Determination of Complex Magnetic Structures From Polarized Neutron Reflectivity Data by Flexible Modeling of Depth-Dependent Vector Magnetization

In multilayer systems with exchange-coupled layers such as exchange-spring magnets, interfacial pinning can give rise to spiral domain walls and other complex magnetic structures that are sensitive to temperature, relative layer thicknesses, etc. Though these spin structures develop in subsurface layers, the depth-dependent magnetic profile can be fully characterized using polarized neutron reflectivity (PNR). In order to obtain the profile of the vector magnetization as well as the chemical composition, these data are typically analyzed using software in which the sample is described by a series of flat layers. This approach is cumbersome for continuously varying depth profiles, such as magnetic spirals, since the magnetic layers must be artificially subdivided to mimic the smooth changes in the vector magnetization. Thus, we have developed a flexible PNR fitting program in which users can specify a formula for the model (e.g., flat, power law, or piecewise polynomials). The program can easily be extended to handle simultaneous fitting of multiple data sets from measurements made with different techniques (such as PNR and X-rays) with constraints between the models.