Machine Learning Analysis of Tunnel Magnetoresistance of Magnetic Tunnel Junctions with Disordered MgAl₂O₄

S. Ju^1*, Y. Miura, K. Yamamoto, K. Masuda, K. Uchida, J. Shiomi

¹ Shanghai Jiao Tong University, Shanghai, 201306, China

² National Institute for Materials Science, Tsukuba, 305-0047, Japan

³ The University of Tokyo, Tokyo, 113-8656, Japan

ABSTRACT: Magnetic tunnel junctions (MTJs) are important devices for spintronic applications, such as magnetic random access memories and high-sensitivity magnetic sensors. Through Bayesian optimization and LASSO technique combined with first-principles calculations, we investigated the tunnel magnetoresistance (TMR) effect of Fe/disordered−MgAl₂O₄(MAO)/Fe(001) MTJs to determine the structures of disordered-MAO that give large TMR ratios. The optimal structure with the largest TMR ratio was obtained by Bayesian optimization with 1728 structural candidates, where the convergence was reached within 300 structure calculations. Characterization of the obtained structures suggested that the in-plane distance between two Al atoms plays an important role in determining the TMR ratio. The TMR ratio tended to be large when the ratio of the number of Al, Mg, and vacancies in the [001] plane was 2:1:1, indicating that the control of Al atomic positions is essential to enhancing the TMR ratio in MTJs with disordered MAO. The present work reveals the effectiveness and advantage of materials informatics in designing high-performance spintronic devices based on MTJs.

 Figure 1.  Designing magnetic tunnel junctions via Bayesian optimization.

Keywords:  Machine learning; Magnetic tunnel junctions; High tunnel magnetoresistance ratio; Structure optimization