Influence of bias voltage history on conductance properties of YBaCuO/normal metal junctions

Inelastic electron tunneling across magnetically active interfaces in cuprate and manganite heterostructures modified by electromigration processes

“…Then the difference in V * can be ascribed to the fact that in the lanthanum-based cuprate the Cu-O bond is weaker than the Mn-O bond in LCMO. It should be noted that for a parent yttrium-based cuprate (YBCO), where the Cu-O bond is known to be stronger than in LSCO, the bias range where the electromigration effect can be detected is, as a rule, essentially larger [7]. The picture proposed explains also why the activation voltages V * needed to stimulate the transition from one branch of the conductance spectra to another have increased with decreasing temperature (Fig.1).…”

“…The values of V * were different: near 0.8 V for LCMO and about 0.5 V for LBCO. Following previous works on the same subject [6,7], we explain abrupt changes of the point-contact re- sistance at activation voltages by modifications of the oxygen stoichiometry near the intrinsic metaloxide surface under the influence of applied electric field. Then the difference in V * can be ascribed to the fact that in the lanthanum-based cuprate the Cu-O bond is weaker than the Mn-O bond in LCMO.…”

“…In manganite-and cuprate-based contacts the overall dependence of s( ) V does not have the form (1) but rather follows the formula [7] | |…”

“…We relate it to chemical composition changes in the near-interface region that are usually attributed to oxygen ion displacement processes at the cuprate surface, leading, in particular, to strong suppression of superconductivity in the upper layers (see the review [9]). The most important thing for our purposes is the correlation between the strength of electromigration processes in LBCO (compared with YBCO and LCMO) and the quasi-linear over- all behavior of the differential conductance, which is known to serve as a fingerprint of the inelastic-tunneling effect [7]. Next, we consider the validity of the assumption of an oxygen-deficient near-interface layer created after application of corresponding high biases to the metallic injector and the existence of a magnetically active interlayer between the two electrodes that strongly influences electron tunneling processes across the transition region.…”

“…Next, we consider the validity of the assumption of an oxygen-deficient near-interface layer created after application of corresponding high biases to the metallic injector and the existence of a magnetically active interlayer between the two electrodes that strongly influences electron tunneling processes across the transition region. It is believed [7] that the layer appears because of the removal of oxygen in the upper cuprate layers, which modifies the doping of CuO 2 planes and can thus lead to long-range antiferromagnetic correlations between localized Cu spins. An argument for such a supposition comes from the paper [7], where it was shown that the anomalous b| | V contribution to the conductance spectra of a normal metal/YBCO junction vanishes near the expected Neel ¢ temperature T N .…”