Identification of stable alloy with giant magnetization via machine learning based autonomous materials search

doi:10.21203/rs.3.rs-55136/v1

Download PDF

Article

Identification of stable alloy with giant magnetization via machine learning based autonomous materials search

https://doi.org/10.21203/rs.3.rs-55136/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 19 Mar, 2021

Read the published version in Communications Materials →

Version 1

posted

You are reading this latest preprint version

Discovery of new magnets with high magnetization has always been important in human history because it has given birth to powerful motors and memory devices. Currently, the binary alloy Fe₃Co₁ exhibits the largest magnetization of any stable alloys explained by the Slater-Pauling rule. A multi-element alloy system has been expected to include an alloy with magnetization beyond that of Fe₃Co₁, but it has been difficult to identify appropriate elements and compositions because of combinatorial explosion in the multi-element system. In this work, we found a new alloy with giant magnetization beyond that of Fe₃Co₁ via an autonomous materials search system combining machine learning and ab-initio calculation. After autonomous/automated exploration in the large material space of multi-element alloy for six weeks, the system unexpectedly indicated that Ir and/or Pt impurities would enhance the magnetization of FeCo alloys. To confirm this experimentally, we comprehensively synthesized Fe_xCo_yIr_1-x-y and Fe_xCo_yPt_1-x-y alloys and found that some of them have magnetization beyond that of Fe₃Co₁.

Applied Mathematics

Materials Engineering

Fe3Co1

alloys

magnetization

There is NO Competing Interest.

FigS1.png
The Slater-Pauling curve. A graphical summary of the magnetic moments (∝ magnetization) in a binary alloy system calculated ab-initio using the KKR-CPA method. The Fe75Co25 alloy, which is located at the top of the curve, has the largest magnetic moment.
FigS2.png
Local magnetic moment in Fe99X1 binary system. To limit the material search space, we calculated local magnetic moments of various impurities X in Fe (Fe99X1) using the ab-initio (KKR) method. Positive and negative signs indicate parallel and anti-parallel magnetic moments of element X with respect to Fe local magnetic moment. We identified eight elements (Fe, Co, Ni, Ru, Rh, Pd, Ir and Pt) that have local magnetic moments greater than +0.25 μB.
FigS3.png
Combinatorial MOKE data. Results of combinatorial MOKE experiments. a, b, c, Composition maps of FexCoyIr100-x-y, FexCoyPt100-x-y and FexCoyNi100-x-y composition-spread thin films respectively. d,e,f, MOKE curve data of FexCoyIr100-x-y, FexCoyPt100-x-y and FexCoyNi100-x-y composition-spread thin films, respectively. Inserting Ir and Pt impurities enhances not only magnetization but also coercive force. g,MOKE curves of Fe67.3Co26.3Pt6.4.
FigS4.png
Results of combinatorial XRD experiments. a, b, c, Composition maps of FexCoyIr100-x-y, FexCoyPt100-x-y and FexCoyNi100-x-y composition-spread thin films. d,e,f, XRD curves of FexCoyIr100- x-y, FexCoyPt100-x-y and FexCoyNi100-x-y composition-spread thin films. g, XRD curve of Fe67.3Co26.3Pt6.4. There are no diffraction peaks in the area of 2θ>75°. All XRD curves include a small diffraction peak at 2θ=61.7° fromthe X-ray source (Cu-Kβ). All samples have an Fe (200) peak around 2θ>65° although its intensity decreases with increasing amounts of Ir, Pt and Ni impurities.

Download PDF

Journal Publication

published 19 Mar, 2021

Read the published version in Communications Materials →

Version 1

posted

You are reading this latest preprint version

Identification of stable alloy with giant magnetization via machine learning based autonomous materials search

Status:

Journal Publication

Version 1

Abstract

Figures

Full Text

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1