All modern electronic devices use integrated circuits built on silicon. But silicon can’t do everything and one thing it cannot do is give out light. There are no silicon LEDs or lasers. It is rather better at detecting light, and can do so in the ultraviolet and visible regions of the spectrum, but fails in the near infrared, being unable to sense radiation longer than about 1.2 microns.

“Silicon is a very good detector in the visible and near infrared,” says Kevin Homewood of the University of Surrey, “but it lacks the ability to produce light efficiently.” Since 2009 Homewood’s team has been supported by the European Research Council in efforts to coax silicon into emitting light.

Why does this matter? There are two reasons. First, to make circuits that can emit or detect infrared light, manufacturers have to graft other materials onto silicon chips. Making it all in silicon would greatly simplify the process and reduce costs.

But the second and more important reason is that it would open the way to a silicon laser. Such a laser would make it possible to transmit signals within integrated circuits by light rather than by electrons and so replace electronic chips with faster and more efficient optical chips.

New physics

In a paper published in February 2016, Homewood and his colleagues described their discovery of “band-edge modification”, where small quantities of rare earth elements – notably europium, ytterbium and cerium – can be made to alter the electronic properties of silicon, allowing it to absorb and emit light out to mid-infrared wavelengths. “It’s completely new physics,” he says. “And it also has nice applications.”

Among those applications is, potentially, the long-sought silicon laser, but for the time being his group is concentrating on using the technique to develop infrared detectors.

“We decided the detector technology was closest to market,” he says. “We still have work to do on the laser – we’ve got close, but we’re not quite there – and will go back to it at some point. But the detector is something that’s almost usable now.”

The new silicon detector covers those parts of the mid-infrared spectrum – 1.6 to 6 microns – of most interest for sensing molecular gases such as carbon dioxide.

Conventional infrared detectors use toxic materials such as cadmium mercury telluride and need to be cooled with liquid nitrogen to operate efficiently. “We can potentially replace a lot of those materials with silicon which is non-toxic and also much cheaper and easier to manufacture,” Homewood says.

Room-temperature sensors

As the six-year ERC Advanced grant ended in 2014, the ERC awarded Homewood a further 18-month Proof-of-Concept grant to commercialise the detector technology. The Royal Society then followed with a two-year Brian Mercer Award for Innovation which Homewood, now working at Queen Mary University of London, is using to refine the technology to operate at or near room temperature. A new spin-out company is being set up to market the detectors.

It’s a big deal. The world market for mid-infrared detectors is estimated to grow to USD 5 billion by 2018, mainly in security and military thermal imaging applications but also in environmental monitoring.

Homewood says that the long-term support from the ERC allowed him to spend more of his time on the project than he would otherwise have done. “And because the ERC and the Royal Society are high prestige contracts it endorses the technology when you try to go out and get investors.

“We’ve got the patents pretty well pinned down and we’ve got some seed funding to develop a business plan that will enable us to go out and get some serious investment ready for the next stage.”