The efficiency of current energy production methods can be improved by utilizing the thermo-electric effect to extract electricity from waste heat. As a matter of fact, the thermoelectric effect can be exploited anywhere where there are local thermal gradients. However, the efficiency of the thermoelectric conversion materials is still lacking. Additionally, in order to use thermoelectric converters on a global scale, or to wear them on the skin, the materials have to be chosen carefully. The current best performing materials contain lead, bismuth and tellurium, and their inherent toxicity and scarcity on the Earth's crust prohibits their use on a global scale. For this reason, a lot of effort has been put to the research of oxides as thermoelectric materials. Development and fine tuning of such materials is an expensive and time-consuming process, but all the relevant material properties can be predicted with computational methods, cutting time spent on the laboratory. This thesis focuses on the accuracy of the computational approaches, especially on d-metal oxides where pure density functional theory methods are known to perform poorly. Most effort was put to the accuracy of estimating the lattice thermal conductivity of materials. For the first time, the studies combined the hybrid functionals of density functional theory with the lattice thermal conductivity calculations. The results showed that lattice thermal conductivity calculated with hybrid functionals was in much better agreement with the experimental measurements than the results obtained with computationally less costly generalized gradient approximation functionals. Additionally, the first hybrid density functional calculations on the lattice thermal conductivity of magnetic materials were performed, and the effect of the used symmetry on the lattice dynamical properties and lattice thermal conductivity was scrutinized. The hybrid density functionals were found to improve the accuracy on the calculations of the thermoelectric properties of d-metal oxides based on comparisons to previous computational work. The performance of some Earth-abundant d-metal oxides was estimated and they were shown to have promise. The pure oxides, however, are far from efficient, and there is still work to be done to improve their efficiency by doping and/or nanostructuring.
|Translated title of the contribution||Tarkempi d-metallioksidien termosähköisten ominaisuuksien ennustaminen tiheysfunktionaaliteorialla|
|Publication status||Published - 2019|
|MoE publication type||G5 Doctoral dissertation (article)|
- thermoelectric properties
- d-metal oxides
- lattice thermal conductivity
- computational methods