The temperature dependence of giant tunnel magnetoresistance ͑TMR͒ in epitaxial Fe/MgO/Fe magnetic tunnel junctions has been investigated. The resistance in the parallel configuration between the bottom ͑free͒ Fe layer and the top Fe layer, exchange biased by an IrMn antiferromagnetic layer, is nearly independent of the temperature. In contrast, in the antiparallel configuration the resistance increases with decreasing temperature, resulting in an increase in the TMR ratio from 170% at room temperature to 318% at 10 K. The dynamic conductance ͑GЈ = dI / dV͒ in the parallel configuration shows flat bias voltage dependence in the range Ϯ0.4 V, but in the antiparallel configuration it shows typical parabolic behavior as a function of bias voltage. A model, based on the temperature dependence of magnetic disorder in the two electrodes and its effect on the spin-dependent tunneling, is proposed to describe the temperature dependence of the TMR ratio and the resistance, in good agreement with our experimental data.
Well-aligned nanoneedles are found to form in a Fe∕GaAs (001) epitaxial thin film when irradiated with a Ga+ ion beam normal to the sample surface at an ion dose of 3×1017∕cm2 in a focused ion beam. These nanoneedles, with their axes perpendicular to the original sample surface, are composed of GaAs nanorods with Fe nanoislands situated at the top. The GaAs nanorods, formed from the GaAs substrate, are crystalline and in the same crystal orientation as the GaAs substrate. The crystalline nature of the Fe thin film is disrupted by the ion sputtering, becoming nanocrystalline Fe islands. It is the thin Fe layer that is responsible for the formation of the nanoneedles. The ion sputtering of the Fe layer roughens the sample surface due to the uneven sputtering rate of the Fe layer. The removal of the Fe layer exposes the substrate to the incident ions, which then sputter the substrate leading to the formation of the nanoneedles.
Low-voltage spin-dependent tunneling spectroscopy of an epitaxial Fe/MgO/Fe magnetic tunnel junction is measured and compared to first-principles calculation of the tunneling conductance. The measured dynamic conductance ͑dI / dV͒ in the parallel configuration shows distinct asymmetric features as a function of the bias voltage. The peaks are independent of barrier thickness, magnetic field, and temperature. With the help of the first-principles calculations, positive and negative bias spectra can be related to different types of Fe/MgO interfaces.The discovery of tunneling magnetoresistance ͑TMR͒ ͑Refs. 1-3͒ and giant TMR ͑Refs. 4-11͒ in magnetic tunnel junctions ͑MTJs͒ leads to an intense interest in the spindependent tunneling spectroscopy. 12-17 By measuring the first and second derivatives of the I-V curve of a tunnel junction, it is hoped that features such as interface resonant states 18 can be revealed. In tunneling spectroscopy, measured dI / dV curves of common tunnel junctions are rather featureless. Peaks in d 2 I / dV 2 are used to identify inelastic processes in tunneling, commonly called inelastic electron-tunneling ͑IET͒ spectroscopy. Few experiments have been able to produce dI / dV curve with identifiable features that can be compared to theory.A MTJ has two states that are switched by a magnetic field. One is an antiparallel ͑AP͒ configuration where the magnetizations of two electrodes are opposite to each other and the other a parallel ͑P͒ configuration where the magnetizations are aligned. Exception for one or two works, 19 most dI / dV measurements for both configurations are featureless. However, the IET spectra ͑d 2 I / dV 2 ͒ for two configurations are very different. There are usually multiple peaks in the d 2 I / dV 2 of the P configuration, which are generally identified with the electron-magnon-and electron-phonon-scattering processes. For the AP configuration, the dominant features are the peaks in d 2 I / dV 2 due to the logarithmic singularity of the zero-bias anomaly, also attributed to the electron-magnon scattering. 20 Due to the lack of features in the measured dynamic conductance, spin-dependent tunneling spectroscopy has not been able to provide much information on interface structures of MTJs despite the realization that such measurements should be sensitive to the changes at the interfaces. 21 In this paper, we report measurements of the dynamic conductance and IET spectrum in fully epitaxial Fe/MgO/Fe MTJs. By carefully eliminating most of the junction defects that can smear out features in the spectroscopy measurement, clear features emerge in the dynamic conductance ͑dI / dV − V͒ of the P configuration. The peaks in dI / dV are matched to the peaks in the transmission probability calculated from first principles under ͑nonself-consistent͒ finite biases for three types of Fe/MgO interface structure, symmetry Fe/ MgO/Fe junction, asymmetric Fe/MgO/vacancy/Fe, and Fe/ FeO/MgO/Fe junctions. Furthermore, features in d 2 I / dV 2 can be identified from the peaks in dI / dV whi...
We present a detailed structural characterization of the interfaces in Fe/MgO/Fe layers grown by molecularbeam epitaxy using aberration-corrected transmission electron microscopy ͑TEM͒, scanning TEM, and electron energy-loss spectroscopy. When fabricated into magnetic tunnel junctions, these epitaxial devices exhibit large tunnel magnetoresistance ratios ͑e.g., 318% at 10 K͒, though still considerably lower than the values predicted theoretically. The reason for this discrepancy is being debated and has been attributed to the structure of, and defects at the interface, namely, the relative position of the atoms, interface oxidation, strain, and structural asymmetry of the interfaces. In this structural study, we observed that Fe is bound to O at the interfaces. The interfaces are semicoherent and mostly sharp with a minor degree of oxidation. A comparison of the two interfaces shows that the top MgO/Fe interface is rougher.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.