The ratio between the length and thickness of the platelets has to be just right, Studart says. If it is too high, the platelets break when the material is stretched. If it is too low, the material is not very stiff.

The researchers chose to work with aluminum oxide platelets, which are five times as strong as the calcium carbonate platelets found in nacre. They also made their platelets thinner--about 200 nanometers across, as opposed to the 500 to 1,000 nanometers of the naturally occurring platelets--to lower the likelihood of flaws in their structure. The best average length-to-thickness ratio, the researchers calculated, is 40, so they made the platelets 5 to 10 micrometers long. "Stronger platelets allow us to use a higher ratio and therefore achieve higher strength, compared to shells, with a lower concentration of platelets," Studart says. Low concentrations are important, he says, "because that means the composite has more polymer and has a lot of [stretchability]."

This is the closest anyone has come to duplicating the mechanical structure and behavior of a natural material, says Francois Barthelat, a mechanical-engineering professor and biomimetic-materials researcher at McGill University, in Montreal, Quebec. But before the material can be used, he says, the researchers will have to develop a faster way to make it in larger quantities.

Princeton University chemistry professor Ilhan Aksay thinks that the technique should be easy to modify so that it is suitable for bulk manufacture. "You could make large shapes with this technique," he says. He imagines that the material could be useful for bone and dental implants.

Gauckler says that the material needs many improvements before it can be practically used. A better polymer would make the composite stronger. The researchers also need to find a way to get better bonding between the aluminum oxide and the polymer. For now, Gauckler says, "we have shown that we can [come close to] doing as good a job as nature."