Post date: Nov 29, 2012 6:34:02 PM
Physical Review B just published our recent work on thermal transport nanowires and alloys based on Mg2Si and Mg2Sn. You can access the article here, "Thermal conductivity of bulk and nanowire Mg2SixSn1-x alloys from first principles." Thermoelectrics based on these alloys show great promise due to the fact that their base materials are non-toxic and relatively cheap. Current thermoelectrics like PbTe rely on toxic and expensive materials. However, in order for Mg2Si and Mg2Sn thermoelectrics to be competitive, their thermoelectric figure of merit still needs to be improved. One route to do this is to reduce the thermal conductivity of the material through nanostructuring. In this article, we use first principles calculations to examine how alloying and nanostructuring affect the thermal conductivity of these materials. The initial results are encouraging and indicate that there is room for improvement in the figure of merit.
Technical Note: Up to this point, much of our work predicting thermal conductivity in materials based on first principles, has been done using plane wave codes (i.e. Quantum Espresso). In this study, we wanted to see if we could also get reliable results from localized orbital approaches like Siesta. Overall, we found that both approaches gave good agreement for the phonon dispersions of these materials as well as thermal transport. However, some care is needed with Siesta calculations to make sure you have a reasonable basis set and a good mesh cutoff.
This work was done in collaboration with Natalio Mingo's group at CEA Grenoble, David Broido at Boston College, and Lucas Lindsay at the Naval Research Laboratory. The research was funded by the NSF.