Electric Field Induced Magnetic Anisotropy in a Ferromagnet

We report the first observation of a transient all electric field induced magnetic anisotropy in a thin film metallic ferromagnet. We generate the anisotropy with a strong ($\sim {10}^9\mathrm{V}/\mathrm{m}$) and short (70 fs) $\overrightarrow{E}$-field pulse. This field is large enough to distort the valence charge distribution in the metal, yet its duration is too brief to change the atomic positions. This pure electronic structure alteration of the sample generates a new type of transient anisotropy axis and strongly influences the magnetization dynamics. The successful creation of such an anisotropy opens the possibility for all $\overrightarrow{E}$-field induced magnetization reversal in thin metallic films—a greatly desired yet unachieved process.

Figure 1

(a) A Gaussian shaped electron bunch of 28 GeV and indicated standard (rms) deviations traverses a thin metallic ferromagnetic film perpendicular to the surface. The magnetization $\overrightarrow{M}$ is initially set uniformly as shown. (b) The $\overrightarrow{E}$ and $\overrightarrow{B}$ fields are confined to a flat disk perpendicular to the beam and lie entirely in the $x\mathrm{\text{−}}y$ plane. (c) Plot of the maximum temporal $E$- and $B$-field amplitudes versus distance $r$ from the beam center using the experimental longitudinal, $\tau =70\mathrm{fs}$, and transverse, ${\sigma }_r=20\mu \mathrm{m}$, rms beam sizes.

Figure 2

Experimental magnetic pattern generated by a single $\tau =70\mathrm{fs}$ electron bunch in the experimental geometry of Fig. 1 for an in-plane uniaxial 10 nm thick ferromagnetic film of ${\mathrm{Co}}_{70}{\mathrm{Fe}}_{30}$ on MgO(110). The pattern was determined by SEMPA [7]. In the light gray regions, $\overrightarrow{M}$ points into the preset direction as shown while in the dark regions $\overrightarrow{M}$ has switched into the opposite direction.

Figure 3

Magnetic patterns calculated with the LLG equation. (a) Assuming that only a $\overrightarrow{B}$ field is inside the sample and that the damping of the precession is constant in time, (b) $\overrightarrow{B}$ field only but damping increases exponentially with time, (c) $\overrightarrow{B}$ and $\overrightarrow{E}$ field are present and the damping increases according to the text. Pattern (c) reproduces the main features of the experiment Fig. 2.