We have observed oscillations with non-Magnetic spacer layer thickness in the saturation fields and characteristic exchange fields of CO/Cu Multilayers grown by MBE. The Specimens consisted of a set of samples in which the thickness of copper layers took a series of values between 5Å and 20Å, while the cobalt layer thickness was set at 15Å throughout. The Multilayers were grown epitaxially on GaAs (110) with Ge, Co (110) and Au in the buffer region, resulting in CO/Cu bilayers in the (111) orientation. Although oscillations were observed in these magnetic measurements, no oscillations were observed in the magnetoresistance itself. These observations are somewhat at odds with similar measurements made on samples prepared by sputtering. We conclude with a discussion involving exchange coupling, in-plane anisotropy and defects in the magnetic structure in an attempt to account for these conflicting results.

We have fabricated pillar-like microstructures of high vacuum sputtered Fe/Cr Magnetic Multilayers and measured the giant magnetoresistance effect in the configuration where the measuring current is perpendicular to the film plane from 4.2 K to 300 K. At 4.2 K we find a magnetoresistance of 108 % for multilayers with a Fe thickness of 3 nm and a Cr thickness of 1 nm. The pronounced temperature dependence of the perpendicular magnetoresistance is studied for samples with different Cr thicknesses and tentatively explained by electron-Magnon scattering. The low-temperature data are compared with existing low-temperáture models.

The Boltzmann equation is solved for a system consisting of alternating ferromagnetic -normal metallic layers. The in-plane conductance of the film is calculated for two configurations: successive ferromagnetic layers aligned (i) parallel and (ii) antiparallel to each other. The results explain the giant negative magnetoresistance encountered in these systems when an initial antiparallel arrangement is changed into a parallel configuration by application of an external magnetic field. The calculation depends on (A) geometric parameters (the thicknesses of the layers); (B) intrinsic metal parameters (number of conduction electrons, Magnetization and effective masses in the layers); (C) bulk sample properties (conductivity relaxation times); and (D) interface scattering properties (diffuse scattering versus potential scattering at the interfaces). It is found that a large negative magnetoresistance requires, in general, considerable asymmetry in the interface scattering for the two spin orientations. All qualitative features of the experiments are reproduced. Quantitative agreement can be achieved with sensible values of the parameters. The effect can be conceptually explained based on considerations of phase-space availability for an electron of a given spin orientation as it travels through the multilayer sample in the various configurations and traverses the interfaces.

We present evidence for a new type of giant magnetoresistance in (111) cobalt-copper superlattices with atomically smooth interfaces. We propose that the lowered dimensionality of the structure leads to an enhancement of the scattering of conduction electrons from paramagnetic interfaces obeying a Langevin-like saturation at very high fields, well beyond the switching field of the Co layers. The findings help to explain similarities in magnetotransport behavior with recently reported granular systems as well as differences with antiferromagnetically coupled Multilayers.

The Magnetoresistance of MBE deposited Co/Cu (111) superlattices as a function of Cu layer thickness is studied at various growth temperatures. Wedged superlattices are fabricated with uniform Co layers but with Cu layers whose thickness varies with position along the length of the substrate. Cu thickness is varied over the range 7–75Å. Only one Maximum M Magnetoresistance (MR) is observed for a Cu layer thickness ∼ 10Å for all growth temperatures studied (0°, 150°, 200°C). The optimal magnetoresistance values are achieved with 150°C growth temperature.

The Magnetoresistance Measured perpendicular to the plane of the Multilayer, (CPP-MR) has been measured for the Cu/CO and Cu/ (Ni/Fe) systems. The predictions of a two spin-channel model are summarized, and the Cu/CO data are analysed in terms of this theory. The Cu/ (NiFe) data show a more complex behaviour.

We discuss the effect of the superlattice potential on the magnetotransport properties of magnetic multilayers for current parallel and perpendicular to the plane of the layers. While quantum well states affect the Magnetotransport, they are not the primary origin of the giant Magnetoresistance observed in these materials for currents in the plane of the layers. In general, it is necessary to include both spin-dependent scattering and the effects of superlattice potentials in order to explain the magnetoresistance of multilayered structures.

The Layer Korringa Kohn Rostoker-Coherent Potential Approximation technique was used to calculate the low temperature Giant Magnetoresistance from first principles for Co|Cu and permalloy|Cu superlattices. Our calculations predict large giant Magnetoresis-tance ratios for Co|Cu and extremely large ratios for permalloy|Cu for current perpendicular to the layers. Mechanisms such as spin-orbit coupling which mix spin channels are expected to greatly reduce the GMR effect for permalloy|Cu.

Recent advances in molecular beam epitaxy have renewed research on the physics of artificially structured magnetic superlattices. In particular, there has been much theoretical research on the propagation of magnetic spin waves or magnetic polaritons in magnetic superlattices.1 In this work, we have studied the effect of modulating both the period of an antiferromagnetic/non-Magnetic semi-infinite superlattice and the relative thickness of its individual layers to see how the dispersion relationships co (k) for bulk and surface magnetic polaritons are effected. We have also calculated the effect of an external magnetic field on (u (k) and our calculation goes beyond the magnetostatic approximation by taking retardation effects into account.

A Model for the field dependence of giant Magnetoresistance (GMR) in ‘granular’ co-sputtered alloy thin films (based on a relatively simple spin-dependent scattering concept appropriate to superparamagnetic and weakly ferromagnetic films) is applied to new experimental data from the Co90Fe10-Ag system. The Model and the experimental data can be shown to compare very well with the help of a single adjustable parameter related to spin correlation of adjacent Co-Fe clusters. A careful fit of field-dependent MR data and theory leads to a fairly reliable determination of spin-cluster radius. An analysis of the relative permeability of granular GMR films derived from the generalized form of the Clausius-Mossoti relationship is also presented. For a non-Magnetic Matrix the effective relative permeability is shown to be materials independent. The permeability model is applied to Co-Au granular films.

Giant Magnetoresistance (GMR) is reported in as-deposited Ag1−xCox (x = 0.26–0.53) films co-sputtered on Si from separate Ag and Co targets. GMR ratios (10 K Oe Maximum field) exceeding 0.50 and 0.19 at 5 and 295 K, respectively, are observed for the Ag067Co033 films deposited at ≃28 to 175 °C. The Maximum ratios of 0.55 and 0.24 occur at a substrate temperature of ≃125 °C for these films. The ratios decrease rapidly for the films deposited at temperatures > 175 °C and reduce to ≃0.15 and 0.04 at deposition temperatures >300 °C. This deposition temperature dependence of GMR ratios is interpreted in terms of the change in the spin-dependent interfacial electron scatterings due to the change in the size and number of ferromagnetic Co particles within the electron mean free path. The initial increase and the subsequent decrease in GMR ratios with increasing deposition temperature are attributed to the increase in the mean free path, and the Co and Ag particle size, respectively. Changes in mean free path are obtained from the resistivities of these films while changes in Ag and Co particles are deduced mainly from the X-ray diffraction patterns, transmission electron Micrographs, and the coercivities of these films.

Although Much has been published on giant Magnetoresistance (GMR) in co-deposited thin films [1–4], only little [5] has been published on the structure-property relationships limiting the effect. Here, we report the results of microstructural characterization of NiFeAg thin films that exhibit a GMR effect. The as-deposited films show a sizeable GMR effect. The Maximum GMR effect observed was 6.4% with -4k0e FWHM of the 6P/P peak. Upon annealing these films, the GMR at first increases, and then decreases. We present microstructural evidence from TEM and XRD, amongst other techniques, which shows that this is a consequence of the initial NiFeAg thin film agglomerating into NiFe grains in a predominantly Ag segregant Matrix. Upon extended annealing, excessive grain growth leads to a decrease in the GMR as predicted by the model of Berkowitz, et al. [1].

The role of the interlayer Fermi surface in driving oscillatory exchange coupling of ferromagnetic layers has been critically tested in a series of MBE-grown coherent epitaxial sandwiches prepared on single-crystal substrates. The wedge-shaped Cu (or Cu-based) interlayers have been modified in a number of different ways, achieving orientational, compositional and structural alteration of the Cu Fermi surface. Observed phenomena include multiperiodic coupling oscillations in both the ferromagnetically and antiferromagnetically coupled regions of (lOO) -oriented Co/Cu/Co samples, extended oscillatory periods across CuNi alloy interlayers, and extremely short-period oscillations in a bcc (100) Fe/Cu/Fe system. These phenomena are straightforwardly explained in terms of calculated Fermi surface variations.

Co-Monolayers, prepared by MBE on Cu (111) -surfaces at room temperature and covered by Cu, are ferromagnetic with a Curie-temperature of about 430 K. They are magnetized perpendicularly because of a strong perpendicular magnetic surface anisotropy of the Cu/Co (111) -interface. They provide a remarkably good representation of the 2-dimensional Ising Model. The indirect coupling between these perpendicularly magnetized ferromagnetic monolayers was investigated using samples of type Cu (111) /lCo/DçuCu/lCo/Cu, containing Co/Cu/Co-trilayers composed of Co-Monolayers and a spacer consisting of DCu atomic layers of Cu (111). Torsion oscillation magnetometry of these samples showed clearly a coupling between the monolayers with an oscillatory dependence on DCu. The amplitude of the oscillation is strongly reduced if the coupled Co-films consist of 5 ML instead of 1 M.L. The present controversy on the presence or absence of antiferromagnetic and oscillatory indirect coupling in the Co/Cu (111) -system is discussed in the light of these experiments. The discussion shows that the oscillatory coupling is an intrinsic property of ideal (111)-structures, and can be understood by the RKKY-type theory of indirect coupling between ferromagnetic Monolayers. The usual application of this theory to the coupling between thicker films is justified. However, in the fcc (111) -system there is apparently a specific barrier against complete coalescence, resulting in a tendency to retain holes and channels in the Cu-spacer. This tendency is stronger in flat single-crystal samples than in sputtered films with high densities of atomic steps. Apparently, this results in competing ferromagnetic hole coupling which may more or less completely obscure the intrinsic oscillatory coupling, preferentially in samples grown on extremely flat single crystal surfaces.

The conditions for an almost perfect growth of smooth Cr (001) films on an iron whisker substrate have been investigated by means of reflection high energy electron diffraction (RHEED). The exchange interaction between 20 Monolayer thick Fe (001) films separated from a bulk whisker Fe (001) substrate by a variable number of Cr (001) Monolayers (ML) has been investigated by means of Brillouin light scattering experiments (BLS). These experiments show unambiguously that the exchange coupling strength between the iron film and the iron whisker can be described by a short wavelength oscillatory term superposed on a slowly varying antiferromagnetic background. The BLS data enabled one to separate the bilinear and the biquadratic contributions to the antiferromagnetic exchange coupling terms. Both the bilinear and the biquadratic coupling strengths exhibited a short period oscillatory dependence on the Cr interlayer thickness (∼2 Monolayers). Maxima in the bilinear antiferromagnetic coupling strength occur for an odd number of Cr Monolayers. This observation is not in agreement with first principles calculations. The first phase inversion has been found to occur between 4 and 5 ML of Cr.

We investigated the magnetization behavior of [(Co/Pd) 4-Ru]x samples with perpendicular anisotropy and a limited number of repetitions (x = 1,‥,22). In these systems the Co/Pd Multilayers behave as single magnetic entities. A detailed analysis and comparison of the magnetization curves observed by MOKE and VSM permits us to observe the magnetization reversal and hysteresis of the individual layers and to determine the antiferromagnetic coupling J between each pair of layers. A gradual increase in J is observed in all samples when going from the bottom layer to the top layer. Magnetoresistance curves show the same sharp transitions as the magnetization curves. A clear distinction can be made between an outer layer and an inner layer. These results will be compared with model calculations.

We present small angle neutron scattering experiments on (001) Fe/Ir superlattices. A short period oscillated coupling is observed. However, antiferro and ferromagnetic peaks are simultaneously measured on neutron diagram, which cannot be simply explained. A biquadratic coupling model is proposed to account for this behaviour.

We have searched for the electronic states that mediate oscillatory magnetic coupling in superlattices, and have found strong evidence that these are quantum well states, which are created by quantizing the momentum of s,p-band states perpendicular to the interfaces. The quantum well picture also explains how quantum well states in noble metals become spin-polarized, due to a spin-dependent electron reflectivity at the interface with the ferromagnet. The resulting implications for magnetoresistance are discussed.

We report on the Magneto-optical Kerr rotation (<φ>K) spectra of ultrathin Fe films on Au or Ag (100) substrates and the φK oscillation due to interlayer thickness in Fe/Au/Fe sandwich films. In 3.5–4.5 eV, a new φK peak appears in the bcc-Fe (100) ultrathin films on the fcc-Au (100) surface and it shifts towards the higher energy side with increasing Fe layer thickness. The absolute value of eXy for 3Å (2ML) thick Fe layers is twice as large as that of bulk Fe at 3.7 eV. The thickness dependence of the transition energy of this new peak in the spectra is well explained by the concept of quantum well states in the Fe ultrathin layers, attributing the new transition to a transition from the majority spin Δ5 band ({px±i py), {dxz±i dyZ}; M=±l) to the Δ1 quantum well states (s, pz, dz2; M=0). The new peak is also observed in the Fe/Au (100) artificial superlattices. Using the εxy obtained experimentally for the Fe ultrathin films and the εxy of literature, we can reproduce the experimental φK spectra of the artificial superlattices by optical calculation. On the other hand, we cannot observe the same behavior for the ultrathin Fe films grown on a fcc-Ag (100) surface and covered by a Au (100) ultrathin film, although the εXy of Fe is different from that of the bulk and shows some structures in 2–3 eV. These structures around 2.5 eV are thought to be due to polarized Au atoms adjacent to an Fe layer.

An oscillation of φK as a function of interlayer thickness, d, was observed in photon energy region between about 2.5 and 3.8 eV for the Fe (6Å) /Au (dÅ) /Fe (6A) sandwiched film. The oscillation period was about 10Å (5ML) of Au. The oscillation is thought to be closely related with a formation of spin polarized quantum well states of Δ1 band in Au layers sandwiched by magnetic layers.

We study the effects of quantum-well states on the calculated RKKY coupling. We find that the bound states of a finite-size potential well of depth V give an added oscillation period of size For the simplest case of a spherical free-electron Fermi surface, thus two periods appear: the original, “fast,” π/kf oscillation, and the quantum-well one The quantum-well contributions have larger amplitude, and are in fact the predominant oscillation. For physically reasonable V (tenths of an eV) this period is around 8–10Å. We discuss evidence for these effects in experimental systems.