The high transparency of silica at visible wavelengths and longer (up to about 2 μm) makes it possible to guide light at a wavelength of 1550 nm in a 15-km-long optical fiber and still only lose about 50% of the power (a loss of 0.2 dB/km). However, silica also has its weaknesses. The transparency is very low for more infrared wavelengths (>2 μm) owing to vibrational resonances of the material. Also, the nonlinearity is very low, which is usually good for communication purposes, but not an advantage for, e.g., making Raman fiber lasers.
On the other hand, silicon is a semiconductor with a bandgap of only 1.1 eV, making it useful for microelectronic circuits, but also leads to high optical absorption just around the corresponding photon wavelength of 1.1 μm. However, there is a low-loss transmission window from about 1.2 to 6.5 μm, with losses in the order of 0.1 dB/cm. Furthermore, the Raman gain is about 1000 times higher than in silica. For these reasons, it has been suggested that silicon for photonics in the mid-infrared be explored. Until recently, the main vision was to extend the well-known technology of silicon chip-scale electronic devices to planar photonic devices, which usually only allows a few centimeters of interaction length. That changed when John Ballato (Clemson University) and co-workers in 2008 demonstrated the first successful drawing of a silicon core fiber. The door was now open for simple, low-cost fabrication in a robust geometry over significant propagation lengths, thereby bringing within reach the goals of Raman fiber lasers and power delivery in the mid-infrared. The combination of silicon and optical fibers also naturally gives inspiration for future devices that can both perform ultrafast optical processing of the information while fitting seamlessly together with the current silica fibers. Unfortunately, these advances have not been just around the corner owing to remaining technological difficulties. One critical requirement is (transverse) single-mode operation of the silicon fiber. For a “standard” step-index design consisting of a high-index silicon core surrounded by low-index silica cladding, this requires a core size much smaller than the silica core used in telecommunications fiber, thereby hindering integration and low-loss coupling. However, the waveguide properties can be controlled to a high degree when moving from the step-index design to microstructured optical fibers (MOFs). Silica MOFs are commercially available and consist of pure silica with an array of microscopic air holes along the fiber; the holes both confine the light and control the waveguide properties. Researchers from University of Southampton, UK, recently demonstrated that MOFs can be filled with silicon to form a silicon–silica MOF (SSMOF) and that the microstructure design made it possible to only support two modes in the fiber.
Now, researchers from the same group have made another important advancement. They have carefully analyzed the guiding properties of SSMOFs. The understanding gained from this leads them to suggest selectively filling the cladding holes of an SSMOF with a raised-index semiconductor (e.g., doped silicon). The slight increase in refractive index of the cladding relative to the core is found to allow effectively single-mode operation of the fiber while still having a core size comparable to that of telecommunications fiber. This work is therefore another important step on the way to devices that integrate existing fiber networks with new semiconductor fibers and for efficiently pumped Raman fiber lasers for the mid-infrared.
You must log in to add comments.