Elsevier

Thin Solid Films

Volume 520, Issue 6, 1 January 2012, Pages 2184-2189
Thin Solid Films

Tuning exchange spring effects in FePt/Fe(Co) magnetic bilayers

https://doi.org/10.1016/j.tsf.2011.10.143Get rights and content

Abstract

Structural and magnetic properties of exchange spring magnets consisting of hard magnetic (FePt) and soft magnetic (Fe and Co) bilayers, prepared by ion beam sputtering method are studied via X-ray diffraction (XRD), magneto-optic Kerr effect (MOKE) and vibrating sample magnetometry (VSM). Thin tracer layers of 57Fe were introduced in the soft layer in order to observe the Fe spin structure and interfacial diffusion by Conversion Electron Mössbauer Spectroscopy (CEMS). The observed in-plane exchange spring behavior extends also to the magnetic hard layer, whose switching field can be tuned in an unexpected manner via the top soft magnetic layer. To explain the observed phenomenon it is suggested that the increased switching field, found in the system with a Co/Fe bilayer acting as a single soft magnetic layer, is compatible with a peculiar behavior of the stiffness coefficient of the heterogeneous soft magnetic layer. According to this observation, possibilities to maximize the exchange spring effects via suitably chosen non-homogeneous soft magnetic layers are open.

Introduction

Nano-composite materials consisting of exchange coupled soft magnetic (SM) and hard magnetic (HM) phases, known as exchange spring magnets, are the focus of current research because of various technological applications, such as permanent magnets with high energy products [1] or nano-scale magnets for ultrahigh density recording [2]. For example, in the case of information storage, such composite systems combine the high thermal stability of the stored information due to the hard magnetic component with the moderate switching field due to the soft magnetic component.

Composite systems of type HM/SM, where the hard magnetic layer is L10–FePt and the soft magnetic layer is Fe, are being studied both theoretically and experimentally for their exchange spring behavior [[3], [4], [5], [6], [7], [8]]. It may be noted that usually, the as-prepared equiatomic FePt alloy has chemically disordered face-centered-cubic (fcc) symmetry. Upon annealing at about 500–600 °C, it transforms into a face-centered-tetragonal (fct) structure, known as L10, with very high magnetocrystalline anisotropy [9]. In case of thin films, the fct structure can be grown, depending on the deposition conditions, either as an epitaxial phase with high perpendicular to plane anisotropy or as a polycrystalline phase with different textures. Hence, from the magnetic point of view, a soft magnetic film grown on a hard magnetic FePt layer of Ll0 structure may develop an exchange spring structure with either out-of-plane or in-plane behavior.

Depending on the thickness of the soft magnetic layer, various magnetic regimes may occur in such composite systems. For a small thickness of the soft layer, the system behaves as a rigid magnet switching at a single field whereas for a higher thickness of the soft layer, the system behaves as an exchange spring magnet [4], [5]. Goto et al. [10] have also studied the magnetization reversal process in a SM film, ferromagnetically coupled to a HM layer with uniaxial in-plane anisotropy, under the assumption that the hard layer is perfectly rigid and the soft layer has no anisotropy. Upon decreasing the applied magnetic field from positive saturation, the spins of thick enough soft layers (t > 50 nm), exhibit continuous in-plane rotations, with the rotation angle increasing with distance from the hard layer. They always come back along the initial direction (positive magnetization) after releasing negative in-plane applied fields smaller than an exchange field, which is dependent on geometric and magnetic parameters of only the soft layer. For a much thinner soft layer (approaching the domain wall width in the hard layer), the soft layer is expected to be magnetically rigid coupled to the hard layer, switching together at a value of switching (or nucleation) field, which depends on the geometric and magnetic characteristics of both the soft and hard layer. In real situations, the soft layer may have a finite anisotropy and an intermediate thickness between the two situations described above and the hysteresis loop may show both the exchange and the switching (nucleation) fields [11]. Hence, the exchange field Hex in the low-field range reflects the reversible magnetization reversal of the soft Fe layer, whereas the switching field Hirr in the high-field region is indicative for the irreversible switching of the hard layer. Larger are the exchange and switching fields, better are the exchange spring characteristics of the overall system. It is worth mentioning at this point the fundamental difficulty in assigning the exchange and the switching field in real systems presenting relatively thicker soft layers and finite anisotropy of the hard magnetic layer. It has been previously proved [12], [13] that the reversible twist of the spins in the soft layer is not pinned rigidly at the interface but rather propagate significantly into the hard magnetic layer. Therefore, the reversible magnetization is partially stored also in the hard magnetic layer and the usual procedure of using the magnetization reversal curves (hysteresis loops) for defining the exchange field (as related to the steeper decrease of the magnetization at low negative fields) and the switching field (as related to the steeper decrease of the magnetization at high negative fields), most probably fails. A distinction should be made between the exchange field (according to the above definition) and the field corresponding to the steeper decrease of the magnetization at low negative field, which will be called in the fallowing as the switching field of the soft magnetic layer. Accordingly, the last one can be seen as a sort of coercive field assigned to the soft layer (which is not free, but exchange coupled to the hard one). As physical meaning, it would correspond to a continuous rotation of the spins in the soft layer, as in a Bloch wall, with interfacial spins along the magnetization of the hard layer and the most far away spins, in the opposite direction (namely along the negative applied field). In such situations, it becomes very interesting to study both the exchange and the switching fields, with the mutual influence of the two interfacing layers one to each other, and especially the less expected influence of the soft magnetic layer on the hard magnetic one. However, the available classes of systems which may allow such a study is quite limited and request the possibility to tune the magnetic properties (e.g. magnetic hardness) of the both layers. In addition, the quality of the interface may also be decisive in establishing the magnetic properties of the bilayer structure. For example, it has been shown that diffused interface instead of sharp one between the hard magnetic (SmCo) and soft magnetic (Fe) phase improves the exchange spring behavior of the system [14].

In the present work, we propose a study of the mutual influence of the interfacial layers in HM(FePt)/SM(Fe,Co/Fe) systems, with relatively thick soft magnetic layers, in order to induce the exchange spring behavior. The main advantage of the FePt HM layer consists in the possibility to tune its hardness via the processing conditions, as discussed in the following. In addition, the anisotropy of the SM component might be also tuned in the proposed system, while the Co layer presents a different anisotropy as compared to the Fe one. It is worth to mention the continuous change of the crystal structure of the equiatomic FePt phase via thermal treatments, from the chemically disordered A1 fcc phase to the chemically ordered Ll0 fct phase, which can be monitored by the shift of the (111) peak and the evolution of the (001) and (110) peaks in the X-ray diffraction patterns (e.g. [15]). As expected, this evolution is accompanied by a change of the hardness of the FePt overall phase, which behaves magnetically as a single component with a coercive field, μ0HC, increasing from 0.02 T (for mostly chemically disordered fcc phase) to above 0.3 T (for a better formed chemically ordered fct phase) [15].

Section snippets

Experimental details

The studied thin films were prepared on Si(111) substrate by using an ion beam sputtering method. Firstly, the FePt film of 30 nm thickness was grown from a compound target. The compound target was obtained by placing Fe and Pt targets side by side. Relative areas of Fe and Pt targets exposed to the ion-beam were controlled in order to achieve equiatomic composition of the film. The film was in-situ post-deposition annealed at about 500 °C for one hour to induce the L10 phase formation. The

Results and discussion

Fig. 1 shows the GIXRD patterns of the two thin films. As expected from the processing procedure of the samples, they are quite similar with respect to the hard magnetic phase. One can clearly see various peaks corresponding to the ordered L10 FePt. It is worth mentioning the (001) reflection at ~ 24° and the (110) reflection at ~ 32°, the so-called “superlattice” peaks that are the specific signature of the L10 FePt to distinguish it from the cubic A1 FePt phase [[17], [18], [19], [20], [21],

Conclusions

We have reported on the synthesis and characterization of HM(FePt)/SM(Fe and Co/Fe) multilayers presenting in-plane exchange spring behavior. The peculiarities of the interfacial exchange coupling allowed us to tune unexpectedly the exchange spring performances (including the switching field of the magnetically hard layer) only by substituting partially the Fe soft layer by a Co soft layer. The phenomenological explanation for the relationship between the switching field of the hard layer and

Acknowledgments

The work was partly carried out in the frame of the bilateral Indo-Romania project funded by DST (New Delhi, INT/Romania/P-07/07) and the Romanian Ministry of Research (contract CB317/2009). VRR thanks Mr. Sattish Potdar for help in film deposition and AB thanks DST, New Delhi, for funding PPMS at CSR, Indore. Financial support from Romanian Ministry of Research through PN II project nos. , is also gratefully acknowledged.

References (27)

  • J.M.D. Coey

    Solid State Commun.

    (1997)
  • R. Rohlsberger et al.

    J. Magn. Magn. Mater.

    (2004)
  • D. Goll et al.

    Physica B

    (2008)
  • A.D. Crisan et al.

    Mater. Sci. Eng. C: Biomim. Supramol. Syst.

    (2007)
  • A.D. Crisan et al.

    Mater. Sci. Eng. C: Biomim. Supramol. Syst.

    (2007)
  • M. Seqqat et al.

    J. Magn. Magn. Mater.

    (1996)
  • J.P. Liu et al.

    Appl. Phys. Lett.

    (1998)
  • R.H. Victoria et al.

    IEEE Trans. Magn.

    (2005)
  • F. Casoli et al.

    Appl. Phys. Lett.

    (2008)
  • R. Pellicelli et al.

    Phys. Rev. B

    (2008)
  • G. Asti et al.

    Phys. Rev. B

    (2006)
  • D. Goll et al.

    Appl. Phys. Lett.

    (2008)
  • C.P. Luo et al.

    J. Appl. Phys.

    (2000)
    M.R. Visokay et al.

    Appl. Phys. Lett.

    (1995)
  • Cited by (37)

    • Interface magnetization in Y<inf>3</inf>Fe<inf>5</inf>O<inf>12</inf> / Nd<inf>3</inf>Ga<inf>5</inf>O<inf>12</inf> epitaxial heterostructures caused by Fe<sup>3+</sup> - Nd<sup>3+</sup> exchange coupling

      2022, Thin Solid Films
      Citation Excerpt :

      In contrast to YIG / GGG, in YIG / NdGG heterostructures Ha is positive because of the opposite sign of YIG film lattice strain in these heterostructures [14]. One of the most important aspects in the study of thin-film magnetic heterostructures is understanding of structural and magnetic properties of interfaces, since they largely determine the properties of magnetic layers such as, the perfection of their crystal structure, the epitaxial stresses, type of magnetic anisotropy, orientation of the magnetic moment, dynamics of magnetization, etc. [15–23]. To study interfaces, different methods have been used such as synchrotron X-ray diffraction, polarized neutron reflectometry (PNR), scanning electron microscopy (SEM), magneto-optical methods (magneto-optical Kerr effect (MOKE), magnetoabsorption [24–26] in visible light and X-ray ranges (X-ray magnetic linear and circular dichroism (XMCD)).

    • Annealing driven positive and negative exchange bias in Fe–Cu–Pt heterostructures at room temperature

      2020, Journal of Alloys and Compounds
      Citation Excerpt :

      The occurrence of exchange bias phenomenon in these heterostructures on annealing above 300 °C indicates the presence of plausible hard and soft magnetic phases in the system, where Fe and alloy layer (e.g. FePt, FePtCu, FeCu, etc.) act as a soft and hard ferromagnetic layer, respectively. Previous reports also show observation of EB in FePt/Fe systems [35–37]. Temperature-dependent macroscopic magnetization properties of the heterostructures S1 and S2 at different annealing temperatures are shown in Fig. 2 (c).

    View all citing articles on Scopus
    View full text