The potential of microfluidic technology in medicine has been recognised for decades. Put a minute amount of blood or saliva into the tiny, hair-width channel on a microchip and it can interact with a reagent to detect a biomarker and produce a diagnosis. This lab-on-a-chip is – in theory – highly portable and produces an instant result, enabling people to be diagnosed within their own homes, and avoiding the need to send samples to a laboratory. Such a device would be particularly useful in developing countries with limited laboratory facilities.

Over the past year or two, most of us have been using microfluidic devices regularly. The lateral flow tests used to diagnose Covid using just a nasal sample can now be found in almost every home. But these paper tests, while cheap and easy to mass produce, have low sensitivity. Diagnosis of many diseases requires a more sophisticated version of the lab-on-a-chip, using tiny channels encased in glass, silicon or polydimethylsiloxane.

Deploying this type of microfluidic device successfully, however, involves overcoming the problem of frictional drag, caused by the fact that the liquid sticks on the wall of the capillary. The smaller the capillary, the bigger the problem: more of the liquid, relatively speaking, sticks to the solid walls. “When we start to go down to capillaries that are maybe 100 microns in diameter or less, then this effect becomes really strong,” says Chiara Neto, professor of physical chemistry at the University of Sydney. “And then it’s much more relevant to try and minimise that friction.” 

Solving the friction problem

To overcome the friction problem, the solution has so far been to use high pressures to drive the flow. As well as being inefficient, this high pressure can damage delicate samples in the device, such as cells and other soft materials. The solid walls can become fouled by biological molecules or bacteria, leading to fast degradation. A paper published earlier this year in Nature Communications, however, suggests that Neto and her University of Sydney colleagues, have found a way of reducing that frictional drag. The four-strong team consists of Neto herself, Chris Vega-Sanchez, a PhD student whose work focuses on this problem, Dr Sam Peppou-Chapman, an expert in liquid-infused surfaces, and Dr Liwen Zhu, an expert in atomic force microscopy, which enables scientists to see down to a billionth of a metre.

The problem of frictional drag is one that Neto herself has been thinking about since taking her PhD 20 years ago, and she has worked on it on and off ever since. “What I’m interested in is how we can tune the interaction of liquids with solids as we change either the surface structure or the surface composition of the solid,” she says. Over time she and her colleagues have made a lot of progress, culminating in the discovery made last year: “We’ve now probably hit on the most successful story that we’ll find on reducing drag.”

The lotus leaf effect

It is already well-known, Neto explains, that “by applying a microscopic or nanoscopic surface structure on a hydrophobic surface, you can make water become really highly repelled by this surface”. This is known as the “lotus leaf effect” because when water comes into contact with a lotus leaf, the droplets roll off, leaving the surface of the leaf dry. Although this “spectacular” effect has been observed and investigated for 20 years, it is very short-lived, Neto says: “As soon as the surface is created, it tends to become very weak if the surface, for example, is exposed to pressure.” This, she says, is because “this super-repellence comes from the surface trapping more pockets of air, and so when you press water against it, the air typically re-dissolves into the water, and you’ve lost the super hydrophobicity”.

So how can scientists retain the hydrophobic effect? The answer, says Neto, is to “use the same type of structure but replace the air pockets with another liquid, a lubricant that is immiscible with the liquid that you’re trying to repel”. Her team tackled the problem by creating a “highly wrinkled nanostructured surface that traps a silicone oil”. (The wrinkled surface was made of Teflon.) “The silicone oil spreads fully on this surface, and fills all the gaps in between the structure, and creates a smooth liquid interface on which water can then flow really easily,” she explains. In effect, the solid wall of the channel is replaced by a liquid wall, thus reducing the hydrodynamic drag.

A puzzling find

The premise of the study – which proved to be correct – was that using the silicone oil would create a slippery surface that reduced frictional drag. In fact, the effect turned out to be 50 times greater than the team had anticipated. These new slippery surfaces reduced the drag by up to 28% – something that would be expected only if the surface was infused with air rather than a viscous lubricant. It was “really puzzling,” says Neto: “The effect that we were seeing could not be justified based on the lubricant presence alone. The lubricants that we were using were more viscous than the water we were flowing on top of it, and so the existing theory did not explain it.”

The challenge, says Neto, was to find out what was causing the effect. Using atomic-force microscopy to scan the surfaces underwater, they imaged the spontaneous formation of nanobubbles on the surface. (Nanobubbles are only 100nm – that is, 100 billionths of a metre – high).

It was these nanobubbles that accounted for the scale of the effect. “The surface partly converts back to a super hydrophobic state,” Neto says. It does this by capturing trapped air from the water, thus becoming a “mixed lubricant interface, which is even more effective at resisting drag”. The research also provides a solution to the problem of fouling. In separate studies, the team investigated the attachment of bacteria and larger organisms on the same surfaces. The surfaces proved “extremely effective at preventing the attachment of both bacteria and larger algae and organisms,” she says.

For microfluidic medical devices, the benefits are clear. Traditionally, if a delicate sample was placed in a microfluidic channel, the risk was that the high pressure needed to push the flow through the device would damage the sample. Because the method developed by the Sydney team reduces the frictional drag, there is no longer a need to apply such high levels of pressure. Another advantage, says Neto, arises in cases where the microfluidic device requires protein adhesion on the test sample, but not anywhere else.

A compelling reason to change?

So, does Neto think that her team’s discovery will change the way microfluidic devices are manufactured? She agrees that the principle could be adopted but adds that for mass manufacture it would be necessary to use different fabrications, as the ones used by her team were “lab scale” and couldn’t be used commercially.

She is also cautious, pointing out that microfluidic devices are built with traditional materials and that it’s not that easy to get people to change the materials from which they fabricate those devices. There would, she adds, have to be a “very compelling reason to do it” and she thinks the most likely will be a biological study that requires gentle pressure to be applied to the flow, and therefore needs a liquid wall to replace the solid one. In a medical microfluidic device, in which a minute volume of liquid reacts with the solid surface of the channel, she says it is “important that you control how that flows, because you don’t want to lose your samples on the walls”.

Neto’s principal interest is less in the commercial application of her discovery than the science itself: “It’s more the fundamental understanding that has been really rewarding, being able to fully understand what’s happening in this complex system.” She has, however, considered the ways in which the findings could have implications elsewhere, pointing out, for example, that oil is sometimes extracted from rock by pushing water through it. If the walls of the rock are covered in nanobubbles, that would decrease the effectiveness of the removal. Similarly, she says, there is evidence that having gas dissolved in oil or water can destabilise an emulsion: “Emulsions are everywhere, from paint to the food industry to pharmaceuticals.” Ignoring the presence of the gas could lead to errors in modelling these systems.

Yet it’s clear the discovery has major implications for anyone interested in how microfluidic devices could be used in medicine. Until now the problem of frictional drag – along with the cost of mass production – has held back widespread adoption. Solving that particular problem could be an important step on the path to turning the long-held dream of a universally available lab-on-a-chip into a reality.


Using silicone oil creates a slippery surface that can reduce frictional drag by this amount – up to 50 times greater than the team anticipated.

Nature Communications

Slip and no-slip conditions

When a liquid flows through a channel, its velocity at the channel wall is reduced as a consequence of its interaction with the wall, an effect called frictional drag. In macroscopic flows, a no-slip boundary condition is usually assumed, for example, the liquid relative velocity is expected to be zero at the wall. However, in the past two decades evidence of nanoscale interfacial slip has emerged in situations when the flow is highly confined and the wettability of the solid by the liquid is poor. Slip is quantified using the slip length b, the distance beyond the interface at which the liquid velocity linearly extrapolates to zero. The larger the slip length b, the larger the reduction in frictional drag.

Superhydrophobic and lubricant-infused surfaces (LIS) have been shown to reduce drag substantially, which makes them attractive to reduce the energy required to drive flow. Drag reduction by these surfaces is explained with the reduced contact area between the solid and the fluid, which results in an “apparent slip” of the flowing liquid of viscosity μw over the air (in the case of superhydrophobic surfaces) or over a lubricant of lower viscosity μo in LIS, compared to the case of a solid surface.

Source: Vega-Sánchez, C., Peppou-Chapman, S., Zhu, L. and others. ‘Nanobubbles explain the large slip observed on lubricant-infused surfaces’.