guest ::
TY - THES AU - Kipp, Jonathan Martin TI - Machine learning models for chiral transport in magnetic systems PB - RWTH Aachen University VL - Dissertation CY - Aachen M1 - RWTH-2024-10466 SP - 1 Online-Ressource : Illustrationen PY - 2024 N1 - Veröffentlicht auf dem Publikationsserver der RWTH Aachen University N1 - Dissertation, RWTH Aachen University, 2024 AB - Achievements in the field of solid state physics have shaped our way of life profoundly over the last century, propelling mankind into an era of wearable electronics, worldwide access to electronically stored information, and computing power previously unheard of. Mobile phones, which can be found in almost everyone's pocket today, perfectly illustrate the pivotal conflict between simultaneous miniaturization of devices and increases in computing power and storage density, that has been at the heart of solid state research since the fabrication of the world's first transistor. In addition, there is much interest in computing architectures beyond the standard, semiconductor-based computers. Now, materials with non-collinear magnetization textures, such as domain walls or skyrmions, present themselves as prime candidates in the quest for miniature electronics and unconventional computing platforms alike, due to their small size, their stability, and their intrinsic, non-linear characteristics, enabling complex arithmetic operations. However, a clear description of charge, heat, or spin transport in complex magnetization textures is still sought after. Therefore, a systematic way of conquering the complexity in canted magnets or chiral textures is a key objective in the field of spintronics. In this thesis, explicit tight-binding calculations of the anomalous Hall effect on a two-dimensional, magnetic honeycomb lattice, are exploited in a threefold manner in order to introduce the vector chirality of a magnetic texture as a powerful order parameter. First, the chiral Hall effect in canted magnets is established on equal footing with the anomalous Hall effect of collinear ferromagnets and antiferromagnets, by identifying the chiral Hall effect as the contribution to the anomalous Hall effect linear in vector chirality. Second, by classifying different parts of the Hall signal with respect to the vector chirality of the magnetic configuration and the underlying crystal symmetry, the numerical data reproduces the functional form and directional dependence obtained from an expansion of the anomalous Hall effect in gradients of the magnetization. Lastly, numerical data is utilized for training a linear model of the anomalous Hall effect, encompassing effects up to arbitrary order in the magnetization, which is constructed from the symmetric invariants of the lattice symmetry. Through explicitly training non-chiral and chiral models, this constructive method demonstrates the fingerprint of chiral magnetic textures in electric transport properties. LB - PUB:(DE-HGF)11 DO - DOI:10.18154/RWTH-2024-10466 UR - https://publications.rwth-aachen.de/record/996251 ER -