Open Access
Add to BibSonomy
Abstract
We experimentally compared for the first time, two techniques of optical fiber preform shaping based on the mechanical grinding and thermal CO2 laser processing from the point of the inner-cladding losses. The shaped preforms were fabricated of coreless pure silica technical rods as well as high purity silica Heraeus F300 rods and drawn them into coreless multimode fibers with various inner-cladding geometries coated with a low index fluorinated polymers. The background losses of the fibers were measured via the cut-back method and compared to the losses of the unshaped fibers with a circular cross-section. Results show that both preform-shaping techniques would induce additional losses in the inner-cladding. High surface scattering losses were observed in the mechanically-grinded fibers. On the other hand, the mechanical grinding retains the advantage of a significant reduction of attenuation peaks attributed to OH-groups that penetrated into the preform surface during the preform collapse. On the contrary, CO2 laser thermal-shaping provides the advantage of quick, fully automated shaping with smooth surface finish and induces much lower scattering losses, but it is not so effective in removing water penetrated surface layer of the preform so that OH-groups diffuse deeper towards the preform center. Additionally, laser thermal-shaping allows processing the preform to complex shapes which are more effective in scrambling cladding modes. Some of the absorption peaks of OH-groups and fluorinated polymers may be rather close to common pumping wavelengths and this should be considered in the design of the double-clad fibers and selection of proper shaping technology.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
The invention of double-clad (DC) fiber structure has been a major breakthrough in lightwave technology and laser physics as it paved the way for power scaling of fiber laser devices [1–3]. It enables good coupling efficiency of the high-power laser diode pumps into the active fiber thanks to inner cladding with a larger numerical aperture and larger cross-section as compared to the fiber core. Despite significant advances in fiber laser technology, there is a pressing demand for increasing the saturated output power with nearly diffraction-limited beam and high electrical-to-optical efficiency. A number of limits, that are often contradictory, has to be accounted for. The limits are imposed namely by nonlinear optical effects [4], transverse mode instability [5–7], excess heat management [8], spectroscopic properties of the active media [9], photodarkening [10,11], etc. Many design configurations of active fibers have been investigated to overcome these limits. Nonlinear optical effects and excessive background losses can be mitigated by shortening the fiber through the enhancement of pump absorption efficiency by mode scrambling. DC fibers with a non-circular cross-section of inner-cladding became an important requirement in order to scramble pumping modes, which in turn increases their spatial overlap with a fiber core and improves the pump absorption efficiency. Various shapes of the inner-clad of DC fibers have been investigated [12–14]. Tailoring of the longitudinal absorption can be provided by special coiling methods including figure-eight [15], kidney [16], or spiral-shaped coiling [17], twisted and coiled fibers after or while drawing [18,19], volume Bragg gratings [20] and mode scramblers in pump combiners [21]. There still exists a pressing question of what is the best cross-section shape of inner-clad for efficient pump absorption and what method can be used for the fabrication of such shaped fibers. In this paper, we report for the first time an experimental comparison of the impact of optical fiber preforms shaping on the background inner-cladding losses of shaped DC Fibers. Two different techniques of optical fiber preform shaping were investigated, namely the conventional mechanical grinding, and the recently reported technique of thermal shaping using a CO2 laser. The shaped preforms were fabricated of coreless pure silica technical rods as well as high purity silica Heraeus F300 rods and drawn them into coreless multimode DC fibers with various inner-cladding geometries coated with a low index polymer. The background inner-cladding losses attributed to the attenuation peaks of OH-groups and glass-polymer scattering of the fibers were measured via the cut-back method and compared to the losses of the unshaped fibers with a circular cross-section. Constraints and drawbacks of both shaping techniques were also discussed.
2. Shaping techniques of optical fiber preforms
2.1 Mechanical-based grinding
It is a conventional technique that enables the shaping of optical preforms surface by grinding using diamond or steel plates with abrasive compositions. Where the process started by holding a piece of optical fiber preform with few centimeters in length on a pad of weighted-holder using beeswax. The preform must hold evenly on the holder in order to provide even shaping depth. The holder is then placed upside down on a rotating plate made by diamond or steel. The plate must be filled with water and abrasive compositions prior to starting shaping to avoid any stacking or damaging to the preform with the plate as shown in Fig. 1(a). The preform is then removed from the holder and rotated precisely with a particular angle depending on the desired shape to be placed once again on the holder on a different region. Each polished side is, proceeded repeatedly a few times using still finer abrasive composites. In our shaping process, we used white corundum (Al2O3) with different particle sizes of F320, F1200 and CerOx powder. Nevertheless, this technique still can make the surface of the preform rough and sometimes might lead to subsurface damaging. The process of shaping is lengthy and the preform has to be repeatedly carefully and manually adjusted on the plate. The repeatability of the shaping process is quite low. Moreover, this technique is limited to sharp and flat kind of shaping; it is quite impossible to shape the preform with flower or star shapes. Figure 1(b) shows a real shaped optical preform in Hexagon-shape based grinding technique.
Fig. 1. (a) Schematic diagram of the mechanically shaping process (b) real image of a mechanical-shaped preform.
2.2 Thermal-based CO2 laser
It is a sophisticated technology that used CO2 laser irradiation to heat and evaporates the silica out of the preform surface. Shaping optical preforms with CO2 laser profits from a high absorption of the silica glass at wavelengths around 10.6 µm. Processing silica glass by a CO2 laser is widespread and includes silica glass cutting, fiber drawing [22], rapid material evaporation [23]. CO2 laser based optical preform-shaping was earlier reported in 1981 by Brenci et al., [24] and recently highlighted once again by Shardlow et al., [25]. However, there were not enough details on shaping process and no further investigations were published until we recently highlighted it once again in a preliminary-results report [26]. Our CO2 laser-shaping system allows automated, quick and high-quality shaping processes to various preform shapes. Numerous parameters influence results such as laser intensity, speed, and the number of expositions run. The thermal-shaping system is shown in Fig. 2. One of the preforms ends is held by a motorized and precise rotator-holder fixed on a translation stage in order to make uniform shaping with the required number of grooves. Unlike mechanical shaping, this technique of laser processing can provide almost no contamination or any damage to the surface of the preform, which is a key demand in producing high-strength optical fibers. The laser beam evaporates only a thin layer of the preform surface because the absorption path-length in silica glass is about ∼10 µm [23]. However, the silica evaporation can be controlled by the laser intensity as well as the speed of laser exposition. At high laser intensity with a slow speed of laser motion, a thick silica layer can be evaporated out of the preform surface, leaving some dry silica soot at the preform surface.
Reprocessing the preform surface with a few stages of laser exposition and with gradually decreasing laser power and increasing laser speed evaporates the dry silica soots and allows achieving a preform with a smooth surface [26]. The total duration of the shaping process will depend on the desired length, depth, shape, the number of times exposing to the laser, as well as to the applied laser intensity. For instance, shaping preform with six grooves (hexagon shape) with a length of 6 cm and depth 0.5 mm might take 6 hours under laser intensity ranged from 30 W to 15 W. The time can be shortened by using a more powerful laser. Meanwhile, since the shaping of this technique is based on the evaporation process; such an approach enables preform-shaping to concave and convex cross-sections unlike the mechanical process were limited by flat shapes. Different preform shapes can be achieved by focusing the laser beam with spherical or cylindrical ZnSe lenses. We obtained grooves width of 1 mm and 3 mm for a cylindrical focusing lens with a fast axis perpendicular or parallel to the preform axis, as shown in Figs. 3(a) and 3(b).
Fig. 3. (a) Illustrated diagram of orienting the incident laser beam on the preform surface (b) are real images of small-section shaped preforms based on two angles of the laser beam incident.
The repeatability of symmetry shaping can be investigated by measuring the preform contour using a displacement sensor (see Fig. 2). Figure 4 shows contours of three hexagon-shaped pieces of the same preform (coreless Heraeus F300) varied in groove depths of (∼1.0 mm, ∼0.58 mm, and ∼0.36 mm) together with the contour of the unshaped piece. The displacement sensor was recorded while rotating the preform. Figure 4(a) represents the measured groove depths of three shaped pieces in the linear state, while Fig. 4(b) shows the cross-section of the preform pieces by processing the measured depth data into the polar state. As it can see the shaped grooves have excellent symmetry owing to the motorized, precise, and handless shaping process. Note that the slight curvature of the plotted lines in Fig. 4(a) is due to the slight misalign of the preform position while scanning by the sensor.
Fig. 4. Depth measurement of the three shaped pieces of the same coreless Heraeus F300 preform, where (a) and (b) represents the depth in the linear and polar states respectively.
Figure 5 shows the excellent repeatability of the laser shaping process. Where two pieces of the same preform (coreless Heraeus F300) were individually shaped in two different depths of ∼1.0 mm and ∼0.3 mm for this investigation. As it can see from Fig. 5 that both pieces (piece 1 and piece 2) having perfect matches in both shaping depths of ∼1.0 mm, and ∼0.3 mm. Such a measurement confirms the superior repeatability of this technique among the mechanical technique. Figure 6(a) provides a real image example of a longitudinal section of shaped optical preform proceeded by CO2 laser, while Fig. 6(b) shows a cross-section of four shaped optical preforms in different depths. Such complex shapes are quite impossible to prepare by the shaping based grinding technique. However, thermally-induced stress may build up in the laser shaped preforms. Multi-pass processing with low power laser is advantageous in that as it prevents inducing excessive thermal stress.
Fig. 5. Repeatability measurement of two shaped pieces of the same coreless Heraeus F300 preform at two different depths of 1 mm and 0.3 mm.
Fig. 6. (a) Real image of a longitudinal section of the shaped preform. (b) Real images of the cross-sections of various optical shaped preforms (coreless Heraeus F300) where (1) represents preform with hexagon shape, depth ∼1.0 mm, (2) hexagon shape, depth ∼0.5 mm, (3) hexagon shape, depth ∼0.3 mm, and (4) octagon shape, depth ∼0.5 mm.
3. Preform preparation and measurement setup
In order to examine the impact of shaping optical fiber preform on the background losses of the light propagating in the inner-cladding of drawn DC fibers, various preforms were prepared and shaped. The preforms were made of coreless silica glass rods with an average diameter of ∼10 mm. We included three silica glass rods of different purity (trade names: technical Fused Quartz, Heralux-WG, and Heraeus F300) in the investigation. The Heraeus F300 rod was prepared by collapsing a high pure silica tube (Heraeus F300, OH content is 0.2 ppm) in house by MCVD lathe while the other two rods were prepared by the manufacturer. The preforms are then divided into a few pieces so that each piece can be shaped to a different form while keeping one piece without shaping for comparison (see Table 1). After a shaping process, all the preforms were drawn to passive multimode fibers (MMF) with a diameter of 125 ± 5 µm. The average length of the drawn MMFs was ∼100-m. Unlike the circular MMF, the diameter of shaped MMFs fluctuated by ± 5 µm owing to the fact that these fibers were drawn under the manual diameter-control regime. The automated feedback loop control system was unable to cope with varying readings of fiber thickness at different angle projections of the shaped fibers [27]. All the drawn MMFs of each preform were drawn at the same drawing temperature and speed and coated with low refractive index (RI) UV curable polymer to serve as an outer cladding layer of the MMFs. Three types of low-index fluorinated polymers of (Luvantix/EFiRon PC375-AP (RI = 1.382 @852 nm), Fospia/EFiRon SPC373-AP (RI = 1.373 @852 nm), and Fospia/EFiRon XPC373 (RI = 1.373 @852 nm)) were used for coating the fibers of the technical Fused Quartz, Heralux-WG, and Heraeus F300 preforms respectively (see Table 1). The reason for using different polymers for each preform is to compare the consequences of the background losses with respect to the inner-cladding shapes of the fibers of each perform individually. In addition, this will help to investigate the impact of shaping preforms on the background losses of all drawn fibers despite the purity of the preforms or coating materials (see Section 4.1.1). The background inner-cladding losses of the drawn shaped and unshaped MMFs were measured by a cut-back method [28,29]. Figure 7 shows the cut-back measurement setup. A tungsten halogen bulb was used as a broadband light source. The light was coupled into a 2-m long patchcord of standard MMF connected to another 1-m long pigtail of standard MMF prior launched to the fiber under test (FUT). Both standard MMFs (Thorlabs, FG105LCA, AFS105/125Y), have a numerical aperture of 0.22 with close diameter to our fabricated MMFs of 125 µm. Such employing of standard MMFs between the light source and FUT leads to well-homogenized excitation of guided modes in the MMFs. The light was finally launched to the FUT via mechanical alignment using a fusion splicer. The other end of the FUT was stripped 1 cm, uniformly cleaved and flatly inserted into the optical spectrum analyzer (OSA Ando AQ6317) through a bare fiber adapter (BFA) that has a ferrule diameter of 140 µm. The measured transmission spectra are referenced to transmission spectra at the end of FUT shortened to a length of 2-m. The process of the measurement was done under careful handling to the FUT in the setup to avoid any disturbance to the launching light input. It is worth mentioning that the accuracy and repeatability of our measurement were also investigated to be less than 1.5 dB of power variation with respect to the fiber condition in the BFA that was connected to the OSA. A 200-m of circular coreless MMF used for this investigation under seven cases of fiber condition in the BFA. The investigated cases were varied by the striping length of the fiber end, quality of cleaving, flat/non-flat fiber insertion in the BFA, and the ferrule diameter of BFA as shown in Fig. 8(a). Such cases usually have an influence on light scattering between the fiber and the inner face of the adapter ferrule. Figure 8(b) shows the measured peak power output at a wavelength of 880 nm for all the seven cases with a maximum power variation of 1.47 dB. Such a low variation of power measurement in a long fiber of 200-m confirms the accuracy of our measurements and it can be neglected. Note that the launching light condition was the same and fixed for all seven cases while measurement.
Fig. 8. (a) Illustrated diagram of seven cases of fiber condition in the BFA; where (case 1) The fiber end striped 1 mm, uniformly cleaved and flatly inserted in a 300 µm ferrule diameter of the BFA; (case 2) same fiber of case 1 but non-flatly inserted in the BFA, pushed 1 mm outside the ferrule face; (case 3) the fiber striped 1 cm, uniformly cleaved and flatly inserted; (case 4) the fiber striped 2 cm, non-uniformly cleaved and flatly inserted; (case 5) same fiber of case 4 but non-flatly inserted in the BFA; (case 6) the fiber striped 3 cm, non-uniformly cleaved and flatly inserted in 140 µm ferrule diameter; (case 7) the fiber striped 1.5 cm, uniformly cleaved and flatly inserted again in a 300 µm ferrule diameter. (b) It is the output peak power at 880 nm of the measured 200-m passive circular MMF for all seven cases.
Table 1. List of the fabricated MMFs examined in Fig. 9. All the fibers are drawn at an outer diameter of ∼125 µm.
4. Results and discussion
4.1 Inner-cladding losses at wavelengths band from 700 nm to 900 nm
4.1.1 Losses in the shaped and unshaped DC fibers
It was found that both shaping-techniques would induce additional losses in the inner-cladding losses of about ≥ 20 dB. Higher inner-cladding losses were observed in shaped fibers drawn of the shaped preform pieces compared to the circular fibers drawn of unshaped preform pieces, despite the technique of shaping and purity of initial preform material. Figure 9(a) shows the attenuation spectra at wavelengths (700 nm – 900 nm) of various MMFs drawn of three individual coreless preforms (technical Fused Quartz rod, Heralux-WG rod, and Heraeus F300 rod) as described in Table 1. The microscope images of fiber cross-sections are shown in Fig. 9(b). As can be seen in Fig. 9(a) as well as in Table 1, the circular fiber of each preform has always lower attenuation compared to shaped fibers made of the same preform. However, the circular fiber made of Heraeus F300 rod has the lowest attenuation of 7 dB/km among the other circular fibers of technical and Heralux rods; it is owing to its high purity silica glass with low OH content as well as to its low RI outer-cladding material (RI = 1.37). The increased light attenuation in the shaped fibers was attributed to the surface scattering and to the absorption on OH hydroxyl groups (at wavelengths of 0.945 µm, 1.24 µm, and 1.38 µm) and in a polymer (at wavelengths of 1.18 µm, 1.47 µm, and 1.65 µm).
Fig. 9. (a) Attenuation spectra of the circular and shaped fibers drawn of coreless technical glass, Heralux-WG, and Heraeus F300 rods. The abbreviations “T.shaped” and “M.shaped” denote fibers drawn of thermally-, and mechanically-shaped preforms, respectively. (b) Microscope images of the cross-sections of the shaped fibers with a description of their respective shaped preforms.
4.1.2 Losses in shaped fibers with respect to the inner-clad geometry
It was found that the depth of shaping in both techniques has no significant effect on the attenuation as it can be seen for the case of hexagonal-shaped fibers of Heraeus F300. The fiber was drawn by the thermally-shaped preform piece with grooves depth of ∼1.0 mm has just 1 dB higher attenuation than the fiber drawn of the thermally-shaped preform piece with the depth of a groove of ∼ 0.3 mm. While both thermally-shaped fibers exhibit slightly higher attenuation by ∼ 1-2 dB at 793 nm than the mechanically-shaped fiber made of the same preform. On the other hand, it was also found that the number of grooves in the inner-clad of thermally-shaped fibers has also no significant effect on the attenuation. This can be seen in the Heralux-WG fibers where the fiber with a decagonal inner-clad shape has lower attenuation of ∼2 dB at a wavelength of 793 nm compared to the octagonal fiber. Such a low variation in the attenuations of ≤ 2 dB between the fibers can be related to the error measurements of our setup as we discussed this in Section 3. Thus, we assume that there are no significant differences in the attenuations based on this matter.
4.2 Inner-cladding losses at wavelengths band from 400 nm to 900 nm
4.2.1 Losses in the shaped fibers with respect to their preform surface roughness
At a shorter wavelength band (400 nm – 900 nm), the strongest attenuation was observed in a fiber drawn of a mechanically-shaped preform piece. This can be seen in the shaped fibers of Heraeus F300 preform as shown in Fig. 10(a). We attributed this higher attenuation to the scattering caused by the surface roughness of mechanically-shaped preform piece which can be seen in Fig. 10(b). Note that the spectra of Fig. 10(a) are vertically aligned to the same value at an offset wavelength of 850 nm to facilitate their comparison. The inset of Fig. 10(a) shows the attenuation spectra without offset measured in whole spectral interval available with the OSA used. Attenuation growth rates at short wavelengths were much slower for all thermally-shaped fibers. The deep thermal-shaping (1.0 mm) gives a slightly higher attenuation growth rate at short wavelengths compared to shallow thermal-shaping (0.3 mm). Sooth deposition or thermal history [30–33] can account for these minute changes.
Fig. 10. (a) The scattering spectra at short wavelengths of the Heraeus F300 MMFs were vertically aligned to the same value at an offset wavelength of 850 nm. The inset figure represents the whole spectra interval available with the OSA used (without offset). (b) Real snapshots of the mechanical and thermal shaped pieces of the same coreless Heraeus F300 rod with a hexagonal shape. The mechanically shaped piece exhibits a bit rough surface compared to the surface of the thermally shaped piece.
4.2.2 Losses in the mechanical-shaped fibers with respect to their depth of shaping
In order to investigate further the rising scattering in the mechanical-shaped fibers with respect to the depth of shaping, another coreless preform of Heraeus F300 was prepared especially for this investigation. A new sample of Heraeus F300 tube was collapsed to the rod in house using MCVD lathe and divided into three pieces prior to shaping. Two pieces were mechanically-shaped in two different depths of 0.6 mm and 0.3 mm. While the third piece left without shaping to draw as circular fiber. All three pieces were drawn to three passive MMFs with an outer diameter of 125 µm and coated with a fluorinated polymer (Fospia/EFiRon SPC373-AP (RI = 1.373 @852 nm)). The measured attenuation of the three MMFs at the short wavelength band 400 nm – 900 nm is shown in Fig. 11(a). We observed that the attenuation was increased strongly for the fiber drawn of the deeply ground preform piece (0.6 mm), while there is no evidence of attenuation growth for shallow ground (0.3 mm). This indicates that not only the surface roughness may give rise the surface scattering but also the irregularities of sharp preform edges. Figure 11(b) shows longitudinal and cross-section images of the two mechanically-shaped pieces with circular edges separating the ground planes of the shallowly ground piece and sharp edges on the deeply ground piece.
Fig. 11. (a) The scattering spectra at short wavelengths of the second coreless Heraeus F300 MMFs were vertically aligned to the same value at an offset a wavelength of 850 nm. The inset images represent the cross-sections of the respective fibers. (b) Real longitudinal images of the mechanical-shaped perform pieces. The insets represent their cross-section.
4.3 OH-groups reduction with respect to the shaping technique
Interestingly, it was found that the mechanical-shaping technique can preserve the advantage of OH-groups reduction. The water penetrates to a depth of several hundred microns from the surface in the course of preform fabrications using the hydrogen-oxygen torch [34]. This surface layer can be removed by mechanical grinding. In contrast, the thermal-shaping was found to reduce OH-groups but also causes their diffusion deeper towards the preform center. This was verified by observing second and third overtones of the fundamental OH absorption peaks of (2v3 @1.38 µm) and (3v3 @0.945 µm) indifferently treated fibers drawn of two Heraeus F300 preforms. The results are summarized in Figs. 12(a) and 12(b), respectively.
Fig. 12. Spectra of OH attenuation peaks, where (a) represent OH attenuation peak of 2v3 @1.38 µm were vertically aligned to the same value at an offset wavelength of 1320 nm, and (b) represent the OH attenuation peak of 3v3 @0.945 µm were vertically aligned to the same value at an offset wavelength of 964 nm. The abbreviations TS and MS denote fibers drawn of thermally-, and mechanically-shaped preforms, respectively.
Attenuation spectra were vertically aligned to the same values at an offset wavelength of 1320 nm and 964 nm in Figs. 12(a) and 12(b) respectively, to facilitate their comparison. The circular-unshaped fibers of the coreless Heraeus F300 preforms have obviously the highest attenuation peaks owing to the high contain OH-groups. While the mechanically-shaped fibers exhibit a strong reduction OH-group attenuation peaks owing to the consequences of the grinding process to their preforms. The mechanically-shaped fibers with a depth of (0.6 mm) of both preforms have significantly reduced the OH absorption peak at 1383 nm by 250 dB/km, and the shallow ground fiber with a depth of (0.3 mm) reduced the OH absorption by 216 dB/km. On the other hand, thermal-shaping reduced the OH absorption peak at 1383 nm by 121 dB/km for deeply shaped preform (1.0 mm) and by ∼ 158 dB/km for shallow shaping (0.3 mm). Quite interestingly, the fiber with shallow thermal-shaping featured larger OH reduction compared to the thermally-shaped fiber with 1.0 mm depth. This is probably due to the farther OH diffusion towards the preform axis at the case of shaping with 1.0 mm depth owing to the longer exposition to the CO2 laser. Figure 12(b) shows similar results for the third OH overtone laying at a wavelength of 945 nm. Note that this wavelength is rather close to pumping wavelength (∼940 nm) of ytterbium-doped silica-phosphate fibers [35,36] and this effect should be considered when appropriate shaping technique is selected for these fibers.
4.4 Polymer absorption with respect to the depth shaping technique
It was found that the deeper shaping leads to higher attenuation attributed to a polymer (outer-cladding) absorption regardless of the shaping technique and coating materials. Figure 13 shows the absorption peaks at a wavelength of 1.475 µm, attributed to polymer coatings, as measured in fibers drawn from both coreless Heraeus F300 preforms. As can see from Fig. 13, the thermally-shaped fiber with 1.0 mm depth has a higher polymer absorption peak compared to all the other drawn fibers. This is possibly owing to the fact that fiber with 1.0 mm deep grooves has a longer perimeter. The fiber with longer perimeter has a larger surface area in contact with the polymer, and hence more absorption into the polymer. We have discussed this in [26]. The mechanical-shaped fiber with 0.6 mm deep grooves, coated with SPC 373 shows higher polymer absorption compared to circular fiber (considered has larger perimeter) coated with the same polymer, which is unexpected. This can be possibly explained by imperfections of the fiber sharp edges.
Fig. 13. Spectra of polymer absorption peaks were vertically aligned to the same value at an offset wavelength of 1538 nm. The absorption peak at 1475 nm corresponds to the overtone 2vCH+δCH [37].
5. Conclusions
The impact of optical fiber preforms shaping on the inner-cladding losses of DC fibers was experimentally investigated. Two shaping techniques of mechanical based grinding and thermal-based CO2 laser shaping were employed. The losses were measured by the cut-back method in multimode coreless fibers compatible with a multimode pump fiber of 125 µm in diameter. The outer-cladding was formed by different fluorinated polymers of PC375-AP, SPC-373AP or high-temperature resilient polymer XPC-373 (Fospia) for comparison. Inner-cladding losses of shaped fibers were compared to that of circular fibers. Results show that preform shaping induces significant additional losses. The surface scattering dominated at the short wavelengths (400 nm -900 nm) for the fibers drawn of the mechanically-shaped preform. While the sharp edges imperfections had been identified as another loss factor. Such losses sources were absent in fibers drawn of thermally-shaped preforms.
The mechanical-shaping technique retains the advantage of a significant reduction of absorption peaks related to OH content by removing the surface layer precipitated by water in the course of preform collapse in the flame of the hydrogen-oxygen torch. In thermal-shaping, the OH-groups diffuse deeper towards the preform center.
We observed growths of specific absorption peaks attributed to the polymer of outer-cladding. The magnitude of the peak was proportional to the perimeter of the fiber cross-section. The technique of preform shaping and coating must be carefully considered with respect to the pumping wavelengths. One of the OH-groups absorption peaks, laying at 945 nm, is rather close to the pumping wavelength of ytterbium-doped phosphate fibers, while one of the polymer absorption peaks laying at 1475 nm, could be of importance for the in-band-pumped erbium-doped DC fibers. The detrimental effect of these absorption peaks can be offset by proper fiber design and selection of the preform shaping technique.
Funding
Grantová Agentura České Republiky (GAP19−03141S).
Acknowledgments
We thank Dr. Peter Shardlow from the Optoelectronics Research Centre, the University of Southampton for valuable discussions and advice in this work. Portions of this work were presented at the conference Micro-structured and Specialty Optical Fibers, which took place 1 - 4 April 2019 in Prague [26].
Disclosures
The authors declare no conflicts of interest.
References
1. D. J. Richardson, J. Nilsson, and W. A. Clarkson, “High power fiber lasers: current status and future perspectives [Invited],” J. Opt. Soc. Am. 27(11), B63–B92 (2010). [CrossRef]
2. R. Maurer, “Optical waveguide light source,” U.S. Patent 3,808,549 (1974).
3. E. Snitzer, H. Po, F. Hakimi, R. Tuminelli, and B. C. McCollumn, “Double-clad, offset core Nd fiber laser,” in Proc. Opt. Fiber Sensors, PD5, 533–535 (1988).
4. M. N. Zervas and C. A. Codemard, “High Power Fiber Lasers: A Review,” IEEE J. Sel. Top. Quantum Electron. 20(5), 219–241 (2014). [CrossRef]
5. V. Smith and J. J. Smith, “Mode instability in high power fiber amplifiers,” Opt. Express 19(11), 10180–10192 (2011). [CrossRef]
6. T. Eidam, C. Wirth, C. Jauregui, F. Stutzki, F. Jansen, H. J. Otto, O. Schmidt, T. Schreiber, J. Limpert, and A. Tünnermann, “Experimental observations of the threshold-like onset of mode instabilities in high power fiber amplifiers,” Opt. Express 19(14), 13218–13224 (2011). [CrossRef]
7. C. Jauregui, J. Limpert, and A. Tünnermann, “High-power fibre lasers,” Nat. Photonics 7(11), 861–867 (2013). [CrossRef]
8. F. Kong, G. Gu, T. W. Hawkins, M. Jones, J. Parsons, M. T. Kalichevsky-Dong, S. P. Palese, E. Cheung, and L. Dong, “Efficient 240W single-mode 1018nm laser from an Ytterbium-doped 50/400µm all-solid photonic bandgap fiber,” Opt. Express 26(3), 3138–3144 (2018). [CrossRef]
9. A. Sincore, J. D. Bradford, J. Cook, L. Shah, and M. C. Richardson, “High Average Power Thulium-Doped Silica Fiber Lasers: Review of Systems and Concepts,” IEEE J. Sel. Top. Quantum Electron. 24(3), 1–8 (2018). [CrossRef]
10. S. Jetschke, S. Unger, M. Leich, and J. Kirchhof, “Photodarkening kinetics as a function of Yb concentration and the role of Al codoping,” Appl. Opt. 51(32), 7758–7764 (2012). [CrossRef]
11. S. Suzuki, H. A. McKay, X. Peng, L. Fu, and L. Dong, “Highly ytterbium-doped silica fibers with low photo-darkening,” Opt. Express 17(12), 9924–9932 (2009). [CrossRef]
12. P. Koska, P. Peterka, and V. Doya, “Numerical modeling of pump absorption in coiled and twisted double-clad fibers,” IEEE J. Sel. Top. Quantum Electron. 22(2), 55–62 (2016). [CrossRef]
13. P. Leproux, S. Février, V. Doya, P. Roy, and D. Pagnoux, “Modeling and optimization of double-clad fiber amplifiers using chaotic propagation of the pump,” Opt. Fiber Technol. 7(4), 324–339 (2001). [CrossRef]
14. D. Kouznetsov and J. V. Moloney, “Efficiency of pump absorption in double-clad fiber amplifiers. II. Broken circular symmetry,” J. Opt. Soc. Am. 19(6), 1259–1263 (2002). [CrossRef]
15. J. Nilsson, S. U. Alam, J. A. Alvarez-Chavez, P. W. Turner, W. A. Clarkson, and A. B. Grudinin, “High-power and tunable operation of erbium-ytterbium co-doped cladding-pumped fiber lasers,” IEEE J. Quantum Electron. 39(8), 987–994 (2003). [CrossRef]
16. H. Zellmer, A. Tünnermann, H. Welling, and V. Reichel, “Double-Clad Fiber Laser with 30 W Output Power,” in Optical Amplifiers and Their Applications, Opt. Soc. Am., 16, p. FAW18 (1997).
17. C. A. Codemard, A. Malinowski, and M. N. Zervas, “Numerical optimisation of pump absorption in doped double-clad fiber with transverse and longitudinal perturbation,” in Fiber Lasers XIV: Technology and Systems, Int. Soc. Opt. Photonics, 10083, p. 1008315 (2017).
18. P. Koska, P. Peterka, J. Aubrecht, O. Podrazky, F. Todorov, M. Becker, Y. Baravets, P. Honzatko, and I. Kasik, “Enhanced pump absorption efficiency in coiled and twisted double-clad thulium-doped fibers,” Opt. Express 24(1), 102–107 (2016). [CrossRef]
19. A. A. Jasim, J. Aubrecht, P. Peterka, M. Kamrádek, O. Podrazký, F. Todorov, I. Kašík, and P. Honzátko, “Efficient Pump Absorption in Twisted Double Clad Thulium-Doped Fibers Drawn of CO2 Laser Shaped Preform,” In Proc. CLEO-Europe and EQEC, München, CJ-p-8 (2019).
20. L. Talbot, T. Boilard, and M. Bernier, “Fiber laser pump reflector based on volume Bragg grating,” In Proc. CLEO-Europe and EQEC, München, p. CJ. 11, 5 (2019).
21. D. Toral-Acosta, A. Martinez-Rios, R. Selvas-Aguilar, F. Chavez-Gutierrez, T. E. Porraz-Culebro, K. Madrazo de la Rosa, G. Salceda-Delgado, and J. A. Guerrero-Viramontes, “Asymmetric taper scrambler for improved pump homogeneity and absorption in side-pumped double-clad fibre lasers,” Laser Phys. Lett. 16(8), 085105 (2019). [CrossRef]
22. V. V. Buchanov, M. O. Kaptakov, E. N. Murav’ev, V. I. Revenko, and E. F. Solinov, “Theoretical Research of Laser Glass Cutting,” Glass Phys. Chem. 44(3), 254–259 (2018). [CrossRef]
23. V. K. Sysoev, V. I. Masychev, B. P. Papchenko, S. Y. Rusanov, A. A. Yakovlev, and N. P. Glukhoedov, “High-rate IR laser evaporation of silica glass,” Inorg. Mater. 39(5), 532–537 (2003). [CrossRef]
24. M. Brenci, P. F. Checcacci, R. Falciai, and A. M. Scheggi, “Multimode optical fibers with noncircular cross section,” Radio Sci. 16(4), 535–540 (1981). [CrossRef]
25. P. C. Shardlow, R. Standish, J. Sahu, and W. A. Clarkson, “Cladding Shaping of Optical Fibre Preforms via CO2 Laser Machining,” in European Conference on Lasers and Electro-Optics, Opt. Soc. Am., CJ-P-29 (2015).
26. A. A. Jasim, O. Podrazký, P. Peterka, F. Todorov, and P. Honzátko, “Experimental investigation and characterization of fabrication shaped clad optical fiber by thermally polishing optical fiber preforms with CO2 laser,” in Micro-structured and Specialty Optical Fibres VI, Int. Soc. Opt. Photonics 11029, 1102909 (2019).
27. O. Podrazký, I. Kašík, P. Peterka, J. Aubrecht, J. Cajzl, J. Proboštová, and V. Matějec, “Preparation of optical fibers with non-circular cross-section for fiber lasers and amplifiers,” In Photonics, Devices, and Systems VI, Int. Soc. Opt. Photonics 9450, 94501A (2015).
28. E. Desurvire., and M. N. Zervas, Erbium-doped fiber amplifiers: principles and applications. Wiley Series in Telecommunications and Signal Processing, (1994).
29. M. Bukshtab, Photometry, radiometry, and measurements of optical losses. Springer Series in Optical Sciences, 2nd Ed., Chap. 11, 209 (2019).
30. H. Kakiuchida, K. Saito, and A. J. Ikushima, “Precise determination of fictive temperature of silica glass by infrared absorption spectrum,” J. Appl. Phys. 93(1), 777–779 (2003). [CrossRef]
31. Y. L. Peng, A. Agarwal, M. Tomozawa, and T. A. Blanchet, “Radial distribution of fictive temperatures in silica optical fibers,” J. Non-Cryst. Solids 217(2-3), 272–277 (1997). [CrossRef]
32. K. Tsujikawa, K. Tajima, and M. Ohashi, “Rayleigh scattering reduction method for silica-based optical fiber,” J. Lightwave Technol. 18(11), 1528–1532 (2000). [CrossRef]
33. P. Guenot, P. Nouchi, and B. Poumellec, “Influence of drawing temperature on light scattering properties of single-mode fibers,” in Optical Fiber Communication Conference, Opt. Soc. Am., p. ThG2 (1999).
34. J. Paul and S. Shah, “A case step study for reducing water peak in optical fiber,” In Proc. Photonics, New Delhi, India (2006).
35. J. Świderski, A. Zając, M. Skorczakowski, Z. Jankiewicz, and P. Konieczny, “Rare-earth-doped high-power fiber lasers generating in near infrared range,” Opto-Electron. Rev. 12(2), 169–173 (2004).
36. Y. W. Lee, S. Sinha, M. J. F. Digonnet, R. L. Byer, and S. Jiang, “20 W single-mode Yb3+-doped phosphate fiber laser,” Opt. Lett. 31(22), 3255–3257 (2006). [CrossRef]
37. T. Matsuura, S. Ando, S. Sasaki, and F. Yamamoto, “Polyimides derived from 2, 2'-bis (trifluoromethyl)-4, 4'-diaminobiphenyl. 4. Optical properties of fluorinated polyimides for optoelectronic components,” Macromolecules 27(22), 6665–6670 (1994). [CrossRef]
