Structure and magnetism of metastable y-Fe nanoparticles in SrTiO3

Iron (Fe), one of the most abundant elements on Earth, can appear in different structural phases associated with contrasting magnetic properties, depending on temperature and pressure. The most common phase is a-Fe, which has a body-centered cubic (bcc) structure and is ferromagnetic. Another iron allotrope, y-Fe, a high temperature phase in bulk, has a face-centered cubic structure (fcc). However, this iron allotrope has been stabilized at room temperature in nanostructures, namely in thin films or nanoparticles. In these structures, where one or more dimensions are in the nanoscale regime, the structural and magnetic properties can be different from those of bulk y-Fe. Whereas bulk y-Fe is antiferromagnetic, different magnetic states have been reported for y-Fe thin films. When ferromagnetism was observed, this was associated with a face-centered tetragonal (fct) distortion in the y-Fe thin film.
In this thesis, the coupling between structure and magnetism in embedded y-Fe nanoparticles is investigated, i.e. when strained in three dimensions. We have successfully stabilized y-Fe nanoparticles in Sr(Ti,Fe)O3, an oxide perovskite-type matrix. A detailed structural and magnetic characterization showed that these embedded y-Fe nanoparticles have an fct distortion, in analogy to ferromagnetic y-Fe thin films.
%The small tetragonal distortion of the Sr(Ti,Fe)O3 matrix, possibly due to the lattice site location of non-precipitated iron, is too small to solely explain the fct distortion in these embedded y-Fe nanoparticles. 
This strongly suggests that the fct structure is a general property of low-dimensional y-Fe. Moreover, these embedded y-Fe nanoparticles are superparamagnetic, i.e. ferromagnetic below their blocking temperature. Based on these findings we propose a unified model of y-Fe in a strain-induced ferromagnetic ground state in either thin films or nanoparticles.
This model picture of coupling between structure and magnetism in y-Fe builds on the 2y-state model introduced by Weiss in the early 60's, where different electronic structures were associated with different lattice volumes and different magnetic ordering states. Under tensile strain, y-Fe is in a high-volume ferromagnetic state (Weiss' y2 state). When relaxed, it is in a low-volume antiferromagnetic state, corresponding to the y1 state. We show that both y states can coexist in y-Fe at a finite temperature, by thermal excitation between ferromagnetic y2 and antiferromagnetic y1 (paramagnetic above its Néel temperature) states. This causes a non-Curie phase transition in y-Fe. In this thesis, we generalize this model of the 2y-state model to include the role of the fct distortion in determining the magnetic ground state: starting from relaxed fcc y-Fe, there is a crossover from an antiferromagnetic to a ferromagnetic ground state with increasing fct distortion. Our results thereby motivate and constitute valuable experimental input for a theoretical reassessment of the coupling between structure and magnetism in y-Fe, in particular in y-Fe nanoparticles.
In addition to its fundamental interest, this rich interplay between crystal structure, electronic structure and magnetism in y-Fe opens interesting prospects to use y-Fe as a functional material, for example, as the ferromagnetic constituent of an artificial multiferroic system. Showing that y-Fe nanoparticles can be embedded in an perovskite-type structure (i.e. Sr(Ti,Fe)O3) indicates that y-Fe nanostructures can be combined with perovskite-type ferroelectric materials. In such artificial multiferroics, the strain induced on y-Fe, via an electric field applied on the ferroelectric host, would allow to reversibly change the magnetic ground state.