Imagine if your computer could run on electricity that flows forever without overheating. This isn’t magic: It’s the potential future of a real phenomenon called superconductivity, which today underpins everything from cutting-edge magnetic research to MRIs.

Now, scientists have found that they can make a superconductor that’s different from others that have come before. It lets electricity flow in only one direction: Like a train pointing downhill, it slides freely one way but faces a daunting uphill in the other. It sounds arcane, but this ability is critical to making electronic circuits like the ones that power your computer. If these scientists’ results hold, it could bring that future one step closer.

“There are so many fun possibilities available now,” says Mazhar Ali, a physicist at Delft University of Technology in the Netherlands, and one of the authors who published their work in the journal Nature on April 27.

Superconductivity flies in the face of how physics ought to work. Normally, as electric current flows along a wire, the electrons inside face stiff resistance, brushing up against the atoms that form the wire. The electrical energy gets lost, often as heat. It’s a large part of why your electronics can feel hot to the touch. It’s also a massive drain on efficiency.

But if you deep-chill a material that conducts electricity, you’ll reach a point that scientists call the critical temperature. The precise critical temperature depends on the substance, but it’s usually in the cryogenic realm, barely above absolute zero, the coldest temperature allowed by physics. At the critical point, the material’s resistance plunges off a cliff to functionally nil. Now, you’ve created a superconductor.

What does resistance-free electricity look like? It means that current can flow through a wire, theoretically for an eternity, without dissipating. That’s a startling achievement in physics, where perpetual motion shouldn’t be possible.

“It violates our current understanding of how one-way superconductivity can occur.”

Mazhar Ali

We’ve known about this magical-sounding quirk of quantum physics since a student in the Netherlands happened across it in 1911. Today, scientists use superconductivity to watch extremely tiny magnetic fields, such as those inside mouse brains. By coiling superconducting wires around a magnet, engineers can craft low-energy, high-power electromagnets that fuel everything from MRI machines in hospitals to the next generation of Japanese bullet trains.

Bullet trains were probably not on the minds of Ali and his colleagues when they set about their work. “My group was not approaching this research with the goal of realizing one-way superconductivity actually,” says Ali.

Ali‘s group, several years ago, had begun investigating the properties of an evocatively named metal, Nb3Br8, made from atoms of niobium (a metal often used in certain types of steel and specialized magnets) and bromine (a halogen, similar to chlorine or iodine, that’s often found in fire retardants).

As the study team made thinner and thinner sheets of Nb3Br8, they found that it actually became more and more conductive. That’s unusual. To further investigate, they turned to a tried technique: making a sandwich. Two pieces of a known superconductor were the bread, and Nb3Br8 was the filling. The researchers could learn more about Nb3Br8 from how it affected the sandwich. And when they looked, they found that they’d made a one-way superconductor.

What Ali’s group has created is very much like a diode: a component that only conducts electricity in one direction. Diodes are ubiquitous in modern electronics, critical for underpinning the logic that lets computers operate.

Yet Ali and his colleagues don’t fully know how this effect works in the object they created. It also, as it turns out, “violates our current understanding of how one-way superconductivity can occur,” says Ali. “There is a lot of fundamental research as well that needs to be done” to uncover the hidden new physics.

It isn’t the first time physicists have built a one-way superconducting road, but previous constructions generally needed magnetic fields. That’s common when it comes to manipulating superconductors, but it makes engineers’ lives more complicated.

“Applying magnetic fields is cumbersome,” says Anand Bhattacharya, a physicist at Argonne National Laboratory in suburban Chicago, who was not one of the paper authors. If engineers want to manipulate different parts within a superconductor, for instance, magnetic fields make a formidable challenge. “You can’t really apply a magnetic field, very locally, to one little guy.”

For people who dream of constructing electronics with superconductors, the ability to send electricity in one direction is a powerful inspiration. “You could imagine very cool device applications at low temperatures,” says Bhattacharya.

Such devices, some scientists believe, have some obvious hosts: quantum computers, which harness particles like atoms to make devices that do things conventional computers can’t. The problem is that tiny amounts of heat can throw quantum computers off, so engineers have to build them in cryogenic freezers that keep them barely above absolute zero. The problem compounds again: Normal electronics don’t work very well at those temperatures. An ultra-cold superconducting diode, on the other hand, may thrive.

Conventional computers could benefit, too: Not your personal computer or laptop, most likely, but larger behemoths like industrial supercomputers. Other beneficiaries could be the colossal server racks that line the world’s data centers. They account for a whopping 1 percent of the world’s energy consumption, comparable to entire mid-sized countries. Bringing superconductors to data servers could make them thousands of times more energy-efficient.

There is some way to go before that can happen. One next step is finding how to produce many superconducting diodes at once. Another is to find how to make them operate above -321°F, the boiling point of liquid nitrogen: That temperature sounds extremely low, but it’s easier to achieve than the even colder temperatures, supplied by liquid hydrogen, that current devices might need.

Despite those challenges, Ali is excited about the future of his group’s research. “We have very specific ideas for attacking both of these avenues and hope to see some more ground-breaking results in the next couple of years,” he says.