The role of adsorption and diffusion in improving the selectivity and reactivity of zeolite catalysts

Abstract

This review provides a comprehensive overview of the fundamental principles, characterization techniques, and recent advances in understanding molecular adsorption and diffusion behaviors within zeolite materials. By examining the distinctive microporous frameworks, tunable pore sizes, and adjustable acid site distributions of zeolites, we highlight how adsorption and diffusion processes critically govern catalytic activity and selectivity. We discuss state-of-the-art experimental approaches alongside multi-scale computational methods, which collectively shed light on the molecular-level transport dynamics, interaction mechanisms, and energy barriers within zeolite channels. Focusing on exemplary topologies, we detail their performance and mechanistic insights in key applications including hydrocarbon adsorption, catalytic cracking, methanol conversion, and molecular separation. We further explore how tuning the Si/Al ratio, incorporating metal ions, engineering hierarchical pore structures, and regulating acid site distributions can synergistically optimize adsorption and diffusion, thereby enhancing catalytic efficiency and selectivity. These advancements pave the way for precise molecular-level control over transport phenomena and reaction pathways, underpinning the development of sustainable zeolite-based catalysts for clean energy, chemical process, and environmental applications.

---

Daoning Wu

Daoning Wu received his bachelor's degree from the Department of Chemistry at Heilongjiang University and conducted eletrocatalysis research in Professor Honggang Fu's group. He is currently a PhD student at the School of Materials Science and Engineering, Nankai University, in Professor Weili Dai's group. In 2023, he joined Professor Guangjin Hou's group for a joint PhD training program at the Dalian Institute of Chemical Physics, Chinese Academy of Sciences. His research focuses on the synthesis of zeolite catalysts, with emphasis on elucidating structural features and mechanistic pathways through advanced solid-state NMR spectroscopy.

Mingfeng Li

Mingfeng Li received his PhD in Chemical technology from the SINOPEC Research Institute of Petroleum Processing Co., Ltd (RIPP) in 2001. He is a professorate senior engineer, doctoral supervisor, member of the Chinese Society of Chemical Industry, president of SINOPEC RIPP, and director of “Special Committee on Hydrocarbon Resource Evaluation, Processing and Utilization” of the Chinese Society of Chemical Industry and “Special Committee on Carbon Neutrality” of the Chinese Society of Petroleum. His research focuses on the fields of gasoline and diesel quality improvement, chemical recycling of waste plastics, hydrogen energy production and utilization, and dual-carbon accounting in the petrochemical industry.

Guangjin Hou

Guangjin Hou is a Professor at the Dalian Institute of Chemical Physics (DICP), Chinese Academy of Sciences (CAS), heading the solid-state NMR research group. He received his PhD in 2007 from the National Centre for Magnetic Resonance in Wuhan, CAS. He then undertook postdoctoral research at the Max Planck Institute for Polymer Research in Germany and the University of Delaware in USA. Dr Hou's research focuses on the development of advanced solid-state NMR methodologies and in situ dynamic characterization techniques and investigation of microscopic structures, host–guest interactions, and reaction mechanisms of solid materials relevant to heterogenous catalysis, energy storage, etc.

Michael Hunger

Michael Hunger studied physics at the University of Leipzig and gained his doctorate in 1984 and his habilitation in 1992 at the Faculty of Physics. During this time, he investigated solid acids, in particular zeolites, by utilizing modern techniques of solid-state NMR spectroscopy. In 1992, after research stays in Cambridge and Manchester, he moved to the University of Stuttgart and was appointed Professor for Chemical Technology at the Faculty of Chemistry. In Stuttgart, his research focused on the development of novel methods for the characterization of micro- and mesoporous solid catalysts and for in situ spectroscopy of heterogeneously catalyzed reactions.

Weili Dai

Weili Dai is a Professor at the School of Materials Science and Engineering & National Institute for Advanced Materials, Nankai University, mainly engaged in research on the synthesis of zeolites and their catalytic applications. He has developed various in situ spectroscopic coupling techniques to enable real-time monitoring of zeolite synthesis and catalytic reaction processes, elucidated the mechanisms of zeolite crystallization and catalysis, and guided the design of a range of highly efficient zeolite catalysts.

---

1. Introduction

Zeolite materials represent a versatile class of porous solids, distinguished by their highly ordered microporous frameworks, large specific surface areas, adjustable pore structures and surface acidity.^1–3 The well-defined channel architectures offer significant potential for diverse catalytic applications. The adsorption and diffusion behaviors within zeolitic pores play a pivotal role in determining the efficiency of reactant molecules accessing the active sites, thereby fundamentally influencing catalytic activity and selectivity.^4–6 Gaining a comprehensive understanding of these processes is crucial for the rational design and optimization of zeolite-based catalytic systems, enabling enhanced reaction performance and precise control over selectivity.

1.1 Basic principles of adsorption

The adsorption behavior of zeolites is governed by their well-defined pore sizes, geometric structures, tunable surface chemistry, and high-density active centers.⁵ As the initial step in catalytic reactions, adsorption plays a crucial role in the enrichment and activation of reactants. It profoundly influences reaction pathway selectivity by enabling precise capture, targeted adsorption, and activation of specific molecules, thereby creating an optimal microenvironment for complex catalytic processes.^7–10

The adsorption mechanism in zeolite materials can be classified into physical adsorption and chemisorption. Physical adsorption, driven by weak intermolecular interactions, is typically reversible and benefits from the high surface area and ordered porosity of zeolites, which enhance molecular interactions and increase adsorption capacities. This mechanism is prevalent for small gaseous molecules or nonpolar substances, which do not engage in chemical reactions but instead interact with the zeolite surface through weak interactions.^11,12 In contrast, chemisorption involves a chemical interaction between the zeolite surface and the adsorbate, resulting in the formation of covalent or partially covalent bonds with active sites on the catalyst surface. Compared to physisorption, chemisorption generally exhibits significantly stronger interactions, characterized by pronounced electron redistribution and localized polarization within the adsorbed species. These effects effectively lower the activation barriers required for subsequent catalytic transformations. It is crucial to clarify that although chemisorption interactions are typically stronger and less readily reversible than those involved in physisorption, they are not inherently irreversible. Under suitable conditions, chemisorption processes can be reversible, as illustrated by the interaction between methanol and Brønsted acid sites (BAS) to form methoxy species and water, a reaction known to reverse upon the addition of water. Additionally, chemisorption is fundamental to various catalytic reactions, such as hydrocarbon catalytic cracking and aromatic isomerization, where it mediates selective cleavage of C–H or C–C bonds, thereby facilitating essential molecular rearrangements integral to catalytic activity. As a result, it is regarded as a pre-activation step in the reaction. In the initial stages, physisorption enables the rapid accumulation of reactant molecules on the catalyst surface or within its pores, creating localized high-concentration regions. This molecular enrichment effect increases the collision frequency between reactants and active sites and supplies ample precursors for chemisorption. Altogether, these processes synergistically establish a hierarchical, stepwise reaction pre-activation mechanism.^13–19 Desorption refers to the process in which adsorbed molecules detach from the zeolite surface or channels after completing the catalytic reaction cycle as the reverse of absorption.¹³ As product molecules desorb from the catalyst surface and exit the zeolite channels, desorption provides fresh active sites for subsequent reactions. The rate of desorption can be optimized through external control measures, such as adjusting temperature or pressure, to ensure efficient product release without hindering the reactant adsorption.²⁰

Adsorption selectivity in zeolites is determined by multiple factors. The pore size of zeolites strictly dictates the molecular dimensions that can access their channels. The determination of molecular diameters, kinetic diameters, and access to pores was thereby barely unified, thus frequently opposing values of unknown origin are extracted from the literature. A recent unification by Gugeler et al. determines molecular diameters on a computational basis and integrates experimental data to empirically calculate their accessibility to pores of specific sizes.²¹ Only reactant molecules with diameters smaller than the zeolite pores can enter the channels and reach the active sites, while larger molecules are effectively excluded.^16,22 This size-selective mechanism ensures preferential adsorption and reaction of specific molecules. Beyond size, the shape of molecules influences their adsorption and reaction within zeolites, i.e. the shape selectivity.^23,24 Only reactant molecules with appropriate shapes and sizes can penetrate the zeolite channels and interact with active sites (reactant shape selectivity). During the reaction process, only product molecules of specific shapes and sizes can readily diffuse out of the channels. Those that are unsuitable may become trapped and undergo further transformations (product shape selectivity). Additionally, the channels of zeolites also constrain the formation of reaction intermediates or transition states, permitting only specific reaction pathways and thereby enhancing the reaction selectivity (restricted transition state shape selectivity). Furthermore, adsorption selectivity can be influenced by regulating the chemical composition and cation types of zeolites, adjusting the distribution and strength of acid sites, and modifying surface properties through Si/Al ratio modulation.^25–27 These aspects will be discussed in detail in Section 4 of this review.

1.2 Basic principles of diffusion

The diffusion behavior of molecules within the pore channels of molecular sieves plays a critical role in determining overall performance. The diffusion process directly influences the efficient transport of reactants to active sites, the timely removal of products, and consequently affects the catalyst activity, selectivity, and stability. In catalytic reactions, reactant molecules must diffuse into the zeolite pores to reach active sites for reaction, while the generated products need to diffuse out of the channels. If the diffusion rate is slow, reactants may not adequately reach the active sites, and accumulation of products within the pores can lead to channel blockage, reducing the catalysis efficiency and lifespan.²⁸ Therefore, understanding and controlling molecular diffusion within molecular sieves are essential for enhancing the efficiency of catalytic processes.

Diffusion in zeolites can primarily be categorized into the following types.

(1) Fickian diffusion:^29,30 when the pore size significantly exceeds the mean free path of molecules, the collisions between molecules dominate the diffusion process, and the diffusion can be described by Fick's first law:

J = −D_τ∇C (1.1)

where J stands for the diffusive flux, D_τ for the diffusion coefficient, and ∇C for the concentration gradient. During catalytic cracking, the transport of reactants and products primarily proceeds via molecular diffusion within the macroporous or mesoporous pores of zeolites, where the pore dimensions substantially exceed the molecular mean free path, resulting in a diffusion regime that conforms to Fickian behavior.^31,32

(2) Knudsen diffusion: this type of diffusion occurs when the pore size is comparable to or smaller than the mean free path of the molecules.³³ The collisions between gas molecules and the pore walls dominate over intermolecular collisions. The diffusion coefficient (D_K) for Knudsen diffusion is given by:

In eqn (1.2), d represents the pore diameter, R is the gas constant, T stands for the temperature, and M signifies the molecular weight of the gas.

(3) Surface diffusion: it involves the migration of molecules adsorbed on the pore wall surfaces along the surface. This type of diffusion is particularly significant under conditions of low concentration or strong adsorption.³⁴ The migration of adsorbate molecules along adjacent active sites on the pore walls (i.e., inner surface) proceeds via a hopping mechanism, which at the microscopic level can be viewed as a continuous cycle of adsorption–desorption events. The surface diffusion flux (J) is described by:

J = −D_s∇θ (1.3)

In eqn (1.3), D_s denotes the surface diffusion coefficient, and θ signifies the surface coverage. During the removal of organic pollutants, surface diffusion markedly influences the mass transfer rate. In surface diffusion, the diffusion coefficient can often be described by the Arrhenius equation, where increasing temperature enhances the ability of adsorbed particles to overcome potential energy barriers, thereby facilitating their migration. However, unlike Fickian diffusion, surface diffusion is also governed by adsorption behavior. For typical exothermic adsorption systems, higher temperatures tend to reduce the surface coverage, which in turn decreases the number of mobile adsorbed species. As a result, although thermal activation promotes diffusivity, the concurrent decline in adsorbate concentration may counterbalance or even outweigh this effect, leading to a non-monotonic dependence of the surface diffusion coefficient on temperature.²⁸

(4) Configurational diffusion: in microporous materials where the molecular dimensions are comparable to the pore sizes, molecular mobility is spatially restricted, requiring the surmounting of significant energy barriers. For instance, during the isomerization of aromatic hydrocarbons, the diffusion of molecules within ZSM-5 zeolites is governed by configurational diffusion.³⁵

Given that each diffusion mechanism represents distinct molecular transport behaviors within molecular sieves, a further analysis of the kinetic and thermodynamic properties of diffusion coefficients will aid in quantifying the impact of these mechanisms. Based on the nature of the diffusion process in catalysis, diffusion coefficients are categorized into self-diffusion coefficients, transport diffusion coefficients, and corrected diffusion coefficients.^36,37 The self-diffusion coefficient describes the random motion of individual molecules in the absence of a concentration gradient reflecting the intrinsic mobility of the molecules. The corrected diffusion coefficient accounts for the diffusion behavior of molecules within the zeolite channels, adjusting for the effects of channel structure, adsorption, and other factors on diffusion.

In multicomponent systems diffusing through nanoporous zeolites, nonlinear interactions and intense competitions among components often occur. These complexities render the Fick diffusion coefficient inherently inadequate for accurately describing the actual diffusion behavior. The Maxwell–Stefan theory, however, provides a more fundamental framework for describing diffusion in such systems by treating it as a process driven by chemical potential gradients, wherein components interact through mutual frictions and coupled motions.^38–41 The Maxwell–Stefan equation is expressed as follows:

where μ_i is the chemical potential of component i, R is the gas constant, T is the absolute temperature, x_i is the mole fraction of component i, J_i is the diffusion flux vector of component i, and D_ij is the Maxwell–Stefan diffusion coefficient. This theoretical framework offers a more intrinsic approach to understand diffusion in multicomponent, non-ideal systems.

When the pore dimensions closely match the molecular sizes, molecules are unable to pass each other within the channels, rendering the classical assumptions of Fickian diffusion invalid. For example, in single-file diffusion, the restricted motion leads to a sublinear temporal evolution of the mean-square displacement, a clear deviation from the linear time dependence predicted by Fick's law.^42,43

Beyond single-file diffusion, a suite of anomalous transport phenomena have emerged under specific pore architectures and interaction regimes. Incommensurate diffusion arises when the periodicity of the pore network is mismatched with the molecular dimensions, leading to resonance effects that modulate the effective diffusion coefficient.⁴⁴ Similarly, levitation effects occur when the attractive interactions between the molecule and the pore wall reach an optimal balance, effectively reducing friction and promoting unexpectedly rapid transport.⁴⁵ The concept of molecular traffic captures the idea that high-density conditions within narrow channels can lead to congestion effects analogous to vehicular traffic, dramatically influencing net flux.⁴⁶

Due to the inhomogeneity of the pore structure, the diffusion coefficient often changes with the variation of position and concentration. Therefore, in order to describe the molecular diffusion process in porous media more accurately, it is usually necessary to extend Fick's second law and introduce the influence of changes in local diffusion coefficients. A common extension is the following equation:

Eqn (1.5) consists of two parts. The first term represents the traditional Fick's second law, which describes the relationship between molecular diffusion and the concentration gradient, where D_T is the transport diffusion coefficient, representing the rate of molecular diffusion within the material. The second term accounts for the influence of the concentration gradient on the local diffusion coefficient, reflecting how the diffusion coefficient may vary in regions of high concentration gradient. This term is particularly suitable for describing diffusion in porous materials with structural inhomogeneities or local barriers, as it accurately captures the diffusion behavior resulting from the heterogeneity of pore structures.^47,48

1.3 Adsorption and diffusion in zeolite catalysis

As above mentioned, the adsorption process determines the rate and efficiency with which reactant molecules access the zeolite channels, while the diffusion process governs their movement within these channels to reach active sites and facilitates the desorption of products from the catalyst surface. In practical catalytic systems, the characteristics of adsorption and diffusion are often interdependent and intertwined, directly influencing the overall efficiency of the catalytic reaction.

Once adsorption occurs, the reactant molecules must diffuse through the channels to reach the active sites where the reaction takes place, and the resulting products must subsequently diffuse out for desorption. Excessively strong adsorption of reactants on catalytic surfaces can significantly impede molecular transport phenomena, potentially resulting in the confinement of reactive species at active sites or within the internal pore structure. Under such conditions, mass transfer limitations become the primary kinetic bottleneck, limiting the overall reaction rate. The complexity of the pore architectures and confined spatial dimensions can impose substantial diffusional constraints, leading to pronounced mass transport restrictions that ultimately affect reaction kinetics and catalytic efficiency. By establishing reaction–diffusion coupling equations, the concentration distribution of reactants and the reaction rate within the catalyst can be described. The commonly used Thiele modulus (ϕ) and the Weisz–Prater criterion are employed to quantitatively assess the impact of internal mass transfer limitations.^49–51 When the ϕ value is large, molecular diffusion within the catalyst becomes the primary bottleneck restricting the reaction rate. In such cases, enhancing pore connectivity and introducing hierarchical pore structures are effective strategies to improve mass transfer efficiency. Conversely, when the ϕ value is small, the reaction rate is predominantly governed by catalytic activity, making the optimization of surface adsorption properties more critical.

In earlier reviews, Rey et al. summarized the synthesis methods and modification techniques of zeolites and their applications in adsorption and ion exchange.⁵ In 2008, Smit and Maesen provided a detailed theoretical summary of the advancements in molecular simulation techniques applied to adsorption, diffusion, and selectivity in molecular sieves. They highlighted adsorption mechanisms, diffusion behaviors, and various mechanisms of shape selectivity.⁴ Pérez-Ramírez et al. conducted an in-depth analysis of the impact of zeolite structural optimization on catalytic reactions, highlighting two key strategies, i.e., the enlargement of micropore sizes and the incorporation of hierarchical pore architectures. They thoroughly evaluated these approaches focusing on the synthesis methodologies and respective advantages in enhancing catalytic performance.²³ Sastre et al. focused on the properties of small-pore silica zeolites and their molecular sieving effects in hydrocarbon separation, while systematically reviewing studies on the structural, mechanical, and dynamic properties of pure silica zeolites based on classical force field methods. The importance of molecular simulations in evaluating the framework flexibility of zeolites and its impact on diffusion dynamics were emphasized, providing theoretical support for optimizing the separation performance.⁵²

Over the past several decades, the adsorption and diffusion properties of zeolites have been extensively summarized from distinct perspectives. However, with the continuous evolution of advanced characterization techniques and catalytic performance evaluation strategies, a holistic integration of zeolite microstructure, molecular transport dynamics, and catalytic activity is crucial for elucidating the complex transport-reaction interplay within zeolites. In this review, we provide a thorough overview of the mechanisms underlying adsorption and diffusion, alongside a theoretical summary of widely used advanced characterization methods (Fig. 1). These techniques are contextualized through their application in practical catalytic reaction systems to offer deeper insights. Zeolites mentioned in this review are categorized according to their distinct topological structures, and their structural and performance attributes are systematically summarized in key applications, including methane activation,^53–56 Fischer–Tropsch (FT) synthesis,^57–59 catalytic cracking^60–62 and adsorption separation.^63–66

Fig. 1 A schematic representation of the main topics covered in this review. The inset shows ¹H–¹H DQ-SQ NMR spectroscopy, providing the acid site interaction of ZSM-5.

Beyond the various forms of diffusion coefficients, this review also discusses additional fundamental parameters commonly employed in the study of adsorption, diffusion, and reaction kinetics, which are essential for understanding molecular transport and transformation within zeolitic frameworks. One such parameter is Henry's constant, which characterizes the initial affinity of adsorbate molecules toward the zeolite surface at low partial pressures and serves as a benchmark for evaluating adsorption strength and selectivity under dilute conditions. The enthalpy of adsorption and entropy of adsorption respectively reflect the energetic and configurational changes upon sorption and are central to describing the thermodynamic driving forces governing equilibrium uptake. The mass transfer rate determines how rapidly molecules can migrate from the external fluid phase to the internal pore system and is often influenced by both external film resistance and internal pore diffusion. This process is quantitatively linked to the diffusion rate constant, which encapsulates molecular mobility within the zeolite pores and can vary significantly with pore topology, loading, and temperature. Under site saturation conditions, zero-order rate constants accurately describe the kinetics of surface reactions involving adsorbed molecules, etc. Altogether, these parameters form a quantitative basis for analyzing and modeling adsorption equilibria, intracrystalline diffusion, and reaction kinetics in zeolite catalysts. Additionally, we provide an in-depth discussion of the critical factors influencing adsorption and diffusion behaviors, offering a fresh perspective on their role in catalysis. Our systematic compilation of extant research findings delineates the fundamental structure and property relationships governing adsorption and diffusion performance and establishes a theoretical framework and practical foundation for the future development of catalytic adsorbents and separation materials. This review also identifies contemporary challenges and opportunities within the field, providing novel perspectives for future research directions and technological advancement.

2. Characterization methods for adsorption and diffusion

Researchers have employed various advanced characterization techniques to gain deeper insight into the adsorption and diffusion properties of zeolites. These methods elucidate the fundamental physical and chemical characteristics underlying adsorption and diffusion processes, playing a critical role in guiding the optimization of zeolite performance in catalytic reactions. In this section, we provide a comprehensive overview of these state-of-the-art techniques, highlighting their mechanisms, applications and advantages in the study of zeolite properties.

2.1 Adsorption described by the Langmuir model

The Langmuir adsorption isotherm model, one of the earliest and most widely implemented theoretical frameworks in adsorption science, serves as a fundamental cornerstone for understanding molecular adsorption phenomena in zeolitic systems.⁶⁷ This classical model operates under several idealized assumptions: the formation of a monolayer adsorbate coverage on the adsorbent surface, the energetic equivalence of all adsorption sites, and the absence of lateral interactions between adjacent adsorption sites. Although these idealized postulates may not fully hold in practical zeolitic adsorption systems, the Langmuir model continues to provide invaluable theoretical insights into the underlying mechanisms of zeolitic adsorption processes. The Langmuir isotherm equation can be expressed as:where q represents the amount of adsorbate per unit mass of adsorbent at equilibrium, q_max is the maximum monolayer adsorption capacity, K is the Langmuir adsorption constant associated with the affinity between the adsorbate and the adsorbent, and p denotes the equilibrium pressure in the gas phase or the equilibrium concentration in the liquid phase.

The Langmuir model provides a quantitative evaluation of the contributions of pore size and distribution to the selective adsorption and elucidates the role of acid sites in governing the adsorption strength. For example, the adsorption constant quantifies the strength of interactions between reactants and acid sites. For surface-modified zeolites, such as metal cation exchanged or organically modified zeolites, the Langmuir model can evaluate changes in adsorption selectivity induced by modifications, providing a scientific basis for the catalytic optimization. Furthermore, the Langmuir model is widely used to investigate the selective adsorption of gas separation applications, such as CO₂, CH₄, and N₂.⁶⁸ In hierarchical zeolites with both micro- and mesoporous structures, micropores provide adsorption sites to enrich reactants, while mesopores enhance mass transport by reducing diffusion resistance. By combining the Langmuir model with diffusion models, researchers can elucidate the synergistic effects between adsorption and diffusion, shedding light on how the dynamic equilibrium between these processes determines the overall catalytic performance.

The classical Langmuir model idealizes adsorption as a reversible one-to-one interaction between adsorbate molecules and energetically identical surface sites, yielding a monolayer saturation limit. In contrast, the Brunauer–Emmett–Teller (BET) theory extends this framework to allow weak, delocalized sorbate–surface interactions that support successive sorbate layers; nevertheless, it still treats the surface as energetically uniform. In real zeolite systems, surface heterogeneity is common due to the variations in site energies arising from framework defects, extra-framework cations, or external surface steps, and more flexible empirical or semi-empirical expressions, such as the Freundlich and Temkin isotherms, are commonly employed to better capture this surface heterogeneity.⁶⁹ For multicomponent adsorption systems, the competitive adsorption behavior often deviates from predictions based on single-component systems, necessitating the adoption of extended competitive Langmuir models or dual-site Langmuir models. The predictive accuracy may be compromised due to the emergence of multilayer adsorption phenomena at elevated pressures. For systems exhibiting significant pore constraints and diffusional limitations, the conventional Langmuir model may fail in accurately describing the complete adsorption process.

Lill et al. applied the Langmuir model to individual adsorption sites and identified four primary locations in the MFI zeolite, where CO₂ molecules are adsorbed (Fig. 2a): two within the straight channels (STR1 and STR2) and two within the sinusoidal channels (SINU1 and SINU2).⁷⁰ Only the STR2 site within the straight channels exhibited classical Langmuir behavior, characterized by a relatively high adsorption constant and the absence of neighboring molecular interactions. In contrast, the other three sites (STR1, SINU1, and SINU2) showed significant intermolecular interactions, rendering the Langmuir model inadequate, and the Toth isotherm models were required to accurately capture their adsorption characteristics. The Langmuir model not only offers classical thermodynamic evidence for adsorption but also provides quantitative support for the dual-site adsorption model on the external surface of zeolites. For instance, Moscatelli et al. identified the coexistence of two energetically distinct types of adsorption sites on the external surface of MFI zeolites, high-affinity sites located at pore mouths and lower-affinity sites distributed across inter-pore regions.⁷¹ By employing a set of structurally defined probe molecules (including o-MeDBK, o,o′-diMeDBK, 4-oxo-TEMPO and TEMPO), surface adsorption was quantitatively characterized using Langmuir isotherms. The adsorption behavior of 4-oxo-TEMPO displayed a clear two-site profile in Fig. 2b, requiring a dual-site Langmuir model for accurate fitting. In contrast, TEMPO, lacking a carbonyl group, exhibited selective adsorption only on the low-energy sites, with its isotherm fitting well to a single-site model. The distinction underscores the role of C=O functionality in strong-site binding. These Langmuir-derived parameters were further validated by EPR measurements of critical loading, collectively confirming the pore opening as the structural origin of the strong adsorption sites.

Fig. 2 (a) The change in occupancy of the four sites with varying CO₂ loading pressure and Rietveld refinement patterns. Reproduced with permission from ref. 70. Copyright 2022, American Chemical Society. (b) Langmuir isotherm for 4-oxo-TEMPO and TEMPO adsorbed onto silicalite S₇. Reproduced with permission from ref. 71. Copyright 2008, American Chemical Society.

2.2 Adsorption and diffusion studied by quasi-elastic neutron scattering (QENS)

QENS offers precise insights into diffusion as well as rotational and vibrational motions at the nanoscale, as a highly specialized technique for probing molecular dynamics. Its exceptional sensitivity to hydrogen atoms and ability to resolve both temporal and spatial scales make it uniquely suitable for investigating adsorption and diffusion processes in zeolites. By measuring minute energy transfers between neutrons and atomic nuclei, QENS provides critical quantitative data on adsorption and diffusion, including molecular residence times at adsorption sites, self-diffusion coefficients, and jump distances. The precise quantification of self-diffusion and collective diffusion coefficients reveals the limiting effects imposed by pore geometry and topology.²⁸

In terms of temporal resolution, QENS can capture dynamic events ranging from ultrafast vibrations and localized jumps on the picosecond scale to slow diffusion processes occurring over hundreds of nanoseconds (10⁻³–500 ns). This broad temporal window allows the technique to probe both atomic-level relaxations and cooperative molecular motions. In terms of spatial resolution, QENS is sensitive to displacements spanning from approximately 0.1 nm (the scale of individual chemical bonds) up to about 100 nm, encompassing molecular segments and diffusion pathways in nanostructured environments. Such a wide spatiotemporal coverage enables seamless investigation of dynamics from local atomic fluctuations to mesoscale collective behavior.⁷²

The transport diffusion coefficient D_t is determined using QENS through coherent scattering measurements:

while the self-diffusion coefficient D_s is derived using Einstein's relation:where S(Q) is a structure factor that describes density fluctuations in the molecular distribution, the diffusion transport coefficient D_t is related to the wave vector Q, which represents the scattering angle of neutrons, while ω denotes the frequency involved in the scattering process, and r_i(t) represents the position of molecule i at time t.⁷³

Eqn (2.2) indicates that the scattering intensity increases with the growth of D_t, and the diffusion transport coefficient D_t is directly proportional to both the frequency and wave vector in the scattering spectrum, providing valuable insights into the molecular dynamics within the pores. By fitting experimental data, the transport diffusion coefficient of molecules within the zeolite channels can be extracted, reflecting both the diffusion rate and the limitations imposed by the pore structure. The coefficient D_s is derived from the long-term displacement average, representing the molecular diffusion behavior. Analyzing the molecular motion in the QENS spectrum allows for the determination of the self-diffusion coefficient, thereby revealing the random walk characteristics of molecules within the channels. For example, the D_s value of normal alkanes decreases significantly as the chain length increases, e.g. in silicalite-1. Short-chain alkanes, such as methane and ethane, diffuse more rapidly within the pores, whereas long-chain alkanes experience restricted diffusion due to pore size limitations. D_t varies significantly with the concentration of the gas mixture. Jobic et al. extracted molecular diffusion kinetic parameters using QENS technology to study the diffusion behavior of alkanes in zeolites.⁷⁴ As shown in Fig. 3a, the side-on and end-on hydrogen bond configurations indicate a strong interaction between methanol molecules and the acidic sites of the HY zeolite, which significantly influences the diffusion rate of methanol within the zeolite.⁷⁵ The diffusion behavior of methanol follows an Arrhenius relationship with the temperature, and compared to the NaX zeolite, the diffusion coefficient in HY is notably higher. The QENS spectra of catechol in the beta zeolite show a significant elastic component, indicating that a large proportion of molecules remain immobile on the timescale probed by the experiment (Fig. 3b).⁷⁶ This motion is consistent with an isotropic rotation model. The D_s of methane in the NaY zeolite shows a non-monotonic trend at different loadings, with a peak observed at a loading of 32 CH₄ molecules per unit cell. Initially, at low loadings, the strong interaction between methane molecules and Na⁺ ions restricts the diffusion, resulting in a lower diffusivity (Fig. 3c). As the loading increases, the interactions between methane molecules become more pronounced, leading to an increase in the diffusion rate. However, at higher loadings, steric hindrance due to overcrowding of molecules causes a reduction in the diffusion rate.⁷⁷

Fig. 3 (a) Common hydrogen-bonded configurations of methanol in HY, showing both end-on and side-on geometries. Reproduced with permission from ref. 75. Copyright 2016, Royal Society of Chemistry. (b) QENS spectra as a function of Q for catechol at 393 K in the beta zeolite. Reproduced with permission from ref. 76. Copyright 2021, Springer Nature. (c) Self-diffusion coefficients as a function of CH₄ loading at 200 K, obtained through QENS for NaY. Reproduced with permission from ref. 77. Copyright 2010, American Chemical Society.

2.3 Adsorption studied by gas chromatography (GC)

GC is a highly efficient analytical technique used for the separation and analysis of volatile or semi-volatile compounds. By leveraging the differences in distribution coefficients of the sample components between the stationary phase and the mobile phase, GC enables the effective separation and quantitative analysis of complex mixtures. Breakthrough curve experiments involve injecting a gas mixture into a column packed with zeolites and using GC to monitor the outlet gas concentration over time.^78–80 This generates breakthrough curves that reveal key parameters, such as adsorption capacities, saturation points, and selectivities for different gases. In multicomponent gas systems, GC can distinguish the adsorption behavior of individual components, identifying preferential adsorption mechanisms that are essential for developing efficient gas separation materials. GC can construct adsorption isotherms under controlled gas flow and temperature conditions, enabling accurate determination of equilibrium adsorption capacities and adsorption site saturation. By fitting experimental data to adsorption models, such as the Langmuir isotherm (Section 2.1), GC allows a detailed analysis of adsorption site distributions and mechanisms, providing valuable insights for optimizing zeolite structures and surface functionalities. Common detectors include the flame ionization detector (FID),⁸¹ flame photometric detector (FPD),⁸² thermal conductivity detector (TCD),⁸³ and mass spectrometry detector (MS).⁸⁴ These detectors convert the signals of separated components into analyzable electrical signals. The output signals from the detectors form chromatographic peaks over time, with each peak corresponding to a specific component. The height and area of the peaks are utilized for quantitative analysis, while the retention time of the peaks is used for qualitative analysis.

The pulse injection method involves injecting a pulse of adsorbate with a known concentration into a chromatographic column packed with a molecular sieve material. This method monitors the retention times and changes in peak shapes of different components to analyze adsorption and diffusion behaviors. Peak broadening in chromatography is primarily caused by diffusion, mass transfer, and system effects. By analyzing the peak width, diffusion-related parameters, such as the effective diffusion coefficient, can be extracted.

Eqn (2.4) represents the Van Deemter equation,⁸⁵ where H is the height equivalent to a theoretical plate, A is the eddy diffusion term, related to the particle size and packing of the column filler, B is the longitudinal molecular diffusion term, associated with the gas diffusion coefficient, C is the mass transfer resistance term, related to the intraparticle diffusion coefficient, and u is the carrier gas velocity. By measuring the theoretical plate height H at various flow rates u and fitting the Van Deemter equation, the B and C terms can be extracted to calculate the effective diffusion coefficient D_eff.

The zero-length column (ZLC) method utilizes an extremely short chromatographic column, making external mass transfer resistance negligible.⁸⁶ Consequently, the desorption process of the adsorbate is primarily controlled by the diffusion within the molecular sieve particles. The analysis of the curve is performed using the following equation:

where C(t) is the concentration at time t, C₀ is the initial concentration, k is the desorption rate constant, and β is the gas–solid mass transfer ratio. By using the desorption rate constant and the mass transfer ratio, the intraparticle diffusion coefficient D_p can be calculated.

Denayer et al. employed pulse gas chromatography to systematically probe the low-coverage adsorption thermodynamics of C₁–C₁₄n-alkanes on the isostructural cage-based zeolites SAPO-34 and Na–CHA, elucidating the nuanced interplay between pore confinement and window-mediated transport limitations (Fig. 4a).⁸⁷ Through precise measurement of retention times and first-moment analysis of breakthrough curves, temperature-dependent Henry's constants were obtained, enabling extraction of zero-coverage adsorption enthalpies and entropies via Van’t Hoff analysis. The results reveal a distinctly non-monotonic relationship between the alkane chain length and the adsorption strength: while short alkanes exhibit an exponential increase in Henry's constant with the chain length due to cumulative dispersion interactions, a marked decline occurs for intermediate alkanes as molecular coiling becomes necessary to accommodate the cage volume. For longer chains (C₁₁–C₁₄), a secondary rise in adsorption strength is observed, attributed to partial chain extension through 8-MR windows despite increased steric penalties.

Fig. 4 (a) Henry adsorption constants of C₁–C₁₄n-alkanes on SAPO-34. Reproduced with permission from ref. 87. Copyright 2008, American Chemical Society. (b) ZLC switches between a flowing carrier gas with and without benzene across a thin layer of silicalite-1 crystals. (c) ZLC release profiles for benzene in silicalite-1. Linear segments of the ZLC profiles were used for short-time analytical fitting (black, 50 °C; red, 70 °C; green, 90 °C; blue, 110 °C). Reproduced with permission from ref. 88. Copyright 2015, American Chemical Society.

The diffusion behavior of benzene in silicalite-1 crystals of varying particle sizes was examined using the ZLC method. Fig. 4b illustrates that the ZLC method enables the capture of desorption dynamics through an instantaneous switch from benzene-containing to pure carrier gas. Fig. 4c presents the desorption profiles of benzene from silicalite-1 crystals ranging from 62 nm to 3 μm at different temperatures.⁸⁸ As temperature increases, the desorption rate accelerates, indicating a thermally activated diffusion process. Notably, smaller crystals exhibit markedly slower desorption, with diffusion rates decreasing far more rapidly than predicted by classical intracrystalline models, highlighting that surface permeation becomes the rate-limiting step. Quantitative comparison with simulation results further reveals that such behavior can only be reproduced when the fraction of open surface pores falls below 0.1% (over 99.9% of the surface pores are effectively blocked). This suggests that surface pore occlusion is an inherent structural feature in nanosized MFI zeolites, significantly hindering molecular transport. As a result, even when the internal microporous framework remains intact, overall material performance may be severely compromised by surface-level structural constraints.

2.4 Diffusion studied using pulsed field gradient nuclear magnetic resonance (PFG NMR)

PFG NMR is a powerful technique for measuring self-diffusion coefficients of molecules without altering or damaging the sample.^42,43 This method is particularly well-suited for studying diffusion in porous materials, as it can directly detect the migration and diffusion of molecules within pore structures. In liquid-phase diffusion studies, PFG-NMR has been extensively employed to measure the diffusion of water and organic molecules in hydrophobic or hydrophilic zeolites, providing quantitative data that guide the design of zeolites for liquid-phase catalysis or separation.⁸⁹Fig. 5a illustrates the pulsed gradient spin echo (PGSE) sequence. PFG NMR introduces controllable magnetic field gradient pulses into a conventional spin-echo NMR sequence, thereby correlating the phase of nuclear spins with their spatial positions. By measuring changes in the phase, molecular displacement information can be obtained, enabling the calculation of diffusion coefficients.⁹⁰

Fig. 5 (a) Pulse sequence of PGSE. (b) The relationship between effective diffusivities and mean square displacements for n-octane, as measured by PFG NMR (data points) and predicted by dynamic Monte Carlo simulations (lines). Reproduced with permission from ref. 97. Copyright 2005, American Chemical Society. ¹H PFG NMR attenuation data for DNL-6 (c) and SAPO-42 (d) zeolites during methane adsorption at 298 K under 1.0 bar pressure. (e) The loading dependence of self-diffusion coefficients for methane in both zeolites at 298 K, measured via PFG NMR. (f) 2D ¹²⁹Xe EXSY NMR spectrum of xenon in DNL-6 at 154 K. Reproduced with permission from ref. 100. Copyright 2025, American Chemical Society.

This method is particularly effective in distinguishing between free and restricted diffusion behaviors and captures their time and spatial dependencies. The diffusion coefficient D can be calculated from the relationship between signal intensity and the gradient strength g, which follows the Stejskal–Tanner equation (eqn (2.6)), where S represents the signal intensity and S₀ is the unattenuated signal intensity.^91–96

Fig. 5b illustrates the relationship between the diffusion coefficient of n-octane and the root mean square displacement in NH₄-Y and USY zeolites, demonstrating that the diffusion coefficient decreases as the displacement increases, which is attributed to limitations imposed by the crystal size and surface defects. The diffusion rate in USY declines more rapidly, reflecting its smaller average crystal size and greater diffusion restrictions.⁹⁷ Hunger et al. investigated the relationship between organic deposition, acid sites, diffusion properties, and catalytic performance of SAPO-34 zeolites in the MTO reaction.⁹⁸ The diffusion activation energy for ethene was determined to be 4.2 kJ mol⁻¹, significantly lower than 7.6 kJ mol⁻¹ for ethane. This indicates that ethene diffusion in SAPO-34 is more sensitive to temperature variations and can sustain higher diffusion rates even at lower temperatures. The pore structure of SAPO-34 exhibits excellent selectivity toward smaller molecules like ethene, which enhances its catalytic performance. Liu et al. reported the diffusion coefficients of methane in different zeolites with the RHO zeolite exhibiting the highest diffusion coefficient (5.4 × 10⁻¹⁰ m² s⁻¹), followed by CHA- (1.2 × 10⁻¹⁰ m² s⁻¹), while LEV zeolites showed the lowest diffusion coefficient (2.7 × 10⁻¹¹ m² s⁻¹). Notably, the diffusion coefficient of methane increases with loading, especially in the LEV zeolite, where a marked enhancement is observed as the loading increases. The size and shape of the 8-MR windows in this zeolite play a crucial role in governing methane transport. Larger pores, like those in the RHO zeolite, facilitate faster diffusion, whereas smaller pores, such as those in LEV, restrict methane movement. CHA zeolites show intermediate diffusion rates, lying between RHO and LEV zeolites.⁹⁹

These differences highlight the strong influence of the zeolite pore structure on molecular diffusion and underscore the importance of the pore architecture in shape-selective catalysis, particularly in processes like methane conversion. As shown in Fig. 5c–f, the behavior of methane adsorption and diffusion in DNL-6 (RHO) and SAPO-42 (LTA) zeolites was systematically investigated using ¹H PFG NMR and 2D ¹²⁹Xe exchange spectroscopy (EXSY) NMR techniques. DNL-6 exhibits higher methane diffusivity, whereas SAPO-42 shows more significant diffusion constraints. By analyzing the relationship between methane loading and self-diffusion coefficients, it was observed that as methane loading increased, the diffusion coefficient of DNL-6 significantly increased due to a completely new “self-gating” mechanism proposed by Liu et al., while SAPO-42 displayed a more gradual change. This highlights the substantial differences in the diffusion behavior of gas molecules due to the distinct pore structures of the two zeolites.¹⁰⁰

2.5 Theoretical simulation of adsorption and diffusion

The theoretical computational methods range from density functional theory (DFT) at the quantum mechanics level to molecular dynamics (MD) and Monte Carlo (MC) simulations at the classical force field level. DFT is typically used to solve the electronic structure, while MC simulations are employed to statistically determine the equilibrium adsorption distribution, and MD is used to capture the dynamic diffusion pathways of molecules within the pores. Through multiscale integration, these methods can deeply reveal the intrinsic mechanisms of adsorption and diffusion at the atomic or molecular scales.^101–105

E_ads = E_complex − (E_zeolite + E_molecule) (2.7)

ΔE = E_TS − E_Reactants (2.8)

As shown in eqn (2.7) and (2.8), DFT provides critical insights into the most stable configurations of adsorbed molecules within zeolites, as well as their adsorption energies, reaction energy barriers, and electronic structures. The reaction energy barrier governs the reaction rate, with lower barriers corresponding to faster reaction kinetics. Meanwhile, adsorption energy influences the concentration and selectivity of reactants on the catalyst surface, playing a pivotal role in catalytic efficiency and product distribution.

The MC method obtains the equilibrium properties of a system by randomly sampling system configurations and state space to satisfy the probability distribution of a specific statistical ensemble. For adsorption, the commonly used approach is the grand canonical Monte Carlo (GCMC) method, which allows the number of particles in the system to vary. This enables the direct determination of adsorption isotherms, component selectivity, and pore occupancy. In eqn (2.9), θ represents the adsorption amount, 〈N〉 is the statistically averaged number of particles, and N_total is the total number of adsorption sites within the zeolite channels. MD simulations, based on Newton's equations of motion, directly track the trajectories of molecules over a given timescale using known molecular force field parameters, thereby extracting dynamical properties, such as diffusion coefficients, molecular conformational changes, and vibrational frequencies. The diffusion coefficient in MD simulations is also determined using eqn (2.3).

As shown in Fig. 6a, Xe and Kr predominantly adsorb within the CHA cages of the Ag-SSZ-13 zeolite, with a higher concentration near the Ag⁺ sites, indicating that Ag⁺ ions significantly enhance the adsorption of Xe. Compared to Kr, Xe exhibits a much higher adsorption density at the Ag⁺ sites, suggesting a stronger interaction between Xe and Ag⁺. In Ag-ZSM-5 zeolites, Xe and Kr are primarily adsorbed in the 10-MR channels, with Ag⁺ sites further enhancing the adsorption of both gases.¹⁰⁶ Through DFT-MD simulations, Brogaard et al. revealed that ethene molecules dynamically coordinate to the nickel active site in Ni-SSZ-24 (AFI), promoting its reversible detachment from the ion-exchange position within the zeolite framework (Fig. 6b).¹⁰⁷ This coordination leads to the formation of mobile [(ethene)₂–Ni–alkyl]⁺ complexes, which adopt a square-planar geometry favored by Ni(II) and are stabilized by both enthalpic and entropic contributions. Importantly, the dynamic detachment of the Ni center significantly enhances catalytic performance by lowering the intrinsic free-energy barriers for C–C bond formation (77 to 37 kJ mol⁻¹) and for β-hydride transfer, facilitating selective ethene dimerization. The mobile nature of the active site also improves butene desorption kinetics and introduces characteristics typically associated with homogeneous catalysis, thereby highlighting a mechanistic regime that bridges the heterogeneous–homogeneous catalysis within zeolite-confined spaces.

Fig. 6 (a) Density distribution profiles of Xe and Kr adsorbed on Ag-SSZ-13 and Ag–ZSM-5 at 10 kPa, 298 K. Reproduced with permission from ref. 106. Copyright 2022, Elsevier. (b) The free-energy profiles for ethene coordination to [ethene–Ni–ethyl]⁺ (left) and [ethene–Ni–butyl]⁺ (right) were obtained through DFT-MD simulations at 25 bar and 120 °C. The histogram in the left inset demonstrates the Ni–Al distance distributions; the right insets present simulation snapshots depicting the configurations of the Ni–ethyl species. Reproduced with permission from ref. 107. Copyright 2019, American Chemical Society. (c) The 623 K FESs for isobutene conversion in cage B. (d) Density maps illustrating the position of the tertiary isobutene carbon atom during 1 ns of PBE-D3(BJ) and RI-RPA (random phase approximation) MD simulations of isobutene and tert-butyl cations at 623 K. Reproduced with permission from ref. 108. Copyright 2019, Wiley-VCH.

Bocus et al. investigated the coordination energetics of ethene binding to Ni–ethyl and Ni–butyl complexes confined within the SSZ-13 zeolite framework (Fig. 6c).¹⁰⁸ The coordination of ethene to the Ni–ethyl complex entails only moderate changes in free energy, indicating a relatively smooth and entropically accessible process. In contrast, the Ni–butyl complex shows a substantially higher free energy barrier, which can be attributed to steric constraints imposed by the bulky butyl group within the confined 8-MR cages of the CHA structure. This contrast highlights that both the size of substituents and the geometric constraints of the zeolite topology play a critical role in governing coordination dynamics. Moreover, simulation snapshots (Fig. 6d) capture the transient configurations during the ethene coordination to Ni–ethyl, revealing the synergistic interactions between the Ni center and ethene. By comparing these two studies, it becomes evident that the large-pore AFI topology offers sufficient spatial freedom and residence volume for metal complexes to migrate and undergo cooperative interactions, thereby enhancing catalytic efficiency. In contrast, the small-cage, narrow-window CHA framework intensifies steric hindrance and electrostatic confinement, which significantly destabilizes carbocationic intermediates. Therefore, the design of zeolite catalysts should simultaneously consider both the framework architecture and the coordination saturation of active sites. In wide-pore systems, mobile complexes can be leveraged to achieve high selectivity, whereas in confined microporous environments, the stabilization of short-lived cationic species may require the introduction of ancillary ligands or defect engineering strategies.

3. Zeolite materials in adsorption and diffusion applications

The topological structure of zeolites determines the size, shape, connectivity, and aperture dimensions of the channels. These factors directly influence the positioning of adsorbed species within the channels, their diffusion pathways, and intermolecular interactions. In this section, we systematically summarized the diverse applications of typical zeolite materials (e.g., FAU, CHA, MFI, MOR and *BEA) in adsorption and diffusion, focusing on their catalytic performance across various topological frameworks, offering critical insights into their optimization for specific industrial applications. Table 1 lists the zeolites discussed in this review, including their framework topologies and typical Si/Al ratio. Specifically, to achieve precise control over Si/Al ratios, Yu et al. synthesized a highly siliceous FAU-type zeolite with the Si/Al ratio reaching 42.56 by utilizing low charge-density organic structure-directing agents (OSDAs).¹⁰⁹ High-silica AEI zeolites with Si/Al ratios ranging from 13 to 20 can be prepared via interzeolite transformation from FAU-type precursors.¹¹⁰ Zeolites with different pore sizes exhibit distinct adsorption selectivity and diffusion properties, reflecting pronounced size effects and structure-dependent behaviors. The influence of pore size and structural variations on catalytic performance will be comprehensively explored. Fig. 7 illustrates the characteristic structural features and key interaction sites within zeolites, including the well-ordered microporous framework, BAS, Lewis acid sites (LAS), extra framework aluminum (EFAl), hydroxyl groups, hydrogen bonding, van der Waals interactions, etc. These diverse interactions not only govern the adsorption and diffusion behaviors of guest molecules within the pores, but also play a pivotal role in modulating reactant activation and pathway selectivity during catalytic processes.

Table 1 framework type and typical Si/Al ratio of the zeolite catalysts covered in this review

Code	Trivial name^a	Framework structure	Typical Si/Al ratio^b	Ref.
a The trivial names denote only the specific materials discussed in this review and do not encompass all zeolites associated with each IZA code. b It represents typical values and exclude those resulting from post-synthetic modifications such as dealumination or desilication.
AEI	SAPO-18, SSZ-39	8-MR, the alternating orientation of double 6-rings (D6Rs)	SSZ-39: 6–20	110–112
AFI	SSZ-24	One-dimensional 12-MR	50–∞	107 and 113
*BEA	Beta	Intersecting three-dimensional 12-MR	Al-rich: <5; high-silica: >10	114
CHA	SSZ-13, SAPO-34	8-MR, layers of tilted D6Rs with the same orientation	SSZ-13: 3–150	87, 106, 108 and 115
DDR	Deca-dodecasil 3 rhombohedral	8-MR 19-hedral cages	All-silica	116
EMT	ZSM-20 (FAU/EMT)	12-MR c axis: circular 0.73 nm pores a axis: 0.65 nm × 0.75 nm channels	1–3	117
FAU	X, Y	12-MR 1.3 nm supercage	X: 1–1.5 Y: 1.5–3	75, 109, 118 and 119
FER	Ferrierite	Two-dimensional 10-MR with intersecting 8-MR	8–75	120
LEV	Levyne	Single 6-MR, 0.48 nm × 0.36 nm pores of 8-MR	16–22	121
LTA	A, SAPO-42	8-MR, α-cage, β-cage	1–∞	100, 122 and 123
MER	Merlinoite	Three intersecting 8-MR	2–10	124
MFI	ZSM-5, silicalite-1	10-MR, straight and zig-zag channels	10–∞	70, 71, 106 and 125
MOR	Mordenite	Straight 12-MR, crossed 8-MR	2–150	126 and 127
MWW	MCM-22, SSZ-70	10-MR sinusoidal channels, 12-MR supercages	7–50	128
RHO	ZSM-25	α-Cage with d8r	1–13	129
SOV	SCM-15 (germanosilicate)	12 × 12 × 10-ring channels	—	130

---

Fig. 7 Integrated view of structural features and host–guest interactions in zeolite catalysis, exemplified by MFI/MEL intergrowth.

3.1 FAU zeolites

3.1.1 Gas adsorption on FAU-type zeolites.
3.1.1.1 Hydrocarbon adsorption on FAU-type zeolites. The FAU topology features large pores (12-MR) and a relatively open three-dimensional channel network. The supercage structure provides sufficient space for the smooth diffusion and reaction of large molecules. The design of materials with synergistically active sites has long been a key strategy for enhancing catalytic efficiency. Zhao et al. developed a rhenium-doped Y zeolite catalyst with Re–O–H⁺ synergistic active sites for the cross-metathesis of ethene and trans-2-butene. By anchoring individual Re species on the inner surface of the zeolite framework and positioning them near BAS, the system effectively promoted the adsorption of olefin molecules. This precise alignment with the Re sites promotes the formation of reaction intermediates, significantly boosting catalytic activity. The immobilization of Re species does not follow a conventional ion-exchange mechanism. Instead, molecular ReO₄ units are covalently anchored onto the zeolite framework by replacing strong Brønsted acid protons (H⁺) located on O atoms (specifically O₄ sites) within the supercages of the USY zeolite. This anchoring process is likely facilitated by a dehydration reaction between terminal Re–OH groups and BAS, resulting in a covalent Re–O–zeolite bond. Nearby unreacted normal BAS remains proximal to the anchored Re, forming defined Re OMS–H⁺ (olefin metathesis site) active pairs that exhibit synergistic catalytic behavior. The ReO_x/USY catalyst demonstrated remarkable propene selectivity (up to 90%) and 2-butene conversion (79%) at a mild temperature of 75 °C, outperforming both conventional WO_x and other Re-based catalysts.¹³¹

Li et al. have achieved a breakthrough in the efficient separation of alkynes and olefins by regulating the topology and pore of FAU zeolites.¹³² By incorporating isolated and active Ni(II) sites into the six membered rings of FAU zeolites, the chemoselective adsorption toward alkynes was significantly enhanced (Fig. 8a). At 298 K and 1 bar, Ni@FAU exhibited an impressive adsorption capacity of 3.48 mmol g⁻¹ for acetylene, far surpassing its capacity for ethylene (2.36 mmol g⁻¹). The acetylene uptake remained as high as 2.0 mmol g⁻¹, meeting the requirements for industrial ethylene stream purification even at a low pressure. The adsorption heat measurements demonstrated that the heat of acetylene adsorption on Ni@FAU was 48.6 kJ mol⁻¹, substantially exceeding that for ethylene (25.8 kJ mol⁻¹), indicating the formation of strong yet reversible chemical interactions between alkynes and the Ni(II) sites. Selective and reversible adsorption is achieved through the formation of metastable [Ni(alkyne)₃] complexes.

Fig. 8 (a) In situ FTIR spectra of Ni@FAU upon adsorption of C₂H₂ and C₂H₄ at 298 K. Reproduced with permission from ref. 132. Copyright 2020, AAAS. (b) Crystallographic model of WO_x/USY derived from Rietveld refinement reveals the immobilization of WO₄ MAS via hydrogen bonding. Adsorption of trans-2-butene on WO_x/USY demonstrates significant interaction between the alkene double bond, which leads to the formation of a stereospecific adsorption geometry. Reproduced with permission from ref. 134. Copyright 2018, American Chemical Society.

Subsequently, they also discussed the potential of uniform and stable extra-framework transition metal ions (TMIs) in FAU zeolites as active sites for adsorption and catalysis.¹³³ By precisely incorporating TMIs such as Co, Ni, Cu, Rh, and Pt into specific cationic locations within the FAU framework, particularly at the 6-MR centers, these active ions were effectively stabilized. These TMIs exhibit well-defined coordination environments, achieve high adsorption efficiency through electronic structure modulation and possess superior performance in applications, such as selective hydrogenation, olefin oxidation, and alkyne/olefin separation, benefiting from their synergistic interaction with the zeolite framework.

When metal oxides are precisely dispersed within the zeolitic pore channels, the synergistic effect between the active sites of the two components can significantly improve the catalytic performance. Tsang et al. uniformly distributed WO_x active sites within a USY zeolite as a bifunctional olefin disproportionation catalyst, exhibiting a high propylene yield of 79% with 80% selectivity under mild conditions (200 °C, 1 bar).¹³⁴ SXRD analysis revealed that trans-2-butene adopts a unique adsorption configuration on the WO_x/USY catalyst. The C=C bond interacts with the BAS and forms oriented interactions with the WO bonds of WO₄ active sites, displaying structural features of a metal-containing cyclic intermediate. The weak acid sites of the USY molecular sieve can effectively suppress cracking and coking side reactions, thereby ensuring high stability and selectivity of the catalyst (Fig. 8b).

3.1.1.2 NO_X and CO_X adsorption on FAU-type zeolites. In addition to the adsorption of olefins or alkynes, FAU-type zeolites combined with metal ions have also been studied in NO_x and CO_x adsorption. When using alkali and alkaline earth metal cation-exchanged FAU zeolites in N₂O adsorption, Li et al. discovered that N₂O molecules preferentially anchor to extra-framework cations through terminal oxygen atoms while being stabilized by electrostatic interactions between the central nitrogen atoms and framework oxygens. At extremely low pressures, N₂O molecules predominantly occupy the sodalite (t-toc) cages of the FAU zeolite, and as the pressure increases to 0.05 bar, they occupy also the supercages (t-fau). Under high concentration conditions, the dynamic adsorption capacity of the zeolite for N₂O correlates with its specific surface area and adsorption strength. In contrast, at low concentrations, diffusion resistance exerts a significant negative impact. As shown in Fig. 9a–g, Ba-FAU and Na–CHA zeolites exhibit particularly remarkable adsorption capacities at higher concentrations. In the low-pressure range of 0.05 bar, both Ba-FAU and Na–CHA outperform other samples in terms of dynamic adsorption performance.¹³⁵ Similarly, for CO₂ adsorption, Ghassabzadeh et al. utilized post-synthesis modification in a fluoride environment, which enables the incorporation of alkali metals, alkaline earth metals, and transition metals into FAU zeolites.¹³⁶ Selectivity calculations based on Henry's constants and ideal adsorbed solution theory (IAST) demonstrate that for the same metal ion incorporation, the X zeolite with a lower Si/Al ratio exhibits superior separation performance for CO₂/CH₄ and CO₂/N₂ mixtures compared to the Y zeolite (Fig. 9h).

Fig. 9 N₂O adsorption isotherms on (a) M-FAU and (b) M–CHA zeolites at 298 K. (c) Comparison of N₂O uptake upon various adsorbents. (d) N₂O, N₂ and O₂ adsorption isotherms on the Ba-FAU zeolite at 298 K. (e) Calculated IAST selectivity of N₂O/N₂ and N₂O/O₂ mixtures (0.1/99.9) on the Ba-FAU zeolite. (f) N₂O, N₂ and O₂ adsorption isotherms on the Na–CHA zeolite at 298 K. (g) Calculated IAST selectivity of N₂O/N₂ and N₂O/O₂ mixtures on the Na–CHA zeolite. Reproduced with permission from ref. 135. Copyright 2023, Elsevier. (h) Henry's selectivity and IAST predicted selectivity for CO₂/CH₄ (10 : 90) and CO₂/N₂ (15 : 85) gas mixtures of a metal incorporated FAU zeolite. Reproduced with permission from ref. 136. Copyright 2023, Elsevier. (i) FTIR spectra following carbon dioxide adsorption at 180 Torr (left) and CO₂ adsorption capacity of Na- and Cu–X zeolites measured at 180 Torr (right). Reproduced with permission from ref. 137. Copyright 2019, Elsevier.

A comparison of the adsorption properties of Na–X and Cu–X zeolites revealed that Cu–X exhibits a significantly higher affinity for NO. The enhanced adsorption is attributed to the introduction of copper ions, which create additional active sites, strengthening the interaction between NO and the zeolite surface. In contrast, Na–X demonstrates a substantially greater adsorption capacity for CO₂, particularly in the physisorption region, with an uptake eight times higher than that of Cu–X (Fig. 9i).¹³⁷ The adsorption of NO on Cu–X primarily occurs through the formation of Cu²⁺–NO coordination bonds, while CO₂ adsorption on Na–X predominantly involves the formation of monodentate and bidentate carbonate species. Yahiro et al. utilized electron paramagnetic resonance (EPR) spectroscopy to investigate the magnetic interactions between Cu²⁺ ions in copper-exchanged Y zeolites and paramagnetic NO and O₂ molecules at low temperatures. The adsorption of paramagnetic NO and O₂ can significantly alter the EPR signals of Cu²⁺ ions, primarily due to magnetic interactions between the adsorbed molecules and the metal centers, leading to line broadening and nonlinear variations in signal intensity. Temperature-dependent EPR analysis enables the assessment of adsorbate stability, desorption behavior, and migration characteristics within the pore network, thereby providing indirect evidence for understanding diffusion processes.¹³⁸ Within the temperature range of 110 to 270 K, significant changes in EPR signal intensity and linewidth were observed upon the introduction of NO or O₂. For NO, adsorption was accompanied by spin–spin dipolar interactions, leading to broadened EPR signals and a minimum signal intensity at 150 to 170 K. This behavior corresponds to the proportion of paramagnetic NO monomers formed within this temperature range. In contrast, O₂ exhibited a distinct behavior, with adsorption increasing significantly below 230 K, enhancing magnetic interactions. The adsorption of NO was further analyzed using the Langmuir model, yielding an estimated adsorption enthalpy of approximately 6.0 kJ mol⁻¹, consistent with previously reported values for NO adsorption on activated carbon.

Introducing coordinatively unsaturated Cu²⁺ sites into FAU zeolites enables the efficient separation of C₂H₂ and CO₂. Li et al. demonstrated that Cu@FAU can effectively remove most adsorbed CO₂ at low temperatures using an “adsorption–purging–desorption” three-step process, while retaining approximately 66% of the adsorbed C₂H₂, achieving high-purity acetylene (97.8%). The C₂H₂ molecules are stably adsorbed through strong hydrogen bonding and coordination interactions between the Cu²⁺ active sites and the framework oxygen atoms, whereas CO₂ molecules are bound only by weaker physical interactions.¹³⁹

3.1.2 Specific adsorption and diffusion properties of FAU-type zeolites. For FAU zeolites in diffusion and catalysis, Mintova et al. focused on how microporosity modification influences molecular diffusion and catalytic performance.¹⁴⁰ The use of NH₄F solution for unbiased etching enables the removal of silicon or aluminum atoms from the framework, leading to the formation of mesoporous structures. The etched FAU-type zeolites (FY series) exhibit significantly larger surface area and pore volume compared to the original samples (PY), with mesopore volume accounting for approximately 6%, while the original micropore volume is around 25%. 3D-TEM and nitrogen physical adsorption measurements revealed that the mesopore sizes in the FY series range from 2 to 5 nm, with the majority of mesopores in the FY20 sample concentrated in the 2–3 nm range. The hyperpolarized ¹²⁹Xe NMR (Fig. 10a and b) studies demonstrated that these hierarchical structures significantly enhanced mass transfer between micropores and mesopores. The connectivity of the mesopores was also notably improved, with the FY60 sample showing a mesopore connectivity of up to 90%, indicating an increased connection between the SOD cages and supercages during the etching process, forming a more efficient mesopore network. DFT calculations show that the introduction of silanol nest defects significantly reduces the diffusion energy barrier for gases. In the pure silica FAU structure, the diffusion energy barrier for Xe from the SOD cage to the supercage is 402 kJ mol⁻¹, while the energy barrier is significantly reduced upon introducing silanol nest defects. For example, a model with one silanol nest defect shows a diffusion energy barrier of 83 kJ mol⁻¹, while a model with four silanol nest defects shows a diffusion energy barrier of 34 kJ mol⁻¹, indicating that the introduction of defects significantly facilitates gas diffusion. In the TiPBz dealkylation reaction, the FY5 sample shows higher product selectivity, particularly in the formation of 1,4-diisopropylbenzene (1,4-DIPB) and 1,2-diisopropylbenzene (1,2-DIPB). In the n-C₈ hydrogenation cracking reaction, the FY60 zeolite exhibited a higher conversion rate compared to the PY zeolite, mainly due to the improvement in contact between the reactant molecules and the active sites owing to the introduction of mesoporosity.

Fig. 10 (a) 1D HP ¹²⁹Xe spectra of FY5-FY60 zeolites at 253 K. (b) Difference in chemical shifts between the peaks associated with the xenon in the sodalite cages and supercages in the hierarchical zeolites. Reproduced with permission from ref. 140. Copyright 2021, Wiley-VCH. (c) Methanol dehydration to DME in an in situ water removal process within a bifunctional catalytic membrane reactor (CMR) featuring a zeolite FAU-LTA double-layer membrane: methanol dehydration occurs on the acidic H–FAU layer, while water is removed by the Na–LTA layer. Reproduced with permission from ref. 147. Copyright 2016, Wiley-VCH.

The adsorption efficiency of molecules on BAS is determined by their basicity and geometric structure and constrained by the diffusion path length and spatial restrictions within the channels. Pang et al. reported that the selectivity for methyl acrylate and acrylic acid increased from 61% to 84% upon the addition of amines using the NaY zeolite as the catalyst. Through selective adsorption, amines effectively passivated the in situ generated BAS during the reaction, significantly suppressing side reactions, such as decarbonylation. Amines with high proton affinity and molecular sizes smaller than 85% of the NaY pore diameter were most effective in enhancing the dehydration selectivity. However, as the size of amines increases, steric hindrance gradually becomes a limiting factor, leading to significant reduction in BAS accessibility due to internal diffusion limitations and local steric constraints. Furthermore, bulkier amines, such as 2,6-diisopropylpyridine, exhibit significantly slower diffusion rates within NaY channels compared to smaller amines, and this internal diffusion limitation severely compromises their effectiveness in passivating BAS.¹⁴¹ On the other hand, the solvent composition can be optimized to modulate the molecular distribution and reaction rate within the FAU catalyst cages. For instance, in various γ-valerolactone (GVL) with water mixed solvents, particularly at low GVL concentrations, glucose adsorption is significantly reduced leading to a decrease in the isomerization rate. However, as the GVL concentration increases, glucose preferentially partitions into the cages, elevating the local glucose concentration and substantially enhancing isomerization activity. The microheterogeneity of the solvent system, reflected in the non-monotonic trend of mixing enthalpy, affects the mobility and adsorption of carbohydrates, water, and GVL, as shown by PFG NMR diffusivity measurements. At low GVL concentrations, glucose is less concentrated within the zeolite cages than in the bulk solution, and variations in coadsorbed water further hinder the isomerization process. As the GVL content increases, the reduced reactivity is largely offset by the significant accumulation of glucose in the pores, leading to a dramatic increase in the local sugar concentration at the solid–liquid interface.¹⁴²

To systematically elucidate the adsorption and diffusion behaviors of target molecules within FAU zeolites, both the zeolite framework and the adsorbed molecules can be modeled as flexible entities. The diffusion coefficient inversely correlates with molecular weight, with n-butane exhibiting significantly lower diffusion rates compared to ethane. As the loading increases, the diffusion coefficients markedly decrease, primarily due to the aggregation effects of the adsorbed molecules. At low loadings, molecules maintain high mobility, facilitating omnidirectional diffusion within the cage structures of the FAU-type zeolite. Conversely, molecules preferentially diffuse within individual supercages rather than migrating between cages at high loadings, particularly for n-butane.¹⁴³

Regarding the diffusion behavior of isomers, Xu et al. employed Arrhenius fitting to reveal that the diffusion mechanisms and activation energies of dimethylbenzene isomers are temperature dependent. At elevated temperatures (600 K and 900 K), the dimethylbenzene isomers predominantly reside within the central regions of supercages and diffuse through the centers of 12-MR apertures. In contrast, at temperatures below 300 K, the isomers are more uniformly distributed within the supercages and preferentially traverse the edges of the 12-MR windows, exhibiting a “jump-like” diffusion mechanism. Orientation analysis reveals that at low temperatures, the molecular planes of the dimethylbenzene isomers maintain specific angles relative to the FAU framework axes, whereas at high temperatures, their orientations become more randomized. The π-electrons of aromatic rings are represented as quadrupolar electrostatic interaction sites, shaping the molecular electrostatic potential rather than exerting any intrinsic π–π attractive force. The observed enhancement in molecule–framework interaction stems primarily from electrostatic complementarity between the anisotropic charge distribution of the π-system and the negatively charged oxygen atoms in the zeolite lattice. This interaction strengthens local adsorption, particularly near critical topological features, thereby increasing residence times and stabilizing preferential molecular orientations. Collectively, these effects give rise to localized electrostatic potential wells that restrict translational freedom and reduce overall diffusivity. When the π-electron sites are removed from the model, the electrostatic landscape becomes less structured. Diffusion energy barriers are diminished, and molecular mobility is correspondingly enhanced, as reflected in steeper MSD profiles. These findings support a refined view in which π-electrons modulate diffusion not through inherent “π-stacking” forces, but by sculpting the electrostatic field landscape within the microporous environment.^144,145

In the fluid catalytic cracking (FCC) process, FAU-type zeolites crack large molecular hydrocarbons into lighter hydrocarbon products, such as gasoline, propylene, butylene, and other light to medium hydrocarbons. The EMT/FAU intergrowth zeolite exhibits a finely tuned synergy of pore structure and acidity, enabling an optimal balance between hydrodesulfurization efficiency and retention of research octane number, highlighting its potential for advanced catalytic applications.¹⁴⁶ Huang et al. presented an innovative bifunctional catalytic membrane reactor designed for the efficient dehydration of methanol to dimethyl ether (DME).¹⁴⁷ This reactor features a sandwich-like membrane structure with an H–FAU top layer and a Na–LTA bottom layer (Fig. 10c). The H–FAU layer, with its moderate acidity, facilitates the catalytic conversion of methanol into DME and water, while the Na-LTA layer, leveraging its hydrophilic nature and precise pore size (∼0.4 nm), rapidly adsorbs and diffuses water molecules. Meanwhile, the larger DME molecules (a kinetic diameter of 0.43 nm) are effectively excluded, enabling selective water removal. This configuration achieves a methanol conversion rate of up to 90.9%, near-complete selectivity for DME at 310 °C and ensures remarkable long-term operational stability. In the NH₃ selective catalytic reduction (SCR) reaction,^148–150 the introduction of co-cations directs copper ions to preferentially occupy the FAU supercages, thereby enhancing both the efficiency and selectivity of the SCR process. By favoring localization within the sodalite cages, co-cations effectively suppress active sites associated with the N₂O formation, leading to an optimal reaction pathway. Among the co-cation-modified Cu-FAU catalysts, Ba, Sr, and Ca ions exhibit significant improvements in catalytic activity and selectivity.¹⁵¹

Jousse et al. conducted a series of systematic investigations to elucidate the adsorption and diffusion behavior of benzene in FAU zeolites.¹⁵³ The underlying kinetics are modulated by a complex interplay of factors including framework topology, charge distribution, cation type and concentration, structural flexibility, and intermolecular interactions. In the NaY zeolite, benzene migration between the S_II and W adsorption sites has been extensively studied using transition state theory (TST) and correlation function theory (CFT). The S_II site is located near Na⁺(II) ions, while the W site resides in the 12-MR window between adjacent supercages. The energy barrier between these sites ranges from 16 to 20 kJ mol⁻¹, and the corresponding rate constants can be accurately determined by combining minimum energy path calculations with TST and CFT approaches. Notably, CFT incorporates configurational entropy and vibrational contributions, making it more suitable for describing the migration of non-spherical molecules.¹⁵² Although the benzene–Na⁺ interaction leads to a redshift in low-frequency center-of-mass vibrational modes, indicating strong guest–framework coupling, the impact on diffusion rates is limited. This is because benzene's internal degrees of freedom are sufficient to dissipate excess kinetic energy, enabling thermal equilibration even under rigid lattice conditions.

In contrast, benzene diffusion in acidic HY zeolite is considerably more intricate. The presence of BASs allows benzene to adsorb in face-on configurations through interactions with framework protons in the supercage. These sites have been confirmed by both IR spectroscopy and neutron diffraction. As these adsorption geometries are located along the cage wall rather than at its center, benzene migration proceeds incrementally along the inner surface. At low temperatures (< 150 K), this wall-bound diffusion involves energy barriers of 10–20 kJ mol⁻¹, whereas at higher temperatures (> 300 K), enhanced thermal agitation enables transitions among multiple quasi-degenerate states, giving rise to non-Arrhenius temperature dependence. Nonequilibrium molecular dynamics simulations initiated at the transition state reveal that benzene often relaxes not into a single final state but rather into a distribution of accessible sites. It challenges the assumptions of classical TST and necessitates the integration of kinetic Monte Carlo (KMC) methods and path sampling techniques for accurate modeling.¹⁵⁴

In the NaX zeolite with an extremely low Si/Al ratio (∼1.2), the framework hosts up to 86 negative charges per unit cell, neutralized by an equivalent number of Na⁺ cations. This results in an exceptionally dense electrostatic environment where benzene diffusion is not merely influenced by framework–cation interactions but is also dominated by cooperative migration behavior. Here, collective effects play a central role, arising from intermolecular exclusion and cation-mediated guidance. At low loadings (1–3 molecules per supercage), both self-diffusion (D_self) and Maxwell–Stefan diffusivity (D_MS) increase with loading due to enhanced pathway connectivity facilitated by weak benzene–benzene attraction. However, at higher loadings (≥4.5 molecules per cage), diffusion coefficients sharply decline, reflecting severe steric hindrance and channel crowding. Detailed trajectory analyses further reveal that benzene transport is frequently accompanied by localized rearrangements of Na⁺ ions, forming transient “cooperative gating” configurations in which benzene–Na⁺ complexes dynamically open migration channels. These cooperative jumps are prominent at elevated temperatures but become less significant near ambient conditions, implying that extrapolations from high-temperature MD simulations must be approached with caution and may require mechanistic corrections.¹⁵⁵

To validate these simulation results, Jobic et al. employed neutron spin-echo (NSE) spectroscopy to experimentally measure benzene diffusivity in NaX. The extrapolated D_MS values from MD simulations showed excellent agreement with the NSE data in both magnitude and loading dependence, successfully reproducing the observed non-monotonic trend.¹⁵⁶ However, when these diffusion coefficients were used to estimate membrane permeation fluxes, the predicted values significantly exceeded experimental measurement. This discrepancy is likely due to microstructural imperfections in polycrystalline NaX membranes, such as grain boundaries, electrostatic shielding, or disconnected pore networks.

As above mentioned, FAU zeolites stand out in adsorption and diffusion applications due to their large 12-MR windows and supercage structures, which endow them with exceptional permeability, high capacity, and excellent molecular accessibility. These features make FAU zeolites particularly advantageous for processes involving the treatment of large molecules, complex mixtures, or reactions and separations that demand high diffusion rates. Compared to small-pore and medium-pore topologies, FAU-type zeolites demonstrate superior efficiency and stability under such challenging conditions.

3.2 CHA-type zeolites

In zeolite synthesis, the transformation from FAU- to CHA-type materials involves a complex conversion process. Both FAU and CHA share the double six-ring (d6r) cage as a common structural unit. However, instead of being transferred intact, the d6r undergoes significant rearrangement during the transformation, accompanied by the breaking and reformation of Si–O–Al bonds, resulting in intricate structural reconstruction. The aluminum distribution within FAU plays a pivotal role in determining the final atomic configuration of CHA, leading to a lower energy, preferentially ordered aluminum arrangement in the resulting framework.^157,158 CHA features an ordered 8-MR structure with a small window size, which restricts the access of large molecules. This distinctive characteristic renders it particularly suitable for small molecule conversions, offering exceptional shape selectivity. Such attributes make it an ideal catalyst for reactions demanding precise control over the molecular size and shape.

3.2.1 Gas adsorption on CHA-type zeolites. Fu et al. explored the enhancement of low-concentration CO₂ adsorption through zinc ion modification of CHA-type small-pore zeolites (Fig. 11a and b).¹⁵⁹ Zn²⁺ ions act as key active sites when positioned in the 6-MR within the CHA framework, facilitating effective physical interaction with CO₂ molecules. Adjusting the framework composition to achieve a lower Si/Al ratio increased the density of paired aluminum sites (2Al: Al–O–Si–O–Al), leading to a higher capacity for the zinc ion exchange and a CO₂ adsorption capacity of 0.67 mmol g⁻¹. Additionally, the Zn–CHA zeolites demonstrated strong hydrophobic properties and lower water adsorption energy, making them particularly effective in humid conditions and complex gas streams.

Fig. 11 (a) CHA topology structure. (b) The capacities for CO₂ adsorption in ion-exchanged CHA zeolites. Reproduced with permission from ref. 159. Copyright 2021, Wiley-VCH. (c) Schematic illustration of the sunflower morphology for sunflower-shaped CHA with crystal structure. Reproduced with permission from ref. 162. Copyright 2023, American Chemical Society. CO₂ adsorption isotherms of Na/K/Cs–CHA–M (d) and Na/K/Cs–CHA45 (45 nm) (e) at 298 K. Reproduced with permission from ref. 164. Copyright 2024, Elsevier. (f) The most energetically favorable configuration for CO₂ in the Si–CHA framework, optimized at the PBE-D2 level. Reproduced with permission from ref. 165. Copyright 2012, American Chemical Society.

The morphology can be regulated through synthetic methods to enhance its adsorption capacity. For instance, the CH₄ separation productivity of donut-like K–CHA zeolites in an equimolar CH₄/N₂ mixture was improved to 6939 mL kg⁻¹, with CH₄ purity reaching 84%.¹⁶⁰ Cao et al. developed a unique sunflower-like CHA zeolite superstructure by employing an oriented attachment mechanism of nanocrystals, which combines the inherent microporous features of CHA zeolites with newly introduced mesopores. In CO₂ adsorption tests, the material achieved a capacity of 71.14 cm³ g⁻¹ at 298 K and 0.15 bar, significantly surpassing the 44.44 cm³ g⁻¹ capacity of conventional bulk CHA. Additionally, calculations by IAST revealed CO₂/N₂ and CO₂/CH₄ selectivities of 815 and 264, respectively.¹⁶¹ The exceptional performance of the sunflower-shaped CHA superstructure arises from the improved diffusion, increased surface area, and stronger host–guest interactions. (Fig. 11c). Zhang et al. developed a novel mesoporous In-SSZ-13_(MP) catalyst to enhance the performance of electrochemical CO₂ reduction reaction at low CO₂ concentrations.¹⁶² By combining alkaline etching with ion exchange, indium active sites were uniformly anchored within the mesoporous SSZ-13 framework. This approach significantly increased the catalyst surface area and the dispersion of active metal sites. Chen et al. successfully synthesized CHA-type zeolites without using organic structure-directing agents, achieving significant improvements in CO₂ adsorption performance.¹⁶³ The incorporation of phenol was found to stabilize aluminum dimer intermediates, promoting the CHA nucleation while suppressing MER phase formation. Real-time Raman spectroscopy revealed that phenol increased the proportion of aluminum dimers and optimized the connectivity of Si–O and Al–O bonds. The etching effect of the oxidative medium introduced hierarchical porosity, improving mass transfer properties. The as-prepared CHA zeolite exhibited outstanding performance in CO₂ adsorption tests, achieving a CO₂ uptake of 3.8 mmol g⁻¹ at 100 kPa, with negligible CH₄ adsorption, which highlights the zeolite's distinctive “molecular trapdoor effect”.

Ghojavand et al. investigated the dynamic CO₂ adsorption performance of CHA-type zeolites with varying alkali metal ions (Na⁺, K⁺, and Cs⁺) and crystal sizes in multicomponent gas mixtures, such as CO₂/N₂/He and CO₂/CH₄/He (Fig. 11d and e).¹⁶⁴ The study revealed that reducing the crystal size significantly improved CO₂ separation performance. A nano-sized (45 nm) K–CHA zeolite achieved a selectivity of 108 for CO₂/N₂ and 78 for CO₂/CH₄, with the corresponding CO₂ uptake capacities of 841 mmol g⁻¹ and 721 mmol g⁻¹, outperforming the 500 nm CHA crystals. Smaller crystals provided shorter diffusion pathways, leading to an enhanced mass transfer efficiency within the zeolite particles. Cations with smaller radii, such as K⁺, resulted in larger pore volumes within the framework, enhancing CO₂ adsorption capacity. Additionally, the higher charge density of K⁺ enabled stronger interactions with the CO₂ quadrupole moment, increasing adsorption selectivity.

Fang et al. developed a novel force field for predicting CO₂ adsorption in zeolites, utilizing dispersion-corrected periodic DFT.¹⁶⁵ As shown in Fig. 11f, they analyzed CO₂ adsorption configurations within Si–CHA zeolites and constructed a classical force field (D2FF) through extensive DFT-D2 calculations. By exploring numerous randomly generated adsorption configurations, it identified the 8-MR region within Si–CHA zeolites as the most favorable adsorption site for CO₂, with a calculated adsorption energy of −31.3 kJ mol⁻¹. Further GCMC simulations confirmed the reliability of the D2FF force field, accurately predicting adsorption isotherms and heats of adsorption across various siliceous zeolites, including Si–MFI and Si–DDR.

Environmental factors such as water and NO play a pivotal role in regulating Pd-catalyzed reactions. Mandal et al. revealed that Pd²⁺ ions predominantly occupy 6-MR within the SSZ-13 zeolite framework, with their charge balanced by 2Al sites.¹⁶⁶ At temperatures below 573 K, water molecules solvate Pd²⁺ ions, facilitating their detachment from the framework oxygen atoms and enabling their migration within the zeolite. This process leads to the formation of tetra-coordinated Pd²⁺ complexes, resembling coordination chemistry observed in homogeneous systems, which enhances the catalytic performance of Pd/SSZ-13 for CO oxidation. Under NO-rich conditions, Pd²⁺ ions transition from their 2Al coordination sites to form water-solvated Pd–nitrosyl complexes. At elevated temperatures, these complexes release NO, effectively inhibiting the absorption and oxidation of CO.

However, harmful substances can significantly impair catalytic activity by passivating active sites. Chen et al. demonstrated that phosphorus strongly interacts with [Pd(OH)]⁺ and Pd²⁺ species, driving their relocation from ion-exchange sites within the SSZ-13 zeolite framework to the external surface.¹⁶⁷ This relocation promotes the formation of inactive species, such as PdO_x and metaphosphates. The depletion of active sites caused by this migration and aggregation severely compromises the catalytic efficiency of the zeolite in low-temperature NO_x and CO removal.

Through the study of NO_x adsorption performance in Pd-based CHA zeolites, Li et al. developed Mn/CHA@Pd/CHA core-shell structured zeolite monoliths using coaxial 3D printing technology, providing a novel approach for designing efficient low-temperature passive NO_x adsorbers.¹⁶⁸ The Pd/CHA zeolite shell demonstrated exceptional NO_x capture capability under cold-start conditions while effectively shielding the Mn/CHA core from H₂O interference. Simultaneously, the Mn/CHA core offered efficient storage sites, enhancing NO_x desorption during temperature ramping. Such core-shell structures exhibited comparable NO_x adsorption performance to single Pd/CHA but achieved a 35% increase in desorption efficiency at 350 °C. NO_x adsorption/desorption tests on Pd–CHA with varying Pd loadings (0.5–5.4 wt%) revealed that higher Pd²⁺ loading significantly enhances both NO_x adsorption capacity and desorption temperature. When the distribution of 2Al sites in 6-MR is limited, weak NO_x adsorption becomes more pronounced, whereas a higher proportion of paired aluminum sites in 8-MR promotes a stronger NO_x adsorption. This adsorption behavior is primarily attributed to orbital interactions and charge transfer between Pd²⁺ cations and NO_x molecules. Furthermore, the NO_x adsorption and desorption properties of Pd–CHA zeolites can be precisely tuned by adjusting the Pd loading and the distribution of paired aluminum sites, offering a strategic approach to optimizing passive NO_x adsorption performance.

3.2.2 Adsorption and diffusion of hydrocarbons on CHA-type zeolites. The adsorption of alkanes in zeolites involves two primary interactions: van der Waals forces with the micropore walls, which dominate in zeolites with larger pore sizes, and weak chemical bonding with BAS, resulting in a relatively stable chemisorbed state. Using molecular dynamics simulations, Jiang et al. demonstrated that at lower temperatures (250 K), alkanes preferentially reside at active sites, forming chemisorbed states. However, as the temperature increases to 350 K, the thermal motion of alkane molecules intensifies, significantly reducing the probability of chemisorption at active sites. Instead, the molecules tend to interact more through physisorption with the micropore walls or exhibit free diffusion. For example, the methane chemisorption probability decreases from 50% at 250 K to 22% at 350 K. Similarly, ethane and propane show marked reduction in chemisorption probabilities, though their temperature dependence varies due to differences in molecular size and polarity. Rising temperatures lower the chemisorption energy of alkanes and increase the average distance between alkanes and BAS, enhancing physisorption behavior near pore walls.¹⁶⁹

Denayer et al. investigated the adsorption behavior of C₁–C₁₄n-alkanes on SAPO-34 and Na–CHA molecular sieves, highlighting the interplay of cage confinement and window effects on adsorption thermodynamics and diffusion characteristics.¹⁷⁰ Using pulse chromatographic methods, the research reveals a non-monotonic dependence of the Henry adsorption constant on the alkane chain length. For shorter chains, the adsorption constant increases exponentially with the chain length, but beyond the confinement capacity of the cages, a decrease is observed, followed by a subsequent rise for longer alkanes. It identifies significant configurational changes in alkanes, ranging from coiling within the cages to partial protrusion through the interconnecting windows, as critical factors influencing the adsorption enthalpy and entropy. Particularly on SAPO-34, pronounced discontinuities in entropy changes indicate a transition in the adsorption mechanisms. These findings underscore the importance of molecular-scale interactions between adsorbates and the structural features of zeolites, offering new perspectives for selective separation processes and catalysis. Olson et al. highlighted that Si–CHA zeolites exhibit an exceptionally high diffusivity ratio between propene and propane.¹⁷¹ This remarkable selectivity and diffusion disparity are attributed to its unique three-dimensional channel network and a relatively small pore size.

The MTO reaction represents a highly efficient pathway for converting methanol into light olefins, serving as a pivotal route for producing fundamental chemicals from natural gas, coal, and biomass. In this process, the topology of the zeolite plays a decisive role in governing reactant adsorption, intermediate diffusion, and product distribution.¹⁷² The small size of methanol molecules enables facile entry into CHA cages, where they are adsorbed onto high-density acid sites. This efficient adsorption facilitates intimate interactions between reactants and active sites, thereby accelerating the reaction kinetics. The spatial constraints imposed by the small-pore structure hinder the formation or persistence of large molecular intermediates, favoring the selective enrichment of light olefins (C₂–C₄). Concurrently, the restricted formation and diffusion of bulkier byproducts, such as aromatics or long-chain hydrocarbons, mitigate carbon deposition and extend catalyst longevity.

Light olefin molecules, such as ethylene and propylene, benefit from their small molecular diameters, diffusing rapidly through CHA pores to be efficiently removed from the reaction system. This rapid diffusion enhances both the yield and selectivity of light olefins. In contrast, larger molecular products, including higher hydrocarbons and aromatic macromolecules, face significant diffusion barriers, preventing their accumulation within the zeolite framework. The synergistic interplay of diffusion constraints and shape selectivity in CHA zeolites thus enables unparalleled precision in controlling product distribution during the MTO reaction.

Our group developed a strategy that integrates precoking and steaming to simultaneously enhance the selectivity for light olefins and the catalyst's operational lifespan.¹⁷³ By strategically constructing active naphthalenic species within the core of SAPO-34 crystals, the formation of ethylene and propylene is significantly enhanced. This targeted decomposition and preservation mechanism effectively mitigates rapid coke accumulation on the crystal surface, enhances the accessibility of active sites within the crystal framework, and optimizes the overall catalyst performance. By introducing mesopores and increasing the pore volume, the molecular diffusion rate is significantly enhanced, which reduces catalyst deactivation and prolongs catalyst lifespan, thereby improving the overall performance of the MTO reaction. Fig. 12 presents the adsorption results of methanol and other molecules (such as n-butanol, i-butanol, ethylene, and propylene). The data reveal that the methanol adsorption rate for SSZ-13-F-M25 is notably higher than that for SSZ-13-R and SSZ-13-F, indicating that SSZ-13 samples with more mesopores facilitate faster methanol transport, thereby improving adsorption capacity and catalytic activity. In the adsorption tests of n-butanol and i-butanol, SSZ-13-F and SSZ-13-F-M25 display faster adsorption rates than SSZ-13-R, highlighting the critical role of mesopores in adsorbing larger molecules. Propylene adsorption experiments show that SSZ-13-F and SSZ-13-F-M25 exhibit stronger propylene adsorption than SSZ-13-R, further supporting the idea that mesopore structures improve the transport of reactants.¹⁷⁴

Fig. 12 (a) Methanol, (b) n-butanol, (c) i-butanol, (d) ethylene, and propylene uptake experiments performed at 30 °C on zeolite samples (black) SSZ-13-R (synthesis gel was stirred at room temperature), (green) SSZ-13-F (synthesis gel with sodium fluoride), and (red) SSZ-13-F-M25, which were dehydrated before the adsorption. Confocal fluorescence images of (f) SSZ-13-R and SSZ-13-F crystals collected after the propylene adsorption. Reproduced with permission from ref. 174. Copyright 2016, American Chemical Society.

Surface barriers and intracrystalline diffusion are the two critical factors affecting propane diffusion performance in SAPO-34 zeolites. As shown in Fig. 13a, smaller crystal sizes result in faster adsorption rates, suggesting enhanced surface accessibility or higher surface permeability. Fig. 13b and c indicate that both higher Si content and reduced crystal size correlate with increased adsorption rates. Dual-resistance model (DRM) analysis reveals that the intracrystalline diffusivity of propane remains nearly constant regardless of crystal size, whereas the effective diffusivity decreases significantly as crystal size diminishes (Fig. 13d). Additionally, as the density of external acidity increases, surface permeability declines markedly (Fig. 13e). The surface permeability decreased significantly with the increase of external acid density, suggesting that stronger interactions between surface molecules and acidic sites reinforce surface barriers, thereby impeding propane permeation (Fig. 13f).¹⁷⁵

Fig. 13 (a) The uptake rate, measured by IGA at 313 K (0 to 9 mbar), is shown with experimental data (scattered points) and fitting results (solid lines). (b) and (c) Initial uptake rates for different crystal sizes and Si content, including experimental data (scattered points) and fitting results (dashed lines). (d) Intracrystalline diffusivities obtained from DRM with effective diffusivities represented by half-filled points. (e) Surface permeability variations with SAPO-34's external acidity. (f) The relationship between the characteristic factor L and Henry's constant K for various guest molecules (methanol, ethane, propylene, and propane) at 313 K. Reproduced with permission from ref. 175. Copyright 2019, Springer Nature.

In practical catalytic reactions, surface barriers typically control the rate at which reactants enter the zeolite pores, while intracrystalline diffusion influences the migration of molecules within the pores. A deeper understanding of these two factors can help optimize the design of zeolites to enhance catalytic efficiency. Simply reducing the crystal size may not always improve catalytic performance and may, in fact, increase surface resistance. High surface acidity can strengthen interactions between the surface and molecules, but it also intensifies the surface barrier, limiting reactant diffusion. Introducing mesopores can effectively increase the diffusion rate in zeolites, reduce the impact of surface barriers, and thus improve catalytic performance.

3.3 MFI-type zeolites

The MFI topology comprises two intersecting 10-MR channel systems: one set of channels run linearly along the b-axis, while the other exhibits a zigzag configuration in the orthogonal direction. This unique structural arrangement establishes MFI-type zeolites as some of the most extensively studied and widely utilized zeolite frameworks in both industrial applications and academic research.^176,177

3.3.1 Gas adsorption on MFI-type zeolites. Understanding the adsorption configurations and interaction mechanisms between small molecules and the acid site within zeolites is a significant challenge. Li et al. investigated the adsorption of water vs. methanol on silicalite-1, H- and Na-form MFI compared with mesoporous and non-porous analogues. They identified a substantial contribution of the MFI micropore confinement enhancing the stability of, in particular, protonated clusters at acid sites, crucial adsorbate structures for subsequent chemical reactions.¹⁷⁸ Shen et al. utilized pyridine and thiophene as probe molecules and leveraged in situ integrated differential phase contrast scanning transmission electron microscopy (iDPC-STEM) to directly observe the adsorption and desorption processes of these molecules inside ZSM-5 zeolites.¹⁷⁹ Pyridine molecules can be adsorbed vertically via van der Waals interactions with the skeletal oxygen atoms, while also forming strong charge-transfer interactions in a horizontal orientation with acidic aluminosilicate sites.

Lo et al. utilized synchrotron X-ray powder diffraction (SXRD) to investigate the adsorption behavior of pyridine, methanol, and ammonia on H–ZSM-5.¹⁸⁰ The results revealed that these molecules form acid–base adducts with BAS, i.e. with the primary adsorption sites located at the intersection of the straight and bent channels of H–ZSM-5. The secondary adsorption species of methanol and ammonia preferentially occupy the bent channels, while the second adsorption site for pyridine molecules is predominantly located in the less sterically hindered straight channels. The formation of these adsorption configurations is dictated by a combination of electrostatic interactions between the molecules and the acid sites, along with the geometric constraints imposed by the channel architecture.

Yang et al. investigated the reaction pathways and mechanisms in a tandem catalyst system comprising an Fe-based Fischer–Tropsch catalyst and zeolites for the direct conversion of syngas to aromatics.¹⁸¹ The steady-state isotopic transient kinetic analysis experiments revealed a strong adsorption of CO on the HZSM-5 zeolite. Under reaction conditions, the surface concentration of CO on HZSM-5 was found to be 34 mmol g⁻¹. The introduction of CO significantly enhanced the selectivity towards aromatics. Specifically, in a pure olefin atmosphere, the conversion of C₂H₄ was 81%, whereas in a mixed atmosphere of C₂H₄ and CO, the conversion increased to 84%, with a CO conversion rate of approximately 7%. The adsorption process enhanced the catalytic activity and optimized the pathway for aromatic formation through the synergistic effect of the acid sites.

During high-temperature catalytic reactions, metal particle sintering leads to a reduction in surface area and a loss of active sites, ultimately causing catalyst deactivation. Investigating the sintering mechanism provides insights into optimizing catalytic performance by controlling the location and size of metal particles, enhancing metal utilization and improving selective adsorption and catalytic efficiency. In wet-impregnated samples, larger Pt particles are distributed in mesopores and on the external surface, while smaller particles are confined within the micropores. 2D pressure-jump IR spectroscopy results further reveal the dynamic behavior of CO adsorption, confirming the spatial distribution of Pt particles across different pore structures.¹⁸²

3.3.2 Role of adsorption and diffusion in hydrogenation and hydrocracking reactions on MFI-type zeolites. By strategically selecting or designing tailored zeolite microenvironments, precise control over product selectivity in catalytic reactions can be achieved. For MFI-type zeolites in catalytic adsorption, our research group successfully encapsulated sub-nanometer Pd nanoparticles within their pores, revealing distinct product selectivity in the hydrogenation of furfural.¹⁸³ In the Pd@S-1 catalytic system, furfural was predominantly converted to furan with selectivity exceeding 70%, while the formation of fully hydrogenated products, such as tetrahydrofuran and 2-methyltetrahydrofuran, was effectively suppressed. This result indicates that the confined microporous environment of the S-1 zeolite selectively inhibits over-hydrogenation and promotes decarbonylation to generate furan. Conversely, the Pd@Na–ZSM-5 catalytic system exhibited an exceptionally high selectivity for furfuryl alcohol, achieving selectivity greater than 90%. In the Pd@H–ZSM-5 catalytic system, the product distribution was more complex, including furfuryl alcohol, furan, and 1,5-pentanediol (PDO). This is attributed to the abundant BAS in H–ZSM-5, which facilitates electrophilic attack on the oxygen atom of the furan ring, driving ring-opening reactions and subsequent PDO formation.

Fig. 14 presents a comparison of the diffusion behaviors of C₄ olefins over three distinct morphologies of H–ZSM-5 catalysts (Z-cS, Z-cM, and Z-cL).¹⁸⁴ As the c-axis length increases, the diffusion rate of C₄ olefins significantly improves, indicating that the extent of [010] facet exposure plays a crucial role in molecular transport (Fig. 14c). A higher exposure of the [010] facet correlates with faster diffusion, suggesting that in H–ZSM-5, straight channels contribute more effectively to molecular transport than sinusoidal channels (Fig. 14d). The consistently higher diffusion rate observed for Z-cL compared to Z-cS further confirms that an extended c-axis reduces diffusion resistance and enhances molecular mobility (Fig. 14e and f). As shown in Fig. 14g, the activation energy for diffusion in Z-cS is 26.4 kJ mol⁻¹, whereas in Z-cL, it is reduced to 20.3 kJ mol⁻¹. This lower activation energy in Z-cL indicates a more favorable environment for molecular diffusion, reinforcing the critical influence of crystal morphology on transport properties.

Fig. 14 (a)–(c) Uptake curves of C⁼₄ over H–ZSM-5 samples with different lengths of the c-axis: Z-cS, Z-cM, Z-cL. (d) Correlation of the [010] exposure degrees of H–ZSM-5 zeolites. (e) and (f) Uptake curves of C₄⁼ over Z-cS and Z-cL at different temperatures. (g) Arrhenius plot of diffusion rates over Z-cS and Z-cL. Reproduced with permission from ref. 184. Copyright 2021, Springer Nature.

The competitive adsorption of water molecules effectively regulates the hydrocracking of n-C₁₆ hydrocarbons over MFI zeolites. Brosius et al. reported that Pt/MFI catalysts achieved a remarkable 80% selectivity for linear alkanes at an 80% n-C₁₆ conversion rate. By preferentially adsorbing onto acid sites, water molecules prevent further cracking of primary cracking products and facilitate their desorption from the micropores, thereby increasing the yield of linear alkanes. Nanosheet MFI zeolites, with shorter micropore pathways and a higher proportion of mesopores, demonstrate superior mass transfer efficiency compared to conventional MFI zeolites, enhancing both hydrocracking activity and selectivity. Under humid conditions, water molecules significantly reduce the activation energy of n-C₁₆ hydrocracking, optimizing the primary cracking pathway and improving the efficiency and selectivity of both MFI and nanosheet catalysts. Additionally, low-temperature pretreatment effectively reduces the Pt particle size, further enhancing the catalytic performance.¹⁸⁵ A two-step dry gel conversion method was developed to encapsulate Pt nanoparticles within MFI zeolite crystals. In hydrogenation reactions, this catalyst demonstrated a remarkable activity for the reduction of nitrobenzene. The catalyst exhibited exceptional selectivity in the hydrogenation of nitrostyrene, producing only 4-aminostyrene, in contrast to conventional Pt/ZSM-5 zeolites, which lacked such selectivity entirely.¹⁸⁶

By tailoring the nanoporous environment of zeolites, the catalytic performance of encapsulated rhodium nanoparticles can be effectively optimized. Encapsulating rhodium nanoparticles within silicalite-1 zeolite significantly enhances the CO selectivity while suppressing CH₄ formation. In contrast, HZSM-5, with its pronounced hydrogen spillover effect and strong CO adsorption capacity, preferentially promotes methane production. The acid sites in HZSM-5 facilitate hydrogen spillover and enhance the reactivity of hydrogen species, leading to a high methane selectivity. The strong adsorption properties of HZSM-5 restrict the diffusion of products and intermediates, thereby favoring deeper hydrogenation pathways.¹⁸⁷

The catalytic and adsorption properties of ZSM-5 zeolites exchanged with different non-noble metal ions exhibit pronounced variations. For instance, Mahyuddin et al. demonstrated that the nanoporous framework of ZSM-5 significantly modulates the geometry and electronic environment of metal active centers, thereby markedly reducing the activation barrier for C–H bond cleavage in methane.¹⁸⁸ Among the studied catalysts, CuO⁺–ZSM-5 exhibits a superior performance, achieving the lowest activation energy (6.4 kcal mol⁻¹) and the highest methanol selectivity. In contrast, although FeO⁺ shows relatively low activation energy, its methanol selectivity is diminished due to competing pathways leading to byproducts. Structural optimization results reveal that the position of MO⁺ cations within ZSM-5 is constrained by coordination interactions with framework oxygen atoms, resulting in a distinct bent configuration. This bent geometry weakens the adsorption strength of methane, facilitating efficient activation of the C–H bond.

The design of hierarchical zeolites with interconnected pore structures effectively enhances diffusion and improves catalytic activity. When engineering hierarchical pores, it is crucial to introduce mesopores to ensure effective connectivity between different pore domains, facilitating the efficient transport of reactants to active sites located within micropores. The hierarchical catalyst ZSM-5-A exhibits exceptional micro-mesopore connectivity, as evidenced by the close alignment between its global diffusion time constant and the micropore diffusion time constant. This indicates that the seamless integration of micro- and mesopores significantly improves the accessibility of catalytic active sites. In contrast, the multicomponent catalyst ZSM-5-AR shows a markedly higher global diffusion time constant compared to the micropore diffusion time constant, reflecting the obstruction or poor connectivity of the pore network caused by non-zeolitic components, which restricts the utilization of micropore acid sites.¹⁸⁹ Combining micropore structures with morphology-directed channel orientation could further optimize diffusion performance. Xiao et al. investigated the low-temperature catalytic conversion of polyethylene (PE) over ZSM-5 zeolites.¹⁹⁰ PE undergoes initial cracking on the catalyst surface to generate intermediates, which subsequently diffuse into micropores for further cracking into light olefins. ZSM-5 zeolites with an optimized b-axis thickness of 80–100 nm achieved a remarkable yield of 74.6% for light hydrocarbons (C₁–C₇), with 83.9% of the products being C₃–C₆ olefins (Fig. 15). ZSM-5 nanosheet catalysts efficiently minimize intermediate accumulation on the external surface by reducing the diffusion pathway length, thereby significantly suppressing coke formation.

Fig. 15 (a) Schematic illustration of the carbon flow during the catalytic depolymerization of PE over the s-ZSM-5 catalyst and subsequent product utilization. (b) Variation in average PE molecular weight over time for s-ZSM-5 and n-ZSM-5 catalysts under identical reaction conditions. (c) Time-dependent yields of C₁–C₇ products during PE depolymerization over different zeolite catalysts. Reproduced with permission from ref. 190. Copyright 2022, American Chemical Society.

3.3.3 Role of adsorption and diffusion in the methanol conversion on MFI-type zeolites. Taking methanol conversion as an example, the MFI nanosheets exhibit a significantly increased number of acid sites on their external surfaces. These accessible acid sites create more active regions for reactions involving large molecules, enhancing the catalyst performance. The improved adsorption capacity effectively boosts catalytic efficiency, particularly in reactions with bulky reactants, resulting in higher reaction rates and conversion levels. The reduced crystal thickness notably shortens diffusion pathways, improving the transport efficiency of reactants and products within the micropores. Additionally, this optimized structure minimizes coke formation, with deposition predominantly occurring on the external surface. This external localization of coke significantly reduces the risk of catalyst deactivation, maintaining prolonged catalytic activity.

In the methanol-to-gasoline (MTG) reaction, Shen et al. reported that the c-axis oriented ZSM-5 hollow fibers demonstrated superior methanol conversion rates and enhanced selectivity for C₅⁺ hydrocarbons compared to conventional ZSM-5 zeolites (Fig. 16a–c).¹⁹¹ Additionally, they exhibited improved resistance to coke deposition, further extending their catalytic lifetime. In the application of MFI zeolites for methanol-to-hydrocarbon (MTH) reactions, the introduction of acetaldehyde initiates multistep aldol condensation reactions on BAS, leading to the formation of higher aromatic compounds. These aromatics are strongly adsorbed within the zeolite pores, enhancing the propagation of the aromatics-based catalytic cycle. Through subsequent methylation and dealkylation reactions, these aromatic intermediates facilitate ethene production while suppressing side reactions that generate other hydrocarbons. Conventional MFI zeolites with their longer diffusion pathways (∼250 nm) allow extended residence times for aromatic species, favoring the continuation of the aromatics-based cycle and resulting in higher ethene selectivity. In contrast, self-pillared pentasil MFI zeolites, characterized by much shorter diffusion pathways (∼1.5 nm), exhibit reduced diffusion constraints for aromatics, thereby diminishing the effectiveness of the aromatics-based catalytic cycle (Fig. 16d).¹⁹² Cation-exchanged ZSM-5 zeolites enable the simultaneous conversion of methane and methanol. Methanol is activated at BAS to form oxygen ion intermediates, while methane undergoes C–H bond activation facilitated by metal oxides, generating methyl intermediates. These methyl intermediates synergistically react with oxygen ion intermediates, producing aromatics and higher hydrocarbons. Compared to conventional H–ZSM-5, H–GaAl–ZSM-5 demonstrates superior methane conversion efficiency and enhanced selectivity for C₇–C₁₂ aromatic hydrocarbons, while minimizing the excessive formation of toxic benzene.

Fig. 16 (a) Schematic illustration of the growth mechanism of the ZSM-5-CHF (hollow fibers) sample. (b) Methanol conversions over conventional ZSM-5 and ZSM-5-CHF as a function of time on stream. (c) Product selectivity of MTG reactions over the conventional ZSM-5 (red) and the ZSM-5-CHF (blue) after 2 h time on stream. Reproduced with permission from ref. 191. Copyright 2013, American Chemical Society. (d) Inset: (a) the variation of net carbon conversion (blue) over time-on-stream, and (b) the selectivity trends for ethene (orange), propene (blue), C₄–C₈ aliphatics (green), methylbenzenes (purple), and C₉+ hydrocarbons (red) as a function of net carbon conversion during the catalytic transformation of acetaldehyde over (i) conventional MFI (Conv MFI) and (ii) SPP MFI. Reproduced with permission from ref. 192. Copyright 2016, American Chemical Society.

In isomerization and alkylation reactions, the microporous structure of MFI zeolites facilitates the formation of target products through molecular sieving effects, while mesoporous modifications further enhance diffusion efficiency and accelerate reaction rates. In biomass sugar conversion, metal-modified MFI zeolites leverage the synergistic effects of framework LAS and BAS, demonstrating exceptional catalytic performance in sugar isomerization and dehydration reactions.^193–197 By optimizing pore structures and surface chemistry, MFI zeolites can accommodate various molecular sizes and reaction requirements, offering versatile solutions for industrial catalysis.

3.4 MOR-type zeolites

Introducing small organic electron withdrawing reagents into the microporous walls of zeolites can enable uniform modification and precise control over the effective pore diameter within the micropores. Le et al. reported that on MeOPh-modified MOR-type zeolites, the adsorption capacity of ethylene reaches 1.1 mmol g⁻¹ at 30 °C and 100 kPa, while ethane adsorption is only 0.021 mmol g⁻¹. For propylene, the adsorption capacity is 0.58 mmol g⁻¹ at 30 °C and 100 kPa, while for propane it is only 0.028 mmol g⁻¹, demonstrating high selectivity for propylene/propane separation as well. The isosteric heats of adsorption for ethylene and propylene, in the coverage range of 0.1–1 mmol g⁻¹, were found to be in the range of 38–12 kJ mol⁻¹, which falls within the typical physical adsorption.¹⁹⁸

Effectively leveraging adsorption properties enables precise modulation of catalytic performance. Pyridine molecules were applied to detect BAS at pore mouths of the 8-MR side pockets of H-MOR zeolites, interacting with Brønsted hydroxyls within these confined spaces. This adsorption behavior is closely associated with framework defects in H-MOR, which effectively enlarge the pore openings of the 8-MR side pockets. At controlled pyridine desorption temperatures of 300–400 °C, the H-MOR zeolite with a Si/Al ratio of 13.8 exhibits exceptional carbonylation performance, achieving a methyl acetate production rate as high as 7.2 mmol (h g)⁻¹.¹⁹⁹

To enhance the accessibility of micropores, the synthesized mordenite was subjected to oxalic acid treatment followed by ion exchange to produce H-MOR.²⁰⁰ The oxalic acid treatment effectively removed Na⁺ cations obstructing the channels, while the subsequent ion exchange and calcination further improved the pore accessibility. These modifications were achieved without introducing mesopores or altering the Si/Al ratio and crystal morphology. Adsorption experiments with n-butane and isobutane demonstrated a significant increase in adsorption capacity after the treatments. Notably, n-butane exhibited a higher saturation capacity, while isobutane showed a greater heat of adsorption due to its stronger interactions with the channel walls.

Three distinct migration behaviors of the acetyl cations (CH₃–C≡O+) within different channels of mesoporous zeolite may occur. In the absence of reactants, the migration of the acetyl cations from the side pockets to the 12-MR channels requires overcoming an energy barrier of 57.9 kJ mol⁻¹, accompanied by an energy compensation of 30.8 kJ mol⁻¹, indicating that this process is unfavorable both kinetically and thermodynamically. However, in the presence of dimethyl ether molecules in the 12-MR channels, long-distance attraction (approximately 5 Å) reduces the migration barrier to 47.9 kJ mol⁻¹, leading the system toward the thermodynamic equilibrium. When methanol is present, the barrier further decreases to 38.8 kJ mol⁻¹, making the migration process favorable in both kinetic and thermodynamic terms. Initially, the acetyl cations are formed via carbon–carbon bond coupling in the 8-MR channel. These cations are then stabilized in the side pockets, gaining a higher migration activity. Finally, they undergo a carbon–oxygen bond coupling with the reactant in the 12-MR channels, leading to the formation of the target product, methyl formate, and effectively preventing the accumulation of active intermediates within the channels, which could lead to coke formation (Fig. 17a–d).²⁰¹

Fig. 17 (a) Pore and channel sizes of mordenite, along with the kinetic diameters of methanol, dimethyl ether, and methyl acetate. The 8-MR window connects the 12-MR channel to the 8-MR side pocket (SP) and catalytic cycle of DME carbonylation. (b) Free energy profile of MOR-8-MR, MOR-12-MR, and MOR-SP at 473 K using the PBE-D3/dgdzvp method. (c) Energy decomposition analysis at TS3 in MOR-8-MR and MOR-12-MR. (d) Host–guest interactions at TS3 visualized via a reduced density gradient, with repulsion in red and attraction in green. Reproduced with permission from ref. 201. Copyright 2022, Springer Nature. (e) Comparison of volumetric CO₂ uptake at 298 K. (f) Comparison of CO₂/N₂ and CO₂/CH₄ IAST selectivities at 1 bar and 298 K. (g) Experimental column breakthrough curves for dry and humid CO₂/N₂(CH₄) separation. (h) Recyclability under humid CO₂/CH₄ column breakthrough tests. Reproduced with permission from ref. 204. Copyright 2021, AAAS.

Chemical etching of MOR-C using an HF solution effectively mitigated surface resistance, significantly enhancing mass transfer properties and catalytic stability, bringing its performance close to that of MOR-T.²⁰² Furthermore, the study identified that surface resistance originates from localized structural disorder on the crystal surface and emphasized the critical influence of template agents used during synthesis on the resultant crystal morphology, surface structure, and diffusion properties of MOR zeolites. The highly ordered alignment of the unique 12-MR channels of MOR zeolites along the c-axis results in a significant morphology-dependent impact on molecular diffusion. Using n-butane as a probe molecule, the apparent diffusivity decreases dramatically from 22.7 × 10⁻³ s⁻¹ in nanometer scale crystals to 4.2 × 10⁻³ s⁻¹ in micrometer scale counterparts. Similarly, in catalytic performance, nanometer sized Ti-MOR zeolites achieve an impressive cyclohexanone conversion rate of 94.4%, in strong contrast to just 16.6% for larger micrometer sized crystals.²⁰³

Zhou et al. investigated the adsorption and separation performance of MOR. Fe-MOR (0.25), synthesized by self-assembled strategy, exhibited a significantly higher CO₂ uptake compared to conventional zeolites and metal–organic frameworks, while maintaining superior selectivity (Fig. 17e). Moreover, this material demonstrated exceptional water stability (Fig. 17f), retaining its high CO₂ separation efficiency even under humid conditions (Fig. 17g and h).²⁰⁴

3.5 *BEA zeolites

Incorporating Pd into the micropores of BEA zeolites also results in remarkable selective adsorption properties. Zhang et al. demonstrated that Pd@beta exhibits a pronounced preference for adsorbing nitroaromatic compounds while significantly suppressing the adsorption of other functional groups, such as chloro groups (Fig. 18a).²⁰⁵ The unique architecture of Pd@beta encapsulates Pd nanoparticles entirely within the zeolite micropores, ensuring that target molecules are subjected to stringent spatial screening and directional alignment during diffusion, thereby enabling nitro groups to preferentially interact with the active metal sites. Patet et al. compared the adsorption capabilities of H–BEA zeolites substituted with different ions.²⁰⁶ H–[B]–BEA with its weaker acidity allows furan molecules to desorb more readily, resulting in higher selectivity and stability (Fig. 18b). H–[Al]–BEA zeolites exhibit a stronger adsorption strength and higher catalytic activities but more severe catalyst deactivation due to side reactions. The catalytic adsorption selectivity varies depending on the different acid sites. Kruger et al. found that BASs of BEA zeolites effectively catalyze the dehydration of fructose to HMF, while LASs, derived from octahedral aluminum species, facilitate the isomerization of fructose to glucose and the conversion of HMF into secondary byproducts.²⁰⁷ H–BEA-18 exhibits strong selective adsorption properties for HMF, furfural, and levulinic acid compared to fructose and glucose, making these strongly adsorbed products more susceptible to secondary reactions. The hydration of HMF produces acidic byproducts, such as formic acid, which can dissolve the framework of aluminosilicates, generating homogeneous catalytic species that further accelerate undesirable side reactions. To mitigate the strong adsorption of HMF and suppress side reactions, methyl isobutyl ketone was introduced as an organic extraction phase, significantly enhancing the selectivity for HMF and levulinic acid.

Fig. 18 (a) Substrate conversions and product selectivities for the hydrogenation of 4-nitrochlorobenzene and 4-nitrobenzaldehyde on various catalysts. Adsorption models for 4-nitrochlorobenzene on Pd/C and Pd@beta. Reproduced with permission from ref. 205. Copyright 2017, Wiley-VCH. (b) Electrostatic potential mapped onto the electron density isosurface with a value of 0.05 a.u. for the active-site-coordinated oxanorbornene intermediates and C–O cleavage barrier as a function of the density and the negative of the Laplacian of the density evaluated at the C–O bond critical point. Reproduced with permission from ref. 206. Copyright 2017, American Chemical Society. (c) Lewis acid site densities of Sn–beta, Ti–beta, and Zr–beta probed by FTIR spectroscopy of pyridine adsorption and evacuation at different temperatures. Reproduced with permission from ref. 208. Copyright 2014, American Chemical Society. (d) Characteristic diffusion time constants from uptake rates of 2,2-dimethylbutane for beta-NS and beta-C. Reproduced with permission from ref. 213. Copyright 2025, Wiley-VCH.

Using an organometallic precursor, Sn species can be incorporated into the Si–beta zeolite framework by reacting with vacant T-sites and subsequent calcination. When Sn–beta zeolites are employed as catalysts in the ring-opening hydration of epoxides to produce 1,2-diols under mild, solvent-free conditions with a H₂O/epoxide molar ratio of 2 : 1, they exhibit remarkable catalytic activity and selectivity. Compared to H–beta, Sn–beta shows superior catalytic performance, with a lower apparent activation energy of 34.8 kJ mol⁻¹, in contrast to H–beta's 55.5 kJ mol⁻¹. The catalytic activities of Zr–beta and Ti–beta were systematically evaluated under identical conditions and found to correlate strongly with their Lewis acidity, following the order Sn–beta > Zr–beta > Ti–beta, as determined by FTIR spectroscopy of pyridine adsorption.²⁰⁸ Roy et al. investigated the Lewis acidity and adsorption behavior of framework Sn sites in Sn–beta zeolites using three probe molecules: acetonitrile, diethyl ether, and tert-butyl alcohol.²⁰⁹ The framework Sn sites form a 1 : 1 adsorption stoichiometry with each probe molecule, indicating the presence of uniform, well-defined LASs. These Sn sites facilitate the selective dehydration of tert-butyl alcohol, producing butenes and water within a narrow temperature window around 410 K, highlighting the elevated catalytic activity of Sn as a Lewis acid. Moreover, the framework Sn sites exhibit a strong affinity for water, which effectively suppresses acetonitrile adsorption, while exerting negligible influence on the adsorption of alcohol-functionalized species. This behavior further emphasizes the distinctive role of framework Sn sites in mediating selective interactions with specific adsorbates.

Kwon et al. investigated the effects of solvent–pore interactions on the vapor-phase epoxidation of alkenes over Ti–BEA zeolite catalysts, demonstrating how the reorganization of intrapore solvent molecules modulates reaction kinetics and selectivity.²¹⁰ Their findings revealed that the density of (SiOH)_x defects in Ti–BEA significantly influences the condensation behavior of solvents, such as acetonitrile (CH₃CN), within the zeolite pores, which in turn impacts the stability of transition states and the rate of 1-hexene epoxidation. Infrared spectroscopy and dynamic vapor sorption analysis indicated that CH₃CN molecules condense to near-liquid densities (up to 0.3 g cm⁻³) within Ti–BEA zeolites with higher (SiOH)_x densities. This condensation effect provides additional enthalpic and entropic contributions to the stabilization of epoxidation transition states through van der Waals forces and hydrogen bonds, resulting in a 20-fold increase in the reaction rates and a twofold improvement in the epoxide selectivity. In contrast, the most hydrophobic Ti–BEA–F catalyst exhibited a significantly weaker CH₃CN condensation, lower adsorption capacities, and the slowest reaction rates.

The interactions between solvent and pore structures in Ti–BEA zeolites were investigated with respect to their effects on the adsorption and catalytic behavior of 1,2-epoxyoctane (C₈H₁₆O).²¹¹ The study revealed that the density and distribution of silanol nests ((SiOH)₄) within Ti–BEA significantly influence the thermodynamic parameters of adsorption during catalysis. In systems containing trace amounts of water, Ti–BEA with a higher density of silanol nests exhibited an increase of 19 kJ mol⁻¹ in adsorption enthalpy (ΔH_Ads) and 75 J mol⁻¹ K⁻¹ in adsorption entropy (ΔS_Ads) for C₈H₁₆O compared to materials with fewer defects. This indicates that water molecule reorganization causes an additional free energy during adsorption. The adsorption of C₈H₁₆O was found to significantly disrupt the hydrogen-bond network in hydrophilic Ti–BEA zeolites, while such disruption was not observed in more hydrophobic Ti–BEA materials, suggesting that water molecules within hydrophobic pores have minimal interactions with the reactive intermediates at active Ti sites. These findings highlight the distinct roles of solvent–pore interactions in shaping adsorption.

The polarity of the BEA framework and the choice of solvent play a pivotal role in modulating the catalytic performance.²¹² In hydrophilic frameworks, silanol nests form hydrogen bonds with substrate molecules, significantly increasing the adsorption enthalpy and activation energy barriers. In contrast, hydrophobic frameworks tend to minimize such hindrances. Adsorption isotherm experiments for EP revealed that its uptake in toluene is substantially higher than in acetonitrile, indicating that non-polar solvents promote substrate enrichment within the pores. Conversely, the higher polarity of acetonitrile favors competitive occupation of active sites, thereby suppressing the reaction.

Synthesizing sheet-like beta zeolites remains a significant challenge. Recently, our group reported a novel intergrown B–C structured beta zeolite nanosheet (beta-NS) synthesized using a polycationic organic structure-directing agent (OSDA) in a mild hydroxide medium.²¹³ This strategy effectively circumvents the conventional A + B intergrowth commonly observed in beta zeolites, in which polymorph A introduces tortuous micropore channels. Instead, through a specially designed polycationic OSDA, the resulting nanosheets are composed exclusively of straight 12-MR channels by selectively stabilizing polymorphs B and C (BEC), leading to enhanced molecular diffusion. Beta-NS features a distinctive “house-of-cards” architecture with a thickness of approximately 16 nm, substantially enhancing molecular diffusion, as evidenced by a 35-fold increase in diffusion efficiency compared to conventional beta zeolite (beta-C) (Fig. 18d). Catalytic evaluation in n-heptane cracking at 873 K demonstrates that beta-NS achieves a significantly higher selectivity towards light olefins (C₂–C₄), while exhibiting an extended catalyst lifetime with a turnover number (TON) of 8.43 × 10⁴. The strong interaction between the OSDA and the zeolite framework facilitates the preferential localization of Al sites within the straight channels, leading to increased Brønsted acidity and enhanced catalytic stability.

3.6 Other zeolites

Smit et al. confirmed the adsorption sites of n-alkanes in the FER zeolite, revealing that the chain length significantly influences both the adsorption location and strength.²¹⁴ Specifically, longer-chain n-alkanes preferentially adsorb towards the center of the zeolite channels, while shorter chain n-alkanes are more likely to be adsorbed near the channel entrances. Kärger et al. revealed a concentration-dependence of the methanol diffusion in 8-MR FER channels. Fick's second law was applied, and a surprisingly good match between experimental and theoretical data was achieved.²¹⁵

Cordero-Lanzac et al. compared the reaction kinetics of MAPO-18 under MTO conditions.²¹⁶ MgAPO-18 exhibited more stable methanol conversion activity at 1 bar, while SAPO-18 underwent rapid deactivation under the same conditions. When high-pressure CO/H₂ was introduced, the performance of MgAPO-18 was further enhanced, whereas SAPO-18 continued to deactivate rapidly. This improvement in MgAPO-18 is attributed to the inhibition of olefin hydrogenation and coke formation by CO and H₂, which enhanced its stability. Both MAPO-18 catalysts were also tested in the CO₂ hydrogenation to hydrocarbons process, in combination with a ZnO–ZrO₂ mixed oxide catalyst. During CO₂ hydrogenation, MgAPO-18 and SAPO-18 exhibited a similar catalytic performance, partly due to the influence of water and methanol concentrations during the reaction, which diminished the differences between the two catalysts. During the initial stages of the MTO reaction, the two catalysts displayed distinct reaction mechanisms that MgAPO-18 predominantly underwent a decarbonylation reaction, while SAPO-18 was more prone to decarboxylation.

To fabricate mesoporous structures, Cho et al. successfully engineered hierarchically structured LTA zeolites featuring interconnected three-dimensional mesoporous networks integrated within the microporous framework, which demonstrates remarkable enhancement in xenon diffusion kinetics, exhibiting an increase in the diffusion rate compared to conventional LTA zeolites.²¹⁷ Moreover, the mesopore dimensions can be precisely tuned through systematic modulation of the surfactant concentration or through the strategic addition of triblock copolymers as pore-expanding agents, enabling expansion of pore diameters up to 24 nm. The enhanced molecular transport facilitated by the mesoporous network promotes rapid diffusion of dimethyl ether intermediates, resulting in superior catalytic performance characterized by elevated methanol conversion efficiency, reduced hydrocarbon selectivity, and significantly extended catalyst lifetime.

In low Si/Al ratio MER zeolites, CO₂ adsorption operates through the “cooperative cation gating-breathing mechanism”.²¹⁸ The interplay between the structural flexibility of the framework and the dynamic migration of extra-framework cations is critical in facilitating the CO₂ capture. As CO₂ molecules interact with cation-blocked windows, these interactions drive the repositioning of cations and induce pronounced breathing-like structural adjustments in the zeolite framework, enabling selective adsorption.

The positioning of Al atoms within the zeolite framework plays a crucial role in determining the catalytic properties. Using solid-state NMR and synchrotron X-ray diffraction techniques, Berkson et al. revealed that 94% of the Al atoms are located on the surfaces of the large pores, while only 6% are found in the sub-nanometer-sized intralayer channels of the Al-SSZ-70 (MWW) zeolite.²¹⁹ The accumulation of Al atoms on the surface of the large pores enhances the adsorption of reactant molecules within the channels and facilitates the diffusion process. For catalytic cracking reactions, the larger pore size and the surface accumulation of Al heteroatoms provide a more efficient diffusion path for larger molecules, thus improving the catalytic efficiency. In contrast, the Al atoms located in the sub-nanometer channels cause limited diffusion pathways, preventing larger molecules from entering these pores and resulting in slower diffusion rates.

Surface hydrophilicity and surface defects play a critical role in the water transport through zeolite membranes. Measurements of diffusion coefficients under varying conditions, such as temperature, hydrophilicity, and surface roughness, revealed that an increase in the surface hydrophilicity significantly enhances the diffusion coefficient of water molecules. Under low pore hydration conditions, enhanced surface hydrophilicity strengthens hydrogen bonding between water molecules and the framework, thereby restricting molecular mobility and resulting in a reduced diffusion coefficient. In contrast, at high hydration levels, water–water interactions dominate, forming a continuous hydrogen-bonded network that facilitates molecular transport. In this regime, the influence of surface–water interactions diminishes, leading to an overall increase in diffusivity. However, surface defects, including irregular pore sizes and uneven surfaces, notably hinder the diffusion process. These defects introduce additional barriers that hinder the unobstructed passage of water molecules through the zeolite membrane. The irregularity in pore geometry, including variations in pore size and surface roughness, can result in partial blockages or pore constriction, which forces water molecules to follow tortuous and longer diffusion pathways. This, in turn, increases the resistance to molecular transport. As a consequence, the free diffusion of water molecules is impeded, undermining the overall efficiency of mass transfer. Ultimately, this leads to a reduction in membrane permeability, as the water flux is diminished not only by the slower diffusion rates but also by the additional resistance imposed by surface imperfections and pore restrictions.²²⁰

Liu et al. first proposed the innovative “continuum-intersecting-channels” mechanism, which essentially guides the design of porous materials with high-throughput diffusion capabilities.²²¹ In the SCM-15 (SOV) zeolite, the z-direction channels characterized by strong adsorption sites and sufficient molecular rotation space facilitate the rapid diffusion of p-xylene molecules t low loading. However, as the loading increases, molecules preferentially diffuse along the x-direction due to the significantly reduced diffusion barriers provided by the continuum intersecting channels. In the x-direction, weak adsorption regions minimize resistance, making it favorable for high-loading diffusion. Conversely, while the strong adsorption sites in the z-direction aid in the initial positioning of molecules, the diffusion barriers progressively increase with loading, ultimately slowing down the diffusion rate. This interplay between adsorption strength and diffusion resistance highlights the critical role of the channel structure in governing molecular transport.

4. Influencing factors of the adsorption and diffusion on zeolite catalysts

In the discussion of the preceding sections, it has been observed that the catalytic performance of zeolites, regardless of their topology, the catalytic reaction type, or other kinds of porous materials, is typically influenced by several factors. Therefore, in this section, we will provide a detailed examination of these influencing factors.

4.1 Acid sites as influencing factors

The strength of acid sites plays a critical role in determining molecular adsorption energies within pores and the ease of desorption. Strong acid sites can significantly enhance the chemisorption, establishing stable interactions between the adsorbates and the material, thereby improving the molecular capture efficiency. However, excessively strong adsorption may impede diffusion by increasing the transport resistance within confined pores, ultimately limiting overall mass transfer efficiency. Thus, the acidity dictates the adsorption behavior and further indirectly governs the molecular transport and catalytic performance by influencing diffusion barriers and reaction pathways. Through precise tuning of the strength and spatial distribution of acid sites, a dynamic balance between adsorption and diffusion can be achieved, unlocking enhanced catalytic efficiency.

EFAl can modulate the acid strength of BAS. As shown in Fig. 19a–c, upon PA adsorption, the ZY60 sample predominantly exhibited the interactions between teterahedral AL^IV sites and PA molecules. Due to the low EFAl content in ZY60, PA interacts primarily with framework aluminum Al^IV sites. In contrast, the ZY5S7 sample, which contains a higher concentration of EFAl species, displays distinct interactions between PA and not only the Al^IV sites but also pentacoordinate Al^V and octahedral Al^VI aluminum species. These findings suggest that EFAl species engage in significant interactions with the amine, methylene, and methyl protons of PA. The synergistic effect between EFAl species and BAS likely contributes to the enhanced catalytic activity of the ZY5S7 sample. In catalytic cracking reactions, the synergistic interaction between EFAl and BAS significantly enhances the acid strength and catalytic activity of the active sites. Conversely, in hydrocracking reactions, the catalytic performance is predominantly determined by the quantity of BAS, with little dependence on the presence of EFAl or its proximity to BAS.²²² By precisely tuning the location and concentration of EFAl, it is possible to design catalyst systems with tailored product selectivity, enabling more efficient and targeted catalytic processes. The surface LAS and Si–OH groups in zeolites can work synergistically to enhance catalytic activity. Josephson et al. proposed that Sn–O–Si–OH moieties in Sn–SPP facilitate the stabilization of the transition state in the fructose etherification reaction through hydrogen bonding interactions.²²³ In contrast, the Sn–beta zeolite, which lacks adjacent Si–OH groups, cannot provide such stabilization, leading to weaker catalytic activity. The adsorption of fructose on Sn–SPP is more favorable for etherification, whereas Sn–beta tends to favor the isomerization of fructose.²²⁴ Furthermore, Sn–SPP exhibits significantly higher selectivity for the formation of ethyl fructoside compared to Sn–beta. The binding of framework tri-coordinated aluminum with hydroxyl groups exhibits high reactivity for methoxylation, facilitating the conversion of methanol at low temperatures through the formation of methoxy species. This process reveals the interaction between methanol molecules and the framework Al sites, and these methoxy species can further transform into hydrocarbons at elevated temperatures, serving as key C1 species in the reaction.

Fig. 19 (a) ²⁷Al DE and 2D ¹H–²⁷Al CP LG HETCOR NMR results for the ZY60 and ZY5S7 samples upon PA adsorption. Schematic illustrations of the interaction of PA protons with Al sites of the zeolite (b) ZY60 and (c) ZY5S7. Reproduced with permission from ref. 222. Copyright 2024, American Chemical Society. Integrated reflection intensities of experimentally observed H-ZSM-5 (d) absolute intensity and (e) change in intensity relative to lowest incident energy 1540 eV. arb., arbitrary. (f) Rietveld refinement against experimental diffraction data at 1569 eV. (g) The overall refined Al occupancy from the five energies with the constraint on total occupancy. (h) The refined framework Al locations at T4 (sinusoidal channel), T6 (intersection void), and T8 (straight channel) are viewed along the [010] and [100] projections. Reproduced with permission from ref. 227. Copyright 2025, AAAS.

In 2005, Peng et al. successfully employed the ¹⁷O MAS NMR experiment to detect oxygen signals directly associated with the BAS, and these signals were further confirmed using ¹H–¹⁷O double-resonance NMR techniques.²²⁵ By subjecting the zeolite samples to various temperature treatments and the adsorption of hydrogen-bonding substances, changes in the local environment of the acid sites were observed, specifically manifested as an increase in the O–H bond length and alterations in the geometric structure of the acid sites. Yu et al. employed high field ²⁷Al DQ-MAS NMR spectroscopy to investigate the evolution of aluminum species during the dealumination process of HY zeolite, which reveals interactions between FAL and EFAL species.²²⁶ The formation of EFAL species is closely linked to temperature, highlighting the temperature-dependent nature of this transformation. Controlling the distribution of aluminum within zeolites to regulate their acidity and achieve a controlled tuning of acid sites has long been a challenging task. In commercial H–ZSM-5 (Si/Al = ∼17) synthesized under specific conditions, the distribution of aluminum sites was found to be highly non-random, primarily concentrated at the T8, T6, and T4 positions.²²⁷ The T8 site is located in the straight channel, while the T6 and T4 sites are situated in the intersecting and sinusoidal channels, respectively. Additionally, the T6 and T4 sites form aluminum pairs (AlT6–O–SiT5–O–AlT4). Through Rietveld refinement and ammonia adsorption isotherm analysis, the occupancy rates of these aluminum sites were quantified, revealing that the T8 site has the highest aluminum occupancy (0.320), while T6 and T4 sites have lower occupancy rates of 0.116 and 0.244, respectively. Ammonia molecules adsorb singly at the T8 site, while at the T6 and T4 sites, they form bridged adsorbates. This indicates that the aluminum pairs at T6 and T4 sites may play a synergistic role in catalysis, facilitating bimolecular reactions (Fig. 19d–h).

Liu et al. employed DFT to investigate the variations in the acidity of FAU, CHA, and MFI zeolites at different Si/Al ratios and under extra-framework cation modifications.²²⁸ Acidity, characterized by acid site density, acid site strength and accessibility, is usually investigated by applying common probe molecules for zeolites, such as ammonia, pyridine, and acetonitrile.²²⁹ Within the same zeolite topology, strong correlations were observed between the acidity, adsorption energies, and reaction activation energies. However, these scaling relationships do not hold across different framework topologies, as they are significantly influenced by van der Waals interactions and steric effects. Boronat and Corma systematically investigated the interactions of probe molecules with BAS in MOR and MFI zeolites, uncovering the profound influence of confinement effects from the microporous structure on acidity measurements.²³⁰ For weakly basic probe molecules such as CO, the formation of neutral ZH–CO hydrogen-bonded complexes requires a specific spatial orientation. In zeolite frameworks with narrow or irregular micropores, this configuration may be distorted due to geometric constraints, leading to a reduction in interaction strength. However, when the molecule tightly fits within the pore structure, dispersion forces are significantly enhanced, resulting in stronger adsorption. In contrast, strongly basic molecules like NH₃ and pyridine are protonated at the acid sites, with the resulting cations stabilized by multiple interactions with the framework oxygen atoms, forming stable Z⁻–BH⁺ ion pairs, which substantially influence the acidity measurements. The confinement effect in acidity measurements depends on the geometric matching between the molecules and the pore structure, and the stabilizing role of framework oxygen atoms and spatial repulsive forces. These effects explain the variation in the acidity of the same acid site depending on its position or the specific framework structure. When the H-zeolite samples are heated to 750 °C, the generation of hydrogen indicates the dehydrogenation of the BAS. This process begins with the dissociation of the O–H bond at the BAS, followed by the diffusion of hydrogen atoms between the framework oxygen atoms, eventually recombining to form molecular hydrogen. During this process, the BAS transforms into [AlO₄]⁰ sites, which act as non-acidic one-electron acceptors and interact with adsorbed molecules. These sites may become active centers for hydrocarbon catalytic reactions at high temperatures.²³¹ Notably, the solid state NMR technique in combination with suitable probe molecules with desirable and observable NMR-sensitive nuclei has been demonstrated to be a reliable and practical approach for qualitatively and quantitatively characterizing the detailed acidic properties (e.g., type (Brønsted vs. Lewis acidity), distribution, strength, concentration, and spatial interactions) of acid sites in various zeolite catalysts in terms of its uniqueness and unprecedented spectroscopic resolution and sensitivity, particularly in conjunction with advanced 2D homo-/heteronuclear correlation spectroscopy.^232–235 For example, by combining DFT calculations with ³¹P NMR using TMP as the probe molecule, Yi et al. successfully disclosed the origin, fine structure, and possible location of tri-coordinated Al³⁺ species with ultra-strong Lewis acidity in a dealuminated HY zeolite.²³⁶ Our group employed solid-state NMR and FTIR to investigate the structural evolution and acidity of partially framework-coordinated aluminum species (Al(IV)-2) in zeolites using TMP as the probe molecule. Unlike conventional BAS and LAS, Al(IV)-2 exhibits strong adsorption interactions with TMP via Al–OH groups. 2D ³¹P–³¹P correlation NMR experiments revealed that the spatial proximity between Al(IV)-2 and BAS is significantly shorter than that between BAS and LAS, suggesting a potential synergistic effect between Al(IV)-2 and BAS in catalytic processes.²³⁷ Additionally, 1,2,4-trimethylbenzene (1,2,4-TMB) was selected as a probe molecule due to its molecular size (0.74 nm),²¹ which exceeds the diameter of the 10-MR channels in MFI zeolites but remains smaller than the channel intersections. This allows for a systematic investigation of its diffusion and dynamic behavior under different pore confinement conditions. In siliceous silicalite-1, which lacks acid sites, 1,2,4-TMB predominantly diffuses along the straight channels and preferentially resides at the channel intersections in well-defined orientations. However, in H–ZSM-5, the presence of BAS induces strong adsorption of 1,2,4-TMB, further restricting its confined motion. This pronounced interaction significantly alters the adsorption and diffusion properties of aromatic species within the zeolite framework, potentially influencing the behavior of methylated benzene intermediates in catalytic processes, either as active centers or as precursors to catalyst deactivation.²³⁸

To understand the regulatory effect of extra-framework silica (EFSi) on the catalytic performance of BAS in H–ZSM-5 zeolites, Zhao et al. found that EFSi incorporation does not alter the intrinsic acidity of BAS but significantly enhances their adsorption capacity for basic molecules through van der Waals interactions, particularly for probe molecules with alkyl chains or aromatic rings, such as pyridine and N-methylpropylamine.²³⁹ EFSi-modified BASs exhibit markedly improved catalytic activity in n-pentane cracking, with activation enthalpies for the C₁ + C₄⁼ (P1), C₂ + C₃⁼ (P2), and C₃ + C₂⁼ (P3) pathways reduced by 51, 24, and 42 kJ mol⁻¹, respectively (Fig. 20a). This effect is primarily attributed to enthalpic stabilization of the transition state, which facilitates the formation of carbenium ions and lowers the reaction energy barrier. Unlike EFAl, which enhances catalytic performance through entropic stabilization, EFSi creates a more confined local environment that optimally stabilizes the transition state (Fig. 20b and c). This study proposes a novel strategy for improving zeolite-based catalysis by tailoring the microenvironment of active sites rather than modifying their intrinsic acid strength, offering new insights into the rational design of high-performance zeolite catalysts.

Fig. 20 (a) Reaction pathways for BAS-catalysed protolytic n-pentane cracking and dehydrogenation on zeolites. The red line indicates the location of the bond scission. The P1, P2 and P3 cracking pathways of n-pentane are illustrated as going through the C₂ or C₃ carbonium ion. TOF of overall cracking and dehydrogenation (b) and different cracking pathways (c) as a function of the EFSi-BAS portion of the total BAS on the samples. Reproduced with permission from ref. 237. Copyright 2023, Springer Nature. (d) First-principles MD simulations, PFG-NMR spectroscopy and TAP experiments for SAPO-34 crystal analysis. (e) Free energy profile for ethene, ethane, propene, propane diffusion at 600 K through an 8-ring window of SAPO-34 containing 0, 1, or 2 BASs. Reproduced with permission from ref. 94. Copyright 2021, Wiley-VCH.

As shown in Fig. 20d, Cnudde et al. systematically investigated the impact of BASs on the diffusion behavior of light olefins (ethylene and propylene) and alkanes (ethane and propane) in small-pore zeolites.⁹⁴Fig. 20e reveals that olefin diffusion is significantly influenced by BASs, whereas alkane diffusion remains largely unaffected. At 600 K, the diffusion barriers for ethylene in the absence of BASs, with a single BAS, and with two BASs are 38, 20, and 10 kJ mol⁻¹, respectively, while for propylene they are 57, 44, and 34 kJ mol⁻¹, demonstrating that an increased BAS density substantially reduces the diffusion barrier and enhances the transport of olefins through the zeolite pores. In contrast, the diffusion barriers for ethane (40 kJ mol⁻¹) and propane (68 kJ mol⁻¹) remain nearly constant, regardless of BAS density. This effect is primarily attributed to the formation of π–H interactions between olefins and the BAS, which stabilize olefins at the 8-ring windows, thereby facilitating their diffusion. In contrast, alkanes, lacking π-electrons, do not exhibit such interactions, resulting in diffusion that is independent of BAS presence. PFG-NMR experiments confirm that the self-diffusion coefficient of ethylene increases with BAS concentration, whereas ethane diffusion remains unaffected. Subsequently, they conducted a comprehensive investigation into the diffusion behavior of light olefins within the H-SAPO-34 catalyst during the MTO process. Utilizing enhanced sampling molecular dynamics simulations based on both force field and DFT, they elucidated the diffusion mechanism of ethene and propene through the 8-ring windows of H-SAPO-34. The findings revealed that olefin diffusion is a hindered process, occurring via discrete hopping events between adjacent nanocages. The presence of methanol, olefins, and aromatic species within the cages significantly modulates diffusion rates. BASs on the 8-ring windows facilitate diffusion by stabilizing transient π-complex interactions, whereas the accumulation of hydrocarbon pool species severely impedes mass transport and may further influence product selectivity through spatial distribution effects.

4.2 Pore size, channel type, and morphology as influencing factors

As shown in Fig. 21, a “hyperloop-like” diffusion model has been first proposed to describe the behavior of long-chain molecules in confined spaces.²⁴⁰ In smaller pore channels, the repulsive interactions with the pore walls constrain the molecules to move along the central axis, maintaining a linear trajectory and enabling rapid diffusion. In contrast, in larger pores, the molecules are more prone to deviating from the central path, leading to increased collisions with the pore walls and a reduction in the diffusion rate. Additionally, molecular bending and deformation exacerbate diffusion resistance, further hindering efficiency.

Fig. 21 (a) Illustrative examples of movement observed in daily life. (b) Schematic representation of n-dodecane (C₁₂) transport within zeolite channels, depicting both movement along the designated path (top) and deviations from it (bottom). (c) Visualization of van der Waals interactions between C₁₂ molecules and the zeolite framework. (d) Three-dimensional model illustrating the positioning of C₁₂ molecules within nanoscale channels. (e) Diffusion coefficients of C₁₂ molecules in confined channels of varying diameters. (f) Measurement of the deviation angle between the axial direction of the channels (inset line 1) and the end-to-end alignment of C₁₂ molecules (inset line 2). Reproduced with permission from ref. 240. Copyright 2023, Springer Nature.

Krishna and Van Baten proposed a theoretical framework based on the Maxwell–Stefan equations to describe the diffusion behavior of multicomponent mixtures in zeolites with different topologies under near-pore saturation conditions.²⁴¹ MFI zeolites exhibit strong diffusion correlations with significant decreases in the diffusion coefficients as saturation is approached. In contrast, LTA and DDR zeolites, composed of cage-like structures connected by narrow pore windows, restrict the diffusion to a stepwise hopping mechanism for individual molecules, leading to substantially lower diffusion rates. FAU zeolites with their larger pores and wider windows, allow multiple molecules to pass simultaneously, resulting in minimal diffusion limitations. In large-pore structures, molecular jumps are less obstructed, weakening correlation effects and increasing diffusion rates. Conversely, small-pore structures intensify molecular interference during diffusion, strengthening correlation effects and significantly reducing diffusion rates.

To gain deeper mechanistic insights into molecular transport phenomena within different zeolite frameworks, Tan et al. utilized a combination of grand canonical MC and molecular dynamics simulations to compare the adsorption and diffusion characteristics of DMF and p-xylene.²⁴² IFR demonstrated the lowest apparent activation barrier (30.85 kcal mol⁻¹), indicating that its catalytic activity remains relatively stable regardless of reactant pressure variations. MOZ exhibited the lowest intrinsic activation barrier (21.67 kcal mol⁻¹), suggesting that its inherent catalytic activity could be optimized through careful tuning of reaction conditions like temperature and pressure. MOR displayed superior diffusion kinetics, with the unique characteristic that product (p-xylene) diffusion outpaces reactant (DMF) diffusion. This favorable transport behavior helps minimize undesired side reactions and coke formation by facilitating rapid product removal from the catalytic sites (Fig. 22a). The design of zeolite catalysts with dual porosity combining mesopores and micropores offers a promising strategy to enhance molecular diffusion and improve the efficiency of reactant delivery to active sites.²⁴³ Investigations into the catalytic behavior of BASs in MWW and MFI zeolites with hierarchical meso–microporous structures reveal that mesopores significantly enhance the accessibility of larger organic molecules, such as di-tert-butylpyridine, to the acid sites. For smaller molecules like ethanol and propane, where diffusion limitations are less pronounced, the microporous environment remains the primary determinant of the reaction pathway (Fig. 22b).

Fig. 22 Screening zeolites based on reactant diffusion, reaction, product diffusion and synthesis difficulty. Reproduced with permission from ref. 242. Copyright 2024, Elsevier. (b) Arrhenius plots of monomolecular conversion of propane over MWW and MFI zeolite catalysts: cracking, dehydrogenation and temperature dependence of monomolecular propane cracking-to-dehydrogenation (Cr/De) rate ratios on MWW and MFI zeolites. Reproduced with permission from ref. 243. Copyright 2011, American Chemical Society. (c) The Si/Al ratio is denoted as the number following the framework-type code. H₄SiW and H₃PW stand for tungstosilicic and phosphotungstic acids. Rates were determined at conversions <10%. Solid lines are fits to the Arrhenius equation. Reproduced with permission from ref. 245. Copyright 2017, Springer Nature.

To elucidate the relationship between the adsorption behavior and diffusion mass transfer characteristics of hierarchical porous molecular sieves and zeolites, adsorption experiments alone are insufficient to effectively differentiate hierarchical porous materials from traditional molecular sieves. In contrast, PFG NMR can clearly reveal the interconnectivity within the materials.²⁴⁴ The variation in pore size exerts a profound impact on the kinetics and thermodynamics of catalytic reactions. Smaller pores, such as those in MFI zeolites, exhibit lower activation enthalpies due to stronger dispersive interactions but are accompanied by less favorable activation entropies. In contrast, larger pores, as found in FAU zeolites, favor an entropy-driven reaction pathway. The adsorption behavior of the substrate cyclohexanol within zeolite pores follows a Langmuir isotherm, revealing a trend of decreasing adsorption enthalpy with increasing pore size. The nanoscale dimensions and geometric constraints of the pores significantly enhance the association between H₃O⁺ ions and the substrate, thereby lowering the intrinsic free energy of activation for the dehydration reaction. Compared to aqueous-phase acidic solutions, reactions within medium sized pores achieve rate enhancements of up to two orders of magnitude. This dramatic increase in reaction rates is primarily attributed to positive entropy contributions compensating for the higher activation enthalpy, highlighting the critical role of the pore size in optimizing catalytic performance (Fig. 22c).²⁴⁵

Witte et al. revealed the mass transfer and adsorption behaviors of hydrocarbon molecules in small-pore zeolites.²⁴⁶ For small hydrocarbon molecules containing adjacent single bonds, such as propane and 1-butene, significant mass transfer resistance arises due to the size constraints of 8-MR channels in zeolites. In contrast, molecules with alternating single and double bonds, such as propene and butadiene, exhibit better diffusion performance. This effect diminishes with increasing carbon chain length, likely because longer-chain molecules can compensate for the energy barrier encountered at the pore entry through multiple interactions outside the pore channels. In Si–LTA, Henry's constant for n-alkanes and 1-alkenes increases monotonically with the carbon number. However, in Si–CHA, due to its unique cage-like structural effects, Henry's constant exhibits a non-monotonic behavior. Similarly, in AEI zeolites, Henry's constant for n-alkanes exhibits a nonlinear variation with increasing carbon chain length, showing a local minimum at C₈, while for CHA-type zeolites, the local minimum occurs at C₉. Thermodynamic analysis of the adsorption enthalpy and entropy reveals that this non-monotonic behavior arises from the mismatch between the cage cavity space and the molecular size. When the molecular chain length is short, the molecules preferentially adsorb in a fully trans configuration within a single cage. However, as the chain length exceeds the capacity of a single cage, the molecules undergo multi-cage adsorption behavior via the 8-MR windows, interacting with adjacent cages.²⁴⁷ Travert et al. proposed a novel methodology for investigating mass transport properties in acidic zeolite materials.²⁴⁸ This approach integrates gravimetric analysis to assess diffusion from the gas phase into the entire porous structure and infrared spectroscopy to selectively probe diffusion at active sites within the micropores. Furthermore, an integral equation inversion method was employed to reconstruct the distribution of diffusion domains and elucidate the evolution of adsorption curves and infrared spectra. Using a mechanical mixture of FAU and MFI zeolites as a model system, the study examined the adsorption and diffusion behavior of isooctane, revealing that H-FAU exhibits a significantly higher effective diffusion rate constant compared to H-MFI. This method effectively distinguishes diffusion behaviors within different pore structures, providing valuable theoretical insights for optimizing zeolite-based materials in catalytic and adsorption processes.

Using interference microscopy (IFM), Saint Remi et al. conducted real-time imaging of methanol adsorption kinetics in individual SAPO-34 crystals and uncovered significant variations in mass transfer rates due to surface resistance.²⁴⁹ Despite similarities in crystal size and morphology, adsorption and diffusion behaviors varied considerably, with surface barriers playing a crucial role in restricting transport (Fig. 23a and b). The presence of these barriers led to reduced permeability and a substantial decrease in overall mass transfer efficiency (Fig. 23c). Analysis of adsorption curves further confirmed that the adsorption kinetics of both crystals were primarily governed by surface permeability rather than being purely diffusion-limited (Fig. 23d). Furthermore, they investigated molecular mass transfer in metal-free AlPO-LTA zeolites and conducted a comprehensive assessment of light hydrocarbon diffusion using PFG NMR and infrared microscopy (IRM). Fig. 23e provides a detailed visualization of the diffusion profiles of propylene and propane within AlPO-LTA crystals. The concentration distribution of propylene closely aligns with theoretical models, indicating that its transport is primarily governed by intracrystalline diffusion. In contrast, the concentration profile of propane deviates from the ideal cubic symmetry, suggesting that its diffusion is significantly hindered, likely due to surface resistance or pore blockage effects. Propylene exhibits a relatively high surface permeability (2 × 10⁻⁷ m s⁻¹) and intracrystalline diffusivity (3 × 10⁻¹³ m² s⁻¹), whereas propane demonstrates a markedly lower mass transfer rate. It underscores the interplay between surface and intracrystalline transport mechanisms in nanoporous materials and highlights that even in zeolites with cubic symmetry, guest molecule size effects can lead to distinct diffusion behaviors.²⁵⁰

Fig. 23 (a) IFM pictures with crystal dimensions of SAPO-34 (crystal A). (b) 3D concentration plots over entire crystals. (c) 1D plots with fitted concentration profiles. (d) Integral uptake curves with analytical expressions for diffusion-limited (1) and surface-barrier-limited (2) uptake, respectively. Reproduced with permission from ref. 249. Copyright 2015, Springer Nature. (e) IFM profiles of the concentration integrals measured perpendicular to the crystal face during three-dimensional uptake of propylene (middle) and propane (bottom) within a single AlPO-LTA crystal following a pressure increase from 0 to 250 mbar at 295 K, and their representation for an ideal cubic structure (top). Reproduced with permission from ref. 250. Copyright 2015, American Chemical Society. (f) The top row presents confocal fluorescence microscopy images of individual FCC particles exposed to thiophene at 373 K across various deactivation stages. The middle row offers magnified views of the highlighted regions, revealing micrometer-sized zeolitic domains with progressively diminishing fluorescence intensity as deactivation progresses. The bottom row displays fluorescence intensity histograms corresponding to the zeolite domains in each of the four catalyst samples. Reproduced with permission from ref. 251. Copyright 2011, Springer Nature.

Buurmans et al. employed selective staining combined with confocal fluorescence microscopy (CFM) to investigate the deactivation process of FCC catalysts (Fig. 23f).²⁵¹ Their findings revealed a progressive decline in fluorescence intensity as the catalysts underwent steam deactivation (ST), cyclic deactivation (CD), and metal impregnation-steam deactivation (MI), indicating a systematic reduction in BAS density. Concurrently, the extent of structural degradation increased with more severe deactivation, as evidenced by a pronounced collapse of the zeolite framework and a significant reduction in micropore volume, suggesting partial structural deterioration or framework aluminum leaching. Catalytic cracking experiments further demonstrated a strong correlation between BAS depletion and declining catalytic activity, with the MI-treated catalyst exhibiting the lowest conversion efficiency. FCC catalyst deactivation arises from both the depletion of BAS and the progressive degradation of the zeolite framework, accompanied by pore blockage. This underscores the critical need to optimize acid site distribution while preserving structural integrity to enhance catalyst longevity and sustain high catalytic efficiency.^252–256

By integrating quasi-equilibrium temperature-programmed desorption and adsorption (QE-TPDA) experimental techniques with MC molecular simulations, C₅–C₁₀ alkanes exhibit a two-step adsorption process within LTA-type zeolites.²⁵⁷ As saturation conditions are approached, the addition of the final alkane molecule entering an LTA cage leads to a reduction in the conformational entropy of the other adsorbed molecules. To compensate for this entropy loss, the system must increase the thermodynamic driving force for adsorption by lowering the temperature. In contrast, the adsorption of alkanes in FAU zeolites manifests as a single-step process. With increasing cation content, the desorption temperature of alkanes in FAU zeolites progressively rises, aligning with the enhanced interactions between alkanes and cations.

The 12-MR channels of BEA zeolites create an optimal confined environment for methanol dehydration to DME, with a van der Waals dispersion interaction enthalpy reaching 63 kJ mol⁻¹, which is significantly higher than that of FAU zeolites (37 kJ mol⁻¹).²⁵⁸ This enhanced dispersion energy arises from the relatively smaller geometric dimensions of the BEA channels, which amplify interactions between the transition state and the pore walls, thereby stabilizing these reaction intermediates. Furthermore, the adsorption enthalpy of n-hexane (Q_ads), closely tied to molecular diffusion pathways, underscores the suitability of BEA for intermediate-sized molecules. BEA exhibits a Q_ads value of ∼56 kJ mol⁻¹, which surpasses that of FAU with larger pore sizes yet remains slightly lower than that of MOR with tighter pores, indicating a balanced environment favoring efficient diffusion. Kinetic studies of methanol dehydration reveal that the first-order rate constant of BEA increases exponentially with rising Q_ads, underscoring the pivotal role of adsorption in facilitating transition state stabilization. In contrast, the zero-order rate constant, influenced less by confinement due to the comparable sizes of intermediates and transition states, remains relatively insensitive to such effects.

4.3 Extra-framework cations as influencing factors

4.3.1 Active metal cations. The selection of active sites in zeolite catalysis and adsorption studies is inherently complex, particularly concerning the stability of palladium ions under varying Al distributions and oxidation states. Using a high-throughput computational approach, the study reveals that Pd cations preferentially occupy 6-MR, where the spatial arrangement of adjacent oxygen atoms provides optimal stabilization. Pd⁺H⁺ sites exhibit significantly stronger binding affinities for NO compared to Pd²⁺, highlighting the critical role of solvent interactions with the transition state. The stability of Pd ions across different Al configurations notably influences the adsorption selectivity and capacity of zeolites for NO (Fig. 24a and b).²⁵⁹

Fig. 24 (a) NO binding energy versus Pd formation energy on CHA. (b) Dotted lines refer to fitted data for each gas adsorbate. Reproduced with permission from ref. 259. Copyright 2022, Springer Nature.

In CHA and BEA zeolites, the spatial distribution of aluminum pairs significantly impacts the stability of palladium ions. 6-MRs offer the most favorable environment, particularly when aluminum pairs adopt an NNNN (second-nearest neighbour) configuration. In contrast, 4-MRs and 8-MRs provide lower stability, likely due to their surrounding oxygen arrangement being less effective at stabilizing Pd ions.²⁶⁰ Theoretical simulations of the adsorption behaviors of CO and NO in transition metal exchange zeolites revealed that using a single or two-parameter regression approach is not insufficient to accurately capture the complex interactions between molecules and metal centers, while multi-parameter models can significantly improve the prediction performance.^261,262

Wannakao et al. investigated the role of gold-ion-exchanged zeolites in methane activation, specifically focusing on ZSM-5, FAU, FER, and MCM-22.²⁶³ The ZSM-5 structure features a bicoordinated cation site, which shows a higher catalytic activity compared to the FAU structure with a tricoordinated cationic site. In the Au–ZSM-5 catalyzed reaction, the activation energy is 13.2 kcal mol⁻¹, which is lower than that of the bare Au⁺ ion catalyzed reaction (21.3 kcal mol⁻¹). This is attributed to the structural constraints of the ZSM-5 zeolite, which leads to an earlier transition state and a greater charge difference between the C and H atoms during the C–H bond cleavage.

The proximity between metal and acid sites does not necessarily correlate with enhanced catalytic performance. For instance, Cheng et al. loaded Pt onto an Al₂O₃ binder, maintaining a nanoscale distance between the metal sites and acid sites, and achieved higher selectivity for the target product (C₇ isomers) compared to confining platinum within zeolite crystals.²⁶⁴ This is because restricting platinum nanoparticles within the molecular sieve significantly limits the diffusion of reaction intermediates, leading to undesired secondary cracking reactions. This phenomenon is particularly pronounced in zeolites with narrower micropores or larger crystal sizes.

4.3.2 Trapdoor effects caused by extra-framework cations. Among the various manifestations of cation mobility upon adsorption, the trapdoor mechanism represents an especially intriguing and selective gating process.²⁶⁵ In certain small-pore zeolites, extra-framework cations are positioned near the pore entrances and function as mobile gatekeepers that dynamically regulate access to the internal framework. This phenomenon is not governed by geometric size exclusion alone, but instead relies on the specific interactions between guest molecules and these cations. When certain guest molecules (such as CO₂) interact strongly enough with these cations, they can temporarily displace the cation from the pore mouth, effectively “opening” the trapdoor and allowing access to the internal channels. In contrast, other molecules (such as CH₄ or N₂), which interact more weakly with the cation, are unable to dislodge it and are thus excluded. This mechanism differs fundamentally from conventional molecular sieving, as molecular access is governed not solely by size exclusion but by the molecule's ability to induce cation displacement.

Lozinska et al. investigated the CO₂ and CH₄ adsorption behavior of zeolite RHO exchanged with univalent cations Na⁺, K⁺, and Cs⁺.²⁶⁶ In these systems, Na⁺ exhibits only slight displacement within the S8R sites; K⁺ undergoes relocation from window positions to S6R cage sites, accompanied by a marked expansion of the unit cell; while Cs⁺ migrates from D8R to S8R sites, inducing a symmetry change. At elevated CO₂ pressures, all three cation forms converge toward a similar distribution, highlighting the framework's adaptive structural response to CO₂ adsorption. Recently, Jia et al. reported a novel zeolite material, Sr/K-HEU, that leverages the molecular trapdoor effect to achieve the challenging inverse separation of CO₂ and C₂H₂. Under ambient conditions, it exhibits outstanding separation performance with a dynamic CO₂/C₂H₂ selectivity of 48, a high CO₂ uptake of 0.96 mmol g⁻¹, and excellent cycling stability. Three-dimensional electron diffraction revealed that Sr²⁺ cations reside at the entrances of 10-MR channels and can transiently shift in response to CO₂, thereby opening the pore. In contrast, C₂H₂ is effectively excluded due to its inability to induce such cation displacement, primarily as a result of repulsive interactions with Sr²⁺.²⁶⁷

Zhao et al. proposed a novel trapdoor-type adsorbent, K⁺–ZSM-25, whose framework comprises multiple cages interconnected by 8-MR apertures, characteristic of small-pore zeolites.²⁶⁸ By incorporating a high density of gatekeeping K⁺ cations, they established a temperature-modulated diffusion mechanism, wherein the cations undergo periodic, thermally induced displacements that transiently open molecular-scale channels, enabling selective diffusion of small guest molecules. As illustrated in Fig. 25a and b, four distinct diffusion pathways traverse the ZSM-25 unit cell, each formed by paired cage structures: grc-opr, pau-opr, phi-oto, and plg-oto, with 8-MRs serving as the critical diffusion bottlenecks. First-principles DFT calculations were employed to determine the minimum energy trajectories for K⁺ displacement from its primary to secondary coordination sites, revealing the activation barriers associated with cation motion in each pathway. Among them, the pau-opr segment exhibited the highest barrier (1.13 eV), nearly nine times that of grc-opr (0.13 eV) and notably higher than those in phi-oto (1.70 eV) and plg-oto (1.60 eV), identifying it as the rate-limiting site for intracrystalline transport. For N₂, passage through the pau-opr segment required a K⁺ displacement of approximately 2.8 Å, associated with a cooperative barrier of just 0.99 eV; in contrast, CH₄ required a larger displacement of 3.1 Å and faced a higher barrier of 1.16 eV. This energy difference stems from the subtle but critical variation in kinetic diameters. Importantly, diffusion of both gases occurs prior to K⁺ reaching its secondary equilibrium position, indicating that permanent cation migration is unnecessary; rather, the thermal oscillation alone suffices to generate transient windows for molecular access. This establishes the essence of the trapdoor mechanism as a thermally driven, dynamically modulated gating process mediated by the reversible vibration of extra-framework cations.

Fig. 25 (a) 3D representation of the ZSM-25 unit cell and the primary gas diffusion pathways formed by four distinct channels interconnected via 8-MR, comprising four doubly linked cages. (b) Minimum energy pathways for door-keeping K⁺ migration from the most stable site to the nearest secondary site in grc-opr, pau-opr, phi-oto, and plg-oto frameworks. Reproduced with permission from ref. 268. Copyright 2021, American Chemical Society. (c) CO₂/CH₄ and CO₂/N₂ selectivities and CO₂ working capacities in columns of zeolites. CO₂/CH₄ and CO₂/N₂ breakthrough curves on K–MER-2.3, respectively. Reproduced with permission from ref. 188. Copyright 2020, Wiley-VCH.

It is noteworthy that in some zeolitic systems, the trapdoor mechanism is not an isolated phenomenon, but frequently couples with framework breathing, forming a cooperative response that modulates both adsorption behavior and molecular transport. Georgieva et al. found that MER zeolites bearing different extra-framework cations undergo a “breathing” transition upon CO₂ uptake, expanding from a dehydrated narrow-pore to a wide-pore phase through cooperative interactions among CO₂, the framework, and the cations.²⁶⁹ Notably, K–MER triggers rapid, reversible channel opening at very low CO₂ partial pressures (<0.1 bar), exhibiting a classic “triggered gate-opening” mechanism with exceptionally fast CO₂ uptake and excellent CO₂/CH₄ separation. By contrast, Na–MER and Cs–MER require higher CO₂ pressures to induce the phase change and show slower adsorption kinetics. During CO₂ adsorption, Cs⁺ ions display pronounced migration (e.g. from D8R to S8R sites), and the resulting two-step isotherm further highlights the cation-modulated flexibility of the MER framework. Choi et al. further revealed the diversity of CO₂ adsorption behaviors in MER-type zeolites (Fig. 25c).¹⁸⁸ Samples with low Si/Al ratios (2.3) exhibit high extra-framework cation concentrations, leading to significant cation migration from the 8-MR windows and concurrent framework distortion during adsorption, features characteristic of a cooperative cation gating–breathing mechanism (e.g., Na–MER-2.3, K–MER-2.3, and Rb–MER-2.3). In contrast, high Si/Al ratio (3.8) samples, with lower cation occupancy, primarily rely on framework flexibility for adsorption, corresponding to the breathing mechanism (e.g., K–MER-3.8). For Cs⁺-exchanged MER, the high cation occupancy and rigid framework enable selective adsorption solely via the cation gating mechanism (e.g., Cs–MER-2.3). It further demonstrates the pronounced impact of these mechanisms on CO₂ selectivity and capacity, with K–MER-2.3 exhibiting the best overall performance in terms of CO₂/CH₄ and CO₂/N₂ selectivity and working capacity, thereby confirming the advantage of the cooperative mechanism in achieving selective CO₂ capture. Interestingly, a “cation crowding” adsorption mechanism distinct from trapdoor effects has been identified in PHI-type zeolites. In Cs–PHI-2.5 (Si/Al ratio), strong repulsive interactions among densely packed Cs⁺ ions within the t-phi cages lead to a structurally closed state; only when the CO₂ partial pressure exceeds a critical threshold do the Cs⁺ ions undergo concerted migration, triggering a framework transition from a narrow-pore to a wide-pore configuration and resulting in a highly cooperative, step-like adsorption profile.²⁷⁰

Coudert and Kohen proposed a distinct interpretation of the “trapdoor” adsorption mechanism of CO₂ in zeolite Na–RHO using ab initio molecular dynamics (AIMD) and free energy calculations.²⁷¹ Their study revealed that the presence of CO₂ does not significantly reduce the free energy barrier for Na⁺ migration from the S8R site, which contradicts the earlier hypothesis of CO₂ actively “pulling open” the trapdoor. Instead, their simulations demonstrated that Na⁺ exhibits substantial thermal fluctuations at ambient temperature, allowing it to transiently vacate its blocking position and create a momentary passage through the narrow window. In this framework, CO₂ diffuses not by actively displacing the cation, but by opportunistically passing through transiently open channels, which they describe as a “swinging door”. While the study reported by Wright et al. captures ensemble-averaged structural responses over longer timescales,²⁶⁶ analysis by Coudert et al. focuses on sub-nanosecond dynamic events. Thus, these two perspectives can be considered complementary views of the same phenomenon observed at different temporal resolutions: thermal fluctuations of Na⁺ create fleeting diffusion windows for CO₂, whose adsorption, in turn, stabilizes the framework and governs the macroscopic uptake behavior and selectivity. This comparison highlights the importance of integrating experimental structural characterization with high-fidelity dynamical modeling to uncover the microscopic mechanisms of adsorption in flexible zeolites and suggests that the design of future materials should account for the amplitude of cationic motion and its coupling with the framework as key parameters in tuning diffusion pathways and molecular selectivity.

Calero et al. employed configurational-bias Monte Carlo (CBMC) simulations to systematically investigate how the type and concentration of nonframework cations affect the adsorption behavior of C₁–C₄ alkanes and the separation selectivity of their isomers in MFI-type zeolites. Despite identical framework structures, variations in cation size, charge, and population were shown to markedly influence adsorption capacity and molecular discrimination. Higher cation densities reduce accessible pore volume and hinder adsorption, while smaller or more sparsely distributed cations facilitate molecular uptake by minimizing steric obstruction. In binary mixtures such as n-butane/isobutane, a moderate increase in cation density can enhance the preferential adsorption of linear alkanes by selectively blocking the intersectional adsorption sites favored by branched isomers. However, excessive cation loading (e.g., 8 cations per unit cell) reverses this trend by over-constricting diffusion pathways. Moreover, at a fixed Si/Al ratio, lighter alkali cations (e.g., Li⁺) more closely reproduce the adsorption characteristics of all-silica MFI compared to heavier species like Cs⁺ or Ba²⁺, underscoring the critical role of cationic identity in tuning zeolite adsorption performance.²⁷²

4.4 Hydrophilicity and framework defects as influencing factors

The hydrophilicity of zeolites plays a pivotal role in modulating the interfacial interactions between guest molecules and the external surface, thereby exerting a profound influence on adsorption affinity, diffusion resistance, and consequently on adsorption capacity, molecular selectivity, and mass transport kinetics. In parallel, structural defects, such as framework interruptions, silanol nests, and intercrystalline boundaries, give rise to spatial heterogeneities in adsorption energy landscapes and introduce non-ideal diffusion pathways near pore mouths and along external surfaces. These effects collectively contribute to a more intricate adsorption–diffusion behavior than that predicted by idealized frameworks. Therefore, a mechanistic understanding of how hydrophilicity and structural defects shape the adsorption and diffusion characteristics of zeolites is essential from the surface to the internal microporous environment. This perspective not only complements the earlier discussions on framework geometry, cation exchange, and acid site distribution, but also provides a crucial foundation for the rational design and engineering of zeolite materials tailored for high performance separation and catalytic processes under realistic operational conditions.

Hydrophilicity typically refers to the affinity of zeolite materials towards adsorption of polar guest molecules. Zeolites with high hydrophilicity possess a greater number of polar surface sites, which strongly bind and interact with polar species. In contrast, highly hydrophobic zeolites tend to exclude water and other polar molecules. The hydrophilic or hydrophobic character of a zeolite is predominantly governed by its chemical composition and framework structure. In particular, the concentration of framework aluminum and the density of surface hydroxyl groups (Si–OH) are key determinants of hydrophilicity. An increase in the Si/Al ratio enhances the proportion of nonpolar Si–O–Si linkages within the framework and reduces the number of surface –OH groups, thereby rendering the zeolite more hydrophobic. Conversely, a lower Si/Al ratio, often associated with a higher density of polar hydroxyl sites, results in markedly enhanced hydrophilicity.^273,274 For example, in MFI-type zeolites, the incorporation of Al atoms into an otherwise pure-silica framework introduces negatively charged AlO₄⁻ units that require charge compensation by extra-framework cations, often generating BASs or associated defect hydroxyls, both of which contribute to localized hydrophilic sites. Furthermore, the presence of exchangeable cations within the zeolite framework enhances its electrostatic field and overall polarity, imparting stronger hydrophilicity to low-silica zeolites with high cation content. But also aluminum-free zeolites have strongly hydrophilic sites. For example, silicalite-1 is able to strongly bind water and methanol in defects in the form of silanol nests that result from its synthesis.¹⁷⁸ Defects are commonly introduced during the synthesis or post-synthetic modification of zeolites.²⁷⁵ Procedures such as high-temperature calcination, steam treatment, and chemical leaching are frequently employed to modulate the acidity and pore architecture of zeolites. However, these treatments can cleave framework bonds, resulting in the generation of hydroxyl-associated defects or EFAl species.

Through controlled acid-induced dealumination, abundant silanol nests are introduced into the zeolite framework without compromising its structural integrity, providing well-defined anchoring sites for subsequent heteroatom incorporation. In the case of Ti–beta synthesis, titanium precursors selectively interact with these hydroxyl nests and are efficiently immobilized within the framework via high-temperature calcination, leading to the formation of isolated, framework-integrated Ti species. This approach markedly enhances the catalytic performance of Ti–beta in both CO₂ photoreduction and selective oxidation reactions. As the Ti loading increases, a systematic decrease in surface hydrophilicity is observed, corresponding to the progressive transformation of the Ti coordination environment from hydrophilic Si–OH nests to hydrophobic Si–O–Ti linkages within the lattice. Importantly, by modulating the initial Si/Al ratio and Ti incorporation level, the hydrophilic/hydrophobic character of the material can be independently tuned without significantly altering its textural properties. This capability offers a powerful tool for the rational design of local polarity environments in complex catalytic systems, thereby broadening the scope of Ti–beta zeolites in advanced catalytic applications.²⁷⁶

Sun et al. developed a strategy combining in situ ligand protection with hydrogen reduction to encapsulate subnanometric Pd–Mn bimetallic clusters within amino-functionalized MFI-type zeolites. The incorporation of amino groups at framework defect sites endowed the zeolite with basic character and markedly enhanced its hydrophilicity. This modification also reinforced metal–support electronic interactions, leading to the stabilization of electron-enriched Pd active centers. In contrast to hydrophobic silicalite-1 zeolites, the hydrophilic amino-functionalized analogues exhibited substantially improved catalytic performance in formic acid dehydrogenation, delivering higher hydrogen evolution rates and lower apparent activation energies. These enhancements are primarily attributed to the increased affinity of polar formic acid molecules for the hydrophilic channels, facilitating both molecular enrichment and access to confined active sites. Contact angle analysis provided compelling evidence for the tunability of zeolite surface hydrophilicity via organic functionalization, offering mechanistic insights into the interplay between surface polarity and catalytic behavior. The unmodified pure-silica silicalite-1 exhibited moderate hydrophobicity, as indicated by a contact angle of 37.2°, consistent with the nonpolar nature of the siliceous framework. Progressive incorporation of amino silane precursors resulted in a marked reduction in the contact angle for the amino-functionalized variants, confirming a substantial increase in surface hydrophilicity. In contrast, grafting of methyl groups led to a sharp rise in the contact angle up to 85.6°, reflecting enhanced hydrophobic character. These findings demonstrate that surface functionalization with polar groups significantly improves the affinity of the zeolite surface toward polar reactants, thereby promoting their diffusion and accumulation within the micropores. This surface property modulation is directly correlated with enhanced catalytic activity, underscoring the importance of interfacial polarity in governing molecular accessibility and catalytic efficiency in zeolite-based systems.²⁷⁷

Conrad et al. systematically investigated the catalytic performance of Sn–MFI zeolites with varying degrees of hydrophilicity in the Baeyer–Villiger oxidation of cyclohexanone. Compared to their more hydrophobic counterparts, hydrophilic Sn–MFI materials exhibit enhanced affinity for both the reactant (cyclohexanone) and the by-product (acetic acid), thereby increasing the local concentration of reactants within the micropores and promoting higher reaction rates. However, excessive hydrophilicity can lead to the accumulation of polar by-products, resulting in pore blockage that hinders the diffusion of reactants and the desorption of products, ultimately compromising catalytic stability.²⁷⁸ This study employs molecular simulation methods to systematically evaluate the reverse osmosis separation performance of zeolite nanosheets in ethanol/water systems, revealing the key role of hydrophilic characteristics of both the material surface and the framework structure in the separation mechanism. Zeolites with strong hydrophobicity exhibit excellent ethanol selectivity, preferentially adsorbing the larger ethanol molecules while repelling the smaller and more polar water molecules, thereby achieving a counterintuitive separation process. This adsorption-dominated mechanism is distinct from the conventional reverse osmosis pathway based on molecular sieving and reflects the fundamental influence of the hydrophilic/hydrophobic nature of zeolites on selective adsorption and diffusion. The hydrophilicity of the external surface of the zeolite, particularly at the pore entrances, also has a certain impact on the initial selective adsorption process prior to molecular entry into the channels. Surface silanol groups tend to form strong hydrogen bonds with water molecules, resulting in preferential water accumulation on the surface, thereby reducing the ethanol concentration at the pore openings and decreasing its probability of permeation.²⁷⁹ Using first-principles grand canonical Monte Carlo (FP-GCMC) simulations, Bai et al. revealed that proton exchange significantly enhances the hydrophilicity of MFI-type zeolites. The number of adsorbed water molecules increases with the aluminum content in the framework, but at higher Al concentrations, the number of water molecules per BAS reaches a saturation point, indicating that adsorption is limited by the spatial constraints of the pore structure. Across all temperatures and humidity levels, protons spontaneously dissociate from the framework and integrate into the hydrogen-bonding network of the adsorbed water to form hydronium ions (H₃O⁺), a process that proceeds without additional energetic barriers. This finding highlights the pivotal role of water cluster formation in facilitating proton mobility within highly hydrophilic acidic zeolites and provides crucial insights into the altered reactivity of BASs under aqueous conditions.²⁸⁰

For the SAPO-34 zeolite, iDPC-STEM images reveal the presence of disrupted TO₄ units at the external surface (Fig. 26a and b), while AFM measurements show a surface thickness reduction of approximately 1.0 nm after acid etching, confirming the effective removal of the silicon-rich outer layer (Fig. 26c). XPS and EDS analyses further demonstrate a significant decrease in surface silicon content in the etched sample, with a steep silicon concentration gradient observed from the exterior to the interior in the pristine sample, indicating that the outer layer is enriched with Si–OH defects. DRIFT spectra and solid-state NMR provide additional evidence that acid treatment markedly reduces the defective surface layer (Fig. 26d–f). First-principles calculations show that the adsorption free energy of water at Si–OH sites is −2.9 kJ mol⁻¹, much lower than at Al–OH sites (−10.1 kJ mol⁻¹), suggesting that polar molecules such as water and olefins exhibit stronger interactions with Si–OH-rich surfaces (Fig. 26g). As shown in Fig. 26h, contact angle measurements further support this conclusion, showing the pristine sample to be more hydrophilic (42.4°) and the etched sample more hydrophobic (58.9°). This enhanced adsorption favors guest molecule accumulation at the interface but simultaneously increases surface diffusion resistance. The authors delineate a three-step pathway for guest molecules traversing the external layer: adsorption from the gas phase onto the outer surface, overcoming the surface diffusion barrier, and entry into the intracrystalline framework. Si–OH defects stabilize guest molecules within the external adsorption zone and induce a narrowing of the surface 8-MR pores to an average of 10.2 Å, compared to 11.1 Å in the etched counterpart, thereby elevating both forward and backward diffusion barriers and reducing surface permeability. In the etched sample, the reduction in Si–OH content (with a Si/Al ratio decreasing to 0.75) leads to a notable decrease in the activation energy for surface diffusion of propane and propylene, enhancing their permeation into internal catalytic sites (Fig. 26i). Additionally, the adsorption–desorption process exhibits clear asymmetry: propylene adsorption is kinetically more favorable than desorption, attributed to the non-uniform spatial distribution of BAS within the CHA cages, which creates a pronounced molecular trapping effect.²⁸¹

Fig. 26 (a) Illustration of how isolated silanol groups influence local framework structure, molecular uptake, and transport dynamics at zeolite surfaces. (b) iDPC-STEM image of the H-SAPO-34 zeolite. (c) AFM images of pristine and etched H-SAPO-34 zeolites. (d) DRIFT analysis for hydroxyl functional groups of pristine and etched H-SAPO-34 zeolites. ²⁹Si (e) and ¹H (f) MAS NMR spectra, (g) proposed models of the crystal boundary, (h) contact angle of water droplets at pristine and etched H-SAPO-34 zeolites. (i) Isosurface plots of the reduced density gradient and adsorption energy for the water adsorbed at ≡Si–O–H (left) and ≡Al–O–H (right) sites. Reproduced with permission from ref. 281. Copyright 2023, American Chemical Society.

Ye et al. developed a three-dimensional anisotropic reaction–diffusion model to comprehensively investigate the coupled effects of surface, interface, and defect structures on the catalytic performance of ZSM-5 in ethylbenzene synthesis. Anisotropic diffusion significantly alters molecular migration pathways and reactant concentration profiles within micropores, thereby impacting overall reaction rates and effectiveness factors. Reduced surface and intercrystalline permeabilities introduce substantial mass transport barriers, which emerge as key rate-limiting factors. Introducing an appropriate amount of structural defects, such as secondary or through-type mesopores, can effectively shorten the diffusion path within micropores and mitigate diffusion limitations. However, excessive defect volume may dilute the number of active catalytic sites and compromise performance. The porosity, size, and spatial distribution of these defects play a crucial role in catalytic efficiency, with through-type mesopores aligned along the direction of slowest diffusion showing the greatest enhancement in mass transport and reaction performance.²⁸² The synergistic effect of secondary mesopores and internal defects plays a critical role in governing coke deposition and catalyst deactivation. Taking ZSM-5 as an example, the introduction of mesoporosity via alkaline treatment effectively suppresses coke accumulation within micropores, thereby delaying the blockage of active acid sites and moderately enhancing catalytic stability. However, this strategy primarily shifts the location of coke from the internal to the external surface, with limited impact on the total coke amount. In contrast, defect site healing through ammonium hexafluorosilicate treatment significantly reduces overall coke formation, particularly within the micropores, as coke precursors (alkylated aromatics) are predominantly generated at internal defect sites. Suppressing these defects limits precursor generation and consequently mitigates downstream coke polymerization. Thus, an optimal approach involves coupling tailored mesoporosity to facilitate molecular transport with the minimization of internal defects to control coke precursor formation.

4.5 Experimental conditions of adsorption and diffusion studies

Water molecules alter the adsorption properties of reactants by affecting the hydrophobicity of the zeolite surface. On the other hand, water also regulates the diffusion behavior of reactants within the zeolite channels, thereby optimizing the selectivity and efficiency of the reaction.²⁸³ The adsorption of benzene molecules on the SMS–ZSM-5 surface was characterized using ¹³C–¹³C proton-driven spin diffusion (PDSD) MAS NMR spectroscopy, revealing the interactions between the benzene molecules and the zeolite surface (Fig. 27). The variations in the correlation peak signals for benzene and the zeolite with different water loadings indicates the influence of water molecules on benzene adsorption behavior. The changes in the NIS values of the correlation peaks with water loading quantify how water molecules regulate benzene adsorption and diffusion behaviors.

Fig. 27 Influence of water on the construction of a reactive SMS–benzene complex over H–ZSM-5. (a) 2D ¹³C–¹³C PDSD MAS NMR spectra of ¹³C benzene adsorbed over ¹³C SMS–ZSM-5 (SMS: surface methoxy species). (b) Enlarged correlation peak signals with various water loadings. (c) NIS value of correlation peak (59, 130) ppm vs. loaded water amounts. (d) Schematic of water driving benzene toward SMS to form the SMS–benzene complex. Reproduced with permission from ref. 283. Copyright 2023, Wiley-VCH.

The presence of water molecules proximate to acid sites in H–ZSM-5 catalysts exhibits a remarkable dual functionality in modulating C–H bond activation that water can either promote reactivity through transition state stabilization or suppress it via competitive adsorption. Chen et al. reported that when water loading remains at or below stoichiometric levels, the rate constant for isobutane C–H bond activation increases by an order of magnitude relative to anhydrous conditions.²⁸⁴ However, as the water loading exceeds 2–3 molecules per acid site, reaction kinetics are dramatically attenuated, ultimately leading to complete suppression of reactivity. In situ solid-state NMR spectroscopy reveals that under low water loading conditions, water molecules likely facilitate enhanced proton transfer through transition state barrier reduction. Conversely, at elevated water concentrations, the formation of water clusters with substantially higher proton affinities compared to isolated water molecules or isobutane effectively inhibits interaction between alkane molecules and catalytic acid sites.²⁸⁵

For protonated water clusters formed in aqueous solvent, Xu et al. confirmed that proton transfer to pyridine was more favorable in low-polarity solvents like acetonitrile and 1,4-dioxane compared to water.²⁸⁶ By measuring adsorption free energies at different temperatures, the adsorption enthalpy and entropy of pyridine on Si–ZSM-5 zeolite were determined to be −30 kJ mol⁻¹ and −53 J mol⁻¹ K⁻¹, respectively, which were significantly lower than those observed for gas-phase adsorption.²⁸⁷

Temperature can dynamically modulate the active site structures in zeolites. Using SAPO-34 as an example, there exist two distinct adsorption modes for acetone: one involving bonding to BAS and the other an induced FLP adsorption mode. As the temperature increases, acetone undergoes a progressive transformation from the BAS adsorption mode to the FLP adsorption mode. This transition is driven by temperature-induced reconstruction of the zeolite framework, which stabilizes the FLP sites relative to the BASs.²⁸⁸

The influence of the temperature on diffusion varies with different pore sizes and structures. For intermediates, such as ethanol, ethylene, and propylene, their diffusion behaviors within molecular sieves differ between zeolites. In ZSM-5 zeolites, ethanol molecules are primarily concentrated in the straight-channel regions and gradually diffuse into other areas as the temperature increases. Conversely, in FAU zeolites, small molecules like ethylene are predominantly located within the two supercage regions at low temperatures and progressively disperse to other sites with rising temperatures. For larger molecules, such as isopentene, diffusion throughout the entire FAU unit cell is achievable even at high temperatures, while difficult to penetrate the tortuous channels of ZSM-5 even at elevated temperatures.²⁸⁹

Through molecular dynamics simulations of Maxwell–Stefan (D_i) and self-diffusion coefficients (D_i,self), it was demonstrated that for cage-type zeolites with narrow window apertures, such as LTA, CHA, and DDR, the loading dependence of diffusivity exhibits a significant temperature sensitivity when guest molecules experience strong confinement effects at the window regions. Under these conditions, the activation energy for diffusion displays a systematic variation with the molecular loading. This phenomenon is particularly pronounced in systems where the molecular dimensions approach the window dimensions, leading to enhanced diffusional barriers at the inter-cage interfaces.^290,291

At specific calcination temperature, internal strain fields within zeolite microcrystals significantly modulate the spatial distribution of adsorption sites and molecular diffusion pathways. Through coherent X-ray diffraction imaging (CDI), Kim et al. revealed a distinctive core–shell strain architecture within ZSM-5 crystals, characterized by contrasting thermal expansion behaviors that the core region, containing residual organic template molecules, exhibits positive thermal expansion, while the shell demonstrates intrinsic negative thermal expansion characteristics. This thermal expansion heterogeneity induces a pronounced triangular strain field distribution throughout the crystal framework, fundamentally altering its structural properties.²⁴⁷

In zeolite composites, variations in temperature and diffusion time can significantly alter the diffusion behavior of molecules. For example, in the FAU/EMT system, at a lower temperature (264 K), the effective diffusion coefficient of iso-octane continuously decreases with increasing diffusion time, while its root mean square displacement (MSD) remains essentially unchanged. This indicates the presence of diffusion and mass transfer barriers within the FAU/EMT particles, restricting the movement of iso-octane molecules. These barriers are likely situated at the interfaces between FAU and EMT structures, hindering mass transfer among the different components within the particles. At a higher temperature (289 K), iso-octane diffusion in FAU/EMT composites may involve both confined diffusion within the particles and long-range diffusion between particles. By performing biexponential fitting of the PFG NMR attenuation curves, the diffusion coefficients and respective molecular fractions for these two diffusion modes can be separately determined. As the diffusion time increases, the molecular fraction associated with the long-range diffusion mode increases significantly, indicating that elevated temperatures facilitate the ability of iso-octane molecules to overcome mass transfer barriers within the particles.²⁹²

When investigating the characteristics of multicomponent diffusion under saturated conditions within zeolites, the researchers employed pure-component Maxwell–Stefan diffusion coefficients to calculate the diffusion flux under saturation. Lobo et al. proposed that, under saturated conditions, molecular hopping processes become highly correlated, leading to a significant increase in the ratio of single-component diffusion coefficients to intercomponent exchange coefficients. For narrow-pore structures, such as LTA and DDR zeolites, this correlation is relatively weak, allowing for the approximation of independent diffusion equations. In contrast, for large-pore structures like MFI and FAU zeolites, the correlation is markedly enhanced.²⁹³ In conclusion, we summarized representative zeolite materials for selectivity and reactivity in various applictions (Table 2), providing a reference for the future design and application optimization of zeolite materials. Hydrothermal synthesis and seed-assisted crystallization remain the primary methods for preparing zeolite-based catalysts. These are often complemented by techniques such as incipient wetness impregnation, ion exchange, and post-synthetic modification to introduce diverse metal centers and finely tune active sites. The construction of hierarchical structures, along with the use of binder-free and template-directed strategies, can significantly enhance diffusion properties and intercrystalline surface activity, thereby improving performance in large-molecule reactions or mass transfer processes.

Table 2 Several representative zeolite materials with adsorption and diffusion property applications

Topology	Sample	Synthesis method	Performance	Section	Ref.
FAU	Ni@FAU	Hydrothermal reaction	Dynamic separation selectivities of 100 in acetylene/ethylene	3.1.1.1/4.3	132
	Cu@FAU	Hydrothermal reaction	98–99% acetylene in acetylene/carbon dioxide	3.1.1.1/4.3	139
	Pt–FAU	Incipient wetness impregnation	Ethyl propyl ether selectivity of 34%.	3.1.2/4.3	294
	Ru–Na/FAU	Exchange and impregnate	>99% CH4 carbon yield in reactive carbon dioxide capture	3.1.2/4.3	295
	H–FAU zeolites	Moderate ion exchange	Dehydrogenation to cracking selectivity (kD/kC ratio) of ca. 7	3.1.2/4.3	296
	Hierarchical USY	Post-synthetic strategy	Uptake of ca. 2 mmol g^−1, p/p0 < 0.1	3.1.2/4.2	297
	Hierarchical Y	Unbiased leaching with NH4F	Higher conversion in n-C8 hydroconversion	3.1.2/4.2	140
	Ba–Y	Ion exchange	80% NOx conversion at 463–513 K	3.1.1.2/4.3	298
	Hydrophobic FAU-supported TiO2	Two-step method	Degradation rate k = 10.1 in photocatalytic degradation of 2-propanol	3.1.2/4.2	299
CHA	SAPO-34	High-pressure steam	Methanol conversion 100%	3.2.2/4.4	300
	Hierarchical SSZ-13	C22-4-4·Br2 addition in a precursor gel	Methanol conversion 98%	3.2.2/4.2	174
	Zn^2+ containing SSZ-13	Ion exchange	0.67 mmol CO2/g-zeolite	3.2.1/4.3	159
	Pd/H–CHA	Incipient wetness impregnation	57 μmol NO per g	3.2.1/4.3	301
	In-, Ga–CHA	Incipient wetness impregnation	98% C2H4 selectivity of ethane dehydrogenation	3.2.2/4.3	302
	Fe–CHA	Hydrothermal reaction	95% (MeOH + DME) selectivity	3.2.2/4.3	303
MFI	Pt/MFI nanosheet	Secondary nucleation assisted hydrothermal reaction	80% selectivity of linear alkanes in n-hexadecane hydrocracking	3.3.2/4.3	185
	MFI membrane	Nanocrystal-seeded growth	p-Xylene permeance of 0.56 × 10^−6 mol Pa^−1 m^−2 s^−1 at 150 °C	3.3.2/4.5	304
	NbAlS-1	Hydrothermal reaction	96% GVL conversion	3.3.2/4.3	305
	MFI membrane	b-Oriented MFI seed layer	CO2 permeance of 51 × 10^−7 mol m^−2 s^−1 Pa^−1	3.3.1/4.5	306
	Pd–SnOx@MFI	Hydrothermal reaction	Hydrogen peroxide productivity of ca. 10 170 mmol g Pd^−1 h^−1.	3.3.3/4.3	307
	Pt/In–ZSM-5	Two-step method	Ca. 48% propane conversion	3.3.3/4.3	308
	Na–ZSM-5	Ion exchange	3.5 mmol g^−1, 2.1 mmol g^−1 of water and methanol loadings	3.3.3/4.4	309
MOR	Mordenite with ZIF fragments	Hydrothermal reaction	139 kinetic selectivity at 25 °C of propene/propane	3.4/4.2	310
	Fe–MOR	One-pot template and binder-free process	CO2 uptake of 5.68/3.89 mmol g^−1	3.4/4.3	204
	MOR-nrod	Hydrothermal reaction	Initial activity of 85% in the cumene synthesis reaction	3.4/4.2	311
	Plate-MOR	Hydrothermal reaction under tumbling conditions	44% DME conversion	3.4/4.2	312
*BEA	Ga–BEA	Incipient wetness impregnation	Propene selectivity of 82%	3.5/4.3	313
	Pd@beta	Seed-directed synthesis	>99% selectivity of 4-aminobenzaldehyde	3.5/4.3	205
	Dealuminated BEA	Two-step process	Higher ethene hydroformylation activity	3.5/4.1	314
	Sn–Ni–beta	Hydrothermal reaction	71.2% yield of glucose to methyl lactate	3.5/4.3	315
	M–BEA	Ion exchange	101.3 kPa at 273, 298 K of CO2 adsorption isotherms	3.5/4.3	316

---

5. Conclusions and outlook

Zeolite materials, with their highly ordered microporous structure, abundant framework acidity, and tunable pore sizes, have long been central to catalysis, adsorption, and separation. A deeper understanding of the adsorption and diffusion processes within zeolites can significantly enhance the selectivity of existing industrial processes and provide a vital conceptual foundation for the development of the next generation of high-performance porous materials. By integrating advanced characterization techniques with multiscale simulations, we can more comprehensively explore how these processes drive performance in various catalytic systems.

(1) First, adsorption and diffusion are established as essential steps in catalytic processes, distinguishing between physical and chemical adsorption and outlining various diffusion mechanisms. By applying models, such as Maxwell–Stefan and IAST, researchers have addressed complex phenomena like competitive adsorption and non-equilibrium diffusion, offering deeper insights into these intricate processes.

(2) Techniques, such as Langmuir model, QENS, PFG NMR, GC, and ZLC methods, combined with advanced in situ spectroscopy, diffraction, electron microscopy, and integrated theoretical approaches of DFT, MC and MD simulation, offer unprecedented resolution in tracking molecular transport. These methodologies yield detailed maps of diffusion pathways, potential energy barriers, and adsorbate configurations, greatly refining our grasp of pore environments and host–guest interactions.

(3) Different pore architectures confer distinct diffusion properties and adsorption strengths. For instance, a large-pore FAU zeolite facilitates bulky molecule transformations and separation, whereas a small-pore CHA zeolite excels in precise molecular sieving and methane activation. The interplay of pore geometry, defect engineering, and acid site tuning in MFI and other frameworks drives performance enhancements in key reactions. Multifunctional design, such as incorporating transition metals, creating hierarchical porosities, and forming bifunctional composites enables fine-tuned interactions, improved selectivity and longevity in catalysts.

(4) This review underscores the interplay of acid site strength and distribution, extra-framework cations, pore size modulation, and environmental factors in governing adsorption and diffusion behaviors. By employing post-synthetic modifications and cation exchange strategies, researchers can regulate adsorption and diffusion kinetics, leading to enhanced catalytic performance.

Industrial processes often involve multicomponent systems with large flow rates and fluctuating operational conditions (e.g., temperature, pressure, and solvent environment), leading to variations in adsorption saturation, diffusion pathways, and acid site states within the pores. To address this, more refined spatiotemporal characterization techniques, such as photon-homologous diffraction imaging, transient isotopic spectroscopy, ultrafast laser spectroscopy, and multiphysical-field coupled simulations are needed to capture real-time changes in the zeolite pore environment under operational conditions. This will lay the foundation for designing more intelligent, multi-channel cooperative zeolite microenvironments.

Under liquid-phase or high-humidity gas-phase conditions, competition between solvent and water molecules for adsorption, hydrogen-bond network formation, and localized polarity of the pore environment significantly impacts the zeolite mass transfer and acid site modulation. For example, water or polar molecules can promote C–H bond activation through protonation or cluster formation but may also compete with target molecules for adsorption and inhibit reactions. The aggregation and rearrangement of organic solvents can alter pore volume and diffusion pathways. Future work could employ variable temperature in situ infrared/Raman spectroscopy, advanced NMR, and X-ray absorption spectroscopy, coupled with interface dynamics simulations, to quantitatively dissect the effects of solvents, guest molecules, and framework interactions on mass transport and reaction pathways.

Numerous studies have shown that the complex multidimensional relationships between zeolite structures (topology, Si/Al ratio, metal ion exchange, pore size distribution, etc.) and catalytic performance can be mined and predicted using machine learning and high-throughput computing. By building data platforms, performing feature engineering, and training models, we can quickly identify the optimal zeolite synthesis routes or modification strategies and accurately predict adsorption, diffusion, and product distribution performances. This could provide the theoretical and algorithmic foundation for the custom design of efficient zeolites. Artificial intelligence, coupled with high-throughput laboratory experiments, holds the potential for groundbreaking breakthroughs.

The adsorption and diffusion properties of zeolites under extreme conditions, such as electrocatalysis, photocatalysis, and high-temperature or supercritical environments remain an area of ongoing exploration. For example, in CO₂ electro-reduction reactors, zeolite nanofilms or microporous layers can be coated onto electrode surfaces to selectively adsorb CO₂ and suppress H₂ formation, enhancing the selectivity for target products. In biomass platform molecule conversions, zeolites face challenges in efficient mass transport and coke resistance in highly polar and viscous reaction media. In-depth analysis of structural evolution and molecular diffusion pathways under these extreme conditions will open new applications for zeolites in renewable energy conversion and green chemical processes.

In summary, the core of zeolite research lies in the precise design of pores and active sites, a deep understanding of adsorption–diffusion mechanisms, and dynamic optimization under various operating conditions. Current research is shifting away from merely pursuing higher surface areas or stronger acidity towards the coordinated regulation of pore environments, guest molecules, and solvent media. This shift is further enhanced by the integration of novel porous materials, nanotechnology, and intelligent algorithms. Looking ahead, the continuous advancements in in situ and quasi-in situ characterization, data-driven multiscale simulations, and controllable integration of porous structures will enable the development of the next generation of zeolites and composite porous materials. These innovations promise to deliver high efficiency and stability in complex chemical and energy systems, laying a solid foundation for greener, more sustainable chemical and energy conversion processes.

Conflicts of interest

There are no conflicts to declare.

Data availability

No primary research results, software or code have been included and no new data were generated or analysed as part of this review.

Acknowledgements

This work was supported by the National Key Research and Development Program of the Ministry of Science and Technology of China (2021YFA1501201), the National Natural Science Foundation of China (Grant No. 22272083, 22325405, 22432005 and 22321002), the Fundamental Research Funds for the Central Universities (Nankai University) and the Dalian Science and Technology Talent Innovation Program (Grant No. 2024RG009).

Notes and references

1. Y. Li and J. Yu, Chem. Rev., 2014, 114, 7268–7316.
2. M. E. Davis, Nature, 2002, 417, 813–821.
3. B. M. Weckhuysen and J. Yu, Chem. Soc. Rev., 2015, 44, 7022–7024.
4. B. Smit and T. M. Saesen, Chem. Rev., 2008, 108, 4125–4184.
5. E. Perez-Botella, S. Valencia and F. Rey, Chem. Rev., 2022, 122, 17647–17695.
6. R. Shah, M. C. Payne, M.-H. Lee and M.-H. Lee, Science, 1996, 271, 1395–1397.
7. D. Barthomeuf, Catal. Rev., 1996, 38, 521–612.
8. G. E. Boyd, A. W. Adamson and L. S. Myers, J. Am. Chem. Soc., 1947, 69, 2836–2848.
9. P. Nugent, Y. Belmabkhout, S. D. Burd, A. J. Cairns, R. Luebke, K. Forrest, T. Pham, S. Ma, B. Space, L. Wojtas, M. Eddaoudi and M. J. Zaworotko, Nature, 2013, 495, 80–84.
10. S. Wang and Y. Peng, Chem. Eng. J., 2010, 156, 11–24.
11. C. Wu and T. Bein, Science, 1994, 264, 1757–1759.
12. W. Steele, Chem. Rev., 1993, 93, 2355–2378.
13. D. Saha, Z. Bao, F. Jia and S. Deng, J. Am. Chem. Soc., 2005, 127(51), 17998–17999.
14. K. Cheng, B. Gu, X. Liu, J. Kang, Q. Zhang and Y. Wang, Angew. Chem., Int. Ed., 2016, 55, 4725–4728.
15. S. M. Kuznicki, V. A. Bell, S. Nair, H. W. Hillhouse, R. M. Jacubinas, C. M. Baunbarth, B. H. Toby and M. Tsapatsis, Nature, 2001, 412, 720–724.
16. Y. Tao, H. Kanoh and K. Kaneko, J. Am. Chem. Soc., 2003, 125, 6044–6045.
17. A. Corma, M. J. Díaz-Cabañas, J. L. Jordá, C. Martínez and M. Moliner, Nature, 2006, 443, 842–845.
18. M. R. Hudson, W. L. Queen, J. A. Mason, D. W. Fickel, R. F. Lobo and C. M. Brown, J. Am. Chem. Soc., 2012, 134, 1970–1973.
19. P. Cnudde, R. Demuynck, S. Vandenbrande, M. Waroquier, G. Sastre and V. Van Speybroeck, J. Am. Chem. Soc., 2020, 142, 6007–6017.
20. V. L. Sushkevich, D. Palagin, M. Ranocchiarim and J. A. Bokhoven, Science, 2017, 356, 523–527.
21. K. Gugeler, J. Kästner and M. Dyballa, ACS Catal., 2025, 15, 4798–4816.
22. J. Shang, G. Li, R. Singh, Q. Gu, K. M. Nairn, T. J. Bastow, M. Medhekar, C. M. Doherty, A. J. Hill, J. Z. Liu and P. A. Webley, J. Am. Chem. Soc., 2012, 134, 19246–19253.
23. J. Pérez-Ramírez, C. H. Christensen, K. Egeblad, C. H. Christensen and J. C. Groen, Chem. Soc. Rev., 2008, 37, 2530–2542.
24. S. M. Csicsery, Pure Appl. Chem., 1986, 58, 841–856.
25. B. Smit and T. Maesen, Nature, 2008, 541, 671–678.
26. M. Hartmann, Angew. Chem., Int. Ed., 2004, 43, 5880–5882.
27. J. Huang, Y. Jiang, V. R. R. Marthala and M. Hunger, J. Am. Chem. Soc., 2008, 130, 12642–12644.
28. J. Kärger, R. Valiullin, S. Brandani, J. Caro, C. Chmelik, B. F. Chmelka, M.-O. Coppens, S. Farooq, D. Freude, H. Jobic, M. Kruteva, E. Mangano, R. Pini, W. S. Price, A. Rajendran, P. I. Ravikovitch, G. Sastre, R. Q. Snurr, A. G. Stepanov, S. Vasenkov, Y. Wang and B. M. Weckhuysen, Pure Appl. Chem., 2025, 97, 1–89.
29. A. Zhokh and P. Strizhak, Microporous Mesoporous Mater., 2019, 282, 22–28.
30. F. J. Keil, R. Krishna and M. Coppens, Rev. Chem. Eng., 2000, 16, 71–197.
31. C. M. Bidabehere and U. Sedran, Ind. Eng. Chem. Res., 2001, 40, 530–535.
32. J. Kärger and S. Vasenkov, Microporous Mesoporous Mater., 2005, 85, 195–206.
33. J. Xiao and J. Wei, Chem. Eng. Sci., 1992, 47, 1123–1141.
34. P. F. Zito, A. Caravella, A. Erunetti, E. Drioli and G. Barbieri, J. Membr. Sci., 2018, 564, 166–173.
35. J. Liu, Z. Cheng, J. Wei, Q. Zhang, X. Chen, Y. Cen and L. Li, Chem. Eng. Sci., 2017, 172, 117–124.
36. S. Fritzsche, M. Wolfsberg and R. Haberlandt, Chem. Phys., 2003, 289, 321–333.
37. S. Fritzsche, M. Wolfsberg, R. Haberlandt, P. Demontis and G. B. Suffritti, Chem. Phys. Lett., 1998, 296, 253–258.
38. F. Kapteijn, J. A. Moulijn and R. Krishna, Chem. Eng. Sci., 2000, 55, 2923–2930.
39. R. Krishna and R. Baur, Chem. Eng. J., 2004, 97, 37–45.
40. R. Krishna and J. A. Wesselingh, Chem. Eng. Sci., 1997, 52, 861–911.
41. Y. Hasegawa, K. Kusakabe and S. Morooka, Chem. Eng. Sci., 2001, 56, 4273–4281.
42. K. Hahn, J. Kärger and V. Kukla, Phys. Rev. Lett., 1996, 76, 2762.
43. V. Kukla, J. Kornatowski, D. Demuth, I. Girnus, H. Pfeifer, L. V. C. Rees, S. Schunk, K. K. Unger and J. Kärger, Science, 1996, 272, 702–704.
44. D. Dubbeldam, S. Calero, T. L. M. Maesen and B. Smit, Phys. Rev. Lett., 2003, 90, 245901.
45. P. K. Ghorai, S. Yashonath, P. Demontis and G. B. Suffritti, J. Am. Chem. Soc., 2003, 125, 7116–7123.
46. L. A. Clark, G. T. Ye and R. Q. Snurr, Phys. Rev. Lett., 2000, 84, 2893.
47. A. Einstein, Ann. Phys., 1905, 322, 549–560.
48. F. Feil, S. Naumov, J. Michaelis, R. Valiullin, D. Enke, J. Kärger and C. Bräuchle, Angew. Chem., Int. Ed., 2012, 51, 1152–1155.
49. R. Baur and R. Krishna, Chem. Eng. J., 2004, 99, 105–116.
50. M. Hartmann, A. G. Machoke and W. Schwieger, Chem. Soc. Rev., 2016, 45, 3313–3330.
51. C. H. Christensen, K. Johannsen, E. Törnqvist, I. Schmidt, H. Topsøe and C. H. Christensen, Catal. Today, 2007, 128, 117–122.
52. A. F. Combariza, D. A. Gomez and G. Sastre, Chem. Soc. Rev., 2013, 42, 114.
53. V. R. Choudhary, K. C. Mondal and S. A. R. Mulla, Angew. Chem., Int. Ed., 2005, 44, 4381–4385.
54. V. R. Choudhary, A. K. Kinage and T. V. Choudhary, Science, 1997, 275, 1286–1288.
55. J. He, T. Xu, Z. Wang, Q. Zhang, W. Deng and Y. Wang, Angew. Chem., Int. Ed., 2012, 51, 2438–2442.
56. Z. Jin, L. Wang, E. Zuidema, K. Mondal, M. Zhang, J. Zhang, C. Wang, X. Meng, H. Yang, C. Mesters and F. S. Xiao, Science, 2020, 367, 193–197.
57. D. Fraenkel and D. C. Gates, J. Am. Chem. Soc., 1980, 102, 2478–2480.
58. F. Jiao, J. Li, X. Pan, J. Xiao, H. Li, H. Ma, M. Wei, Y. Pan, Z. Zhou, M. Li, S. Miao, J. Li, Y. Zhu, D. Xiao, T. He, J. Yang, F. Qi., Q. Fu and X. Bao, Science, 2016, 351, 1065–1068.
59. J. Bao, J. He, Y. Zhang, Y. Yoneyama and N. Tsubaki, Angew. Chem., Int. Ed., 2008, 47, 353–356.
60. I. Buurmans, J. Ruiz-Martínez, W. Knowles, D. Beek, J. Bergwerff, E. Vogt and B. Weckhuysen, Nat. Chem., 2011, 3, 862–867.
61. G. J. Kramer, R. A. van Santen, C. A. Emeis and A. K. Nowak, Nature, 1993, 363, 529–531.
62. G. T. Whiting, N. Nikolopoulos, L. Nikolopoulos, A. D. Chowdhury and B. M. Weckhuysen, Nat. Chem., 2019, 11, 23–31.
63. T. Baerdemaeker and D. Vos, Nat. Chem., 2013, 5, 89–90.
64. P. Xie, T. Pu, G. Aranovich, J. Guo, M. Donohue, A. Kulkarni and C. Wang, Nat. Catal., 2021, 4, 144–156.
65. S. Horike, S. Shimomura and S. Kitagawa, Nat. Chem., 2009, 1, 695–704.
66. H. Zeng, M. Xie, T. Wang, R. Wei, X. Xie, Y. Zhao, W. Lu and D. Li, Nature, 2021, 595, 542–548.
67. I. Langmuir, J. Am. Chem. Soc., 1918, 40, 1361–1403.
68. D. Saha, Z. Bao, F. Jia and S. Deng, Environ. Sci. Technol., 2010, 44, 1820–1826.
69. U. Habiba, A. M. Afifi, A. Salleh and B. C. Ang, J. Hazard. Mater., 2017, 322, 182–194.
70. J. Lill, C. Dejoie, C. Giacobbe and A. N. Fitch, J. Phys. Chem. C, 2022, 126, 2214–2225.
71. A. Moscatelli, Z. Liu, X. Lei, J. Dyer, L. Abrams, M. F. Ottaviani and N. J. Turro, J. Am. Chem. Soc., 2008, 130, 11344–11354.
72. M. Kruteva, Adsorption, 2021, 27, 875–889.
73. H. Jobic and D. N. Theodorou, Microporous Mesoporous Mater., 2007, 102, 21–50.
74. F. Leroy and H. Jobic, Chem. Phys. Lett., 2005, 406, 375–380.
75. A. J. O’Malley, V. G. Sakai, I. P. Silverwood, N. Dimitratos, S. F. Parker and C. R. A. Catlow, Phys. Chem. Chem. Phys., 2016, 18, 17294–17302.
76. C. Hernandez-Tamargo, I. P. Silverwood, A. J. O’Malley and N. H. D. Leeuw, Top. Catal., 2021, 64, 707–721.
77. I. Déroche, G. Maurin, B. J. Borah, S. Yashonath and H. Jobic, J. Phys. Chem. C, 2010, 114, 5027–5034.
78. J. A. Dunne, R. Mariwala, M. Rao, S. Sircar, R. J. Gorte and A. L. Myers, Langmuir, 1996, 12, 5888–5895.
79. N. Chang, Z.-Y. Gu and X.-P. Yan, J. Am. Chem. Soc., 2010, 132, 13645–13647.
80. Q. Lu, Z.-F. Zhang, C.-Q. Dong and X.-F. Zhu, Energies, 2010, 3, 1805–1820.
81. M. Bjørgen, U. Olsbye, S. Svelle and S. Kolboe, Catal. Lett., 2004, 93, 37–40.
82. A. J. Hernández-Maldonado and R. T. Yang, J. Am. Chem. Soc., 2004, 126, 992–993.
83. H. J. Park, H. S. Heo, J. K. Jeon, J. Kim, R. Ryoo, K. E. Jeong and Y. K. Park, Appl. Catal., B, 2010, 95, 365–373.
84. J. M. Halket, D. Waterman, A. M. Przyborowska, R. K. P. Patel, P. D. Fraser and P. M. Bramley, J. Exp. Bot., 2005, 56, 219–243.
85. I. Gusev, X. Huang and C. Horváth, J. Chromatogr. A, 1999, 855, 273–290.
86. F. Brandani and D. M. Ruthven, Ind. Eng. Chem. Res., 2004, 43, 8339–8344.
87. J. F. M. Denayer, L. I. Devriese, S. Couck, J. Martens, R. Singh, P. A. Webley and G. V. Baron, J. Phys. Chem. C, 2008, 112, 16593–16599.
88. A. R. Teixeira, X. Qi, W. C. Conner, T. J. Mountziaris, W. Fan and P. J. Dauenhauer, Chem. Mater., 2015, 27, 4650–4660.
89. W. Dai, X. Wang, G. Wu, N. Guan, M. Hunger and L. Li, ACS Catal., 2011, 1, 292–299.
90. X. Ma, S. Velu, J. H. Kim and C. Song, Appl. Catal., B, 2005, 56, 137–147.
91. J. Kärger and H. Pfeifer, Zeolites, 1987, 7, 90–107.
92. J. Kärger and H. Spindler, J. Am. Chem. Soc., 1991, 113, 7571–7574.
93. S. Vasenkov, W. Böhlmann, P. Galvosas, O. Geier, H. Liu and J. Kärger, J. Phys. Chem. B, 2001, 105, 5922–5927.
94. P. Cnudde, E. A. Redekop, W. Dai, N. G. Porcaro, M. Waroquier, S. Bordiga, M. Hunger, L. Li, U. Olsbye and V. Van Speybroeck, Angew. Chem., Int. Ed., 2021, 60, 10016–10022.
95. M. Holz, S. R. Heil and A. Sacco, Phys. Chem. Chem. Phys., 2000, 2, 4740–4742.
96. D. Sinnaeve, Concepts Magn. Reson., Part A, 2012, 40, 39–65.
97. P. Kortunov, S. Vasenkov, J. Kärger, R. Valiullin, P. Gottschalk, M. F. Elía, M. Perez, M. Stöcker, B. Drescher, G. McElhiney, C. Berger, R. Gläser and J. Weitkamp, J. Am. Chem. Soc., 2005, 127, 13055–13059.
98. W. Dai, M. Scheibe, L. Li, N. Guan and M. Hunger, J. Phys. Chem. C, 2012, 116, 2469–2476.
99. S. Gao, Z. Liu, S. Xu, A. Zheng, P. Wu, B. Li, X. Yuan, Y. Wei and Z. Liu, J. Catal., 2019, 377, 51–62.
100. Z. Liu, C. Lou, J. Yuan, X. Tang, Y. Fan, J. Qi, R. Zhang, P. Peng, G. Liu, S. Xu and A. Zheng, J. Am. Chem. Soc., 2025, 147, 6126–6136.
101. L. Liu, M. Lopez-Haro, J. A. Perez-Omil, M. Boronat, J. J. Calvino and A. Corma, Nat. Commun., 2022, 13, 821.
102. V. Van Speybroeck, M. Bocus, P. Cnudde and L. Vanduyfhuys, ACS Catal., 2023, 13, 11455–11493.
103. J. Nicholas, Top. Catal., 1997, 4, 157–171.
104. A. Zheng, L. Chen, J. Yang, M. Zhang, Y. Su, Y. Yue, C. Ye and F. Deng, J. Phys. Chem. B, 2005, 109, 24273–24279.
105. T. Ammouli, J. L. Paillaud, H. Nouali, R. Stephan, M. C. Hanf, P. Sonnet and I. Deroche, J. Phys. Chem. C, 2021, 125, 19405–19416.
106. Y. Li, X. Min, W. Li, Q. Wang, M. Shang, Y. Yu, P. Tao and T. Sun, Microporous Mesoporous Mater., 2022, 330, 111631.
107. R. Y. Brogaard, M. Komurcu, M. M. Dyballa, A. Botan, V. Van Speybroeck, U. Olsbye and K. D. Wispelere, ACS Catal., 2019, 9, 5645–5650.
108. M. Bocus, S. Vandenhaute and V. Van Speybroeck, Angew. Chem., Int. Ed., 2025, 64, e202413637.
109. Z. Mi, S. Li, W. Liu, J. Wang, D. Deng, P. Liu, P. Tian, C. Liu, W. Yan, K. Zhu, Z. Wang, Z. Liu and J. Yu, J. Am. Chem. Soc., 2025, 147, 16922–16934.
110. T. Sonoda, T. Maruo, Y. Yamasaki, N. Tsunoji, Y. Takamitsu, M. Sadakane and T. Sano, J. Mater. Chem. A, 2015, 3, 857–865.
111. M. Dusselier, J. E. Schmidt, R. Moulton, B. Haymore, M. Hellums and M. E. Davis, Chem. Mater., 2015, 27, 2695–2702.
112. A. Simmen, L. B. McCusker, Ch Baerlocher and W. M. Meier, Zeolites, 1991, 11, 654–661.
113. Y. Kubota, H. Maekawa, S. Miyata, T. Tatsumi and T. Sugi, Microporous Mesoporous Mater., 2007, 101, 115–126.
114. M. A. Camblor, A. Corma and S. Valencia, Microporous Mesoporous Mater., 1998, 25, 59–74.
115. C. Fan, Z. Chen, L. Pang, S. Ming, X. Zhang, K. B. Albert, P. Liu, H. Chen and T. Li, Appl. Catal., A, 2018, 550, 256–265.
116. J. Zhang, M. Li, Y. Lin, C. Liu, X. Liu, L. Bai, D. Hu, G. Zeng, Y. Zhang, W. Wei and Y. Sun, Microporous Mesoporous Mater., 2016, 219, 103–111.
117. E.-P. Ng, D. Chateigner, T. Bein, V. Valtchev and S. Mintova, Science, 2012, 335, 70–73.
118. S. Ferdov, K. Tsuchiya, N. Tsunoji and T. Sano, Microporous Mesoporous Mater., 2019, 276, 154–159.
119. D. Verboekend, N. Nuttens, R. Locus, J. Van Aelst, P. Verolme, J. C. Groen, J. Perez-Ramirez and B. F. Sels, Chem. Soc. Rev., 2016, 45, 3331–3352.
120. E. Catizzone, A. Aloise, M. Migliori and G. Giorrdano, J. Energy Chem., 2017, 26, 406–415.
121. T. Inoue, M. Itakura, H. Jon, Y. Oumi, A. Takahashi, T. Fujitani and T. Sano, Microporous Mesoporous Mater., 2009, 122, 149–154.
122. M. Palomino, A. Corma, F. Rey and S. Valencia, Langmuir, 2010, 26, 1910–1917.
123. F. Collins, A. Rozhkovskaya, J. G. Outram and G. J. Millar, Microporous Mesoporous Mater., 2020, 291, 109667.
124. H. J. Choi, D. Jo and S. B. Hong, Chem. Eng. J., 2022, 446, 137100.
125. J. M. Chezeau, L. Delmotte, J. L. Guth and Z. Gabelica, Zeolites, 1991, 11, 598–606.
126. H. G. Karge and E. Boldingh, Catal. Today, 1988, 3, 379–386.
127. Y. Wang, Y. Zhang, W. Liu, Q. Yu, Z. Liu, S. Xu, J. An, X. Li and X. Zhu, Microporous Mesoporous Mater., 2022, 346, 112282.
128. W. Chu, X. Ji, S. Liu, X. Zhu, S. Xie, F. Chen, Y. Wang, W. Xin and L. Xu, J. Mater. Chem. A, 2018, 6, 12244–12249.
129. Q. Ke, T. Sun, H. Cheng, H. Chen, X. Liu, X. Wei and S. Wang, Chem. – Eur. J., 2017, 12, 1043–1047.
130. Y. Luo, S. Smeets, Z. Wang, J. Sun and W. Yang, Chem. – Eur. J., 2019, 25, 2184–2188.
131. P. Zhao, L. Ye, G. Li, C. Huang, S. Wu, P. L. Ho, H. Wang, T. Yoskamtorn, D. Sheptyakov, G. Cibin, A. I. Kirkland, C. C. Tang, A. Zheng, W. Xue, D. Mei, K. Suriye and S. C. E. Tsang, ACS Catal., 2021, 11, 3530–3540.
132. Y. Chai, X. Han, W. Li, S. Liu, S. Yao, C. Wang, W. Shi, I. da-Sillva, P. Manuel, Y. Cheng, L. D. Daemen, A. J. Ramirez-Cuesta, C. C. Tang, L. Liang, S. Yang, N. Guan and L. Li, Science, 2020, 368, 1002–1006.
133. W. Li, Y. Chai, G. Wu and L. Li, J. Phys. Chem. Lett., 2022, 13, 11419–11429.
134. P. Zhao, L. Ye, Z. Sun, B. T. W. Lo, H. Woodcock, C. Huang, C. Tang, A. I. Kirkland, D. Mei and S. C. E. Tsang, J. Am. Chem. Soc., 2018, 140, 6661–6667.
135. B. Yue, X. Lian, S. Liu, G. Wu, J. Xu and L. Li, Chem. Eng. J., 2023, 462, 142300.
136. A. M. Najafi, S. Soltanali and H. Ghassabzadeh, Chem. Eng. J., 2023, 468, 143719.
137. K. Goldyn, C. Anfray, S. Komaty, V. Ruaux, C. Helaine, R. Retook, S. Valable, V. Valtchev and S. Mintova, Microporous Mesoporous Mater., 2019, 280, 271–276.
138. H. Yahiro, Y. Ohmori and M. Shiotani, Microporous Mesoporous Mater., 2005, 83, 165–171.
139. S. Liu, X. Han, Y. Chai, G. Wu, W. Li, J. Li, I. da Silva, P. Manuel, Y. Cheng, L. L. Daemen, A. J. Ramirez-Cuesta, W. Shi, N. Guan, S. Yang and L. Li, Angew. Chem., Int. Ed., 2021, 60, 6526–6532.
140. Z. Qin, S. Zeng, G. Melinte, T. Bucko, M. Baodawi, Y. Shen, J. P. Gilson, O. Ersen, Y. Wei, Z. Liu, X. Liu, Z. Yan, S. Xu, V. Valtchev and S. Mintova, Adv. Sci., 2021, 8, 2100001.
141. Y. Pang, M. A. Ardagh, M. Shetty, A. Chatzidimitriou, G. Kumar, B. Vlaisavljevich and P. J. Dauenhauer, ACS Catal., 2021, 11, 5718–5735.
142. L. Qi, R. Alamillo, W. A. Elliott, A. Andersen, D. W. Hoyt, E. D. Walter, K. S. Han, N. M. Washton, R. M. Rioux, J. A. Dumersic and S. L. Scott, ACS Catal., 2017, 7, 3489–3500.
143. P. Wongthong, R. Rungsirisakun, M. Probst and J. Limtrakul, Microporous Mesoporous Mater., 2007, 100, 160–166.
144. D. Zhai, L. Zhao, J. Gao and C. Xu, Phys. Chem. Chem. Phys., 2012, 14, 7296–7303.
145. C. R. Martinez and B. L. Iverson, Chem. Sci., 2012, 3, 2191–2201.
146. D. Gao, A. Duan, X. Zhang, Z. Zhao, H. E, Y. Qin and C. Xu, Chem. Eng. J., 2015, 270, 176–186.
147. C. Zhou, N. Wang, Y. Qian, X. Liu, J. Caro and A. Huang, Angew. Chem., Int. Ed., 2016, 55, 12678–12682.
148. T. Takata, N. Tsunoji, Y. Takamitsu, M. Sdakane and T. Sano, Microporous Mesoporous Mater., 2017, 246, 89–101.
149. N. Martin, M. Moliner and A. Corma, Chem. Commun., 2015, 51, 9965–9968.
150. U. Deka, I. Lezcano-Gonzalez, B. M. Weckhuysen and A. M. Beale, ACS Catal., 2013, 3, 413–427.
151. S. Kieger, G. Delahay and B. Coq, Appl. Catal., B, 2000, 25, 1–9.
152. F. Jousse and S. M. Auerbach, J. Chem. Phys., 1997, 107, 9629–9639.
153. F. Jousse, D. P. Vercauteren and S. M. Auerbach, J. Phys. Chem. B, 2000, 104, 8768–8778.
154. F. Jousse, S. M. Auerbach and D. P. Vercauteren, J. Phys. Chem. B, 2004, 104, 2360–2370.
155. H. Ramanan, S. M. Auerbach and M. Tsapatsis, J. Phys. Chem. B, 2004, 108, 17171–17178.
156. H. Jobic, H. Ramanan, S. M. Auerbach, M. Tsapatsis and P. Fouquet, Microporous Mesoporous Mater., 2006, 90, 307–313.
157. K. Muraoka, Y. Sada, A. Shimojima, W. Chaikittisilp and T. Okubo, Chem. Sci., 2019, 10, 8533.
158. M. Debost, P. B. Klar, N. Barrier, E. B. Clatworthy, J. Grand, F. Laine, P. Brázda, L. Palatinus, N. Nesterenko, P. Boullay and S. Mintova, Angew. Chem., Int. Ed., 2020, 59, 23491–23495.
159. D. Fu, Y. Park and M. E. Davis, Angew. Chem., Int. Ed., 2022, 61, e202112916.
160. J. Yang, J. Liu, P. Liu, L. Li, X. Tang, H. Shang, J. Li and B. Chen, Angew. Chem., Int. Ed., 2022, 61, e202116850.
161. S. Cao, L. Li, M. Guo, W. Xu, T. Liu, T. Xing, Z. Li, M. Wang, M. Wang and M. Wu, Chem. Mater., 2023, 35, 10119–10128.
162. X. Zhang, Z. Wang, Z. Chen, Y. Zhu, Z. Liu, F. Li, W. Zhou, Z. Dong, J. Fan and L. Liu, Appl. Catal., B, 2022, 317, 121771.
163. Y. Chen, R. Zhu, G. Chen, M. Zhao, M. Ju, R. Cai, J. Gao, M. Hong and S. Yang, Chem. Eng. J., 2024, 497, 154727.
164. S. Ghojavand, E. B. Clatworthy, B. Coasne, D. H. Piva, R. Guillet-Nicolas, I. C. Medeiris-Costa, M. Desmurs, V. Ruaux, V. Pugnet, P. Kumar-Gandhi and S. Mintova, Chem. Eng. J., 2024, 500, 157101.
165. H. Fang, P. Kamakoti, J. Zhang, S. Cundy, C. Paur, P. I. Ravikovitch and D. S. Sholl, J. Phys. Chem. C, 2012, 116, 10692–10701.
166. K. Mandal, Y. Gu, K. S. Westendorff, S. Li, J. A. Pihl, L. C. Grabow, W. S. Epling and C. Paolucci, ACS Catal., 2020, 10, 12801–12818.
167. D. Chen, H. Lei, W. Xiong, Y. Li, X. Ji, J. Y. Yang, B. Peng, M. Fu, P. Chen and D. Ye, ACS Catal., 2021, 11, 13891–13901.
168. D. Li, Y. Wei, T. Zhang, Y. Bai, Y. Qi, J. Han, L. Li and J. Yu, Adv. Funct. Mater., 2024, 2409837.
169. T. Jiang, F. Göltl, R. E. Bulo and P. Sautet, ACS Catal., 2014, 4, 2351–2358.
170. J. F. M. Denayer, I. Devriese, S. Couck, J. Martens, R. Singh, P. A. Webley and G. V. Baron, J. Phys. Chem. C, 2008, 112, 16593–16599.
171. D. H. Olson, M. A. Camblor, L. A. Villaescusa and G. H. Kuehl, Microporous Mesoporous Mater., 2004, 67, 27–33.
172. L. Yang, C. Wang, L. Zhang, W. Dai, Y. Chu, J. Xu, G. Wu, M. Gao, W. Liu, Z. Xu, P. Wang, N. Guan, M. Dyballa, M. Ye, F. Deng, W. Fan and L. Li, Nat. Commun., 2021, 12, 4661.
173. C. Wang, L. Yang, M. Gao, X. Shao, W. Dai, G. Wu, N. Guan, Z. Xu, M. Ye and L. Li, J. Am. Chem. Soc., 2022, 144, 21408–21416.
174. X. Zhu, J. P. Hofmann, B. Mezai, N. K osinov, L. Wu, Q. Qian, B. M. Weckhuysem, S. Asahina, J. Ruiz-Martinez and E. J. M. Hensen, ACS Catal., 2016, 6, 2163–2177.
175. M. Gao, H. Li, M. Yang, S. Gao, P. Wu, P. Tian, S. Xu, M. Ye and Z. Liu, Commun. Chem., 2019, 2, 43.
176. K. Na, M. Choi, W. Park, Y. Sakamoto, O. Terasaki and R. Ryoo, J. Am. Chem. Soc., 2010, 132, 4169–4177.
177. J. Kim, M. Choi and R. Ryoo, J. Catal., 2010, 269, 219–228.
178. Z. Li, D. Dittmann, C. Rieg, M. Benz and M. Dyballa, Catal. Sci. Technol., 2022, 12, 5189–5202.
179. B. Shen, H. Wang, H. Xiong, X. Chen, E. G. T. Bosch, I. Lazic, W. Qian and F. Wei, Nature, 2022, 607, 703–721.
180. B. T. W. Lo, L. Ye, J. Qu, J. Sun, J. Zheng, D. Kong, C. A. Murray, C. C. Tang and S. C. E. Tsang, Angew. Chem., Int. Ed., 2016, 55, 5981–5984.
181. X. Yang, R. Wang, J. Yang, W. Qian, Y. Zhang, X. Li, Y. Huang, T. Zhang and D. Chen, ACS Catal., 2020, 10, 3797–3806.
182. M. Rivallan, E. Seguin, S. Thomas, M. Lepage, N. Takagi, H. Hirata and F. Thibault-Starzyk, Angew. Chem., Int. Ed., 2010, 49, 785–789.
183. Y. Chai, S. Liu, Z. Zhao, J. Gong, W. Dai, G. Wu, N. Guan and L. Li, ACS Catal., 2018, 8, 8578–8589.
184. X. Liu, J. Shi, G. Yang, J. Zhou, C. Wang, J. Teng, Y. Wang and Z. Xie, Commun. Chem., 2021, 4, 107.
185. R. Brosius, P. J. Kooyman and J. C. Q. Fletcher, ACS Catal., 2016, 6, 7710–7715.
186. J. Gu, Z. Zhang, P. Hu, L. Ding, H. Xue, L. Peng, X. Guo, M. Lin and W. Ding, ACS Catal., 2015, 5, 6893–6901.
187. C. Wng, E. Guan, L. Wang, X. Chu, Z. Wu, J. Zhang, Z. Yang, Y. Jiang, L. Zhang, X. Meng, B. C. Gates and F. S. Xiao, J. Am. Chem. Soc., 2019, 141, 8482–8488.
188. M. H. Mahyuddin, A. Staykov, Y. Shiota and K. Yoshizawa, ACS Catal., 2016, 6, 8321–8331.
189. P. Peng, D. Stosic, A. Aitblal, A. Vimont, P. Bazin, X.-M. Liu, Z.-F. Yan, S. Mintova and A. Travert, ACS Catal., 2020, 10, 6822–6830.
190. J. Duan, W. Chen, C. Wang, L. Wang, Z. Liu, X. Yi, W. Fang, H. Wang, H. Wei, S. Xu, Y. Yang, Q. Yang, Z. Bao, Z. Zhang, Q. Ren, H. Zhou, X. Qin, A. Zheng and F. S. Xiao, J. Am. Chem. Soc., 2022, 144, 14269–14277.
191. K. Shen, W. Qian, N. Wang, C. Su and F. Wei, J. Am. Chem. Soc., 2013, 135, 15322–15325.
192. R. Khare, S. S. Arora and A. Bhan, ACS Catal., 2016, 6, 2314–2331.
193. C. Song, Y. Chu, M. Wang, H. Shi, L. Zhao, X. Guo, W. Yang, J. Shen, N. Xue, L. Peng and W. Ding, J. Catal., 2017, 349, 163–174.
194. J. Zhang, W. Qian, C. Kong and F. Wei, ACS Catal., 2015, 5, 2982–2988.
195. D. Dubbeldam, S. Calero, T. L. M. Maesen and B. Smit, Phys. Rev. Lett., 2003, 90, 245901.
196. K. Hahn, J. Kärger and V. Kukla, Phys. Rev. Lett., 1996, 76, 2762–2765.
197. G. J. L. D. Reijer, A. Schaefer, A. Hellman and P. A. Carlsso, Ind. Eng. Chem. Res., 2025, 64, 2025–2035.
198. H. S. Lee, N. S. Kim, D. Kwon, S. Lee, M. Numan, T. Jung, K. Cho, M. Mazui, H. S. Cho and C. Jo, Adv. Mater., 2021, 33, 2105398.
199. K. Cao, D. Fan, L. Li, B. Fan, L. Wang, D. Zhu, Q. Wang, P. Tian and Z. Liu, ACS Catal., 2020, 10, 3372–3380.
200. D. Lozano-Castello, W. Zhu, A. Linares-Solano, F. Kapteijn and J. A. Moulijn, Microporous Mesoporous Mater., 2006, 92, 145–153.
201. W. Chen, K. A. Tarach, X. Yi, Z. Liu, X. Tang, K. Góra-Marek and A. Zheng, Nat. Commun., 2022, 13, 7106.
202. K. Cao, D. Fan, M. Gao, B. Fan, N. Chen, L. Wang, P. Tian and Z. Liu, ACS Catal., 2022, 12, 1–7.
203. Y. Yang, J. Ding, C. Xu, W. Zhu and P. Wu, J. Catal., 2015, 325, 101–110.
204. Y. Zhou, J. Zhang, L. Wang, X. Cui, X. Liu, S. S. Wong, H. An, N. Yan, J. Xie, C. Yu, P. Zhang, Y. Du, S. Xi, L. Zheng, X. Cao, Y. Wu, C. Wang, H. Wen, L. Chen, H. Xing and J. Wang, Science, 2021, 373, 315–320.
205. J. Zhang, L. Wang, Y. Shao, Y. Wang, B. C. Gates and F. S. Xiao, Angew. Chem., Int. Ed., 2017, 56, 9747–9751.
206. R. E. Patet, M. Koehle, R. F. Lobo, S. Caratzoulas and D. G. Vachos, J. Phys. Chem. C, 2017, 121, 13666–13679.
207. J. S. Kruger, V. Choudhary, V. Nikolakis and D. G. Vlachos, ACS Catal., 2013, 3, 1279–1291.
208. B. Tang, W. Dai, G. Wu, N. Guan, L. Li and M. Hunger, ACS Catal., 2014, 4, 2801–2810.
209. S. Roy, K. Bakhmutsky, E. Mahmoud, R. F. Lobo and R. J. Gorte, ACS Catal., 2013, 3, 573–580.
210. O. Kwon, E. Z. Ayla, D. S. Potts and D. W. Flaherty, ACS Catal., 2023, 13, 6430–6444.
211. D. T. Bregante and D. W. Flaherty, ACS Catal., 2019, 9, 10951–10962.
212. A. A. Khechfe, T. B. Matha and Y. R. Leshkov, ACS Catal., 2023, 13, 6474–6485.
213. G. Qie, R. Liu, Q. Deng, D. Deng, M. Zhai, W. Liu, W. Dai, L. Han and K. Zhu, Small, 2025, 2411639.
214. W. J. M. V. Well, X. Cottin, J. W. D. Haan, R. A. V. Santen and B. Smit, Angew. Chem., Int. Ed., 1998, 37, 1081–1083.
215. J. Kärger, P. Kortunov, S. Vasenkov, L. Heinke, D. B. Shah, R. A. Rakoczy, Y. Traa and J. Weitkamp, Angew. Chem., Int. Ed., 2006, 45, 7846–7849.
216. T. Cordero-Lanzac, I. Capel Berdiell, A. Airi, S.-H. Chung, J. L. Mancuso, E. A. Redekop, C. Fabris, L. Figueroa-Quintero, J. C. Navarro de Miguel, J. Narciso, E. V. Ramos-Fernandez, S. Svelle, V. Van Speybroeck, J. Ruiz-Martínez, S. Bordiga and U. Olsbye, JACS Au, 2024, 4, 744–759.
217. K. Cho, H. S. Cho, L.-C. Menorval and R. Ryoo, Chem. Mater., 2009, 21, 5664–5673.
218. H. J. Choi, D. Jo, J. G. Min and S. B. Hong, Angew. Chem., Int. Ed., 2021, 60, 4307–4314.
219. Z. J. Berkson, M. F. Hsieh, S. Smeets, D. Gajan, A. Lund, A. Lesage, D. Xie, S. I. Zones, L. B. McCusker, C. Baerlocher and B. F. Chmelka, Angew. Chem., Int. Ed., 2019, 58, 6255–6259.
220. M. Fasano, T. Humplik, A. Bevilacqua, M. Tsapatsis, E. Chiavazzo, E. N. Wang and P. Asinari, Nat. Commun., 2016, 7, 12762.
221. Z. Liu, J. Yuan, J. M. Van Baten, J. Zhou, X. Tang, C. Zhao, W. Chen., X. Yi, R. Krishna, G. Sastre and A. Zheng, Sci. Adv., 2021, 7, eabf0775.
222. S. Askarli, I. Mukhambetov, S. H. Chung, M. Awwad, J. N. Almeida, S. Komaty, J. C. Navarro de Miguel, T. Li and J. Ruiz-Martínez, ACS Catal., 2024, 14, 13630–13639.
223. T. R. Josephson, R. F. DeJaco, S. Pahari, L. Ren, Q. Guo, M. Tsapatsis, J. I. Siepmann, D. G. Vlachos and S. Caratzoulas, ACS Catal., 2018, 8, 9056–9065.
224. M. Hu, C. Wang, Y. Chu, Q. Wang, S. Li, J. Xu and F. Deng, Angew. Chem., Int. Ed., 2022, 61, e202207400.
225. L. Peng, Y. Liu, N. Kim, J. E. Readman and C. P. Grey, Nat. Mater., 2005, 4, 216–219.
226. Z. Yu, A. Zheng, Q. Wang, L. Chen, J. Xu, J. P. Amoureux and F. Deng, Angew. Chem., Int. Ed., 2010, 49, 8657–8661.
227. G. Li, C. Foo, R. Fan., M. Zheng, Q. Wang, Y. Chu, J. Li, S. Day, P. Steadman, C. Tang, T. W. B. Lo, F. Deng and S. C. E. Tsang, Science, 2025, 387, 388–393.
228. C. Liu, I. Tranca, R. A. V. Santen, E. J. M. Hensen and E. A. Pidko, J. Phys. Chem. C, 2017, 121, 23520–23530.
229. M. Dyballa, Energy Fuels, 2023, 37, 18517–18559.
230. M. Boronat and A. Corma, ACS Catal., 2019, 9, 1539–1548.
231. M. J. Nash, A. M. Shough, D. W. Fickel, D. J. Doren and R. F. Lobo, J. Am. Chem. Soc., 2008, 130, 2460–2462.
232. X. Yi, H.-H. Ko, F. Deng, S.-B. Liu and A. Zheng, Nat. Protoc., 2020, 15, 3527–3555.
233. X. Yi, Y.-K. Peng, W. Chen, Z. Liu and A. Zheng, Acc. Chem. Res., 2021, 54, 2421–2433.
234. X. Yi, K. Liu, W. Chen, J. Li, S. Xu, C. Li, Y. Xiao, H. Liu, X. Guo, S.-B. Liu and A. Zheng, J. Am. Chem. Soc., 2018, 140, 10764–10774.
235. X. Yi, W. Chen, Y. Xiao, F. Liu, X. Yu and A. Zheng, J. Am. Chem. Soc., 2023, 145, 27471–27479.
236. X. Yi, K. Liu, W. Chen, J. Li, S. Xu, C. Li, Y. Xiao, H. Liu, X. Guo, S.-B. Liu and A. Zheng, J. Am. Chem. Soc., 2018, 140, 10764–10774.
237. Z. Wang, D. Xiao, K. Chen, C. Lou, L. Liang, S. Xu and G. Hou, ACS Catal., 2023, 13, 4960–4970.
238. Y. Ji, Z. Liu, Z. Zhao, P. Gao, X. Bao, K. Chen and G. Hou, ACS Catal., 2022, 12, 15288–15297.
239. R. Zhao, R. Khare, Y. Zhang, M. Sanchez-Sanchez, R. Bermejo-Deval, Y. Liu and J. A. Lercher, Nat. Catal., 2023, 6, 68–79.
240. J. Yuan, M. Gao, Z. Liu, X. Tang, Y. Tian, G. Ma, M. Ye and A. Zheng, Nat. Commun., 2023, 14, 1735.
241. R. Krishna and J. M. Van Baten, J. Phys. Chem. C, 2010, 114, 11557–11563.
242. J. Tan, H. Liang, W. Chen, J. Yuan, W. Dong, W. Liu, X. Yu, H. Shi and A. Zheng, Microporous Mesoporous Mater., 2024, 365, 112887.
243. D. Liu, A. Bhan, M. Tsapatsis and S. A. Hashimi, ACS Catal., 2011, 1, 7–17.
244. A. Galarneau, F. Guenneau, A. Gedeon, D. Mereib, J. Rodriguez, F. Fajula and B. Coasne, J. Phys. Chem. C, 2016, 120, 1562–1569.
245. H. Shi, S. Eckstein, A. Vjunov, D. M. Camaioni and J. A. Lercher, Nat. Commun., 2017, 8, 15442.
246. N. D. Witte, S. Robijns, J. F. M. Denayer, M. Dusselier and T. R. C. Van Assche, J. Phys. Chem. C, 2023, 127, 24393–24402.
247. W. Cha, N. C. Jeong, S. Song, H. Park, T. Pham, R. Harder, B. Lim, G. Xiong, D. Ahn, I. McNulty, J. Kim, K. B. Yoon, I. K. Robinson and H. Kim, Nat. Mater., 2013, 12, 729–734.
248. A. A. Blal, D. Stosic, P. Bazin, A. Vimont and A. Travert, Phys. Chem. Chem. Phys., 2023, 25, 27170.
249. J. C. Saint Remi, A. Lauerer, C. Chmelik, I. Vandendael, H. Terryn, G. V. Baron, J. F. M. Denayer and J. Kärger, Nat. Mater., 2016, 15, 401–406.
250. F. Hibbe, J. Caro, C. Chmelik, A. Huang, T. Kirchner, D. Ruthven, R. Valiullin and J. Kärger, J. Am. Chem. Soc., 2012, 134, 7725–7732.
251. I. L. C. Buurmans, J. Ruiz-Martínez, W. V. Knowles, D. V. D. Beek, J. A. Bergwerff, E. T. C. Vogt and B. M. Weckhuysen, Nat. Chem., 2011, 3, 862–867.
252. C. Liu, I. Tranca, R. A. V. Santen, E. J. M. Sensen and E. A. Pidko, J. Phys. Chem. C, 2017, 121, 23520–23530.
253. E. T. C. Vogt and B. M. Weckhuysen, Chem. Soc. Rev., 2015, 44, 7342–7370.
254. A. V. Yakimov, M. Ravi, R. Verel, V. L. Sushkevich, J. A. V. Bokhoven and C. Copéret, J. Am. Chem. Soc., 2022, 144, 10377–10385.
255. E. Grifoni, G. Piccini, J. A. Lercher, V. A. Glezakou, R. Rousseau and M. Parrinello, Nat. Commun., 2021, 12, 2630.
256. D. B. Lukyanov, T. Vazhnova, N. Cherkasov, J. L. Casci and J. J. Birtill, J. Phys. Chem. C, 2014, 118, 23918–23929.
257. A. Slawek, J. M. Vincent-Luna, K. Ogorzaly, S. Valencia, F. Rey, W. Makowski and S. Calero, J. Phys. Chem. C, 2019, 123, 29665–29678.
258. A. J. Jones, S. I. Zones and E. Iglesia, J. Phys. Chem. C, 2014, 118, 17787–17800.
259. H. A. Aljama, M. Head-Gordon and A. T. Bell, Nat. Commun., 2022, 13, 2910.
260. K. J. Coopersmith, K. L. Broadwater, S. M. E. Demers, R. Monikandan, M. Torcivia, L. Roy and D. Hunter, J. Phys. Chem. C, 2023, 127, 1598–1606.
261. F. Berger, M. Rybicki and J. Sauer, J. Catal., 2021, 395, 117–128.
262. F. Göltl, P. Muller, P. Uchupalanun, P. Sautet and I. Hermans, Chem. Mater., 2017, 29, 6434–6444.
263. S. Wannakao, C. Warakulwit, K. Kongpatpanich, M. Probst and J. Limtrakul, ACS Catal., 2012, 2, 986–992.
264. K. Cheng, L. I. van der Waals, H. Yoshida, J. Oenema, J. Harmel, Z. Zhang, G. Sunley, J. Zečević and K. P. de Jong, Angew. Chem., Int. Ed., 2020, 59, 3592–3600.
265. J. Shang, G. Li, R. Singh, Q. Gu, K. M. Nairn, T. J. Bastow, N. Medhekar, C. M. Doherty, A. J. Hill, J. Z. Liu and P. A. Webley, J. Am. Chem. Soc., 2012, 134, 19246–19253.
266. M. M. Lozinska, J. P. S. Mowat, P. A. Wright, S. P. Thompson, J. L. Jorda, M. Palomino, S. Valencia and F. Rey, Chem. Mater., 2014, 26, 2052–2061.
267. J. Jia, N. Yan, X. Lian, S. Liu, B. Yue, Y. Chai, G. Wu, J. Xu and L. Li, Angew. Chem., Int. Ed., 2025, 64, e202419091.
268. J. Zhao, S. H. Mousavi, G. Xiao, A. H. Mokarizadeh, T. Moore, K. Chen, Q. Gu, R. Singh, A. Zavabeti, J. Z. Liu, P. A. Webley and G. K. Li, J. Am. Chem. Soc., 2021, 143, 15195–15204.
269. V. M. Georgieva, E. L. Bruce, M. C. Verbraeken, A. R. Scott, W. J. Casteel, Jr., S. Brandani and P. A. Wright, J. Am. Chem. Soc., 2019, 141, 12744–12759.
270. H. J. Choi, E. L. Bruce, K. S. Kencana, J. Hong, P. A. Wright and S. B. Hong, Angew. Chem., Int. Ed., 2023, 135, e202305816.
271. F.-X. Coudert and D. Kohen, Chem. Mater., 2017, 29, 2724–2730.
272. E. Beerdsen, D. Dubbeldam, B. Smit, T. J. H. Vlugt and S. Calero, J. Phys. Chem. B, 2003, 107, 12088–12096.
273. C. Wang, S. Leng, H. Guo, L. Cao and J. Huang, Appl. Surf. Sci., 2019, 478, 319–326.
274. Y. Wang, X. Liu, X. He, F. Yang and X. Zhu, Microporous Mesoporous Mater., 2023, 351, 112491.
275. B. Tang, S. Li, W.-C. Song, E.-C. Yang, X.-J. Zhao and L. Li, ACS Sus. Chem. Eng., 2020, 8, 3796–3808.
276. N. Hoeven, G. Mali, M. Mertens and P. Cool, Microporous Mesoporous Mater., 2019, 288, 109588.
277. Z. Zhao, T. Zhang, Y. Feng, N. Wang and Q. Sun, Chem. Eng. J., 2024, 496, 154194.
278. S. Conrad, P. Wolf, P. Müller, H. Orsted and I. Hermans, ChemCatChem, 2016, 9, 175–182.
279. Y.-Y. Wu and L.-C. Lin, Phys. Chem. Chem. Phys., 2024, 26, 19854–19862.
280. P. Bai, M. Neurock and J. I. Siepmann, J. Phys. Chem. C, 2021, 125, 6090–6098.
281. M. Gao, H. Li, Y. Tian, M. Ye and Z. Liu, ACS Catal., 2023, 13, 11598–11609.
282. G. Ye, L. Zhou, Q. Zhang, J. Meng, J. Weng and X. Zhou, Ind. Eng. Chem. Res., 2023, 62, 9983–9992.
283. C. Wang, Y. Chu, D. Xiong, H. Wang, M. Hu, Q. Wang, J. Xu and F. Deng, Angew. Chem., Int. Ed., 2024, 63, e202313974.
284. K. Chen, J. Damron, C. Pearson, D. Resasco, L. Zhang and J. L. White, ACS Catal., 2014, 4, 3039–3044.
285. D. E. Resasco, S. P. Crossley, B. Wang and J. L. White, Catal. Rev., 2021, 63, 302–362.
286. J. Tang and Y. Yamauchi, Nat. Chem., 2016, 8, 638–639.
287. N. S. Gould, S. Li, H. J. Cho, H. Landfield, S. Caratzoulas, D. Vlachos, P. Bai and B. Xu, Nat. Commun., 2020, 11, 1060.
288. G. Li, T. Yoskamtorn, W. Chen, C. Foo, J. Zheng, C. Tang, S. Day, A. Zheng, M. M. Li and S. C. Tsang, Angew. Chem., Int. Ed., 2022, 61, e202204500.
289. J. Yin, X. Guo, Y. Sun, S. Han and Q. Li, J. Phys. Chem. C, 2021, 125, 310–334.
290. R. Krishna and J. M. Van Baten, Microporous Mesoporous Mater., 2009, 125, 126–134.
291. R. Krishna, Chem. Eng. J., 2001, 84, 207–214.
292. A. Menjoge, S. A. Bradley, D. B. Galloway, J. J. Low, S. Prabhakar and S. Vasenkov, Microporous Mesoporous Mater., 2010, 135, 30–36.
293. U. Navarro, C. A. Trujillo, A. Oviedo and R. Lobo, J. Catal., 2002, 211, 64–74.
294. C. E. Berdugo-Díaz, M. T. Manetsch, Y. S. Yun, J. Lee, J. Luo, X. Chen and D. W. Flaherty, Angew. Chem., Int. Ed., 2023, 62, e202216165.
295. J. M. Crawfold, M. J. Rasmussen, W. W. McNeary, S. Halingstad, S. C. Hayden, N. S. Dutta, S. H. Pang and M. M. Yung, ACS Catal., 2024, 14, 8541–8548.
296. X. Li, H. Han, W. Xu, S. J. Hwang, P. Lu, A. Bhan and M. Tsapatsis, Angew. Chem., Int. Ed., 2022, 61, e202111180.
297. D. Verboekend, G. Vilé and J. Pérez-Ramírez, Adv. Funt. Mater., 2012, 22, 916–928.
298. J. H. Kwak, J. Szanyi and C. H. F. Peden, J. Catal., 2003, 220, 291–298.
299. Y. Kuwahara, J. Aoyama, K. Miyakubo, T. Eguchi, T. Kamegawa, K. Mori and H. Yamashita, J. Catal., 2012, 285, 223–234.
300. C. Zhang, X. Wu, Y. Zhang, L. Wang, Y. Jin, M. Gao, M. Ye, Y. Wei and Z. Liu, ACS Catal., 2025, 15, 1553–1562.
301. P. Kim, J. V. D. Mynsbrugge, H. Aljama, T. M. Lardinois, R. Gounder, M. Head-Gordon and A. T. Bell, Appl. Catal., B, 2022, 304, 120992.
302. J. Pan, J. Lee, M. Li, B. A. Trump and R. F. Lobo, J. Catal., 2022, 413, 812–820.
303. P. Xiao, L. Wang, H. Toyoda, Y. Wwang, K. Nakamura, J. Huang, R. Osuga, M. Nishibori, H. Gies and T. Yokoi, J. Am. Chem. Soc., 2024, 146, 31969–31981.
304. M. Y. Jeon, D. Kim, P. Kumar, P. S. Lee, N. Rangnekar, P. Bai, M. Shete, B. Elyassi, H. S. Lee, K. Narasimharao, S. N. Basahel, S. Al-Thabaiti, W. Xu, H. J. Cho, E. O. Fetisov, R. Thyagarajan, R. F. DeJaco, W. Fan, K. A. Mkhoyan, J. I. Siepmann and M. Tsapatsis, Nature, 2017, 543, 690–694.
305. L. Lin, A. M. Sheveleva, I. D. Silva, C. M. A. Parlett, Z. Tang, Y. Liu, M. Fan, X. Han, J. H. Carter, F. Tuna, E. J. L. McInnes, Y. Cheng, L. L. Daemen, S. Rudić, A. J. Ramirez-Cuesta, C. C. Tang and S. Yang, Nat. Mater., 2020, 19, 86–93.
306. M. Zhou, D. Korelskiy, P. Ye, M. Gehn and J. Hedlund, Angew. Chem., Int. Ed., 2014, 53, 3492–3495.
307. Y. Liu, Z. Liu, J. Zhang, F. S. Xiao, X. Cao and L. Wang, Angew. Chem., Int. Ed., 2023, 62, e202312377.
308. Y. Yuan, W. Huang, S. Hwang, P. Liu and J. G. Chen, Nat. Commun., 2024, 15, 6529.
309. Z. Li, C. Rieg, A.-K. Beurer, M. Benz, J. Bender, C. Schneck, Y. Traa, M. Dyballa and M. Hunger, Adsorption, 2021, 27, 49–68.
310. Y. Wu, S. Zeng, D. Yuan, J. Xing, H. Liu, S. Xu, Y. Wei, Y. Xu and Z. Liu, Angew. Chem., Int. Ed., 2020, 59, 6765–6768.
311. C. Jo, J. Jung, H. S. Shin, J. Kim and R. Rroo, Angew. Chem., Int. Ed., 2013, 52, 10014–10017.
312. Z. Xiong, G. Qi, E. Zhan, Y. Chu, J. Xu, J. Wei, N. Ta, A. Hao, Y. Zhou, F. Deng and W. Shen, Chem, 2023, 9, 76–92.
313. L. Ni, R. Khare, R. Bermejo-Deval, R. Zhao, L. Tao, Y. Liu and J. A. Lercher, J. Am. Chem. Soc., 2022, 144, 12347–12356.
314. L. Qi, S. Das, Y. Zhang, D. Nozik, B. C. Gates and A. T. Bell, J. Am. Chem. Soc., 2023, 145, 2911–2929.
315. P. Sun, C. Liu, H. Wang, Y. Liao, X. Li, Q. Liu, B. F. Sels and C. Wang, Angew. Chem., Int. Ed., 2023, 62, e202215737.
316. S. T. Yang, J. Kim and W. S. Ahn, Microporous Mesoporous Mater., 2010, 135, 90–94.

---