Brillouin light scattering (BLS) spectroscopy is the primary key technique in our laboratory to investigate the dynamic properties of magnetic materials and devices. It is based on the interaction of photons with the fundamental excitations of a solid such as magnons, the quanta of magnetic excitations. The interaction can be understood as an inelastic scattering process of the incident photons with magnons, taking into account energy and momentum conservation as indicated in Fig. 1. This technique is sensitive for incoherent, thermal magnons as well as for externally excited, coherent excitations.

The detection of the inelastically scattered photons, i.e. the separation from the elastically scattered photons and the determination of the transferred energy, requires an interferometric technique with extremely high contrast and sensitivity. In our laboratory we implemented the (3+3) Tandem-Fabry-Pérot-Interferometer, designed by John R. Sandercock and schematically shown in Fig. 2. It consists of two Fabry-Pérot interferometers (FPI), each one passed three times by the inelastically scattered light. This approach results in a contrast better than 10¹⁰ for the separation of the elastically and inelastically scattered photons in a frequency range from 500 MHz up to 1 THz.

In the last decades we made significant progress in the improvement of BLS spectroscopy. The spatial resolution was pushed to near the fundamental limit of classical optics by constructing a BLS microscope (Fig. 3) with sophisticated active stabilization methods. In this way, spin-wave transport phenomena can be investigated in micro-structures including time- and phase resolved measurements.

However, in the BLS microscope, the wavevector resolution is lost since the angle of light incidence is not well defined anymore. Thus, wavevector resolved measurements like the direct determination of the dispersion relation are conducted with conventional BLS which has a typical spatial resolution of about 20 μm. During the past two years, wavevector resolution in conventional BLS geometry was significantly improved and pushed to its limits by utilization of novel scattering geometry (Fig. 4). It provides the possibility to achieve wave-vector resolution δk ≈ 0.2 rad/μm with the maximal wavevector limited only by the wave number of the used probing laser light (for 532 nm laser k_max = 4π/λ_laser = 23 rad/μm). Moreover, the described setup enables for an easy change in the magnetization angle of the sample from in-plane to completely out-of-plane geometries as well as the full control of the bias magnetic field due to the utilization of a conventional high-field electromagnet. Another significant improvement here is the utilization of a dielectric mirror to enhance the reflectivity of the sample surface. It provides a significant increase in sensitivity without disturbing the magnetic boundary conditions of the sample. Also this technique provides a constant BLS sensitivity and wavevector resolution regardless of the used excitation geometry.

In a further development, our BLS equipment has been upgraded towards the investigation of new material systems. Integrating a blue laser with a wavelength of 491 nm into our micro-focus BLS setups opened up access to the dynamics of thin YIG films (with thicknesses of 100 nm and below) by largely increasing the sensitivity since the effective scattering cross section is much higher at this wavelength.

A further advancement regarding our standard BLS techniques is the pulsed probe laser beam technique allowing for the detection of various evolution processes such as, e.g., the temperature evolution inside a sample. The special feature of this technique is that the focused laser beam combines the role of the magnon probe with the role of a local sample heater. The heating time is adjusted using amplitude modulation of the probing laser beam by an acousto-optic modulator (AOM). The scheme of the corresponding experimental setup, which consists of a YIG film sample, a microwave circuit, and a BLS laser system, is shown in Fig. 5.