Photonic crystal fibers (PCFs) or "holey" fibers were first developed in 1996 and have subsequently generated enormous interest.1 Important features associated with such fibers are their ability to remain single-moded over a very large frequency range, to be single-moded with a large mode area, and to guide light in air (see oemagazine, June 2002, page 25). Modifications of the hole structure have allowed a number of specialty applications to be realized; for example, highly birefingent or polarization-maintaining fibers. Their dispersion properties have also attracted attention, particularly because of the possibilities they offer for dispersion compensation. By changing the core size of the fibers, it is also possible to make them have either very low or very high optical nonlinearity. A recent use of this property was the development of a fiber system for supercontinuum generation.
Last year our group at the Optical Fibre Technology Centre, University of Sydney (Sydney, Australia) succeeded in producing a PCF in polymer for the first time.2 This development complements the glass PCF technology in several ways. Polymer PCFs leverage the tailorizable properties and processing opportunities of polymer materials. Moreover, polymer may allow PCFs to be used in application areas in which flexibility, weight, and connectivity issues are crucial, for example, automotive and short-haul data communications. microstructured optical fibers
Glass PCFs are made using a method based on capillary stacking with successive drawing to produce a structure of the appropriate dimensions. This technique yields structures that reflect the stacking properties of the capillaries, generally having either hexagonal or square lattices. Although polymer PCFs have been made by capillary stacking, it has generally been found more convenient to use the large variety of polymer processing techniques such as casting, forming, molding, and extrusion to produce a "monolithic" preform in which all the desired structure is included in a single piece of material. This preform is then drawn to the desired dimensions.
GIMPOFs use interpenetrating rings of holes to approximate a parabolic refractive index profile.
With this process, it is possible to make structures that are hard, if not impossible, to make by capillary stacking. Capillary stacking produces structures that reflect the stacking properties, for example for circular capillaries you always get structures with hexagonal symmetry. With polymer processes, we can easily make circular rings of holes, for example, or other arbitrary hole structures (see figure). Although these structures do undergo deformation as a result of the drawing process, the process is in general a more forgiving one in polymer than in glass because of polymer chain alignment effects and the reduced surface tension, both of which help preserve the structures.
The combination of the monolithic preform and the robustness of the polymer structure to the draw process has allowed new types of microstructures to be produced, including those with noncircular holes and nonperiodic structures. One area that has been a particular focus of interest is holey ring structures, in which the holes are arranged in the form of a ring.3 These microstructured polymer optical fibers (MPOFs) have been shown to approximate Bragg fibers, and even a form of structural graded-index fiber (GIMPOF). GIMPOFs in particular are of great interest because they represent the microstructured version of one of the major markets for polymer fibers. dealing with loss
High loss is the main disadvantage of optical polymers. For short-haul applications, however, polymer fibers have a real role to play because ease of connectivity becomes a determining factor for such distances. Conventional polymer fibers have very large coresalmost a millimeter in diameterand use a graded index profile to reduce the dispersion.
GIMPOFs are such a new concept that their properties are still being unraveled, but initial experimental results, still very preliminary, are promising (see figure, below). Although modeling their properties poses considerable challenges, MPOFs offer potential advantages over conventional polymer fiber technology. MPOFs do not rely on chemical doping to achieve guidance and thus avoid some of the problems associated with diffusion and refractive-index fluctuations due to local variations in the dopant concentration. Scattering caused by these fluctuations is a major source of loss in polymer fiber. As an example, Asahi Glass Co. (Tokyo, Japan), which makes the lowest-loss polymer fibers on the market, has quoted a loss of 17 dB/km for its fibers at 850 nm, of which 13 dB/km is due to scattering.4
Microstructures themselves introduce new sources of loss, such as confinement loss (due to the bridges between the holes) and surface scattering. However, the progress of glass PCFs, which have achieved losses as low as 0.6 dB/km, gives cause for cautious optimism. Similar progress would be relatively more dramatic in polymer fibers, as the switch to microstructured fibers alone affords a significant loss improvement over standard POF, materials improvements notwithstanding. Essentially, the losses introduced by the microstructure (surface scattering and so on) should not exceed the reduction in loss we achieve by eliminating dopants and therefore reducing material scattering. The figures suggest that this is true. Avoiding dopants may also raise the working temperature range of the fibers, which is currently limited by diffusion. We hope to be able to produce a relatively large-core GIMPOF (probably still smaller than 1 mm) with loss and dispersion properties that exceed those of conventional graded-index polymer fibers. The impact of such a fiber in local area networks could be dramatic indeed. polymer materials
From a materials perspective, MPOF also opens up new possibilities. In glass fibers, the possibilities for modifying the properties by doping are limited both by the need to avoid phase separation and by high processing temperatures (2000° C), which cause many materials to decomposefor example, organic materials such as dyes decompose above 400°C. The processing temperatures of MPOF are on the order of 200°C, which expands the choice of available dopants.
This temperature-based material flexibility has been available in conventional polymer fibers for years but the challenges involved in producing single-mode polymer fiber have limited their use. Microstructured fiber, on the other hand, can be designed to be inherently single mode. In addition to the obvious benefits, this characteristic allows designers to separate the guidance mechanism from the presence of dopants. Thus, MPOFs allow designers to vary optical properties independent of each otherdispersion and optical nonlinearity, or gain and birefringence, for exampleto an unprecedented degree.
Polymers are intrinsically modifiable. Using guest-host systems, co-polymers of surfactants, and inclusions, it is possible to substantially modify the properties of the polymer and include high levels of active components. There has been, for example, a sustained effort to develop polymers for nonlinear optics or the electro-optic effect. We have recently shown that it is possible to incorporate tungsten electrodes into an MPOF as part of the fiber-drawing process without significant damage to the microstructure. It is therefore straightforward to produce poling in a doped MPOF system. Depending on the size of the effect and its thermal stability, applications include switching, sensing applications, and second harmonic generation. Our group continues to focus on this work.
A more exotic material modification we are currently working on is producing a chiral polymer system, which will allow us to produce a circularly birefringent fiber. MPOF technology may also be well suited to a variety of other applications, such as sensing and polarization control. oe
1. J. Knight, T. Birks, et al., Opt. Lett. 21, No. 19, pp. 1547-1549, (1996).
2. M. van Eijkelenborg, M. Large, Optics Express, 9 (7), 319-327, September 2001.
3. A. Argyros, I. Bassett, Optics Express Vol. 9, No. 13 (2001).
4. T. Onishi, "Low Loss Perfluorinated GI-POF," Proceedings of POF (2001), 337-40, 2001.
taking fibers from nature
When Maryanne Large sets her mind to something, look out. Less than a year after beginning work on microstructured polymer optical fibers (MPOFs), Large's team from the Optical Fibre Technology Centre at the University of Australia (Sydney, Australia) not only produced the first single-mode fiber MPOF but established the methodologies and facilities for preform and fiber production.
"The whole area is very new and realistically is still in the R&D phase," says Large. The work is most likely to have a key impact for large-core plastic fibers with good dispersion properties for short-distance, high-data-rate communication. "We have patents relating to this, but there is still quite a bit of work to be done understanding and optimizing the performance," she adds.
Large's interest in this field grew from her fascination with biology. "I had previously done some work on 'animal optics,' looking at structural color in nature," she says. This is color that is caused by microstructures such as diffraction gratings, and thin films rather than pigment. The striking color of a hummingbird's feathers, for example, is largely due to diffractive effects. Butterflies have a huge range of optical effects. Large's work on butterflies that have 3-D "photonic crystals" in their wing scales will soon be published.
"There are other reasons why this field of study might be considered interesting from a physics perspective, namely the question of optimization," says Large. Structures have evolved to be efficient even with a much more limited range of materials, so they may offer some interesting hints about which approaches are best. "I have a parallel interest in optimizing structures by using evolutionary strategies," she says. "In some cases there are natural structures available that cannot or have not been fabricated yet, and this is a chance to study their optical properties." -Laurie Ann Toupin
Maryanne Large is with the Optical Fibre Technology Centre, University of Sydney.