Speaker
Description
Micro-focused Brillouin light scattering (BLS) spectroscopy is one of the most versatile methods to probe spin-wave dynamics. Typically, magnons are measured by detecting scattered light whose polarization is rotated by 90$^\circ$ relative to the incident beam. However, this conventional approach assumes the presence of only linear magneto-optical coupling. Generally, this assumption is not true, especially in magnetic garnets, where the contribution of the quadratic magneto-optical effect can be significant.
In this study, performed on coherently excited spin waves in a 100 nm-thick bismuth-doped yttrium iron garnet (BiYIG) thin film, we systematically analyzed the polarization dependence of the BLS signal by modifying the angle of both the polarizer and analyzer [1]. By sweeping their angles, we obtained polarization-dependent spectra for all basic spin-wave geometries: backward volume, Damon–Eshbach, and forward volume. To interpret the measurements, we employed a semi-analytical model that combines calculated electric-field distributions with spin-wave dispersion and dynamic magneto-optical susceptibility including both linear (Kerr) and quadratic (Cotton–Mouton) contributions. Fitting the experimental data reveals that the Cotton–Mouton effect in BiYIG is comparable in magnitude to the linear Kerr effect, highlighting the importance of nonlinear contributions in accurately describing magneto-optical interactions.
Furthermore, we resolved the dispute about the feasibility of measuring the forward volume spin waves in BLS. We demonstrated that a high numerical aperture objective significantly modifies the local electric field, introducing non-negligible out-of-plane and transverse components even for nominally linearly polarized input light. This field redistribution enables sensitivity to all components of dynamic magnetization, making micro-focused BLS measurements in the forward volume geometry possible.
[1] K. Szulc et al. arXiv:2602.15760 (2026).