Abstract
The continuing miniaturization of electronic devices drives the advancement of modern technology. As the device dimension shrinks to the nanoscale, thermal management becomes the bottleneck for further performance improvements of these devices. The performance and the lifetime are reliant on the efficiency of heat dissipation. Poor heat dissipation leads to high operating temperatures, and consequently leads to device performance degradation and device lifetime reduction. To achieve better heat dissipation efficiency in nanoscale devices, a fundamental understanding of thermal transport at the nanoscale is essential. One of the biggest challenges to understanding the phonon transport processes at the nanoscale is the involvement of size effects and transport regimes. In nanostructures or low dimensional solids, the size of a system can be comparable to the phonon characteristic lengths (phonon mean free paths and coherence length), and consequently the thermal conductivity is highly dependent on material sizes and transport regimes.
In this dissertation, I present the study of nanoscale size effects in three different systems with varied levels of complexities. These three projects center around varying the most important material transport properties: number of propagating modes and transmission per mode. I first discuss thermal transport from the ballistic to the diffusive regime in an intrinsic two-dimensional material, specifically monolayer transition metal dichalcogenides (TMDs). In this work, number of modes will not change from the ballistic to the diffusive regime, and the only factor that would vary with the transport regime and the material size is the transmission. We show that with varying sample sizes, the ordering of thermal conductivity among monolayer transition metal dichalcogenides MX2 (M: Mo, W; X: S, Se) changes as phonon transport transits from the ballistic to the diffusive regime, driven by the competition between the phonon conduction frequency range and the scattering strength (or reduction of phonon transmission). Then I study size effects of thermal transport across an impedance bridged interface, with both number of modes and transmission changing from the harmonic to the strongly anharmonic limit. In this project, we introduce a quantity called "conserving modes" which is the upper-bound of number of modes that conserves momentum and energy in the harmonic limit. It shows that in the harmonic limit, the enhancement of conductance is limited due to fewer available channels, while in the anharmonic limit, the thermal conductance is strongly enhanced due to added inelastic channels. In the last project, I study the size effects demonstrating phonon coherence in superlattices. In these structures, number of modes can be tuned by varying the periodic dimension of superlattices to form new phonon band structures. We propose a method built in Non-Equilibrium Green's Function(NEGF) formalism to illustrate the scattering mechanisms and transport regimes. In addition, we introduce phenomenological models of momentum and phase scattering with the flexibility of adjusting the strength of these scatterings. We find momentum scattering is the only source of thermal resistance while phase scattering only removes wave interference and oscillations in the transmission.
These three projects we show in this dissertation extend our understanding of the size effects in nanoscale thermal transport, and the physics behind these systems can be further applied to other structures as well. The work in this dissertation will benefit the design of better heat dissipation in nanoscale devices.
