Merging GW and dynamical mean-field theory

点击翻译按钮获取中文摘要

Material calculations on the basis of the local density approximation (LDA) plus dynamical mean field theory (DMFT) approach represent a major step forward, since strong electronic correlation effects are included. For truly parameter-free ab initio calculations, however, two severe shortcomings persist: (i) the screened Coulomb interaction is usually considered to be an adjustable parameter in LDA+DMFT and (ii) the so-called double counting problem, e.g., it is difficult to determine the electronic correlations already accounted for on the LDA level. While one might mitigate these problems by constrained LDA calculations, a conceptionally preferable and well-defined approach is to substitute LDA by GW. The main goal of this project is to implement the GW+DMFT approach and to make the program package available to the scientific community. We will apply this approach in cases where GW+DMFT is superior from a physical point of view, i.e., if the exchange is important besides strong electronic correlations. We will study semiconductors where GW estimates band gap values much better than LDA and where, if transition metals are present, strong correlations need to be included as well. A second application are novel superconductors such as cuprates and iron pnictides, where the exchange interactions between transition metal d electrons and O or As p holes might be important.