The already buzzing field of topological physics could be about to explode. For the first time, researchers have systematically scoured through entire databases of materials in search of ones that harbour topological states — exotic phases of matter that have fascinated physicists for a decade. The results show that thousands of known materials probably have topological properties — and perhaps up to 24% of materials in all. Previously, researchers knew of just a few hundred topological materials, and only around a dozen have been studied in detail.
“I’m shocked by the number,” says Reyes Calvo, an experimental physicist at the nanoGUNE Cooperative Research Center in San Sebastián, Spain.
In late July, several teams posted preprints1,2,3 detailing their scans of tens of thousands of materials and their predicted topological classifications, which are based on algorithms that use a material’s chemistry and symmetry to calculate their properties. Two teams have already integrated their algorithms into searchable databases. “You can put in a compound name and, with one click, get whether there is topology or not. For me, this is wonderful,” says Chandra Shekhar, a condensed-matter physicist at the Max Planck Institute for Chemical Physics of Solids in Dresden, Germany.
The resulting haul of topological materials could bring scientists closer to finding practical applications for these exotic phases — which have the potential to revolutionize electronics and catalysis. “The more materials with unusual properties we know, the more chance there will be of a breakthrough,” says Oleg Yazyev, a physicist at the Swiss Federal Institute of Technology in Lausanne.
Topological materials derive their unusual features from their topology. In mathematics, this branch studies objects that have spatial properties that remain unchanged when they are smoothly deformed and not torn. In materials, topology applies not to the shape of a solid object but to the geometry of an abstract description of its electrons’ quantum states. Their topological nature means these states are resistant to change, and so stable to temperature fluctuations and physical distortion — features that could make them useful in devices.
The electronic properties of these materials are also unusual, and physicists have been investigating one class, known as topological insulators, since the property was first seen experimentally in 2D in a thin sheet of mercury telluride in 20074 and in 3D in bismuth antimony a year later5. Topological insulators are strange because they consist mostly of insulating material, yet their surfaces are great conductors. And because currents on the surface can be controlled using magnetic fields, physicists think they could find uses in energy-efficient ‘spintronic’ devices, which encode information in a kind of intrinsic magnetism of particles known as spin. But despite a decade of study, physicists have yet to find a topological insulator that has properties suitable for use in devices — for example, a material that is easy to grow, non-toxic and with tunable electronic states at room temperature.
The newly released catalogues classify all non-magnetic materials with known crystal structures by their topology, using methods published last year. Until now, physicists had largely relied on complex theoretical calculations to predict whether a specific material should harbour topological states. But in 2017, teams led by Andrei Bernevig, a physicist at Princeton University in New Jersey, and Ashvin Vishwanath, at Harvard University in Cambridge, Massachusetts, separately pioneered approaches6,7 that greatly speed up the search process. They use algorithms to sort materials automatically into databases on basis of their chemistry and properties that result from symmetries in their structure.
The symmetries — which can occur across mirror, rotational or translational axes — and their directions define where and how electrons can move in the lattice of a crystal structure. They can thus be used to predict how electrons will behave, and so whether a material is likely to host topological states.
Open for exploration
Applying Bernevig’s principles, a team led by researchers at the Beijing National Laboratory for Condensed Matter Physics scanned 39,519 materials and found more than 8,000 that are likely to have topological states. This includes both topological insulators and topological semi-metals — materials in which electrons, under certain conditions, act collectively like massless particles. The latter allow for the study of new quantum phenomena, and are being explored for use as catalysts. The team’s database is available for anyone to access and is searchable by a range variables, such as component elements or the size of each repeating unit in the crystal.
Bernevig’s team also applied its method to create a topological catalogue. His team used the Inorganic Crystal Structure Database and found 5,797 “high-quality” topological materials. The researchers plan to add the ability to check a material’s topology, and certain related features, to the popular Bilbao Crystallographic Server. A third group — including Vishwanath — also found hundreds of topological materials, many of which they deem promising. “As an experimentalist, I find this really thrilling,” says Calvo, who works on stacking layers of 2D materials with topological properties to make next-generation electronic devices.
The current processes are somewhat limited: they can’t be applied to magnetic materials, or those in which electrons interact with each other strongly, says Yazyev. Some of these materials might also have useful topological qualities, so Vishwanath and colleagues are exploring ways to apply similar symmetry-based methods to identify topological magnetic materials8.
Experimentalists now have their work cut out. Researchers will be able to comb the databases to find new topological materials to explore. But until they are probed and measured in experiments, the classifications assigned to each material are still only predictions, Yazyev says. “We now have a large database of candidate materials, and it’s up to experimentalists to uncover new exciting physical phenomena,” he says.
Not every new topological material will prove interesting. So even more useful, says Judy Cha, an experimental physicist at Yale University in New Haven, Connecticut, would be if theorists could factor into the databases other practical information about the materials, such as how defects in the crystal affect the flow of electrons through it; this would help to whittle the list down to only the most practical. “That would be really fantastic,” she says.
Nature 560, 151-152 (2018)