Spectroscopic probe of ultrafast magnetization dynamics in the extreme ultraviolet spectral range

TL;DR

This work addresses the need for a simple, intensity-based XUV probe of ultrafast magnetization in systems with out-of-plane anisotropy grown on opaque substrates. It introduces two geometries, L-MOKE and P-MOKE, that leverage Kerr-induced polarization changes together with strong Brewster-angle reflectance anisotropy, allowing analyzer-free measurements with linearly polarized XUV light. Through HHG-based XUV sources, simulations using the udkm1Dsim toolkit, and time-resolved pump-probe experiments on FeNi and Pt/GdCo/Pt structures, the authors map energy- and polarization-dependent magneto-optical signals and demonstrate sizable, linear responses up to about 10% asymmetry. The methods reveal distinct ultrafast demagnetization dynamics, including element-specific delays, and offer a versatile platform for imaging nanoscale magnetic textures on opaque substrates, with broad implications for ultrafast spintronics and magnetic microscopy.

Abstract

The development of spectroscopic techniques in the extreme ultraviolet (XUV) spectral range has significantly advanced the understanding of ultrafast interactions in magnetic systems triggered by optical excitation. In this work, we introduce a previously missing geometry that facilitates the observation of the ultrafast magnetization dynamics of magnetic systems with an out-of-plane magnetization grown on XUV opaque substrates. This approach to probing ultrafast magnetization dynamics combines the magneto-optical Kerr effect with the strong dependence of a sample's reflectance near its Brewster angle. It therefore works with linearly polarized light and does not require any additional polarizing optics. We provide a comprehensive analysis of the technique by presenting both simulations and experimental data as a function of the energy and the polarization of the XUV probe radiation as well as of the delay time after optical excitation.

Figures (6)

• Figure 1: Illustration of the experimental geometries for the (a) longitudinal (L-MOKE), (b) transverse (T-MOKE), and (c) polar (P-MOKE) magneto-optical Kerr effect. The respective alignment of the direction of the magnetization is controlled by an external magnetic field. For all three geometries the incidence angle is $\theta=45°$. For L/P-MOKE the polarization angle $\phi \neq 0$, for T-MOKE we measure at $\phi \approx 0$, i.e. for p-polarized radiation.
• Figure 2: Measured and modeled asymmetry spectra in (a) L-MOKE and (b) T-MOKE geometries for the FeNi sample and in (c) P-MOKE geometry for the Pt/GdCo/Pt sample. The spectra were acquired at polarization angles, $\phi$, of a) 10.0°, (b) 1.4° and (c) 7.0° and at an angle of incidence $\theta=45°$. The dots mark the energies of the odd HHG orders, H33 to H43, for which we have measured polarization- and time-dependent data.
• Figure 3: Measured and simulated magnetic asymmetries as a function of polarization angle $\phi$ for FeNi in (a) L-MOKE and (b) T-MOKE geometry, as well as for (c) Pt/GdCo/Pt in P-MOKE geometry. The data for the FeNi sample is shown for photon energies H37 = 57.3eV and H42 = 65.3eV as well as H35 = 54.2eV and H43 = 66.6eV with a predominant sensitivity to Fe and Ni, respectively. For Pt/GdCo/Pt, we show data at H33 = 51.1eV and H39 = 60.5eV, which probe the magnetization of Co and Pt, respectively. The shaded bands represent the standard error of the measurement.
• Figure 4: (a) Simulations of the reflectance and asymmetry as a function of the polarization angle $\phi$ of the Pt/GdCo/Pt sample in the P-MOKE geometry at a photon energy 60.5eV. The calculation is shown for three different values of the magnetization, $M=0\%, 50\%$ and 100%. Panel (b) focuses on $\phi$ angles close to zero with arrows indicating the location of the minimum reflectance for the different magnetization values. Panel (c) shows the P-MOKE asymmetry as a function of both the incident angle $\theta$ and the polarization angle $\phi$.
• Figure 5: Simulations of the reflectance and asymmetry for the Pt/GdCo/Pt sample as a function of varying magnetization in P-MOKE geometry at $\theta = 45°$ for polarization angles (a) $\phi = 2.0°$ and (b) $\phi = 7.0°$.