Experimental and modeling evidence for the reduction of thermal conductivity in Mg2Si by fine tuning the nano & micro-structural features

Abstract

In this paper, experimental and modeling evidence for reduced thermal conductivity in Bi-doped Mg2Si is presented. Upon doping, lattice thermal conductivity was reduced monotonically and, at the same time, MgO content was increased while grain size was decreased. A model has been developed, which takes into account various phonon-scattering mechanisms, the 3-phonon (Umklapp) process, electron-phonon interaction and phonon-scattering by the grain-boundaries or nano-scale MgO inclusions. The model was successfully applied and excellently describes the temperature dependence of lattice thermal conductivity as well as its evolution vs. doping. Application of the model suggests that lattice thermal conductivity is mainly affected by both nano-MgO precipitates and Umklapp processes. Exploring the effect of the grain size and nano-precipitates, our model predicts that the two phonon-scattering mechanisms are interconnected; for a given grain size there is an optimal value for nano-precipitate concentration, and vice-versa. This conclusion may lead to a general approach in the optimization of lattice thermal conductivity in thermoelectric materials, by fine tuning both the nano- as well as the micro-structural features.