Anomalously Damped Heat-Assisted Route for Precessional Magnetization Reversal in an Iron Garnet

A heat-assisted route for subnanosecond magnetic recording is discovered for the dielectric bismuth-substituted yttrium iron garnet, known for possessing small magnetic damping. The experiments and simulations reveal that the route involves nonlinear magnetization precession, triggered by a transient thermal modification of the growth-induced crystalline anisotropy in the presence of a fixed perpendicular magnetic field. The pathway is rendered robust by the damping becoming anomalously large during the switching process. Subnanosecond deterministic magnetization reversal was achieved within just one-half of a precessional period, and this mechanism should be possible to implement in any material with suitably engineered dissimilar thermal derivatives of magnetization and anisotropy.

Figure 1

(a) The static temperature dependence of the out-of-plane magnetocrystalline anisotropy field $H_{\mathrm{ani}}$ and the magnetization $M$ (both indicated in the inset) characteristic of bismuth-substituted yttrium iron garnet. (b) In the ground magnetic state, the magnetization $\mathit{M}$ lies parallel to the effective magnetic field, which is given by the combination of $H_{\mathrm{ani}}$ and the in-plane bias magnetic field $H_B$. (c) The anisotropy field is temporarily reduced to ${H^{\prime }}_{\mathrm{ani}}$ by a thermal load, triggering large-amplitude precession of the magnetization about a new transient effective magnetic field (dominated by $H_B$). (d) After sufficient thermal dissipation and large-amplitude precession, the magnetocrystalline anisotropy recovers its original strength, rendering the magnetization switched and stabilized.

Figure 2

(a) Snapshots of the numerically calculated spatial distributions of the out-of-plane component of magnetization, taken 3.5 ns after the instantaneous modification of the magnetocrystalline anisotropy field. The images shown were obtained by varying the strength of the in-plane bias magnetic field $H_B$ and the maximum temperature change $\mathrm{\Delta }{T}_{\max }$ as indicated. (b) Typical raw snapshots of the out-of-plane component of magnetization, acquired 3.5 ns after exposure to the ultrashort optical pulse. These snapshots were obtained with the optical fluence and $H_B$ as indicated.

Figure 3

Background-corrected time-resolved snapshots of the magnetization after exposure to the optical pulse are shown, obtained when the ground-state magnetization initially had positive projection on to the sample normal (light gray region). The snapshots are presented in order of increasing time (relative to the arrival of the pumping optical pulse at time zero) as indicated and were acquired with a pump fluence of $0.37\mathrm{J}{\mathrm{cm}}^{-2}$ and in-plane magnetic field $H_B=5.9\mathrm{kOe}$.

Figure 4

The raw pump-probe signals (of comparable amplitude) are plotted for different pump fluences as indicated. From the results of imaging, we identify three distinct regimes, whereby the signals are collected when no switching occurs (red curve) and when a single spot (black curve) and ring domain (blue curve) of reversed magnetization is formed. All measurements shown here were obtained when $H_B=5.9\mathrm{kOe}$ was applied within the plane of the film.