Date

2019-11-03

Author

Malcıoğlu, Osman Barış

Metadata

Show full item record

Item Usage Stats

186
views

0
downloads

Cite This

Point defects in 2D materials, among other novel prospects, show promise as polarized and ultra-bright single-photon emitters at room temperature, with similar or better properties than nitrogen-vacancy center (NV) in diamond. A robust quantitative description of the charged defects is a prerequisite to explore the physics and band gap engineering in such cases using material discovery techniques. Such charged defects can be approached using ab Initio methods (i.e. [1-4]), however, the popular supercell approach used mandates charge neutrality per supercell. For charged systems this is enforced by adding fictional compensation charges, which introduces a spurious interaction. A posteriori correction schemes exist in the literature (i.e. [4-6]), however, model fitting the dielectric response, especially in the case of 2D materials, is far from straightforward and introduces additional parameters. The recent advances in finite elements techniques [7] allow a fully automated ab initio approach, where the microscopic polarizability is read from ab-initio calculation/code of choice. In this talk, we present a fully automated and efficient multiscale code for handling long-reach fields. The response of the medium and the localized charge density of the ab-initio calculations are represented by a finite element function space (discontinuous Galerkin). The relevant physical observables are obtained using an autoadaptive mesh solver. We demonstrate applications of our approach to select cases.

Subject Keywords

Charged state defects, FEM/FET, Inverse PDE problem, Machine learning

URI

https://hdl.handle.net/11511/74977

Collections

Department of Physics, Conference / Seminar