For the first time to our knowledge, polymer-based microstructured fibers with complex cross-sections are directly produced via melt extrusion. Two principal types of fibers were fabricated: a microstructured fiber of a single polymer with a hexagonal array of air holes and a bicomponent fiber consisting of approximately 60 coaxial rings. From the latter, strong visible iridescence was observed and is shown to exhibit a mechanochromic response. This approach, the mainstay of the textile trade for decades, offers a means of continuous high-volume low-cost manufacturing of polymer (and conceivably soft-glass) fibers. For example, in the present effort, 128 coaxially microstructured fibers were fabricated simultaneously at rates exceeding 1200 m/min from industrially mainstream polymers. This approach offers an important step forward towards commoditizing microstructured fibers and open new doors for optical engineering in fashion, marking/identification, and numerous military applications.
© 2007 Optical Society of America
There has been considerable effort undertaken over the past few years towards the production of microstructured and photonic crystal fibers. These efforts principally have focused on glass rods and capillaries that are drawn into cane, stacked, possibly fused and subsequently redrawn down into a fiber that then possesses a desired cross-sectional geometry and symmetry.[1–4] The choice of silica glass as a starting material has largely been dictated by its prevalence among the optical fiber community, which arises from silica’s high strength and thermal stability coupled with its optical performance and low loss at wavelengths that have enabled the growth in global telecommunications. As opportunities outside of telecommunications continue to expand, other materials have received greater attention for application-specific fiber designs. To-date, this includes soft oxide glasses,[5–7] polymers,[8, 9] and infrared glasses. Although this research still is at a nascent stage, the reduced fabrication temperatures associated with polymers and soft-glasses, the high indices of chalcogenides and other selected soft glasses, and the large nonlinearities possible with chalcogenides and chromophore-doped polymers, with respect to silica, mark these specialty materials as strong candidates for future applications.
In this paper we describe a continuous approach to produce microstructured fibers from conventional and fluorinated polymers and soft-glass systems using high-volume low-costs methods adopted from the textile industry. As is shown, fibers exhibiting established cross-sectional geometries are fabricated at high speeds as are more complicated concentric ring fibers that would be nearly impossible to make using stack-and-redraw techniques. Such fibers, especially in the case of the polymeric analogs that diffract UV and visible light, are of interest for remote sensing and anti-counterfeiting needs. Further, these fibers are uniquely suited for direct integration as flexible textile composite with engineered optical functionalities.
Fibers that possess the types of cross-sections that the optics community now refer to as microstructured or photonic crystal have long been named as “islands in the sea” fibers by those in the textile trade. Such fibers are generally achieved through the co-extrusion of two polymers using a double screw melt spinning extruder. For the fibers discussed here, polyethylene terephthalate (Eastman F61HC, “PET”), polypropylene (22 MI, “PP”), polymethyl methacrylate (Syro L 40-002-000, “PMMA”), and nylon (Basf BS 400 PA6; “Nylon”) were extruded with ethylene-vinyl alcohol (Kuraray Exceval CP; “EvOH”), which was used for the formation of islands in selected fibers. Figure 1 provides a schematic of the bicomponent extrusion process used in this work. In general, pieces of the desired polymers are softened or melted and subsequently pressurized through a die (or series of dies called a pack). There are several requirements for the production of microstructured bicomponent fibers. The fiber structure and its orientation is achieved through the construction of an extrusion die pack and is maintained through the spinneret and winding stages. The primary qualification is that the polymer is amenable to a melt spinning process and, as such, the melt must be sufficiently stable over a broad temperature range. In semi-crystalline polymers, such as poly(ethylene terephthalate) (PET) and polypropylene (PP), the processing range extends from the melting transition to the degradation temperature. For glassy non-crystallizable polymers, such as PMMA, the processing range extends from above the glass transition temperature to the thermal stability limit. Thus many polymers and copolymer combinations are viable for melt spinning into microstructured bicomponent fibers and with careful temperature selection the fiber microstructure can be controlled and maintained. In the present case, for the PP/PET fibers (discussed below in regards to Figs. 3 and 5), each polymer traversed four heated zones that raised the temperature from 220 to 250 °C for the PP and from 280 to 305 °C for the PET. The extruder pressure was 750 psi for both the PP and the PET.
The formation of such microstructured fibers via bi-component (possibly multi-component) melt-spinning is a careful balance between the miscibility of the polymer melts, polymer melt viscosity and the relative rate of metering the polymers to the bicomponent spinneret. As a general rule it is desirous to have poor miscibility between components in order to form well-defined islands or coaxial layers.
Optical microscopy was performed using a Vista-Vision microscope with the images being taken using top illumination with the fiber samples on a black background. Color images were taken using a Tripix camera attached to the microscope. Reflectance spectra were taken using and Ocean Optics USB2000 spectrophotometer attached to the Vista-Vision microscope.
Fiber structures were examined using a HITACHI S3500-N scanning electron microscope (SEM) at 20 kV. In order to allow high vacuum mode in SEM, sample surfaces were coated with platinum for 2 minutes at a current of 15 milli-amperes using a Hummer 6.2 Sputtering System (Anatech Ltd). In order to determine the exact layer thicknesses, a Hitachi T–7600 transmission electron microscope (TEM) was used. Samples were observed at 100 kV voltage.
3. Results and Discussion
The combination of PMMA and EvOH [not shown] did not give particularly good result for any of the conditions examined. The EvOH tended to blend into the PMMA matrix preventing the formation of the structure, and hence the use of this polymer was dismissed. Nylon and EvOH did give better results and indicated the feasibility of the bicomponent extrusion approach to directly fabricating microstructured fibers. However, the resultant structure still possessed some unwanted results including islands that were out-of-round and not equally spaced as generally would be required for useful application. Some mixing between the EvOH and nylon was present presumably due to the relatively hydrophilic nature of nylon. In view of the difficulty in producing a well-defined structure the next direction was to use polymers with higher levels of chemical incompatibility, such as PET and EvOH.
The relatively hydrophobic character of PET against to the hydrophilic nature of the EvOH polymer resulted in a better separation of the island (hole precursor) leaving an improved and better-defined structure. The next step was to increase the number of islands from 16 to 127 therein replicating a hexagonal structure, which already has been proven to be effective as a photonic crystal design. Further engineering of pack design allowed the fabrication of fibers with a greater degree of symmetry and island (hole) isolation; see Fig. 2. The islands are regularly arranged and equally spaced in the polymer matrix. The EvOH polymer was dissolved out in this case to leave the microstructure shown.
Although PET and EvOH are not particularly good optical polymers, the point is that high quality fibers with appropriately structured cross-sections can be directly fabricated using high-volume, low-cost approaches borrowed from the textile industry. Since the purpose of this work was to develop iridescent fibers for consideration in sensing and anti-counterfeiting applications, the synthesized fibers did not possess a “core” in the conventional description of an optical fiber. This could easily be done through redesign of the extrusion die pack to fill the central region with PET.
With fiber exhibiting a periodic cross-section of islands (and holes once dissolved away), efforts turned to the more complicated task of coaxial rings. There has been previous work in such fibers though, as mentioned above, these efforts have focused on combinations of high index contrast materials (e.g., chalcogenide glasses and polyether sulfone polymers) with relatively few layers rather than lower index contrast materials and a greater number of layers.[11,12] Proprietary extrusion dies were redesigned to permit the bicomponent extrusion of concentric ring structures as shown in Fig. 3. This fiber exhibits 60 concentric rings of alternating PP and PET. This particular fiber, when viewed laterally, exhibited a blue-green iridescence.
As can be seen, the fiber’s cross-sectional dimensions are commensurate with the conventional sizing of textile fibers thus making these iridescent fibers immediately able to be integrated into apparel and related commercial, fashion, and military products. Interestingly, although the outer diameter in this present case is about one order of magnitude smaller than a standard telecommunication optical fiber, the center “core” region is of correct order of magnitude for single mode operation (though not the focus of this particular work). Increasing the core size slightly and scaling the outer diameter to more appropriate dimensions is straight-forward given proper extrusion designs.
In order to evaluate mechanochromic effects in these iridescent coaxially microstructured polymer fibers, they were mechanically strained to various degrees along their length and changes in their reflectivity were observed visually and quantitatively in the radial direction. Figure 3 provides visual images of the fiber under magnification showing the iridescent coloration intrinsic to this otherwise optically passive fiber.
Figure 4 provides a more quantitative examination of the fiber’s spectral reflectivity. As can be seen, the spectral measurements corroborate the Fig. 3 optical micrographs. The unstrained fiber has a reflectivity that peaks in the green whereas the fiber strained by 120% has a reflectivity that peaks in the red. This is, of course, counter-intuitive since a lesser amount of strain (40%; not shown in Fig. 4 but seen in Fig. 3) yields the appropriate blue-shift in reflection as one would expect with positive strain along the fiber length yielding the observed decrease in lateral dimension courtesy of Poisson’s ratio. This diminution in lateral (or radial in the present case) dimensions does lead to a blue-shift in the reflectivity spectrum. In reality, there are multiple reflectivity peaks that arise from the several periodicity scales and diffraction orders. The strain leads to a predicted blue-shift in the cross-sectional reflectivity which causes the peak in the green to shift to the blue and out of the measurement spectral window. Similarly, a near-infrared reflectivity peak similarly blue-shifts towards the red and into the measurement spectral window. Simple computations based on a one-dimensional series of dielectric stacks bear out this postulate.
4. Summary and Conclusions
In summary, for the first time to our knowledge, fibers possessing complex periodic cross-sections have been made directly using polymer extrusion. Such processes, foundationally based on textile manufacturing, open the door to high-volume low-cost production of microstructured and photonic crystal fiber.
In order to further support this point, Fig. 5 shows a highly concentric microstructured polymer fiber made following a modification to the extrusion die stack used to make the Fig. 3 fibers. As with those fibers, using this present approach, 128 fibers are made simultaneously at a rate exceeding 1200 m/min (> 20 m/s). Industrially, similar processes operate in excess of 5000 m/min and, given the direct generation of 128 filaments at the same time as employed here, one could make enough fiber to circle to world in 1 day.
Although commodity polymeric materials were the focus of this work, the process is amenable to more exotic materials such as fluoropolymers and soft-glasses, which can be intrinsically passive or chemically modified to provide “enhanced” optical effects such as light emission or nonlinearities through appropriate active dopants or chromophores.
More specifically, and in conclusion, microstructured and coaxial Bragg fibers were fabricated via direct bicomponent extrusion. The fibers exhibited optical reflectivities in the visible portion of the electromagnetic spectrum orthogonal to the fiber axis due to the structure-induced periodicity. In other words, the coloration arises solely from the structural periodicity and not, as is conventional in apparel, a dye or colorant added to the polymer. This also means that such a fiber will never bleach or lose its coloration with wear. The elastic nature of the polymers was used to show strain-induced mechanochromism. Such potentially low cost polymer fibers are of great interest for use in fashion, anti-counterfeiting, and numerous military applications.
The authors wish to acknowledge the National Textile Center for financial support and Dr. Frank Barnes (University of Colorado) for insightful discussions and setting the authors initially on the path to the concentric ring fibers.
References and Links
4. J. Broeng, “Photonic Crystal Fibers: A new class of Optical Waveguides,” Opt. Fiber Technol. 5, 305 – 330 (1999). [CrossRef]
5. V. V. Ravi Kumar, A. George, W. Reeves, J. Knight, P. Russell, F. Omenetto, and A. Taylor, “Extruded soft glass photonic crystal fiber for ultrabroad supercontinuum generation,” Opt. Express 10, 1520 – 1525 (2002). [PubMed]
6. K. Kiang, K. Frampton, T. Monro, R. Moore, J. Tucknott, D. Hewak, D. Richardson, and H. Rutt, “Extruded singlemode non-silica glass holey optical fibres,” Electron. Lett. 38, 546 – 547 (2002). [CrossRef]
7. X. Feng, T. Monro, V. Finazzi, R. Moore, K. Frampton, P. Petropoulos, and D. Richardson, “Extruded singlemode, high-nonlinearity, tellurite glass holey fibre,” Electron. Lett. 41, 835 – 837, (2005). [CrossRef]
8. M. van Eijkelenborg, A. Argyros, J. Zagari, S. Manos, N. Issa, I. Bassett, S. Fleming, R. McPhedran, C. M. de Sterke, and N. A. P. Nicorovici, “Microstructured polymer optical fibre,” Opt. Express 9, 319 – 327 (2001). [CrossRef] [PubMed]
9. A. Argyros, I. Bassett, M. van Eijkelenborg, M. Large, J. Zagari, N. Nicorovici, R. McPhedran, and C. Martijn de Sterke, “Ring structures in microstructured polymer optical fibres,” Opt. Express 9, 813 – 820 (2001). [CrossRef] [PubMed]
10. T. Monro, Y. West, D. Hewak, N. Broderick, and D. Richardson, “Chalcogenide holey fibres,” Electron. Lett. 36, 1998 – 2000 (2000). [CrossRef]
12. S. Hart, G. Maskaly, B. Temelkuran, P. Prideaux, J. D. Joannopoulos, and Y. Fink, “External reflection from Omnidirectional Dielectric Mirror Fibers,” Science 296, 510 – 513 (2002). [CrossRef] [PubMed]