A general approach is presented for the epitaxial growth of rare-earth-transition-metal films and heterostructures via sputtering onto single-crystal substrates suitably coated with buffer layers. The approach permits the phase and orientation of the films to be controlled. In this manner new, nanostructured permanent-magnet film configurations are being explored by interleaving magnetically hard and soft layers to realize "exchange-spring" magnets.
Thin film techniques to fabricate magnetic systems for research purposes are maturing to the point where much-enhanced magnetic properties can be realized in "atomically" engineered nanostructures. In the present examples thin-film permanent-magnet multilayers and superlattices are realized with striking properties of strong "spring" coupling at the interfaces that highlights innovative pathways to new permanent-magnet materials.
Permanent magnet research and development will benefit from artificially nanostructured composites that "atomically" engineer and couple desirable properties into new materials for the future. Model studies using thin films provide guidance to further develop and optimize the processing of high-performance permanent magnets which will have enhanced opportunities to impact and bolster the U.S. economy.
A General Approach to the Epitaxial Growth of Rare-Earth-Transition-Metal Films
E. E. Fullerton, C. H. Sowers, J. Pearson, X. Z. Wu, D. Lederman, and S. D. Bader
Appl. Phys. Lett. 69, 2438 (1996)
Structure and Magnetic Properties of Exchange-spring Sm-Co/Co Superlattices
E. E. Fullerton, J. S. Jiang, C. H. Sowers, J. Pearson, and S. D. Bader
Appl. Phys. Lett. 72, 380, (1998).
Exchange-Spring Behavior in Epitaxial Hard/Soft Magnetic Bilayers
E. E. Fullerton, J. S. Jiang, M. Grimsditch, C. H. Sowers, and S. D. Bader
Phys. Rev. B (submitted).
Magnetic films can be artificially structured on the nanometer-scale into new configurations with enhanced properties. Recent interest has focused on the realization of a new class of novel permanent magnets known as "exchange-spring" magnets. These are systems consisting of nanodispersed `hard' and `soft' ferromagnetic phases, the `hard' phase provides the high coercivity and pins the `soft' phase, which enhances the magnetization. In the present work epitaxial bilayers and superlattices are synthesized and used to investigate the exchange hardening mechanism.1,2 Epitaxy permits the phase and orientation of the films to be controlled.3-5 Systematic interleaving of hard and soft phases into layered structures provides the advantages of nanometer-scale control of the thicknesses, and interfaces that are structurally coherent. This facilitates quantitative studies of basic physical properties, such as the magnetic anisotropy6, and modeling studies2. We have achieved coercivities for single, hard layers in excess of 4 T at room temperature and 7 T at low temperature, and document epitaxial and morphological considerations that influence properties. We use our model bilayer and superlattice systems to explore and extend numerical simulations that describe exchange hardening and realistically estimate ultimate performance based on this mechanism.
First it is useful to provide a brief background on the relatively new concept of exchange-spring magnets. In general, these composite systems utilize a hard phase that includes a rare earth (RE) to enhance the magnetic anisotropy, and a rare-earth-free `soft' phase that is transition-metal (TM) based to enhance the magnetization. This boosts the maximum energy product (BH)max which is the important engineering figure of merit of a permanent magnet. The exchange coupling at the interfaces pins the magnetization of the soft phase to that of the hard phase, making it resist magnetization reversal. Even if an external magnetic field is able to reverse the magnetization of the soft phase, as long as the field strength is not high enough to switch the hard phase, the magnetization of the soft phase will "spring" back in place when the field is removed. (In contrast, a conventional permanent magnet would trace a minor loop and enter a remanent state with reduced magnetization.) Also, spring magnets contain less RE content than occur in single-component, hard RE-TM intermetallics. This lowers materials' cost and improves corrosion resistance.
For many permanent-magnet applications bulk materials are needed. However, bulk-processing techniques invariably result in nanocomposites whose random nature prevents them from achieving the full potential of exchange hardening. The present goal is foremost to explore the basic properties of new interfacial materials. Systematic interleaving of hard and soft phases into layered structures provides an idealized model exchange-spring system for which detailed information about the underlying physics can be obtained. While this is highly beneficial, there are also daunting scientific challenges associated with an epitaxial film approach to the realization of new permanent magnets. RE-TM intermetallics tend to have complex, low-symmetry crystal structures that lead to inherent ambiguities in the phase identification. This is because of the thermodynamic stability of many phases with similar compositions, and the possible existence of polymorphic forms, and even metastable phases. Thus, the problem at hand is substantial, and well matched to the significant impact that advances in this field can make in energy technologies and conservation.
We present our recent advances in the synthesis of epitaxial films of RE-TM intermetallics via d.c. magnetron sputtering.3,4,5 We also allude to our results on the growth of textured films of Nd2Fe14B via ultrahigh vacuum (UHV) evaporation. In both cases, the substrates are coated with a buffer material that serves as a template for the desired epitaxial growth. The functions of the buffer layer include providing strain relief from any lattice mismatch between the film and substrate, and serving as a chemical barrier to interdiffusion. The buffer layer approach is of general applicability in thin-film synthesis, although the present achievement concerns the fabrication of ferromagnets, and, in particular, exchange-spring magnets. The important point is that the epitaxial buffer layer can be selective of a particular phase and composition.
Figure 1: XRD results for Sm-Co films grown simultaneously onto (a) an MgO(100) substrate with a Cr(100) buffer layers, and (b) an MgO(110) substrate with a Cr(211) buffer layer. The inset is a rocking curve scan of the Sm-Co () peak.
We have examined RE-Co and RE-Fe intermetallics grown on MgO single-crystal substrates with either W [Refs. 3,4] or Cr [Ref. 5] buffer layers. We highlight here one representative example of Sm-Co epitaxial films grown using Cr buffer layers. Figure 1 shows x-ray diffraction scans 5 for two films grown simultaneously, side-by-side in the d.c. magnetron sputtering chamber on (a) MgO(100) and (b) MgO(110), which provides Cr(100) and Cr(211) templates,9 respectively. The intermetallics are cosputtered from separate sources with rates adjusted to provide a nominal composition Sm2Co7, which has an hexagonal crystal structure. In each case the x-ray data illustrate that a single orientation is stabilized, consistent with a-axis and b-axis Sm-Co growth, respectively, with hexagonal lattice constants a=5.035 Å and c=4.10 Å. Interestingly, these orientations are the same as those observed for Hcp Co films on Cr(100) and Cr(211), respectively.
Figure 2: Room-temperature Hc values vs. film thickness for Sm-Co() and () films grown on MgO(110) and (100) substrates, respectively. The lines are guides to the eye.
Room-temperature Hc values are shown in Fig. 2 for the two substrate orientations as a function of film thickness. The easy axis of magnetization is in-plane. For example, the 75-Å thick film on MgO(110) has Hc=4.1 T; the 300-Å one has Hc=3.4 T (which increases to 7.3 T at 25 K.) The a-axis films have coercivities that are independent of thickness, while the b-axis films have coercivities that vary with thickness. For these latter films the anisotropy fields also are very high, but relatively insensitive to thickness; estimates yield 20-25 T anisotropy fields based on extrapolations of the hard-axis hysteresis loops. Thus, we turn to the microstructure to understand the Hc trends.10 For the a-axis films the microstructure is dominated by twinning, which yields two orthogonal in-plane easy axes, due to the four-fold symmetry of the Cr(100) buffer layer. This can stabilize a uniformly high Hc, since the density of twin boundaries is insensitive to thickness. However, for the b-axis films there is uniaxial magnetic anisotropy, and a tilted epitaxy with two domains that are offset from the film normal by ±5°, probably to accommodate the 0.5% lattice misfit between the Sm-Co (4.10 Å) and the Cr(110) at 4.08 Å. As the b-axis films thicken, the initial islanding (observed in our atomic force microscopy) that stabilizes the high Hc, gives way to island coalescence and Hc decreases.
Figure 3: Room-temperature maximum energy product of Sm-Co/Co superlattices as a function of the Co layer thickness. The curve is a guide to the eye.
These type of hard magnet films can be incorporated into epitaxial hard/soft bilayer and superlattice configurations to explore the exchange-hardening mechanism.1,2 Figure 3 shows the room-temperature maximum energy product for a series of a-axis growth Sm-Co/Co superlattices. A single layer of Sm-Co hard magnet has a (BH)max value of 11 MGOe. When interleaved with Co layers, (BH)max of the superlattices increases initially with increasing Co thickness, peaks , and then decreases. (BH)max peaks at a value 30% larger than that of a single hard layer. This illustrates the importance of the exchange hardening mechanism.
The magnetic behavior of the exchange-spring superlattice can be exemplified by a bilayer structure for which a numerical model has been developed.2 The inset of Fig. 4 shows the comparison between the modeling and the experimentally measured demagnetization curves for a Sm-Co/Fe bilayer structure. The symbols represent the measured longitudinal (circles) and transverse (triangles) magnetization components. The solid curves represent atomic-level simulations of the spring coupling that utilize anisotropy and exchange constants characteristic of the individual layers, and an interfacial exchange that is intermediate in magnitude. The simulation not only quantitatively describes the salient features (except the width of the hard-layer reversal), but does so for a whole series of curves for different soft-layer thicknesses. The physical property that governs the interplay of the demagnetization with applied field is the energetics of the magnetic domain-walls in the two layers. As the applied field increases, the magnetization of the soft layer rotates increasingly with distance from the hard interface, but is pinned at the interface. The pinned region in the soft layer can be likened to a domain wall; when its energy becomes comparable to that of a domain wall in the hard layer, the wall can sweep through and reverse the hard layer. These results demonstrate that spring magnets with optimal coupling is realized in our epitaxial bilayers, and this represents a general and robust phenomenon.
Figure 4: Calculated (BH)max curves for Sm-Co/Fe bilayers with different layer thicknesses. Inset: Demagnetization curves for a Sm-Co/Fe(200 Å) film compared to the model calculation (solid line). The longitudinal and transverse components of the magnetization are given by the circles and triangles, respectively.
With an understanding of the magnetization process, we can elucidate the behavior of (BH)max in exchange-spring magnets. Shown in Fig. 4 are calculated (BH)max curves for Sm-Co/Fe bilayers with different combinations of layer thicknesses. Also shown as dashed curves is the ideal energy-product (BH)max = (2pMs)2. At low Fe thicknesses (less than the Block wall width in the hard layer), the Fe layer couples rigidly to the hard layer and (BH)max increases as a result of the enhanced saturation magnetization, following the ideal curve. With increasing thickness, the Fe-layer magnetization reverses at lower fields, and (BH)max decreases. The calculation shows that for bilayers with suitably thin constituent layers, (BH)max can even be greater than that of Nd-Fe-B, which is currently the best available permanent magnet material. Thus, our well-characterized epitaxial layered structures permit the modeling and testing of the exchange-hardening principle. Such investigations serve as guidelines for the development of complex nanostructures with optimized properties.
1 E. E. Fullerton, J. S. Jiang, C. H. Sowers, J. E. Pearson, and S. D. Bader, "Structure and Magnetic Properties of Exchange-spring Sm-Co/Co Superlattices", Appl. Phys. Lett. 72, 380, (1998).
2 E. E. Fullerton, J. S. Jiang, M. Grimsditch, C. H. Sowers, and S. D. Bader, "Exchange-Spring Behavior in Epitaxial Hard/Soft Magnetic Bilayers", Phys. Rev. B (submitted).
3 E. E. Fullerton, C. H. Sowers, X. Z. Wu, and S. D. Bader, "Growth of Oriented Rare Earth-Transition Metal Thin Films", IEEE Trans. Magn. 32, 4434 (1996).
4 E. E. Fullerton, C. H. Sowers, J. P. Pearson, X. Z. Wu, D. Lederman, and S. D. Bader, "Structure and Magnetism of Epitaxial Rare-Earth-Transition-Metal Films", J. Appl. Phys. 81, 5637 (1997).
5 E. E. Fullerton, J. S. Jiang, Christine Rehm, C. H. Sowers, S. D. Bader, J. B. Patel, and X. Z. Wu, "High Coercivity, Epitaxial Sm-Co Films with Uniaxial Anisotropy", Appl. Phys. Lett. 71, 1579 (1997).
6 D. J. Keavney, E. E. Fullerton, D. Li, C. H. Sowers, S. D. Bader, K. Goodman, J. G. Tobin, and R. Carr, "Enhanced Co Orbital Moments in Co-Rare-Earth Permanent Magnet Films", Phys. Rev. B 57, 5291 (1998).
7 D. J. Keavney, E. E. Fullerton, J. E. Pearson, and S. D. Bader, "High-coercivity, c-axis oriented Nd2Fe14B Films Grown by Molecular Beam Epitaxy", J. Appl. Phys. 81, 4441 (1997).
8 D. J. Keavney, E. E. Fullerton, J. E. Pearson, and S. D. Bader, "Magnetic Properties of c-axis Textured Nd2Fe14B Thin Films", IEEE Trans. Magn. 32, 4440 (1996).
9 As in the earlier work: E. E. Fullerton, M. J. Conover, J. E. Mattson, C. H. Sowers, and S. D. Bader, "Oscillatory interlayer coupling and giant magnetoresistance in epitaxial Fe/Cr(211) and (100) superlattices", Phys. Rev. B 48, 15755 (1993).
10 Mohamed Benaissa, Kannan Krishnan, Eric E. Fullerton and J. S. Jiang, "Magnetic Anisotropy and Its Microstructural Origin in Epitaxially Grown SmCo Thin Films" J. Appl. Phys. 83, (June 1998).