Machine learning for the development of new materials for a magnetic tunnel junction

Abstract In materials science, we have been increasing the number of constituent elements in an alloy and compounds to improve their properties. For example, in magnetism and spintronics, ternary alloys, such as NdFeB and CoFeB have been developed and widely used in permanent magnets and memories/se...

Full description

Saved in:
Bibliographic Details
Main Authors: Atsufumi Hirohata, Hiroki Koizumi, Tufan Roy, Masahito Tsujikawa, Shigemi Mizukami, Kenji Nawa, Masafumi Shirai
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:npj Spintronics
Online Access:https://doi.org/10.1038/s44306-025-00094-z
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849238229443674112
author Atsufumi Hirohata
Hiroki Koizumi
Tufan Roy
Masahito Tsujikawa
Shigemi Mizukami
Kenji Nawa
Masafumi Shirai
author_facet Atsufumi Hirohata
Hiroki Koizumi
Tufan Roy
Masahito Tsujikawa
Shigemi Mizukami
Kenji Nawa
Masafumi Shirai
author_sort Atsufumi Hirohata
collection DOAJ
description Abstract In materials science, we have been increasing the number of constituent elements in an alloy and compounds to improve their properties. For example, in magnetism and spintronics, ternary alloys, such as NdFeB and CoFeB have been developed and widely used in permanent magnets and memories/sensors, respectively. It has now been considered to be a time to add more elements to further explore their horizon. For such a complicated development, a manual systematic study is no longer practical, leading to the utilisation of machine learning to predict a candidate. These candidates can then be additionally screened by ab initio calculations before experimental confirmation, which can be performed routinely. Additional use of quantum annealing may also broaden the adoptability of machine learning on the materials development. In this perspective, we plan to offer a standardised process for such a development with some requirements for improvement.
format Article
id doaj-art-6df2cdd8de984f09b95ba6227124ef2d
institution Kabale University
issn 2948-2119
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series npj Spintronics
spelling doaj-art-6df2cdd8de984f09b95ba6227124ef2d2025-08-20T04:01:42ZengNature Portfolionpj Spintronics2948-21192025-07-01311910.1038/s44306-025-00094-zMachine learning for the development of new materials for a magnetic tunnel junctionAtsufumi Hirohata0Hiroki Koizumi1Tufan Roy2Masahito Tsujikawa3Shigemi Mizukami4Kenji Nawa5Masafumi Shirai6Center for Science and Innovation in Spintronics, Tohoku UniversityCenter for Science and Innovation in Spintronics, Tohoku UniversityCenter for Science and Innovation in Spintronics, Tohoku UniversityResearch Institute of Electrical Communication, Tohoku UniversityCenter for Science and Innovation in Spintronics, Tohoku UniversityGraduate School of Engineering, Mie UniversityCenter for Science and Innovation in Spintronics, Tohoku UniversityAbstract In materials science, we have been increasing the number of constituent elements in an alloy and compounds to improve their properties. For example, in magnetism and spintronics, ternary alloys, such as NdFeB and CoFeB have been developed and widely used in permanent magnets and memories/sensors, respectively. It has now been considered to be a time to add more elements to further explore their horizon. For such a complicated development, a manual systematic study is no longer practical, leading to the utilisation of machine learning to predict a candidate. These candidates can then be additionally screened by ab initio calculations before experimental confirmation, which can be performed routinely. Additional use of quantum annealing may also broaden the adoptability of machine learning on the materials development. In this perspective, we plan to offer a standardised process for such a development with some requirements for improvement.https://doi.org/10.1038/s44306-025-00094-z
spellingShingle Atsufumi Hirohata
Hiroki Koizumi
Tufan Roy
Masahito Tsujikawa
Shigemi Mizukami
Kenji Nawa
Masafumi Shirai
Machine learning for the development of new materials for a magnetic tunnel junction
npj Spintronics
title Machine learning for the development of new materials for a magnetic tunnel junction
title_full Machine learning for the development of new materials for a magnetic tunnel junction
title_fullStr Machine learning for the development of new materials for a magnetic tunnel junction
title_full_unstemmed Machine learning for the development of new materials for a magnetic tunnel junction
title_short Machine learning for the development of new materials for a magnetic tunnel junction
title_sort machine learning for the development of new materials for a magnetic tunnel junction
url https://doi.org/10.1038/s44306-025-00094-z
work_keys_str_mv AT atsufumihirohata machinelearningforthedevelopmentofnewmaterialsforamagnetictunneljunction
AT hirokikoizumi machinelearningforthedevelopmentofnewmaterialsforamagnetictunneljunction
AT tufanroy machinelearningforthedevelopmentofnewmaterialsforamagnetictunneljunction
AT masahitotsujikawa machinelearningforthedevelopmentofnewmaterialsforamagnetictunneljunction
AT shigemimizukami machinelearningforthedevelopmentofnewmaterialsforamagnetictunneljunction
AT kenjinawa machinelearningforthedevelopmentofnewmaterialsforamagnetictunneljunction
AT masafumishirai machinelearningforthedevelopmentofnewmaterialsforamagnetictunneljunction