Parameter prediction of factors influencing methane adsorption for coal macromolecules under different water molecular morphologies

2.1. Model construction and internal connection

In this manuscript, the coal samples from the 2^# coal seam of the Chiyu coal mine were selected as the object. The following procedure was employed initially: grinding→drying→acid washing→agitation→filtering→inspection→re-drying, etc. The contained structures (carbon spectrum, functional groups and heteroatoms, etc.) were determined through relevant experiments [23], the molecular formula and molecular weight constructed by reusing Chemdraw Software were C₂₈₆H₂₀₆O₅N₄S₄ and 3902, respectively, wherein the calculated C, H, O, N, and S with elemental percentages of 87.95%, 5.28%, 2.05%, 1.44%, and 3.28%. The proportions of H/C, O/C, N/C, and S/C were 0.72, 0.017, 0.014, and 0.014, respectively. Given that coal macromolecules exhibit an”approximate” polymeric state structure, the model was constructed and optimized in Materials Studio (MS) through a series of steps, including energy minimization and multiple annealing cycles, to obtain a structure with the lowest possible non-periodic energy (Figure 1a). Subsequently, 12 coal macromolecules were incorporated to form a periodic cell. By applying successive NPT and NVT ensembles combined with several annealing optimizations, the system converged to a coal molecular density of 1.2 g∙cm^-3, which falls within the actual density range of coal [24]. The final cell dimensions were a=b=c=40.18 Å. At this stage, the relative atomic coordinates, bond angles, molecular conformations, and pore distributions of the coal macromolecular chains were completely stabilized, thereby establishing a “benchmark invariant structure” for subsequent water molecular insertion.

Close

Figure 1.

(a) optimization and annealing of coal macromolecular model. (b) CY-CW-CMM. (c) CY-RW-CMM. Construction process of coal macromolecular models under different water molecular morphologies.

Export to PPT

Under the premise of maintaining an identical unit cell size, two models were established: the Chiyu cluster water–coal macromolecular model (CY-CW-CMM) and the Chiyu random water–coal macromolecular model (CY-RW-CMM). In the CY-CW-CMM, water molecules form stable hydrogen-bonded clusters that anchor near the polar functional groups (hydroxyl and carboxyl groups) of coal macromolecules. This model reflects the multilayer aggregation adsorption mechanism of moisture in coal: water molecules are first adsorbed at the polar sites of coal via electrostatic interactions, and subsequently aggregate into ordered clusters through intermolecular hydrogen bonding, without altering the structural integrity of coal macromolecules. Even after energy minimization, the water molecules remain in a stable “clustered state” (Figure 1b). According to Eq. (1), prior to methane adsorption, increasing water content leads to a simultaneous increase in both the maximum negative van der Waals energy and the maximum negative electrostatic potential energy of the system, with the latter exceeding the former (Table 1). This indicates that intense intermolecular collisions occur among water molecules, accompanied by the aggregation of hydrogen and hydroxide ions. Due to spatial constraints, water molecules form multimolecular aggregates, a fraction of which react with certain coal macromolecular sites. These aggregates and reaction products occupy the pore space within coal macromolecules, thereby reducing the available adsorption volume to 0.16-0.18 cm³∙g^-1, as determined by Connolly N₂ and He probe analyses.

In contrast, in CY-RW-CMM, water molecules are randomly distributed, without a fixed aggregated structure. They freely diffuse to fill the free space in coal pores and enable coal-water molecules to form an organic whole and optimize. After optimization, they are in a dispersed state and fully integrate into the pore structure (Figure 1c). With the increase in water molecule content, the maximum negative van der Waals energy and maximum negative electrostatic potential energy also increase. However, the maximum electrostatic potential energy of CY-RW-CMM is lower than that of CY-CW-CMM, while the maximum negative van der Waals energy shows the opposite trend (Table 1). This suggests that ions decomposed from water molecules may undergo vigorous reactions with certain functional groups in coal macromolecules, occupying high-energy adsorption sites and the entire pore structure, which leads to an even smaller adsorption space (with a value of 0.15 cm³∙g^-1-0.17 cm³∙g^-1 as calculated using the Connelly N₂ probe and He probe). Therefore, when considering the fugacity state of water molecules, a single water molecule may be anchored around hydroxyl or carboxyl groups, which does not necessarily reflect the trend of water molecule state changes. If multiple water molecules can aggregate with each other and do not completely flow into the pore structures of other coal macromolecules, when intense collisions occur, the water molecules will diffuse into other pore structures; this indicates that the two models represent different occurrence states of water molecules in coal. The water molecular energy after interaction was calculated according to Eq. (1)

Where: E_water - Water molecular energy after interaction; E_coal-water- Energy from mixing coal and water molecules; E_coal - coal molecular energy.

2.2. Simulation method and adsorption process

To simulate and analyze the methane adsorption law of the two types of models. Using the giant canonical monte carlo (GCMC) method and considering methane is as an incompressible gas, the interactions of coal macromolecules with the two types of water molecule models were represented through the L-J (12-6) potential function, and then the interatomic distance parameter (${\sigma }_{ij}$) and energy parameter (${\epsilon }_{ij}$) were taken arithmetically and geometrically averaged Eqs. (2-4), respectively [25].

where: $r_{ij}$- distance between atoms i and j; $r_{cut}$- truncation radius of the L-J potential function (12-6), 15.5 Å; and ${\epsilon }_{ij}$ and ${\sigma }_{ij}$ - depth and radius of the potential well of L-J (12-6); and ${\sigma }_{ii}$ and ${\sigma }_{jj}$- methane molecular diameter and intermolecular distance; ${\epsilon }_{ii}$ and ${\epsilon }_{jj}$- interaction energies between molecules i and j of methane.

The adsorption process was chosen to use the fix pressure setting in the Sorption module to select the range of force field and number of steps, in which the force field and charge were chosen to be COMPASS II and Use Current, respectively, the electrostatic potential energy and van der Waals energy were chosen to be Ewald and Atom Based, and the number of equilibrium steps and production steps were both 4 × 10⁸, and the range of pressures was set to be 0.001 MPa-10 MPa (interval 2 MPa), and the conversion relationship between pressure and fugacity was considered [26]; the temperature range was set at 283.15 K-303.15 K (interval: 5 K), and each pressure and temperature were simulated five times, and the average value was taken as the final absolute adsorption simulation result (Eq. 5), which provided a guarantee for determining the three main parameters of water molecular content, temperature and pressure.

Where: $\overline{x}$: absolute adsorption capacity (mmol·g⁻¹); n: number of methane adsorbed for the corresponding pressure; M: relative molecular weight of individual coal molecules (g·mol⁻¹); N: Number of coal molecules added

3.1. Methane absolute adsorption capacities by coal macromolecules under water molecular morphologies

Whether for the CY-CW-CMM model or the CY-RW-CMM model, at the same temperature, methane adsorption capacity basically reached a steady state once a specific pressure was attained as pressure increased. When the water molecule content was identical, the methane adsorption capacity decreased with rising temperature. With an increase in water molecular content, the absolute methane adsorption capacity consistently exhibited a decreasing trend. These observations demonstrate that the presence of water molecules and elevated temperature exert a significant inhibitory effect on methane adsorption by coal macromolecules, while pressure plays a promotional role in enhancing methane adsorption capacity. These results are consistent with the findings reported in previous studies [27,28]. Meanwhile, under high-pressure and low-temperature conditions, the driving force for methane adsorption is strong enough to allow methane to access the limited remaining space not occupied by water molecules. In contrast, under low-pressure and high-temperature conditions, the competitive advantage of water molecules is overwhelming, and the increased thermal motion of methane further impedes its adsorption (Figures 2a-l).

Close

Figure 2.

(a) CW,2MPa. (b) RW,2MPa. (c) CW,4MPa. (d) RW,4MPa. (e) CW,6MPa. (f) RW,6MPa. (g) CW,8MPa. (h) RW,8MPa. (i) CW,10MPa. (j) RW,10MPa. (Absolute methane adsorption capacity is shown in Figure a-j.) (k) Methane adsorption under different conditions in CW. (l) Methane adsorption under different conditions in RW. (m) Pore structure distribution in the CW. (n) Pore structure distribution in the RW and attached the modified version next to the original image.

Export to PPT

Close

Figure 2.

(a) CW,2MPa. (b) RW,2MPa. (c) CW,4MPa. (d) RW,4MPa. (e) CW,6MPa. (f) RW,6MPa. (g) CW,8MPa. (h) RW,8MPa. (i) CW,10MPa. (j) RW,10MPa. (Absolute methane adsorption capacity is shown in Figure a-j.) (k) Methane adsorption under different conditions in CW. (l) Methane adsorption under different conditions in RW. (m) Pore structure distribution in the CW. (n) Pore structure distribution in the RW and attached the modified version next to the original image.

Export to PPT

Close

Figure 2.

(a) CW,2MPa. (b) RW,2MPa. (c) CW,4MPa. (d) RW,4MPa. (e) CW,6MPa. (f) RW,6MPa. (g) CW,8MPa. (h) RW,8MPa. (i) CW,10MPa. (j) RW,10MPa. (Absolute methane adsorption capacity is shown in Figure a-j.) (k) Methane adsorption under different conditions in CW. (l) Methane adsorption under different conditions in RW. (m) Pore structure distribution in the CW. (n) Pore structure distribution in the RW and attached the modified version next to the original image.

Export to PPT

The difference was that when water molecular contents were increased from 1% to 5%, absolute adsorption capacity decreased in both models, with the decline in absolute adsorption capacity being more pronounced in CY-RW-CMM than in CY-CW-CMM. The two main reasons were as follows:

(1) with the addition of water molecules, the free pore structure of the CY-CW-CMM model basically remained unchanged, indicating that the larger the range of pore size structure, the corresponding peaks were close to 0. In the skeleton macroporous structure after immersion, the corresponding peaks then decreased, proving to be favorable for methane adsorption. In the small pore structure of the skeleton, water molecules were completely active in these ranges, and if water molecules competed with each other, the transport effect was greatly limited. The immersed water molecules caused the specific surface areas of coal macromolecules to be reduced, absolute adsorption capacities to be decreased, and corresponding peak values to be relatively increased (Figure 2m). Similarly, in CY-RW-CMM, with the increase of water molecules, the range of free pore structure was gradually smaller, then the resulting skeleton large and small pore structure were shifted to smaller region, and peaks corresponding to the range of the skeleton large pore structure were relatively lower, on the contrary, the peak corresponding to the range of small pore in the skeleton increased, indicating that the accumulation for a large number of water molecules produced certain “rough structure” or existed in closed form, which caused transport interference and reached an inaccessible state, then specific surface areas were weaker and occupied larger proportion of the pore structure, resulting in lower absolute adsorption capacity capacities and further enhancement of the peak (Figure 2n).

(2) Adding the same number of water molecules to the two types of models, the ratios of H/C and O/C increased significantly, and oxygen atoms of water molecules with strong electronegativity and hydrogen atoms reacted with coal macromolecules, so that not only occupied the pore structure, but also changed the aromaticity, which led to reduction in absolute methane adsorption capacities of and weakened the polar interactions and affinities between the methane and coal macromolecules. Both of them, for H/C, a higher ratio means that coal macromolecular structures were richer in hydrogen-rich groups. In CY-CW-CMM, hydrogen-rich groups may form interactions with water molecules, such as hydrogen bonding. Due to the intervention of water molecules, some of the active sites that can interact with methane molecules were occupied, resulting in a decrease in absolute methane adsorption capacities; in CY-RW-CMM, the interaction between water and coal macromolecules was wider, and a large amount of random water molecules formed “water film,” which further impeded the methane diffusion to the pore interior and adsorption, and the reduction of absolute methane adsorption capacities was even more significant compared with those in CY-CW-CMM. Meanwhile, for O/C, the larger proportion indicated that coal macromolecules contained more oxygen-containing functional groups (carboxyl and hydroxyl groups). In CY-CW-CMM, water molecules preferentially combined with oxygen-containing functional groups to form local water-functional group complexes, which changed the distribution of electron clouds and the spatial environment around the functional groups, thus weakening the adsorption between methane and functional groups and decreasing the absolute methane adsorption capacities. When in CY-RW-CMM, the water molecules not only combined with the oxygen-containing functional groups, but also filled into the pore structure, which greatly compressed the methane adsorption space, and changed the polarity and surface energy of the whole pore system due to the presence of water, which were not conducive to the non-polar methane adsorption, resulting in a significant decrease in absolute methane adsorption. This process analysis provided an important reference for which specific functional group had an effect on methane adsorption.

3.2. Parameter prediction of factors influencing methane adsorption for two types of models

3.2.2. Analysis of parameter prediction for main influencing factors on methane adsorption

3.2.2.1. Extreme variance analysis

Based on the above theory of orthogonal tests, the method of extreme variance analysis (EVA) (also known as visual analysis) is used to reveal the effects of each type of factor on the results of the two types of models in orthogonal tests. That is to say, the results corresponding to the same level of each factor are averaged, and then the minimum average value is subtracted from the maximum average value of each level to obtain the extreme deviation values. These values can reflect the influence of different levels of a given factor on the index of interest; the larger the values, indicate that the degree of influence (weight) of the factor on the test results were more significant [30]. The formula for their calculations is as follows (Eqs. 7-8):

(7)

$R_j=\max \left(K_i\right)-\min \left(K_i\right)$

(8)

$k_i=\frac{K_i}{s}$

Where: $R_j$- the extreme deviation in the column$j$; $k_i$- the arithmetic mean of the test results corresponding to the factor at the level of the orthogonal table; $K_i$- the sum of the test indicators corresponding to the level of the factor $i$ of the orthogonal table; $s$- the number of occurrences of a factor $i$ level in the orthogonal table.

After completing the analysis of the orthogonal tests for the two types of models, Table 3 presented the interactions of three-level parameters (including water molecular contents, pressures, and temperatures). As to select specific levels of the three variables, the variables for each parameter were shown in Figure 3, among which in Figure 3(a), it was found that the absolute methane adsorption capacities of both types of models induced by the water molecular content were linearly negatively correlated, and the larger the slopes of the linear negative correlation were, the smaller the absolute methane adsorption capacities were, then the larger the linear fits were and the more precise they were, which mean that under the conditions of same water molecular content, CY-RW-CMM reduced the absolute methane adsorption capacities more than CY-CW-CMM, and the inhibition effect was the most obvious; The pressure-induced absolute methane adsorption capacities for both models were found to be positively correlated and consistent with the Langmuir type I curve (Eq. 9) in Figure 3(b), and there fits were compatible with the patterns of previous studies in both experimental and simulation results [31]. The difference was found that from the two types of models and Eq. 9 that when the multiplications of ab were smaller and the values of c were larger, then the absolute methane adsorption capacities were even larger, suggesting that more methane was adsorbed by the CY-CW-CMM than by the CY-RW-CMM over a certain range of pressures, whereas the fitting degrees were both the same (R²=0.995); Linear negative correlations were found in Figure 3(c) for the absolute methane adsorption capacities of both models caused by temperature, and the larger the slopes of the linear negative correlation and the higher the degrees of fit, the smaller the absolute methane adsorption capacities, indicating that CY-RW-CMM similarly reduced more absolute adsorption capacity capacities than CY-CW-CMM under the same temperature conditions, and the thermal movement of methane was intensified, leading to decrease in the attraction between them and the adsorbent surface. However, the magnitude of the negative correlation slope corresponding to temperature was not as significant as the magnitude of the negative correlation slope corresponding to water molecular content, proving that water molecular content had the greatest effect on the absolute methane adsorption capacity.

(9)

$Y=\frac{a\cdot b\cdot p^{1-c}}{1+b{p}^{1-c}}$

Table 3. Extreme variance analysis of parameters affecting absolute methane adsorption capacities.

CY-CW-CMM	Water molecular content	Pressure	Temperature	CY-RW-CMM	Water molecular content	Pressure	Temperature
K1	8.00	5.43	7	K1	6.76	4.24	5.37
K2	7.10	6.25	6.72	K2	6.03	4.67	5.06
K3	6.33	6.83	6.58	K3	4.82	5.06	4.88
K4	6.04	7.14	6.35	K4	3.68	5.25	4.8
K5	5.45	7.26	6.27	K5	3.33	5.4	4.51
k1	1.60	1.09	1.4	k1	1.35	0.85	1.07
k2	1.42	1.25	1.34	k2	1.21	0.93	1.01
k3	1.27	1.37	1.32	k3	0.96	1.01	0.98
k4	1.21	1.43	1.27	k4	0.74	1.05	0.96
k5	1.09	1.45	1.25	k5	0.67	1.08	0.9
Rj	0.51	0.36	0.15	Rj	0.69	0.23	0.17

Note: Ki and ki are the sum and mean of absolute methane adsorption capacity at level i, respectively; Rj is the range. All K₁-K₅, k₁-k₅ and R values are in mmolg^-1.

Close

Figure 3.

(a) Water molecular content. (b) Pressure. (c) Temperature. (d) Extreme variance analysis. Variation in absolute methane adsorption capacities due to three influential parameters and extreme variance analysis.

Export to PPT

To effectively determine whether the results were reasonable or not, based on the above Eq. 9 and the results of calculating the extreme deviation ($R_j$) in the above table (Table 4), the analysis of the factors influencing the absolute methane adsorption of the two types of models were obtained from the strongest to the weakest order of the water molecular content, pressure and temperature (Figure 3d). From the results of the extreme difference values of the two types of models, it can be seen that the extreme difference values of CY-RW-CMM under the conditions of water molecular content and temperature were larger than those of CY-CW-CMM, which indicates that the inhibition of methane adsorption by the former was obvious. However, for pressure, the higher the CY-CW-CMM extreme difference value than that of CY-RW-CMM, the larger the adsorbed methane adsorption, which verified the validity of the above results. Similarly, the larger the absolute value of the results of the extreme differences in the water molecular content corresponding to the three factors in the two types of models, the more the water molecular content influences and inhibits the absolute methane adsorption capacity. To more intuitively analysis the effect of each factor on absolute methane adsorption capacities for coal macromolecules, according to the results of the extreme variance analysis in the above table (Table 4) to do the analysis of the effect of each factor on the change of absolute methane adsorption capacities of the analysis of the graph (Figures 3a-c), respectively, the optimal combination of the level of the three factors was determined to be the water molecular content of 1%, the pressure of 10 MPa and the temperature at 283.15 K.

Table 4. ANOVA results.

Model	$j$	$S{S}_j$	dfj	$V_j$	$F$	P
CY-CW-CMM	Water molecular content	0.781	4	0.195	198	<0.01
	Pressure	0.451	4	0.113	114	<0.01
	Temperature	0.069	4	0.017	17	<0.01
CY-RW-CMM	Water molecular content	1.736	4	0.434	674	<0.01
	Pressure	0.177	4	0.044	68	<0.01
	Temperature	0.081	4	0.02	31	<0.01

Note: Statistical significance was assessed at α = 0.05. P < 0.05 indicates statistical significance (all effects shown are significant; P < 0.01 as reported in the table).

3.2.2.2. Analysis of variance

The results of the extreme variance analysis described above were limited to distinguishing data fluctuations that occur as a result of the effects of the level of each variable, and did not allow for the determination of the real error, the evaluation of the differences between the means, or the parametric dominance. To overcome their limitations, the results of the experiment were subjected to ANOVA in order to perform error analysis more accurately, as well as to make a judgment on the significance of the factors. This approach (ANOVA) involved consulting the table $F$ to derive the critical value $F_{\alpha }$, given the significance level $\alpha$. The value $F$ of each factor was then calculated and compared with the critical value. If $F$ > $F_{\alpha }$ and the greater the difference between the two, the greater the influence of the factor on the indicator and the higher the significance [32]. The sum ($S{S}_j$) and the variance ($V_j$) of squared differences of each parameter were calculated as follows (Eqs. 10-11). The freedom degree of each parameter was $d{f}_j=i-1$.

(10)

$S{S}_j=\frac{1}{i}\sum _{i=1}^i{k}^2-\frac{1}{i^2}{\left(\sum _{i=1}^{i^2}{W}_{tota{l}_i}\right)}^2$

(11)

$V_j=\frac{S{S}_j}{d{f}_j}$

Based on ANOVA and the fitted equations (Eq. 6) for absolute methane adsorption capacities under the three parameters, Figure 4 and Table 4 observed that the P-values of water molecular content, pressure, and temperature were less than 0.05 in both types of models, which were proved to be the key factors affecting the absolute methane adsorption capacities. The F test critical value table determined F_0.05 (4, 4) = 6.3882. For CY-CW-CMM corresponding to water molecular content, pressure and temperature, the magnitudes of F-value relationship were 198>114>17>6.3882, among the three significant influencing factors, the water molecular content was the significant influencing factor; The F-value magnitude relationships of CY-RW-CM corresponding to water molecular content, pressure and temperature were 674>68>31>6.3882, which also proved that the water molecular content was a principal influencing factor for the absolute methane adsorption capacity. In addition, by comparing the sum of squares of water molecular content, pressure, and temperature with their errors in the two types of models, it can be concluded that the dominant factor affecting the absolute methane adsorption capacities was the water molecular content.

Close

Figure 4.

(a) CY-CW-CMM. (b) CY-RW-CMM. Regression standardized prediction.

Export to PPT

3.3. Experimental changes in functional groups before methane adsorption

3.3.1. Fourier transform infrared indoor experiment

Based on the aforementioned prediction of 1% moisture content and the identified key controlling factors, fresh coal samples were prepared as follows. First, a raw coal block was crushed using a crusher and sieved to obtain particles with a size of 100 mesh. A 5 g portion was placed into the first sealed bag as the original coal sample. Another 5 g portion was then taken, and distilled water was dropped onto its surface until the total weight reached 5.05 g. This sample was transferred to a beaker and subsequently dried in an oven at 60°C for 24 h; it was designated as the water-sprayed coal sample (corresponding to the molecular model CY-CW-CMM). Finally, a portion of coal exceeding 5 g was soaked in a beaker of pure water for 7 days, followed by vacuum filtration and natural air-drying in a fume hood at room temperature for 72 h. After drying, the sample was stored in the third sealed bag and designated as the water-soaked coal sample (corresponding to the molecular model CY-RW-CMM) (Figure 5).

Close

Figure 5.

Three modes of coal sample handling.

Export to PPT

FT-IR tests were carried out on the three coal samples, and the overall trends of the transmittance curves were basically the same (the spectral features were basically in the range of wave numbers from 400 cm^-1 to 4000 cm^-1), which were mainly divided into four regions (Figure 6), including hydroxyl groups (3600 cm^-1∼3000 cm^-1), aliphatic hydrocarbons (3000 cm^-1∼2800 cm^-1), oxygen-containing functional groups (1800 cm^-1∼1000 cm^-1) and aromatic hydrocarbons (900 cm^-1∼700 cm^-1).

Close

Figure 6.

Results of FT-IR test for three coal samples.

Export to PPT

3.3.2. Distribution characteristics of major functional groups

3.3.2.1. Hydroxyl groups (3600 cm^-1∼3000 cm^-1)

The hydroxyl groups in the three coal samples (raw coal, water-sprayed coal, and water-soaked coal) primarily exist in the forms of OH–π hydrogen bonds, OH–OH hydrogen bonds, cyclic hydrogen bonds, and OH–N hydrogen bonds (Figures 7a-c). Among these, OH–OH hydrogen bonds constitute the largest proportion: 51.28% in raw coal, 57.48% in water-sprayed coal, and 61.09% in water-soaked coal, with the highest value observed in the latter. This increase can be attributed to the fact that, when water molecules reach suitable distances and spatial orientations, they readily form self-associated hydrogen bonds with hydroxyl groups in coal. These OH–OH hydrogen bonds occupy adsorption sites that would otherwise be available for methane, thereby reducing methane adsorption capacity. In contrast, OH–π hydrogen bonds, which facilitate the interaction between methane and the aromatic ring structures of coal through van der Waals forces and thus enhance adsorption, are most abundant in raw coal (24.03%), but decrease in the water-treated samples (16.78% in water-sprayed coal and 15.45% in water-soaked coal), further weakening methane adsorption. Hydrogen bonds with cyclic structures (≈22%) and OH–N hydrogen bonds (the least abundant type) exhibit no significant differences among the three samples, and thus exert relatively minor effects on adsorption (Figure 7d).

Close

Figure 7.

(a) Raw coal. (b)Water-sprayed coal. (c) Water-soaked coal. (d) Percentage of –OH bonding types (%). FT-IR fitting spectra of hydroxyl structures.

Export to PPT

3.3.2.2. Aliphatic hydrocarbons (3000 cm^-1∼2800 cm^-1)

Coal contains methyl (-CH₃), methylene (-CH₂), and methine (-CH) groups, which exhibit characteristic peaks corresponding to asymmetric/symmetric stretching vibrations (Figures 8a-c). The key change is that the proportion of -CH₂ asymmetric stretching vibrations in water-soaked coal increases significantly: 31.38% (raw coal) →34.90% (water-sprayed coal)→47.48% (water-soaked coal) Figure 8(d). This indicates that water treatment may promote the formation of short-chain alkanes in coal; these short-chain alkanes fill the pore structure of coal and compress the methane adsorption space.

Close

Figure 8.

(a) Raw coal. (b) Water-sprayed coal. (c) Water-soaked coal. (d) Percentage of aliphatic C–H structures (%). Fourier fitting spectra of aliphatic hydrocarbon structures.

Export to PPT

In contrast, the symmetric stretching vibration of -CH in water-soaked coal decreases to the lowest value (12.96%). This may be attributed to the enhanced hydrophilicity of coal; water molecules occupy the surface of the coal matrix, which weakens the methane adsorption capacity.

3.3.2.3. Oxygen-containing functional groups (1800 cm^-1∼1000 cm⁻¹)

Oxygen-containing functional groups in coal are primarily composed of phenols, alcohols, ethers, and esters (Figures 9a-c), with the most significant variations observed in two categories: (1) conjugated C=O, whose proportion increases markedly after water treatment, reaching 38.21% in water-soaked coal compared with 28.11% in raw coal; and (2) phenolic/ether C–O, which reaches its maximum proportion in water-sprayed coal (25.82%), representing an increase of 9.24% relative to raw coal (16.58%) (Figure 9d).

Close

Figure 9.

(a) Raw coal. (b) Water-sprayed coal. (c) Water-soaked coal. (d) Percentage of oxygen-containing structures (%). Fourier fitting spectra of oxygen-containing functional group structures.

Export to PPT

Overall, the proportion of polar oxygen-containing functional groups follows the order: water-soaked coal (80.25%) > water-sprayed coal (73.78%) > raw coal (66.51%). These polar groups may alter the electron cloud distribution within coal macromolecules, thereby weakening the non-polar interactions between methane (a non-polar molecule) and coal [33]. Moreover, the increased aromatic C=O stretching vibrations observed in water-soaked coal (23%) may occupy high-energy adsorption sites, which further reduces methane adsorption capacity.

3.3.2.4. Aromatic structure (900 cm^-1 to 700 cm^-1)

Aromatic structures in coal primarily consist of pentasubstituted benzene rings (864–865 cm⁻¹), tetrasubstituted benzene rings (833–836 cm⁻¹), trisubstituted benzene rings (785–807 cm⁻¹), and disubstituted benzene rings (747–748 cm⁻¹) (Figure 10). Among these, the most pronounced variation is observed in the proportion of trisubstituted benzene rings, which increases progressively from 34.25% in raw coal to 36.04% in water-sprayed coal and 37.41% in water-soaked coal. This trend suggests that higher degrees of substitution may alter the electron cloud density of aromatic rings, thereby weakening the van der Waals interactions between methane molecules and the aromatic structures. Consequently, raw coal, with its relatively lower degree of ring substitution, exhibits a stronger methane adsorption capacity compared with the water-treated samples.

Close

Figure 10.

(a) Raw coal. (b) Water-sprayed coal. (c) Water-soaked coal. (d) Percentage of aromatic structures (%). Fourier fitting spectra of aromatic structures.

Export to PPT

In addition, the proportion of tetrasubstituted benzene rings increases in water-soaked coal (10.93%) relative to raw coal (8.21%), while the proportions of pentasubstituted and disubstituted benzene rings remain essentially unchanged. These findings indicate that water-induced variations in substituent groups within the aromatic structures contribute to the overall reduction in methane adsorption capacity in the water-sprayed and water-soaked coal samples.