Regensburg 2019 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
MM: Fachverband Metall- und Materialphysik
MM 4: Methods in Computational Materials Modelling (methodological aspects, numerics)
MM 4.4: Vortrag
Montag, 1. April 2019, 11:00–11:15, H45
Flexible and User-friendly Symmetry-Constrained Optimization and its Application to Study Different SiO2 Polymorphs — •Sara Panahian Jand1, Maja-Olivia Lenz1, David Hicks2, Stefano Curtarolo2, Matthias Scheffler1, and Christian Carbogno1 — 1Fritz-Haber-Institut der Max-Planck-Gesellschaft, Berlin — 2Duke University, Durham NC, USA
In computational solid-state physics, crystal symmetries are frequently exploited to accelerate and improve electronic-structure calculations. However, the standard techniques are usually not applicable when the global symmetry is broken, e.g, in defective materials or thermodynamic investigations. To overcome this problem, we have developed and implemented flexible geometry-constrained relaxation in the all-electron code FHI-aims. This approach preserves the local symmetries on a per-atom basis via user-defined symmetry-reduced parameters as presented in the AFLOW library of prototypes [1]. We demonstrate the flexibility and power of the approach by studying the stabilities of different SiO2 polymorphs. For this compound, it is generally difficult to model structural transitions at very high temperatures and/or pressures [2]. However, our method provides a rapid way to compute the phase diagram of SiO2, including the temperature-dependent lattice expansion. This demonstrates the ability of the developed approach to aid and accelerate the systematic search for metastable structures and the calculation of their thermodynamic properties.
[1] M. J. Mehl, et al., Comp. Mater. Sci. 136, S1 (2017).
[2] C. Catlow, et al., Phys. Chem. Chem. Phys., 12, 786 (2010).