Interplay between magnetic and dielectric phenomena at transition metal oxide interfaces

The present work is concerned with the preparation, characterization and analysis of thin film heterostructures of perovskite oxide materials. Two different systems have been analyzed in detail: La0.66Sr0.33MnO3/SrTiO3 (LSMO/STO) heterostructures have been investigated in order to understand the unusual occurrence of an exchange bias effect in multilayers of these two oxides. Monocrystalline LSMO single and LSMO/STO bilayers have been grown on STO by both High Oxygen Pressure Sputter Deposition (HSD) and Pulsed Laser Deposition. It was possible to reproduce the Exchange Bias effect in the samples grown by HSD by reducing the oxygen pressure during the layer growth. In fact, the size of the effect can be increased by further reduction of the oxygen pressure. The macroscopic sample analysis by X-ray Diffraction and Vibrating Sample Magnetometry suggests that the occurence of the Exchange Bias effect is linked to oxygen deficiencies in the LSMO layer. By combining X-ray Reflectometry, Polarized Neutron Reflectometry and X-ray Resonant magnetic Scattering (XRMS), the magnetic depth profile of the samples has been determined. By this, a region in LSMO at the interface to STO has been detected, where the magnetic moment is strongly suppressed. By putting together the results of the macroscopic sample analysis and the scattering experiments, an explanation for the occurence of the effect can be given: It is proposed, that a combination of strain and oxygen deficiencies shifts the LSMO at the interface in the antiferromagnetic phase of the LSMO strain vs. doping phase diagram. This interface region thus couples to the ferromagnetic part of the LSMO causing the Exchange Bias effect. The second heterostructure system under investigation in this work are bilayers of La0.5Sr0.5MnO3 (LSMO) and BaTiO3 (BTO). A possible dependence of the interface near magnetic structure of La(1-x)Sr(x)MnO3 having a doping level x close to the ferromagnetic-antiferromagnetic phase diagram boundary on the orientation of the electric polarization in the BTO film was proposed theoretically. Therefore, here it was the aim to analyze to what extend these predictions can be reproduced experimentally. Epitaxial La0.5Sr0.5MnO3/BaTiO3 bilayers have been grown by Molecular Beam Epitaxy on conductive Nb doped STO substrates. It was found that the magnetic properties of the LSMO thin film highly depend on the stacking order of the bilayer: the sample with LSMO on top of BTO shows antiferromagnetic, the one with BTO on top of LSMO ferromagnetic behavior. This can be explained by the different strain states present in the two configurations and proves that the doping level x in the La(1-x)Sr(x)MnO3 layer is very close to the nominal value of 0.5. The magnetic depth profile in the samples has been determined by XRMS. The simulations reveal a continuous magnetization in the ferromagnetic sample and a small net magnetization in the antiferromagnetic sample probably is created by canted magnetic moments due to the applied magnetic field during the experiment. The experiment reveals that this small net moment is suppressed at the interface to air, which can be explained by the reduced dimensionality close to the surface. Therefore, this XRMS experiment reveals that it is possible with this technique to resolve magnetic profiles in samples with very small magnetic moments. So far, no dependence of the magnetic profile on the application of electric fields has been detected. Besides the possibility that the proposed effect cannot be reproduced experimentally in general, it is also possible that the electric polarization in BTO is not coherently switchable in the samples prepared in this work. The characterization and improvement of the ferroelectric properties needs to be the focus for future investigations on this system