With Laser-Doping, Silicon Responds to IR Light
CAMBRIDGE, Mass., Jan. 7, 2014 — New IR imaging systems could be possible now that a new method has demonstrated that silicon is more responsive to IR light when laser-doped with one of its most dangerous impurities: gold.
Detectors that are responsive to a broad range of IR light could form imaging arrays for security systems or solar cells that use a broader spectrum of the sun's energy. But efforts to develop such detectors using silicon have faced limitations — until now.
A laser beam is used in the lab to test the gold-hyperdoped sample of silicon and confirm its IR-sensitive properties. Courtesy of the researchers.
The method, developed at five institutions, including MIT, uses silicon, an abundant material that is relatively cheap, easy to process, and forms the basis of most semiconductor and solar cell technology. Normally, silicon lets most IR light pass right through it because its bandgap requires an energy level greater than that carried by photons.
"Silicon usually has very little interaction with infrared light," said Tonio Buonassisi, associate professor of mechanical engineering at MIT.
The usual work-around for this is to create a waveguide with structural defects or to dope the silicon with certain other elements. But these and other such methods have significant negative effects on silicon's electrical performance, work only at very low temperatures, or only make silicon responsive to a very narrow band of IR wavelengths.
The new system works at room temperature and provides a broad IR response, Buonassisi said. It works by implanting gold into the top 100 nm of silicon and using a laser to melt the surface for a few nanoseconds. The silicon atoms recrystallize into a near-perfect lattice, and the gold atoms don’t have time to escape before getting trapped in the lattice.
The material contains about 1 percent gold, an amount more than 100 times greater than silicon’s solubility limit — a layer of silicon supersaturated with gold atoms.
“It’s still a silicon crystal, but it has an enormous amount of gold near the surface,” Buonassisi said. Although others have tried similar methods with materials other than gold, this work is the first clear demonstration that the technique can work with gold as the added material, he said.
“It’s a big milestone; it shows you can do this,” said MIT graduate student Jonathan Mailoa. “This is especially attractive because we can show broadband infrared response in silicon at room temperature.”
This use of gold was a surprise: Usually, gold is incompatible with anything involving silicon, Buonassisi said. Even the tiniest particle of it can destroy the usefulness of a silicon microchip — so much so that in many chip-manufacturing facilities, the wearing of gold jewelry is strictly prohibited. “It’s one of the most dangerous impurities in silicon,” he said.
But at the very high concentrations achieved by laser doping, Buonassisi said, gold can have a net positive optoelectronic impact when IR light shines on the device.
Although this is early-stage work, for some specialized purposes — such as a system for adjusting infrared laser alignment — it might be useful relatively quickly. But its efficiency is probably too low for use in silicon solar cells, Buonassisi said. However, this laser-processing method might be applicable to different materials that would be useful for making solar cells.
The work was published online Jan. 2 in Nature Communications
) by Mailoa, Buonassisi and 11 others.
It was funded by the US Army Research Office, the National Science Foundation, the Department of Energy, and the MIT-KFUPM Center for Clean Water and Energy, a joint project of MIT and the King Fahd University of Petroleum and Mining.
For more information, visit: www.mit.edu