📄 chemistry

Topology-Controlled Solvent Eviction and Co-Occupation in Liquid-Filled Zeolite Micropores

This study demonstrates that zeolite pore topology acts as a molecular switch governing solvent behavior in liquid-phase catalysis, where narrow MFI channels expel solvent molecules to favor reactant adsorption while larger FAU supercages permit solvent-reactant co-occupation, a mechanism elucidated through combined NMR, thermodynamic, and machine learning-enhanced simulation analyses.

Original authors: JianZhi Hu, Wenda Hu, Princy Jarngal, Nicholas Jaegers, Sungmin Kim, Benjamin Jackson, Iljun Chung, Feng Chen, Oliver Gutiérrez, Ruixue Zhao, Yue Liu, Qiang Liu, Donald Camaioni, Cong Zhou, Huamin Wan

Published 2026-07-07

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: JianZhi Hu, Wenda Hu, Princy Jarngal, Nicholas Jaegers, Sungmin Kim, Benjamin Jackson, Iljun Chung, Feng Chen, Oliver Gutiérrez, Ruixue Zhao, Yue Liu, Qiang Liu, Donald Camaioni, Cong Zhou, Huamin Wang, Mal-Soon Lee, Yong Wang, GiovanniMaria Piccini, Johannes Lercher

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a zeolite catalyst as a microscopic, sponge-like building made of tiny rooms and hallways. Inside these rooms, chemical reactions happen. But these rooms aren't empty; they are often filled with liquid solvent (like a fluid filling the sponge), and the reactants (the chemicals we want to change) have to squeeze in to find the "active sites" (the workers) where the magic happens.

The big question this paper answers is: When a reactant tries to enter these liquid-filled rooms, does it kick the solvent out, or do they squeeze in together?

The researchers discovered that the answer depends entirely on the shape and size of the rooms (the "topology") inside the zeolite. They used two different types of zeolites as their test subjects:

MFI: Think of this as a building with narrow, twisting hallways.
FAU: Think of this as a building with huge, open ballrooms (supercages).

Here is what they found, using simple analogies:

1. The "Forced Eviction" in Narrow Hallways (MFI)

In the MFI zeolite, the hallways are so narrow that there is barely enough room for one person to walk through, let alone two.

The Scenario: Imagine a crowded, narrow subway car (the solvent, called decalin) is already inside. A new passenger (the reactant, cyclohexanol) wants to get to the driver's seat (the active site).
The Result: Because the hallway is so tight, the new passenger cannot squeeze in next to the solvent. To get to the driver, the new passenger must push the solvent out.
The Evidence: The researchers used a special "motion camera" (NMR) to watch the solvent molecules. When the reactant arrived, the solvent molecules suddenly started moving freely outside the zeolite, as if they had been kicked out of the building. The reactant took the spot, and the solvent was evicted.
The Cost: This eviction is energetically "expensive" in terms of order (entropy). It's like forcing a chaotic crowd to stand in a single file line; it takes a lot of energy to make that happen, but once it's done, the reactant holds the spot very tightly.

2. The "Co-Occupancy" in Big Ballrooms (FAU)

In the FAU zeolite, the rooms are massive ballrooms.

The Scenario: Imagine a huge, empty concert hall (the solvent) is already inside. A new VIP guest (the reactant) wants to get to the stage (the active site).
The Result: Because the room is so big, the VIP doesn't need to kick anyone out. They can walk onto the stage, and the solvent molecules can just hang out in the rest of the ballroom. They co-exist in the same space.
The Evidence: When the researchers added the reactant to this system, the solvent molecules didn't get kicked out. They stayed inside the zeolite, just moving around a bit more. The "motion camera" showed that the solvent was still confined inside, sharing the space with the reactant.
The Benefit: This is a more relaxed situation. There isn't a huge penalty for order because the room is big enough for everyone to move around freely.

The "Molecular Switch"

The paper concludes that the shape of the zeolite acts like a switch:

Narrow Shape (MFI): Switches to "Eviction Mode." The reactant forces the solvent out to get to the active site.
Wide Shape (FAU): Switches to "Sharing Mode." The reactant and solvent share the space comfortably.

How They Knew This

The scientists didn't just guess; they used a toolkit of high-tech methods:

NMR (The Motion Camera): They looked at how fast the solvent molecules were wiggling. If they were stuck inside, the signal was blurry. If they were kicked out and moving freely, the signal became sharp.
Heat Measurements: They measured how much heat was released when the reactant entered. In the narrow hallways, a lot of heat was released (strong bonding), but it came with a high "order penalty." In the big ballrooms, the heat release was weaker, but the "order penalty" was low.
Computer Simulations (The Virtual Lab): They used supercomputers and artificial intelligence to build a 3D movie of the molecules. These simulations confirmed that in the narrow hallways, the math says the reactant must push the solvent out to win. In the big ballrooms, the math says they can both stay.

The Bottom Line

This study shows that you can't just look at what chemicals are reacting; you have to look at where they are reacting. The architecture of the tiny pores decides whether the solvent is a rival that gets kicked out or a roommate that stays put. This helps scientists design better catalysts by choosing the right "room shape" for the job.

Technical Summary: Topology-Controlled Solvent Eviction and Co-Occupation in Liquid-Filled Zeolite Micropores

Problem Statement
In liquid-phase catalysis within nanoporous materials, solvent molecules are not passive media; they occupy pore volume, compete with reactants for active sites, and reshape adsorption thermodynamics. A critical unresolved question in zeolite catalysis is determining the structural rules that dictate whether a solvent is expelled from a liquid-filled micropore upon reactant adsorption or remains co-confined with the reactant near the active site. While prior studies have noted that reaction rates and activation parameters vary significantly with solvent identity and zeolite topology (e.g., cyclohexanol dehydration in decalin), the molecular origin of these solvent effects—specifically the redistribution of solvent between confined and mobile states—remains unclear. Conventional adsorption measurements report net uptake or heat release but fail to directly resolve how solvent molecules redistribute within the micropore network under liquid-phase conditions.

Methodology
This study employs a multi-modal approach combining experimental characterization and advanced computational simulations to investigate solvent-reactant competition in two distinct zeolite topologies: MFI (narrow intersecting channels) and FAU (large supercages). The model system consists of cyclohexanol (reactant) and deuterated decalin (solvent).

In Situ 2H Magic-Angle-Spinning (MAS) NMR: Motion-sensitive 2H NMR was used to distinguish between confined (restricted motion) and mobile (bulk-like) decalin populations. Line broadening indicates confinement within micropores, while line narrowing indicates mobility. Experiments tracked decalin mobility in MFI and FAU before and after cyclohexanol addition, and at varying temperatures.
Infrared (IR) Spectroscopy: Used to confirm interactions between decalin and Brønsted acid sites (BAS) by monitoring shifts in hydroxyl group vibrations.
Liquid-Phase Adsorption Calorimetry and Isotherms: Adsorption isotherms were measured to determine equilibrium capacities and constants. Calorimetry provided standard adsorption enthalpies ( $\Delta H^\circ_{ads}$ ), allowing the calculation of standard adsorption entropies ( $\Delta S^\circ_{ads}$ ).
Machine-Learning-Enhanced Simulations: Graph neural network machine-learning potentials (MLP), trained on ab initio data using the MACE architecture, were employed to perform enhanced-sampling molecular dynamics (well-tempered metadynamics). These simulations mapped the free-energy landscapes of competitive adsorption between cyclohexanol and decalin isomers at BAS in both MFI and FAU frameworks.

Key Results

Topology-Dependent Solvent Behavior:
- MFI (Narrow Channels): In MFI, decalin is strongly immobilized within the micropores. Upon cyclohexanol adsorption, the 2H NMR linewidth of decalin narrows dramatically (from ~27.6 Hz to ~7.7 Hz), indicating that cyclohexanol effectively displaces decalin from the micropores. This establishes a "forced-exchange" regime where reactant adsorption is directly coupled to solvent eviction.
- FAU (Large Supercages): In FAU, decalin remains significantly confined even after cyclohexanol addition. The NMR linewidth narrows only moderately (from ~59.1 Hz to ~23 Hz), indicating that a substantial fraction of decalin remains co-occupied with cyclohexanol within the supercages. This defines a "co-occupation" regime.
Thermodynamic Signatures:
- MFI: Cyclohexanol adsorption is enthalpy-dominated ( $\Delta H^\circ_{ads} \approx -43$ kJ mol $^{-1}$ ) with a large entropy penalty ( $\Delta S^\circ_{ads} \approx -93$ J mol $^{-1}$ K $^{-1}$ ). This reflects strong, specific binding of cyclohexanol dimers to BAS under severe geometric constraints.
- FAU: Adsorption exhibits weaker enthalpic stabilization ( $\Delta H^\circ_{ads} \approx -29$ kJ mol $^{-1}$ ) and a smaller entropy loss ( $\Delta S^\circ_{ads} \approx -60$ J mol $^{-1}$ K $^{-1}$ ). The thermodynamics suggest that cyclohexanol clusters form via self-association, and the larger pore volume allows for greater configurational flexibility, reducing the entropic cost.
Free-Energy Landscapes:
- MFI: Metadynamics simulations reveal a free-energy preference of approximately 25 kJ mol $^{-1}$ for cyclohexanol over decalin at BAS. The narrow channels impose steric constraints that prevent co-adsorption; the system must undergo a forced exchange where decalin is displaced to accommodate cyclohexanol.
- FAU: The free-energy landscape is broader and shallower. The large supercages relax steric constraints, allowing cyclohexanol and decalin to co-adsorb in multiple configurations near the BAS without a strict exchange mechanism. The enthalpic stabilization is weaker, but the entropic contribution is more favorable due to increased molecular freedom.

Significance and Claims
The authors claim that this work establishes pore topology as a molecular switch that controls solvent participation in liquid-filled zeolite micropores. The study provides a direct experimental and computational demonstration that:

In narrow channels (MFI), reactant adsorption induces efficient solvent eviction due to steric constraints and strong enthalpic driving forces.
In large cages (FAU), the topology permits solvent-reactant co-occupation, resulting in a less constrained competitive free-energy landscape.

The paper posits that these findings offer a mechanistic criterion for designing catalytic microenvironments. By selecting specific zeolite architectures, chemists can choose to either exclude solvent from active-site environments (favoring forced exchange) or retain solvent-reactant co-confinement. The work resolves the molecular origin of solvent effects in liquid-phase zeolite chemistry by linking macroscopic thermodynamic observations to microscopic solvent redistribution dynamics.