New Insights into Phonon and Photon Heat Carrier Dynamics

This talk consists of two parts. In the first part, I will talk about photon heat carrier transport in solid materials at extreme temperatures. We present a comprehensive study of the photon thermal conductivity (κrad) in 13 III-IV semiconductor materials using a modified Rosseland model that incorporates both ballistic and diffusive photon transport. Leveraging first-principles calculations and the Lorentz oscillator model, we developed a unified framework for calculating κrad across a wide range of optical thicknesses and surface emissivity. Our findings reveal a strong correlation between the optical phonon frequency and photon absorption, higher optical phonon frequencies correlate to absorption peaks in the lower infrared spectrum. Also, a robust inverse relationship between optical phonon linewidth difference and photon MFP was identified, consistent across all materials and temperatures. We observed that most of the examined materials exhibit photon mean free paths (MFPs) in the millimeter range within the near-infrared spectrum, with notable exceptions like BAs and GaSb, which demonstrate significantly longer MFPs (in meters). At room temperature, the κrad is considerably small compared to phonon thermal conductivity and can be safely ignored. However, its contribution increases significantly with temperature and becomes significant (around 20% for most materials) at higher temperatures. The study also highlights the transition between ballistic and diffusive photon transport across different optical regimes, with ballistic contributions dominating in optically thin materials and diffusive mechanisms prevailing in optically thick materials. In all intermediate cases, both ballistic and diffusive photon contribute based on emitting surfaces and media’s optical properties.
In the second part, I will talk about phonon transport across successive interfaces. At the nanoscale, thermal boundary conductance (TBC) and thermal conductivity are not intrinsic properties of interfaces or materials but depend on the nearby environment. However, most studies focus on single interfaces or superlattices, and the thermal transport across heterostructures formed by multiple different materials is still mysterious. In this study, we demonstrate how much the TBC of an interface is affected by the existence of a second interface, as well as how much the thermal conductivity of a material is affected by the nearby materials. Using Si and Ge modeled by classical molecular dynamics simulations, the following phenomena are demonstrated. (1) The existence of a nearby interface can significantly change the TBC of the original interface. For example, by adding an interface after Si/Ge, the TBC can be increased from 400 to 700MWm−2 K−1. This is because the nearby interface serves as a filter of phonon modes, which selectively allows particular modes to pass through and affect the TBC of the original interfaces. This impact will disappear at the diffusive limit when the distance between interfaces is much longer than the phonon mean free path so that phonon modes recover equilibrium statistics before arriving at the second interface. (2) The thermal conductivity of a material can be significantly changed by the existence of neighboring materials. For example, the thermal conductivity of standalone 30 nm thick Si can be increased from 50 to 280 Wm−1 K−1, a more than fourfold increase, beating the bulk thermal conductivity of Si, after being sandwiched between two Ge slabs. This is because the Ge slabs on the two sides serve as filters that only allow low-frequency phonons to transport heat in Si; these phonons carry more heat than optical phonons. This work opens up an area of successive interface thermal transport and is expected to be important for nanoscale thermal characterization and thermal management of semiconductor devices