Multiferroics are compounds that show ferroelectricity and magnetism. BiFeO3, by far the most studied, has outstanding ferroelectric properties, a cycloidal magnetic order in the bulk, and many unexpected virtues such as conductive domain walls or a low bandgap of interest for photovoltaics. Although this flurry of properties makes BiFeO3 a paradigmatic multifunctional material, most are related to its ferroelectric character, and its other ferroic property--antiferromagnetism--has not been investigated extensively, especially in thin films. Here we bring insight into the rich spin physics of BiFeO3 in a detailed study of the static and dynamic magnetic response of strain-engineered films. Using Mössbauer and Raman spectroscopies combined with Landau-Ginzburg theory and effective Hamiltonian calculations, we show that the bulk-like cycloidal spin modulation that exists at low compressive strain is driven towards pseudo-collinear antiferromagnetism at high strain, both tensile and compressive. For moderate tensile strain we also predict and observe indications of a new cycloid. Accordingly, we find that the magnonic response is entirely modified, with low-energy magnon modes being suppressed as strain increases. Finally, we reveal that strain progressively drives the average spin angle from in-plane to out-of-plane, a property we use to tune the exchange bias and giant-magnetoresistive response of spin valves.

In the search for multiferroic materials magnetic compounds with a strongly elongated unit-cell (large axial ratio c/a) have been scrutinized intensely. However, none was hitherto proven to have a switchable polarization, an essential feature of ferroelectrics. Here, we provide evidence for the epitaxial stabilization of a monoclinic phase of BiFeO3 with a giant axial ratio (c/a=1.23) that is both ferroelectric and magnetic at room temperature. Surprisingly, and in contrast with previous theoretical predictions, the polarization does not increase dramatically with c/a. We discuss our results in terms of the competition between polar and antiferrodistortive instabilities and give perspectives for engineering multiferroic phases.

First-principles-based simulations are used to simulate the electrocaloric effect (ECE) in Ba(0.5)Sr(0.5)TiO(3) alloys. In analogy with experimental studies we simulate the effect directly and indirectly (via the use of Maxwell thermodynamics). Both direct and indirect simulations utilize the same atomistic framework that allows us to compare them in a systematic way and with an atomistic precision for the very first time. Such precise comparison allows us to provide a bridge between the atomistic and macroscopic descriptions of the ECE and identify the factors that may critically compromise or even destroy their equivalence. Our computational data reveal the intrinsic features of ECE in ferroelectrics with multiple ferroelectric transitions and confirm the potential of these materials to exhibit giant electrocaloric response. The coexistence of negative and positive ECE in one material as well as an unusual field-driven transition between them is predicted, explained at an atomistic level, and proposed as a potential way to enhance the electrocaloric efficiency.

An effective Hamiltonian scheme is developed to study finite-temperature properties of multiferroic BiFeO3. This approach reproduces very well (i) the symmetry of the ground state, (ii) the Néel and Curie temperatures, and (iii) the intrinsic magnetoelectric coefficients (that are very weak). This scheme also predicts (a) an overlooked phase above Tc approximately 1100 K that is associated with antiferrodistortive motions, as consistent with our additional x-ray diffractions, (b) improperlike dielectric features above Tc, and (c) that the ferroelectric transition is of first order with no group-subgroup relation between the paraelectric and polar phases.

No abstract