Thermal Transport in Metal-Organic Frameworks—A Molecular Dynamics Approach

Thermal transport properties in metal-organic frameworks (MOFs) are critical for their applications in areas such as gas storage, separation, and catalysis. In this work, we adopt a combination of classical molecular dynamics (MD) simulations, lattice dynamics calculations, and ab initio simulations to investigate thermal transport in MOFs. We use Density Functional Theory (DFT) and lattice dynamics calculations to characterize the material from first principles. Then, we adopt the recently developed Chebyshev Interaction Model for Efficient Simulation (ChIMES)[1,2], a linearly parametrized machine-learned force field that, leveraging the Chebyshev polynomials of the first kind, allows us to obtain a complete description of the total energy and the forces in the system starting from ab initio calculations. This approach gives the possibility of obtaining an interaction model for model potential MD simulations using a relatively small set of data compared to other machine learning-based approaches. We perform equilibrium and non-equilibrium MD simulations on extended structures, providing a comprehensive analysis of thermal conductivity in MOFs, and highlighting the influence of structural characteristics, defect presence, and material composition. Our findings demonstrate the potential of ChIMES in advancing the simulation capabilities for complex materials like MOFs and offer valuable insights for the design of MOFs with tailored thermal properties.