Insertion layer magnetism detection and analysis using transverse magneto-optical Kerr effect (T-MOKE) ellipsometry

Abstract

This experimental study demonstrates that with transverse magneto-optical Kerr effect (T-MOKE) ellipsometry, it is possible to determine the magneto-optical and magnetic properties of insertion layers, even if they are superimposed onto much larger magnetic signals from the surrounding structure. Hereby, it turns out to be especially valuable that with T-MOKE ellipsometry one has full and precise quantitative access to the complex value of the magneto-optical reflection matrix component β, because small magneto-optical insertion layer signals do not necessarily increase the absolute size of β, but can lead to observable phase changes of this complex quantity instead. We demonstrate the ability of T-MOKE ellipsometry to precisely detect such small effects and hereby allow for an accurate determination of the alloy concentration dependent onset of ferromagnetism in ultrathin Co x Ru 1−x insertion layers, that are embedded into a much thicker ferromagnetic structure. In addition, a detailed and quantitative signal analysis allowed us to demonstrate that the Co x Ru 1−x insertion layers in our samples exhibit a magnetization reversal behavior that is independent of the adjacent Y 3 Fe 5 O 12 layers, clearly indicating that both magnetic entities are either not or only very weakly coupled.

Sensitivity and reproducibility of transverse magneto-optical Kerr effect (T-MOKE) ellipsometry

“…Fundamentally, the βiω values can be defined for each iω contribution as βiω = β • m iω , with m iω being the normalized magnetization reversal contribution at frequency iω. This relation in between the magneto-optical frequency components and the magnetization dynamics is valid due to the fact that the m vs. time t evolution exhibits time-reversal symmetry under sufficiently large field excitation [17]. Figure 4(b) displays the measurement extracted | βiω | values for ∆I/I and A iω /I, as data points, which were obtained from the least-squares fits represented in figures 3(f)-(j), and which show the expected substantial decrease for higher frequencies.…”

“…So, to understand the noise level and signal-to-noise performance of our T-MOKE ellipsometry measurements, it is meaningful to analyze the noise performance of the individually contributing frequency components. Moreover, for multilayer systems, different Fourier components do not necessarily provide the exact same information but can actually be utilized to distinguish layers [17]. Thus, there is an added value in investigating the Fourier component performance in addition to understanding the compounded performance of ∆I only.…”

“…The experimental grid that is used in any specific measurement sequence is defined by particular combinations of angles θ 2 and ϕ 2 . The grid used so far in this work covers 21 × 21 (θ 2 , ϕ 2 ) equidistant settings in an angular range of ±4 • and thus defines a square grid, which was also used in prior works [17,37]. The square grid has been used for simplicity, but the specific measurement grid is not actually optimized or formally justified.…”

“…MOKE techniques have been demonstrated to be especially useful for materials characterization and the study of magnetization properties in thin films, surfaces, and layered structures [13][14][15]. Furthermore, magneto-optical effects have not only been used as characterization techniques for magnetization properties [16,17] but also for ultra-fast magnetization dynamics detection [18,19] and spintronic process and device measurements [2,20]. Moreover, optical isolation and modulation devices [21,22], magnetoplasmonics [23,24], data storage technologies [3,25] and bio-sensing [26,27], among other fields [28][29][30][31], utilize magneto-optical effects.…”

“…Therefore, the present work explores these crucial experimental aspects and overall, aims to understand the achievable sensitivity of T-MOKE ellipsometry by precisely characterizing the achieved signal and signal-to-noise ratios in detail, and explore the physical origins of the noise levels to understand fundamental and practical limits of this methodology. To achieve these goals, our comprehensive study of T-MOKE ellipsometry is conducted by using three separate experimental approaches, which are each described in their own respective results sections: (a) the best operational point is determined by analyzing the impact of the polarization angle of the incident light; (b) given that different Fourier components of the total measured signal pattern can provide different information [17], a comparative study of their individual sensitivities and signal-to-noise performances is conducted; and (c) a search for an improved experimental measurement sequence has been performed to explore another pathway of obtaining either higher sensitivity and/or reduced measurement times, similar to work conducted towards the optimization of the generalized magneto-optical ellipsometry (GME) method [41]. Moreover, the quality and reproducibility of the T-MOKE ellipsometry data is analyzed in precise detail by repeating the same type of measurement 100 times under identical conditions and quantitatively analyze deviations of each individual measurement from the average.…”