Patrice Chantrenne

Molecular dynamics simulations for the prediction of thermal conductivity of bulk silicon and silicon nanowires: Influence of interatomic potentials and boundary conditions

The reliability of molecular dynamics (MD) results depends strongly on the choice of interatomic potentials and simulation conditions. Five interatomic potentials have been evaluated for heat transfer MD simulations of silicon, based on the description of the harmonic (dispersion curves) and anharmonic (linear thermal expansion) properties. The best interatomic potential is the second nearest-neighbor modified embedded atom method potential followed by the Stillinger-Weber, and then the Tersoff III. However, the prediction of the bulk silicon thermal conductivity leads to the conclusion that the Tersoff III potential gives the best results for isotopically pure silicon at high temperatures. The thermal conductivity of silicon nanowires as a function of cross-section and length is calculated, and the influence of the boundary conditions is studied for those five potentials.

Thermal conductivity of GaAs/AlAs superlattices and the puzzle of interfaces

We present a molecular dynamics investigation of the cross-plane thermal conductivity of superlattices using the non-equilibrium molecular dynamics method. The purpose is to investigate the influence of the interfaces, which is expected to be important in those nanostructures where the superlattice period is smaller than the phonon mean free path. In contrast to previous studies, more realistic interfaces are considered: interfacial roughness is modeled using atomic rectangular islands and interdiffusion is taken into account. It is shown that thermal conductivity is very sensitive to the detailed interfacial shape and to the presence of interdiffusion. This may be relevant to recent experiments.

Amorphization and reduction of thermal conductivity in porous silicon by irradiation with swift heavy ions

In this article, we demonstrate that the thermal conductivity of nanostructured porous silicon is reduced by amorphization and also that this amorphous phase in porous silicon can be created by swift (high-energy) heavy ion irradiation. Porous silicon samples with 41%-75% porosity are irradiated with 110 MeV uranium ions at six different fluences. Structural characterisation by micro-Raman spectroscopy and SEM imaging show that swift heavy ion irradiation causes the creation of an amorphous phase in porous Si but without suppressing its porous structure. We demonstrate that the amorphization of porous silicon is caused by electronic-regime interactions, which is the first time such an effect is obtained in crystalline silicon with single-ion species. Furthermore, the impact on the thermal conductivity of porous silicon is studied by micro-Raman spectroscopy and scanning thermal microscopy. The creation of an amorphous phase in porous silicon leads to a reduction of its thermal conductivity, up to a factor of 3 compared to the non-irradiated sample. Therefore, this technique could be used to enhance the thermal insulation properties of porous Si. Finally, we show that this treatment can be combined with pre-oxidation at 300 °C, which is known to lower the thermal conductivity of porous Si, in order to obtain an even greater reduction.

A microscopic thermal model for dry sliding contact

Abstract A microscopic thermal model for dry sliding contact that accounts for a volume generation of the friction heat is proposed. A numerical procedure that allows the determination of the two parameters α and R s1 of macroscopic thermal models for dry sliding contact is given. The results are compared in the case of a simple contact geometry, with an analytical solution for α and R s1 . The influences of the microscopic parameters, the contact geometry and the velocity on the two macroscopic parameters are shown.

Thermal conductivity and thermal boundary resistance of nanostructures

AbstractWe present a fabrication process of low-cost superlattices and simulations related with the heat dissipation on them. The influence of the interfacial roughness on the thermal conductivity of semiconductor/semiconductor superlattices was studied by equilibrium and non-equilibrium molecular dynamics and on the Kapitza resistance of superlattices interfaces by equilibrium molecular dynamics. The non-equilibrium method was the tool used for the prediction of the Kapitza resistance for a binary semiconductor/metal system. Physical explanations are provided for rationalizing the simulation results.PACS68.65.Cd, 66.70.Df, 81.16.-c, 65.80.-g, 31.12.xv

Study of a macroscopic sliding contact thermal model from microscopic models

Abstract The macroscopic sliding contact thermal model studied here involves two parameters: a sliding contact thermal resistance and a heat generation coefficient which can be determined from a microscopic model. For the two proposed microscopic models, the aim is to provide correlations which allow fast and easy determination of the two macroscopic parameters. The various definitions of the thermal resistance are also studied and our results are compared with other published work. The influence of the microscopic model on the macroscopic parameters is studied. It is shown that the heat generation coefficient depends on the location of the heat generated by friction.

Characterization of the thermal conductivity of insulating thin films by scanning thermal microscopy

This paper reports on the abilities of a Scanning Thermal Microscopy (SThM) method to characterize the thermal conductivity of insulating materials and thin films used in microelectronics and microsystems. It gives a review of the previous works on the subject and gives new results allowing showing the performance of a new method proposed for reducing the thermal conductivity of meso-porous silicon by swift heavy ion irradiation. Meso-porous silicon samples were prepared by anodisation of silicon wafers and underwent irradiation by 845MeV ^2^0^8Pb ions, with fluences of 4x10^1^1 and 7x10^1^1cm^-^2. Thermal measurements show that irradiation reduced thermal conductivity by a factor of up to 2.

Parameter identification of a lumped parameter thermal model for a permanent magnet synchronous machine

In the thermal modeling of electric machines by lumped parameters, an important step is the tuning of influential poorly known parameters to calibrate the model. The use of Inverse methods based on the minimization of residuals between measured and calculated temperatures is therefore of great help. In this paper, the Gauss-Newton (GN) method, the Levenberg-Marquardt (LM) method and the Genetic Algorithms (GA) method are used to solve this optimization problem in order to identify 10 parameters of a lumped parameter thermal model for a permanent magnet synchronous machine (PMSM).

Molecular Dynamics Simulations and Kapitza Conductance Prediction of Si/Au Systems Using the New Full 2NN MEAM Si/Au Cross-Potential

Superlattices made by superposing dielectric and metal nanolayers are of great interest as their small size restricts the thermal energy carrier mean free path, decreasing the thermal conductivity and thereby increasing the thermoelectric figure of merit. It is, therefore, essential to predict their thermal conductivity. Potentials for Au and Si are discussed, and the potential of second nearest-neighbor modified embedded atom method (2NN MEAM) is chosen as being the best for simulating heat transfer in Si/Au systems. Full 2NN MEAM Si/Au cross-potential parameterization is developed, and the results are compared with ab initio calculations to test its ability to reproduce local density approximation (LDA) calculations. Volume-constant (NVT) molecular dynamics simulations are performed to deposit Au atoms on an Si substrate by physical vapor deposition, and the results of the intermixing zone are in good agreement with the Cahn and Hilliard theory. Nonequilibrium molecular dynamics simulations are performed for an average temperature of 300 K to determine the Kapitza conductance of Si/Au systems, and the obtained value of 158 MW/m 2 K is in good agreement with the results of Komarov for Au deposited on isotopically pure Si- 28 and natural Si, with values ranging between 133 and 182 MW/m2 K.

Application of classic and T lumped parameter thermal models for Permanent Magnet Synchronous Machines

Heat Transfer simulation in Permanent Magnet Synchronous Machines is mandatory in order to predict temperatures during the design phase. The temperatures in the machine condition the choice of the correct materials as well as the cost of the machine. As these machines are complex, a compromise between accuracy and computational cost is necessary. In this paper, we present two thermal Permanent Magnet Synchronous Machine models developing using nodal approach. The first model corresponds to a classic nodal method, with a fine meshing of the machine. The second is made of a network of thermal resistors and capacitors set up using the T-nodal method with a crude meshing. This approach is quite popular as the computation time is quite short. Both models take into account the variation of copper heat losses with copper and magnet temperatures.