Spin noise spectroscopy from the perspective of scattered light and noise formation principles in variety of systems

Abstract

This work is dedicated to a tool to get information about the spin system ground state with the help of optics: spin noise spectroscopy. Its primary goal is to characterize the virtually undisturbed spin dynamics by obtaining parameters of spin fluctuations in thermal equilibrium.

The homodyne detection scheme with phase stabilization is utilized to improve the responsiveness of the polarization analysis in spin noise spectroscopy. By providing power to the local oscillator, it is possible to overcome the electronic noise and effectively increase the acquired signal - without additional disturbance in the probed spin system while the beam bypasses the sample. This possibility allows the work with lower probing power densities, approaching the desired nonperturbing regime. An improvement larger than half an order of magnitude is present for a bulk n-doped GaAs for small probe intensities. Phase manipulation and stabilization make it achievable to choose the desired parameter - Faraday rotation, ellipticity or a mixture - from the experiment. It avoids otherwise necessary modifications in the arrangement of the optical components.

This improved technique examine further the fundamental characteristics of the spin noise signal construction by obtaining various angular dependencies of the scattered light. The distributed Bragg reflector forming a microcavity amplifies the light-matter interplay, making it possible to detect spin noise from an ensemble of $n$-doped (In, Ga)As/GaAs quantum dots reliably and extract weak effects. As a proof of principle, we performed an observation of pure scattered field (outside of the transmitted light aperture), as well as extraction of the primary electron spin properties, g-factor and spin dephasing time. We also studied the impact of the microcavity on the spatial and spectral dispersion of the scattered light intensity. Additionally, the interplay of two beam resonance excitation was considered for the potential signal enlargement capabilities.

From that point, the interest in the work is shifted in the direction of probing new systems and their characterization. The fundamentals of the spin noise spectroscopy is transcription of the spin-system magnetization on the angle of the Faraday rotation, which should be sufficient to be measurable. The presence of such a sufficiency can not be extracted from linear magneto-optical effects, which is especially crucial in inhomogeneously broadened systems, exhibiting the spin noise gain effect. In this part, the connection between the spin noise gain effect and the behavior of the nonlinear resonant Faraday effect is established, allowing us to predict the applicability of the spin noise spectroscopy to this type of paramagnet. The experimental evidence is based on intraconfigurational (4f-4f) transitions of the trivalent rare-earth ions of neodymium and ytterbium in fluorite-based crystals, approving the theoretical estimations.

At last, spin noise spectroscopy is applied to materials conventionally studied by means of the electronic paramagnetic resonance spectroscopy -- dielectrics with paramagnetic impurities, which were thought of as inapplicable for the spin noise spectroscopy before. Such belief was founded on their low specific Faraday rotation for strong optical transitions. This work demonstrates that for forbidden intraconfigurational transitions, one can see the spin noise spectroscopy due to the spin noise gain effect, which is proportional to a relation of inhomogeneous linewidth to the homogeneous one and can be as high as 10^8 in the mentioned system. The requirements for the optical setup to unlock such measurements are discussed, along with the potential applications. Finally, a method is present to simplify the identification of obtained spectra components. Due to the discovered fact that the sum of squares of the magnetic resonance frequencies stays the same for any direction of the magnetic field in cubic crystals with anisotropic impurity centers, if the magnitude of the field is constant, one can match the peaks to corresponding types of centers. The relation between the invariant and the g-tensor components were derived for various kinds of centers and proofed experimentally with the spin noise spectroscopy on a cubic CaF2-Nd3+ crystal.