Numerical evaluation of the effect of the twist angle on phonon hydrodynamics in twisted bilayer graphene

The influence of twist angles on the thermal transport properties of bilayer graphene has attracted considerable attention within the scientific community. However, the relationship between twist angle and phonon hydrodynamics remains insufficiently explored. Additionally, the limitations of empirical potential fields have often restricted the accuracy of previous calculations. In this study, we systematically examine the effect of twist angle on phonon hydrodynamics in twisted bilayer graphene using a machine learned neuroevolution potential combined with the Boltzmann transport equation. Our results reveal a distinctive ``$\mathsf{V}$-shaped'' trend in thermal conductivity, driven by variations in N-type and U-type phonon scattering rates, underscoring the significant influence of twist angle on phonon hydrodynamics. Particularly, the flexural mode plays a critical role, with its contribution to thermal conductivity decreasing as the twist angle changes. This analysis highlights the importance of machine learning potentials in accurately capturing interlayer interactions, providing a more precise understanding of thermal transport in two-dimensional materials and offering alternative possibilities for material manipulation.