A guideline for atomistic design and understanding of ultrahard nanomagnets

Abstract

magnetic nanoparticles are of immense current interest because of their possible use in biomedical and technological applications. Here we demonstrate that the large magnetic anisotropy of FePt nanoparticles can be significantly modified by surface design. We employ X-ray absorption spectroscopy offering an element-specific approach to magnetocrystalline anisotropy and the orbital magnetism. Experimental results on oxide-free FePt nanoparticles embedded in Al are compared with large-scale density functional theory calculations of the geometric-and spin-resolved electronic structure, which only recently have become possible on world-leading supercomputer architectures. The combination of both approaches yields a more detailed understanding that may open new ways for a microscopic design of magnetic nanoparticles and allows us to present three rules to achieve desired magnetic properties. In addition, concrete suggestions of capping materials for FePt nanoparticles are given for tailoring both magnetocrystalline anisotropy and magnetic moments.

Effect of structure on the magnetic anisotropy ofL10FePt nanoparticles

“…Thus, it is not surprising that we find that bi-pyramidal structures with large central Pt layer has significantly higher MCA per atom compared to the cuboctahedral ones. 18,25 For example, we get an MCA per atom value 1.14 meV for the Fe 64 Pt 68 versus 0.86 meV reported in the papers above for the 147 atom FePt cluster reported in Ref [25]. This difference can be traced to the larger percentage of Pt atoms in the center for the bi-pyramidal case compared to the cuboctahedral structure.…”

“…And given the large spin-orbit coupling constant of Pt it is this enhanced orbital moment of Pt that is responsible for the higher anisotropy of the system as a whole. We also found that when the central layer of the bipyramidal cluster is Pt, the cluster has higher MCA than a cluster of the same size but with Fe as the central layer, in contrast to the cuboctahedral cases, 18,25 in which it is the surface layers that play crucial role in high MCA . This stems from the fact that when Pt atoms comprise the central layer they have more Fe atoms neighboring them, so that hybridization increases, lending them higher orbital moments than are possessed by Pt atoms in other layers.…”

“…The explanation for this is that the central layer has more atoms than other layers in the cluster, and when these atoms are of high orbital moment -Pt atoms hybridizing with the Fe atoms below and above -the system as a whole exhibits higher anisotropy than when is central layer consists of Fe atoms, whose orbital moment, in hybridizing with the Pt atoms above and below, is markedly lower. In contrast to the cuboctahedral case 18,25 bi-pyramidal nanoparticles possess (similar magnitude) MCA mostly at the (large) central Pt layer. This fact may eliminate the need to cap them in order to preserve MCA.…”

“…FernandezSeivane and Ferrer 24 studied the correlation of the magnetic anisotropy with the geometric structure and magnetic ordering of small atomic clusters of sizes up to 7 atoms. Gruner et al 18,25 demonstrated that in cuboctahedral nanoparticles the high anisotropy of the layers increases as one moves towards the surface, and the anisotropy can be even enhanced by embedding the material in some suitable other metal (e.g., Au in the case of Pt-terminated structures). Various experimental studies, using X-ray Magnetic Circular Dichroism Spectroscopy (XMCD) or X-ray Absorption spectra (XAS), have recently confirmed the enhancement of MCA in free or supported clusters.…”

“…17 Another way to obtain L1 0 phase experimentally is by annealing alternating multi-layers of Fe and Pt. 11 Stable FePt L1 0 nanoparticles have been prepared experimentally 18 both without any covering and with Al encapsulation.…”