Contact
Juan Maria García Lastra: jmgla@dtu.dk
Tejs Vegge: teve@dtu.dk
Challenge: Density Functional Theory (DFT) has widely and successfully used to describe the electronic properties of a vast number of materials. However it is well-known that conventional DFT suffers from a coulombic self-interaction (SI) error, which tends to delocalized electrons and holes. For metallic systems the SI error does not cause many troubles. However in insulator materials, where electrons and holes tend to be localized, the SI error may lead to a completely wrong description of the material. There are various ways to overcome the SI error of conventional DFT. The easiest way is to use Hubbard corrections, i.e. the DFT+U approximation, where a correction U parameter is applied to specific atoms. The DFT+U technique is often considered as a semi-empirical method because the U value can be tuned to reproduce a specific experimental property of the studied system. The question here is which experimental property makes more sense to reproduce by using the U correction from the physical point of view.
Idea: We think that the band gap is the experimental property that it should be reproduced by tuning the U parameter. We want to corroborate that if we make this choice then other important properties of material relevant in energy applications (for instance, cathode materials in batteries or polymers in organic solar cells), such as the energy barriers involved in the transport of localized electrons, will be also correctly described.
Your task: Your task here will be to use a tight-binding model to prove that if the U parameter is used to match the experimental bandgap of a material then the energy barriers involved in the transport of localized electrons will be also well described.
