Strongly magnetic materials called antiferromagnets have an intriguing property. Below a certain temperature, the spins (intrinsic angular momenta) of neighbouring atoms in the material point in opposite directions, so that the material exhibits no response to an external magnetic field. The discovery and pioneering studies of antiferromagnets were made by the physicist Louis Néel, who stated in his 1970 Nobel Lecture that the materials “are extremely interesting from the theoretical viewpoint, but do not seem to have any applications” (see go.nature.com/2lzrly8). This pessimistic view of antiferromagnets would change dramatically a few decades later. Today, the materials have practical applications, and promise to deliver several more. In a paper in Nature, Lebrun et al.1 report the observation of long-distance spin transport in an antiferromagnetic insulator, demonstrating that such materials could be used for spin-based electronics (spintronics).
The debut of antiferromagnets in technology was made possible through the 1988 discovery of the giant magnetoresistance effect2,3, which resulted in the 2007 Nobel Prize in Physics. This discovery showed that, in magnetic structures containing nanometre-thick multilayers (stacks of different ultrathin films), electron transport could be controlled by the spin of the electrons — rather than by their electric charge, as in conventional electronics. This finding triggered research into magnetic multilayers, and gave birth to the field of spintronics, which has revolutionized magnetic recording techniques and promises to bring about advances in information technologies4.
Antiferromagnets proved to be essential in sensors that use the giant magnetoresistance effect. Since the late 1990s, the read heads for computer-disk drives have been based on such sensors. These read heads are much more sensitive to changes in magnetic fields than are conventional ones. However, although antiferromagnets are important for spintronic devices, they have had a passive role. Ferromagnets — materials in which all of the atomic spins are aligned — have had the active role.
In the past decade, this situation has begun to change because of experimental and theoretical results showing that antiferromagnets have several advantages over ferromagnets in spintronic devices. One advantage is the insensitivity of antiferromagnets to perturbations in external magnetic fields. Another is their ultrafast dynamics, which could enable devices to operate at terahertz-scale (1012 Hz) frequencies and therefore facilitate faster electronics. Developments in the past few years have given rise to the field of antiferromagnetic spintronics5–7, which now gains a boost, thanks to Lebrun and colleagues.
The authors used a thin, flat sample of single-crystal haematite, α-Fe2O3 — an electrical insulator that is the most common antiferromagnetic iron oxide. They deposited two thin platinum wires onto the surface of the sample, such that the two wires were parallel to each other, and injected an electric current into one of the wires (Fig. 1). This electric current was converted into a spin current, which in metals is produced by electrons with opposite spins moving in opposite directions, by means of a phenomenon called the spin Hall effect8. The spin current flowed into the sample and was transported laterally towards the second wire, where it was converted into an electric current by the inverse spin Hall effect9. Finally, this electric current produced a voltage signal that was picked up by a detector.
By reversing the direction of the injected electric current and calculating the difference in the voltages produced by currents in opposite directions, Lebrun et al. separated the signals arising from the spin Hall effect from those produced by thermal effects. This voltage difference varies linearly with the magnitude of the injected current and can be used to decode information carried by the spin current. The authors measured the voltage difference as a function of several parameters, such as the spatial separation of the wires, the intensity of an applied external magnetic field, and the angle between the direction of this field and the wires.
There are several key achievements in Lebrun and colleagues’ work. First, the authors demonstrated spin transport in an antiferromagnetic insulator over a relatively long distance — tens of micrometres at a temperature of 200 kelvin. This finding suggests that such spin transport could be feasible at room temperature, which would be useful for practical applications. The authors argue that antiferromagnets have an advantage over the well-studied ferromagnetic insulator yttrium iron garnet (Y3Fe5O12), which has been used in similar spin-transport experiments10. In devices based on this material, it is not possible to separate the signals produced by the spin Hall effect from those generated by thermal effects.
A second substantial achievement for device application is the authors’ demonstration that the flow of the spin current can be controlled by an external magnetic field. The results also resolve a controversial issue: the physical mechanism responsible for spin transport in antiferromagnets. The dependence of the voltage difference on the spatial separation of the wires shows that the spin transport is mainly caused by the diffusion of excitations called magnons11, rather than by the spins forming an exotic state of matter known as a superfluid12.
Despite these achievements, some obstacles must be overcome before antiferromagnets can be used in practical devices requiring the control of spin currents. One challenge is pointed out by Lebrun and colleagues: finding antiferromagnets in which spins can be transported over even longer distances than demonstrated in the current work. Another, more formidable, challenge is to find materials amenable to spin-current control using weaker magnetic fields than those used by the authors. Nevertheless, the current findings provide additional impetus to the emerging field of antiferromagnetic spintronics.
Nature 561, 181-182 (2018)