Logo Logo
Switch language to English
Lott, Dieter (2001): Magnetic X-ray Reflectivity. Dissertation, LMU München: Fakultät für Physik



The scope of the thesis is to demonstrate the feasibility to examine magnetization profiles of thin films and multilayer systems via magnetic soft and hard x-ray reflectivity. The focus here is on 3d transition metals, which are used mainly for development of numerous noval magnetic devices, that are both technologically and scientifically interesting. Complementary to Neutron diffraction, which is the standard tool for the examination of magnetic structures in matter, magnetic x-ray diffraction permits to study small samples and exhibits better Qz-resolution due its small and only slightly divergent beam. The biggest advantage is its element specificity, which enables one to probe different magnetic sites separately. The method of magnetic x-ray reflectivity combines the strong magnetic circular dichroism (MCD) effect, significantly enhancing the magnetic sensitivity of x-rays, with the technique of conventional specular reflectivity, a well established tool for the structural studies of the chemical makeup of thin films and artificial multilayer systems. The theory of resonant magnetic scattering within dipole approximation combined with the specular reflectivity condition suggests that the strongest effects are in the lower incident angle regime using circularly polarized x-rays. By using soft and hard x-rays structures on a scale of a few to several hundreds of Å are probed, which is the dimensions of the thicknesses of the layers of most thin film and multilayers systems. In order to retrieve quantitative information from the measured magnetic reflectivity curves, an approach for visible light magneto-optical effects based on known dielectric tensors of the sample has been adopted and applied for soft and hard x-ray resonant scattering. Sample absorption and polarization changes in the sample are accounted for. Besides the structural composition, the thickness of the individual layers and the index of refraction, also the magnetic spin configuration can be chosen with arbitrary moment direction and magnitude by modifying the off-diagonal terms in the dielectric tensor. The magnetic optical constants, which determine the magnitude of the magnetic moments, are experimentally determined via MCD absorption measurements and then retrieving the real part through the Kramers-Kronig transformation of the measured imaginary part. This is shown in this work for several 3d transition metals and edges. The simulations are sensitive to a variety of different spin configurations: spiral spin structures, magnetic dead layers and of collinear alignment. Experimentally the magnetic reflectivity of 3d transition metals has to distinguish between the two available possible absorbtion edges, L and K, lying in different x-ray regions. The L-edges are situated in the soft x-ray region and exhibit large enhancements of the magnetic cross section, while the K-edges lie in the hard x-ray regime and show much smaller effects. In spite of this handicap, the latter can be important due to the much larger penetration depth and better Qz-resolution. The X13 beamline at the NSLS at Brookhaven National Laboratory consisting of two branches for soft and hard-x ray operations, respectively, uses an elliptical polarized wiggler (EPW), which produces circularly polarized x-rays in the orbit plane and allows fast switching between left and right circular polarization. Lock-in detection is used to improve the signal-to-noise ratio at the soft x-ray branch and single photon detection at the hard x-ray branch to measure the magnetic signal. The EPW and the experimental setup was commissioned to demonstrate the feasibility of magnetic x-ray experiments. Especially at the hard x-ray beamline branch the small magnetic effects, less than 0.1% of the charge scattering, were possible to detect. In order to satisfy the need for high flux the CMC-CAT beamline at the APS in Argonne was used for magnetic hard x-ray reflectivity, providing an undulator beamline where the high flux of linear polarized photons was converted into circular polarization via a diamond phase plate, delivering much higher flux and better circular polarization. The sample used to demonstrate the feasibility of the method of magnetic reflectivity consists of two multilayer structures of Fe/Cr on top of each other, where the iron spins of the upper are ferromagnetically and of the lower antiferromagnetically coupled, representing an exchange bias system. The sample was characterized with conventional x-ray reflectivity and MOKE measurements in order to accurately determine the structural composition and magnetic configuration (hysteresis loops), respectively. Magnetic reflectivity experiments on the L-edges at the X13A beamline showed strong magnetic effects, which could be clearly identified as ferromagnetic and antiferromagnetic Bragg peak contributions and simulation confirmed the collinear alignment and full magnetization of the iron spins throughout the iron layers. Energyand magnetic field dependent measurements complete the picture. By tuning the x-ray energy to the chromium L-edge, a signal 20 times weaker compared with iron, demonstrates that the weak magnetic moment in the chromium layers could be detected. Especially the AFM contribution shows strong effects which could be qualitatively and quantitatively evaluated. Simulation show clearly that the magnetic moment is concentrated at the interfaces and could be approximated to a magnetic layer with an effective thickness of about 0.5 Å assuming a step function in the magnetization profile. Soft x-ray data usually suffer from strong absorption and the limited Qz-range and resolution and therefore the use of hard x-rays seems desirable to probe the whole sample. Magnetic hard x-ray reflectivity measurements on the Fe/Cr double multilayer carried out at the CMC beamline by switching the magnetic field on the sample show clear magnetic Bragg reflection at the ferromagnetic structural peaks. They are very well reproduced by simulations and thus confirm the collinear alignment of the iron spins. In order to probe the AFM spin configuration the helicity of the photon beam has to be switched with constant magnetic field. In spite of complications in the reflectivity spectra it was possible to extract the relative orientation of the AFM to FM spin configuration in the two multilayers. In summary the work showed for the example of an Fe/Cr double multilayer that magnetic soft and hard x-ray reflectivity can be applied to retrieve information about the magnetization profile of thin magnetic films and multilayer, and can compliment polarized neutron scattering.