Magnetic field effects on total and partial conductance histograms in Cu and Ni nanowires

Abstract

Conductance histograms have become a powerful tool for studying transport properties of metallic nanowires. However, the individual conductance curves display a very rich structure that might be concealed by the statistical procedure of finding preferred conductance values by building conductance occurrence histograms using consecutive nanocontact breakage experiments. This is particularly true when it comes to discerning 1/2G0=e2/hquantization in magnetic nanowires. The effect of disorder, added to possible magnetic sources of scattering, and different magnetic states of different nanowires, might hide its appearance as histogram peaks. This work analyzes and compares Ni and Cu nanowire experimental histograms at room temperature (RT). Those obtained with no curve selection criteria are basically unaffected by the presence of a magnetic field. A selection of particular sets of conductance curves shows that conductance quantization could occur in steps of e2/h and 2e2/h in Ni as well as in Cu in the presence or absence of a magnetic field. Sorting out curves in sets that present conductance plateaus at half integer and integer values, and compiling statistics on the number of such curves that appear, depending on the applied magnetic field, results in differences between the behaviour of Cu and Ni. While for Cu, the magnetic field keeps the ratio of curves that present plateaus at 1/2G0with respect those presenting G0 plateaus unchanged; for Ni, the number of curves which exhibit plateaus at just G0 almost disappears with the applied field. This experimental fact might indicate that the magnetic field removes spin degeneracy in these magnetic nanowires.