Spintronic devices and their operation are governed by the microstructure of magnetic fields. These magnetic field structures undergo complex, drastic changes when an external magnetic field is applied to the system. The resulting fine structures cannot be reproduced, and it is difficult to measure the complexity of magnetic field structures. Our understanding of the magnetization reversal phenomenon is therefore limited to rough visual inspections and qualitative methods, which represent a serious bottleneck in material design. Even the stability and shape of magnetic field structures in Permalloy, a well-known material that has been studied for over a century, has been difficult to predict.
Addressing this issue, a team of researchers led by Professor Masato Kotsugi of Tokyo Science University in Japan has recently developed an artificial intelligence-based method to analyze material functions in a more quantitative way. In their work published in Science and Technology of Advanced Materials: Methods, the team used topological data analysis and developed a super-hierarchical and descriptive analysis method for magnetic reversal processes. In simple terms, superhierarchy, according to the research team, refers to the link between micro and macro features that are often treated in isolation but that in the grand scheme of things jointly contribute to the physical explanation.
The team measured the complexity of magnetic field structures using persistent homology, a mathematical tool used in computational topology that measures topological properties of persisting data at multiple scales. The team further visualized the magnetization reversal process in two-dimensional space using principal component analysis, a data analysis procedure that summarizes large datasets with smaller “summary indices” facilitating better visualization and analysis. prof. As Kotsugi explains, “Topological data analysis can be used to explain the complex magnetization reversal process and to quantitatively evaluate the stability of the magnetic field structure.” The team discovered that small changes in structure invisible to the human eye could be detected by this analysis, indicating a latent feature that dominates metastable/stable reversal processes. They also successfully determined the reason for the branching of the macroscopic reversal in the original microscopic magnetic field structure.
The novelty of this research lies in the ability to freely link magnetic field microstructures and macroscopic magnetic functions across hierarchies by applying the latest mathematical advances in topology and machine learning. This enables the detection of subtle microscopic changes and subsequently the prediction of stable/meta-stable states that were hitherto impossible. prof. “This super hierarchical and descriptive analysis will improve the reliability of spintronic devices and our understanding of the stochastic/deterministic magnetization inverse phenomenon,” says Kotsugi.
Interestingly, the new algorithm with superior annotation ability can also be applied to study chaotic phenomena such as the butterfly effect. On the technological front, next-generation magnetic memory could potentially improve the reliability of writing and help develop new hardware for next-generation devices.
materials provided by Tokyo University of Science. Note: Content can be edited for style and length.