Researchers at Jilin University in China have created tiny, simple "robots" that are powered entirely by moisture. Made out of sheets of graphene oxide treated with a flash of light, the team made a four-legged bug bot that crawled and a claw that closed in response to changes in humidity.

To make their robots respond to moisture, the team treated sheets of graphene oxide with the chemical process of reduction, where oxygen is removed from molecules. But the key here is to only "reduce" one side of the material – that way, when the air becomes more humid, the non-reduced side absorbs more water, expanding and curling the material towards the reduced side. When the air dries out, the material flattens again.

Previous research has found that strong light is an effective way to induce reduction in graphene oxide, so the Jilin team used a quick, simple and cheap light source: the flash of a camera. To make sure only the top side of the material is reduced, the sheets need to be more than 5 microns thick, and the flash needs to be activated between 20 and 30 cm (7.9 and 11.8 in) away from it.

Doing so results in sheets of graphene oxide that will react to changes in relative humidity between 33 and 86 percent, by bending to angles between zero and 85 degrees. To help visualize the effects, the team cut the treated material into four-legged critter shapes, about 1 cm (0.4 in) wide. Switching the humidity on and off a few times, they could make the little bug crawl a distance of 3.5 mm (0.1 in) in 12 seconds, without the help of any other power source.

Using the same principle, the team also made a claw that opened and closed as the humidity changed. Eight ribbons of the material, each 5 mm (0.2 in) long, were stuck into a star shape, which would completely close after 12 seconds of exposure to moisture. When the air dried out, the claw would flatten out again in under a minute.

"These robots are simple and can be flexibly manipulated by changing the environmental humidity," says Yong-Lai Zhang, lead researcher on the study. "These designs are very important because moving and capturing/releasing are basic functions of automated systems."

In future, the researchers plan to explore ways to more precisely control the robots, in order to use them for more complex tasks. The study was published in the journal Optical Materials Express.