A research collaboration between the University of Chicago, Argonne National Laboratory, and The Pennsylvania State University demonstrated that Cr-(Bi,Sb)₂Te₃, a topological insulator, grown on a strontium titanate (SrTiO₃)  substrate can be reversibly patterned using light and an external magnetic field. These results, published recently in the Proceedings of the National Academy of Sciences (PNAS), demonstrate that topological insulators can be engineered into devices for magnetic storage or quantum computing applications.

Topological materials belong to a class of quantum materials with unusual band structures, where the spin-orbit interaction produces energy shifts large enough to “invert” the conduction and valence bands. Going from a topological insulator to any other material (including vacuum) requires these bands to “unwind,” producing unique symmetry-protected surface conductivity on materials that would otherwise be insulators. Topological insulators specifically have insulating interiors but are conducting on their surface. As a result, electrons do not flow in the bulk of the material but move freely along the surface. “It’s that surface that could be ultimately used to construct a wire in a nanoelectric device,” says N. Peter Armitage of Johns Hopkins University. Armitage also studies quantum materials but was not connected with this work.

Many devices have been proposed to take advantage of this surface conductivity of topological insulators, but their fabrication has been hindered by the inherent fragile nature of topological insulators. A topological insulator must be carefully grown for its properties to manifest. Furthermore, “topological materials tend to degrade rapidly when exposed to atmosphere, water, and cleanroom processing such as lithography, deposition, and etching,” says David Awschalom, corresponding author for the article.

In this work, Awschalom’s team manipulated both the magnetic and charge distribution, or chemical potential, in Cr-(Bi,Sb)₂Te₃ to produce micron-sized magnetic patterns and p-n­ junctions in a single thin film. High-quality films of Cr-(Bi,Sb)₂Te₃ were grown by molecular beam epitaxy in the laboratory of Nitin Samarth at The Pennsylvania State University. The technique used to control the film’s magnetization resembles that used in commercial magneto-optical storage devices from two decades ago. Control over the chemical potential in topological insulators had been previously demonstrated by this group. According to Andrew L. Yeats, a postdoctoral fellow in Awschalom’s group and first author on the PNAS publication, “We took that method and combined it with magneto-optical recording techniques to show independent local control over both the chemical potential and magnetization of these materials.”

Even more advantageous, Yeats and Awschalom showed that these patterns persist at low temperatures for over 24 hours and can be completely erased. By either applying a strong magnetic field or shining a different color of light on the topological insulator films, existing patterns could be erased completely and the film returned to its original state. This could prove powerful in the prototyping of novel devices using topological insulators: to redesign a device “you can erase it and write another one with the same material, on request,” Awschalom says.  

In principle, this article represents the demonstration of a functioning read-write device made from a topological insulator. Additional experiments are needed to make it practical for everyday use. First, miniaturizing the device below the micron-scale must be achieved. Yeats and Awschalom are interested in fully understanding the mechanism by which these patterns are created to determine its fundamental microscopic limit. Second, most magnetic properties of topological insulators are only visible at temperatures close to absolute zero, around 5 K. Awschalom mentions that if topological insulators with room-temperature magnetism can be synthesized, they could likely be patterned using these procedures. 

This research introduces many interesting directions beyond those necessary for commercialization. Yeats and Awschalom suspect that many other topological insulators can be patterned with these methods. Many of the interesting properties of topological insulators have been predicted by theory but not yet observed in experiments. “These techniques may be useful in a lot of those efforts,” Yeats says. Armitage thinks these techniques could be used to engineer a topological-insulator-based metamaterial by patterning an array of magnetic or charge domains. The topological insulator could be “patterned in such a way that it displays novel optical properties,” Armitage says.  

Read the article in PNAS.