The ability to use an electric or magnetic field to manipulate the orientation of electric dipoles or magnetic moments associated with atoms, ions or molecules in a material provides a vast array of functions. In rare materials called magnetoelectric multiferroics, the dipoles are intimately coupled to the moments, and a single field can control both1. After the field is applied, however, the dipoles and moments typically all have the same orientation, and the original pattern that they formed is lost. In a paper Nature, Leo et al.2 show that, in two particular materials, a magnetic field can flip each of the dipoles or moments while preserving the structure of the original pattern. The work illustrates how the complex coupling in these materials could be used to uncover other, previously unobserved electric and magnetic effects.
When most materials are placed in an electric field, their positive and negative charges shift by a tiny amount (less than 0.1 nanometres, which is about the radius of an atom). This microscopic movement leads to a macroscopic, measurable response: an electric polarization. In ferroelectric materials, however, clusters of ions assemble in a way that results in electric dipoles and a macroscopic polarization, even in the absence of an electric field.
Ferroelectrics are typically composed of domains — mesoscopic regions, often 100 nm to several micrometres in size, in which dipoles are aligned. Applying a strong electric field to a ferroelectric material causes all of the dipoles to point in a single direction, erasing both the original domain pattern and any engineered functions of the domain structure or of the boundaries between domains3.
There is a magnetic analogue to this phenomenon. A ferromagnetic material contains concerted arrangements of electron magnetic moments, which are located on specific sites of the material’s atomic lattice. These moments generate a macroscopic magnetization that can be controlled using a magnetic field. Most ferromagnetic materials are also composed of mesoscopic domains.
Despite the apparent macroscopic similarities between ferroelectricity and ferromagnetism, materials that exhibit both phenomena, known as multiferroics, are exceedingly rare4. Magnetoelectric materials — those in which electric and magnetic properties are coupled, but that do not necessarily possess ferroelectric or ferromagnetic order — are also uncommon. Most exotic are magnetoelectric multiferroics, in which ferroelectricity and ferromagnetism are intrinsically coupled. This coupling holds great potential for next-generation devices, such as data-storage units that run on ultra-low power, highly sensitive magnetic-field detectors5 and energy-efficient nanoscale motors6. Much of the research focus on magnetoelectric multiferroics so far has centred on the control of magnetism using electric fields of ever-decreasing strength7.
Identifying multiferroics is a great challenge. In the current work, however, Leo and colleagues recognize that once such a material is identified, the complex parameters that give rise to this state of matter can be combined or manipulated in completely distinct ways. They illustrate this new way of thinking about multifunctional materials by considering the intertwined electric and magnetic properties of two such materials, imaging the domain structure while applying an external magnetic field.
The authors observed domains in the materials using a technique called optical second-harmonic generation. In this approach, two photons interact with a material to produce a single photon that has twice the frequency of the incident photons. The technique is sensitive to the spatial and magnetic (point-group) symmetry of the material’s lattice, making it a powerful probe of structural, electronic and magnetic order. Of particular relevance to the authors’ work is that second-harmonic generation is sensitive to magnetism even when the magnitude of the magnetic moments in the material is 1,000 times smaller than that of the moments in a typical ferromagnet1,8 — a sensitivity that can be matched by few complementary techniques.
Leo et al. studied ferromagnetic domains in one of the materials as a perpendicular magnetic field was swept across the material, and ferroelectric domains in the other material during application of a parallel magnetic field. They found that when the field was gradually changed from one direction to the opposite direction, the boundaries between the domains moved. But, remarkably, when this process was complete, the polarization or magnetization of each domain was reversed and the original domain pattern was recovered (Fig. 1).
Such an effect is similar to switching the black and white squares of a chessboard, without changing the boundaries between the squares. It is in sharp contrast to what is usually observed when a uniform field is applied to a material: an alignment of all the electric dipoles or magnetic moments, or in the chess analogy, a conversion of all the squares to a single colour.
The authors explain the inversion effect as being due to the coupling of three order parameters — variables that describe the alignment of dipoles or moments in a material. The first parameter represents the observed domain distribution. The second parameter, which is unaffected by the applied magnetic field, imprints the original domain pattern onto the first parameter. Finally, the third parameter, which is directly controlled by the field, causes the observed domain distribution to be inverted.
Leo and colleagues’ results suggest that the coupling of multiple order parameters is generic, but it remains to be seen how frequently it manifests in other materials. Perhaps more importantly, however, the study shows how multiple order parameters in certain materials can be exploited. Although magnetoelectric multiferroics have garnered much interest because of their strongly coupled magnetization and ferroelectric polarization, future work might find ways to combine the many order parameters in these materials to derive new functions. Precisely what other relationships might be lurking between these parameters is uncertain. Nevertheless, the authors’ demonstration of domain-pattern inversion resulting from the coupling of three order parameters is a big step forward in our understanding of complex coupling in multiferroic materials.
Nature 560, 435-436 (2018)