Magnetothermopower and magnetoresistance of single Co-Ni/Cu multilayered nanowires

The magnetothermopower and the magnetoresistance of single Co-Ni/Cu multilayered nanowires with various thicknesses of the Cu spacer are investigated. Both kinds of measurement are performed as a function of temperature (50–325 K) and under applied magnetic fields perpendicular to the nanowire axis, with magnitudes up to −15% at room temperature. A linear relation between thermopower $S$ and electrical conductivity σ of the nanowires is found, with the magnetic field as an implicit variable. Combining the linear behavior of the $S$ vs σ relation and the Mott formula, the energy derivative of the resistivity is determined. In order to extract the true nanowire materials parameters from the measured thermopower, a simple model based on the Mott formula is employed to distinguish the individual thermopower contributions of the sample. By assuming that the nondiffusive thermopower contributions of the nanowire can be neglected, it is found that the magnetic-field-induced changes of thermopower and resistivity are equivalent. The emphasis in the present paper is on the comparison of the magnetoresistance and magnetothermopower results and it is found that the same correlation is valid between the two sets of data for all samples, irrespective of the relative importance of the giant magnetoresistance or anisotropic magnetoresistance contributions in the various individual nanowires.

Figure 1

(a) Transmission electron microscope image of a single nanowire (NW) of the 3.5-nm Cu sample with an average bilayer thickness of 8.7 nm. The fringes at the bottom are not related to the bilayered structure, but are rather artifacts that are associated with electron scattering at twin boundaries in the material because of the large nanowire diameter. (b) Magnification of the bilayers and (c) high-resolution transmission electron microscope image of the edge of the same nanowire and the ${\mathrm{SiO}}_2$ shell in the bottom left. (d) Transmission electron microscope image of the 5.2-nm Cu sample.

Figure 2

Plot of the $r_{\mathrm{MR}}\left(H\right)$ curves of the Co-Ni/Cu multilayered nanowires in parallel and perpendicular directions of the magnetic field with respect to the nanowire axis (electrical current direction) at RT for (a) 0.2-nm Cu, (b) 0.8-nm Cu, (c) 0.9-nm Cu, (d) 1.4-nm Cu, (e) 3.5-nm Cu, and (f) 5.2-nm Cu samples. The samples within the dashed black border show AMR-dominated behavior due to pinholes in the nonmagnetic layers. The samples within the red border show significant GMR effects due to continuous bilayers and the sample within the dotted green border shows a dominating GMR effect. The saturation fields are not reached for all of the samples and the actual saturation values are slightly higher.

Figure 3

(a) Scanning electron microscope image of a nanowire and the electrical contact structure. (b) Parallel (probe station setup) and perpendicular (cryostat) resistance measurements at RT and a Seebeck coefficient measurement (cryostat) in a magnetic field for the 3.5-nm Cu sample.

Figure 4

Temperature dependence of the (a) zero-field resistivity, (b) $r_{\mathrm{MR}}$, (c) thermopower, and (d) $r_{\mathrm{MTEP}}$ of the multilayered nanowires. The $r_{\mathrm{MR}}$ and $r_{\mathrm{MTEP}}$ data were measured in perpendicular magnetic fields of 3 T. Also shown are (c) estimated thermopower values for Co/Cu and Ni/Cu multilayer with a layer thickness ratio of 5:1 using the literature bulk data for $S_{\mathrm{Co}},S_{\mathrm{Ni}}$, and $S_{\mathrm{Cu}}$. (b) and (d) Two 3.5-nm Cu samples were measured 21 months apart, which caused a reduction of the effect magnitudes.

Figure 5

(a) Seebeck coefficient versus average temperature times the conductance of the 1.4-nm Cu sample in 25-K steps from 50 to 325 K with the applied magnetic field as an implicit variable. For simplicity, only data for $U_{\mathrm{heater}}=5$ V are shown, which correspond to a $\Delta T$ of 3 K at 25 K to 2 K at 325 K. (b) Energy derivative of the resistivity at the Fermi energy derived from Eq. (5) against the temperature. (c) Offset from Eq. (5) and the absolute literature values of Pt [21, 77].

Figure 6

(a) Absolute thermopower obtained by correcting by the literature values for $S_{\mathrm{Pt}}$ [21, 77], $S+S_{\mathrm{Pt}}^{\mathrm{abs}}$, and (b) the absolute thermopower obtained by correcting by $S_{\mathrm{offset}},S+S_{\mathrm{offset}}$. The dashed lines indicate the estimated absolute thermopower values for Co/Cu and Ni/Cu multilayers with 5:1 layer thicknesses [21, 76, 78].

Figure 7

Temperature dependence of $r_{\mathrm{MR},\inf },r_{\mathrm{MTEP}}$, and $r_{\mathrm{MTP}}$ values, which are corrected under the assumption of $S_{\mathrm{contact}}=S_{\mathrm{offset}}$. The expected relation between $r_{\mathrm{MTP}}$ and $r_{\mathrm{MR},\inf }$ according to Eq. (6) is observed.