Understanding Dynamics of Nanocluster-Organic Frameworks and Gas Diffusion from Machine Learning Potential-Based...

Area of Science:

• Materials Science
• Computational Chemistry
• Chemical Engineering

Background:

• Nanocluster-organic frameworks (NOFs) are advanced photofunctional materials with applications in sensing and optoelectronics.
• Their luminescence properties are sensitive to gases like oxygen (O2) and volatile organic compounds.
• Understanding the atomistic structure and small molecule diffusion mechanisms within NOFs is crucial but remains challenging.

Purpose of the Study:

• To develop accurate machine learning potentials for modeling a specific NOF, [Ag12(StBu)8(CF3COO)4(bpy)4]n.
• To investigate the structural evolution and dynamics of this NOF using molecular dynamics simulations.
• To elucidate the diffusion mechanism and free energy landscape of O2 gas within the NOF pores.

Main Methods:

• Development of machine learning potentials for accurate atomistic modeling of NOFs.
• Molecular dynamics simulations to study NOF structure and dynamics.
• On-the-fly probability-enhanced sampling simulations for O2 diffusion and free energy surface construction.

Main Results:

• Accurate modeling of the flexible NOF structure and dynamics was achieved.
• O2 gas was observed to predominantly localize around the bipyridine linker within the NOF.
• A pore-to-pore diffusion barrier of approximately 16-18 kJ/mol for O2 was determined, indicating feasible room-temperature diffusion.

Conclusions:

• This work presents the first integrated approach for modeling a flexible NOF and its gas diffusion with DFT-level accuracy.
• The findings provide insights into the gas interaction and transport mechanisms in NOFs.
• The developed methodology can be applied to study other porous materials and guest molecule interactions.