We report on the effect of germanium (Ge) coatings on the thermal transport properties of silicon (Si) nanowires using nonequilibrium molecular dynamics simulations. Our results show that a simple deposition of a Ge shell of only 1 to 2 unit cells in thickness on a single crystalline Si nanowire can lead to a dramatic 75% decrease in thermal conductivity at room temperature compared to an uncoated Si nanowire. By analyzing the vibrational density states of phonons and the participation ratio of each specific mode, we demonstrate that the reduction in the thermal conductivity of Si/Ge core-shell nanowire stems from the depression and localization of long-wavelength phonon modes at the Si/Ge interface and of high frequency nonpropagating diffusive modes.

The thermal conductivity of silicon thin films is predicted in the directions parallel and perpendicular to the film surfaces (in-plane and out-of-plane, respectively) using equilibrium molecular dynamics, the Green-Kubo relation, and the Stillinger-Weber interatomic potential. Three different boundary conditions are considered along the film surfaces: frozen atoms, surface potential, and free boundaries. Film thicknesses range from 2to217nm and temperatures from 300to1000K. The relation between the bulk phonon mean free path (Λ) and the film thickness (ds) spans from the ballistic regime (Λ⪢ds) at 300K to the diffusive, bulk-like regime (Λ⪡ds) at 1000K. When the film is thin enough, the in-plane and out-of-plane thermal conductivity differ from each other and decrease with decreasing film thickness, as a consequence of the scattering of phonons with the film boundaries. The in-plane thermal conductivity follows the trend observed experimentally at 300K. In the ballistic limit, in accordance with the kinetic and phonon radiative transfer theories, the predicted out-of-plane thermal conductivity varies linearly with the film thickness, and is temperature-independent for temperatures near or above the Debye’s temperature.

Using nonequilibrium molecular dynamics simulations, we study the thermal diode effect in a system composed of silica, self-assembled monolayers (SAMs) at the silica surface and water surrounding this system, by imposing a series of positive and negative heat currents. We have found that in the limit of large heat currents, the thermal conductance at the SAMs-water interface is about 1000 MW/m2 K at room temperature for heat flowing from the SAMs to the water and 650 MW/m2 K for heat flowing from the water to the SAMs, respectively, resulting in a thermal rectification of up to 54%. Analysis of the radial distribution function of oxygen-oxygen atoms in water indicates that the origin of the thermal rectification resides in the strong temperature dependence of the hydrogen bonds in water.

We report the effect of water nanoconfinement on the thermal transport properties of two neighbor hydrophilic quartz interfaces. A significant increase and a nonintuitive, nonmonotonic dependence of the overall interfacial thermal conductance between the quartz surfaces on the water layer thickness were found. By probing the interfacial structure and vibrational properties of the connected components, we demonstrated that the mechanism of the peak occurring at submonolayer water originates from the freezing of water molecules at extremely confined conditions and the excellent match of vibrational states between trapped water and hydrophilic headgroups on the two contact surfaces. Our results show that incorporation of polar molecules into hydrophilic interfaces is very promising to enhance the thermal transport through thermally smooth connection of these interfaces.

Two mechanisms that enhance heat dissipation at solid-liquid interfaces are investigated from the atomistic point of view using nonequilibrium molecular dynamics simulation. The mechanisms include surface functionalization, where –OH terminated headgroups and self-assembled monolayers (SAMs) with different chain lengths are used to recondition and modify the hydrophilicity of silica surface, and vibrational matching between crystalline silica and liquid water, where three-dimensional nanopillars are grown at the interface in the direction of the heat flux with different lengths to rectify the vibrational frequencies of surface atoms. The heat dissipation is measured in terms of the thermal conductance of the solid-liquid interface and is obtained by imposing a one-dimensional heat flux along the simulation domain. A comparison with reported numerical and experimental thermal conductance measurements for similar interfaces indicates that the thermal conductance is enhanced by 1.8–3.2 times when the silica surface is reconditioned with hydrophilic groups. The enhancement is further promoted by SAMs, which results in a 20% higher thermal conductance compared with that of the fully hydroxylated silica surface. Likewise, the presence of nanopillars enhances the interface thermal conductance by 2.6 times compared with a bare surface (without nanopillars). Moreover, for different nanopillar densities, the conductance increases linearly with the length of the pillar and saturates at around 4.26 nm. Changes in the vibrational spectrum of surface atoms and water confinement effects are found to be responsible for the increase in conductance. The modification of surface vibrational states provides a tunable path to enhance heat dissipation, which can also be easily applied to other fluids and interfaces.

A hierarchical model of heat transfer for the thermal analysis of electronic devices is presented. The integration of participating scales (from nanoscale to macroscales) is achieved by (i) estimating the input parameters and thermal properties to solve the Boltzmann transport equation (BTE) for phonons using molecular dynamics (MD), including phonon relaxation times, dispersion relations, group velocities, and specific heat, (ii) applying quantum corrections to the MD results to make them suitable for the solution of BTE, and (iii) numerically solving the BTE in space and time subject to different boundary and initial conditions. We apply our hierarchical model to estimate the silicon out-of-plane thermal conductivity and the thermal response of an silicon on insulator (SOI) device subject to Joule heating. We have found that relative phonon contribution to the overall conductivity changes as the dimension of the domain is reduced as a result of phonon confinement. The observed reduction in the thermal conductivity is produced by the progressive transition of modes in the diffusive regime (as in the bulk) to transitional and ballistic regimes as the film thickness is decreased. In addition, we have found that relaxation time expressions for optical phonons are important to describe the transient response of SOI devices and that the characteristic transport regimes, determined with Holland and Klemens phonon models, differ significantly.

Molecular dynamics simulations are performed to estimate acoustical and optical phonon relaxation times, dispersion relations, group velocities, and specific heat of silicon needed to solve the Boltzmann transport equation (BTE) at 300 K and 1000 K. The relaxation times are calculated from the temporal decay of the autocorrelation function of the fluctuation of total energy of each normal mode in the ⟨100⟩ family of directions, where the total energy of each mode is obtained from the normal mode decomposition of the motion of the silicon atoms over a period of time. Additionally, silicon dispersion relations are directly determined from the equipartition theorem obtained from the normal mode decomposition. The impact of the anharmonic nature of the potential energy function on the thermal expansion of the crystal is determined by computing the lattice parameter at the cited temperatures using a NPT (i.e., constant number of atoms, pressure, and temperature) ensemble, and are compared with experimental values reported in the literature and with those computed analytically using the quasiharmonic approximation. The dependence of the relaxation times with respect to the frequency is identified with two functions that follow the functional form of the relaxation time expressions reported in the literature. From these functions a simplified version of relaxation times for each normal mode is extracted. Properties, such as group and phase velocities, thermal conductivity, and mean free path, needed to further develop a methodology for the thermal analysis of electronic devices (i.e., from nano- to macroscales) are determined once the relaxation times and dispersion relations are obtained. The thermal properties are validated by comparing the BTE-based thermal conductivity against the predictions obtained from the Green–Kubo method. It is found that the relaxation times closely resemble the ones obtained from perturbation theory at high temperatures; the contribution to the thermal conductivity of the transverse acoustic, longitudinal acoustic, and longitudinal optical modes being approximately 30%, 60%, and 10%, respectively, and the contribution of the transverse optical mode negligible.

No abstract