Understanding the influence of secondary building units on the thermal conductivity of metal–organic frameworks via high-throughput computational screening

The thermal conductivity of metal–organic frameworks (MOFs) has garnered increasing interest due to their potential applications in energy-related fields. However, due to the diversity of building units, understanding the relationship between MOF structures and their thermal conductivity remains an imperative challenge. In this study, we predicted the thermal conductivity (κ) of MOFs using equilibrium molecular dynamics (EMD) simulations and investigated the contribution of structure properties to their thermal conductivity. It is revealed that the arrangement of secondary building units (SBUs) with a closer distance of metal atoms, a larger proportion of metal elements, and transition metal elements (Fe, Mn, and Co) leads to high thermal conductivity. To generally quantify the influence of such factors on thermal conductivity, the pathway factors with SBU influence (Pm) were proposed and can be used to efficiently classify structures into high, medium, and low thermal conductivity types. It was found that Pm indicates that MOFs with met topology tend to have high thermal conductivity, while rna and pcu topologies naturally tend to possess medium and low thermal conductivity. Moreover, it was also suggested that taking Pm as a descriptor in the machine learning algorithms can significantly improve the prediction accuracy for thermal conductivity. This study offers molecular insight into the impact of various SBUs on thermal conductivity in framework-based nanomaterials, which may guide the rational design of nanoporous materials with desirable thermal conductivity.