Microscopic Origin of Linear Magneto.Optical effect in BCC FE

Abstract

The recent rapid development in first principle calculations of solids provided the missing link between the simplified theoretical models and the complex nature of experiments to the point, where the comparison of the calculation with the experiment has become a routine in the solid state research. This opens up novel ways of analyzing the results and new opportunities to explore and learn.

We think that the ab initio calculations are not used to its full potential. Most of the time, we only care about the result and whether it agrees with the experiment and when it does, we say that it is well understood. But our understanding of the physics involved is often stuck within those simplified analytical models. In this thesis, we would like to demonstrate, how the ab initio calculations can be used to connect the results to those analytical models and that there is still a lot to be discovered.

We will focus on the magneto-optical spectra of bcc Fe. We perform ab initio calculations of the permittivity tensor of bcc Fe by the WIEN2k code and compare it to the experiment performed in our laboratory. We then proceed to analyze the results. The magneto-optical spectra are a result of summation over all pairs of bands and integration over the Brillouin zone that all collapse to that single curve. We shall investigate what features form the final curve and how the magneto-optical effect originates in the electronic structure. It involves sorting out all the data, performing additional numerical calculations and developing new ways of visualizing the results.

We also stumble upon the topological aspect that seems to naturally arise alongside the investigation of the magneto-optical effect. We shall see that the origins of the magne\-to-optical effect and the Berry curvature are connected. We will also use our analyzing and visualizing tools to deepen our understanding of the Berry curvature flow in the Brillouin zone and related topological features.