A real-space pseudopotential method for magnetocrystalline anisotropy energies and the search for magnets without rare-earth metals*

We present a real-space pseudopotential formalism for calculating magnetocrystalline anisotropy energies within relativistic density-functional theory (DFT). Our method is implemented in our real-space pseudopotential DFT code, PARSEC, which is designed to run efficiently on massively parallel computing platforms [1]. We show that our formalism works well for prototypical transition-metal compounds, such as YCo5 and Mn2Ga, yielding an accurate magnetization and a magnetocrystalline anisotropy constant consistent with other density-functional methods. We illustrate how our methods can be applied to the search for permanent magnets without rare-earth metals [2,3,4]. In particular, we identified several ZrCo5 and Fe–Ni–B compounds as possible candidate materials that can provide high magnetocrystalline anisotropy energies along with sufficient saturation magnetization.

References:
[1] Phys. Rev. Materials 2, 084411 (2018).
[2] Phys. Rev. Materials 2, 084410 (2018).
[3] Phys. Rev. Materials 3, 044402 (2019).
[4] Phys. Rev. Materials, under review.