Skip to main content

Building the World's Smallest Faucet

June 29, 2015

Joshua E. Brown, University Communications, University of Vermont

This article is reprinted with the permission of the author, Joshua E. Brown, University Communications, University of Vermont. This article, Building the World's Smallest Faucet was originally published on May 20, 2015 at The original publication includes images and a video simulation demonstrating the quantum nature of helium at very low temperature. The computer code, developed at the University of Vermont, which produced the simulation can probe how atoms cooperate to form a superfluid at the nanoscale.

"We all know intuitively that normal liquids flow more quickly as the channel containing them tightens. Think of a river flowing through narrow rapids.

But what if a pipe were so amazingly tiny that only a few atoms of "superfluid" helium could squeeze through its opening at once? According to a longstanding quantum-mechanics model, this bizarre form of helium would behave differently from a normal liquid: far from speeding up, it would actually slow down.

For more than 70 years, scientists have been studying the flow of helium through ever-smaller pipes. But only recently has nanotechnology made it possible to reach the scale required to test the mathematical model — known as the Tomonaga-Luttinger theory (after the scientists who developed it) — in the real world.

Now, Adrian Del Maestro, a professor of physics at the University of Vermont, has collaborated with a team of researchers from McGill University and Leipzig University in Germany, to successfully create the smallest channel yet — less than 30 atoms wide.

In results published May 15 in the journal Science Advances, Del Maestro and the other researchers report that the flow of superfluid helium through this miniature faucet does, indeed, appear to slow down.

"Our results suggest that a quantum faucet does show a fundamentally different behavior," says McGill physics professor Guillaume Gervais, who led the project. "We don’t have the smoking gun yet. But we think this a great step toward proving experimentally the Tomonaga-Luttinger theory in a real liquid."

Where physics change

Insights from the research could someday contribute to novel technologies, such as nano-sensors with applications in GPS systems. But for now, Gervais says, the results are significant simply because "we’re pushing the limit of understanding things on the nanoscale. We’re approaching the grey zone where all physics changes."

UVM’s Adrian Del Maestro used computer simulations — on parallel processors in the Vermont Advanced Computing Core located at the University of Vermont — to understand just how small the faucet has to be before this new physics emerges. "The ability to study a quantum liquid at such diminutive length scales in the laboratory is extremely exciting as it allows us to extend our fundamental understanding of how atoms cooperate to form the superfluid state of matter," he says.

Unlike ordinary liquids — water or maple syrup, for example — "a superfluid has no friction or no viscosity. It's a perfect liquid," Del Maestro says. As a result, it can flow through an extremely narrow channel; and once in motion, its cooperating atoms don’t need any pressure to keep going. Helium is the only element in nature known to become a superfluid; it does so when cooled to an extremely low temperature.

But slippery perfection has quantum limits, it seems. "The superfluid slowdown we observe signals that this cooperation is starting to break down as the width of the pipe narrows to the nanoscale," Del Maestro said, and edges closer to the exotic one-dimensional limit envisioned in the Tomonaga-Luttinger theory.

"This ‘Luttinger liquid,’ as it’s sometimes called, is a very strange state of matter," Del Maestro said. "Because it exists in strictly one dimension, it's not really a liquid, it's not really a superfluid, it's not really a solid — it's everything, all at once." At least that’s one layman-friendly way to describe what the theory suggests. "We’ve thought for a long time: wouldn't it be cool if we could figure out how to make one of these Luttinger liquids in the real world," he said, "instead of just on our computers?"

With this new experiment, the team of scientists is getting close. But building what is probably the world’s smallest faucet has been no simple task. McGill’s Guillaume Gervais hatched the idea during a five-minute conversation over coffee with a world-leading theoretical physicist. That was eight years ago. But getting the nano-plumbing to work took "at least 100 trials — maybe 200," says Gervais.

A beam of electrons as drill bit

Using a beam of electrons as a kind of drill bit, the team made holes as small as seven nanometers wide in a piece of silicon nitride, a tough material used in applications such as automotive diesel engines and high-performance ball bearings. By cooling the apparatus to very low temperatures, placing superfluid helium on one side of the pore and applying a vacuum to the other, the researchers were able to observe the flow of the superfluid through the channel. Varying the size of the channel, they found that the maximum speed of the flow slowed as the radius of the pore decreased.

An inadvertent breakthrough

For years, however, the researchers were frustrated by a technical glitch: the tiny pore in the silicon nitride material kept getting clogged by contaminants. Then one day, while Gervais was away at a conference abroad, a new student in his lab inadvertently deviated from the team’s operating procedure and left a valve open in the apparatus. "It turned out that this open valve kept the hole open," Gervais says. "It was the key to getting the experiment to work. Scientific breakthroughs don’t always happen by design!"

And from this fortunate mistake, science now has an opening small enough that it begins to make visible a "new regime of matter," Adrian Del Maestro says, "that's never been explored before.""

Link to research paper in Science Advances (peer reviewed, open access journal):


This article has been reprinted with the permission of the author, Joshua E. Brown, University Communications, University of Vermont. To view the original publication, visit