Generally, silicon lets most infrared (IR) light pass through with very little interaction. Now, researchers at the Massachusetts Institute of Technology (MIT), with collaborators from the universities of Cambridge, Albany and Canberra, et al., have used pulsed laser recrystallization to trap high concentrations of dopants in silicon, an approach known as hyperdoping. For the first time, they demonstrated that it is possible to generate a photocurrent using sub-band-gap light at room temperature, a feat previously not achieved.
“We implanted gold ions into the top hundred nanometres of the silicon surface,” MIT graduate student Jonathan Mailoa describes the new laser method. His team then used nanosecond pulse laser melting (PLM) to melt and resolidify the silicon in a single-crystal lattice a containing non-equilibrium concentration of gold. The gold atoms induce mid-gap trap states in the silicon band gap; sub-band-gap photons now can excite electrons, making the silicon responsive to infrared light.
The hyperdoping approach constitutes the first ever demonstration of impurity-mediated sub-band-gap photoresponse at room temperature. Rather than impurifying the entire wafer the method enables localised silicon modification, defining the affected area through photolithography and maintaining the silicon crystal’s original structure. The devices showed a much higher infrared absorption coefficient than previous deep-level dopants, with remarkably simple planar photodiodes, no microstructure required.
The hope is that the novel laser doping technique will enable silicon-based broadband infrared imaging arrays that work at room-temperature. “We think it is attractive because the material responds to a broad range of infrared light and simple planar photodiodes can be made out of it,” Mailoa says. Being able to use silicon here is appealing, as the common semiconductor is easy to process, abundant and less expensive than other standard optoelectronic materials. The MIT researcher believes real-time, broadband infrared imaging arrays will be the best application for gold-hyperdoped silicon, and most standard silicon device processing techniques should be compatible with the material. They also predict photoexcitation-based IR response to be faster than that in thermal-based infrared imaging systems. On the account of Mailoa, the prototype’s infrared response efficiency is still small, but already good enough for some possible applications, even without additional improvements. “For instance, I currently use this new photodetector device as an alignment tool for the infrared laser in the lab — it’s cheap, fast and sufficient for the task at hand.”
Going forward, the multi-institution team wants to improve the quantum efficiency of the infrared detector using standard photodetector engineering practices. “We need to improve the infrared detection efficiency more before it can be used to make useful room-temperature broadband infrared imaging array,” Mailoa concludes.
The pioneering hyperdoping approach is described in the paper “Room-temperature sub-band gap optoelectronic response of hyperdoped silicon,” co-authored by Mailoa and published in the journal Nature Communications.
Written by Sandra Henderson, research editor Novus Light Technologies Today
Image: Test with laser beam confirms infrared-sensitive properties of gold-hyperdoped silicon sample at room temperature. Courtesy of Dr Mark Winkler, MIT.