Researchers at Sandia National Laboratories have shrunk silicon solar cells down to the micro scale, opening new possibilities for improved efficiency.

Multi-crystalline silicon, currently the gold standard for solar-cell efficiency, is expensive and produces cells that are heavy and brittle. Sandia’s microscopic silicon solar cells use 100 times less material while operating with the same efficiency.

In addition to lower materials costs, the smaller scale of these cells means they could be incorporated into compact optical systems for cheaper light-tracking and concentration. Researchers might even suspend them in inks that could be printed onto plastic to make efficient, flexible silicon-solar modules.

“In microsystems, you’re looking for things that become cheaper, perform better, and gain new functionalities,” says Gregory Nielson, head scientist on the project.

So far, the Sandia researchers have assembled and tested a single micro solar cell as proof of principle. But they have begun testing functioning solar modules made from multiple tiny cells and are developing techniques for assembling them efficiently.

Sandia’s cells are between 0.25 and one millimeter in diameter. The main benefit of manufacturing such small cells would be lower materials costs, since the tiny cells can be made about 10 times thinner than conventional ones. Ordinarily, solar cells must be 100 micrometers thick to support their surface area–typically about 15 centimeters square.

Sandia makes its cells from silicon that has been processed using conventional chemical methods. Researchers carve the cells out of this silicon using a chemical etching technique that creates negligible waste. They treat the surface of the wafer to create the electrical properties necessary for a functioning cell, then top it with metal contacts. Researchers then etch the top 10 to 20 micrometers of the wafer surface using chemicals that only eat into a particular part of the crystal structure.

The resulting cells are about 20 micrometers thick but have the same efficiency as conventional cells, converting about 14.9 percent of sunlight into electrical energy. It’s also easier to make the cells in a hexagonal shape, which makes the most of the available area without wasting much silicon. “The materials savings are a big deal,” says Nielson.

Microscale solar cells offer new possibilities for light concentrating and tracking, which could further boost the cells’ efficiency. Conventional tracking systems are large and heavy and have to be moved by motors. An array of micro solar cells could be topped with a microlens array that needs to move only a fraction of a millimeter to track the sun.

The microscopic cells could also be combined with more efficient lenses. Instead of Fresnel lenses, which are bulky and capture only about 80 percent of the light that hits them, the micro cells could use refractive lenses, which capture 90 percent of incoming light. It isn’t practical to use refractive lenses with conventional solar cells because such lenses would become too expensive and bulky at the size required (the larger the lens, the farther it must be mounted from the surface of the cell). But for the Sandia cells, a refractive microlens array could match each silicon device with one lens just a few micrometers in diameter. Such arrays are already commercially available.

Nielson says developers could eventually suspend the cells in a liquid to make an ink that could be printed onto flexible substrates coated with electrical contacts to create flexible solar modules.

For solar cells, flexibility usually hurts efficiency. For example, the company Konarka makes flexible solar cells from organic materials, but these only operate at about 4 percent efficiency. “We think we can use high-efficiency materials to provide the same flexibility using five times less area,” says Nielson.

Nielson expects the project, which is funded through the U.S. Department of Energy Solar Technologies Program, to yield modules for military use (for example, in energy-harvesting tents and backpacks) in about three years. The rest of the solar market has stringent lifetime requirements, so it may take another few years to develop modules that are durable enough. The national lab will likely license the technology to a company after it’s more mature.