Hot-Electron-Induced Ultrafast Demagnetization in $\mathrm{Co}/\mathrm{Pt}$ Multilayers

Using specially engineered structures to tailor the optical absorption in a metallic multilayer, we analyze the magnetization dynamics of a $\mathrm{Co}/\mathrm{Pt}$ multilayer buried below a thick Cu layer. We demonstrate that hot electrons alone can very efficiently induce ultrafast demagnetization. Simulations based on hot electron ballistic transport implemented within a microscopic model that accounts for local dissipation of angular momentum nicely reproduce the experimental results, ruling out contribution of pure thermal transport.

Figure 1

(a) Schematic representation of the multilayer structure. (b) Time evolution of the magneto-optical signal measurements for different Cu thicknesses. The laser power was adjusted to keep the maximum demagnetization constant for all samples, allowing for a normalization of the signal.

Figure 2

(top) Cu thickness dependence of the induced time delay in the onset of demagnetization $\mathrm{\Delta }{t}_0=t_0(d)-t_0(d=0)$. The full line is a linear fit to the data. (bottom) Modification of the characteristic demagnetization time $\mathrm{\Delta }\tau (d)=\tau (d)-\tau (d=0)$ as a function of the Cu thickness $d$. The full line is a linear fit to the data.

Figure 3

(a) Power dependence of the induced Kerr rotation variation for different Cu thicknesses $d$. (b) Thickness dependence of the induced Kerr rotation variation for different power. Zone I corresponds to combined direct photon and hot electron excitation. In zone II, only hot electrons are responsible for the demagnetization. Zone III corresponds to a faster attenuation of the demagnetization.

Figure 4

(a) Simulation of ultrafast demagnetization for a laser power of 3.5 mW and different Cu thicknesses. (b) Simulation of the maximum demagnetization as a function of the Cu thickness for different laser power excitations. The hot electron attenuation length was set to 200 nm (see Supplemental Material [30] for details about the parameters used in the simulations).