For the first time, researchers have succeeded in making sound travel across a chip – over long distances, without losing speed or information – and even around sharp corners. The researchers say the technology could be useful in ultrasensitive sensors or in amplifying weak signals to make them easier to decode.
One very important feature of computer chips is that signals can travel across them without being lost and without their energy dissipating as heat along the way.
Signal loss is typically measured by how far a signal can travel before it loses half its information.
Now, researchers have developed a new type of chip that uses sound waves, and the signal loss is not measured in millimetres but in kilometres.
The ability to move sound without loss on a chip is similar to how superconductors conduct electricity – and could prove just as significant.
“We have developed a method for routing signals through a chip using sound. In theory, you could build chips that are kilometres long without any meaningful loss of information in the signal,” says a researcher behind the study, Albert Schliesser, Professor at the Niels Bohr Institute of the University of Copenhagen, Denmark.
Sound can travel as tiny vibrations
You cannot just shout into one of Albert Schliesser’s chips and expect the sound to travel kilometres. The researchers work with a highly specialised waveguide for phonons: tiny vibrations that move through solid materials – like the sound you feel when someone knocks on a pipe.
When the atoms in a material vibrate – say, because it is heated – these vibrations can propagate as small waves from atom to atom.
You can think of a phonon as a microscopic wave of sound or heat that hops from atom to atom – like dominoes falling in a line. Just like photons carry light, phonons carry vibrations.
Special chip guides sound without loss
The scientists created an ultrathin chip made of silicon nitride, which enables them to guide the movement of phonons.
The chip is 10 millimetres wide and so thin that 20,000 of them need to be stacked to reach the thickness of a single sheet of graph paper. It is also perforated with small triangular holes with rounded corners, which help to steer the sound.
This chip channels phonons across its surface – from one end to the other, or even around a sharp 120-degree bend.
Signal loss in phonons refers to a drop in the amplitude of the sound wave – that is, its height – as it travels through the chip. Even around those sharp corners, just one phonon in a million goes missing, which is unprecedented.
“We have worked with these membranes before and knew that they had properties that minimise amplitude loss. In this new study, we built a silicon nitride structure that can guide sound waves along a defined path – and even around corners,” explains Xiang Xi, another researcher behind the study at the Niels Bohr Institute.
Could lead to better sensors and computers
Albert Schliesser and Xiang Xi emphasise that their work is still firmly rooted in basic research. Nevertheless, the ability to move phonons across a chip without signal loss could eventually lead to practical applications.
One possibility is developing ultrasensitive sensors that can detect even the faintest signals. In such cases, the sensor itself must not lose part of the signal during processing.
Another potential use is in connecting quantum systems or quantum sensors, in which minimising information loss is also important.
Finally, the new chips could potentially become fundamental building blocks in central processing units (CPUs), in which their precision could enable more complex processors – and thus more powerful computers.
“They will not replace electronic CPUs in computers, but you can do different things with these chips. They can be integrated into other systems in which signals must travel without losing information – or in which signals need to be amplified to make them easier to read,” says Albert Schliesser.
He notes, however, that much more research is needed before any of this becomes reality.
“For now, we want to explore the method and find out what it can really do. For example, we want to build more complex structures and study how phonons can move across them – or build structures in which phonons crash into each other like cars at a crossroads. This will enable us to better understand what is ultimately possible – and what new applications may emerge,” says Albert Schliesser.
