A structural switch for magnetism

Magnetic materials have been a mainstay of computing technology due to their ability to permanently store information in their magnetic state. Current technologies are based on ferromagnets, whose states can be easily reversed by magnetic fields. Faster, denser and more robust next-generation devices would be made possible by using a different class of materials, called antiferromagnets. Their magnetic state, however, is notoriously difficult to control.

Now, a research team from MPSD and the University of Oxford has successfully driven a prototype antiferromagnet into a new magnetic state using terahertz frequency light. Their revolutionary method produced an effect orders of magnitude larger than before, and on ultra-fast time scales. The team’s work has just been published in Natural Physics.

The strength and direction of a magnet’s “north pole” is referred to as its so-called magnetization. In ferromagnets, this easily reversible magnetization can represent a “bit” of information, which has made them the materials of choice for magnet-based technologies. But ferromagnets are slow to work and react to stray magnetic fields, which means they are error-prone and cannot be very close to each other.

Antiferromagnets represent an interesting alternative. Unlike ferromagnets, they have no macroscopic magnetization, as they are made up of alternating “magnetic moments” pointing up and down, like atomic-sized bar magnets that reverse the direction of an atom to the other. They are not strongly affected by magnetic fields, making them robust for information storage and allowing them to be scaled to much smaller sizes. Moreover, they could respond faster than current devices, with frequencies up to several terahertz. The challenge for researchers is to find ways to reliably change the magnetic state of an antiferromagnet.

In their new paper, the MPSD/Oxford research team took a novel approach, investigating how the magnetic state of an antiferromagnet is affected by its crystal structure. They exploited a property of certain antiferromagnets called piezomagnetism, where a change in the atomic structure leads to magnetization, just like in a ferromagnet. This change is usually achieved by applying uniaxial pressure, but it is a slow process that can break the crystal.

Instead of pressure, the team used light to control the piezomagnetic effect in CoF₂. The method, originating from the Hamburg group in 2011, is based on exciting lattice vibrations, or “phonons”, with carefully matched light pulses. By adjusting the frequency and polarization of light pulses, they could induce the same structural distortions that give rise to piezomagnetism without having to strain the crystal – an experimental idea proposed by co-author Paolo Radaelli of the University of Oxford during his a visit to the MPSD in 2018.

This innovative technique allowed the researchers to create a magnetization 400 times greater than that obtained before. Amazingly, it only took about 100 ps for the magnetization to grow, and the direction of the magnetization could be reversed by changing the polarization of the light. The results represent a major advance in the optical control of material properties.

Lead author Ankit Disa says, “This experiment was the first demonstration of ‘rational’ or ‘intentional’ engineering of a crystal structure with light. We knew what kind of structural distortion was needed to create a transition of phase from an antiferromagnetic to a ferromagnetic-like state.The trick was to figure out how to use light to drive the material into this new crystal structure.

Andrea Cavalleri, who led the experimental team at MPSD and is involved in the CUI: Advanced Imaging of Matter center of excellence, sees vast potential in using light to control the properties of materials: “This technique could lead to optomagnetic switches, for example, to create memories that could be written and read by light.More fundamentally, we now have the tools and understanding to optically engineer the structure of materials at the atomic scale, which can be applied to manipulate functionality in many types of systems, from magnets to ferroelectrics to superconductors.