New Structures and Functionalities in Multiferroic Bismuth Ferrite: First-principles–based studies

Abstract

In this work, we present a comprehensive computational and theoretical study of the structural phase space of multiferroic Bi-Fe-O and the influence of the structure on functional properties. Complex oxides are known for their many different functionalities, including ferroelectricity, ferromagnetism, or piezoelectricity, which are relevant for technological applications. Multiferroicity, in which at least two different ferroic orders (ferromagnetism, ferroelectricity, and ferroelasticity) coexist and couple in a single phase is of particular interest and can enable, for example, control of magnetic properties using an applied electric field. BiFeO3 (BFO) is one of the most studied multiferroic materials because of the coexistence of magnetic and polar orders at room temperature promising for applications. The substantial intrinsic functionalities of complex oxides can be further enhanced by growing them as thin films in superlattices or heterostructures. In most cases identified to date, this is a result of the change in the material’s lattice constants caused by epitaxial strain imposed by the substrate. This change in lattice constants can modify the functionalities, for example causing a phase change to a higher polar state in BFO. In heterostructures of polar and non-polar materials, the polar discontinuity at the interface results in an accumulation of surface charges, which in turn create an electric field, known as the depolarising field. This de- polarising field can cause the formation of domains in the polar material and even change the orientation of the polarisation, in order to reduce the energy penalty of the interfacial charge. We start our study by considering the case in which a phase transition to a non-polar phase with low relative energy is preferred over polar domains formation and present a new phase with a larger unit cell size and surprising properties matching recent experimental results. We then explore the phase space of BFO using density functional theory (DFT) calculations and reveal several low-energy phases with large unit cells which could be stabilised by a polar dis- continuity at the interface. Finally, we consider a heterostructure in which the polar discontinuity is partially reduced by the presence of differently charged ionic layers and show that this causes an isosymmetric phase transition from two phases with the same symmetry but different polar states in highly-strained BFO. Next, we study a new candidate multiferroic material, bismuth hexaferrite, which exhibits net magnetic and possible ferroelectric moments at room temperature, experimentally demonstrated. We show that the net magnetic moment is likely related to the presence of Fe vacancies and compute the energy barrier between opposite orientations of the polarisation in order to evaluate the likelihood for it to be ferroelectric. This result paves the way for the exploration of new materials in the Bi-Fe-O compositional phase space with technologically relevant properties. Finally, we conclude our work by proposing an accelerated method combining DFT, irreducible phonon modes, and machine learning to explore the potential energy surface of BFO and we predict several low-energy structures, suggesting that the phase space of BFO is still far from being completely known. This thesis highlights the richness of the phase space of Bi-Fe-O and the importance of the boundary conditions provided by the heterostructure, in particular, the electrostatics at the interface in determining the properties of functional oxides. Finally, the methods that we developed provide a new accelerated approach to structure discovery. They are broadly applicable and can certainly lead to discoveries in other materials.