» Back to overview
Proposing Institution

Universität Regensburg Institut I - Theoretische Physik, Spintronik
Project Manager

Dr. Martin Gmitra
Universitätsstraße 31
93040 Regensburg
Two-dimensional (2D) materials have revolutionized condensed matter physics, and are starting torevolutionize technology as well. The lowered dimensionality offers the possibility to explore novelproperties on the nanoscale, especially important for device applications. Most of the propertiesoriginate from the electronic structure of the 2D materials. In this project we propose to performsystematic and comprehensive ab initio investigations of the electronic structure of two dimensionallayered materials and their heterostructures. The list of two dimensional materials is growing ata tremendous rate, and various new materials---such as semimetallic graphene, semiconductingtransition-metal dichalcogenides, and recently discovered ferromagnetic CrI3 or MnSe---have beenfound. Due to the layered structure of 2D materials, they are perfectly suited to build van der Waalheterostructures in a lego-like fashion, leading to novel effects, stemming from proximity physics. Our main research interest lies in the field of condensed matter physics within the scope ofspintronics. Spintronics utilizes the intrinsic angular momentum of an electron called spin. Spincan be viewed as an additional degree of freedom (akin to rotation) with a direction. Electrons thuscarry not only charge but also spin. The motion of an electron in matter is coupled to its spin due tothe fundamental interaction called spin-orbit coupling. Within the project we will investigate spinorbitcoupling and exchange effects induced due to proximity of a material to a substrate. This offersthe possibility to enhance, for example, the weak spin-orbit coupling of graphene or even make itmagnetic. Thus, graphene based heterostructures could be employed to build bipolar spintronicselements, such as spin diodes and spin transistors, or spin logic devices, which require to maketheir electronic bands spin dependent. Such calculations are usually based on supercell approaches,including many atoms in complex structures, which demands high-performance computation (HPC)facilities and usage of sophisticated theories and algorithms. The most established state-of-the-artapproaches are Density Functional Theory (DFT) with perturbative many-body corrections theorieslike GW. The current state-of-the art, also established by our consortium, is the application ofQuantum Monte Carlo (QMC) methods in the 2D realm. Such methods require HPC to be able tofind convergence of electronic properties with the increasing supercell sizes. Specifically, DFT andGW are very demanding in terms of memory usage and communication, while QMC is veryefficiently parallelizable (99%) as nearly no communication is required and it is scaling up tothousands of cores.

Impressum, Conny Wendler