Open Access
Add to BibSonomy
Abstract
Results of the fabrication and characterization of optical fiber couplers made of multimode step-index fluoroindate (InF3) fibers are presented. The fabrication setup was customized for this type of glass with a constant source of controlled nitrogen flow heated to a target temperature with an accuracy ±1°C. Combined with a novel fast fusion approach and with excellent control of the viscosity throughout the process, the clean gas flow and well-controlled temperature enable the fabrication of fused fiber couplers absent of any noticeable crystallization. A coupling ratio of 45/55 was achieved, with an excess loss of 0.35 dB, at 1.7 µm. To the best of our knowledge, this represents the first low excess loss (<1 dB), multimode, InF3 fiber couplers.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Sensing in the mid-infrared (MIR) region of the electromagnetic spectrum (2-5 $\mathrm{\mu}$m) is useful in a large variety of applications, including spectroscopy, as typical bonding energies of molecules fall within that range; medical imaging such as optical coherence tomography (OCT); and therapy such as laser-based tissue ablation [1–3]. For instance, targeting the peak absorption of water at around 2.9 $\mathrm{\mu}$m enables minimally-invasive laser surgery in a less damaging way than conventional methods. This is achieved through the use of ultrafast pulses that rapidly superheat the water within the tissue, leading to the expulsion of tissue at a rate faster than the energy can diffuse to the surrounding area [4]. For such applications, there is a clear need for fiber-based components allowing the transport of MIR signals over moderate to large distances with the lowest amount of optical loss possible. Unfortunately, silica-based fibers are unsuitable for MIR light as they are not transparent in this region. In contrast, fluoride glasses such as ZBLAN (ZrF$_4$, BaF$_2$, LaF$_3$, AlF$_3$, NaF) and indium fluoride (InF$_3$), feature a wide transparency window, ranging from 0.3 to 4.3 $\mathrm{\mu}$m for ZBLAN and 0.3 to 5.5 $\mathrm{\mu}$m for InF$_3$ [5]. However, even though the fabrication of fluoride fibers has made great progress in the past and reached commercial availability with low-loss transmission [6], one problem remains: few compact and robust devices based on these fibers exist on the market. Indeed, fluoride fiber components, namely optical fiber couplers—the working horse of fiber-based devices—are difficult to fabricate. In recent years, notable achievements have included the development of single-mode, InF$_3$ side-polished couplers operating at 4.53 $\mathrm{\mu}$m, demonstrating complete power transfer [7]. Additionally, an InF$_3$ multimode combiner recently featured an excess loss of 0.96 dB at 1.55 $\mathrm{\mu}$m [8]. Researchers have also successfully created a range of ZBLAN single-mode fused couplers with different coupling ratios. Initial experiments conducted at 2 $\mathrm{\mu}$m showed a coupling ratio of 50/50 (3 dB excess loss), ratios of 1/99 to 10/90 (excess losses as low as 0.76 dB) [9]; a ratio of 5/95 (1.6 dB excess loss) and a ratio of 14/86 (1.8 dB excess loss) at 2.2 $\mathrm{\mu}$m [10]; and in a more recent development, a ratio of 39/61 (4.3 dB excess loss) at 2.7 $\mathrm{\mu}$m [11]. Other types of glasses, in particular chalcogenides, were also proven useful in the fabrication of fused fiber couplers [12,13]. While these results are promising, the levels of excess loss are not yet sufficient, and further innovation is needed in order to achieve mature, nearly lossless fluoride fiber optical fiber couplers with any desired coupling ratio.
In this paper, we present a novel fabrication method for low-loss (<1 dB) multimode fused optical fiber couplers made of InF$_3$ glass. This method involves a quick fusion process, followed by an adiabatic tapering procedure, where the flame used for silica components is replaced by a heated nitrogen flow with high-accuracy temperature control. The resulting couplers are entirely absent of visible crystallization. Through this method, we fabricated what is, to the best of our knowledge, the first instance of low-loss, multimode, InF$_3$ fiber couplers.
2. Principles of coupler fabrication
2.1 State of the art
Although fabrication methods and design parameters [14] for silica-based fiber couplers have been considerably developed, such parameters for fluoride fiber coupler fabrication have not yet reached maturity because of the many challenges these types of glasses present. These challenges are many and include a propensity for the fluoride fibers to crystallize, a narrow window of temperatures with the right range of viscosity for component fabrication [15,16], hygroscopicity, low glass transition temperature, and higher fragility of the glasses. One avenue of fabrication bypasses some of these difficulties by polishing the fibers, thereby avoiding the need for heat. The technique consists in inserting a fiber into a groove made on a holding block, usually quartz, sealing the fiber to the block, and grinding and polishing the surface of the block to expose the core of the fiber. A coupler is obtained by placing two of these blocks on top of each other. Depending on the grinding depth, various coupling ratios can be achieved [7,17,18].
Another technique, illustrated in Fig. 1, is often used for silica-based devices. It consists in contacting several fibers parallel to one another, fusing them, and finally pulling on them while scanning a source of heat along a defined length, in order to reduce the cross-sectional area of the fused section. The final component is characterized by its degree of fusion and its inverse taper ratio (ITR), $\sqrt {S_f/S_i}$ [16], where $S_i$ is the initial cross-section, and $S_f$ the final cross-section. Small ITR and/or high degree of fusion bring the cores of the fibers closer to one another, thus allowing light to transfer. A key aspect of a successful design is having an adiabatic transition in the tapered regions.
Fig. 1. Different steps of the fusion and tapering technique for optical fiber couplers. (a) The two fibers are next to each other, touching, perfectly parallel. (b) A chosen region is fused by sweeping the heat source. This can be achieved either by moving the heat source while the fibers stay immobile or by having a still heat source while the fibers are moving. (c) The two ends of the fibers are pulled apart while the heat scan continues, creating a tapered shape.
2.2 Necessary improvements
The fusion and tapering technique can in principle also be used with fluoride fibers. However, many hurdles are present. The main challenge in making optical fiber couplers with fluoride fibers is crystallization. Indeed, fluoride fibers, such as ZBLAN and InF$_3$, are prone to crystallization when heated and mechanical tension is applied, like in a fusion and tapering approach. Figure 2 presents a clear example of such a phenomenon.
Moreover, the glass transition temperature T$_g$ plays a crucial role in fiber component fabrication with this technique. Fluoride fibers require a lower temperature to reach glass transition, or temperature process range, than their silica counterpart. Indeed, $T_g$ for ZBLAN and InF$_3$ are 265 $^{\circ }$C and 300 $^{\circ }$C, respectively [19,20]. Compared to the glass transition temperature of silica at 1200 $^{\circ }$C [21], it is clear that a different approach to heating the fibers must be considered.
In contrast to the oxygen-based flame torch often used for silica fibers [22], a constant flow of heated gas can serve as the heat source [16]. The gas must be free of water due to the hygroscopic nature of the glass, and of impurities that act as nucleation sites for crystallization. Ambient air is thus not a great choice [23], and a pure non-reactive gas can be used instead. Nitrogen was chosen due to its commercial purity and relatively low cost. A demonstration of the difference between air and nitrogen can be found in Fig. 3, where the same fiber taper recipe is used with both heated air and hot nitrogen. The resulting tapers clearly showcase the advantage of clean gas over ambient air.
Fig. 3. Visual comparison for crystallization in InF$_3$ tapers at 0.3 ITR of the same recipe with hot air (a) and hot nitrogen (b).
Another difficulty of the fabrication process is the limited temperature transition range, characterized by a rapid alteration of the viscosity of the glass as a function of temperature [15,16,24], which can lead to crystallization, non-adiabaticity, a more fragile structure, or a broken component. Therefore, fluoride fiber components demand accurate control of the temperature.
3. New approach
3.1 Setup
In order to prepare the fused coupler, two pieces of multimode step-index 100/192 $\mathrm{\mu}$m 0.26 NA InF$_3$ fluoride fiber (IRFH10026, Thorlabs) are prepared by stripping a 16 mm section with a removal gel. Once the protective acrylate has been removed, the fibers are cleaned with isopropanol, followed by acetone, to ensure the bare fiber is exempt from dust. Two stripped fibers are then placed side-by-side in a custom-built fusion-tapering system, presented in Fig. 4.
The fusion tapering system comprises two holding blocks (left and right) made of brass and engraved with a V-shaped groove. The grooves are customized to the sizes of the fibers. Combined with two magnetically secured clamps, the grooves ensure a firm hold of the fibers. The blocks are resting onto motorized translation stages and are run by a motion controller (Newport, Model MM3000). The setup offers great flexibility thanks to its removable and replaceable blocks. To ensure proper fiber alignment and contact, small weights are added to each block, in close proximity to the fusion zone, as shown in Fig. 4(c). This method avoids the need to twist the fibers, minimizing unwanted tensions and potential breaks of the fragile fibers. A temperature controller (J-KEM, model 210–TC), in conjunction with a J-type thermocouple, precisely maintains the desired temperature of the constant stream of nitrogen passing through the heating tube, within $\pm$0.5 $^\circ$C. This accuracy level for temperature control is paramount to the whole process.
3.2 Fabrication method
Crystallization can greatly affect the transmission performances of fiber devices. Losses induced by this phenomenon are highly dependent on the localization of crystallization. Losses are much higher if crystallization reaches the core, at any ITR. In contrast, if crystallization stops near the surface, the losses appear only below a specific inverse taper ratio (ITR) threshold. As crystallization is usually seeded at the surface, it is best to keep any heating process as short as possible to mitigate its migration toward the core.
As a consequence, both fibers are fused under high heat for a short time. The temperature at the fibers is kept within the relatively narrow range of 312 $^\circ$C to $316\pm 1^\circ$C (initially measured by a thermocouple replacing the fibers), depending on the desired degree of fusion. The temperature for the tapering step is lowered to about 260 $^\circ$C at the position of the fibers. Although the precision of the temperature readings is high, their accuracy is much lower, and it is possible that other setups will require different values. The temperature window is always higher than T$_g$ to allow both high viscosity and quick fusion, but it must be low enough to avoid the melting of the fibers. No tension is applied at that time and the blocks remain stationary. The fast process only requires a few sweeps of the nozzle over the fused region of 2 to 3 mm. During the fast fusion, due to gravity and lower viscosity of the fibers, a small drop in height can be observed. The blocks are slightly pulled apart at the end of the fusion step while the heat source is slowly returning to its initial position. This straightens the fibers and minimizes losses. A combination of swift heating and low tensile stress helps to avoid crystallization.
To taper down the two fused fibers, the temperature controller is adjusted and stabilized to a lower temperature, close to the glass transition. The pulling speed is kept constant, and the nozzle sweeps the fibers with increasing width as time advances. The whole sweeping process takes only about 1 minute. The right settings for a given ITR and pre-determined pulling speed can be calculated from the theory outlined in Ref. [14].
4. Results
For characterization, both output branches of the coupler are measured with a power meter (PM100USB, Thorlabs) connected to an integrating sphere photodiode (S145C, Thorlabs). Illumination is performed by a broadband (450 – 5500 nm) stabilized tungsten light source (SLS202L, Thorlabs). The input optical fiber was illuminated through a bare fiber adapter connected to a matching adapter at the broadband source output, ensuring that a maximum number of modes were populated. Figures 5 and 6 show insertion losses (losses due to coupling plus excess losses) in the main branch (the fiber in which light is initially injected) of two different fiber couplers during fusion and tapering. Figure 5 shows the power at the output of the main branch, as a function of time, for a coupler with an initial low degree of fusion, while Fig. 6 showcases a component with an initial high degree of fusion.
Fig. 5. Power inside the main fiber, as a function of time, for a low degree of fusion. Section (a) represents the fusion step. Section (ii) is the tapering operation. The sudden apparent rise in power is caused by a short period of stress relief at the beginning of the process.
These Figures show that insertion loss occurs at each step of the process. It is clear that for a higher degree of fusion, as seen in Fig. 6(a), there is more loss occurring in the main branch. This is largely due to the field being transferred to the second fiber, and not to additional excess losses. At the end of the fusion step (see Fig. 5(a) and Fig. 6(a)) insertion losses in the main branches are 1.08 dB and 3.04 dB, respectively. The main cause of insertion loss is again coupling between fibers, but excess losses can already appear and are likely due to small shape deformations of the fibers during the fusion process. Pictures of both couplers at the end of the fusion step are presented in Fig. 7.
Fig. 7. Pictures of two fiber couplers after the fusion step. Coupler (a) has a small degree of fusion, while coupler (b) has a high degree of fusion.
Figure 5(b) and Fig. 6(b) present the behavior of both couplers during the tapering step. In Fig. 5(b), additional insertion losses are observed. They are due to an increase in coupling between the fibers as the cores get closer. In contrast, the tapering process does not increase the coupling in the case of a high degree of initial fusion, as shown in Fig. 6(b). Tapering is still performed to increase the solidity of the fusion between the fibers. The sudden rise in optical power at the beginning of the fusion step is caused by a brief period of stress relief of the fibers. The original stress is caused by the small weights used to ensure uniform contact between the fibers, and the power reference is taken with that stress still in place. The slow rise in power during the tapering step is caused by further straightening of the fibers, leading to slightly lower coupling. The final couplers are shown in Fig. 8.
Figure 8 also shows an absence of visible crystallization after the tapering step. Hence, the components feature low excess losses (0.33 dB and 0.35 dB respectively). The final coupling ratios are 40/60 for the first coupler and 45/55 for the second. The estimated down and up transition lengths for the coupler represented in Fig. 8(a) are both 1.9 mm. The coupler in Fig. 8(b) is more asymmetric after the fusion step, and the down and up transition lengths are estimated at 1.7 mm and 1.4 mm, respectively. All taper recipes follow the principles of Ref. [14] that lead to adiabatic couplers. Repeatability tests for maximum coupling ratios indicate that, while the swift nature of the fusion step makes having symmetrical fusions and identical degrees of fusion difficult, the coupling ratios are relatively constant throughout tests and losses remain below 1.0 dB, as presented in Table 1.
Table 1. Performances of InF$_3$ optical fiber couplers.
5. Discussion
The fast fusion method with heated nitrogen has proven to be effective in the fabrication of low-loss multimode optical fiber couplers made of fluoride glasses transparent in the MIR, without apparent crystallization. The first couplers aimed at achieving a 50/50 coupling ratio. Due to the very high number of modes guided in the structure, multimode couplers are fundamentally incoherent devices. As a result, a symmetric multimode coupler can never couple more than 50% of the average power into the non-illuminated branch. In fact, currently available commercial multimode, silica-based 50/50 couplers generally only guarantee a coupling ratio between 45 and 50%. Our results are on par with that metric. The goal will always be to get closer to the theoretical limit, and we believe additional changes in the setup could lead to higher, better-controlled coupling ratios. One avenue to be explored is to increase the length of fiber that is heated during the fusion step, all the while keeping the fusion time short. A high degree of fusion over a longer section could potentially yield a high coupling ratio without requiring additional tapering. Additionally, improvements in the characterization setup and better control of the temperature and displacement of the nozzle during the fusion step could yield more reproducible results. A lower degree of fusion is expected to lead to a more diverse offering of coupling ratios (90/10, 80/20 $\ldots$).
Funding
Natural Sciences and Engineering Research Council of Canada (RGPIN-2016-06404, RGPIN-2018-06151).
Acknowledgments
The authors would like to thank Thorlabs and Le Verre Fluoré for supplying, respectively, fluoroindate and ZBLAN optical fibers.
Disclosures
Caroline Boudoux and Nicolas Godbout are founding partners of Castor Optics, Inc.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. J. Haas and B. Mizaikoff, “Advances in mid-infrared spectroscopy for chemical analysis,” Annu. Rev. Anal. Chem. 9(1), 45–68 (2016). PMID: 27070183. [CrossRef]
2. R. Su, M. Kirillin, E. W. Chang, E. Sergeeva, S. H. Yun, and L. Mattsson, “Perspectives of mid-infrared optical coherence tomography for inspection and micrometrology of industrial ceramics,” Opt. Express 22(13), 15804–15819 (2014). [CrossRef]
3. I. Ilev and R. Waynant, “Mid-infrared biomedical applications,” Mid-infrared Semicond. Optoelectron. 118, 615–634 (2006). [CrossRef]
4. S. Amini-Nik, D. Kraemer, M. L. Cowan, K. Gunaratne, P. Nadesan, B. A. Alman, and R. J. D. Miller, “Ultrafast mid-ir laser scalpel: Protein signals of the fundamental limits to minimally invasive surgery,” PLoS One 5(9), e13053 (2010). [CrossRef]
5. M. Saad, “Indium fluoride glass fibers,” in Laser Refrigeration of Solids V, vol. 8275R. I. Epstein and M. Sheik-Bahae, eds., International Society for Optics and Photonics (SPIE, 2012), p. 82750D.
6. S. Cozic, S. Poulain, and M. Poulain, “Low loss fluoride optical fibers: Fabrication and applications,” in Advanced Photonics 2018 (BGPP, IPR, NP, NOMA, Sensors, Networks, SPPCom, SOF), (Optica Publishing Group, 2018), p. SoM2H.3.
7. Y. Kim and K. J. Lee, “Fabrication and characterization of side-polished fiber coupler for mid-infrared applications,” in Advanced Photonics 2018 (BGPP, IPR, NP, NOMA, Sensors, Networks, SPPCom, SOF), (Optica Publishing Group, 2018), p. SeW3E.1.
8. A. Annunziato, F. Anelli, P. L. P. D. Teilleul, S. Cozic, S. Poulain, and F. Prudenzano, “Fused optical fiber combiner based on indium fluoride glass: perspectives for mid-ir applications,” Opt. Express 30(24), 44160–44174 (2022). [CrossRef]
9. G. Stevens and T. Woodbridge, “Mid-IR fused fiber couplers,” in Components and Packaging for Laser Systems II, vol. 9730A. L. Glebov and P. O. Leisher, eds., International Society for Optics and Photonics (SPIE, 2016), p. 973007.
10. M. Rezaei and M. Rochette, “Single-mode zblan optical fiber couplers,” in Optica Advanced Photonics Congress 2022, (Optica Publishing Group, 2022), p. ITh2B.3.
11. M. Rezaei and M. Rochette, “Single-mode zblan optical fiber couplers,” in Optica Advanced Photonics Congress 2022, (Optica Publishing Group, 2022), p. ITh2B.3.
12. Y. Xia, L. Yang, F. Yang, L. Xu, Y. Fu, S. Dai, S. Kang, X. Wang, P. Zhang, Z. Liu, and W. Ran, “Fabrication of a tellurite-fiber-based side-pump coupler based on the tapered-fused method,” Opt. Express 31(8), 13169–13181 (2023). [CrossRef]
13. F. Tavakoli, A. Rekik, and M. Rochette, “Broadband and wavelength-dependent chalcogenide optical fiber couplers,” IEEE Photonics Technol. Lett. 29(9), 735–738 (2017). [CrossRef]
14. T. Birks and Y. Li, “The shape of fiber tapers,” J. Lightwave Technol. 10(4), 432–438 (1992). [CrossRef]
15. G. Zhang, J. Jiang, M. Poulain, A. Delben, and J. Delben, “Viscosity of fluoride glasses near the fiber drawing temperature region,” J. Non-Cryst. Solids 256-257, 135–142 (1999). [CrossRef]
16. É. Ducharme, S. Virally, R. I. Becerra-Deana, C. Boudoux, and N. Godbout, “Viscosity of fluoride glass fibers for fused component fabrication,” Appl. Opt. 61(17), 5031–5039 (2022). [CrossRef]
17. K. Goya, Y. Koyama, Y. Nishijima, S. Tokita, R. Yasuhara, and H. Uehara, “A fluoride fiber optics in-line sensor for mid-ir spectroscopy based on a side-polished structure,” Sens. Actuators, B 351, 130904 (2022). [CrossRef]
18. C. A. Schäfer, H. Uehara, D. Konishi, S. Hattori, H. Matsukuma, M. Murakami, S. Shimizu, and S. Tokita, “Fluoride-fiber-based side-pump coupler for high-power fiber lasers at 2.8 µm,” Opt. Lett. 43(10), 2340–2343 (2018). [CrossRef]
19. M. Bernier, V. Fortin, O. Henderson-Sapir, S. Jackson, F. Jobin, J. Li, H. Luo, F. Maes, M. R. Majewski, D. J. Ottaway, R. Vallée, and Y. O. Aydin, “Chapter 9 - high-power continuous wave mid-infrared fluoride glass fiber lasers,” in Mid-Infrared Fiber Photonics, S. Jackson, M. Bernier, and R. Vallée, eds. (Woodhead Publishing, 2022), Woodhead Publishing Series in Electronic and Optical Materials, pp. 505–595.
20. M. Saad, “Fluoride glass fiber: State of the art,” Proc. SPIE 7316, 73160N (2009). [CrossRef]
21. J. A. Bucaro and H. D. Dardy, “High-temperature brillouin scattering in fused quartz,” J. Appl. Phys. 45(12), 5324–5329 (1974). [CrossRef]
22. F. Bilodeau, K. Hill, S. Faucher, and D. Johnson, “Low-loss highly overcoupled fused couplers: fabrication and sensitivity to external pressure,” J. Lightwave Technol. 6(10), 1476–1482 (1988). [CrossRef]
23. E. Ducharme, “Etude des parametres de conception d’un coupleur fusionne en verre fluore,” Master’s thesis, Dep. de gée physique, Ecole Polytechnique de Montreal, Montreal, QC (2022).
24. K. Matusita, M. Koide, and T. Komatsu, “Viscous flow mechanism of fluoride glasses over a wide temperature range,” J. Non-Cryst. Solids 140, 141–145 (1992). Proceedings of the Seventh International Symposium on HALIDE GLASSES. [CrossRef]
