous and porous surface, illustrating the development of micropores.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. XRD profile (a), FT-IR (b), and weight loss (c) of raw coal SJZ.

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. SEM images of raw coal SJZ.

As a key parameter, the pore structure, which includes the specific surface area, pore volume and pore size distribution, directly affects the CBM production. Based on the knowledge of the nature of coal and its pore system, N₂ adsorption/desorption, could characterize the pore structure.^[22]

Figure 3 displays N₂ adsorption/desorption isotherms at 77 K as a function of coal particle size. According to the BET classification, the adsorption isotherms of this kind belong to type III describing the physical adsorption process of N₂. The isotherms exhibit the remarkable hysteresis loops at higher relative pressures (P/P₀ > 0.45). It is obvious that the nitrogen adsorption volume of sample SJZ-0-100 is the highest among the six samples, while the adsorption volume of SJZ-100-0 is the lowest.

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. N2 adsorption/desorption isotherms at 77 K of the coal samples.

Table 2 presents the corresponding textural characteristics related to the specific surface area (SSA), pore volume (PV), and average pore diameter (PD). The variation of coal particle size leads to differences in SSA, PV, and PD. Among the samples, SJZ-0-100 with the smallest particle size displays the highest SSA and PV, indicating that the porosity of SJZ-0-100 is developed in the grinding process, which possibly opens some closed pores, leading to an increased pore accessibility. Compared with the others, the sample SJZ-100-0 has a low SSA of 1.988 m²/g and PV of 5.325 × 10⁻³ cm³/g.

Table 2.

Table 2.

Table 2. Textural characteristics of the coal samples. . Sample Specific surface areaa/(m2/g) Pore volumeb (10−3 cm3/g) PDc/nm SJZ-0-100 5.267 10.380 3.794 SJZ-20-80 3.329 8.950 3.969 SJZ-40-60 3.110 8.354 3.969 SJZ-60-40 2.881 7.718 3.969 SJZ-80-20 2.532 6.902 3.969 SJZ-100-0 1.988 5.325 4.152 a Measured from NLDFT method. b Evaluated from NLDFT method. c Average pore diameter calculated from NLDFT method.

Table 2. Textural characteristics of the coal samples. .

Figure 4 shows the textural characteristics as a function of coal particle size. It can be seen that the sample SJZ-100-0 with the largest coal particle size has a total SSA of 1.988 m²/g, and after being ground into smaller particles, the total SSAs increase up to 2.532 m²/g for SJZ-80-20, 2.881 m²/g for SJZ-60-40, 3.110 m²/g for SJZ-40-60, 3.329 m²/g for SJZ-20-80, and 5.267 m²/g for SJZ-0-100, respectively. At the same time, with the particle size decreasing, the PV increases, while PD decreases. This result shows that the coal sample with reduced particle size has an increased pore accessibility and consequently an increase in the nitrogen amount adsorbed, thus leading to the increases of SSA and PV.

Fig. 4.

	Figure Option ViewDownloadNew Window
	Fig. 4. Relationships between the textural characteristics and particle size distribution of the samples.

Figure 5 displays the pore size distributions of the samples obtained by applying the DFT method, from which the difference in micropores size distributions among the samples can be clearly seen. Apparently, the decrease of coal particle size can increase the volumes of micropores.

Fig. 5.

	Figure Option ViewDownloadNew Window
	Fig. 5. Pore size distributions obtained by applying the DFT equation for the coals.

Figure 6(a) shows the measured data of CH₄ adsorption on the samples at 308 K under pressures up to 5.0 MPa. The CH₄ adsorption isotherms belong to type I of IUPAC. Differences in coal particle size and SSA can result in variation of CH₄ adsorption capacity. The adsorption capacity over coal increases with particle size reducing. Among the samples, the sample SJZ-0-100 possesses the highest methane adsorption capacity while the SJZ-100-0 sample has the lowest methane adsorption capacity.

Fig. 6.

	Figure Option ViewDownloadNew Window
	Fig. 6. CH4 (a) and CO2 (b) adsorption isotherms of the coal samples, obtained by applying the Langmuir model (solid lines) at 308 K.

Figure 6(b) exhibits CO₂ adsorption abilities of the samples evaluated by a volumetric method at 308 K. The CO₂ adsorption isotherms display a zoom at the low pressure region. As expected, the smaller particle size coal sample has a much higher CO₂ uptake than the larger particle size sample due to its developed pore structure. Since the structural properties are significantly affected by coal particle size, it is possible to enhance CO₂ uptake in a certain range.

Because of its simplicity and providing a reasonable fit to most experimental data, the Langmuir model is widely used in the CBM industry and relevant reservoir simulations. Hence, in the present work the Langmuir model is used to fit the adsorption data, which fits experimental results quite well (R² > 0.99). The adsorption parameters of the Langmuir model are listed in Table 3.

Table 3.

Table 3.

Table 3. Adsorption parameters obtained from Langmuir model of the coal samples. . Sample CH4 CO2 VL/(cm3/g) b/MPa−1 R2 VL/(cm3/g) b/MPa−1 R2 SJZ-0-100 27.686 0.585 0.99750 63.818 0.508 0.99506 SJZ-20-80 25.648 0.698 0.99981 48.810 0.618 0.99886 SJZ-40-60 24.752 0.628 0.99974 43.926 0.553 0.99978 SJZ-60-40 23.856 0.581 0.99956 42.896 0.518 0.99937 SJZ-80-20 22.781 0.535 0.99967 37.565 0.509 0.99846 SJZ-100-0 22.422 0.445 0.99678 35.638 0.448 0.99792

Table 3. Adsorption parameters obtained from Langmuir model of the coal samples. .

The parameter V_L in the model is the maximum uptake of the adsorbent, and b is the reciprocal of the pressure when the adsorption capacity reaches 50% of the maximum uptake. The values of maximum adsorption capacity (V_L) of the coal samples vary from 22.422 cm³/g to 27.686 cm³/g for CH₄ and from 35.638 cm³/g to 63.818 cm³/g for CO₂, respectively. It reveals that the particle size has a significant effect on the gas adsorption capacity for the coal samples, i.e., the gas adsorption capacity increases as the coal particle size decreases. From Fig. 7, it can be clearly seen that the maximum adsorption capacities of CH₄ and CO₂ both decrease with the coal particle size increasing, but the declining trend of CO₂ maximum adsorption amount is more obvious than that of CH₄.

Fig. 7.

	Figure Option ViewDownloadNew Window
	Fig. 7. Relationships between maximum adsorption capacity of CH4 (a) and CO2 (b) and particle size distributions of the samples.

Figure 8 shows the relationships between Langmuir constant b and particle size of the coals. As shown in Fig. 8, the particle size has the different effects on Langmuir constant b for the coal samples. For the coal samples ranging from SJZ-0-100 to SJZ-20-80, the Langmuir constant b increases, however, for the coal with larger particle size the parameter b decreases, which means a weaker adsorbent/adsorbate interaction.

Fig. 8.

	Figure Option ViewDownloadNew Window
	Fig. 8. Relationships between Langmuir constant b and particle size of the coals.

Figure 9 shows CH₄ adsorption/desorption isotherms of the coals. It can be seen that the adsorption/desorption isotherms of the tested coals have apparent hystereses for all coal particle sizes. The phenomenon that desorption isotherms generally lie above the adsorption isotherms represents that the adsorption hysteresis is associated with the adsorption/desorption process. Coal, as the adsorbent, when adsorbing and desorbing the adsorbate, produces the hysteresis, which not only happens to gas, but also to water. The hysteresis effect indicates that the adsorbent/adsorbate system is in a metastable state and at reducing pressure the gas is not readily released to the extent corresponding to the thermodynamic equilibrium value.

Fig. 9.

	Figure Option ViewDownloadNew Window
	Fig. 9. CH4 adsorption/desorption isotherms of the coals.

Figure 10 demonstrates CO₂ adsorption/desorption isotherms of the coals, in which the adsorption/desorption hysteresis is similar to that of CH₄. The existence of the hysteresis could be explained by the nature of CO₂ adsorption over the coal. As shown in Fig. 10, larger particle size coal gives rise to a larger hysteresis. Coal with larger particle size has a better chance to trap the gas molecules as it possesses longer path distances for gas to be desorbed from the internal surface. CO₂ molecules may be absorbed/dissolved into the coal structure^[38] and when the desorption process happens, only the adsorbed molecules come out of the pore-spaces leaving behind the dissolved molecules in the coal structure.

Fig. 10.

	Figure Option ViewDownloadNew Window
	Fig. 10. CO2 adsorption/desorption isotherms of the coals.

The ratio of the maximum adsorption capacities V_L(CO2)/V_L(CH4) derived from adsorption and desorption isotherms is consistently greater than 1, reflecting the preferential adsorption of CO₂ compared with CH₄ and implicating the feasibility of CO₂ enhanced methane recovery (CO₂-ECBM). From Fig. 11, the ratio of V_L(CO2)/V_L(CH4) diminishes with coal particle size increasing, and thus the degree of preferential adsorption for CO₂ decreases. With the increase of coal particle size, the selectivity for CO₂ adsorption decreases, which is in line with the trend of V_L(CO2). In the adsorption process, V_L(CO2)/V_L(CH4) is in a range of 1.588–2.306, while V_L(CO2)/V_L(CH4) varies between 1.468 and 2.236 in the desorption process, i.e., the trends for V_L(CO2)/V_L(CH4) along with coal particle size in both adsorption and desorption processes are similar.

Fig. 11.

	Figure Option ViewDownloadNew Window
	Fig. 11. Ratios of the maximum uptake for CO2/CH4.

Figure 12 displays CH₄ and CO₂ breakthrough curves of the coal sample SJZ-0-100 at 298 K under pressures of 0.1 MPa and 0.5 MPa, respectively. As seen from Fig. 12, CH₄ has a shorter breakthrough time than CO₂, and with pressure increasing from 0.1 MPa to 0.5 MPa, the adsorption equilibrium time becomes longer. When the pressure is 0.1 MPa, the breakthrough times of CH₄ and CO₂ are 11 s and 16 s respectively, but as the pressure increases up to 0.5 MPa, the breakthrough times change into 42 s and 108 s respectively, indicating that the difference in breakthrough time between CH₄ and CO₂ is more obvious, for the increasing of pressure will increase the density of gas molecules, thereby improving the interaction between the adsorbate and the coal, which in turn enhances the gas adsorption capacity.

Fig. 12.

	Figure Option ViewDownloadNew Window
	Fig. 12. Breakthrough curves of He, CO2, and CH4 on the coal sample SJZ-0-100 at 298 K under different pressures.

Figure 13 shows breakthrough curves of CO₂ and CH₄ on the typical samples SJZ-0-100 and SJZ-100-0 at 0.5 MPa. The smaller particle size sample SJZ-0-100 has a higher breakthrough time, i.e., the adsorption amount of the SJZ-0-100 sample is higher. The breakthrough times of CH₄ and CO₂ for the sample SJZ-0-100 are 42 s and 108 s, and their difference is 66 s. As particle size increases, the breakthrough times of the sample SJZ-100-0 are shortened, i.e., 39 s for CH₄ and 77 s for CO₂, and the difference in breakthrough time between the CO₂ and CH₄ becomes 38 s. The shorter breakthrough time indicates that the gas uptake of the adsorbent is smaller. Hence, according to the decrease of the breakthrough time difference, it is demonstrated that the increase of particle size makes the ratio of V_CO2/V_CH4 reduced, which is consistent with the results indicated in Figs. 11 and 14.

Fig. 13.

	Figure Option ViewDownloadNew Window
	Fig. 13. Breakthrough curves of He, CO2, and CH4 on the typical samples SJZ-0-100 and SJZ-100-0 at 0.5 MPa.

Fig. 14.

	Figure Option ViewDownloadNew Window
	Fig. 14. Separation factors of CO2/CH4 versus coal sample under pressures of 0.1 MPa and 0.5 MPa.