Determination of optimal experimental conditions for accurate 3D reconstruction of the magnetization vector via XMCD-PEEM.

This work presents a detailed analysis of the performance of X-ray magnetic circular dichroism photoemission electron microscopy (XMCD-PEEM) as a tool for vector reconstruction of magnetization. For this, 360° domain wall ring structures which form in a synthetic antiferromagnet are chosen as the model to conduct the quantitative analysis. An assessment is made of how the quality of the results is affected depending on the number of projections that are involved in the reconstruction process, as well as their angular distribution. For this a self-consistent error metric is developed which allows an estimation of the optimum azimuthal rotation angular range and number of projections. This work thus proposes XMCD-PEEM as a powerful tool for vector imaging of complex 3D magnetic structures.

Endstation for ultrafast magnetic scattering experiments at the free-electron laser in Hamburg.

An endstation for pump-probe small-angle X-ray scattering (SAXS) experiments at the free-electron laser in Hamburg (FLASH) is presented. The endstation houses a solid-state absorber, optical incoupling for pump-probe experiments, time zero measurement, sample chamber, and detection unit. It can be used at all FLASH beamlines in the whole photon energy range offered by FLASH. The capabilities of the setup are demonstrated by showing the results of resonant magnetic SAXS measurements on cobalt-platinum multilayer samples grown on freestanding Si(3)N(4) membranes and pump-laser-induced grid structures in multilayer samples.

Breakdown of the x-ray resonant magnetic scattering signal during intense pulses of extreme ultraviolet free-electron-laser radiation.

We present results of single-shot resonant magnetic scattering experiments of Co/Pt multilayer systems using 100 fs long ultraintense pulses from an extreme ultraviolet (XUV) free-electron laser. An x-ray-induced breakdown of the resonant magnetic scattering channel during the pulse duration is observed at fluences of 5 J/cm(2). Simultaneously, the speckle contrast of the high-fluence scattering pattern is significantly reduced. We performed simulations of the nonequilibrium evolution of the Co/Pt multilayer system during the XUV pulse duration. We find that the electronic state of the sample is strongly perturbed during the first few femtoseconds of exposure leading to an ultrafast quenching of the resonant magnetic scattering mechanism.

Ultrafast optical demagnetization manipulates nanoscale spin structure in domain walls.

During ultrafast demagnetization of a magnetically ordered solid, angular momentum has to be transferred between the spins, electrons, and phonons in the system on femto- and picosecond timescales. Although the intrinsic spin-transfer mechanisms are intensely debated, additional extrinsic mechanisms arising due to nanoscale heterogeneity have only recently entered the discussion. Here we use femtosecond X-ray pulses from a free-electron laser to study thin film samples with magnetic domain patterns. We observe an infrared-pump-induced change of the spin structure within the domain walls on the sub-picosecond timescale. This domain-topography-dependent contribution connects the intrinsic demagnetization process in each domain with spin-transport processes across the domain walls, demonstrating the importance of spin-dependent electron transport between differently magnetized regions as an ultrafast demagnetization channel. This pathway exists independent from structural inhomogeneities such as chemical interfaces, and gives rise to an ultrafast spatially varying response to optical pump pulses.

Holographically aided iterative phase retrieval.

Fourier transform holography (FTH) is a noise-resistant imaging technique which allows for nanometer spatial resolution x-ray imaging, where the inclusion of a small reference scattering object provides the otherwise missing phase information. With FTH, one normally requires a considerable distance between the sample and the reference to ensure spatial separation of the reconstruction and its autocorrelation. We demonstrate however that this requirement can be omitted at the small cost of iteratively separating the reconstruction and autocorrelation. In doing so, the photon efficiency of FTH can be increased due to a smaller illumination area, and we show how the presence of the reference prevents the non-uniqueness problems often encountered with plane-wave iterative phase retrieval. The method was tested on a cobalt/platinum multilayer exhibiting out of plane magnetized domains, where the magnetic circular dichroism effect was used to image the magnetic domains at the cobalt L3-edge at 780eV.

Diffractive imaging using partially coherent x rays.

The measured spatial coherence characteristics of the illumination used in a diffractive imaging experiment are incorporated in an algorithm that reconstructs the complex transmission function of an object from experimental x-ray diffraction data using 1.4 keV x rays. Conventional coherent diffractive imaging, which assumes full spatial coherence, is a limiting case of our approach. Even in cases in which the deviation from full spatial coherence is small, we demonstrate a significant improvement in the quality of wave field reconstructions. Our formulation is applicable to x-ray and electron diffraction imaging techniques provided that the spatial coherence properties of the illumination are known or can be measured.

Experimental measurement of the four-dimensional coherence function for an undulator x-ray source.

A full measurement of the four-dimensional coherence function from an undulator beam line is reported. The analysis is based on the observation that the data are consistent with a coherence function that is mathematically separable. The effective source size can be altered by changing the width of the exit slit, and the complete coherence function is presented for two settings. We find, to within experimental error, that the four-dimensional complex degree of coherence can be described as a real Gaussian function that depends only on the difference of the spatial coordinates.