SKM 2023 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
MA: Fachverband Magnetismus
MA 27: Electron Theory of Magnetism and Correlations
MA 27.7: Vortrag
Mittwoch, 29. März 2023, 11:15–11:30, HSZ 403
Data-driven estimation of spin models in undoped cuprates — •Denys Y. Kononenko1, Ulrich K. Rößler1, Jeroen van den Brink1,2, and Oleg Janson1 — 1Institute for Theoretical Solid State Physics, IFW Dresden, Dresden, Germany — 2Institute for Theoretical Physics, TU Dresden, Dresden, Germany
Undoped cuprates host a wide variety of low-dimensional and frustrated spin models. The typically leading antiferromagnetic contribution to a magnetic exchange can be accurately estimated if the respective transfer integral is known. To date, the computational estimation of the transfer integral involves a well-established but cumbersome computational procedure. We demonstrate how the Gaussian Process Regression (GPR) model, trained on the results of the density functional theory calculations, can be employed to predict the transfer integrals using crystal structure as the only input. The GPR model receives descriptors of the local crystal environment of two copper sites as an input. The descriptors are based on the truncated expansion of the site position functions on the basis of the three-dimensional Zernike functions [1]. In this way, information on the spatial configuration and the chemical composition of the local crystal environment is incorporated into the descriptor. The approach facilitates rapid screening of spin models with desirable features among a broad range of known and unknown cuprates.
[1] M. Novotni and R. Klein, Computer Aided Design 36, 1047 (2004)