Realizing topological stability of magnetic helices in exchange-coupled multilayers for all-spin-based system

www.nature.com/scientificreports OPEN received: 10 February 2016 accepted: 01 September 2016 Published: 28 September 2016 Realizing topological stability of magnetic helices in exchangecoupled multilayers for all-spinbased system Sergej Fust1, Saumya Mukherjee2, Neelima Paul3, Jochen Stahn2, Wolfgang Kreuzpaintner1, Peter Böni1 & Amitesh Paul1 Topologically stabilized spin configurations like helices in the form of planar domain walls (DWs) or vortex-like structures with magnetic functionalities are more often a theoretical prediction rather than experimental realization. In this paper we report on the exchange coupling and helical phase characteristics within Dy-Fe multilayers. The magnetic hysteresis loops with temperature show an exchange bias field of around 1.0 kOe at 10 K. Polarized neutron reflectivity reveal (i) ferrimagnetic alignment of the layers at low fields forming twisted magnetic helices and a more complicated but stable continuous helical arrangement at higher fields (ii) direct evidence of helices in the form of planar 2π-DWs within both layers of Fe and Dy. The helices within the Fe layers are topologically stabilized by the reasonably strong induced in-plane magnetocrystalline anisotropy of Dy and the exchange coupling at the Fe-Dy interfaces. The helices in Dy are plausibly reminiscent of the helical ordering at higher temperatures induced by the field history and interfacial strain. Stability of the helical order even at large fields have resulted in an effective modulation of the periodicity of the spin-density like waves and subsequent increase in storage energy. This opens broad perspectives for future scientific and technological applications in increasing the energy density for systems in the field of all-spin-based engineering which has the potential for energy-storing elements on nanometer length scales. Information processing via the spin degree of freedom can be subjected to internal interactions such as exchange, Ruderman-Kittel-Kasuya-Yosida (RKKY) or long-range dipolar interactions1,2. Involving magnetic anisotropy on top of it, these interactions can be used to topologically stabilize spin configurations like spin helices or vortices3. The chiral helical structures in rare-earths are essentially due to spatially modulated magnetic states in systems with competing exchange interactions. Fundamentally, the helical ground states here are not realized by the Dzyaloshinskii-Moriya (DM) interaction induced by the spin-orbit scattering of electrons in an inversion asymmetric crystal field (e.g. in non-centrosymmetric MnSi or strain induced DM in centrosymmetric crystals) and thus can be manipulated without an electric field or can be implemented in all-spin-based technology. However, creation of such magnetic helices which are often manifested as 2π planar domain walls (DWs) consisting of different chirality4, with stable magnetic properties, has remained an experimental challenge. Theoretically, Vedmedenko et al. and Dzemiantsova et al. have shown the usages and possibility of nano-sized stable helices for magnetic energy storage2,5. In another example, the magnetization process of a two-dimensional random anisotropy system was shown numerically to be directly connected with topologically stable 2π DWs, with vortices at each end6. In this regard, there have been recent growth of interest in metal-rare earth (TM-RE) systems7–10. RE elements such as Gd, Sm, Dy and Tb have been used regularly to form ferrimagnetic alloys with ferromagnetic (FM) elements, which when coupled antiferromagnetically (or ferromagnetically) to alloys show a reasonable positive (or negative) exchange bias field Heb11–14. Recently, it was shown that an antiferromagnetic (AF)-coupling at a 1 Technische Universität München, Physik-Department, Lehrstuhl für Neutronenstreuung, James-Franck-Straße 1, D-85748 Garching b. München, Germany. 2Laboratory for Neutron Scattering and Imaging, Paul Scherrer Institut, CH-5232 Villigen, Switzerland. 3Technische Universität München, Heinz Maier-Leibnitz Zentrum (MLZ) Lichtenberg Straße 1, D-85748 Garching b. München, Germany. Correspondence and requests for materials should be addressed to A.P. (email: ) Scientific Reports | 6:33986 | DOI: 10.1038/srep33986 1 www.nature.com/scientificreports/ Figure 1. XRR (Cu-Kα) patterns of (a) [Dy6/Fe6]10 and (b) [Dy12/Fe6]10 multilayers are plotted with varying Qz at room temperature. The simulations have been shifted in intensity by two orders of magnitude for clarity. pure TM-RE (Fe-Tb) interface helps in forming planar domain walls which remain frozen upon cooling. The maximum exchange bias field in such systems is determined by the energy it takes to form a (π or 2π) planar domain wall (DW) in the soft layer. Formation of 2π-DWs within a multilayer in helical form can form a double hysteresis loop (DHL) with exchange-bias-like shifts along and opposite to the field cooling axis below the ordering temperature of the RE10. Such a multilayer consists of a mixture of regions containing left-handed DW or right-handed DW in the form of a helix leading to DHL. For the AF-coupled individual layers of Fe and Tb, the possible formation of 2π-DWs within the Fe layers, which is blocked by the anisotropic Tb layers on both sides of Fe, was attributed to be the origin of the exchange bias in such systems. Magnetization reversal of a material with no external magnetic field at a compensation point is the key to manipulate magnetic devices15. This can be done either by electric fields or by optical switching of magnetization using femto-or picosecond pulsed lasers. Very recently, it has been shown that manipulation of the domain walls within a ferrimagnetic TM-RE multilayer (antiferromagnetically aligned TM and RE moment configuration through the multilayer stack) can also be caused upon dilution of the TM layers with non magnetic species16 coupled to the RE layers in the stack. In Tb, Dy and Ho there is a presence of large orbital momentum which leads to a strong spin-orbit coupling and larger magnetic anisotropy. The large difference in spin–orbit coupling in different RE elements has significant influence on the demagnetization processes as well. The difference between Tb and Dy, for example, lies in the temperature range where the RE show an AF helical order. While for Tb this range is only around 220 K, the range where Dy exhibits this helical magnetization ranges approximately from 80–180 K. With one of the highest intrinsic magnetic moments (10.6 μB/atom), Dy exhibits a rich magnetic phase diagram, including a few modulated magnetic phases. Aided by the RKKY interaction, the magnetic modulations propagate coherently over a long range, even with intervening nonmagnetic layers17. However, few experiments exists searching for new phases in Dy when interleaved with a ferromagnetic elements like Fe. Thus it is interesting to confirm the coupling and explore the expected formation of planar DWs in Dy (...truncated)