Optical fibers have traditionally been used as light pipes, and it is this function that drives their two primary application areas, telecommunications and remote imaging. However, as has been shown over the past few decades, fibers are actually an excellent platform for making a wide variety of optical devices, such as lasers, amplifiers, diffraction gratings, Bragg mirrors, wavelength converters, beam shapers, pulse compressors, and chemical and physical sensors, to name a few. The advantages of making devices from fiber are manifold: fibers are already manufactured on a industrial scale, mechanically robust, easy to package and interface with existing light sources and detectors, relatively insusceptible to dust contamination, and do not require alignment as with free-space systems. One notable limitation is that fiber devices are typically limited to a 1-D geometry. 2- or 3-D guided wave photonic devices can be made by fusing and tapering multiple fibers, at the expense of device compactness, or with multicore fibers, at the expense of more complex input and output coupling. They are often made in a planar chip platform lithographically or by laser direct writing, but these are more challenging to produce with low propagation loss and to interface with other elements.

Ultrafast laser direct writing can be done in pure bulk silica glass. An optical fiber is essentially a small rod of doped glass embedded in a much larger pure silica rod, so can one direct write a photonic circuit in a fiber? Researchers at the University of Toronto have shown most certainly yes. In a paper published in Optics Express, Lee et al. used a femtosecond laser to write waveguides, evanescent couplers, S-bends, and Bragg gratings in a coreless fiber spliced to a standard single mode fiber. They combine these elements to create a shape sensor, where off-axis sensing elements (Bragg gratings in this case) discriminate the direction of bends. Their target application is for endoscopy, where such a sensor is can detect the path of a catheter and guide it to a target area. The sensor shows a remarkable improvement over previous work in that it is made from a single fiber, does not require any specialized bonding or packaging, and can launch and collect light through a simple splice.

The device splits the single mode fiber input into three waveguides, each of which in turn has three Bragg gratings which reflect discrete wavelengths. Bragg gratings are sensitive to strain/compression and temperature, which perturb both the physical period of the grating and the optical path of the glass, the latter via the stress-optic and thermo-optic effects. By monitoring the nine calibrated reflection peaks on a fast spectrometer and processing the data with an interpolation algorithm and finite element code, the authors demonstrate real-time distributed shape and temperature sensing. They sense curvature up to of 25 m^-1 with a precision of 1.1 m^-1, temperature up to 250° C with a precision of 5° C, and sub-mm accuracy in position. The femtosecond writing process still presents a number of technical hurdles, in particular birefringence, which limits the sensor resolution due to polarization splitting of the Bragg resonances, and high propagation losses of 0.5 dB/cm, which limits the device length. However, even if these issues cannot be readily solved, the ability to direct-write a 1xN splitter in fiber can provide spliceable and efficient excitation of multicore fibers, which have orders of magnitude lower birefringence and loss. If these issues can be solved, this work shows a route towards a novel class of fiber devices having the functionality of integrated optical circuits.

---

---

 Add Comment

You must log in to add comments.