H₂/CH₄ Competitive Adsorption of LTA Zeolite: Effects of Cations, Si/Al Ratio, Adsorption Temperature, and Pressure

Abstract

The efficient separation of H₂ from CH₄ is crucial for hydrogen purification from industrial off-gases using pressure swing adsorption (PSA). In this study, the competitive adsorption behavior of H₂/CH₄ on LTA zeolites was systematically investigated via grand canonical Monte Carlo (GCMC) simulations, with a focus on the effects of cation type (Na⁺, Li⁺, Ca²⁺, Mg²⁺), Si/Al ratio (1–1.5), temperature (298–318 K), and pressure (0.2–2 MPa). The results reveal that CH₄ favors β-cages as excellent adsorption sites with high population density, followed by the regions adjacent to the cations or framework oxygen atoms of the eight-membered rings. In contrast, H₂ is uniformly distributed throughout all the channels. Cations with higher valence and smaller ionic radii (e.g., Mg²⁺) enhance CH₄ adsorption capacity and diffusion more effectively than monovalent or larger cations. Increasing the Si/Al ratio reduces cation content and exposes more framework oxygen atoms, particularly those in Si–O–Si environments, which improve CH₄ adsorption. Elevated temperature weakens CH₄ adsorption while promoting H₂ diffusion and pore occupancy. Although higher pressure increases the uptake of both gases, H₂ adsorption rises more substantially and distributes more widely, leading to a decrease in CH₄/H₂ selectivity.

1. Introduction

Prior to the maturity of low-cost, large-scale water electrolysis for green hydrogen production, pressure swing adsorption (PSA) is a key hydrogen source, purifying H₂ from industrial off-gases (e.g., natural gas-derived hydrogen, coke oven gas). These off-gases have complex compositions, dominated by H₂, CH₄, and CO₂, and other impurities like CO. CO₂ exhibits high adsorption capacity and kinetics on activated carbon and zeolite-based adsorbents, while CH₄ has weak adsorption on conventional adsorbents, resulting in an extended mass transfer zone in the column. Moreover, studies have shown that a large amount of H₂ remains in the adsorption tower after the adsorption step, resulting in low H₂ recovery rates or the necessity to add pressure equalization steps for H₂ recovery [1,2,3,4,5,6,7].

Zeolites are a class of porous adsorbents with diverse compositions and structures, whose frameworks are three-dimensional networks formed by oxygen-bridged TO₄ tetrahedra containing Si, Al, and heteroatoms (e.g., B, Fe, Ga) [8]. They have abundant pore structures, including 4-, 6-, 8-, and 12-membered rings. Some zeolites (e.g., LTA, CHA types) have pore windows dominated by a single characteristic ring (both centered on 8-membered rings), with other framework rings only playing auxiliary roles; others (e.g., FAU, BEA types) are hierarchical porous materials with multi-type interconnected pores (FAU: 12-membered ring pores; BEA: intersecting 12-membered ring pores), forming multifunctional pore systems. For gas purification, conventional aluminosilicate zeolites are widely used owing to their favorable thermal stability [2,3,8,9,10,11]. Classified by Si/Al ratios, zeolites are categorized into low-silica (Si/Al < 2), medium-silica (2 ≤ Si/Al ≤ 5), high-silica (Si/Al > 5), and all-silica types, with pore sizes ranging from micropores to macropores [8,12]. The cation type, distribution, and pore topology together endow zeolites with excellent adsorption selectivity and molecular sieving ability. LTA-type zeolites are extensively applied in H₂/CH₄ separation [7,10,13,14]. Jeong et al. [15] compared the adsorption thermodynamic and kinetic parameters of H₂ and CH₄ on 4A, 5A, and 13X zeolites, and found that 5A achieves a balance of high CH₄ adsorption capacity, low H₂ adsorption capacity, and fast adsorption rate. Liu et al. [14] reported that 13X has slightly higher CH₄ adsorption capacity than 5A, but much higher H₂ adsorption capacity. Lopes et al. [7] calculated the adsorption capacity and diffusion coefficient of H₂ and CH₄ on 5A carbon, revealing that the H₂ diffusion coefficient on 5A is higher than that of the activated carbon studied under the same conditions, which facilitates taking advantage of the H₂/CH₄ adsorption kinetic differences. Yuan et al. [15] investigated the CH₄/H₂ adsorption–separation performance of various zeolites via a high-throughput computational method based on molecular simulation. By evaluating the H₂/CH₄ adsorption–separation performance of 199 zeolite structures, they found that the CH₄ adsorption selectivity of zeolites is independent of bulk pressure and feed ratio. Specifically, CH₄ preferentially occupies the small pore window spaces of zeolites, while H₂ shows no distinct preferential adsorption sites. Liu Yanna [16] synthesized SrA submicron A-type molecular sieves and SrA hierarchical porous A-type molecular sieves (with Na⁺ as the cation before modification). At 0.4~1.0 MPa and 298 K, the CH₄/H₂ selectivity coefficients of the two adsorbents were 9.14~6.17 and 7.99~6.10, respectively, while that of the commercial 5A adsorbent was determined to be 8~5.3. Moreover, the CH₄/H₂ selectivity coefficient decreased significantly with increasing total adsorption pressure. Liu et al. [14] investigated the purification of steam methane reforming (SMR) gas (H₂/CH₄/CO₂/CO₂/CO = 76/3.5/20/0.5, vol%) via a two-stage separation process, first using activated carbon to remove CO₂ and CO, then adopting 5A for H₂/CH₄ separation, which achieved a theoretical H₂ recovery of 85%. Shi et al. [10] studied CH₄ adsorption from the same SMR gas mixture with activated carbon, and the resultant H₂ recovery was only 71.16%. Liu [17] explored CH₄ adsorption from a multi-component gas mixture (H₂/CH₄/CO₂/N₂/CO = 65.71/8.63/3.25/18.98/2.73, vol%) using CAN zeolite, with the corresponding H₂ recovery rate reaching merely 65%. As shown in the above studies, LTA-type zeolites can balance high CH₄ adsorption capacity, low H₂ adsorption capacity, and fast gas adsorption rate. Moreover, their H₂/CH₄ separation efficiency can be further improved via structural optimization strategies, including cation modification and Si/Al ratio tuning.

Currently, some scholars have investigated CH₄/H₂ adsorption on LTA zeolites. Regarding the effect of cations, relevant studies have yielded important findings: Karki et al. [18] investigated Na-LTA zeolite and found that the exchange of Na⁺ with Sr²⁺ significantly enhanced H₂ adsorption capacity in the low-pressure region. Lao et al. [19] studied the influence of Na⁺ and Ca²⁺ adsorption on CH₄ adsorption of LTA molecular sieves (Si/Al = 1) under high pressure (10 MPa) via GCMC simulation, revealing that Ca²⁺ could remarkably increase CH₄ adsorption capacity. Tao et al. [20] demonstrated through molecular simulation that the charge induction of oxygen atoms in the zeolite framework is the key to enhancing CH₄ adsorption capacity. Compared with cations with high charge and small radius, cations with a low charge-to-radius ratio can shield the oxygen sites on the framework, reducing CH₄ adsorption capacity and thus improving the CO₂/CH₄ selectivity coefficient. In terms of Si/Al ratio research, Karki et al. [18] reported that increasing the Si/Al ratio remarkably enhances H₂ adsorption capacity. The Na-LTA zeolite with a Si/Al = 15 achieved a 40% increase in H₂ adsorption compared to 4A (Si/Al = 1) at 77 K. Xu et al. [21] observed distinct NMR signals between Si-O-Si and Si-O-Al oxygen sites, which indicates different electronic environments around these two types of sites; this difference further affects the role of oxygen-containing adsorption sites in gas adsorption and separation. The above studies indicate that the cation type and Si/Al ratio of the LTA structure significantly regulate H₂/CH₄ separation. However, existing research has focused on gas storage under high pressure or low temperature, CH₄/CO₂ separation, or separate characterization of electron clouds at framework adsorption sites. Few studies have investigated H₂/CH₄ competitive adsorption under industrial by-product gas conditions (298 K–318 K, total pressure 0–2 MPa), leaving unclear how LTA structural characteristics affect the adsorption distribution of H₂/CH₄ in pores.

Therefore, this study performs grand canonical Monte Carlo (GCMC) molecular simulations to track variations in H₂/CH₄ adsorption distribution within LTA pores under diverse adsorbent structures and temperature–pressure conditions. It also conducts molecular dynamics (MD) simulations to calculate the diffusion coefficients of H₂/CH₄ in the pores, thus analyzing the factors and mechanisms governing H₂/CH₄ adsorption and separation in LTA zeolites. Clarifying the competitive adsorption behaviors and intrinsic mechanisms of H₂/CH₄ on LTA zeolites will lay a solid theoretical foundation for developing high-performance adsorbents and optimizing gas mass transfer processes in the column.

2. Modeling Method

2.1. Construction of LTA Framework

Zeolite molecular sieves are composed of SiO₄ and AlO₄ tetrahedra connected alternately through oxygen bridges, adhering to the Lowenstein rule (Al atoms are not directly linked). The AlO₄ tetrahedra in the framework carry negative charges, which require cations for charge neutralization. These cations are mainly distributed at key positions of pore channels, such as 6-membered and 8-membered rings. In this study, a single-factor investigation was conducted focusing on cation types and Si/Al ratio in A-type zeolites. The Material Studio 2020 software was employed to explore the effects of cations (Na⁺, Li⁺, Ca²⁺, Mg²⁺) and Si/Al ratios (1~1.5) on the competitive adsorption of H₂/CH₄. The standard LTA unit cell from the International Zeolite Association (IZA) was expanded into a supercell ([Al₇₆₈Si₇₆₈O₃₀₇₂], Si/Al = 1), as presented in Figure 1. In subsequent studies, a self-developed script was used to replace Al atoms with Si atoms in the supercell, obtaining supercell structures with Si/Al ratios ranging from 1 to 1.5. After incorporating metal cations, the Grand Canonical Monte Carlo (GCMC) method was applied to calculate the adsorption and separation performance of H₂/CH₄ on different LTA structures.

To investigate the effect of cation types, a fixed Si/Al ratio of 1 was adopted, and four cations with distinct charges and ionic radii (Na⁺, Li⁺, Ca²⁺, Mg²⁺) were selected for comparative studies. Their ionic radii are 0.102 nm, 0.076 nm, 0.100 nm, and 0.072 nm, respectively. The models of these cation-exchanged LTA zeolites are shown in Figure 2. The Connolly Surface Parameters of the four cation-exchanged LTA zeolites are provided in Tables S1 and S2. A detailed description of their structural characteristics is available in Section S1.

When constructing models with different Si/Al ratios, the LTA supercell with Si/Al = 1 was used as the basic structure. In compliance with the Lowenstein rule, Al atoms in the framework were randomly replaced by Si atoms according to the target Si/Al ratios. Since Na⁺ is the base cation of zeolites, it was employed to balance the framework charge. Table S3 presents the calculated quantities of Si, Al, O, and Na⁺ atoms in the supercells with Si/Al ratios of 1.1, 1.25, and 1.5. Their structural molecular formulas are Na₇₃₁[(AlO₂)₇₃₁(SiO₂)₈₀₅], Na₆₈₂[(AlO₂)₆₈₂(SiO₂)₈₅₄], and Na₆₁₄[(AlO₂)₆₁₄(SiO₂)₉₂₂]. The corresponding models are illustrated in Figure S1. The Connolly Surface Parameters of LTA with different Si/Al ratios are provided in Tables S4 and S5, with a detailed description of structural characteristics available in Section S2.

2.2. GCMC Simulation Details

All GCMC simulations were classical. The Lennard-Jones (LJ) potential was employed to describe fluid–fluid and gas–wall interactions. The parameters used for calculating LJ interactions and Coulomb interactions are listed in Table 1. The calculations followed Lorentz–Berthelot mixing rules [22,23], as Equations (1) and (2) show. Here, ${\epsilon }_{\mathrm{ij}}$ represents the LJ energy parameter between component i and component j; ${\epsilon }_{\mathrm{ii}}$ and ${\epsilon }_{\mathrm{jj}}$ are the self-interaction LJ energy parameters of component i and j, respectively; ${\sigma }_{\mathrm{ij}}$ denotes the LJ size parameter between component i and j, while ${\sigma }_{\mathrm{ii}}$ and ${\sigma }_{\mathrm{jj}}$ are the self-interaction LJ size parameters of component i and j.

$${\epsilon }_{\mathrm{ij}}={\left({\epsilon }_{\mathrm{ii}}{\epsilon }_{\mathrm{jj}}\right)}^{1/2}$$

(1)

$${\sigma }_{\mathrm{ij}}=\left({\sigma }_{\mathrm{ii}}{\sigma }_{\mathrm{jj}}\right)/2$$

(2)

In GCMC simulations, chemical potential is treated as a fugacity function and calculated via the SRK (Soave–Redlich–Kwong) equation of state. The method was employed to simulate H₂/CH₄ adsorption in graphite slit pores, with calculations performed at 11 pressure-controlled pressure nodes. The COMPASS II force field was adopted; Colombic interactions were handled by the Ewald summation method, van der Waals interactions by the atom-based method, and the Lennard-Jones (LJ) potential cutoff radius was set to 1.25 nm. To reduce initial configuration errors and improve result accuracy, simulations consisted of two stages: equilibrium steps (for confirming adsorption equilibrium) and production steps (for calculating adsorption capacity via gas average adsorption amount). Each model in this study ran a maximum of 2 × 10⁷ steps (1 × 10⁷ equilibrium steps and 1 × 10⁷ production steps), with data from the latter 1 × 10⁷ steps used for statistical averaging. Adsorption capacity from molecular simulations was initially expressed as molecules per unit cell and converted to mol/kg for unit unification using Equation (3). The adsorption selectivity is calculated as in Equation (4). Here, for Equation (3), n denotes the adsorption capacity (in mol/kg); N is the number of adsorbed molecules per unit cell; N_A represents Avogadro’s constant; ${\mathrm{M}}_{\mathrm{zeolite}}$ is the molar mass of the zeolite unit cell. For Equation (4), ${\mathrm{S}}_{{\mathrm{CH}}_4/{\mathrm{H}}_2}$is the adsorption selectivity of CH₄ relative to H₂; ${\mathrm{X}}_{{\mathrm{CH}}_4}$ and ${\mathrm{X}}_{{\mathrm{H}}_2}$ are the mole fractions of adsorbed CH₄ and H₂ in the zeolite, respectively;${\mathrm{Y}}_{{\mathrm{CH}}_4}$ and ${\mathrm{Y}}_{{\mathrm{H}}_2}$ are the mole fractions of CH₄ and H₂ in the bulk gas phase, respectively.

$$\mathrm{n}={\displaystyle \frac{\mathrm{N}}{{\mathrm{N}}_{\mathrm{A}}{\mathrm{M}}_{\mathrm{zeolite}}}}$$

(3)

$${\mathrm{S}}_{{\mathrm{CH}}_4/{\mathrm{H}}_2}={\displaystyle \frac{\frac{{\mathrm{X}}_{{\mathrm{CH}}_4}}{{\mathrm{X}}_{H_2}}}{{\mathrm{Y}}_{{\mathrm{CH}}_4}/{\mathrm{Y}}_{{\mathrm{H}}_2}}}$$

(4)

Since molecular simulations output absolute adsorption capacity (N_abs), while experiments yield excess adsorption capacity (N_exc), conversion was performed via Equation (5). ${\rho }_{\mathrm{bulk}}$ (gas-phase adsorbate molar density) was obtained from the NIST chemical handbook. ${\mathrm{V}}_{\mathrm{pore}}$ (gas-phase accessible volume) was determined using Materials Studio’s Accessible Volume tool combined with probe atom radii [24]. Probe radii were set to gas molecules’ van der Waals radii (0.12 nm for H₂, 0.18 nm for CH₄) to match actual gas-phase molecular spatial occupancy in pores [25].

$${\mathrm{N}}_{\mathrm{exc}}={ \mathrm{N}}_{\mathrm{abs}}-{\rho }_{\mathrm{bulk}}{\mathrm{V}}_{\mathrm{pore}}$$

(5)

The model’s accuracy was verified by simulating single-component H₂ and CH₄ adsorption capacities at 298 K and pressures of 0–1 MPa, with details provided in Figure 3. A 50% Na⁺/50% Ca²⁺ co-doped LTA zeolite model was constructed in this study. For CH₄ adsorption, the simulated capacity at 298 K and 1 MPa was approximately 1.54 mol·kg⁻¹, representing a 5% r decrease relative to the experimental value of 1.4586 mol·kg⁻¹ measured at 298 K and 1.014 MPa. Moreover, the adsorption isotherm exhibited a trend consistent with the 0–1 MPa CH₄ isotherm profiles reported in the literature [7,13]. For H₂ adsorption, the simulated uptake at 298 K and 1 MPa (0.114 mol·kg⁻¹) was 3.5% lower than the literature-reported experimental H₂ capacity of~0.10 mol·kg⁻¹ for 5A zeolites, and 8.7% lower than the experimental value of 0.104 mol·kg⁻¹ obtained at 0.994 MPa. All these errors lie within the generally accepted reasonable range (±15%) for GCMC simulations in porous material adsorption research [26]. These deviations mainly result from the mismatch between the model’s 50% Na⁺/50% Ca²⁺ cation ratio and the actual cation distribution in commercial 5A zeolites [27], as well as minor variations in the framework Si/Al ratio caused by differences in raw materials and synthesis processes during practical production. Nevertheless, the adsorption capacity trends with increasing pressure were in complete agreement between simulations and experiments. Combined with the model’s Connolly Surface Parameters discussed in Sections S1 and S2, the established LTA model can effectively reflect the H₂/CH₄ adsorption behaviors of actual 5A zeolites.

Table 1. Lennard-Jones (LJ) force field parameters for H₂/CH₄ in LTA zeolite.

Fluid/Functional Group	Site	$\mathit{\sigma }$, Å	$\mathit{\epsilon }/{\mathit{k}}_{\mathit{B}}$, K	$\mathit{q}$, e	References
H2	H	2.96	36.7	0.468	[28]
CH4	C	3.40	55.05	−0.612	[29]
	H	2.65	7.9	0.153
Atoms of the LTA zeolite framework	Si	2.30	22.0	1.5	[30]
	O	3.30	53.0	−0.75
	Al	2.30	22.0	1.4
H2-LTA	H-Si	1.95	27.9	0.702	[30,31]
	H-Al	1.29	27.9	0.655
	H-O	1.593	50.1	−0.351
CH4-LTA	C-Si	3.675	57.7	+1.5	[30,31]
	C-O	3.345	109.1	−0.8
	C-Al	3.015	57.7	1.4
	H-Si	3.103	42.2	1.5
	H-O	2.768	82.0	−0.8
	H-Al	2.438	42.2	1.4

Table 1. Lennard-Jones (LJ) force field parameters for H₂/CH₄ in LTA zeolite.

Fluid/Functional Group	Site	$\mathit{\sigma }$, Å	$\mathit{\epsilon }/{\mathit{k}}_{\mathit{B}}$, K	$\mathit{q}$, e	References
H2	H	2.96	36.7	0.468	[28]
CH4	C	3.40	55.05	−0.612	[29]
	H	2.65	7.9	0.153
Atoms of the LTA zeolite framework	Si	2.30	22.0	1.5	[30]
	O	3.30	53.0	−0.75
	Al	2.30	22.0	1.4
H2-LTA	H-Si	1.95	27.9	0.702	[30,31]
	H-Al	1.29	27.9	0.655
	H-O	1.593	50.1	−0.351
CH4-LTA	C-Si	3.675	57.7	+1.5	[30,31]
	C-O	3.345	109.1	−0.8
	C-Al	3.015	57.7	1.4
	H-Si	3.103	42.2	1.5
	H-O	2.768	82.0	−0.8
	H-Al	2.438	42.2	1.4

Molecular dynamics (MD) is a simulation method for calculating a system’s equilibrium and transport properties. First, the equilibrated adsorbent–adsorbate H₂/CH₄ system from adsorption simulations was imported into the Forcite module, with H₂ and CH₄ designated as separate atomic groups to avoid interference from adsorbent framework atoms in diffusion behavior analysis. MD simulations were carried out under the NVT ensemble (constant atomic number N, volume V, and temperature T), with temperature maintained via the Nose thermostat method. The model was run with a 1 fs time step for a total duration of 300 ps, and one trajectory frame was output every 500 steps to generate a continuous dynamic trajectory file.

Next, the Mean Squared Displacement (MSD) of target gas molecules was calculated based on the formula: $\mathrm{MSD}\left(\mathrm{t}\right) = <{\left|{\mathrm{r}}_{\mathrm{i}}\left(\mathrm{t}\right)-{\mathrm{r}}_{\mathrm{i}}\left(0\right)\right|}^2>$ (where ${\mathrm{r}}_{\mathrm{i}}\left(\mathrm{t}\right)$ and ${\mathrm{r}}_{\mathrm{i}}\left(0\right)$ are the position vectors of the $i-th$ gas molecule at time t and the initial time, respectively, and <·> represents the ensemble average over all target molecules and time origins). The linearly increasing segment of the MSD–time curve was linearly fitted to obtain slope k. Per Einstein’s diffusion theory, the diffusion coefficient D was derived as $\mathrm{D} = \underset{\mathrm{t}\to \infty }{\mathit{lim}}{\displaystyle \frac{1}{6}}·{\displaystyle \frac{\mathrm{d}\left(\mathrm{MSD}\right)}{\mathrm{dt}}}$ [32,33]. Notably, a coefficient of determination (R²) greater than 0.995 is required for the linear fitting of MSD-t data points.

3. Results and Discussion

Calculating CH₄/H₂ selectivity (${\mathrm{S}}_{{\mathrm{CH}}_4/{\mathrm{H}}_2}$) and competitive adsorption capacities in adsorbent pores allow direct analysis of their competitive adsorption behaviors. Notably, gas competitive adsorption is a complex issue driven by temperature, pressure, and adsorbent structure. To elaborate, adsorption density map analysis clarifies the distribution ranges and adsorption strengths of gas molecules in pores and at adsorption sites. Calculating van der Waals and electrostatic forces during adsorption intuitively reveals interaction strengths between cations and framework adsorption sites. Finally, determining gas diffusion coefficients for different structures helps analyze diffusion rates post-structural modification, thereby enabling the investigation of H₂/CH₄ adsorption pathways in pores.

3.1. Synergistic Effect of Cation Charge Density and Radius

In zeolites, cations with distinct valences modulate the surface electronic properties and adsorption site stability of the framework, thereby regulating gas adsorption capacity. Meanwhile, cation radius fine-tunes the effective adsorption space within the channels, influencing the diffusion and accommodation of gas molecules [19,34,35], and consequently tailoring the adsorption and separation performance of H₂/CH₄. This section focuses on four cations commonly employed in zeolite adsorbents (Na⁺, Ca²⁺, Li^+, and Mg²⁺) to elucidate the effects of cation properties on the competitive adsorption and separation of H₂/CH₄, as well as their underlying mechanisms. The Shannon radii of these four cations in the zeolite framework are 0.102 nm, 0.100 nm, 0.076 nm, and 0.072 nm, respectively [36].

3.1.1. CH₄ Adsorption Evolution with Different Cations

As shown in Figure 4, the competitive adsorption capacity of CH₄ follows the order: Mg²⁺ (1.706 mol·kg⁻¹) > Ca²⁺ (1.535 mol·kg⁻¹) > Li⁺ (1.406 mol·kg⁻¹) > Na⁺ (1.328 mol·kg⁻¹), while the Shannon radii of the cations show the opposite trend: Na⁺ (0.102 nm) > Ca²⁺ (0.100 nm) > Li⁺ (0.076 nm) > Mg²⁺ (0.072 nm). These results demonstrate that Mg²⁺-LTA achieves the highest CH₄ adsorption capacity. This is attributed to Mg²⁺ having both a higher valence state and a smaller ionic radius, which reduces steric hindrance in zeolite pores and creates larger adsorption spaces. The Connolly Surface Parameters of Na⁺-LTA, Li⁺-LTA, Ca²⁺-LTA, and Mg²⁺-LTA listed in Table S1 further confirm this conclusion. Additionally, Joeong et al. [13], Daouli et al. [37], and Tao et al. [20] identified a positive correlation between CH₄ adsorption capacity and micropore volume and specific surface area via experimental measurements and DFT calculations.

These results demonstrate that high-valence cations in the zeolite framework promote CH₄ competitive adsorption, and for cations of the same valence, smaller Shannon radii correspond to more favorable CH₄ competitive adsorption amounts.

To clarify the mechanism of cation valence on CH₄ competitive adsorption, we analyzed the CH₄ adsorption distribution density in the 4MR, 6MR, and 8MR of zeolites (including α and β cages), as shown in Figure 5. Negligible CH₄ adsorption occurred in the α-cage 4MR of Na⁺-LTA and Li⁺-LTA, whereas stable CH₄ adsorption was detected at the junctions between α-cage 4MR and 8MR of Mg²⁺-LTA and Ca²⁺-LTA. These results show that high-valence cations broaden the CH₄ adsorption distribution range within zeolite pores.

As presented in

Table 2. CH₄ adsorption density and diffusion coefficient (different cations, 29 8 K, 2 MPa).

	Adsorption Density of CH4 (Molecules/Å3)		Diffusion Coefficient of CH4 (cm2/s)
	Maximum	Average
Na+-LTA	(6.0681 ± 0.0012) × 10−1	(1.1895 ± 0.0011) × 10−3	(1.5661 ± 0.0004) × 10−6
Li+-LTA	(6.7756 ± 0.0008) × 10−1	(1.2594 ± 0.0005) × 10−3	(1.6938 ± 0.0006) × 10−6
Ca2+-LTA	(9.2500 ± 0.0010) × 10−1	(1.5276 ± 0.0009) × 10−3	(2.6228 ± 0.0007) × 10−6
Mg2+-LTA	1.2516 ± 0.0007	(1.3750 ± 0.0004) × 10−3	(3.6474 ± 0.0005) × 10−6

, the maximum CH₄ adsorption densities of Mg²⁺-LTA and Ca²⁺-LTA reached 0.925 molecules/ų and 1.2516 molecules/ų, respectively, compared with 0.60681 molecules/ų and 0.67756 molecules/ų for Na⁺-LTA and Li⁺-LTA. This further confirms that divalent Mg²⁺ and Ca²⁺ induce stronger local electric fields in zeolite channels, extend the CH₄ adsorption interaction distance, and thus enhance pore CH₄ adsorption capacity. Additionally, CH₄ diffusion coefficients in divalent cation-exchanged zeolites were roughly twice those in monovalent cation-exchanged counterparts. This is because one divalent cation (Ca²⁺/Mg²⁺) replaces two monovalent cations (Na⁺/Li⁺), reducing total channel cation content, lowering spatial occupancy, increasing pore volume, and mitigating CH₄ diffusion resistance.

To investigate how cation radius affects CH₄ competitive adsorption, we compared CH₄ distributions in the channels of Na⁺-LTA vs. Li⁺-LTA and Mg²⁺-LTA vs. Ca²⁺-LTA zeolites. As shown in Figure 5, Li⁺-LTA and Mg²⁺-LTA (with smaller cation radii) have fewer CH₄-free regions in pores, with CH₄ more uniformly packed in channels. Table 2 data show the average CH₄ adsorption densities are 0.0011895 molecules/ų (Na⁺-LTA), 0.0012594 molecules/ų (Li⁺-LTA), 0.0015276 molecules/ų (Mg²⁺-LTA), and 0.001375 molecules/ų (Ca²⁺-LTA). These results indicate that for same-charge cations, a smaller ionic radius leads to more uniform CH₄ adsorption in pores, higher channel space utilization, and thus greater CH₄ adsorption capacity. CH₄ diffusion coefficients are as follows: 3.6474 × 10⁻⁶ cm²/s (Mg²⁺-LTA) > 2.6228 × 10⁻⁶ cm²/s (Ca²⁺-LTA) > 1.6938 × 10⁻⁶ cm²/s (Na⁺-LTA) > 1.5661 × 10⁻⁶ cm²/s (Li⁺-LTA), as Table 2 shows. This further confirms that a smaller cation radius reduces CH₄ diffusion resistance, facilitating diffusion into deeper microspores and enhancing CH₄ adsorption capacity.

These inferences are derived from the measured CH₄ adsorption densities and diffusion coefficients in different cation-modified zeolites, combined with intrinsic cation properties. Divalent cations (Mg²⁺, Ca²⁺) in zeolite channels may generate a stronger local electric field than monovalent ones (Na⁺, Li⁺), potentially inducing transient dipoles in CH₄ molecules and enhancing adsorption interactions. This may explain why divalent cation-modified zeolites show significantly higher CH₄ adsorption capacity than monovalent counterparts. Cation radius also regulates CH₄ adsorption by affecting the effective pore space and the adsorption site accessibility [19,38], with smaller radii correlating to higher adsorption capacity. Notably, CH₄ adsorption density is higher in the compact 4MR channels of small β-cages—likely owing to concentrated cation charges strengthening inductive interactions—whereas spacious 8MR channels disperse cations and weaken such interaction [39].

3.1.2. H₂ Adsorption Evolution with Different Cations

As shown in Figure 6, the competitive adsorption capacity of H₂ follows the order: Mg²⁺ (0.07815 mol·kg⁻¹) > Li⁺ (0.0779 mol·kg⁻¹) > Ca²⁺ (0.0720 mol·kg⁻¹) > Na⁺ (0.0704 mol·kg⁻¹), which is opposite to the increasing order of cation radius.

To analyze the effect of cation properties on H₂ competitive adsorption in zeolites, H₂ adsorption density distributions in Mg²⁺-LTA, Li⁺-LTA, Ca²⁺-LTA, and Na⁺-LTA were compared (Figure 7). In α-cage 4MR, Mg²⁺-LTA (the smallest cation radius) exhibited full H₂ channel filling, with adsorption density nearly matching that of adjacent 8MR. In contrast, H₂ adsorption in Na⁺-LTA was confined to regions near 8MR, while Li⁺-LTA and Ca²⁺-LTA showed no significant differences in H₂ adsorption density profiles. In 6MR pores, stable H₂ adsorption was observed in Mg²⁺-LTA and Li⁺-LTA, whereas Ca²⁺-LTA and Na⁺-LTA had markedly lower H₂ adsorption density in 6MRs. In β-cage 4MRs (Figure 7), H₂ distribution was highly restricted with extensive H₂-free areas, as this region serves as the primary CH₄ adsorption site (Figure 4).

Table 3. H₂ adsorption density and diffusion coefficient (different cations, 298 K, 2 MPa).

	Average Adsorption Density (Molecules/Å3)	Diffusion Coefficient (cm2/s)
Na+-LTA	(6.3100 ± 0.0005) × 10−5	(4.7939 ± 0.0006) × 10−5
Li+-LTA	(6.9750 ± 0.0013) × 10−5	(4.2885 ± 0.0005) × 10−4
Ca2+-LTA	(6.4500 ± 0.0008) × 10−5	(2.9366 ± 0.0007) × 10−4
Mg2+-LTA	(6.9980 ± 0.0006) × 10−5	(4.8540 ± 0.0004) × 10−4

lists the H₂ adsorption density and diffusion coefficient of zeolites with different cations. Synthesizing the above H₂ competitive adsorption phenomena, smaller cation radii (Li⁺, Mg²⁺) expand the H₂ adsorption range in α-cage 4MR and 6MR. For cations with similar radii, Mg²⁺ (higher valence) imparts zeolites with higher average H₂ adsorption density, diffusion coefficient, and adsorption capacity. These results indicate that framework cation radius is the primary factor governing H₂ adsorption, while high valence plays a secondary role. It can be inferred that H₂ has weak interactions within adsorbent pores and undergoes extensive free diffusion; smaller framework cations provide larger pore space, thereby directly enhancing H₂ adsorption capacity [17].

Figure 8 presents the CH₄/H₂ selectivity. At 298 K and 0–2 MPa, the ${\mathrm{S}}_{{\mathrm{CH}}_4/{\mathrm{H}}_2}$ of Na⁺-, Ca²⁺-, Li⁺- and Mg²⁺-LTA are 51.63–18.85, 41.94–21.32, 40.00–18.06 and 43.13–21.83, respectively. Li⁺-LTA exhibits the lowest overall ${\mathrm{S}}_{{\mathrm{CH}}_4/{\mathrm{H}}_2}$, since it enhances H₂ adsorption while suppressing CH₄ adsorption (consistent with previous results). Below 1 MPa, Na⁺-LTA shows significantly higher selectivity than the other zeolites; the S values are comparable at 1–1.6 MPa. At 2 MPa, the selectivity of Na⁺-LTA is slightly lower than that of divalent cation-exchanged zeolites (18 vs. 21).

For cation-regulated CH₄/H₂ adsorption, smaller cations enhance H₂ adsorption in α-cage 4MR (even 6MR adsorption) and confer advantages, with H₂ density of 0.001–0.01 (molecules/ų). In contrast, CH₄ accumulates densely at β-cage small windows via cationic electric fields and pore wall van der Waals forces; 8MR adsorbs CH₄ with much weaker interactions. Higher cation valence elevates CH₄ distribution and density in 8MR (0.1–1.0, molecules/ų), favoring high-valence cations for CH₄ adsorption. Na⁺-LTA achieves optimal CH₄/H₂ selectivity but low CH₄ adsorption capacity. Thus, doping Na⁺-LTA with Ca²⁺/Mg²⁺ is the optimal strategy for LTA zeolites, balancing high CH₄ capacity and selectivity in H₂/CH₄ separation.

3.2. Effect of Si/Al Ratio on Adsorption Sites

3.2.1. CH₄ Adsorption Evolution with Si/Al Ratios

As Figure 9 shows, the competitive adsorption capacity of CH₄ increases with Si/Al ratio (1 to 1.5): Si/Al = 1 (1.328 mol·kg⁻¹) < Si/Al = 1.1 (1.406 mol·kg⁻¹) < Si/Al = 1.25 (1.498 mol·kg⁻¹) < Si/Al = 1.5 (1.673 mol·kg⁻¹). Meanwhile, the competitive adsorption capacity of CH4 at Si/Al = ∞ is calculated to be 2.855 mol·kg⁻¹, as shown in Figure S2a. This indicates that a higher Si/Al ratio reduces the cation content and exposes more Si-O-Si groups, and this structural modification enhances the CH4 adsorption capacity.

As shown in Figure 10, the 4MR windows of β-cages are high-density CH₄ aggregation zones, retaining stable and dense CH₄ distribution as the Si/Al ratio increases from 1 to 1.5. In 8MR, CH₄ aggregates around framework oxygen atoms and cations; with the Si/Al ratio rising to 1.5, CH₄ adsorption density in 8MR increases progressively, especially near oxygen atoms. In α-cage 4MR, negligible CH₄ adsorption occurs at Si/Al = 1 and 1.1; when the ratio reaches 1.25 and 1.5, stable CH₄ adsorption emerges with expanded distribution areas. These observations indicate that reduced framework cation content enhances CH₄ distribution scope and aggregation density in 8MR and α-cage 4MR.

As shown in

Table 4. CH₄ competitive adsorption density values and diffusion coefficients at different Si/Al.

	Adsorption Density of CH4 (Molecules/Å3)		Diffusion Coefficient of CH4 (cm2/s)
	Maximum	Average
Si/Al = 1	(6.0681 ± 0.0010) × 10−1	(1.1895 ± 0.0005) × 10−3	(1.5661 ± 0.0009) × 10−6
Si/Al = 1.1	(7.5072 ± 0.0008) × 10−1	(1.2768 ± 0.0004) × 10−3	(1.8647 ± 0.0004) × 10−6
Si/Al = 1.25	(8.1268 ± 0.0005) × 10−1	(1.3467 ± 0.0003) × 10−3	(2.2012 ± 0.0004) × 10−6
Si/Al = 1.5	(8.9389 ± 0.0007) × 10−1	(1.4833 ± 0.0006) × 10−3	(3.3467 ± 0.0010) × 10−6

, at Si/Al ratios of 1, 1.1, 1.25, and 1.5, a higher ratio reduces cation content, which alleviates pore blockage, accelerates CH₄ diffusion, and allows gas molecules to penetrate deeper into pores—one reason for enhanced adsorption capacity [40,41]. The maximum adsorption densities further confirm that fewer cations at higher Si/Al ratios reduce pore “occupancy interference”, expand effective adsorption space, and thus improve CH₄ adsorption capacity.

As shown in Figure 11, the average van der Waals force of CH₄ adsorption rises from 650 kcal·mol⁻¹ to 775 kcal·mol⁻¹ as the Si/Al ratio increases from 1 to 1.5. Combined with the adsorption density evolution, the elevated Si/Al ratio exposes more Si–O–Si moieties in the framework, which form a continuous adsorption surface on pore walls to boost CH₄ adsorption density in pores [42]. Consistent with Ahn et al. [40], high Si/Al ratios enhance the covalent character of framework Si–O bonds, yielding uniform charge distribution on oxygen atoms and a stable adsorption environment. Thus, increased Si–O–Si bonds, reduced cation content, and unshielded oxygen sites maximize adsorption activity, which collectively accounts for the enhanced CH₄ adsorption performance at higher Si/Al ratios [21,43].

3.2.2. H₂ Adsorption Evolution with Si/Al Ratios

As shown in Figure 12, the competitive adsorption capacity of H₂ increases with the Si/Al ratio: Si/Al = 1 (0.705 mol·kg⁻¹) < Si/Al = 1.1 (0.0824 mol·kg⁻¹) < Si/Al = 1.25 (0.998 mol·kg⁻¹) < Si/Al = 1.5 (0.128 mol·kg⁻¹). Meanwhile, H₂ adsorption amount at Si/Al = ∞ is 0.3045 mol·kg⁻¹ (Figure S2b). This trend is consistent with that of CH₄; the pore structure changes caused by a higher Si/Al ratio enhance H₂ adsorption capacity. This is because at temperatures of 298 K and above, the adsorption isotherms of H₂ increase linearly; H₂ adsorption capacity rises continuously with increasing adsorption pressure and available adsorption space, exhibiting a “gas-like compression” behavior within the adsorbent pores [7,13]. For LTA zeolites, an increased Si/Al ratio reduces the number of framework cations, which alleviates pore blockage and increases pore volume, thereby enhancing H₂ adsorption capacity. As presented in the Connolly Surface Parameters in Table S4, the specific surface area and pore volume follow the order of Si/Al = 1 < Si/Al = 1.25 < Si/Al = 1.5, which provides sufficient space for H₂ adsorption.

To clarify the mechanism underlying H₂ adsorption capacity changes, Figure 13 shows that H₂ distributes uniformly across pores without distinct high-aggregation zones or specific adsorption sites; its sparse presence in β-cage 4MR stems from this region being a high-density CH₄ aggregation zone.

As seen in Figure 13a, H₂ aggregation density is markedly lower in the Si/Al = 1 structure than in other configurations. In α-cage 4MR (Figure 13b–d), H₂ gradually fills the pores as the Si/Al ratio increases from 1.1 to 1.5: at Si/Al = 1.1, H₂ adsorbs only near 8MR channels, while at Si/Al = 1.5, it fully occupies α-cage 4MR and connects adjacent 8MR channels.

Table 5. H₂ adsorption density and diffusion coefficients with different Si/Al ratios (298 K, 2 MPa).

	Diffusion Coefficient (cm2/s)	Average Adsorption Density (Molecules/Å3)	Maximum Adsorption Density (Molecules/Å3)
Si/Al = 1	(4.7939 ± 0.0011) × 10−5	(6.3121 ± 0.0008) × 10−5	(1.7189 ± 0.0004) × 10−2
Si/Al = 1.1	(1.8564 ± 0.0012) × 10−4	(6.3835 ± 0.0005) × 10−5	(1.8192 ± 0.0010) × 10−2
Si/Al = 1.25	(1.9634 ± 0.0013) × 10−4	(6.4501 ± 0.0004) × 10−5	(1.9791 ± 0.0006) × 10−2
Si/Al = 1.5	(4.3781 ± 0.0009) × 10−4	(6.5543 ± 0.0009) × 10−5	(2.1986 ± 0.0008) × 10−2

presents H₂ diffusion coefficients. The value at Si/Al = 1 is one order of magnitude lower than that of other structures, explaining the lower overall H₂ density in the Si/Al = 1 system (adsorption density diagrams). Moreover, a higher Si/Al ratio elevates H₂ diffusion coefficients, facilitating H₂ penetration into small deep-pore spaces (e.g., α-cages)—consistent with the adsorption trend in Figure 13. It also increases H₂’s average and maximum adsorption densities (Table 5). These results confirm two key effects: a higher Si/Al ratio expands H₂ spatial distribution and strengthens H₂–zeolite framework interactions. Thus, reduced cation content at higher Si/Al ratios exposes more framework oxygen adsorption sites, further enhancing H₂ adsorption.

Figure 14 shows CH₄/H₂ selectivity of LTA zeolites (Si/Al = 1, 1.1, 1.25, 1.5) over 0.2–2.0 MPa, revealing a distinct Si/Al-dependent trend: selectivity rises continuously as the Si/Al ratio increases from 1 to 1.25, peaking at Si/Al = 1.25 across the full pressure range. Further increasing the Si/Al ratio to 1.5 induces a slight decrease in selectivity, though the values remain higher than those for Si/Al = 1 and 1.1. Concurrently, all samples exhibit a gradual selectivity decline with increasing pressure, and the selectivity differences between samples narrow at low pressures. These results confirm that Si/Al = 1.25 is the optimal ratio for maximizing CH₄/H₂ selectivity in LTA zeolites.

A comprehensive analysis of H₂/CH₄ competitive adsorption and separation on LTA zeolites shows that CH₄ accumulates densely at β-cage 4MR windows and 8MR sites, while H₂ distributes uniformly in channels. Increasing the Si/Al ratio (1–1.5) reduces cations and exposes more Si–O–Si groups, elevating the adsorption capacities of both gases. CH₄/H₂ selectivity peaks at Si/Al = 1.25 but declines at 1.5, as the enhanced H₂ adsorption offsets CH₄’s advantage. Together, these results confirm that LTA zeolites with Si/Al ≈ 1.25 are optimal for H₂/CH₄ adsorptive separation.

3.3. Attenuating Effect of Temperature Elevation on CH₄ Adsorption

3.3.1. CH₄ Adsorption Evolution with Temperature

Comparison of CH₄ competitive adsorption capacities at different temperatures (Figure 15) shows the order: 1.3282 mol·kg⁻¹ (298 K) > 1.2377 mol·kg⁻¹ (308 K) > 1.1258 mol·kg⁻¹ (318 K). This trend indicates that temperature elevation inhibits CH₄ adsorption in pores. Subsequent analysis focuses on the evolution of CH₄ adsorption density in framework adsorption windows to explore the mechanism underlying the reduced CH₄ adsorption.

As shown in the CH₄ adsorption density distributions (Figure 16), CH₄ fully occupies β-cages, which serve as high-density aggregation zones in the framework. However, at 318 K, the area of red–yellow regions (indicating high-density aggregation) in the cages decreases. Additionally, the maximum CH₄ adsorption densities at 298 K, 308 K, and 318 K are 0.60681 molecules/ų, 0.58212 molecules/ų, and 0.53326 molecules/ų, respectively (

Table 6. CH₄ competitive adsorption density and diffusion coefficient at 298–318 K.

	Diffusion Coefficient (cm2/s)	Average Adsorption Density (Molecules/Å3)	Maximum Adsorption Density (Molecules/Å3)
298 K	(1.5804 ± 0.0011) × 10−6	(1.1895 ± 0.0006) × 10−3	(6.2123 ± 0.0015) × 10−1
308 K	(1.5692 ± 0.0009) × 10−6	(1.1085 ± 0.0005) × 10−3	(6.0681 ± 0.0008) × 10−1
318 K	(1.5661 ± 0.0012) × 10−6	(1.0083 ± 0.0005) × 10−3	(5.3326 ± 0.0006) × 10−1

), demonstrating that temperature elevation reduces CH₄ adsorption aggregation intensity in pores. In 8MRs, as the temperature rises from 298 K to 318 K, CH₄ aggregation near oxygen and cation sites shrinks significantly, with expanded adsorption-free areas. Correspondingly, the average CH₄ adsorption densities at these temperatures are 0.0011895 molecules/ų, 0.0011085 molecules/ų, and 0.0010083 molecules/ų, respectively (Table 6). These observations confirm that temperature elevation decreases the overall CH₄ distribution density in the framework. Based on the above, increasing temperature weakens interactions between CH₄ and framework oxygen sites/Na⁺, thereby reducing CH4’s competitive adsorption capacity.

Furthermore, within 298–318 K, temperature elevation has a negligible effect on the CH₄ diffusion coefficient, only slightly increasing its diffusion rate (Table 6). Lopes et al. [7] similarly reported minimal temperature impact on CH₄ diffusion coefficients between 303 and 323 K. This is likely due to the ordered zeolite pore arrangement, which enables unobstructed gas molecular movement; thus, small temperature fluctuations barely affect molecular motion. Combined with the aforementioned CH₄ adsorption density evolution in pores, temperature elevation primarily reduces CH₄’s thermodynamic equilibrium adsorption capacity by weakening its interactions with framework adsorption sites, thereby decreasing overall CH₄ adsorption capacity.

3.3.2. H₂ Adsorption Evolution with Temperature

As shown in Figure 17, the maximum H₂ adsorption capacity follows the order: 0.07047 mol·kg⁻¹ (298 K) < 0.07176 mol·kg⁻¹ (308 K) < 0.07242 mol·kg⁻¹ (318 K). The data indicate that temperature elevation slightly increases H₂ competitive adsorption capacity within 298–318 K. Subsequent analysis focuses on H₂ density distribution and diffusion in pores to clarify the mechanism for the increased adsorption capacity.

As shown in Figure 18, at 298–318 K, H₂ fills the adsorbent pores uniformly with no distinct high-density zones. In 8MR channels, H₂ fully occupies pores, with negligible adsorption density variation across temperatures. In β-cage 4MRs, H2 distributes randomly at 298 K but expands its adsorption area at 318 K. Combined with β-cages being high-density CH₄ adsorption zones—where high temperatures reduce CH₄ aggregation—high temperatures induce CH₄ desorption in H₂/CH₄ competitive adsorption, freeing space for H₂. In α-cage 4MRs, H₂ localizes near 8MRs at 298 K but fully occupies α-cages at 318 K. Thus, high temperatures facilitate H₂ diffusion into deep pores inaccessible at lower temperatures.

Table 7. H₂ competitive adsorption density and diffusion coefficients at 298–318 K.

	Diffusion Coefficient (cm2/s)	Average Adsorption Density (Molecules/Å3)	Maximum Adsorption Density (Molecules/Å3)
298 K	(6.7491 ± 0.0015) × 10−5	(6.3114 ± 0.0004) × 10−5	(1.7189 ± 0.0008) × 10−2
308 K	(5.2229 ± 0.0008) × 10−5	(6.3461 ± 0.0006) × 10−5	(1.7659 ± 0.0010) × 10−2
318 K	(4.7939 ± 0.0010) × 10−5	(6.4173 ± 0.0003) × 10−5	(1.8433 ± 0.0014) × 10−2

shows that the maximum and average H₂ adsorption densities in the framework are consistent with the evolution trend of adsorption density distributions, directly confirming that high temperatures induce CH₄ desorption to provide space and adsorption sites for H₂. Furthermore, the H₂ diffusion coefficients at 298 K, 308 K, and 318 K are 4.7939 × 10⁻⁵ cm²/s, 5.2229 × 10⁻⁵ cm²/s, and 6.7491 × 10⁻⁵ cm²/s, respectively; this directly verifies that high temperatures enhance H₂ kinetic energy, facilitating deeper pore diffusion and thus increasing adsorption capacity.

Figure 19 illustrates the selectivity coefficient of CH₄/H₂. At 298–318 K and 2 MPa total pressure, the values are 51–18, 42–15, and 38–13, respectively, indicating that temperature elevation significantly reduces CH₄/H₂ selectivity. As reported earlier, high temperatures induce CH₄ desorption and enhance H₂ adsorption; this phenomenon determines the selectivity decrease with increasing temperature.

A comprehensive analysis of H₂/CH₄ competitive adsorption in framework pores shows that elevated temperatures weaken CH₄’s interactions with adsorption sites (cations and oxygen atoms) during separation, reducing its adsorption capacity and freeing up additional pore space and active sites. Meanwhile, H₂—with its low adsorption enthalpy and high molecular kinetic energy—can diffuse deeper into pores and interact with available sites, thereby increasing its adsorption capacity. This aligns with reports by Cavenati et al. [44] and Choi et al. [45] who noted that weakly adsorbing species (e.g., H₂) exhibit enhanced high-temperature adsorption due to the desorption of strongly adsorbed CH₄, which provides the necessary sites and space for H₂ uptake.

3.4. Microspore Filling of Hydrogen Under Pressure

As indicated by the previous CH₄/H₂ selectivity coefficients, increasing pressure degrades H₂/CH₄ adsorption and separation performance. Thus, this section investigates the effect of pressure on their competitive adsorption in the framework, with H₂/CH₄ partial pressure ratios of 3:7–9:1 and total pressure of 0.2–2 MPa.

3.4.1. CH₄ Adsorption Evolution with Pressure

As shown in Figure 20, with total pressure increasing from 0.6 to 1, 1.6, and 2 MPa, the maximum CH₄ adsorption capacity increases as 0.885 → 1.121 → 1.466 → 1.576 mol/kg. As the H₂/CH₄ ratio rises from 9:1 to 3:7, increasing CH₄ partial pressure reduces the adsorption isotherm slope and slows CH₄ adsorption capacity growth. Some scholars have also reported that CH₄ adsorption slows and gradually saturates in the high-pressure region [13,45]; the effect of pressure on the evolution of CH₄ adsorption distribution in pores is analyzed in detail below.

The effect of feed gas ratio and pressure on CH₄ competitive adsorption was investigated by analyzing the evolution of CH₄ adsorption density distributions. As shown in Figure 21, β-cage 4MRs are preferential CH₄ adsorption regions: at 0.2 MPa, CH₄ adsorption only occurs in β-cage 4MRs, and as pressure rises to 2 MPa, the adsorption density in β-cages increases with expanded distribution. 8MRs are also CH₄ adsorption regions but with lower aggregation intensity than β-cage 4MRs; stable CH₄ aggregation in 8MRs only appears at 0.6 MPa. With increasing total pressure and CH₄ ratio, the CH₄ adsorption density in 8MRs gradually increases until full channel occupancy. The ranges of 0.2–1 MPa total pressure and 10–50% CH₄ partial pressure correspond to the rapid growth stage of CH₄ adsorption density, where the adsorption density in both 8MRs and β-cages increases sharply. CH₄ adsorption properties remain stable at 1–2 MPa total pressure and 50–70% CH₄ partial pressure.

To further quantify the effects of partial and total pressures on CH₄ adsorption density, data from Supplementary Tables S6 and S7 were analyzed. At a total pressure of 0.2 MPa, CH₄ adsorption density increased by 130.81%, 36.44% and 16.65% as the CH₄ mole fraction rose from 10% to 30%, 50% and 70%, respectively; at 0.6 MPa, 1 MPa and 2 MPa, the increments over the same mole fraction range were 77% ± 10%, 26% ± 1% and 17% ± 10%. With the CH₄ mole fraction fixed at 10%, increasing the total pressure from 0.2 MPa to 0.6 MPa, 1 MPa and 2 MPa enhanced adsorption density by 132.82%, 32.39% and 45.89%; for mole fractions > 30%, the growth rates at these total pressures were 81% ± 8%, 26% ± 1% and 37% ± 1%.

These results indicate that in the low-pressure region (CH₄ mole fraction < 30% or low total pressure), increasing CH₄ pressure (via higher mole fraction or total pressure) rapidly elevates adsorption capacity. However, when the CH₄ partial pressure exceeds 50% or the total pressure reaches 1 MPa, the growth rate of adsorption capacity slows significantly.

3.4.2. H₂ Adsorption Evolution with Pressure

H₂ adsorption capacity was analyzed (Figure 22). With total pressure increasing from 0.6 MPa to 1 MPa, 1.6 MPa, and 2 MPa, the maximum competitive H₂ adsorption capacities are 0.0459 mol/kg, 0.0759 mol/kg, 0.117 mol/kg, and 0.147 mol/kg, respectively. Moreover, the shapes indicate a linear increase in H₂ adsorption capacity with rising H₂ partial pressure.

The evolution of H₂ adsorption density in pores was analyzed (Figure 23). H₂ has no preferential adsorption sites and distributes uniformly across all channels, with its adsorption density at different framework positions increasing with pressure. Notably, β-cages exhibit higher H₂ adsorption density only in the high-pressure region (H₂ mole fraction > 70% or total pressure > 1 MPa), since these regions are high-density CH₄ aggregation zones that limit H₂ adsorption space and sites.

Additionally, H₂ adsorption density in pores is extremely low at 0.2 MPa (low pressure). At 0.6 MPa, H₂ adsorbs uniformly at low density in the framework’s 8MR and β-cage 4MR. As pressure further rises to 1 MPa and 2 MPa, H₂ adsorption density in 8MR and β-cages increases significantly, with stable H₂ adsorption also emerging in α-cages. Tables S8 and S9 quantify the effects of total and partial pressures on H₂ adsorption density: increasing total pressure from 0.2 MPa to 0.6 MPa, 1 MPa and 2 MPa yields growth rates of 180% ± 5%, 62% ± 3% and 88% ± 5%, respectively; raising H₂ mole fraction from 30% to 50%, 70% and 90% corresponds to growth rates of 71% ± 3%, 44% ± 3% and 37% ± 3%.

These results indicate that with increasing adsorption pressure, H₂ distributes over a far wider range in the framework than CH₄, and the growth rate of H₂ adsorption capacity is significantly higher than that of CH₄ as total pressure increases.

The effect of partial pressure on the CH₄/H₂ selectivity coefficient was calculated (Figure 24). The data show that increasing total pressure reduces this coefficient, consistent with the aforementioned changes in competitive gas adsorption density.

Evolution of H₂/CH₄ competitive adsorption with pressure confirms that pressure elevation promotes H₂ adsorption throughout the framework owing to its lack of preferential adsorption sites and uniform pore distribution. In contrast, CH₄ is confined to β-cages and 8MR channels, which causes H₂ adsorption density to grow faster than that of CH₄ and thus reduces CH₄/H₂ selectivity.

4. Conclusions

This study systematically investigates the competitive adsorption of H₂/CH₄ in LTA zeolites. By analyzing H₂/CH₄ adsorption density evolution in pores, the main findings are as follows:

(1)

CH₄ adsorbs preferentially in β-cage windows and eight-membered rings, forming high-density zones near cations and oxygen sites, whereas H₂ distributes uniformly with no distinct adsorption sites.

(2)

High-valence, small-radius cations (e.g., Mg²⁺, Ca²⁺) enhance CH₄ adsorption and diffusion by generating stronger local electric fields and reducing pore blockage compared to monovalent cations. H₂ is less sensitive to cation charge but benefits from smaller cation radii, which increase accessible pore volume and thus H₂ uptake.

(3)

Increasing the Si/Al ratio improves CH₄ selectivity by reducing cation content and exposing more framework oxygen sites, especially in Si–O–Si environments, which strengthens CH₄–framework interactions.

(4)

Temperature elevation weakens CH₄ adsorption but promotes H₂ diffusion into deeper pores. Higher pressure raises both uptakes, but H₂ adsorption increases more, reducing CH₄/H₂ selectivity.