Sub-Micrometer Phonon Mean Free Paths in Metal-Organic Frameworks Revealed by Machine Learning Molecular Dynamics...

Area of Science:

• Materials Science
• Computational Chemistry
• Condensed Matter Physics

Background:

• Metal-organic frameworks (MOFs) possess high porosity and tunable structures, making them promising for various applications.
• Understanding thermal transport properties is crucial for MOF applications.
• Accurate interatomic potentials are challenging to develop for MOF molecular dynamics (MD) simulations.

Purpose of the Study:

• Develop accurate and efficient machine-learned potentials for MOFs.
• Investigate thermal transport properties of MOF-5, HKUST-1, and ZIF-8 using MD simulations.
• Characterize phonon behavior and its influence on thermal conductivity in MOFs.

Main Methods:

• Utilized the neuroevolution potential approach for developing machine-learned potentials.
• Employed the GPUMD package for constructing potentials and performing MD simulations.
• Calculated lattice thermal conductivity and phonon mean free paths (MFPs) for selected MOFs.

Main Results:

• Achieved accurate yet efficient machine-learned potentials for MOF-5, HKUST-1, and ZIF-8.
• Predicted low lattice thermal conductivity (< 1 W/(m K)) at room temperature.
• Observed sub-micrometer phonon MFPs in the low-frequency region, impacting apparent thermal conductivity in nanocrystalline MOFs.

Conclusions:

• Sub-micrometer phonon MFPs in MOFs significantly influence their thermal transport characteristics.
• The findings suggest a moderate temperature dependence of thermal conductivity in MOFs.
• This study advances the understanding of thermal transport mechanisms in MOFs, crucial for their technological applications.