“The forces we can produce on the nanoscale are very large – up to several nanonewtons,” explains team leader Jeremy Baumberg of the NanoPhotonics Centre at Cambridge. “Producing such large forces on the nanoscale has proved difficult to do until now, something that has hindered all practical applications of nanobots – we simply did not have nanomachines that can move practically in liquid environments. Water, for example, feels like treacle at the nanoscale, so you need large forces to push nano-objects through it.”

Actuators (sometimes also called artificial muscles) are needed in applications like microrobotics, sensing, smart windows and walls. Researchers have already succeeded in making devices like micropropellers and DNA “origami” machines, but the forces obtained in these actuators are typically quite small at around 10 femtonewtons per square millimetre. They also work extremely slowly (on second timescales).

Forces too small

Piezoelectric materials have also been used to make actuators – for example in high-end instruments such as atomic force microscopes and nanopositioning stages, but these materials are dense, delicate, expensive, difficult to fabricate and require high voltages (of between 150-300V) to operate. This is also true for materials like electrorestrictive rubbers and relaxor ferroelectrics.

More recently, researchers have taken inspiration from biological systems, such as Escherichia coli, cilia and nematocysts, which are sophisticated nanomachines in their own right. Although molecular motors and artificial muscles have been made from hydrogels, colloids or liquid crystalline elastomers that mimic these natural machines, they are slow (again, working on second time scales) and only generate piconewton-sized forces. This is because either the energy density stored in the system is too low, or because energy release is inefficient.

Producing enormous forces within just a microsecond

Key to Baumberg and colleagues’ new actuator was the development of a new and simple way to coat metal nanoparticles with thin polymer shells (made from amino-terminated poly(N-isopropyl acrylamide, or pNIPAM). These polymers have the special property that at room temperature they attract water and the polymer strands extend out into a solution. However, heat them up just a few degrees (above a critical temperature of 32°C) using a laser beam (here with a wavelength of 525 nm) and they suddenly reject water, become oily and collapse down to a nanothin layer that then coats the nanoparticles.

“The particles then stick onto each other, forming small clusters, but as they cool back down, the water wants to flood back in, which tears the nanoparticles back apart with enormous forces within just a microsecond,” explains Baumberg. The metallic nanoparticles scatter light of different colours depending on their separation (a sort of "optical ruler"), and the researchers can observe these colours and so estimate the forces. The nanoparticles they studied are made of gold, but they could easily be replaced with silver, copper or even nickel. And since the working fluid is water, this means the set up could easily be scaled to industrial volumes.

Nanopumps and valves

According to the team, reporting its work in PNAS doi: 10.1073/pnas.1524209113, the actuating nanotransducers (ANTs) produced in this way could be used to make pumps and valves in microfluidic systems. “Microfluidic chips are really interesting for synthesizing pharmaceuticals, in biomedical sensing and in many other biomedical processes,” Baumberg tells nanotechweb.org. “Until now, all such pumps and valves needed to be made with hydraulics so you had to bring a pipe for each onto a chip, something that strongly limits the complexity of anything you can do with them. We believe we can now make pumps and valves from our ANTs, which are each controlled by a beam of light, and we can have thousands on a single chip.”

The researchers say they are now busy trying to directly measure the forces on their ANTs, which is a challenge, as well as making pistons to ensure that the force produced pushes in just one direction. “We will then look at tethering them to different nanomachine designs (like trapdoors, for instance) and make the first fully-functioning pumps and valves,” says Baumberg.