Unraveling Magnetic Anisotropy Energy in Ferromagnetic Monolayer on Ferroelectric ABO\(_3\) via DFT and Machine Learning

Spin-based devices have attracted attention as an alternative to CMOS-based technology. However, one of the challenges in spintronics devices is reducing the spin-switching energy in ferromagnetic (FM) materials. To address this, we considered ferroelectric (FE) materials, which may affect the magnetic properties of FM materials. We explored various oxide perovskites ABO\(_3\) as FE materials, onto which a Fe monolayer was placed as the FM material. We evaluated the magnetic anisotropy energy (MAE) of the Fe monolayer while varying the polarization of ABO\(_3\). Our analysis showed that the MAE depends on the magnetic dipole moment induced in the FE material at the interface between the FE and FM materials due to structural modifications. Machine learning techniques were also employed to identify universal behaviors of the MAE in the presence of FE layers, confirming the importance of magnetic moments near the interface in explaining the dependence of the MAE on FE materials.