Sage Journals: Discover world-class research

Abstract

In this study, the influence of different functional groups (carbonyl, ether, carboxyl, and hydroxyl) on microporous development and competitive adsorption for CO₂-CH₄ was investigated using Connolly potential theory, molecular mechanics, molecular dynamics, and grand canonical Monte Carlo based on the macromolecular representation of coal vitrinite (CV). Results indicated that microporous diameter induced from the carboxyl (COOH) and hydroxyl (OH) was more sensitive to increasing temperature than carbonyl (CO) and ether (C2O). The COOH and OH could induce more micropores than the other functional groups. The adsorption capacities of OH-CV and COOH-CV were significantly higher than C2O-CV, CO-CV, and C-CV (coal vitrinite purified by carbon atoms) and they all decrease with the decreasing temperature. Both the calculation results of Locate- and Isotherm task suggested that OH and COOH have a significantly higher adsorption amount than other coal macromolecules and the others have a close adsorption amount. The adsorption selectivities of CO₂ over CH₄ (S_A/B) for all the functional groups were higher than 1 here, indicating the adsorption preference of CO₂ over CH₄. For the pressure dependence, the S_A/B first decreases significantly for the pressure of < 4 MPa and then slightly for the pressures of >4 MPa with the increasing pressure, indicating that high pressure was not conductive to the replacement of CH₄ by CO₂. For almost all the pressures here, the S_A/B follows the order of COOH-CV > OH-CV≈C2O-CV≈Ori-CV≈CO-CV≈C-CV, indicating the highest adsorption preference of CO₂ over CH₄ for COOH, and OH-CV.

Keywords

Surface functional groups coal macromolecule microporous distribution competitive adsorption ECBM

Introduction

The reduction of CO₂ emissions could minimize these environmental problems. The CO₂ capture and sequestration (CCS) could provide a bridging strategy to the development of carbon-free energy systems (Zheng et al., 2017). Among the existing CCS technologies, the CO₂-enhanced coalbed methane (CO₂-ECBM) has dual advantages of not only reducing the CO₂ emissions but also enhancing the coalbed methane (CBM) recovery (He et al., 2017; Liu and Wilcox, 2012; Middleton et al., 2012). Recent investigations have demonstrated that coal macromolecule was consisted of three-dimensional cross-linked structure which was composed of aromatic clusters, oxygen functional groups, and aliphatic structures (Mathews and Chaffee, 2012; Mathews et al., 2011). The different oxygen function groups in coal macromolecules possessed adsorption abilities for methane, which should be attributed to their varying microporous size distribution (Song et al., 2017a, 2017b). Thus, the investigations on the effects of surface functional groups on microporous distribution and CO₂-CH₄ adsorption in coal macromolecules were of significance for the theory and engineering of enhancing CBM recovery.

Recent years have witnessed great advancements in highlighting the size distribution and heterogeneity of nanopores via experiments and molecular simulations. The experimental characterizations include the liquid injections, gas adsorption methods, and micrograph analysis, where the former was composed of the high-pressure mercury intrusion (HPMI) (Fu et al., 2021; Huang et al., 2019; Wang et al., 2014). The HPMI was conducted based on the capillarity theory and could detect the size distribution of microfractures and the macropores intuitively (>50 nm) (Li et al., 2020; Zhou et al., 2017). However, the mercury injection amounts at higher pressures (>20 MPa) were composed of two parts, that is, matrix compression and effective pore volume (Cai et al., 2018). Thus, the pore size distribution from HPMI results should be corrected especially at high pressures and the compression coefficient can be calculated. The gas adsorption method mainly includes the low temperature CO₂ (LTCO₂GA) and liquid nitrogen adsorption (LTN₂GA). The LTN₂GA which was based on the adsorption and condensation theory could detect the size distribution and morphology of mesopores (2–50 nm) and larger nanopores (Song et al., 2020). The pore size distribution, pore volume, and specific surface area could be calculated via physical models such as Brunauer-Emmett-Teller model, Barrett-Joyner-Halenda model, and Density-Function-Theory (DFT) model (Bardestani et al., 2019; Li et al., 2016; Wei et al., 2016). In addition, the pore morphology could also be inferred by the hysteresis loop shapes, as well as its corresponding pore diameter (Song et al., 2017a, 2017b). The LTCO₂GA method was efficient for characterizing the microporous size distribution and also the heterogeneity properties with the help from DFT model, volume fractal model, and surface fractal model (Gao et al., 2022). It has been successfully utilized to depict the structural properties of micropores in coal, shale and other geological materials (Hui et al., 2019; Mastalerz et al., 2017). The micrograph analysis method mainly includes the scanning electron microscope, optical microscope, and neutron scattering and could visually detect the development characters of microfractures and mesopores in coal (Gupta et al., 1998; Nie et al., 2015).

The nanoporous size distribution was critical for highlighting the adsorption microscopic mechanism of CO₂-CH₄, where the molecular simulation could provide more microscopic insights than the experimental characterizations (Mosher et al., 2013; Xiang et al., 2014; Zhang et al., 2015). Different functional groups could possess different adsorption energy and affinity for CO₂, CH₄, and H₂O (Song et al., 2017a, 2017b). This could be attributed to their varying electronic interactions, the microporous size distribution, and van der Waals force. Among the molecular simulation methods, the Grand Canonical Monte Carlo (GCMC) and molecular dynamics (MD) were effective to calculating the adsorption amounts and adsorption configurations and have been successfully utilized to depict the adsorption process under geological conditions (Zhang et al., 2015). The GCMC was conducted in the Adsorption module of Materials studio 2017 software and usually utilized the Deriding force field to characterize the interaction energy between coal macromolecule and adsorbates such as CO₂, CH₄, and H₂O (Ekramipooya et al., 2021; Yang et al., 2021). Liu et al. investigated the control effects of micropore evolution in coal vitrinite (CV) during coalification based on molecular simulation and proposed that micropore volume decreases with decreasing aliphatic structures (Liu et al., 2016). Then, he used the density functional theory to investigate the effects of chemical composition, disorder degree, and crystallite structure of coal nanopores (0.4–150 nm), clarifying that 2 to 150 nm pores are formed in the space between different basic structural units. Liu and Wilcox (2012) combined the MD and experiment to highlight the control effects of micropores on adsorption capacity and proposed that the adsorption affinity of methane was determined by the larger inaccessible microporosity.

International scholars have clarified that different surface functional groups have various adsorption affinity for CO₂ and CH₄, which should be attributed to the induced microporous size distributions. However, the contributions of different surface functional groups to the microporous development were still unclear, as well as their influences on the competitive adsorption of CO₂ and CH₄. Only through the GCMC, MM (molecular mechanics), and MD can reveal the microscopic mechanism of microporous development and adsorption process. Thus, based on the Connolly potential theory and MD, the macromolecular representation was used to calculate the microporous size distribution considering the occurrence of carboxyl, hydroxyl, carbonyl, and ether bond. Then, the microscopic mechanism of CO₂ and CH₄ competitive adsorption was investigated using GCMC and MD, as well as the influences from different surface functional groups. The outcomes of this study were of significance for the CO₂-enhancing CBM recovery and also the carbon capture, utilization, and storage.

Models and calculation methods

Coal macromolecular models

Coal macromolecule was consisted of aromatic clusters, aliphatic chains, oxygen functional groups, and heteroatom groups (Given et al., 1986; Liu et al., 2019). After entering the 21st century, the research of coal molecular geochemistry has achieved fruitful results combined with the further development of computer-aided molecular design especially the development and progress of ¹³C nuclear magnetic resonance (NMR), Fourier transform infrared spectroscopy (FTIR), X-ray photoelectron spectroscopy and high-resolution transmission electron microscopy, providing the basis for the simulation of gas behaviors and coal geology (Song et al., 2021). The coal macromolecular representation used here was created via elemental analysis, ¹³C NMR, FTIR, and high-resolution transmission electron microscope from Song et al. (2019), whose predicted ¹³C NMR spectrum was in line with the experimental results. This indicates the rationality of the representation and has been successfully used to calculate and simulate the adsorption, diffusion, and the induced swelling deformation of gas molecules in coal (Song et al., 2017a, 2017b). The plane representations of the coal macromolecule of different functional groups were depicted in Figure 1(a)–(f), respectively.

Figure 1.

The plane representation of the coal macromolecule of different functional groups: (a) origin representation; (b) coal representation purified by COOH; (c) coal representation purified by ether bond; (d) coal representation purified by carbonyl; (e) coal representation purified by hydroxyl; and (f) coal representation purified by carbon (modified from Song et al., 2019).

The origin coal macromolecule (C₂₁₅H₁₇₆N₂O₂₄) was depicted in Figure 1(a), which was composed of the oxygen functional groups such as 6 ethers, 1 hydroxyl, 7 carboxyl and 3 carbonyl, as well as 1 pyrrole groups and 1 pyridine groups. To clarify the influence of different oxygen functional groups on the microporous size distribution and gas adsorption behaviors, the origin coal macromolecule was purified by COOH, C2O (ether bond), C = O (simplified by CO, carbonyl), OH, and C (pure carbon atoms), resulting in the different coal macromolecule of different functional groups, which has the chemical formula of C₂₂₂H₁₈₈O₂₆, C₂₂₀H₂₁₀O₁₆, C₂₂₀H₁₇₈O₁₆, C₂₁₉H₂₀₈O₁₆, and C₂₁₆H₂₄₂, respectively (Song et al., 2019). These coal representations were created to reveal the effects of the specific oxygen functional groups on the microporous size distribution and gas adsorption behaviors, which is attributed to the fact that the quantities and locations of the COOH, C2O, CO, OH, and C are basically identified. Thus, the difference in microporous size distribution and gas adsorption behaviors were induced from the types of various functional groups. Here, they were abbreviated as Ori-CV, COOH-CV, C2O-CV, CO-CV, OH-CV, and C-CV, whose plane configurations were depicted in Figure 1(b)–(f), respectively.

GCMC and MD simulation

The plane configurations in Figure 1 were first hydrogenation saturated and then both the MD and MM were conducted to search the optimized configurations. The MD was conducted in Forcite module using Smart Minimizer as the structural optimization method. After each cycle, MM calculation is performed for the output configuration to ensure the low-energy state. The density simulation in this study was conducted using the Amorphous Cell (AC) module based on the configurations obtained via MM and MD. In combination with the characteristics of the simulation software, the parameter settings are re-corrected. Using the construction task, the medium level was selected as the quality where the density value should be calculated in this module. Then, the molecular structures were imported into the composition.

A number of previous studies on coal and its products have employed the Dreiding force field (Carlson, 1992; Nakamura, 1993; Rogel and Carbognani, 2003), which is a very general force field that can be used for a large number of atom types. It allows calculation of the total energy of a molecular structure as a sum of bonding interactions (E_An = angel, E_Bo = bond, E_Tor = torsion, and E_In = inversion) and nonbonding interactions (E_Van = van der Waals, E_elec = electrostatic, and E_hy = hydrogen bonds). On the other hand, the physical structural parameter such as density and porosity calculated via Dreiding force field were more close to the experimental values than the other types of force field. Thus, the Dreiding fore field was used with the charge calculation method of (charge equilibrium) QEq method. In the AC module, put the structure model into a periodic boundary condition, where the cell size can be automatically adjusted according to the size of the structure model or manually input. The density value calculated by simulation in this study is 0.4–1.6 g/cm³, and the interval is 0.1 g/cm³. The structural model of the calculated results is optimized by MM and MD until the energy value converges. The variations of the total energy and the corresponding configuration during the density simulation were depicted in Figure 2. The total potential energy first decreases and then increases with the increasing density, also exhibiting a significant concussion phenomenon with density >1.2 g/cm³. Since the actual macromolecular structure is not in the lowest energy state due to relaxation caused by tectonic stress and solvent swelling action, thus, 1.24 g/cm³, the density corresponding to local minimum value total potential energy after 1.2 g/cm³ was selected as the true density, producing its corresponding configuration for the latter calculation (Song et al., 2017a, 2017b).

Figure 2.

The variation of the total potential energy and the corresponding configuration during the density simulation (Song et al., 2017a, 2017b).

The microporous morphology and size distribution for the coal macromolecule were obtained through the Atom Volumes & Surface tool in the Materials studio 2017 software. The Connolly surfaces and the Solvent surfaces were both calculated using the ultra-fine grid resolution with a vdW scale factor of 1.0000. The maximum solvent radius and Connolly radius were set as 2.0000 Å and 1.0000 Å, respectively. The occupied volume of aromatic skeleton (V_o^F) and free volume (pore volume, V^F) can be obtained in the output interface. Then, the microporous size distribution can be calculated by gradually enhancing the diameter of the molecular probe (Connolly, 1983; Zhao et al., 2016). Finally, the median pore size was obtained using the normal distribution function to fit the microporous size distribution. The pore size distribution within coal macromolecule was depicted in Figure 3(a)–(f) for different coal macromolecular representations, where the gray zone and blue zone were depicted as occupied and free volume, respectively. The following parameters to characterize the microporous development were used: (1) the median pore size (D^M), (2) occupied volume of aromatic skeleton (V_o^F), and (3) free volume (pore volume, V^F).

Figure 3.

The pore size distribution of different coal macromolecular representations (the gray zone and blue zone were depicted as occupied and free volume, respectively): (a) origin representation; (b) coal representation purified by COOH; (c) coal representation purified by ether bond; (d) coal representation purified by carbonyl; (e) coal representation purified by hydroxyl; and (f) coal representation purified by carbon.

Monte Carlo simulation is based on statistical sampling method which uses the random number, sampling, or simulation of random variable statistics to obtain the digital characteristics of a variable. The GCMC refers to the Monte Carlo simulation of constant temperature, volume, and chemical potential conducted in the Grand Canonical Ensemble (Cheng et al., 2004). The adsorption configuration can be obtained using the Local task (Cheng et al., 2004) and the sampling method is based on the Metropolis rule (Metropolis et al., 1953). Specifically, the probability of exchange being accepted is 39%, the probability of conformational isomer being accepted is 20%, the probability of rotation being accepted is 20%, the probability of translation being accepted is 20%, and the probability of Regrow growth being accepted is 1%, which reduces the energy of the system and forms a new configuration. The calculation quality was set as “Medium” and the number of maximum loading steps was 100,000 with a production step of 100,000 as well. The temperature cycle was 4 for the annealing simulation.

Results and discussion

Microporous size distribution

The variations of D_M and V^O at different temperatures were depicted in Figure 4(a) and (b). The median pore size reflected the median value of the pore diameter within the coal macromolecules. It can be found that the D_M of the different coal macromolecules all decrease with the decreasing temperature. At 298 K, the D_M follows the order of COOH-CV > OH-CV > CO-CV > C2O > Ori-CV > C-CV, indicating that the COOH and OH could induce much more larger micropores than other oxygen functional groups and origin coal macromolecule, as well as the carbon-purified coal macromolecule. The D_M of CO-CV, C2O, Ori-CV, and C-CV was basically the same at this temperature. Then, the D_M of COOH-CV, OH-CV, CO-CV, and C2O decreases with the increase of the temperature and the reduce rate of the former two was significantly higher than the latter two, suggesting that the diameter of micropores induced from the COOH and OH was more sensitive than the CO and C2O. And at the temperature of 358 K, the D_M of the oxygen functional groups still follows the order of COOH-CV > OH-CV > CO-CV > C2O. However, that of Ori-CV and C-CV keeps the stable with the increase of the temperature (Figure 4(a)), suggesting that they are insensitive to the variations of the temperature for Ori-CV and C-CV. It should be noted that the D_M of Ori-CV and C-CV was higher than that of the OH-CV, CO-CV, and C2O.

Figure 4.

The variations of (a) the median pore size (D_M) and (b) solid phase which is occupied by aromatic skeleton (V^O) under different temperatures.

The variations of the V^O at different temperatures were depicted in Figure 4(b). Unlike the D_M, the V^O follows the order of Ori-CV > C-CV > OH-CV > COOH-CV > C2O-CV > CO-CV at 298 K, suggesting that the occupied volume of the Ori-CV and C-CV was the highest compared with the coal macromolecule purified by different oxygen functional groups. The inclusions of the oxygen functional groups reduces the occupied volume of coal macromolecule, which should be attributed the variations of the van der Waals force within the coal macromolecules. Then, the occupied volume of the Ori-CV, OH-CV, and C2O increases gradually with the increasing temperature, however, that of the other coal macromolecule such as the C-CV, COOH-CV, and CO-CV change slightly with the increasing temperature, suggesting that the inclusion of the oxygen functional groups enhances the heat sensibility of the occupied volume. At the temperature of 338 K and 358 K, the occupied volume follows the order of Ori-CV > OH-CV > C-CV > C2O-CV > COOH-CV > CO-CV, also indicating the oxygen functional groups could induced from more occupied volume than the pure carbon atoms. Compared with the variations of the D_M and V^O, it could be concluded that the D_M and V^O have the opposite change trend with the increasing temperature. The increasing V^O for the coal macromolecule should be attributed to the enhancement of the thermoplasticity of the organic molecules. At the highest temperature of 358 K considered here, the V^O follows the order of Ori-CV > OH-CV > C-CV≈COOH-CV > C2O-CV > CO-CV, which is basically contrary to the variation of D_M.

The variations of free volume (V^F) and accessible surface area (S^A) under different temperatures were depicted in Figure 5(a) and (b), respectively. Here the unit of the V^F and S^A were Å³ and Å², respectively, they can be converted into cm³/g and m²/g using the cell size (15.8 Å × 15.8 Å × 15.8 Å) and density (1.24 g/cm²) as mentioned above. The conversion equation was 1 Å³ per cell = 1/(15.8³ × 1.24) cm³/g and 1 Å² per cell = (1 × 10⁴)/(15.8³ × 1.24) m²/g, respectively. Since the influence of different functional groups will be discussed here, the microscopic unit (Å³ and Å²) were still utilized here to analyze the microporous development in this study. At 298 K, the V^F follows the order of OH-CV > Ori-CV≈CO-CV≈COOH-CV > C2O-CV > C-CV, indicating that the inclusion of OH could increase more free volume than the pure carbon atoms. The induced effects of carbonyl and carboxyl functional groups were basically the same and the pure carbon atoms have the negative influence for the development of the free volume of the micropores within coal macromolecule. The free volume (V^F) gradually increases with the increasing temperature, especially for the OH-CV and COOH-CV. The increase rate of the C-CV, Ori-CV, and CO-CV was lower compared with the OH-CV and COOH-CV, indicating the insensitivity of the C-CV, Ori-CV, and CO-CV in response to the increasing temperature. Compared with the variation of the D_M, the COOH and OH could induce higher micropore than the other functional groups such as CO and C2O. The pure carbon atoms have a negative effect on the microporous development. The ether bond has the lowest development among these functional groups. The synergistic effect of the OH, COOH, CO, and C2O results in the higher V^F of the Ori-CV than C-CV (Figure 5(a)), also indicating the induced effects of the OH and COOH for microporous development.

Figure 5.

The variations of (a) a void phase known as free volume (V^F) and (b) an accessible surface area (S^A) under different temperatures.

The variations of S^A for various coal macromolecules purified by different oxygen functional groups were depicted in Figure 5(b). At 298 K, the S^A of OH-CV, Ori-CV, CO-CV, COOH-CV, and C2O-CV were basically the same and they were significantly higher than that of C-CV, also suggesting that the inclusion of oxygen functional groups could enhance the development of the micropores in coal. Then, the S^A increases with the increase of the temperature. At the temperature of 298 K and 318 K, the S^A for various coal macromolecules follows the order of OH-CV > CO-CV > COOH-CV > Ori-CV≈C2O-CV > C-CV. It should be noted that this order was not always consistent with the variation of V^F (Figure 5(b)). With the further increase of temperature, the S^A increases linearly. At 338 K and 358 K, the S^A follows the order of OH-CV > COOH-CV > CO-CV > C2O-CV > Ori-CV > C-CV, suggesting that the OH and COOH could induce more micropores than the other functional groups. The porosity of different coal macromolecule can be calculated the V^F and V^O and the variation was depicted in Figure 6. It was larger than the values reported in previous reference (Islam et al., 2020; Liu et al., 2023; Ramandi et al., 2016) and was basically consistent with the variations of V^F and V^O. The higher porosity of these coal macromolecules should be attributed to two reasons: (1) the molecular probe could detect both the open and closed pores within coal macromolecule, however, the experimental methods can only detect the accessible micropores and (2) the coal contains mineral matters and they can plug and fill into the micropores.

Figure 6.

The variations of porosity under different temperatures.

Single adsorption

The simulation method here used the procedure proposed by Song et al. (2017a). Firstly, the CH₄ molecules were added one by one to the coal macromolecule using the Metropolis rule and Deriding force field, producing the adsorption configuration and potential energy composition. Then, the total potential energy decreases as the increasing number of the CH₄ molecules, suggesting that the adsorption process was an advanced process to approach the lowest energy configuration. The adsorption process could continue as long as the total potential energy decreases, finally, the adsorption process ends up if the total potential energy increases at a certain number of CH₄ molecules, indicating the saturated adsorption state of the adsorption for different coal macromolecules. The saturated adsorption configuration and the corresponding potential energy distribution.

The variations of saturated adsorption capacity for CH₄ for different coal macromolecules were depicted in Figure 7(a). The adsorption capacities of the OH-CV, Ori-CV, and COOH-CV were significantly higher than C2O-CV, CO-CV, and C-CV and they all decrease with the decreasing temperature, which should be attributed to the reduction in activities of molecular kinetic energy. However, there exists significant difference in the decrease rate for different coal macromolecules. At the temperature of higher than 318 K, the CO-CV, and C-CV have a higher reduction rate than the others, suggesting their temperature sensitivity, followed by the Ori-CV, and COOH-CV. The OH-CV and C2O-CV have a lower temperature sensitivity and they both show a volatile decrease with the increasing temperature (Figure 7(a)). The Figure 7(b) depicted the variations of saturated adsorption capacity for CO₂ and it could be clearly showed that the saturated adsorption capacity for CO₂ was significantly higher than that of CH₄ for all the coal macromolecules considered here, indicating the adsorption advantages of CO₂ over CH₄ despite of various functional groups. At lower temperatures, the adsorption capacity of CO₂ follows the order of OH-CV > CO-CV > COOH-CV > C2O-CV > C-CV, which was basically consistent with that of CH₄. The saturated adsorption capacity of CO₂ in coals purified by oxygen functional groups decreases significantly with the increase of the temperature, however, the saturated adsorption capacity of C-CV decreases slightly or even increases with the increase of temperature (Figure 7(b)), indicating its temperature insensitivity in response to the heat.

Figure 7.

The variations of saturated adsorption capacity for (a) CH₄ and (b) CO₂ for different coal macromolecules.

Besides the adsorption configuration, the adsorption isotherm of CO₂ and CH₄ can be calculated using the Isotherm task in the adsorption module. The adsorption isotherm (298 K) of CH₄ and CO₂ at the pure adsorption case was depicted in Figure 8(a) and (b), respectively. The adsorption amounts first increases sharply and then keeps stable with the increasing pressure. At the higher pressures, the maximum adsorption amounts of CH₄ follow the order of OH-CV > COOH-CV > Ori-CV≈CO-CV > C-CV≈C2O > CV, suggesting the higher adsorption affinity of OH and COOH than pure carbon atoms and other oxygen functional groups (Figure 8(a)). The calculation results of Isotherm task were in line with the Locate task. It should be noted that the OH and COOH have a significantly higher adsorption amount than other coal macromolecules and the others have a close adsorption amount, which is consistent with the higher micropores of OH and COOH discussed above. Here, the adsorption curves of CH₄ were more smooth than CO₂, the latter was more complex with more mutation points (Figure 8(a) and (b)). The maximum adsorption amounts of CO₂ follow the order of Ori-CV > OH-CV≈COOH-CV > C2O-CV > CO-CV > C-CV (Figure 8(b)), in line with the results of Locate task.

Figure 8.

The adsorption isotherm of (a) CH₄ and (b) CO₂ at the pure adsorption case.

Although the results in Figure 8(b) were obtained via the Isotherm task, the maximum adsorption amount of CO₂ was still higher than CH₄ for the coal representations such as Ori-CV, OH-CV, COOH-CV, C2O-CV, and CO-CV. However, the maximum adsorption amounts of CO₂ and CH₄ are similar in Figure 8(b), it was insufficient to analyze the adsorption preference of CO₂ and CH₄ due to two reasons: (1) this was the result from the pure gas adsorption and can’t depict the binary competitive adsorption system; (2) the “maximum adsorption amounts” in Figure 8 here were different from the “saturated adsorption capacity” in Figure 7. Thus, the adsorption isotherm of CO₂ and CH₄ here was to depict the adsorption affinity of different functional groups. The adsorption selectivity of CO₂ over CH₄ will be discussed in the following co-adsorption results.

Co-adsorption

The competitive adsorption configuration of CO₂ and CH₄ onto coal macromolecules purified by different oxygen functional groups can be calculated using the Locate Task and Dreiding force field. First, 20 CO₂ molecules were absorbed onto these coal macromolecules and then geometrically and energetically optimized using the MD in the Forcite Task. Song et al. (2017a, 2017b) utilized the GCMC and DFT to clarify the optimal adsorption sites of the CV macromolecules and found that the aromatic clusters have higher adsorption energy for both CH₄ and CO₂, indicating the stronger adsorption affinity of aromatic clusters than aliphatic structures and oxygen functional groups. The outcomes of this study also support this conclusion. Then, these adsorption configurations were used to simulate the adsorption amounts of CH₄ at the case of 20 CO₂ molecules. The saturated adsorption capacity of CH₄ follows the order of COOH-CV > C2O-CV > OH-CV > Ori-CV > CO-CV > C-CV, indicating that the COOH, C2O, and OH have a higher adsorption affinity at the competitive adsorption of CO₂-CH₄, followed by the Ori-CV and CO-CV. The higher adsorption amounts of CO₂ than CH₄ here also suggested the stronger adsorption affinity of CO₂ over CH₄. At the competitive adsorption situation, the optimal adsorption sites could preferentially be occupied by CO₂ and CH₄ trends to settle at the suboptimal adsorption sites.

An important parameter adsorption selectivity was used to characterize the adsorption selectivity of CO₂ over CH₄, which was defined as: S_A/B = (x_A/x_B) _adsorbed /(y_A/y_B) _bulk, where the x_i denotes the average mole fraction of component i and the subscripts ()_adsorbed and ()_bulk refer to the quantities of adsorbed phase in pores and bulk phases, respectively. The variations of adsorption selectivity of CO₂ over CH₄ under different pressures and different temperatures were depicted in Figure 9(a) and (b), respectively. All the S_A/B was higher than 1 here, indicating the adsorption preference of CO₂ over CH₄ for almost all the cases. For the pressure dependence, the S_A/B first decreases significantly for the pressure of < 4 MPa and then slightly for the pressures of >4 MPa with the increasing pressure, indicating that high pressure was not conductive to the replacement of CH₄ with CO₂. For almost all the pressures here, the S_A/B follows the order of COOH-CV > OH-CV≈C2O-CV≈Ori-CV≈CO-CV≈C-CV, indicating the highest adsorption preference of CO₂ over CH₄ for COOH-CV > OH-CV. However, for the temperature dependence, the S_A/B of COOH-CV was significantly higher than that of the other coal macromolecules and the others have the close S_A/B, suggesting that the temperature has an interference effect for the competitive adsorption of CO₂ and CH₄. All the S_A/B here gradually decreases with the decreasing temperatures, indicating that high temperature was unfavorable for the replacement of CH₄ with CO₂ for the temperatures of 298 K to 358 K here.

Figure 9.

The variations of adsorption selectivity of CO₂ over CH₄ under (a) different pressures and (b) different temperatures.

Conclusions

In this study, the microporous development and competitive adsorption of CO₂-CH₄ were investigated using Connolly potential theory, MD, and GCMC, as well as the influences from different surface functional groups. The main conclusions were as follows.

The diameter of micropores induced from the COOH and OH was more sensitive to temperature than CO and C2O. However, the D^M of Ori-CV and C-CV keeps the stable with the increase of the temperature, suggesting their insensitivity to the variations of the temperature for Ori-CV and C-CV. Compared with the variation of the D_M, the COOH and OH could induce higher microporous development than other functional groups such as CO and C2O. The pure carbon atoms have a negative effect on the microporous development. The ether bond has the lowest development among these functional groups. The synergistic effect of the OH, COOH, CO, and C2O results in the higher V^F of the Ori-CV than C-CV, also indicating the induced effects of the OH and COOH for micropores.

The adsorption capacities of OH-CV, Ori-CV, and COOH-CV were significantly higher than C2O-CV, CO-CV, and C-CV and they all decrease with the decreasing temperature. The saturated adsorption capacity for CO₂ was significantly higher than CH₄ for all the coal macromolecules here, indicating the adsorption advantages of CO₂ over CH₄ despite of types of functional groups. Both the calculation results of Locate- and Isotherm task suggested that OH and COOH have a significantly higher adsorption amount than other coal macromolecules and the others have a close adsorption amount.

All the adsorption selectivity of CO₂ over CH₄ (S_A/B) was higher than 1 here, indicating the adsorption preference of CO₂ over CH₄ for almost all the cases. For the pressure dependence, the S_A/B first decreases significantly for the pressure of < 4 MPa and then slightly for the pressures of >4 MPa with the increasing pressure, indicating that high pressure was not conductive to the replacement of CH₄ with CO₂. For almost all the pressures here, the S_A/B follows the order of COOH-CV > OH-CV≈C2O-CV≈Ori-CV≈CO-CV≈C-CV, indicating the highest adsorption preference of CO₂ over CH₄ for COOH-CV > OH-CV.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article

ORCID iD

He Liu

References

Bardestani

Patience

Kaliaguine

(2019) Experimental methods in chemical engineering: Specific surface area and pore size distribution measurements-BET, BJH, and DFT. The Canadian Journal of Chemical Engineering 97(11): 2781–2791.

Cai

Liu

, et al. (2018) Insights into matrix compressibility of coals by mercury intrusion porosimetry and N₂ adsorption. International Journal of Coal Geology 200: 199–212.

Carlson

(1992) Computer simulation of the molecular structure of bituminous coal. Energy & Fuels 6(6): 771–778.

Cheng

Yuan

Zhao

, et al. (2004) GCMC simulation of hydrogen physisorption on carbon nanotubes and nanotube arrays. Carbon 42(10): 2019–2024.

Connolly

(1983) Solvent-accessible surfaces of proteins and nucleic acids. Science (New York, N.Y.) 221(4612): 709–713.

Ekramipooya

Valadi

Latifi pour

, et al. (2021) Effect of the pyrrolic nitrogen functional group in the selective adsorption of CO₂: GCMC, MD, and DFT studies. Energy & Fuels 35(19): 15918–15934.

Fang

, et al. (2021) Accurate characterization of full pore size distribution of tight sandstones by low-temperature nitrogen gas adsorption and high-pressure mercury intrusion combination method. Energy Science & Engineering 9(1): 80–100.

Gao

Song

, et al. (2022) Effects of coalification on nano-micron scale pore development: From bituminous to semi-anthracite. Journal of Natural Gas Science and Engineering 105: 104681.

Given

Marzec

Barton

, et al. (1986) The concept of a mobile or molecular phase within the macromolecular network of coals: A debate. Fuel 65(2): 155–163.

10.

Gupta

Wall

Kajigaya

, et al. (1998) Computer-controlled scanning electron microscopy of minerals in coal-implications for ash deposition. Progress in Energy and Combustion Science 24(6): 523–543.

11.

Linak

Deng

, et al. (2017) Particulate formation from a copper oxide-based oxygen carrier in chemical looping combustion for CO₂ capture. Environmental Science and Technology 51(4): 2482–2490.

12.

Huang

Chen

Dang

, et al. (2019) Discussion on the rising segment of the mercury extrusion curve in the high pressure mercury intrusion experiment on shales. Marine and Petroleum Geology 102: 615–624.

13.

Hui

Pan

Luo

, et al. (2019) Effect of supercritical CO₂ exposure on the high-pressure CO₂ adsorption performance of shales. Fuel 247: 57–66.

14.

Islam

Hasan

Shah

, et al. (2020) High yield activated porous coal carbon nanosheets from Boropukuria coal mine as supercapacitor material: Investigation of the charge storing mechanism at the interfacial region. Journal of Energy Storage 32: 101908.

15.

Tian

Chen

, et al. (2016) Application of low pressure gas adsorption to the characterization of pore size distribution of shales: An example from Southeastern Chongqing area, China. Journal of Natural Gas Geoscience 1(3): 221–230.

16.

Liu

Cai

, et al. (2020) Pore structure and compressibility characteristics of heat-treated coals by N₂ adsorption/desorption and mercury intrusion porosimetry. Energy & Fuels 34(3): 3173–3187.

17.

Liu

Han

, et al. (2023) Modeling of gas migration in a dual-porosity coal seam around a borehole: The effects of three types of driving forces in coal matrix. Energy 264: 126181.

18.

Liu

Song

, et al. (2019) Insight into the macromolecular structural differences between hard coal and deformed soft coal. Fuel 245: 188–197.

19.

Liu

Qiu

, et al. (2016) Molecular simulation of CH₄, CO₂, H₂O and N₂ molecules adsorption on heterogeneous surface models of coal. Applied Surface Science 389: 894–905.

20.

Liu

Wilcox

(2012) Effects of surface heterogeneity on the adsorption of CO₂ in microporous carbons. Environmental Science and Technology 46(3): 1940–1947.

21.

Mastalerz

Hampton

Drobniak

, et al. (2017) Significance of analytical particle size in low-pressure N₂ and CO₂ adsorption of coal and shale. International Journal of Coal Geology 178: 122–131.

22.

Mathews

Chaffee

(2012) The molecular representations of coal—A review. Fuel 96: 1–14.

23.

Mathews

Van Duin

ACT

, et al. (2011) The utility of coal molecular models. Fuel Processing Technology 92(4): 718–728.

24.

Metropolis

Rosenbluth

Marshall

, et al. (1953) Equation of state calculations by fast computing machines. The Journal of Chemical Physics 21: 1087–1092.

25.

Middleton

Keating

Stauffer

, et al. (2012) The cross-scale science of CO₂ capture and storage: From pore scale to regional scale. Energy & Environmental Science 5(6): 7328–7345.

26.

Mosher

Liu

, et al. (2013) Molecular simulation of methane adsorption in micro-and mesoporous carbons with applications to coal and gas shale systems. International Journal of Coal Geology 109: 36–44.

27.

Nakamura

(1993) CAMD study of coal model molecules.1.Estimation of physical density of coal model molecules. Energy & Fuels 7(3): 347–350.

28.

Nie

Liu

Yang

, et al. (2015) Pore structure characterization of different rank coals using gas adsorption and scanning electron microscopy. Fuel 158: 908–917.

29.

Ramandi

Mostaghimi

Armstrong

, et al. (2016) Porosity and permeability characterization of coal: A micro-computed tomography study. International Journal of Coal Geology 154: 57–68.

30.

Rogel

Carbognani

(2003) Density estimation of asphaltenes using molecular dynamics simulations. Energy & Fuels 17(2): 378–386.

31.

Song

Jiang

Lan

(2019) Competitive adsorption of CO₂/N₂/CH₄ onto coal vitrinite macromolecular: Effects of electrostatic interactions and oxygen functionalities. Fuel 235: 23–38.

32.

Song

Jiang

, et al. (2020) A review on pore-fractures in tectonically deformed coals. Fuel 278: 118248.

33.

Song

Jiang

(2017b) Molecular simulation of CH₄/CO₂/H₂O competitive adsorption on low rank coal vitrinite. Physical Chemistry Chemical Physics 19(27): 17773–17788.

34.

Song

Jiang

Wei

, et al. (2021) A review on the application of molecular dynamics to the study of coalbed methane geology. Frontiers in Earth Science 9: 775497.

35.

Song

Zhu

(2017a) Macromolecule simulation and CH₄ adsorption mechanism of coal vitrinite. Applied Surface Science 396: 291–302.

36.

Wang

Zhu

Chen

, et al. (2014) Characteristics of the nanoscale pore structure in Northwestern Hunan shale gas reservoirs using field emission scanning electron microscopy, high-pressure mercury intrusion, and gas adsorption. Energy & Fuels 28(2): 945–955.

37.

Wei

Zhang

Xiong

, et al. (2016) Nanopore structure characterization for organic-rich shale using the non-local-density functional theory by a combination of N₂ and CO₂ adsorption. Microporous and Mesoporous Materials 227: 88–94.

38.

Xiang

Zeng

Liang

, et al. (2014) Molecular simulation of the CH₄/CO₂/H₂O adsorption onto the molecular structure of coal. Science China Earth Sciences 57(8): 1749–1759.

39.

Yang

Yin

Xue

, et al. (2021) Construction of Buertai coal macromolecular model and GCMC simulation of methane adsorption in micropores. ACS Omega 6(17): 11173–11182.

40.

Zhang

Liu

Clennell

, et al. (2015) Molecular simulation of CO₂–CH₄ competitive adsorption and induced coal swelling. Fuel 160: 309–317.

41.

Zhao

Feng

Zhang

(2016) Molecular simulation of CO₂/CH₄ self- and transport diffusion coefficients in coal. Fuel 165: 19–27.

42.

Zheng

Zhang

, et al. (2017) Energy related CO₂ conversion and utilization: Advanced materials/nanomaterials, reaction mechanisms and technologies. Nano Energy 40: 512–539.

43.

Zhou

Liu

Cai

, et al. (2017) Multi-scale fractal characterizations of lignite, subbituminous and high-volatile bituminous coals pores by mercury intrusion porosimetry. Journal of Natural Gas Science and Engineering 44: 338–350.

Influence of functional groups on the microporous development and CO 2 -CH 4 adsorption: A molecular simulation investigation

Abstract

Keywords

Introduction

Models and calculation methods

Coal macromolecular models

GCMC and MD simulation

Results and discussion

Microporous size distribution

Single adsorption

Co-adsorption

Conclusions

Footnotes

Declaration of conflicting interests

Funding

ORCID iD

References