US researchers from Clemson University have made silicon-cored optical fibre using techniques compatible with mass production.
There are two main advantages in silicon cored fibre.
"First is that silicon is transparent further into the infrared than conventional glasses as such this opens up doors to applications, such as selected biomedical and sensor uses, that conventional fibres can not fulfil," Professor John Ballato told Electronics Weekly.
"Secondly, silicon has a significantly higher Raman coefficient, so the silicon core could be used to shift wavelengths out to the infrared where light sources are more difficult to make."
Wavelength extension is considerable. "Conventional glass fibres are great out to about two microns wavelength. This fibre, with further optimisation could be good from about one micron to six microns in wavelength where the silicon is transparent," said Ballato.
His silicon cored fibres are made by drawing. "Yes, that's the important part, " said Ballato. "Silicon core fibres have been made before on a limited scale. What is different here is that industry relevant mass-production fibre draw has been used so this is the important step that takes it from an academic curiosity to a potential commercial opportunity."
Traditional optical fibres have a glass core that is a very similar composition to the cladding glass. Both soften together and are drawn together as a single glass wire.
To make silicon fibres, the team uses a cladding tube as a crucible to confine a molten core.
At the temperature where the glass cladding softens and draws into fibre, the crystalline core is molten and "the liquid goes along for the ride and then solidifies as the fibre cools", explained Ballato. "It is not normal to have such dissimilar materials as done here to co-fibreise together."
With the principle proved, the team is seeking an optimised cladding material to take it to production on existing fibre drawing machines.
Early fibre is far from perfect. "Right now the amount of energy lost when the light waves move down this silicon fibre is no better than for other fibres at the longer wavelengths, said the University.
However, Ballato expects energy losses to decline significantly.
High attenuation, he thinks, is likely due to the high temperature processing and ingress of oxygen from the cladding glass into the core.
Cladding glass optimisation will include searching for one that is matched in temperature and thermal expansion to the silicon. "That should reduce the attenuation considerably," said Ballato.
Clemson is working with collaborators at UCLA, Northrop Grumman and Elmira College.
