Researchers at Eindhoven have finally developed photonic silicon

Cal Jeffrey

Posts: 4,178   +1,424
Staff member
Forward-looking: Scientists have been looking to create light-emitting silicon for over 50 years. It had become something of a "Holy Grail" in the microelectronics community. Such a discovery would mean faster on-chip communication, lower heat production, and higher power efficiency.

Now researchers at Eindhoven University of Technology (TU/e) have figured out this decades-long conundrum by creating a new hexagonal silicon alloy capable of emitting light. The hexagonal shape is key to creating a direct bandgap that will emit photons.

"The crux is in the nature of the so-called bandgap of a semiconductor," said TU/e project lead Erik Bakkers. "If an electron 'drops' from the conduction band to the valence band, a semiconductor emits a photon: light."

In traditional cubic silicon, the conduction and valence bands are displaced creating an indirect bandgap, so no photons can be emitted. However, it was theorized 50 years ago that alloyed silicon and germanium in a hexagonal configuration would have a direct bandgap. The trick is creating such an alloy.

This feat had been impossible until the development of naotubes and wires was discovered. The team was able to create hexagonal silicon in 2015 by growing nanowires from another material and using it as a template to develop hexagon-shaped silicon with a germanium shell.

"We were able to do this such that the silicon atoms are built on the hexagonal template, and by this forced the silicon atoms to grow in the hexagonal structure," said Elham Fadaly, coauthor of the team's paper, which was published in Nature.

The researchers now need to develop a silicon-compatible laser. According to Bakkers, they could have one before the end of this year.

"If things run smoothly, we can create a silicon-based laser in 2020," he said. "This would enable a tight integration of optical functionality in the dominant electronics platform, which would break open prospects for on-chip optical communication and affordable chemical sensors based on spectroscopy."

Since photons are not subject to resistance and have less scatter within the conducting medium, no heat is produced, so power consumption is significantly reduced. Furthermore, the on-chip and chip-to-chip communication speeds could increase by a factor of 1,000 on future photonic silicon. The technology has many applications, including laser-based radar for autonomous vehicles and chemical sensors used in both the medical and food industries.

Image credit: TU/e

Permalink to story.

 
Chip-to-chip data transfer rates could increase 1,000 fold

Well they tell us it "could" happen in 2020 but no mention of the trickle down cost. While it would eventually become the standard for manufacturing of chips the impact on cost could largely determine it's success or failure. No doubt the high end applications will be willing to foot the bill but by the time it gets down to the everyday consumer if it's too high it might not catch on.
 
Chip-to-chip data transfer rates could increase 1,000 fold

Well they tell us it "could" happen in 2020 but no mention of the trickle down cost. While it would eventually become the standard for manufacturing of chips the impact on cost could largely determine it's success or failure. No doubt the high end applications will be willing to foot the bill but by the time it gets down to the everyday consumer if it's too high it might not catch on.

I suspect this technology will be mostly limited to the military and large enterprise sectors for close to a decade after hitting the market. Data centers will easily be able to justify the investment considering the long-term cost savings. It also creates some interesting possibilities if fiber networking were made more practical. Pure optical links between physically distant systems would be a quantum leap for cloud applications.
 
Back