MARINE sponges are humble creatures, but some of them have a special talent. They can harvest silicon from seawater, and use it to build the spiky filaments that cover their body. Now this process has inspired the development of a cheap, low-energy method of manufacturing solar cells.

The usual way of making solar cells involves techniques such as vapour deposition, in which chemicals are laid down on an inert surface, or "substrate", to create a layer of crystalline semiconductor that produces an electric current when light strikes it. This is done at high temperature and very low pressure, making it an expensive and energy-intensive process.

Daniel Morse, a molecular biochemist, and his colleagues John Gomm and Birgit Schwenzer at the University of California, Santa Barbara, knew that certain sea sponges like the orange puffball sponge Tethya aurantia naturally synthesise pristine layers of silica without benefit of high temperatures or low pressures, and at near-neutral pH. T. aurantia does this using an enzyme called silicatein to catalyse the conversion of silicic acid in seawater into its silica spikes.

Structures like this are known to make photovoltaic cells more efficient, so the researchers set about developing an analogous low-energy process that produces structured layers of zinc oxide - a widely used solar cell semiconductor. They replaced the seawater with aqueous zinc nitrate and the silicatein with ammonia, which catalyses the breakdown of zinc nitrate into crystalline zinc oxide.

By regulating the rate at which they diffused the ammonia into the zinc nitrate, and hence the pH of the solution, they were able to control the structure of the zinc oxide film that was slowly deposited on a glass substrate in the reaction chamber. In this way, they have built crystalline layers ranging in thickness from 100 to 300 nanometres, they will tell an American Chemical Society meeting in Chicago next week.

Though they have succeeded in building primitive working solar cells, Schwenzer says the technique still needs further development. "There are still problems but the process seems to be working at really low temperatures and producing devices at really low cost."