First-principles calculation study on large perpendicular magnetic anisotropy by interfacial modulation in the Fe/NiFe2O4 superlattice

Abstract

Magnetic random-access memory (MRAM) is emerging as a pivotal technology for next-generation non-volatile spintronics. At its core lies the magnetic tunnel junction (MTJ). With the advent of the era of artificial intelligence, MTJs are required to have large perpendicular magnetic anisotropy (PMA). Achieving strong PMA is the prerequisite for realizing MRAM devices with high integration density, low power consumption, and long-term data retention. Using first-principles calculations, we investigated the structure and magnetic anisotropy of the Fe/NiFe₂O₄ superlattice. It was found that all Fe/NiFe₂O₄ models exhibited PMA and the most energetically favorable configurations for Fe/NiFe₂O₄ interfaces occurred when the interface O atoms in NiFe₂O₄ were on top of the Fe atoms. Importantly, the Fe/NiFe₂O₄ heterostructure with the Fe–FeO interface showed an interfacial PMA density of up to 0.85 mJ m⁻². Although the interface Fe₃ atoms in NiFe₂O₄ make a relatively significant negative contribution to PMA, the interface Fe₂ atoms in Fe produce a large PMA. The d-orbital-resolved magnetic anisotropy energy of interfacial Fe atoms revealed that, similar to surface Fe, the matrix element differences between the d_z² and d_yz orbitals, as well as the d_x²−y² and d_xy orbitals, make large contributions to the PMA. However, compared with the PMA of surface Fe, the energy difference between d_xz and d_yz for interfacial Fe provides a negative contribution to the PMA, making the PMA of interfacial Fe lower than that of surface Fe. A density of state analysis revealed that the bonding between the Fe and O atoms arises from the increase in the hybridization between the relevant spin orbitals. Our results indicate that the Fe/NiFe₂O₄ heterostructures are promising candidates for achieving large PMA in MTJs.