Erbium manganese oxide is a multiferroic material with interdependent magnetic and electric fields. Using an electric field, you can reverse both the electric polarity of multiferroics and the magnetic polarity, which requires much less energy. | Image: Martin Lilienblum and Manfred Fiebig/ETH Zurich

In the 18th century, Europe was passionate about electricity. In aristocratic salons, scholarly lecturers mesmerised their patrons with demonstrations of ‘amusing physics’. In those early days, it was the wonders of static electricity: sparks, luminescent halos and hair standing on end.

At the same time, in England, the dyer Stephen Gray discovered conductivity. The ‘electric fluid’ is not only static, but also ‘flows’ through assemblies of metal wires, wood and stone. But when the device comes into contact with the earth, the electricity escapes, said Gray. To avoid ‘leaks’, he hung his experiments on silk threads.

This marked a historic turning point. From then on, materials were divided into two categories: conductors and insulators. And since then, this difference marks not only our everyday lives but also our mental space. Almost everyone knows that the metal in electrical sockets conducts, whereas the plastic sheathing of cables insulates.  

Semiconductors, however, are less widely understood. They both insulate and conduct electricity, depending on the circumstances, and first appeared in our homes as early as the 1950s inside transistor radio sets. Today, they are ubiquitous. In the form of silicon chips, they can be found in mobile phones, computers and even toasters. But their properties remain abstract to the general public, probably because the way computers work is not very intuitive.  

We’re not always conscious of it, but silicon defines modernity. The ability to both insulate and conduct electricity is what produces the zeroes and ones of the digital world. But after decades of optimisation, the material is reaching its limits, requiring excessive energy to switch from one state to another.

Objective: energy saving

Several candidates for replacement are rushing to the gate, including graphene and molybdenite, backed by research institutions and companies investing heavily into this area. The Swiss National Science Foundation (SNSF) has launched the second phase of a National Centre of Competence in Research (NCCR) dedicated to this topic, bringing together more than 30 laboratories.

At ETH Zurich, Nicola Spaldin is developing a potential successor to silicon that is more discrete than graphene, but with no less potential: multiferroics. These materials are polarised both magnetically and electrically.

The electrical polarity of multiferroics can be changed using an electric field, but this is nothing exceptional on its own. Yet these materials also have a magnetic polarity - just like magnets. And, as Spaldin has shown, the application of an electric field not only reverses the electrical polarity of the multiferroic but also its magnetic polarity.

This property could change everything. Normally, magnetic materials require a magnetic field to change polarity, e.g., in hard disks. “This requires a large amount of energy. If we can change magnetic polarity with electric fields, we will open the door to much more energy-efficient devices”, says Spaldin.

In theory, multiferroics may enable the development of not only very low-power digital storage solutions but also logic units dedicated to information processing. The field is attracting the attention of the industry. In 2018, Intel produced a first experimental device based on multiferroics.

Fascinating materials

There are also other candidates to succeed silicon that have even stranger properties. One example is topological insulators, which are materials that conduct electricity across their surfaces but not through their cores.

“To understand how they work, just imagine a block of wood covered with a conductive metal foil”, says Luka Trifunovic, a research assistant at the University of Zurich. “Except that when you cut a topological insulator in half, you get new surfaces that are also conductive”.

Other materials exist only in theory. For example, Trifunovic’s mathematical models predict a certain type of cubic crystal: its surfaces and core are insulators, but its edges are conductive. What’s more, and unusual, they are conductive in one direction only! One potential application, according to Trifunovic, is as the basis for quantum memory.

Amazing as they may be, these new materials may never enter our imagination in the same way glass, metal and porcelain did through the parlour shows of the Enlightenment. The silicon chip and its successors are discreet, they do not generate sparks, and they don’t make your hair stand on end. You don’t see them working. The properties of these electronic materials are hidden in the new possibilities offered by computer technologies. In a way, this makes them even more fascinating.