Heat transport at nanoscales in semiconductors is investigated with a statistical method. The Boltzmann Transport Equation (BTE) which characterize phonons motion and interaction within the crystal lattice has been simulated with a Monte Carlo technique. Our model takes into account media frequency properties through the dispersion curves for longitudinal and transverse acoustic branches. The BTE collisional term involving phonons scattering processes is simulated with the Relaxation Times Approximation theory. A new distribution function accounting for the collisional processes has been developed in order to respect energy conservation during phonons scattering events. This non deterministic approach provides satisfactory results in what concerns phonons transport in both ballistic and diffusion regimes. The simulation code has been tested with silicon and germanium thin films; temperature propagation within samples is presented and compared to analytical solutions (in the diffusion regime). The two materials bulk thermal conductivity is retrieved for temperature ranging between 100 K and 500 K. Heat transfer within a plane wall with a large thermal gradient (250 K-500 K) is proposed in order to expose the model ability to simulate conductivity thermal dependence on heat exchange at nanoscales. Finally, size effects and validity of heat conduction law are investigated for several slab thicknesses.

International audienceHeat conduction of nanoporous silicon and germanium thin films is studied thanks to a statistical approach. Resolution of phonon Boltzmann transport equation is performed with a Monte Carlo technique in order to assess thermal conductivity. Sensitivity of this latter property with respect to parameters such as phonon mean free path and characteristics of the pores ( distribution, size, porosity) is discussed and compared to predictions from analytical models. Results point out that thermal properties might be tailored through the design of the porosity and more specifically by the adjustment of the phonon-pore mean free path. Finally, an effective medium technique is used to extend our work to multilayered crystalline-nanoporous structures. Results show that ought to pore scattering, a diffusive Fourier regime can be recovered even when the film thickness is below the bulk limit

The authors study the thermal conductivity of silicon nanowires by simulation of phonon motion and interactions through a dedicated Monte Carlo model. This model solves the Boltzmann transport equation, taking into account silicon acoustic mode dispersion curves and three phonon interactions (the normal and umklapp processes). The confinement, which limits the thermal conductivity in such structures, is described by diffuse reflection at lateral boundaries of the nanowire without any adjustment by a boundary collision time, which depends on a specularity factor. They compare simulation results to experimental measurements on similar nanostructures. A good agreement is achieved for almost all the considered diameters.

A spectroscopic study of a laser-induced plume created during the welding of stainless steel and other materials (iron and chromium) has been carried out. A pulsed Nd:YAG laser of 1000 W average power is used. The evolutions of the electron temperature and electron density have been studied for several welding parameters. We use working powers from 300 to 900 W and pulse durations between 1.5 and 5 ms. The influence of shielding gases like nitrogen and argon has been taken into account. Temperature and density calculations are based on the observation of the relative intensities and shapes of the emission peaks. We assume that the plasma is in local thermal equilibrium. The temperature is calculated with the Boltzmann plot method and the density with the Stark broadening of an iron line. The electron temperatures vary in the range of 4500–7100 K, electron density between 3×1022 and 6.5×1022 m−3. The absorption of the laser beam in the plasma is calculated using the Inverse Bremsstrahlung theory.