"Ultimately this kind of device might form the power source to propel a nanoscale device through a fluid," Richard Jones of Sheffield University told nanotechweb.org. "In the nearer term this type of system will find uses for the triggered release of molecules of many different kinds."

To make the synthetic muscle, Jones and colleagues exploited the shape change that occurs in weak polyacids under varying pH conditions. Bulk materials undergoing repeated expansion and contraction can suffer cracking and tearing. With this in mind, the team created a nanostructured gel that was more resilient.

The gel consisted of a polyacid matrix containing nanoscopic hydrophobic domains. The material formed by self-assembly from a triblock copolymer with end-blocks made up of glassy regions of hydrophobic poly(methyl methacrylate) and a middle block of poly(methacrylic acid).

"The simplest basis for macromolecular shape change we knew about was the collapse and expansion of a weak polyacid when the pH was changed," said Jones. "We could achieve the coupling we needed by finding a non-linear, oscillating chemical reaction that made the pH spontaneously oscillate."

To power the muscle, the team used a reaction involving potassium bromate, sodium sulphite, potassium ferrocyanide and sulphuric acid. One cycle of the reaction took about 20 minutes at room temperature, changing pH from 3 to 7.

At low pH, the matrix polyacid was protonated and neutral. But above pH 5.5, the polyacid became ionized and repulsion of like charges caused the polymer to expand to around three times its original volume. The glassy end-blocks remained unaffected by the pH change, apart from a change in the distance between them.

"Our long-term goals [are to] reproduce some of the tricks that cell biology uses in its nanoscale machines in synthetic materials," said Jones. "Of course, what we are able to do is dreadfully crude by biological standards."

Testing the synthetic muscle by using it to bend a soft cantilever revealed that it produced a power per unit mass of material of at least 20 mW/kg. This makes the synthetic muscle 10 000 times weaker than striated muscle but comparable to a eukaryotic spindle at 30 mW/kg.

"We're exploring ways to use these principles to propel sub-micron size particles," said Jones, "and we'll also be looking to find different driving chemical reactions that would allow us to power devices with higher efficiencies in a wider range of environments."

The researchers reported their work in Nano Letters.