Massachusetts Institute of Technology researchers said Monday they've discovered a way to improve the efficiency of thin-film crystalline silicon solar photovoltaic cells by as much as 50 percent.

But don't expect those gains to hit the commercial market right away, since the materials and processes used in the experiment are still too expensive for mass production, a researcher involved in the project said.

Still, the project's results are an important step forward in efforts to make solar cells that can capture specific wavelengths of light, said Peter Bermel, a postdoctoral researcher in MIT's physics department who worked on the project.

In this case, that's red and near-infrared light, which can make up about a third of the "useful" light from the sun, he said.

To capture that spectrum of light, the MIT researchers combined a multi-layered reflective coating and a tightly spaced array of lines called a diffraction grating to the back of the solar cell, he said. They also used an antireflective layer on the front of the cell.

"Just having reflection coatings isn't that novel," Bermel said. "It's the combination of these different elements in this configuration that makes it unique."

In simple terms, that combination traps light in the cell longer, giving it a better chance of being converted into electricity, he said. It took thousands of computer simulations to perfect the spacing of that diffraction grating, as well as the best thickness of silicon and the reflective layers, Bermel said.

But lab tests conducted by MIT materials science and engineering graduate student Lirong Zeng showed that cells prepared under those specifications could boost the efficiency of turning sunlight into electricity to an average of 15.6 percent – up from a typical 10-percent efficiency for cells made of similar materials and thickness that lacked any light-trapping characteristics, Bermel said.

The process used to make the experimental cell, however, is "not production-ready at this point," he said. First, the experiment high-quality silicon costs about $9.50 for a six-inch wafer – a price more suited to computer chips than solar cells, Bermel said.

Second, the diffraction grating was created using a process called holographic lithography, which today only can be done by a few specialized facilities, he said. Also, the coatings on the experimental solar cell were laid down by high-quality chemical vapor deposition systems that probably aren't economical to use in mass production today, he said.

In the longer term, however, "The goal would be to make the cost low enough so that the installed cost in dollars per watt would be lower using this technology than not using it," he said.

The MIT research results are for crystalline silicon cells of 5 microns in thickness, Bermel said. Commercial crystalline silicon solar cells can have thicknesses ranging from 1.3 microns anywhere up to 300 microns, with thin-film cells generally defined as being 50 microns or less, he said.

That could make the MIT research relevant to the range of thicknesses seen in solar cells from manufacturers including CSG Solar and Oerlikon Solar, he said.

This isn't the first time researchers from MIT have boasted big potential gains in solar cell efficiency. In July, a team of MIT researchers said they had identified organic dyes that could also increase internal refraction within solar cells, increasing their efficiency by as much as 50 percent. The researchers planned to start a company, Covalent Solar, to develop the technology (see Dyeing for More Solar Power).

Another MIT spinoff, 1366 Technologies, plans to seek up to $50 million to build a plant that CEO Frank van Mierlo says could produce multi-crystalline solar modules at costs of 25 percent lower than its competitors (see 1366 CEO Lays Out Strategy, Funding Plans).