Magnetostatic effects in giant magnetoresistive spin‐valve devices (original) (raw)
Related papers
Rotating-field magnetoresistance of exchange-biased spin valves
Journal of Applied Physics, 2005
We investigate the magnetoresistance (MR) of spin valves by (i) varying the strength of the field applied in a fixed direction and (ii) rotating the field with fixed strength. The latter data reflect in general a mixture of giant and anisotropic magnetoresistance (GMR and AMR). We present an experimental procedure to suppress the AMR contributions of all ferromagnetic layers in the spin valve without disturbing the GMR response. The resulting angular MR curves are fitted with a single-domain model to determine with high precision the exchange bias field, the uniaxial anisotropies, the GMR ratio, and the interlayer coupling field. The application of the method to differently prepared Ta͑5.0 nm͒ / NiFe͑3.0 nm͒ / FeMn͑15.5 nm͒ / NiFe͑3.0 nm͒ /Co͑2.0 nm͒ / Cu͑3.5 nm͒ /Co͑2.0 nm͒ / NiFe͑7.0 nm͒ spin valves with GMR ratios of 1.8% and 4% demonstrates the sensitivity and reveals differences of the order of a few percents of the exchange bias field for the uniaxial anisotropy fields of the free and pinned layer as well as for the interlayer coupling field.
High-frequency measurements of spin-valve films and devices (invited)
Journal of Applied Physics, 2003
High-frequency measurements of spin-valve films and devices, made using several different measurement techniques, are presented and compared. Pulsed inductive measurements were made on sheet films and provide insight into the intrinsic dynamical properties of the component films and multilayer stacks. The damping parameter, in the completed spin-valve stack, is larger than in the constituent films. Direct time and frequency domain measurements of the dynamical response of micrometer-size spin-valve devices, made using high-bandwidth magnetoresistance techniques, showed damping parameters comparable to these measured on spin-valve sheet films. The small-angle magnetization response was also determined by high-frequency magnetic noise measurements. The damping parameters were smaller than those obtained by direct susceptibility measurements. The device-level measurements show a different dependence of the damping parameter on the easy-axis field as compared to sheet-level measurements. In addition to the uniform rotation mode, other peaks can be observed in the noise spectra that correspond to fluctuation modes arising from the micromagnetic structure. Electrical device measurements have much greater sensitivity than other high-frequency magnetic measurement techniques, which allow the direct observation of magnetization motion in submicrometer elements without averaging. This technique is used to directly examine thermally activated events and nonrepetitive dynamical motions.
Physical Review B, 2006
We present a theoretical analysis of spin transport and magnetoresistance ͑MR͒ of a penta-layer device, consisting of a basic pseudo-spin-valve trilayer with half-metallic ͑HM͒ insertions. Our analysis is based on two models, i.e., the Fert-Valet equation in the limit of infinite spin relaxation length , and the spin-dependent drift-diffusion equation, with finite . In both cases, we found that MR decreases with increasing HM or ferromagnetic ͑FM͒ layer resistivity, when the spin polarization of the HM or FM material is lower than some critical value. We derived the critical value analytically, and explained the MR effect in terms of the relative contribution of FM and HM layers in providing spin-dependent scattering. More interestingly, the presence of spin relaxation causes MR to be always suppressed at high HM/FM resistivity, irrespective of the spin polarization of the HM/FM layers. This phenomenon may be explained by introducing an effective spin-flip resistance, which provides a less resistive path to minority spin electrons via spin flipping. Finally, we show the validity of our analysis in the presence of realistically long contact leads, provided the lead conductivity is sufficiently high. Thus, the insertion of the HM layers results in an interplay between the resistivities of the HM/FM and spacer layers, and the effects of spin relaxation, leading to a MR behavior which is hitherto undescribed.
Simulating device size effects on magnetization pinning mechanisms in spin valves
Journal of Applied Physics, 1996
The effects of magnetostatic interactions on the giant magnetoresistive ͑GMR͒ response of NiFe/ Cu/NiFe spin valves are studied using an analytical model. The model is applicable to devices small enough for the magnetic layers to exhibit single-domain behavior. Devices having lengths in the track-width direction of 10 m and interlayer separations of 4.5 nm are studied. Stripe heights are varied from 0.5 to 2 m. The magnetization of one magnetic layer is pinned by a transverse pinning field that is varied from 0 to 24 kA/m ͑300 Oe͒. GMR curves for transverse fields are calculated. At zero external field the magnetization of the layers shows a tendency to align themselves antiparallel in the transverse direction. This results in an offset from the ideal biasing of the device. Broadening of the curves due to shape anisotropy occurs with decreasing stripe height and increasing magnetic layer thickness, and the magnetization in the pinned layer becomes less stable.
Enhanced magnetoresistance in lateral spin-valves
Applied Physics Letters, 2010
The effect of feature sizes on the characteristics of lateral spintronic devices have been investigated experimentally and theoretically. It is demonstrated that confining spin-transport in the active region of a device enhances magnitude of the spin-dependent response substantially. Numerical simulation of spin-transport corroborates the experimental observations. Device characteristics are found to be a strong function of spin-polarizer and analyzer dimensions. The response is observed to attain a peak value for an optimum device feature size, and this is seen to be a function of temperature. Spin dependent effects become weaker for very small and very large devices.
IEEE Transactions on Magnetics, 2006
We derived the magnetoresistance (MR) that takes into consideration the effect of spreading resistance (SR) due to the patterned layer in a spin-valve (SV) structure. Our analysis is based on the: 1) spin drift diffusion (SDD) model and 2) finite-element Poisson (FEP) solver. The SDD model does not take into consideration the effect of SR due to patterning, whereas FEP includes the effect of SR. This enables us to compare and analyze the contribution of both patterning and SR to the MR of the device. In a FM-NM-FM structure, the NM layer was patterned into: 1) single and 2) multiple cylindrical structure. We found that the spacer patterning and the resulting SR causes a significant increase (by 50%) of MR at low area ratio (A R) of the patterned layer. Yet, MR of patterned structure is lower compared to the MR of original structure. However, patterning of a FM layer inserted within the NM spacer results in significantly higher MR.
Magnetoresistance of lateral semiconductor spin valves
Journal of Applied Physics, 2010
The magnetoresistance of two terminal lateral semiconductor spin valves with respect to varying mesa size is studied. It is shown theoretically that extended regions outside the spin-current path can act as an additional source of spin-relaxation, decreasing the magnetoresistance response. From a simplified expression of magnetoresistance derived from spin-diffusion equations, we show that it is important to etch away these extended regions for devices with channel lengths much smaller than the spin-diffusion length in order to achieve maximum magnetoresistance. Preliminary experimental data on a two terminal local spin valve are in good agreement with the theory established in this article.
Journal of Magnetism and Magnetic Materials, 2005
We measure the magnetoresistance of Ta/NiFe/FeMn/NiFe/Co(t Co )/Cu/Co(t Co )/NiFe spin-valves as a function of field strength and direction. The exchange bias H EB and the giant magnetoresistance depend on the Co thickness t Co ; and the resistance shows rotational hysteresis when the external field is close to compensating H EB : A single-domain model reproduces the angular dependence of the magnetoresistance in regions without hysteresis, yields H EB values in agreement with MOKE, and allows one to precisely determine the interlayer coupling and uniaxial anisotropies although they are much weaker than H EB : r