To make an optical fiber with low loss in the mid-infrared (MIR) wavelength region, it is absolutely critical to make the fiber of a material which itself has low loss in the MIR, right? Actually, no. Sometimes the structure you make with your building blocks at hand is more important than what material your building blocks are made of. The paper by Yu and colleagues from University of Bath is a good example of this. Their goal was to make a fiber with transmission in the MIR (2-25 µm), because many molecules have distinct strong vibrational absorption lines in this wavelength region. The technology for fabricating fibers is extremely well-developed for silica (SiO2) fibers, but the material itself has losses of more than 60,000 dB/km above 3 µm, so it is usually not even considered for MIR-applications. Other glasses, such as lead-silicate, chalcogenide and fluoride glasses, have lower losses in the MIR, but are notoriously more difficult to process into fiber, even though they can be drawn at lower temperatures. For example, the viscosity of these so-called soft-glasses is much more sensitive to temperature than that of silica, making controlled fabrication difficult.

However, the fiber does not need to confine the propagating light in the solid material of which the fiber is made. The perhaps most well-known example of this, is the concept of photonic bandgap guidance. Fibers guiding light using a photonic bandgap typically confine the light to an air-core running along the length of the fiber, while a carefully structured array of smaller air-holes surrounding the core ensure that the light cannot escape. Since only a tiny fraction of the propagating light is in contact with the solid material of the fiber, the loss is more determined by how well the microstructured air-holes confine the light, than by the bulk loss of the material.

The silica fiber made by Yu and colleagues does not have the microstructure required to form a photonic bandgap; instead it consists of just one ring of air-holes surrounding the air-core, but the silica core wall has a negative curvature. It was explained previously (Wang et al., Opt. Lett. 36, p. 669, 2011), that the negative curvature of the core helps to further reduce coupling from modes in the core to the already low number of modes in the cladding of these types of fibers. This guidance mechanism does not lead to losses as low as could be achieved using a bandgap, but the advantage is that the loss remains reasonably low over a much broader wavelength range than possible in a bandgap guiding fiber.

The Authors report measuring a loss of only 34 dB/km at 3050 nm, which is quite impressive when considering the roughly 60,000 dB/km loss of the bulk silica. One should keep in mind that this loss figure represents a measurement of many modes propagating in the core, since this fiber design also does not lead to single-mode guidance. Another disadvantage is that the fiber is very bend sensitive; the Authors report finding a significant increase in loss when the bend diameter is around 30-40 cm or less. Nevertheless, the work is an important step towards practical fibers for the MIR. It is also an important reminder that one does not always need to choose to work with a material which in itself has the best possible optical properties, if instead its mechanical properties allow the formation of advanced structures with better optical properties than the bulk material.

---

---

 Add Comment

You must log in to add comments.