Computational Modeling of Superconductivity and Out-of-Equilibrium Dynamics

Dnr:

SNIC 2020/1-18

Type:

SNIC Large Compute

Principal Investigator:

Peter Oppeneer

Affiliation:

Uppsala universitet

Start Date:

2020-07-01

End Date:

2021-07-01

Primary Classification:

10304: Condensed Matter Physics

Secondary Classification:

10302: Atom and Molecular Physics and Optics

Tertiary Classification:

10404: Inorganic Chemistry

Webpage:

- Tetralith at NSC: 250 x 1000 core-h/month
- Beskow at PDC: 200 x 1000 core-h/month
- Kebnekaise at HPC2N: 100 x 1000 core-h/month
- Hebbe at C3SE: 100 x 1000 core-h/month
- Rackham at UPPMAX: 50 x 1000 core-h/month
- Tegner at PDC: 3 x 1000 core-h/month

The access to excellent computer resources has become a key ingredient in modern materials research. In the present proposal we are asking for computer time for mainly two research areas in which we are currently working. These research areas are computationally demanding; there is however an enormous scientific gain in terms of new knowledge, fundamental understanding, and materials' predictions that could not have been obtained otherwise.
In the first area we perform fully selfconsistent calculations of temperature-dependent superconductivity in real materials. To do this we have in the last years developed the Uppsala Superconductivity (UppSC) code, which is a full-bandwidth, multi-band, anisotropic Eliashberg code for selfconsistent calculations of unconventional superconductivity using full ab initio input from DFT calculations. This code is worldwide unique. It can e.g. treat phonon-mediated pairing and spin-fluctuations-mediated Cooper pairing, as well as conventional and unconventional multiband superconductivity. Until recent fully ab initio calculations of superconductivity have not been possible; we aim however at changing this situation. There are several computationally intensive steps needed to obtain temperature-dependent superconducting quantities (see proposal).
Another area on which we are currently working is the ab initio modeling of out-of-equilibrium dynamics to simulate ultrafast pump-probe experiments. Here we perform ab initio calculations of the strongly enhanced electron and lattice dynamics that occur in solid-state systems upon excitation with an ultrashort laser pulse. To this end we have developed a theory for describing the out-of-equilibrium ultrafast relaxation dynamics, which is completely ab initio based and takes all non-equilibrium processes (e.g., electron-phonon, phonon-phonon interactions) into account. For the modeling we need to compute a number of quantities, such as phonon dispersions and lifetimes and wavevector and the mode dependent electron-phonon coupling.
A further area (not mentioned in the title) concerns state-of-the-art calculations to study the electronic structure, magnetic and structural properties of spin-bearing metal-organic materials (spin-crossover materials) and single molecule magnets. We are investigating the possible spin switching of metalorganic molecules on a metallic surface and in contact with a reagent (e.g., nitric oxide) in order to discover conditions under which the spin can be switched in a molecular electronic device. Another research direction here is the ab initio molecular dynamics description and prediction of spin-crossover systems, invoking temperature effects through molecular dynamics simulations.
Notably, the scientific gain achieved by performing these calculations is enormous which emphasizes the strength of computational materials modeling.