August 1, 2014 / Vol. 39, No. 15 / OPTICS LETTERS
4561
Highly efficient higher-order modes filtering into aperiodic very large mode area fibers for single-mode propagation Aurélien Benoît,1,3,* Romain Dauliat,2 Raphaël Jamier,1 Georges Humbert,1 Stephan Grimm,2 Kay Schuster,2 François Salin,3 and Philippe Roy1 1
Xlim Research Institute, UMR CNRS/University of Limoges no. 7252, Photonics Department, 123 Avenue Albert Thomas, 87060 Limoges, France 2 Institute of Photonic Technology, Albert-Einstein-Straβe 9, 07745 Jena, Germany 3
Eolite Systems, 11 Avenue de la Canteranne, 33600 Pessac, France *Corresponding author: [email protected]
Received March 13, 2014; revised June 1, 2014; accepted June 30, 2014; posted June 30, 2014 (Doc. ID 207872); published July 29, 2014 We report here on the first experimental demonstration, to the best of our knowledge, of a new generation of very large mode area (VLMA) fibers intended to strengthen single-mode propagation. The originality of this work relies on an aperiodicity of the inner cladding microstructuration exacerbating the spatial rejection of higher-order-modes (HOMs) while preserving a significant confinement of the fundamental mode. The single-mode behavior was demonstrated using an optical low-coherence interferometry measurement based on the group-velocity dispersion. As suggested through a preliminary numerical approach, this outstanding characteristic/behavior is evidenced over a large spectral range spanning from 1 to 2 μm for a core diameter of 60 μm. Core scalability was also investigated. © 2014 Optical Society of America OCIS codes: (060.4005) Microstructured fibers; (060.2430) Fibers, single-mode; (140.3510) Lasers, fiber. http://dx.doi.org/10.1364/OL.39.004561
The performances of fiber amplifier/laser systems have experienced an impressive rise during the last decade thanks to the development of new fiber structures (leakage channel fibers [1], large pitch fibers (LPFs) [2], or distributed mode filtering fibers [3]…). The latter have notably contributed to pushing the threshold of the appearance of nonlinear processes by ensuring effective single-mode operation associated with large core sizes (core diameters exceeding 50 μm, inducing mode field areas (MFA greater than 10.000 μm2 for instance [2]). Thus, fiber laser sources appear to be serious candidates for high power applications required in various fields, e.g., micro-machining or surgery, and compete undeniably with traditional technologies (bulk or thin-disk sources). Output powers of 10 kW in continuous-wave regime [4], as well as milliJoules level in pulsed regime [5], have been achieved in ytterbium-doped very large mode area (VLMA) fiber lasers with a near-diffractionlimited output beam. However, disruptive phenomena limit the expected power scaling in such VLMA fibers. In particular, the significant increase in fiber core sizes implies the onset of thermally induced modal instabilities when pump/signal powers overcome a certain threshold, strongly damaging the quality of the emitted beam [6]. In this context, a new approach consists of optimizing the waveguide architecture in order to increase the effective single-mode operation in such a VLMA fiber by enhancing the delocalization of higher-order-modes (HOMs). A thorough theoretical investigation has already allowed us to design several aperiodic triple-clad LPFs in which the modal discrimination between the targeted Gaussian-like fundamental mode (FM) and the HOMs is based on a delocalization of HOMs out of the gain region whereas the FM exhibits a large overlap with the core [7]. Our approach consists of breaking the inner 0146-9592/14/154561-04$15.00/0
cladding symmetry to prohibit the existence of the resonant modes thereon which can be substantially located inside the core. Here, intending to ensure an efficient single-mode operation through a short length of interaction with the gain, the modal discrimination has to be maximized. An example of the cross section of a triple-clad all-solid aperiodic LPF that we explored is shown in the inset of Fig. 1. Unlike air/silica LPFs, the core refractive index of our aperiodic LPF does not have to match that of pure silica.
Fig. 1. Comparison of the computed spectral evolution of the modal overlap factor in a standard air-silica LPF (see Ref. [5]) and in the aperiodic LPF exhibiting an MFA of about 2500 μm2 . The solid lines represent the overlap of the FM and the dashed lines, the most confined HOM. ΔΓ represents the modal discrimination, difference between the two overlap factors. Inset, 2D refractive index profile of the two fiber structures (dark blue region, pure silica; clear blue area, index-matching passive material; red hexagons, rare-earth doped material; yellow region, air). © 2014 Optical Society of America
4562
OPTICS LETTERS / Vol. 39, No. 15 / August 1, 2014
The lake of such a restriction implies that a high rareearth ion concentration can be reached, across a larger area, inducing the achievement of higher levels of gain and possibly shortening the fiber length. Furthermore, it also offers the benefit of freely codoping the fiber core with photodarkening mitigating dopants (e.g., P or Ce [8,9]). Thus, the background inner cladding material (clear blue area in Fig. 1) is passively doped to match the index of the active doped-core medium (red area in Fig. 1). The red area depicted in Fig. 1 represents here the section of the fiber core that is intended ultimately to be actively doped while carefully ensuring its perfect index-matching with the inner cladding (in clear blue). Finally, low index inclusions (dark blue area in Fig. 1) are introduced in an aperiodic way to enable a strong confinement of the FM as well as the exacerbation of the HOMs’ delocalization out of the core. In this Letter, we show evidence of the robustness of the single-mode behavior that occurs in the proposed aperiodic all-solid LPF by experimentally validating the waveguide properties of a double-clad passive LPF structure. These rod-type passive fibers (∼1 mm outer diameter) are composed of an inner microstructuration surrounding the fiber core and an outer-clad created by the ambient air around the rod-type fiber. Consequently, FM and HOMs have negligible propagation losses that can be measured, contrary to that of leakage channel fibers reported in [10] and [11]. Figure 1 compares, across a large spectral range (extending from 800 to 2200 nm), the computed modal overlap factors (defined as overlap factor of the FM and that of the most confined HOM with the core region intended to be actively doped; see details in [7]) of the proposed aperiodic LPF with that of the state-of-the-art air-silica LPF structure presented in Ref. [5]. One can underline through the comparison of the theoretical modal discrimination for both structures that the aperiodic lattice enables, for a fixed MFA, an enhancement of the fundamental confinement into the core by 10% while reinforcing the delocalization of HOM out of the core. Indeed, the modal discrimination is improved from 40 to 55% across the entire near-IR range considered for the proposed aperiodic LPF. Despite the fact that in our aperiodic LPF we consider the overlap factor of modes with the whole core section, unlike in air/silica LPF, our proposed fiber structure numerically demonstrates a higher propensity to discriminate the HOMs. Most of the time, VLMA LPFs are based on a slightly multimode waveguide and the HOMs’ delocalization is emphasized by the so-called preferential gain effect. Here, in order to first experimentally observe the relevance of the proposed microstructure regarding the robustness of the single-mode behavior, we have fabricated a passive aperiodic LPF based on the 2D theoretical index profile represented in Fig. 1. Figure 2 presents the cross section of the fabricated rod-type aperiodic fiber. Here, the core and the background materials (corresponding to the red and clear blue areas of the picture depicted in Fig. 1), are made of pure silica. The inner aperiodic microstructuration (corresponding to the dark blue areas of the picture depicted in Fig. 1) has been obtained using 18 low-index boron-doped silica rods.
Fig. 2. Cross section of the fabricated passive rod-type LPF with a close-up of the inner aperiodic structure. The white arrow indicates the small axis of the core diameter.
One may note that the fabricated fiber is a rod-type structure (external diameter equal to 1.4 mm), the goal of which is to mitigate the shrinking of the mode area observed conventionally in bent VLMA fibers [12]. The non-circular core (area defined by the first layer of low-index boron-doped inclusions) exhibits a transversal size of 60 μm (on the small axis of the core as shown by the white arrow in Fig. 2) and an MFA of about 2900 μm2 at a wavelength of 1 μm. The near-field intensity pattern at the output of a 40-cm-long piece of rod-type LPF was measured at different wavelengths. To do that, we used either a supercontinuum laser radiation delivered by a step-index SMF-28 (Dcore 8.2 μm) with different bandpass filters (BPF with a 10 nm bandpass spectrum) or a home-made 2 μm fiber laser based on an 8 μm step-index thulium-doped fiber. In both cases, the modal content of the rod-type aperiodic LPF was excited using a butt-joint coupling between the step-index fiber and the rod-type LPF under test. It is important to underline here that we have deliberately chosen a mismatched mode coupling between the two fibers: a core size mismatch, a numerical aperture mismatch, a mode-field diameter mismatch, and a modal content mismatch. The goal is to favor the excitation conditions of the HOMs in the aperiodic LPF in order to observe the modal behavior of this waveguide. The results are depicted in Fig. 3 for four particular wavelengths (1, 1.2, 1.6, and 2 μm). We observed a good intensity contrast between the core region and the inner cladding. One may also note that boron-doped inclusions appear clearly (dark areas) on the different images that represent the near-field intensity distributions of the emitted light (see Fig. 3). This illustrates the presence of background intensity because of delocalized HOMs. For each operating wavelength (1, 1.2, 1.6, and 2 μm), only the quasi-Gaussian FM appears at the output of the 40-cm-rod-type aperiodic fiber. Whatever the launching conditions (transverse translation), no distortion on the intensity pattern was noted, which indicates highly efficient single-mode behavior with highly efficient delocalization of the HOMs. Media 1 shows the evolution of a near-field intensity pattern at a working wavelength of 1 μm for a transversal displacement between the step-index and the aperiodic LPF. We can see a single-mode propagation evolution in the core associated to a strong leakage of HOMs. Furthermore, in order to validate the propagation of the sole FM through the fiber core of this aperiodic waveguide, we implemented optical low coherence interferometry (OLCI) measurements in time domain.
August 1, 2014 / Vol. 39, No. 15 / OPTICS LETTERS
4563
Fig. 3. Measured near-field intensity patterns at the output of a 60 μm aperiodic passive rod-type LPF for different working wavelengths: (a) 1 μm (cf. Media 1), (b) 1.2 μm, (c) 1.6 μm, and (d) 2 μm.
The latter allows showing the influence of the groupvelocity dispersion (GVD) on the respective weight of propagated modes [13]. The set-up used is based on a Mach–Zehnder interferometer. The two different arms used to create the coherence trace are respectively a reference arm composed of a delay stage and a probe arm integrating the 40-cm-long-piece of aperiodic LPF under test. The light source is a supercontinuum laser source coupled with a bandpass filter to select the working wavelength. As done previously, we have coupled the launched light in a 20-cm-long piece of step-index SMF-28 and used a butt-joint coupling with the rod-type LPF under test to avoid the selective excitation of the sole FM. Figure 4 shows the intensity of the coherence trace versus either the difference of group index or the localization of two mirrors in the delay stage. The recorded coherence trace in Fig. 4(a) shows one and only one interference peak for a difference of group index of 30.10−3 , corresponding to a delay stage variation of more than 6 mm for a bandpass filter of 950 nm. This trace confirms the efficiency of the rejection of HOMs and the single-mode behavior of the proposed aperiodic VLMA LPF. Figure 4(b) represents the coherence trace recorded with a bandpass filter at 850 nm. We can see in this graphic a first black interference peak for the interference of the FM and a second red peak for the HOMs. These two coherence traces of Fig. 4 corroborate with the observations presented in Fig. 5. During the fabrication stage of the above mentioned 60 μm aperiodic fiber, we have drawn other fiber samples exhibiting different core diameters (29, 40, 47, 75, and 87 μm) from the same preform. In this way, we have studied the scalability of the aperiodic inner microstructuration. All of the samples have been characterized regarding their emitted modal shape at 12 working wavelengths extending from 500 nm to 2 μm (using also a
Fig. 4. (a) Coherence trace recorded using a 10-nm bandpass filter at 950 nm versus the difference of group refractive index and the mirror position with only one pattern interference (Inset: close-up in the peak). (b) Coherence trace recorded using a 10 nm bandpass filter at 850 nm. The black trace represents the FM and the red one represents the HOM’s trace.
Fig. 5. Graphic representation of the modal behavior of the aperiodic LPF for different core diameters versus the working wavelength: red Xs mean a multimode propagation whereas green checkmarks mean efficient single-mode propagation. The green area represents the spectral range where robust single-mode propagation is observed. Insets: measured nearfield intensity pattern, displayed at the true coordinates (wavelength, core diameter).
4564
OPTICS LETTERS / Vol. 39, No. 15 / August 1, 2014
broadband supercontinuum laser source associated with 10 nm bandpass filters and a 2 μm fiber laser source). The length of the different fiber samples is still equal to 40 cm. The 72 measuring points are displayed on the graph using the coordinates (wavelength, core diameter) presented in Fig. 5. The red Xs represent the operating points for which a multimode behavior was observed at the output of the fiber being tested. Conversely, the operating points corresponding to robust single-mode propagation are marked by green checkmarks. Some near-field intensity patterns measured at the output of the fiber being tested are shown also in the inset of Fig. 5 to highlight the modal behavior. Each inset image is displayed on the graphic at the real coordinates. Thanks to all of these measurements, we evidenced an operating spectral domain (green area in Fig. 5) where robust single-mode propagation is obtained. Single-mode propagation was observed across a large spectral range for different core diameters, demonstrating the scalability of the fiber structure and the broadband of single-mode behavior. In conclusion, we present in this Letter the first experimental demonstration, to the best of our knowledge, of a robust single-mode propagation into an undoped-core double-clad aperiodic all-solid LPF. This single-mode behavior is because of a strong modal selection thanks to an innovative aperiodic inner microstructuration which causes an improved leakage of the HOMs. The propagated modes in the fiber were carefully studied using a method based on the optical low-coherence interferometry, confirming the robustness of single-mode propagation. Moreover, this behavior was demonstrated across a large spectral range spanning from λ 1 μm to λ 2 μm for a 60 μm core fiber. Finally, the spectral range of single-mode operation was studied for different core diameters. This proves the scalability of the HOMs’ filtering effect in the proposed aperiodic VLMA optical fibers.
This work, conducted as part as the AVANTAGE project, was co-funded by the European Union and Eolite Systems. EC is involved in the Région Limousin with the “Fonds européen de développement économique et régional.” References 1. L. Dong, J. Li, H. A. McKay, L. Fu, and B. K. Thomas, Proc. SPIE 7195, 71950N (2009). 2. J. Limpert, F. Stutzki, F. Jansen, H.-J. Otto, T. Eidam, C. Jauregui, and A. Tünnermann, Light Sci. Appl. 1, e8 (2012). 3. M. Laurila, M. M. Jørgensen, J. Lægsgaard, and T. T. Alkeskjold, in Conference on Lasers and Electro-Optics Europe, (2013), paper CJ_3_5. 4. M. O’Connor, V. Gapontsev, V. Fomin, M. Abramov, and A. Ferin, in Conference on Lasers and Electro-Optics/ International Quantum Electronics Conference, OSA Technical Digest (CD) (Optical Society of America, 2009), paper CThA3. 5. F. Stutzki, F. Jansen, A. Liem, C. Jauregui, J. Limpert, and A. Tunnermann, Opt. Lett. 37, 1073 (2012). 6. T. Eidam, C. Wirth, C. Jauregui, F. Stutzki, F. Jansen, H.-J. Otto, O. Schmidt, T. Schreiber, J. Limpert, and A. Tünnermann, Opt. Express 19, 13218 (2011). 7. R. Dauliat, D. Gaponov, A. Benoit, F. Salin, K. Schuster, R. Jamier, and P. Roy, Opt. Express 21, 18927 (2013). 8. S. Jetschke, S. Unger, A. Schwuchow, M. Leich, and J. Kirchhof, Opt. Express 16, 15540 (2008). 9. P. Jelger, M. Engholm, L. Norin, and F. Laurell, J. Opt. Soc. Am. B 27, 338 (2010). 10. L. Dong, T.-W. Wu, H. A. McKay, L. Fu, J. Li, and H. G. Winful, IEEE J. Lightwave Technol. 27, 1565 (2009). 11. G. Gu, F. Kong, T. W. Hawkins, P. Foy, and K. Wei, Opt. Express, 21, 24039 (2013). 12. J. M. Fini and J. W. Nicholson, Opt. Express 21, 19173 (2013). 13. D. Schimpf, R. Barankov, K. Jespersen, and S. Ramachandran, in Conference on Lasers and ElectroOptics—Laser Applications to Photonic Applications, OSA Technical Digest (CD) (Optical Society of America, 2011), paper CFM6.
4561
Highly efficient higher-order modes filtering into aperiodic very large mode area fibers for single-mode propagation Aurélien Benoît,1,3,* Romain Dauliat,2 Raphaël Jamier,1 Georges Humbert,1 Stephan Grimm,2 Kay Schuster,2 François Salin,3 and Philippe Roy1 1
Xlim Research Institute, UMR CNRS/University of Limoges no. 7252, Photonics Department, 123 Avenue Albert Thomas, 87060 Limoges, France 2 Institute of Photonic Technology, Albert-Einstein-Straβe 9, 07745 Jena, Germany 3
Eolite Systems, 11 Avenue de la Canteranne, 33600 Pessac, France *Corresponding author: [email protected]
Received March 13, 2014; revised June 1, 2014; accepted June 30, 2014; posted June 30, 2014 (Doc. ID 207872); published July 29, 2014 We report here on the first experimental demonstration, to the best of our knowledge, of a new generation of very large mode area (VLMA) fibers intended to strengthen single-mode propagation. The originality of this work relies on an aperiodicity of the inner cladding microstructuration exacerbating the spatial rejection of higher-order-modes (HOMs) while preserving a significant confinement of the fundamental mode. The single-mode behavior was demonstrated using an optical low-coherence interferometry measurement based on the group-velocity dispersion. As suggested through a preliminary numerical approach, this outstanding characteristic/behavior is evidenced over a large spectral range spanning from 1 to 2 μm for a core diameter of 60 μm. Core scalability was also investigated. © 2014 Optical Society of America OCIS codes: (060.4005) Microstructured fibers; (060.2430) Fibers, single-mode; (140.3510) Lasers, fiber. http://dx.doi.org/10.1364/OL.39.004561
The performances of fiber amplifier/laser systems have experienced an impressive rise during the last decade thanks to the development of new fiber structures (leakage channel fibers [1], large pitch fibers (LPFs) [2], or distributed mode filtering fibers [3]…). The latter have notably contributed to pushing the threshold of the appearance of nonlinear processes by ensuring effective single-mode operation associated with large core sizes (core diameters exceeding 50 μm, inducing mode field areas (MFA greater than 10.000 μm2 for instance [2]). Thus, fiber laser sources appear to be serious candidates for high power applications required in various fields, e.g., micro-machining or surgery, and compete undeniably with traditional technologies (bulk or thin-disk sources). Output powers of 10 kW in continuous-wave regime [4], as well as milliJoules level in pulsed regime [5], have been achieved in ytterbium-doped very large mode area (VLMA) fiber lasers with a near-diffractionlimited output beam. However, disruptive phenomena limit the expected power scaling in such VLMA fibers. In particular, the significant increase in fiber core sizes implies the onset of thermally induced modal instabilities when pump/signal powers overcome a certain threshold, strongly damaging the quality of the emitted beam [6]. In this context, a new approach consists of optimizing the waveguide architecture in order to increase the effective single-mode operation in such a VLMA fiber by enhancing the delocalization of higher-order-modes (HOMs). A thorough theoretical investigation has already allowed us to design several aperiodic triple-clad LPFs in which the modal discrimination between the targeted Gaussian-like fundamental mode (FM) and the HOMs is based on a delocalization of HOMs out of the gain region whereas the FM exhibits a large overlap with the core [7]. Our approach consists of breaking the inner 0146-9592/14/154561-04$15.00/0
cladding symmetry to prohibit the existence of the resonant modes thereon which can be substantially located inside the core. Here, intending to ensure an efficient single-mode operation through a short length of interaction with the gain, the modal discrimination has to be maximized. An example of the cross section of a triple-clad all-solid aperiodic LPF that we explored is shown in the inset of Fig. 1. Unlike air/silica LPFs, the core refractive index of our aperiodic LPF does not have to match that of pure silica.
Fig. 1. Comparison of the computed spectral evolution of the modal overlap factor in a standard air-silica LPF (see Ref. [5]) and in the aperiodic LPF exhibiting an MFA of about 2500 μm2 . The solid lines represent the overlap of the FM and the dashed lines, the most confined HOM. ΔΓ represents the modal discrimination, difference between the two overlap factors. Inset, 2D refractive index profile of the two fiber structures (dark blue region, pure silica; clear blue area, index-matching passive material; red hexagons, rare-earth doped material; yellow region, air). © 2014 Optical Society of America
4562
OPTICS LETTERS / Vol. 39, No. 15 / August 1, 2014
The lake of such a restriction implies that a high rareearth ion concentration can be reached, across a larger area, inducing the achievement of higher levels of gain and possibly shortening the fiber length. Furthermore, it also offers the benefit of freely codoping the fiber core with photodarkening mitigating dopants (e.g., P or Ce [8,9]). Thus, the background inner cladding material (clear blue area in Fig. 1) is passively doped to match the index of the active doped-core medium (red area in Fig. 1). The red area depicted in Fig. 1 represents here the section of the fiber core that is intended ultimately to be actively doped while carefully ensuring its perfect index-matching with the inner cladding (in clear blue). Finally, low index inclusions (dark blue area in Fig. 1) are introduced in an aperiodic way to enable a strong confinement of the FM as well as the exacerbation of the HOMs’ delocalization out of the core. In this Letter, we show evidence of the robustness of the single-mode behavior that occurs in the proposed aperiodic all-solid LPF by experimentally validating the waveguide properties of a double-clad passive LPF structure. These rod-type passive fibers (∼1 mm outer diameter) are composed of an inner microstructuration surrounding the fiber core and an outer-clad created by the ambient air around the rod-type fiber. Consequently, FM and HOMs have negligible propagation losses that can be measured, contrary to that of leakage channel fibers reported in [10] and [11]. Figure 1 compares, across a large spectral range (extending from 800 to 2200 nm), the computed modal overlap factors (defined as overlap factor of the FM and that of the most confined HOM with the core region intended to be actively doped; see details in [7]) of the proposed aperiodic LPF with that of the state-of-the-art air-silica LPF structure presented in Ref. [5]. One can underline through the comparison of the theoretical modal discrimination for both structures that the aperiodic lattice enables, for a fixed MFA, an enhancement of the fundamental confinement into the core by 10% while reinforcing the delocalization of HOM out of the core. Indeed, the modal discrimination is improved from 40 to 55% across the entire near-IR range considered for the proposed aperiodic LPF. Despite the fact that in our aperiodic LPF we consider the overlap factor of modes with the whole core section, unlike in air/silica LPF, our proposed fiber structure numerically demonstrates a higher propensity to discriminate the HOMs. Most of the time, VLMA LPFs are based on a slightly multimode waveguide and the HOMs’ delocalization is emphasized by the so-called preferential gain effect. Here, in order to first experimentally observe the relevance of the proposed microstructure regarding the robustness of the single-mode behavior, we have fabricated a passive aperiodic LPF based on the 2D theoretical index profile represented in Fig. 1. Figure 2 presents the cross section of the fabricated rod-type aperiodic fiber. Here, the core and the background materials (corresponding to the red and clear blue areas of the picture depicted in Fig. 1), are made of pure silica. The inner aperiodic microstructuration (corresponding to the dark blue areas of the picture depicted in Fig. 1) has been obtained using 18 low-index boron-doped silica rods.
Fig. 2. Cross section of the fabricated passive rod-type LPF with a close-up of the inner aperiodic structure. The white arrow indicates the small axis of the core diameter.
One may note that the fabricated fiber is a rod-type structure (external diameter equal to 1.4 mm), the goal of which is to mitigate the shrinking of the mode area observed conventionally in bent VLMA fibers [12]. The non-circular core (area defined by the first layer of low-index boron-doped inclusions) exhibits a transversal size of 60 μm (on the small axis of the core as shown by the white arrow in Fig. 2) and an MFA of about 2900 μm2 at a wavelength of 1 μm. The near-field intensity pattern at the output of a 40-cm-long piece of rod-type LPF was measured at different wavelengths. To do that, we used either a supercontinuum laser radiation delivered by a step-index SMF-28 (Dcore 8.2 μm) with different bandpass filters (BPF with a 10 nm bandpass spectrum) or a home-made 2 μm fiber laser based on an 8 μm step-index thulium-doped fiber. In both cases, the modal content of the rod-type aperiodic LPF was excited using a butt-joint coupling between the step-index fiber and the rod-type LPF under test. It is important to underline here that we have deliberately chosen a mismatched mode coupling between the two fibers: a core size mismatch, a numerical aperture mismatch, a mode-field diameter mismatch, and a modal content mismatch. The goal is to favor the excitation conditions of the HOMs in the aperiodic LPF in order to observe the modal behavior of this waveguide. The results are depicted in Fig. 3 for four particular wavelengths (1, 1.2, 1.6, and 2 μm). We observed a good intensity contrast between the core region and the inner cladding. One may also note that boron-doped inclusions appear clearly (dark areas) on the different images that represent the near-field intensity distributions of the emitted light (see Fig. 3). This illustrates the presence of background intensity because of delocalized HOMs. For each operating wavelength (1, 1.2, 1.6, and 2 μm), only the quasi-Gaussian FM appears at the output of the 40-cm-rod-type aperiodic fiber. Whatever the launching conditions (transverse translation), no distortion on the intensity pattern was noted, which indicates highly efficient single-mode behavior with highly efficient delocalization of the HOMs. Media 1 shows the evolution of a near-field intensity pattern at a working wavelength of 1 μm for a transversal displacement between the step-index and the aperiodic LPF. We can see a single-mode propagation evolution in the core associated to a strong leakage of HOMs. Furthermore, in order to validate the propagation of the sole FM through the fiber core of this aperiodic waveguide, we implemented optical low coherence interferometry (OLCI) measurements in time domain.
August 1, 2014 / Vol. 39, No. 15 / OPTICS LETTERS
4563
Fig. 3. Measured near-field intensity patterns at the output of a 60 μm aperiodic passive rod-type LPF for different working wavelengths: (a) 1 μm (cf. Media 1), (b) 1.2 μm, (c) 1.6 μm, and (d) 2 μm.
The latter allows showing the influence of the groupvelocity dispersion (GVD) on the respective weight of propagated modes [13]. The set-up used is based on a Mach–Zehnder interferometer. The two different arms used to create the coherence trace are respectively a reference arm composed of a delay stage and a probe arm integrating the 40-cm-long-piece of aperiodic LPF under test. The light source is a supercontinuum laser source coupled with a bandpass filter to select the working wavelength. As done previously, we have coupled the launched light in a 20-cm-long piece of step-index SMF-28 and used a butt-joint coupling with the rod-type LPF under test to avoid the selective excitation of the sole FM. Figure 4 shows the intensity of the coherence trace versus either the difference of group index or the localization of two mirrors in the delay stage. The recorded coherence trace in Fig. 4(a) shows one and only one interference peak for a difference of group index of 30.10−3 , corresponding to a delay stage variation of more than 6 mm for a bandpass filter of 950 nm. This trace confirms the efficiency of the rejection of HOMs and the single-mode behavior of the proposed aperiodic VLMA LPF. Figure 4(b) represents the coherence trace recorded with a bandpass filter at 850 nm. We can see in this graphic a first black interference peak for the interference of the FM and a second red peak for the HOMs. These two coherence traces of Fig. 4 corroborate with the observations presented in Fig. 5. During the fabrication stage of the above mentioned 60 μm aperiodic fiber, we have drawn other fiber samples exhibiting different core diameters (29, 40, 47, 75, and 87 μm) from the same preform. In this way, we have studied the scalability of the aperiodic inner microstructuration. All of the samples have been characterized regarding their emitted modal shape at 12 working wavelengths extending from 500 nm to 2 μm (using also a
Fig. 4. (a) Coherence trace recorded using a 10-nm bandpass filter at 950 nm versus the difference of group refractive index and the mirror position with only one pattern interference (Inset: close-up in the peak). (b) Coherence trace recorded using a 10 nm bandpass filter at 850 nm. The black trace represents the FM and the red one represents the HOM’s trace.
Fig. 5. Graphic representation of the modal behavior of the aperiodic LPF for different core diameters versus the working wavelength: red Xs mean a multimode propagation whereas green checkmarks mean efficient single-mode propagation. The green area represents the spectral range where robust single-mode propagation is observed. Insets: measured nearfield intensity pattern, displayed at the true coordinates (wavelength, core diameter).
4564
OPTICS LETTERS / Vol. 39, No. 15 / August 1, 2014
broadband supercontinuum laser source associated with 10 nm bandpass filters and a 2 μm fiber laser source). The length of the different fiber samples is still equal to 40 cm. The 72 measuring points are displayed on the graph using the coordinates (wavelength, core diameter) presented in Fig. 5. The red Xs represent the operating points for which a multimode behavior was observed at the output of the fiber being tested. Conversely, the operating points corresponding to robust single-mode propagation are marked by green checkmarks. Some near-field intensity patterns measured at the output of the fiber being tested are shown also in the inset of Fig. 5 to highlight the modal behavior. Each inset image is displayed on the graphic at the real coordinates. Thanks to all of these measurements, we evidenced an operating spectral domain (green area in Fig. 5) where robust single-mode propagation is obtained. Single-mode propagation was observed across a large spectral range for different core diameters, demonstrating the scalability of the fiber structure and the broadband of single-mode behavior. In conclusion, we present in this Letter the first experimental demonstration, to the best of our knowledge, of a robust single-mode propagation into an undoped-core double-clad aperiodic all-solid LPF. This single-mode behavior is because of a strong modal selection thanks to an innovative aperiodic inner microstructuration which causes an improved leakage of the HOMs. The propagated modes in the fiber were carefully studied using a method based on the optical low-coherence interferometry, confirming the robustness of single-mode propagation. Moreover, this behavior was demonstrated across a large spectral range spanning from λ 1 μm to λ 2 μm for a 60 μm core fiber. Finally, the spectral range of single-mode operation was studied for different core diameters. This proves the scalability of the HOMs’ filtering effect in the proposed aperiodic VLMA optical fibers.
This work, conducted as part as the AVANTAGE project, was co-funded by the European Union and Eolite Systems. EC is involved in the Région Limousin with the “Fonds européen de développement économique et régional.” References 1. L. Dong, J. Li, H. A. McKay, L. Fu, and B. K. Thomas, Proc. SPIE 7195, 71950N (2009). 2. J. Limpert, F. Stutzki, F. Jansen, H.-J. Otto, T. Eidam, C. Jauregui, and A. Tünnermann, Light Sci. Appl. 1, e8 (2012). 3. M. Laurila, M. M. Jørgensen, J. Lægsgaard, and T. T. Alkeskjold, in Conference on Lasers and Electro-Optics Europe, (2013), paper CJ_3_5. 4. M. O’Connor, V. Gapontsev, V. Fomin, M. Abramov, and A. Ferin, in Conference on Lasers and Electro-Optics/ International Quantum Electronics Conference, OSA Technical Digest (CD) (Optical Society of America, 2009), paper CThA3. 5. F. Stutzki, F. Jansen, A. Liem, C. Jauregui, J. Limpert, and A. Tunnermann, Opt. Lett. 37, 1073 (2012). 6. T. Eidam, C. Wirth, C. Jauregui, F. Stutzki, F. Jansen, H.-J. Otto, O. Schmidt, T. Schreiber, J. Limpert, and A. Tünnermann, Opt. Express 19, 13218 (2011). 7. R. Dauliat, D. Gaponov, A. Benoit, F. Salin, K. Schuster, R. Jamier, and P. Roy, Opt. Express 21, 18927 (2013). 8. S. Jetschke, S. Unger, A. Schwuchow, M. Leich, and J. Kirchhof, Opt. Express 16, 15540 (2008). 9. P. Jelger, M. Engholm, L. Norin, and F. Laurell, J. Opt. Soc. Am. B 27, 338 (2010). 10. L. Dong, T.-W. Wu, H. A. McKay, L. Fu, J. Li, and H. G. Winful, IEEE J. Lightwave Technol. 27, 1565 (2009). 11. G. Gu, F. Kong, T. W. Hawkins, P. Foy, and K. Wei, Opt. Express, 21, 24039 (2013). 12. J. M. Fini and J. W. Nicholson, Opt. Express 21, 19173 (2013). 13. D. Schimpf, R. Barankov, K. Jespersen, and S. Ramachandran, in Conference on Lasers and ElectroOptics—Laser Applications to Photonic Applications, OSA Technical Digest (CD) (Optical Society of America, 2011), paper CFM6.
