Vortex-dynamics-mediated low-field magnetization switching in an exchange-coupled system

An aspect of a magnetic vortex whose dynamics strongly affects the magnetic structures of the environment is experimentally shown. We exploit a nanodot of an exchange-coupled bilayer with a soft magnetic $\mathrm{N}{\mathrm{i}}_{81}\mathrm{F}{\mathrm{e}}_{19}$ [permalloy (Py)] having a magnetic vortex and a perpendicularly magnetized $L{1}_0\text{−}\mathrm{FePt}$ exhibiting a large switching field $\left(H_{\mathrm{sw}}\right)$. The vortex dynamics with azimuthal spin waves makes the excess energy accumulate in the Py, which triggers reversed-domain nucleation in $L{1}_0\text{−}\mathrm{FePt}$ at a low magnetic field. Our experimental and numerical results shed light on the essence of reversed-domain nucleation, and provide a route for efficient $H_{\mathrm{sw}}$ reduction.

Figure 1

(a) Schematic illustration of a microfabricated dot of a $L{1}_0-\mathrm{FePt}$ |permalloy (Py) exchange-coupled system. (b) Scanning electron microscope image for the array of dots. (c) Simulated magnetization ($M$) vs the dc magnetic field ($H$), where $M$ was normalized by the saturation value and $H$ was applied along the $z$ direction. (d) Simulated structures of magnetic moments ($m$) of the $x$ component ( $m_{\mathrm{x}}$, upper panel) and the $z$ component ( $m_{\mathrm{z}}$, lower panel) at $H=0\mathrm{Oe}$. (e) and (f) Simulated magnetic structures of $m_{\mathrm{x}}$ in the $y-z$ plane (left panel), $m_{\mathrm{z}}$ in the $y-z$ plane (middle panel), and $m_{\mathrm{x}}$ in the $x-y$ plane (right panel) at $H=2.6$ and 5 kOe.

Figure 2

(a) $M$ as a function of $H$ for microfabricated dots with $L{1}_0-\mathrm{FePt}|\mathrm{Py}$, where the value of $M$ was normalized by the saturation magnetization. (b) Device resistance ($R$) as a function of $H$ without the rf magnetic field $\left(H_{\mathrm{rf}}\right)$. (c) $\mathrm{\Delta }R-H$ curves with $H_{\mathrm{rf}}$ (solid circles) and without $H_{\mathrm{rf}}$ being applied (open circles). $\mathrm{\Delta }R$ is the resistance change from $R$ at $H=-11\mathrm{kOe}$. 22 dBm of rf power was applied to the device, which corresponded to $H_{\mathrm{rf}}=200\mathrm{Oe}$. The frequency ($f$) of $H_{\mathrm{rf}}$ was set at 11 GHz. For both $M-H$ and $R-H$ curves, $H$ was applied perpendicularly to the device plane. The thick arrows denote the switching fields $\left(H_{\mathrm{sw}}\right)$ of $L{1}_0-\mathrm{FePt}$. (d) $H_{\mathrm{sw}}$ (top panel) and $R$ at $H=-11\mathrm{kOe}$ (bottom panel) as a function of $f$. The dotted line denotes $H_{\mathrm{sw}}$ without $H_{\mathrm{rf}}$, and the solid (open) circles represent the experimental (simulated) results.

Figure 3

(a)–(e) Snapshots for $m_{\mathrm{z}}$ at $H=3.4\mathrm{kOe}$ under an application of $H_{\mathrm{rf}}$ with $f=11\mathrm{GHz}$. The top and bottom panels are $x-y$ plane images for Py and FePt, respectively. The planes of the Py and FePt slices are 2 and 5 nm, respectively, away from the FePt|Py interface.

Figure 4

(a)–(e) Snapshots for deviation of $m_{\mathrm{z}}\left(d{m}_{\mathrm{z}}\right)$ for Py at $H=3.4\mathrm{kOe}$ under the application of $H_{\mathrm{rf}}$ with $f=11\mathrm{GHz}$. The plane of the Py slice was 2 nm away from the FePt|Py interface.

Figure 5

(a) $H_{\mathrm{sw}}$ as a function of measurement temperature ($T$) experimentally obtained without $H_{\mathrm{rf}}$. The solid line denotes the result of fitting. (b) $m_{\mathrm{z}}$ at $H=3.4\mathrm{kOe}$ as a function of the position in the dot along the in-plane $y$ direction $\left(D_{\mathrm{y}}\right)$ at various times ($t$). $D_{\mathrm{y}}$ is denoted in the inset. The plane of the FePt slice was 5 nm away from the FePt|Py interface. (c) Top panel: Calculated $t$ dependence of total $m_{\mathrm{z}}\left(m_{\mathrm{z},\mathrm{total}}\right)$ and total energy $\left(E_{\mathrm{total}}\right)$. Since $H$ was set at 3.4 kOe, $m_{\mathrm{z},\mathrm{total}}$ shows a value of $\sim 0.7$ before the reversed-domain nucleation. Middle panel: $t$-dependent energy difference $\left(\mathrm{\Delta }E\right)$ for Zeeman energy $\left(E_{\mathrm{Z}}\right)$, demagnetizing energy $\left(E_{\mathrm{d}}\right)$, exchange energy $\left(E_{\mathrm{ex}}\right)$, and anisotropy energy $\left(E_{\mathrm{ani}}\right)$. Bottom panel: $t$ dependence of $\mathrm{\Delta }{E}_{\mathrm{ex}}$ in Py and $L{1}_0-\mathrm{FePt}$.