Thinking Beyond The Phonon Gas Model
September 18, 2014 - 10:00am
Understanding the thermal conductivity of bulk crystalline solids is essentially a solved problem and it is well described by the phonon gas model (PGM). The PGM treats individual phonons (e.g., quanta of lattice vibration energy) as gas molecules that carry energy at a certain speed for some averaged distance, termed the mean free path (MFP). This model does an excellent job at explaining the thermal conductivity of crystalline solids and due to advancements in modeling over the last decade, one can now calculate phonon energies, velocities and MFPs fully from first principles. This now allows one to predict the thermal conductivity of virtually any crystalline material with excellent agreement with experiments at virtually all temperatures of technological interest. By employing Monte Carlo methods or the Boltzmann Transport Equation, one can also accurately predict the thermal conductivity of micro and nanostructures due to quantum or classical size effects. As a result of the great success of this model, it has prevailed as the primary physical picture used to understand and interpret all phonon transport related phenomena. However, there are a number of technologically important material classes and molecules that are not well described by the PGM. This talk will discuss several instances where the PGM is inconsistent with the atomic level behaviors observed in molecular dynamics simulations. The talk will also cover several new theoretical modeling developments that offer a different perspective on phonon-phonon interactions and allows for direct calculation of phonon contributions to thermal conductivity and interface conductance.