Understanding the thermal transport of various metals is crucial for many energy-transfer applications. However, due to the complex transport mechanisms varying among different metals, current research on metallic thermal transport has been focusing on case studies of specific types of metallic materials. Ageneral understanding of the transport mechanisms across a broad spectrum of metallic materials is still lacking. In this work, we perform first-principles calculations to determine the thermal conductivity of 40 representative metallic materials, within a range of 8-456 W/mK. Our predicted values of electrical and thermal conductivity are in good agreement with available experimental results. Based on the data of separated electron and phonon thermal conductivity, we employ a statistical approach to examine nine factors derived from previous understandings and identify the critical factors determining these properties. For electrons, although a high electron density of states around the Fermi level implies more conductive electrons, we find it counterintuitively correlates with low electron thermal conductivity. This is attributed to the enlarged electron-phonon scattering channels induced by substantial electrons around the Fermi level. Regarding phonons, we demonstrate that among all the studied factors, Debye temperature plays the most significant role in determining the phonon thermal conductivity, despite the phonon-electron scattering being non negligible in- sometransition metals. Correlation analysis suggeststhat Debye temperature has the highest positive correlation coefficient with phonon thermal conductivity, as it corresponds to a large phonon group velocity. Additionally, Young's modulus is found to be closely correlated with high phonon thermal conductivity and contribution. Our findings of simple factors that closely correlate with the electron and phonon thermal conductivity provide a general understanding of various metallic materials. They may facilitate the discovery of novel materials with extremely high or low thermal conductivity, or be used as descriptors in machine learning to accurately predict the thermal conductivity of metals in the future.
.