Berlin 2005 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Downloads | Hilfe

MA: Magnetismus

MA 20: Poster:Schichten(1-29),Spintrsp(30-45),Ex-Bias(46-63),Spindyn(64-75),Mikromat.(76-80),Cluster(81-94),Abbv.(95-99),Obflm(100-02),SpElek.(103-09),E-Theo(110-14),Mikromag.(115-16),Spin+PÜ(117-26),Mag.Mat.(127-51),Meth.(152-55),Mol.Mag(156-59),Kondo(160-65

MA 20.14: Poster

Montag, 7. März 2005, 14:00–18:00, Poster TU C

Advanced fitting algorithm for resonant magnetic and non magnetic soft and hard X-ray reflectometry simulations — •Max Harlander, Jochen Geisler, Eberhard Goering, Uwe Grüner, Sebastian Macke, and Gisela Schütz — Max-Planck-Insitut für Metallforschung, 70569 Stuttgart

We developed a new fit program to simulate and fit the reflectivity of multilayer systems, based upon a modified Parratt-formalism. Conventional Parratt programs allow simulations and fits of the reflectivity of magnetic and non magnetic multilayers, where the reflection curves depend on the thickness, roughness and non resonant and resonant optical constants (δ and β) of the layered system. With the advanced algorithm, even for single buried layers the resulting resonant and element specific asymmetry ratios for reversed magnetization directions could be fitted and optimized. The resonant magnetic scattering has been introduced by superposed magneto optical profiles, which are approximated by Gaussian lines and/or splines at each individual interface. Some additional aspects have been introduced to improve the fit quality beyond the small angle scattering limit, even for large scattering angles as observed in the soft X-ray region. The π- and σ components and the degree of circular polarization of the incident X-Ray beam could be taken into account and independently chosen. The major improvement is related to the fitting speed. After a careful optimization, the magnetic fitting is roughly 300 times faster than the previous implementation. In addition, a simulated annealing algorithm has been integrated to avoid local minima convergence.