Explainable AI-based physical theory for designing advanced materials

Explainable AI-based physical theory for designing advanced materials

Explainable AI-based physical theory for designing advanced materials

An image illustrating the extended Landau free-energy model developed by a research team at Tokyo University of Science, which allows causal analysis of magnetization reversal in nanomagnets. Using this model, the team was able to effectively visualize the images of the magnetic domain and succeeded in the reverse design of low-power nanostructures. Credit: Tokyo University of Science Kotsugi Laboratory, Japan.

Microscopic analysis of materials is essential to achieve desirable performance in next-generation nanoelectronic devices, such as low power consumption and high speeds. However, the magnetic materials involved in such devices often exhibit incredibly complex interactions between nanostructures and magnetic domains. This, in turn, makes functional design difficult.

Traditionally, researchers have performed visual analysis of microscopic image data. However, this often makes the interpretation of this data qualitative and highly subjective. What is missing is a causal analysis of the mechanisms underlying complex interactions in nanoscale magnetic materials.

In a recent breakthrough published in Scientific reports, a team of researchers led by Professor Masato Kotsugi of Tokyo University of Science, Japan, has succeeded in automating the interpretation of microscopic image data. They achieved this by using an “extended Landau free energy model” which they developed using a combination of topology, data science and free energy.

The model illustrated the physical mechanism as well as the critical location of the magnetic effect, and proposed an optimal structure for a nanodevice. The model used physics-based features to draw energy landscapes in information space, which could be applied to understand complex nanoscale interactions in a wide variety of materials.

Explainable AI-based physical theory for designing advanced materials

Scatterplot of results of principal component analysis dimensionality reduction. The color represents total energy. The relationship between magnetic domain and total energy is connected in the space of explainable features. Credit: Masato Kotsugi from Tokyo University of Science, Japan.

“Conventional analysis is based on visual inspection of images under a microscope, and relationships with material function are expressed only qualitatively, which is a major bottleneck for material design. Our extended model of Landau’s free energy allows us to identify the physical origin and location of complex phenomena within these materials.This approach overcomes the problem of explainability faced by deep learning, which is like reinventing new laws of physics,” says Professor Kotsugi.

While designing the model, the team used the advanced technique in the fields of topology and data science to extend the Landau free energy model. This led to a model that allowed causal analysis of magnetization reversal in nanomagnets. The team then carried out an automated identification of the physical origin and visualization of the original images of the magnetic domain.

Their results indicate that the demagnetization energy near a defect gives rise to a magnetic effect, which is responsible for the “pinning phenomenon”. Additionally, the team was able to visualize the spatial concentration of energy barriers, a feat that had not been achieved until now. Finally, the team proposed a topologically inverse design of the recording devices and low-power nanostructures.

Explainable AI-based physical theory for designing advanced materials

TUS scientists were able to visualize slight changes in microscopic images and understand mechanisms that are difficult to analyze visually. Moreover, they succeeded in designing upside-down nanostructures with low energy consumption. Credit: Masato Kotsugi from Tokyo University of Science, Japan.

The model proposed in this study is expected to contribute to a wide range of applications in the development of spintronic devices, quantum information technologies and Web 3.

“Our proposed model opens up new possibilities for optimizing magnetic properties for materials engineering. The extended method will finally allow us to clarify ‘why’ and ‘where’ the function of a material is expressed. materials, which rely on visual inspection, can now be quantified to enable accurate functional design,” concludes Professor Kotsugi.

More information:
Causal Analysis and Visualization of Magnetization Reversal Using Extended Landau Free Energy, Scientific reports (2022). DOI: 10.1038/s41598-022-21971-1

Provided by Tokyo University of Science

Quote: AI-Based Explainable Physical Theory for Advanced Materials Design (2022, November 29) Retrieved November 30, 2022, from https://phys.org/news/2022-11-ai-based-physical-theory-advanced -materials.html

This document is subject to copyright. Except for fair use for purposes of private study or research, no part may be reproduced without written permission. The content is provided for information only.

#Explainable #AIbased #physical #theory #designing #advanced #materials

Leave a Comment

Your email address will not be published. Required fields are marked *