Figure 1

(a) $8\mu \mathrm{m}\times 8\mu \mathrm{m}$ MFM image of the reversed domain at a given stage of magnetization reversal. The domain walls are pinned on the microtwins (along $⟨110⟩$ axes). (b) Schematic diagram of the top view of a 200 nm wide cross superimposed with a $1\mu \mathrm{m}$ AFM image of the emerging microtwins (taken from a similar sample). The gray scale extends over 10 nm. There are microtwins parallel and perpendicular to the wire. (c) Optical image of the device fabricated by $e$-beam lithography and ion etching (down to the MgO substrate). Two Hall crosses are separated by a 200 nm wide and $10\mu \mathrm{m}$ long wire. Electrical contacts are taken on gold pads deposited above the FePt film (out of the image). (d) Hysteresis loop measured by EHE for the reversal of a 200 nm wide Hall cross by DW motion. The field sweep rate is $0.2\mathrm{T}·{\mathrm{mn}}^{-1}$.Reuse & Permissions

Figure 2

(a) Two independent measurements of the sum of Hall voltages $V_1+V_2$ as a function of time, both realized at the same temperature $T=300\mathrm{K}$, and under the same constant field $H_{\mathrm{appl}}=0.11\mathrm{T}$, and leading to propagation times $t_a$ and $t_b$. (b) Cumulative distribution functions of the propagation time at $T=300\mathrm{K}$ and 200 K, resulting from 400 independent measurements. At 300 K, the curve can be fitted to an exponential law. The uncertainties are around $\pm 0.1\mathrm{s}$ for the time measurement, and $\pm 0.05$ for the values of $F(t)$.Reuse & Permissions

Figure 3

(a) Scheme of the possible reversal states involved in the Markov process, around a barrier. (b) Cumulative distribution functions of the propagation time for different values of the experimental parameters ($H_{\mathrm{appl}}$, $T$). Grey lines correspond to numerical fits to the experimental data using a Markow law [Eq. (5)]. The actual distribution extends far beyond 10 s.Reuse & Permissions

Figure 4

(a) Characteristic times vs applied field and temperature, deduced from the fit of the cumulative distribution function of the propagation time. Stars indicate values that cannot be confidently extracted from the fit. (b) Semilog representation of ${\tau }_{13}$ (direct transition) at 200 and 300 K vs field, with Arrhenius fits.Reuse & Permissions