Tina Düren

Responsive porous materials can outperform more rigid analogues in applications requiring precise triggering of molecular uptake/release, switching or gradual change in properties. We have uncovered an unprecedented dynamic response in the diamondoid MOF SHF‐62 , (Me ₂ NH ₂ )[In(BDC‐NHC(O)Me) ₂ ] (BDC = 1,4‐benzenedicarboxylate), by using pressure as a stimulus. SHF‐62 exhibits two distinct framework “breathing” motions involving changes in 1) cross‐section and 2) length of its 1D pores. Our study using synchrotron single‐crystal X‐ray diffraction in sapphire‐capillary ( p < 0.15 GPa) and diamond‐anvil (0.15 < p < 5 GPa) cells reveals that different pressure regimes trigger positive and negative correlation between these two motions, requiring an unprecedented mechanical decoupling. Specifically, the DMF‐solvated framework SHF‐62‐DMF , in DMF as pressure‐transmitting medium, undergoes initial hyperexpansion of pore cross‐section ( p ≤ 0.9 GPa), due to DMF ingress, followed by reversal/reduction at p > 0.9 GPa while pore length contracts for all pressure increases, revealing decoupling of the two framework deformations. By contrast, nonpenetrating medium FC‐70 imposes correlated compression ( p < 1.4 GPa) of pore cross‐section and length, resembling framework activation/desolvation motions but of greater magnitude. Similar behavior occurs for SHF‐62‐CHCl ₃ in CHCl ₃ ( p < 0.14 GPa), suggesting minimal ingress of CHCl ₃ . These findings change our understanding of MOF dynamic responses and provide a platform for future responsive materials development.

Responsive porous materials can outperform more rigid analogues in applications requiring precise triggering of molecular uptake/release, switching or gradual change in properties. We have uncovered an unprecedented dynamic response in the diamondoid MOF SHF‐62 , (Me ₂ NH ₂ )[In(BDC‐NHC(O)Me) ₂ ] (BDC = 1,4‐benzenedicarboxylate), by using pressure as a stimulus. SHF‐62 exhibits two distinct framework “breathing” motions involving changes in 1) cross‐section and 2) length of its 1D pores. Our study using synchrotron single‐crystal X‐ray diffraction in sapphire‐capillary ( p < 0.15 GPa) and diamond‐anvil (0.15 < p < 5 GPa) cells reveals that different pressure regimes trigger positive and negative correlation between these two motions, requiring an unprecedented mechanical decoupling. Specifically, the DMF‐solvated framework SHF‐62‐DMF , in DMF as pressure‐transmitting medium, undergoes initial hyperexpansion of pore cross‐section ( p ≤ 0.9 GPa), due to DMF ingress, followed by reversal/reduction at p > 0.9 GPa while pore length contracts for all pressure increases, revealing decoupling of the two framework deformations. By contrast, nonpenetrating medium FC‐70 imposes correlated compression ( p < 1.4 GPa) of pore cross‐section and length, resembling framework activation/desolvation motions but of greater magnitude. Similar behavior occurs for SHF‐62‐CHCl ₃ in CHCl ₃ ( p < 0.14 GPa), suggesting minimal ingress of CHCl ₃ . These findings change our understanding of MOF dynamic responses and provide a platform for future responsive materials development.

Semiochemicals can be used to manipulate insect behaviour for sustainable pest management strategies, but their high volatility is a major issue for their practical implementation. Inclusion of these molecules within porous materials is a potential solution to this issue, as it can allow for a slower and more controlled release. In this work, we demonstrate that a series of Zr(IV) and Al(III) metal‐organic frameworks (MOFs) with channel‐type pores enable controlled release of three semiochemicals over 100 days by pore size design, with the uptake and rate of release highly dependent on the pore size. Insight from grand canonical Monte Carlo simulations indicates that this is due to weaker MOF‐guest interactions per guest molecule as the pore size increases. These MOFs are all stable post‐release and can be reloaded to show near‐identical re‐release profiles. These results provide valuable insight on the diffusion behaviour of volatile guests in MOFs, and for the further development of porous materials for sustainable agriculture applications.

Visualisation of MOF activation mechanisms, showing gas-like solvent behaviour leading to successful activation (top), and capillary-like solvent behaviour leading to collapse (bottom).

Metal organic framework (MOF) materials are a vast family of nanoporous solids with potential applications ranging from drug delivery to environmental remediation. Application of MOFs in these scenarios is hindered, however, by difficulties in MOF ‘activation’ after initial synthesis – removal of the synthesis solvent from the pores to make the pore space accessible – leading to framework collapse if improperly performed. While experimental studies have correlated collapse to specific solvent properties and conditions, the mechanism of activation-collapse is currently unknown. Developing this understanding would enable researchers to create better activation protocols for MOFs, accelerating discovery and process intensification. To achieve this goal, we simulated solvent removal using grand-canonical Monte Carlo and free energy perturbation methods. By framing activation as a fluid desorption problem, we investigated activation processes in the IRMOF family of MOFs for different solvents. We identified two pathways for solvent activation – the solvent either desorbs uniformly from each individual pore or forms coexisting phases during desorption. These mesophases in turn lead to large capillary stresses within the framework, corroborating experimental hypotheses for the cause of activation-collapse. Finally, we found that the activation energy of solvent removal increased with pore size and connectivity due to the increased stability of solvent mesophases, matching experimental findings. Using these simulations, it is possible to screen MOF activation procedures, enabling rapid identification of ideal solvents and conditions and thus enabling faster development of MOFs for practical applications.

The aggregation of polycyclic aromatic compound (PAC) molecules is sensitive to the solvent they are dissolved or suspended in. By using both dynamic light scattering and diffusion-ordered nuclear magnetic resonance spectroscopy, in combination with molecular dynamics simulations, the effect of chemical structure on the aggregation of PACs in both aromatic and alkane solvents were systematically investigated. A suite of triphenylene-based PACs offers a robust platform to understand the driving forces of aggregation mechanism across both nanometer and micrometer scales. Both the configuration, either parallel or otherwise, and the arrangement, whether compact or loose, of molecules in their aggregates are determined by a fine balance of different interactions such as those between the polar groups, π–π interactions between the aromatic cores, steric hindrance induced by the side chains, and the degree of solvation. These results suggest that molecular architecture is the major factor in determining how the model compounds aggregate. The shift from aromatic to aliphatic solvent only slightly increases the likelihood of aggregation for the model compounds studied while subtle differences in molecular architecture can have a significant impact on the aggregation characteristics.

Cancer remains hard to treat, partially due to the non-specificity of chemotherapeutics. Metal–organic frameworks (MOFs) are promising carriers for targeted chemotherapy, yet, to date, there have been few detailed studies to systematically enhance drug loading while maintaining controlled release. In this work, we investigate which molecular simulation methods best capture the experimental uptake and release of cisplatin from UiO-66 and UiO-66(NH₂). We then screen a series of biocompatible, pH-sensitive zeolitic imidazolate frameworks (ZIFs) for their ability to retain cisplatin in healthy parts of the patient and release it in the vicinity of a tumor. Pure-component GCMC simulations show that the maximum cisplatin loading depends on the pore volume. To achieve this maximum loading in the presence of water, either the pore size needs to be large enough to occupy both cisplatin and its solvation shell or the MOF–cisplatin interaction must be more favorable than the cisplatin–shell interaction. Both solvated and non-solvated simulations show that cisplatin release rates can be controlled by either decreasing the pore limiting diameters or by manipulating framework–cisplatin interaction energies to create strong, dispersed adsorption sites. The latter method is preferable if cisplatin loading is performed from solution into a pre-synthesized framework as weak interaction energies and small pore window diameters will hinder cisplatin uptake. Here, ZIF-82 is most promising. If it is possible to load cisplatin during crystallization, ZIF-11 would outcompete the other MOFs screened as cisplatin cannot pass through its pore windows; therefore, release rates would be purely driven by the pH triggered framework degradation.

MOFs and COFs are porous materials with a large variety of applications including gasstorage and separation. Synthesised in a modular fashion from distinct building blocks, anear in?nite number of structures can be constructed and the properties of the material canbe tailored for a speci?c application. While this modularity is a very attractive feature it alsoposes a challenge. Attempting to identify the best performing material(s) for a given appli-cation is experimentally intractable. Current research e?orts combine molecular simulationsand machine learning techniques to evaluate the simulated performance of hundreds of thou-sands of materials to identify top performing MOFs and COFs for a given application. Theseapproaches typically rely on moderated brute-force screening which is still resource-intensiveas typically between 70 - 100 % of the hundreds of thousands of materials must be simulatedto create a training set for the machine learning models used, restricting screening to rela-tively simple molecules. In this work we demonstrate our novel Bayesian mining approachto materials screening which allows 62 - 92 % of the top 100 porous materials for a range ofapplications to be readily identi?ed from large materials databases after only assessing lessthan one percent of all materials. This is a stark contrast to the 0 - 1 % achieved by conven-tional brute-force screening where porous materials are just chosen at random during a highthroughput screening. Through this accelerated virtual screening process, the identi?cation ofhigh performing materials can be used to more rapidly inform experimental e?orts and hencelead to an acceleration of the entire research and development pipeline of porous materials.

We present an open source Python 3 library aimed at practitioners of molecular simulation, especially Monte Carlo simulation. The aims of the library are to facilitate the generation of simulation data for a wide range of problems; and to support data analysis methods which enable one to make the most of previously generated data. The library contains a framework for automating the task of measuring target physical properties (e.g. density) over a range of thermodynamic parameters (e.g. temperature) calculated using a molecular simulation program, in particular the Monte Carlo program DL_MONTE. The library also supports analysis methods including block averaging, equilibration detection and histogram reweighting. Here we describe the library and provide examples to demonstrate its key functionality: we use the library to automatically calculate isotherms to a specified precision; and to calculate the surface tension and liquid-vapour coexistence properties of methane.

Flexible MOF nanoparticles, i.e. MOF nanoparticles that change their structure upon external stimuli such as guest uptake, are promising for numerous applications including advanced gas adsorption, drug delivery and sensory devices. However, the properties of MOFs are typically characterised based on the bulk material with no consideration of how the particle size and external surface influences their performance. This combined computational and experimental contribution investigates the influence of the particle size and surface functionalisation on the flexibility of DUT-8(Ni) (Ni₂(2,6-ndc)₂ dabco, ndc = naphthalene dicarboxylate, dabco = 1,4-diazabicyclo[2.2.2]octane, DUT = Dresden University of Technology). DUT-8 nanoparticles remain rigid in their open pore form while microparticles, synthesised under slightly different conditions, undergo gate opening upon nitrogen adsorption suggesting that the particle size has an important role to play in the flexibility of DUT-8. While the adsorption environment at the surface capped with modulators smaller than the 2,6-ndc ligand is very different compared to the bulk of the crystal with considerably weaker guest–framework interaction, simulations reveal that the nanoparticles should close. We conclude that the size of the nanoparticles is not the major contributor for keeping DUT-8 nanoparticles open but that it is more likely that defects or nucleation barriers dominate. Moreover, our work reveals for the first time that functionalising the external surface of nanoparticles with different modulators or capping groups offers the opportunity to manipulate the gate opening/closing pressure. This principle is generally applicable and could be exploited to tune the gate opening/closing pressure for the application of interest.