Elucidating CO₂ dynamics in high-entropy MOF-74 via machine learning interatomic potentials

Area of Science:

• Materials Science
• Chemical Engineering
• Computational Chemistry

Background:

• High entropy variants of MOF-74 exhibit enhanced CO2 adsorption.
• The intrinsic contribution of homogenous metal mixing to CO2 transport in MOFs is not well understood.

Purpose of the Study:

• To investigate CO2 diffusion in single-metal and mixed-metal MOF-74 frameworks.
• To determine if homogenous metal mixing alone enhances CO2 transport properties.

Main Methods:

• Development of a transferable machine-learned interatomic potential.
• Large-scale molecular dynamics simulations of CO2 diffusion in MOF-74 with various metal compositions (Mg, Co, Cu, Ni, Zn).
• Analysis of temperature-dependent diffusion from 300 K to 500 K.

Main Results:

• Single-metal MOF-74 axial diffusivities at 300 K range from 0.322 to 1.211 × 10^-8 m^2 s^-1 (Mg slowest, Cu fastest).
• A near-equimolar high-entropy MOF-74 framework showed CO2 diffusivity (Dz = 0.899 × 10^-8 m^2 s^-1) predictable by a composition-weighted rule-of-mixtures.
• This rule-of-mixtures accurately predicted diffusion for other mixed-metal compositions and across the simulated temperature range.

Conclusions:

• Homogenous metal mixing in MOF-74 does not intrinsically enhance CO2 transport beyond the behavior of single-metal parents.
• A simple, predictive rule-of-mixtures accurately describes CO2 diffusivity in homogeneous mixed-metal MOF-74.
• Establishes a baseline for transport in complex MOFs and aids in evaluating experimental factors influencing diffusion.