| Summary: | Multiferroic materials are a very exotic type of materials which present simultaneously two or more ferroic properties. Magnetoelectric multiferroics, in particular, are a very prominent class of materials, mainly due to their outstanding foreseen applications such as magnetic sensors, energy harvester/conversion devices, and high efficiency memories. However, intrinsic magnetoelectric materials are quite rare and do not have, yet, the adequate properties to the everyday applications. One of the reasons for this to occur is due to the requirements for magnetism and ferroelectricity in matter being a priori contradictory, since the former needs unfilled dn orbitals, while the latter favours d0 orbitals. Nevertheless, extrinsic magnetoelectric multiferroics do not suffer from this problem because they do not share the same phase, hence being a very promising approach to engineer adequate magnetoelectric multiferroics. This thesis focus on the study of Fe and BaTiO3 systems as a means of achieving novel magnetoelectric effects. It is shown that a peculiar type of BaTiO3:Fe auto-composite presents an ordered magnetic behaviour, despite the concentration of Fe being as low as 113 atomic ppm. The Fe magnetization displays two abrupt changes in its spontaneous value, one with M/M ≈ 32% and the other with M/M ≈ 14%. These magnetic transitions are correlated the BaTiO3 orthorhombic↔tetragonal and tetragonal↔cubic ferroelectric phase transitions. This magnetoelectric auto-composite was the motivation to resort to Density Functional Theory (DFT) modeling as a means to discover the microscopic mechanism(s) behind such a strong magnetoelectric effect. The study of an iron monolayer placed upon several possible BaTiO3 unit cells lead to the discovery of several interfaces with abrupt changes in their spontaneous magnetization, either through the enhancement and reduction of the Fe magnetic moments, or through the change between antiferromagnetic and ferromagnetic order of the Fe monolayer. However, the highlight of these DFT studies lies in the discovery of a particular kind of interfaces, namely in the BTO221_2ndFe and BTO99_2ndFe supercells, where there is a High-Spin–Low-Spin state transition which can quench completely the atomic magnetic moment of each of Fe atom, depending on the local crystal field felt by the Fe atoms. Based on this specific effect, where it is possible to turn on and off the magnetic moments of the Fe atoms, a magnetoelectric multiferroic device was proposed. Knowing the importance of the crystal field for the High-Spin–Low-Spin state transition, a thorough study regarding the Electric Field Gradient (EFG) of each possible BaTiO3 site was performed, resorting to a combined study of DFT and Perturbed Angular Correlations (PAC) spectroscopy. In this study, it was concluded that the PAC spectroscopy is not the most adequate hyperfine technique to be used in a quantitative study of the BaTiO3/Fe interfaces EFG tensor, due to the non-negligible effects of the radioactive probe on the BaTiO3 matrix. Finally, the deposition of BTO/Fe heterostructures on LaAlO3, MgO, Al2O3 and SrTiO3 substrates using RF-Sputtering, and the Molecular Beam Epitaxy (MBE) deposition of Fe layers on BaTiO3 cut at the (100), (110) and (111) planes were performed as an attempt to recreate the interfaces with the most appealing magnetoelectric effects predicted in the DFT modeling. The thin films deposited using sputtering showed the growth of many Fe, Ba-Ti-O and Fe-Ti-O oxides depending strongly on their substrate, as well as in the deposition and annealing conditions. Still no magnetoelectric coupling was observed in such thin films. On the other hand the Fe thin films deposited on BaTiO3 substrates showed large magnetoelectric couplings between the BaTiO3 ferroelectric phase transitions and the magnetization of the Fe layers (similarly to what happened in the BaTiO3:Fe auto-composite). The magnitude of this magnetoelectric couplings is strongly correlated with the BTO interface where the Fe was deposited, showing a huge change in spontaneous magnetization and coercivity for the rhombohedral↔orthorhombic ferroelectric phase transition up to M/M ≈ 148% and HC/HC ≈ 183% respectively for the (110) case.
|
|---|