-
13
-
-
84927462524
-
-
We define the MR as (Rmax-Rsat)/ Rsat, where Rmax is the maximum resistance at zero field and Rsat the resistance value at saturation of the giant MR effect.
-
-
-
-
18
-
-
84927508252
-
-
Ferromagnetic coupling is indicated by the much smaller coercive field than that found for an uncoupled sample (with large Cu thickness). This coercive field is somewhat temperature dependent, as can be seen from Fig. 1(b).
-
-
-
-
23
-
-
84927508303
-
-
A minimum resistivity of 17 μ Ω cm and a maximum resistivity of 44 μ Ω cm was found for a 150times(1.5 nm Co + 0.9 nm Cu) sample made with superconducting contact leads. For our pillar structures we find that ρsup *min = 45pm7 μ Ω ,cm and ρsup *max = 83pm12 μ Ω ,cm. However, we expect the minimum 'intrinsic' resistivity for both fabrication methods to be comparable. The observation that not only ρsup *min, but also ρsup *max has increased approximately by the same factor indicates that there is no major serial contact resistance involved in our samples, but that the effective contact area on top of the pillar is smaller than expected. The origin for this most probably is a small contamination during the Au top contact fabrication on the pillar (Ref. onlineciteGijs2).
-
-
-
-
24
-
-
84927508066
-
-
According to Ref. onlineciteValet, the low temperature multilayer resistance per unit surface for spin uparrow( downarrow) can be written as R uparrow(downarrow ) = M {2 ρF tF [1-(+) β) + 2 ρN tN + 4 rb (1 -(+) γ] }, with M the number of bilayers in the multilayer, ρF and ρN the spin averaged resistivities of the ferromagnetic and the normal metal, respectively, rb the spin averaged interface resistance per unit surface, tF and tN the thickness of the ferromagnetic and the normal metal layer, respectively, and β and γ the important parameters characterizing the spin dependent bulk and interface scattering, respectively. Due to the small experimental values of tF and tN, spin dependent interface resistance is expected to be dominant (Ref. onlineciteValet).
-
-
-
-
25
-
-
84927480882
-
-
The experimental data have been recalculated for a completely AF coupled sample. Without this procedure, values for the fitting parameters would be α = 5.5 and μ = 10.1.
-
-
-
-
29
-
-
84927487616
-
-
Applying the same analysis to Fe/Cr magnetic multilayers, where the CPP MR strongly decreases with increasing temperature (see Ref. onlineciteGijs2), we find a value for μ = 2.2.
-
-
-
|
