The proposed research project is all about magnetism and its interactions with electronics (spintronics), spin waves (magnonics) or heat (spin caloritronics). These interactions can be employed to facilitate many interesting novel applications, in particular in information technology. Examples are new generation of memory devices, spin transfer torque magnetic random access memory (STT-MRAM) and race-track memory, the latter employing either domain wall motion or skyrmions. These applications are potentially more energy efficient and better performing than the devices used today. Moreover, electromagnetic and/or thermal fields can be used to manipulate spin waves to create an entire new interesting class of materials, magnonic devices, for example transistors and diodes. The basic principle in many of the applications is to influence or control the orientation of magnetic moments with external sources that in turn affects the transport properties of the system. These magnetization dynamics processes will be studied in great detail using theoretical modelling using large-scale computations with a combination of first-principles calculations and atomistic spin dynamics simulations within our own developed UppASD software package that may be downloaded from https://github.com/UppASD/UppASD. The developed theoretical framework is accurate but still fast enough to reach device length scales using massive fine-grained MPI parallelization and has a well proven track record. Contrary to most other modelling efforts, we employ material specific studies that are able to give not only qualitative results but also quantitative results that can be compared and tested against experimental methods. Recent developments in the software are inclusion of longitudinal fluctuations of the magnetic moments that yield even more accurate description of the magnetic properties at finite temperature as well as an implementation of a new computational framework for combined spin and molecular dynamics where the magnon and phonon properties are coupled together. These calculations will allow for novel studies where spin and thermal transport are treated on equal footing, for instance in spincalometric devices where thermal gradients generate spin currents and vice versa (spin-Seebeck and related effects).
The work is an interdisciplinary e-Science project connecting both core areas (visualization, numerical analysis and algorithmic developments) and applied areas (electronic structure and molecular simulations). Application expert support will be an important ingredient for the methodological development of the combined spin and molecular dynamics simulation framework. An attempt to perform multi scale modelling in magnetization dynamics using the hetrogeneous multiscale method (HMM) has recently been initiated as well as an implementation of Fast Ewald method for electrostatic interactions (NA,KTH).