Implementation and application of a real-space pseudopotential method for calculating magnetocrystalline anisotropy

We present a real-space pseudopotential method for calculating magnetocrystalline anisotropy within relativistic density-functional theory. Our formalism is implemented in our real-space pseudopotential code, PARSEC, which is explicitly designed for an efficient implementation on a parallel computing platform (Phys. Rev. Materials 2, 084411 (2018)). We demonstrate 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 previous work. We also use our method to explore possible candidate materials for rare-earth-free permanent magnets. We find that ZrCo₅ compounds can provide moderate magnetocrystalline anisotropy and sufficient work is supported by the National Science Foundation (NSF), DMREF-1729202. HPC resources were provided by the Texas Advanced Computing Center (TACC).saturation magnetization (Phys. Rev. Materials 2, 084410 (2018)).