Strong Bias Effect on Voltage-Driven Torque at Epitaxial Fe-MgO Interface

Torque can be provided to magnetization in nanomagnets directly by electric current and/or voltage. This technique enables electric current (voltage)-to-spin conversion without electromagnetic induction, and has been intensively studied for memory device applications. Among the various kinds of torque, torque induced by spin-orbit splitting has recently been found. However, quantitative understanding of bulk-related torque and interface-related torque is still lacking because of their identical symmetry for current-in-plane devices. In this paper, we propose that a pure interface-related torque can be characterized by spin-torque ferromagnetic resonance with a current-perpendicular-to-plane tunnel junction. Epitaxial Fe-MgO-V tunnel junctions are prepared to characterize the interface-related torque at Fe-MgO. We find that the current-driven torque is negligible, and a significant enhancement of the voltage-driven torque is observed when the MgO barrier thickness decreases. The maximum torque obtained is as large as $2.8\times {10}^{-5}\mathrm{J}/({\mathrm{Vm}}^2)$, which is comparable to the voltage-controlled magnetic anisotropy of $180\mathrm{fJ}/\mathrm{Vm}$. The voltage-driven torque shows strong dc-bias-voltage dependence that cannot be explained by conventional voltage-controlled magnetic anisotropy. Tunnel anisotropic magnetoresistance spectroscopy suggests that the torque is correlated to an interface state at the Fe-MgO. This surface-state-sensitive electric modulation of magnetic properties provides new insight into the field of interface magnetism.

Popular Summary

Sensing and manipulating the magnetic orientation of a nanomagnet is central to technological applications such as efficient microscopic computer memories. One approach relies on spin torque, where electrical currents or voltages consisting of electrons with a predominant spin orientation can induce a torque that flips the orientation of a spin direction in a thin layer of magnetic material. Attempts to measure this torque, however, are hampered by additional torques introduced by measurement devices. We have developed a method to characterize the spin torque at an interface between a magnet and a nonmagnet, and we use this technique to provide insight into spin-torque properties.

We use spin-torque ferromagnetic resonance to analyze the spin-torque vector at an interface between thin layers of iron (Fe) and magnesium oxide (MgO). A current-perpendicular-to-plane tunnel junction allows us to quantitatively characterize the vector components. We find that the current-driven spin torque is negligible and that the perpendicular component of the voltage-driven torque increases as the MgO barrier thickness decreases, reaching a maximum that is larger than any reported voltage-driven torque. A nonlinear dependence on the dc bias voltage indicates that an electronic state in the vicinity of the Fermi level of the system is likely to be of importance.

Fully characterizing spin-torque behavior is critical to future development of nanoscale computer memories. More generally, our observations of the electric modulation of magnetic properties that are sensitive to the surface state also provide insight into interface magnetism.

Figure 1

(a) Schematic illustration of the device structure. (b) In-situ reflection high-energy electron-diffraction images of surfaces of the MgO barrier, Fe, and V buffer.

Figure 2

(a) Device resistance as a function of the MgO barrier thickness ($t_{\mathrm{MgO}}$). (b) Typical TAMR, where $t_{\mathrm{MgO}}$ and resistance are 0.94 nm and $189\mathrm{\Omega }$, respectively. The magnetic-field angle definition (${\theta }_H$) is depicted in Fig. 3. (c) Magnetoresistance ratio as a function of $t_{\mathrm{MgO}}$.

Figure 3

(a) Schematic illustration of the measurement setup. (b) Definition of perpendicular (${\beta }_{\perp }$) and in-plane (${\beta }_{//}$) torque. (c) Typical spectra for spin-torque FMR. The magnetic-field direction is ${\theta }_H=45°$, and the input power ($P_{\mathrm{rf}}$) is $0.1\mu \mathrm{W}$.

Figure 4

(a) Torque vector components. Black and red circles represent perpendicular (${\beta }_{\perp }$) and in-plane (${\beta }_{//}$) torque, respectively. The dashed curve shows the theoretical line of the torque induced by the voltage-controlled magnetic anisotropy of $180\mathrm{fJ}/\mathrm{Vm}$. (b) Input power dependence of peak-to-peak dc voltage in the spin-torque FMR spectra. (c) The dc-bias-voltage dependence of the resonant field. (d) The dc-bias-voltage dependence of the perpendicular torque. All data were obtained with a magnetic-field angle of ${\theta }_H=45°$ and an input microwave frequency of 5 GHz.

Figure 5

(a) The dc-bias-voltage dependence of TAMR. Black circles and red curves are experimental data and fitting curves using Eq. (4), respectively. (b) Twofold component of the TAMR. (c) Fourfold component of the TAMR.