Based on theoretical modelling, the aim of the thesis is to improve our understanding of transport mechanisms in nanostructured semiconductors useful for thermoelectric applications. In particular thermal properties will be calculated using atomistic methods. The long-term issue is to propose innovative solutions to increase the conversion rate of wasted energies. Thermoelectric effects allow converting a temperature gradient into a voltage (Seebeck effect), and therefore allow reusing part of the wasted heat. Till now the small efficiency of the conversion prevents them to be used in large-scale applications. However, recent advances made in the synthesis of nanostructured materials can increase the figure of merit (conversion efficiency), e.g. high quality nanowires with a controlled chemistry, or bulk materials containing spherical inclusions. Those improved geometries creates additional scattering for the phonons carrying heat into those materials, and therefore decrease their thermal conductivity which is a key issue for material optimization. 

The proposed thesis is dedicated to study the impact on the thermal conductivity of spherical inclusions of Ge:Mn in a Ge matrix. Very recently this system has been synthesized by our colleagues in Grenoble (Dr. O. Bourgeois at Néel Institute, see Figure 1), and a complex behaviour was obtained for the thermal conductivity, which, up to now is not understood. One of the peculiarities of this system is that several scales are involved in the transport of heat. Phonons are a direct consequence of the strength of the bonding at the atomic scale, while the scattering of those phonons by the inclusions or surface boundaries involves much larger wavelengths. Therefore, to compute the thermal conductivity, a multi-scale approach is unavoidable. It can be decomposed into several steps. 

1) Study the stability of the material using density functional calculations and classical molecular dynamics. In particular a detailed comparison of the two methods will be performed to discuss the surface states of the Ge/Ge:Mn interface. After this step the phonons properties of the bulk Ge will be known, and the interface characterized. 

2) Transfer the phonon properties obtained quantum mechanically for the bulk Ge into a Monte Carlo solver of the Boltzmann equation. This will allow obtaining the thermal conductivity of the system without any adjustable parameters for Ge/Ge:Mn system, but neglecting the effect of interfaces. 

3) The third step, the effect of interfaces should be considered. This will be done in incorporating the transmission function of the Ge/Ge:Mn interface obtained from classical molecular dynamics into the Monte Carlo solver. 

4) Finally the thermal conductivity of the nanocomposite system will be appraised with both molecular dynamics and Monte Carlo method and compared to experimental measurements carried out at Néel Institute. 

In order to perform the above multi-scale modelling, several computer codes will be used (VASP, LAMMPS, MEDEA…), and/or developed for the occasion. The PhD candidate should therefore have a strong background in solid states physics and numerical simulations. 

Advisers: Dr. Ass. Prof. L. Chaput, Dr. CR/CNRS K. Termentzidis and Prof. Dr. D. Lacroix 

Location: LEMTA, UMR 7563, Un. Of Lorraine and CNRS, Nancy, France 

Keywords: Phonons, nanoinclusions, DFT, MD, MC, Ge:Mn/Ge 

Applications should be submitted till end of July 2016 to: 

Laurent CHAPUT
[email protected]
Tel: +33 3 83 59 55 23

Konstantinos TERMENTZIDIS
[email protected]
Tel: + 33 3 83 59 57 57

David LACROIX
[email protected]
Tel: +33 3 83 59 55 19