Metal-organic frameworks (MOFs), crystalline porous materials composed of metal nodes and organic ligands, show promise in gas storage, separation, and catalysis. MOFs with large cavities are particularly interesting due to their open channels facilitating rapid guest molecule transport and macromolecule encapsulation. However, unmodified MOFs often lack selectivity towards substrates with similar sizes, leading to undesired side products and reduced separation efficiency. Improving selectivity while maintaining porosity is crucial for practical applications. Common strategies involve bottom-up synthesis or post-synthetic modification to create substrate-specific voids. While surface capping agents like polymers, phospholipids, and biomacromolecules have been used, they often reduce porosity. Encapsulation of cages within MOF cavities can improve stability, but non-covalent encapsulation can hinder porosity. This research aims to overcome these limitations by developing a "Cage-on-MOF" strategy to coordinatively functionalize MOF surfaces with porous cages, maintaining porosity while enhancing selectivity.

Existing methods for enhancing MOF selectivity often involve either bottom-up synthesis or post-synthetic modification. Bottom-up approaches focus on creating substrate-specific accessible voids within the MOF framework itself. Post-synthetic modification utilizes various surface capping agents, including polymers, phospholipids, phenylsilanes, and biomacromolecules, to modify the external surface of MOFs. However, non-porous capping agents can block the pores, reducing the effectiveness of the MOF. Some studies have investigated encapsulating small cages inside MOF cavities to enhance properties; however, these non-covalent interactions may compromise porosity and stability. This work distinguishes itself by focusing on a covalent, surface-functionalization approach using porous coordination cages to achieve selectivity enhancement without sacrificing porosity.

The researchers designed and synthesized two porous coordination cages (PCCs), PCC-4 and PCC-5, with opposite charges and distinct secondary coordination sites (-SO3 and -NH2, respectively). Single-crystal X-ray diffraction, 1H NMR, and ESI-MS confirmed their structures. PCN-222 and MIL-101 were selected as model MOFs due to their large surface area and accessible voids. PCCs were coordinatively immobilized onto the MOFs' external surfaces by mixing excess PCCs with MOF nanoparticles in DMF or MeCN at 50 °C for 12 h. The resulting MOF@PCC nanocomposites were characterized using various techniques including TEM, HAADF-STEM, EDX elemental mapping, PXRD, zeta potential measurements, FT-IR, XPS, and N2 adsorption isotherms. The selective adsorption of dyes (methylene blue and eosin Y) was evaluated in aqueous solutions. A sequential reaction involving benzaldehyde dimethyl acetal hydrolysis followed by Knoevenagel condensation with malononitrile was used to assess catalytic selectivity. The influence of cavity blockage by inhibitors on catalytic selectivity was also studied. Molecular dynamics simulations were performed to investigate the location of the cages within the MOF structure. DFT calculations were conducted to understand the adsorption energy of intermediates on PCCs and understand the reaction mechanism. Finally, the reversibility of the surface coordination was demonstrated via acid treatment to remove PCCs, followed by re-functionalization with different PCCs. The process was repeated for MIL-101 MOFs to demonstrate the generality of the strategy. Detailed synthesis procedures for PCC-4, PCC-5, MOF@PCC-4, and MOF@PCC-5, as well as dye adsorption and catalytic reaction protocols are provided in the supplementary materials.

The "Cage-on-MOF" strategy successfully functionalized PCN-222 and MIL-101 MOFs with PCCs without significant porosity loss. Characterization techniques confirmed the uniform surface coating of PCCs on the MOF nanoparticles. The surface charge of the MOFs was effectively tuned by the choice of PCC. In dye adsorption studies, PCN-222@PCC-4 showed enhanced adsorption of positively charged methylene blue, while PCN-222@PCC-5 preferentially adsorbed negatively charged eosin Y. In the model sequential catalytic reaction, PCN-222@PCC-4 favored the formation of the final condensation product, while PCN-222@PCC-5 selectively produced the intermediate aldehyde. Inhibitor studies confirmed that the PCC cavities played a crucial role in the observed catalytic selectivity. DFT calculations showed that the different adsorption energies of the intermediates on PCC-4 and PCC-5 explained the selectivity. The surface modification proved reversible, allowing for repeated tuning of the MOF's catalytic properties by changing PCCs through mild acid treatment. These findings were also successfully replicated for MIL-101 MOF, demonstrating the generality of the Cage-on-MOF strategy. The surface density of the cages on both MOFs was also investigated and found to be approximately 10% of the surface-exposed metal clusters.

The "Cage-on-MOF" strategy successfully addresses the challenge of enhancing MOF selectivity while preserving porosity. The covalent attachment of PCCs avoids pore blockage, unlike methods using non-porous capping agents. The tunable surface charge of the MOF@PCC nanocomposites allows for tailored adsorption and catalytic selectivity. The reversibility of the surface modification offers a unique advantage, enabling the reuse of the MOF and dynamic control over its catalytic properties. The findings highlight the potential of using porous molecular cages as versatile surface-functionalization agents for creating advanced materials with enhanced properties. This approach is potentially broadly applicable to various porous materials and catalytic reactions.

This study successfully demonstrates a "Cage-on-MOF" strategy for enhancing the selectivity of MOFs in both adsorption and catalysis, while preserving their intrinsic porosity. The use of coordinatively attached porous coordination cages allows for the reversible tuning of surface properties and catalytic activity. This approach represents a significant advancement in the field of porous material functionalization and opens new avenues for the design of advanced materials with tailored properties. Future research could explore the application of this strategy to a wider range of MOFs and catalytic reactions, as well as the investigation of different types of porous coordination cages with diverse functionalities.

The study primarily focuses on two model MOFs (PCN-222 and MIL-101) and two types of PCCs. Further investigation is needed to determine the broader applicability of the "Cage-on-MOF" strategy to other MOF structures and diverse PCCs. While the reversibility of the surface modification is shown, the long-term stability and reusability of the MOF@PCC composites under various reaction conditions warrant further exploration. The molecular dynamics simulations simplified the MOF structure and might not fully capture the complexity of the interactions. The study could benefit from more in-depth mechanistic studies using various spectroscopic techniques to confirm the detailed reaction pathways.