Theoretical studies of complex magnetism

Dnr:

SNIC 2018/1-54

Type:

SNIC Large Compute

Principal Investigator:

Lars NordstrÃ¶m

Affiliation:

Uppsala universitet

Start Date:

2018-07-01

End Date:

2019-07-01

Primary Classification:

10304: Condensed Matter Physics

- Tetralith at NSC: 351 x 1000 core-h/month
- Kebnekaise at HPC2N: 300 x 1000 core-h/month
- Kebnekaise Large Memory at HPC2N: 50 x 1000 core-h/month
- Beskow at PDC: 50 x 1000 core-h/month

Within this project the electronic structure of strongly correlated electron materials is calculated realistically in order to determine non-trivial broken symmetry solutions. Especially there exist several materials which exhibit ground states with broken symmetry but where the actual order parameters are not known.
The calculations are based on density functional theory (DFT) where the correlation is taken into account by adding a general term for the strong interaction of the electrons within an open shell. This approach is dubbed DFT+U or DFT+DMFT depending on the sophistication of the treatment of the correlation energy, static or dynamic mean field, respectively. These techniques are combined with an exact decomposition of the correlation energies into so-called multipole contributions. These multipoles in turn can sometimes be identified as order parameters. Especially the DFT+DMFT calculations can be very demanding with its quantum Monte Carlo simulations of the correlation energies.
Materials of interest include so-called heavy fermion materials, with correlated f-electrons, and so-called high Tc super-conducting materials, with correlated d-electrons. All these materials are either magnetic or close to a magnetic order. In both these types of materials there are several cases with anomalous ground states, which properties are with high probability determined by hidden order parameters that arise from non-trivial symmetry breaking. The aim of the project is to identify these states with different types of magnetic multipole ordering.
Another sub-project is to perform realistic spin dynamic simulations on magnetic materials of different geometries. With this studies we aim to understand processes such as spin reversal and formations of topologically stable magnetic structures, so-called skyrmions. The interaction parameters that go into these calculations are calculated by electronic structure methods on the real materials. Depending on the system various type of interactions can com into play. Besides usual bi-linear coupling (Heisenberg interactions), there can be anisotropic interactions, many-body interactions, bi-quadratic and higher order etc and of course the magnetic anisotropy terms. Many processes involve a large number of atoms which may result in very heavy computations.