In this paper we describe a high power narrow-band amplified spontaneous emission (ASE) light source at 1030 nm center wavelength generated in an Yb-doped fiber-based experimental setup. By cutting a small region out of a broadband ASE spectrum using two fiber Bragg gratings a strongly constrained bandwidth of 12 ± 2 pm (3.5 ± 0.6 GHz) is formed. A two-stage high power fiber amplifier system is used to boost the output power up to 697 W with a measured beam quality of M2≤1.34. In an additional experiment we demonstrate a stimulated Brillouin scattering (SBS) suppression of at least 17 dB (theoretically predicted ~20 dB), which is only limited by the dynamic range of the measurement and not by the onset of SBS when using the described light source. The presented narrow-band ASE source could be of great interest for brightness scaling applications by beam combination, where SBS is known as a limiting factor.
© 2011 OSA
Yb-doped high power fibers emitting in the 1 µm wavelength region with diffraction-limited beam quality are well known as a power scalable solid-state laser and amplifier architecture for pulsed and continuous-wave (cw) operation [1–4]. Recently, IPG-Photonics presented a cw-operated master oscillator power amplifier system with ~10 kW of output power . Beside this rapid development in the field of high power fiber lasers, multi-kW cw-emissions usually have several nanometers of bandwidth that might be disadvantageous for some linewidth-sensitive applications. Lately, sub nm bandwidth fiber amplifiers have reached the kilowatt range [6,7].
The primary performance limiting factors for high power cw-operated fiber laser systems can be traced back to two nonlinear effects: stimulated Brillouin scattering (SBS) and stimulated Raman scattering (SRS), depending on the specific radiation linewidth. To circumvent these nonlinearities two different but coexisting approaches are pursuable. First, it is possible to implement advanced large mode area (LMA) fiber designs. By increasing the core size the mode-field area of the guided light is increased too, while the absorption length of the fiber typically is reduced at the same time. Both conditions are beneficial since long propagation lengths and high core intensities support the triggering of SBS and SRS, respectively. For step index fibers it is technologically difficult to scale up the size of the fiber core while preserving the fundamental mode beam quality. Therefore, a number of single-mode large mode area fiber designs have been proposed, among others photonic crystal fibers (PCF) [8,9]. Second, when using narrow-band (fiber) lasers, power scaling techniques like spectral beam combination (SBC) or coherent beam combination of phase locked emissions (CBC) can be of high interest [6,10]. By superimposing the output beams of several emissions the performance of a laser system can be considerably enhanced [11,12].
SBS can appear at much lower output powers if narrow-band single-longitudinal mode signals are used. Thus, SBS is one major problem for beam combining setups even when using such state-of-the-art fiber designs.
Over the last years methods for effective SBS suppression have been proposed which can be categorized basically in two classes, those that take influence on fiber parameters and those related to the signal itself. For instance, a suppression of 4.3 dB can be reached by implementing special acoustically anti-guiding fibers . Also by varying fiber parameters like strain, temperature or core radius along the fiber length a certain SBS suppression can be reached . A proven method to modify the signal is the sinusoidal phase modulation of a narrow-band signal to generate several sidebands with a spectral separation of at least two times the gain bandwidth of SBS (~30 MHz) . But even in this case the accentuated peaks of the sidebands can still trigger SBS.
When using amplified spontaneous emission (ASE) no longitudinal modes or sidebands are present but the photons are equally distributed within the spectral shape (e.g. Gaussian) making SBS most unlikely. 106 W generated in a two-stage fiber-based super fluorescent light source with diffraction-limited beam quality was demonstrated with an optical spectrum spanning from ~1035 to 1100 nm . With a FWHM spectrum of 21 nm this source is mainly directed towards broadband applications. A more spectrally confined (FWHM ~0.5 nm) and tunable ASE source with maximum output power of 135 mW is proposed in . Although phase locking for CBC-purpose was successfully demonstrated even for spectra in the nm range , this source is still not sufficient for e.g. diffraction grating based SBC, where the combined beam quality is strongly depending on the signal bandwidth .
In this contribution we present an ASE source possessing a measured coherence length of 56 mm, which corresponds to a FWHM bandwidth of 12 ± 2 pm (3.5 ± 0.6 GHz), respectively. This spectral bandwidth is too broad to have SBS issues while still narrow enough to meet the requirements needed for SBC  and even CBC . The beam quality at the highest output power level of 697 W is M2 = 1.34 in both directions (4σ-method). To the best of our knowledge the presented fiber-based super fluorescent source has the highest output power while having the smallest linewidth, making it an ideal element for beam combining applications. The performance was limited by the emergence of higher order modes within the final PCF amplifier stage. This power dependent effect is called spatial hole burning and deteriorates the fundamental mode beam quality . A comparison between a single-frequency laser and the presented light source was carried out to prove this approach’s capability to successfully suppress SBS. The measured SBS suppression is at least 17 dB, whereas the dynamic range of the realized measurement is limited by the onset of SRS.
2. Experimental results
2.1. Low power all-fiber front-end
The experimental setup can be split into two functionally different parts, a low-power all-fiber ASE front-end source that molds the spectrum into its final shape and a free-space two-stage amplifier section. The all-fiber front-end source has the advantage of being highly immune from external influences like stress and temperature and is therefore most important for long-term reliable cw-output.
When setting up an ASE source, any cavity build-up has to be extinguished to avoid lasing. Generally, active (especially core-pumped) fibers can have a very high gain. Thus, parasitic cavities will easily lead to laser activity and probably to self-pulsing as well [21,22]. For a setup including spliced all-fiber components it is only possible to minimize but not to avoid the oscillation of a certain number of longitudinal modes due to inherent small back reflections and/or Rayleigh scattered signal light. If the reflectivity becomes too high the lasing threshold can be reached leading to a perturbed ASE generation process at a certain pump level. This can easily be the case when using highly reflective fiber Bragg gratings (FBG) with the intention to spetrally confine the output spectrum. To avoid this generation, spectral shaping by means of two FBG and amplification of the ASE light is done separately. Strong optical isolation between these three stages successfully prevented mutual interactions.
In the experimental setup (shown in Fig. 1 ) all active and passive fibers are polarization-maintaining single-clad transversal single-mode fibers with a mode-field diameter (MFD) of ~7 µm. The single-mode pump laser diode (SM-LD1) at 976 nm is wavelength stabilized by a FBG and additionally includes a current and temperature control. It is split equally (50/50) leaving 200 mW for each pump transmission port (P) of the wavelength division multiplexers WDM1 and WDM2. In the first 2 m long Yb-doped core-pumped fiber (Yb1), which is spliced to the common port (C) of WDM1 a 10 nm broad ASE spectrum reaching from 1025 to 1035 nm is generated. In this crucial section the (4%) Fresnel reflection of one fiber end-facet is used to increase the output power at the reflection port (R) of WDM1 up to 20 mW while staying below the laser threshold. An isolator (ISO1) and the first transition within circulator 1 (CIRC1) is used to isolate the ASE generation process by at least 70 dB. It is worth noting that substituting the Fresnel reflection by a high reflective FBG as well as leaving out ISO1 leads to the described self-pulsing behavior.
FBG1 has a reflectivity of >90% and is reducing the bandwidth by a factor of ~10 down to about 1 nm. The remaining power is subsequently 70 dB-isolated by two further circulator transitions and send to the double pass configuration of the second Yb-doped fiber (Yb2). This also 2 m long core-pumped fiber is amplifying the remaining signal power of 140 µW in a first pass. The relatively small reflectance (32%) of FBG2 and the further reduction of the spectral width down to ~12 pm leads to a low but sufficient seed signal power for the second (counter-propagating) amplification pass through fiber Yb2. The overall gain of this double pass amplifier (including losses caused by FBG2) can be estimated to ~40 dB and is reason for the good efficiency of this front-end.
After the second circulator (CIRC2) the signal power reaches a value of 27 mW, which will then be amplified in an 1 m long active fiber (Yb3) up to an output power of 420 mW (~400 mW after ISO2) when core-pumped with 650 mW by SM-LD2. The all-fiber section is closed with an 8° angle-polished end-facet and a bulk optical isolator (ISO2) to avoid back reflections.
Due to the polarizing nature of the two circulators (four transitions with a polarization extinction ratio of ~25 dB each) the output signal is strongly linear polarized. Usually, when using polarization-maintaining fiber-pigtailed FBG two polarization modes will propagate and cause potentially problems. Thus, in this case no wavelength shifted second polarization mode is observed (Fig. 2 ). At this point it is worth mentioning that there exists the potential of wavelength tunability within the Yb3+ gain spectrum by either heating/cooling the FBG  or by substituting the two matched FBG.
An analysis of the optical spectrum of Fig. 2 shows that 97.3% of the output power is within the center peak. As we will see later, the residual spectrum is further narrowed by an interference filter, which is used to prevent the 10.5 m long main PCF from amplifying background noise and thereby increasing the power within the central peak.
2.2. High power two-stage fiber amplifier system
Fig. 3 shows a schematic illustration of the high power amplifier system. The output of the ASE front-end with a power of 400 mW is used as seed for the first stage, which is a polarizing 1.6 m long single-mode PCF with a 40 µm diameter Yb-doped core region (MFD = 32 µm) and a high NA 200 µm diameter inner cladding (NA~0.6). This fiber is optically pumped by a low brightness fiber-pigtailed (Ø = 200 µm) laser diode at 976 nm wavelength. The output beam with an obtained power of 10 W is spectrally cleaned by the use of a 4 nm bandpass interference filter (IF) to remove residual optical noise and non-signal light lying across the Yb-glass gain spectrum. By this approach, 98.9% of the total output power remains within the steep center peak (Fig. 4 ). The subsequent water-cooled 10.5 m long main amplifier PCF possesses an active core diameter of 42 µm (fundamental mode MFD = 33 µm) and a 500 µm diameter core used for pumping with the 976 nm radiation from a fiber-coupled (Ø = 1mm) laser diode. To prevent the system from lasing both fibers are prepared with angled end-facets and are optically isolated by 35 dB from each other.
Fig. 5 shows the measurement of output power and beam quality factor (4σ-method) versus the launched pump power. The slope efficiency is as high as 69% while the beam quality factor stays stable around M2 = 1.3 for the whole power range. At maximum output power (697 W) a M2 of 1.34 is measured. At higher pump rates high-order modes will appear due to the spatial hole burning effect leading to a decreased beam quality. For instance, at 760 W of output power the M2 value is already degraded to ~1.8. Although the main amplifier fiber is neither polarizing nor polarization maintaining recent experiments showed output emissions with a high degree of linear polarization .
2.3. Experimental investigation of SBS suppression capability
To predict the SBS suppression capability of this system the spectral bandwidth had to be determined first. Since the spectrum is too narrow for high-resolution grating-based optical spectrometer and too broad for most Fabry-Perot interferometers we determined the linewidth by measuring the coherence length Lc on the basis of a Michelson interferometer.
With Lc = 56 mm and assuming a gaussian-shaped spectrum this length corresponds to a spectral bandwidth of Δυ = 0.66 × c0/Lc = 3.5 ± 0.6 GHz (Δλ = 12 ± 2 pm), with c0 being the speed of light. The logarithmic gain of SBS can be described by the following equation:24]. The Brillouin gain bandwidth ΔυB of fused silica is ~30 MHz while Δυ of the presented ASE source is 3.5 GHz. ΔυΒ in relation to Δυ (fraction in equation) shows that our ASE source has a reduced SBS gain by at least a factor of 100 (20 dB) compared to a system with a linewidth of ≤ΔυB. Thus, with this simple assumption 20 dB is the theoretical SBS suppression limit.
To prove this correlation experimentally, the narrow-band ASE source was compared to a single-frequency (SF) laser diode with a measured (self-heterodyne detection) linewidth of 100 kHz. For this purpose the SF light was coupled in a passive 6 km long standard single mode fiber. Already at 4.7 mW of output power the SF light triggers SBS predicting the SBS threshold for the ASE source to be at ~470 mW. This value has not been reached because SRS starts to emerge at 220 mW. Nevertheless, up to this point no indication for SBS was observable. Thus, the dynamic range of this measurement only allows for proving ~17 dB of SBS suppression but the SBS threshold is expected to be even higher and in most cases below the SRS threshold in general. Usually, the Raman gain coefficient is 500 times (27 dB) lower than the Brillouin gain coefficient  and it shows that at a certain point SBS is not the main limiting nonlinear effect anymore but rather SRS.
In this paper we showed a method to generate 697 W of amplified spontaneous emission at 1030 nm wavelength with 12 pm (3.5 GHz) bandwidth and almost diffraction-limited beam quality of M2≤1.34 (4σ-method) by means of a complete fiber-based system. After the main PCF amplifier 98.9% of the output power is within this confined spectral window. The performance of the system was limited by the emergence of higher order modes and hence, the reduction of beam quality. The SBS suppression was at least 17 dB but with the 12 pm linewidth more than 20 dB should be achievable. This measurement was limited by the onset of SRS superseding SBS as the most harmful phenomenon. Other SBS suppression techniques, like the implementation of acoustically anti-guiding fibers and/or phase/amplitude modulation of the optical signal, can be added to further increase the SBS threshold power. Thus, narrow-band super fluorescent sources reveal a high potential for spectral and coherent beam combining applications by avoiding SBS to the greatest extent.
This work was partly supported by Thüringer Ministerium für Bildung, Wissenschaft und Kultur within the Landesprogramm “ProExzellenz” under contract number: PE203-2-1 (“MO-FA – Modenfeldstabile Faserlaser”). A. Kliner acknowledges financial support by the Abbe School of Photonics, Jena (Germany).
References and links
1. J. Limpert, F. Röser, S. Klingebiel, T. Schreiber, C. Wirth, T. Peschel, R. Eberhardt, and A. Tünnermann, “The rising power of fiber lasers and amplifiers,” IEEE J. Sel. Top. Quantum Electron. 13(3), 537–545 (2007). [CrossRef]
2. J. W. Dawson, M. J. Messerly, R. J. Beach, M. Y. Shverdin, E. A. Stappaerts, A. K. Sridharan, P. H. Pax, J. E. Heebner, C. W. Siders, and C. P. J. Barty, “Analysis of the scalability of diffraction-limited fiber lasers and amplifiers to high average power,” Opt. Express 16(17), 13240–13266 (2008). [CrossRef] [PubMed]
4. T. Eidam, S. Hanf, E. Seise, T. V. Andersen, T. Gabler, C. Wirth, T. Schreiber, J. Limpert, and A. Tünnermann, “Femtosecond fiber CPA system emitting 830 W average output power,” Opt. Lett. 35(2), 94–96 (2010). [CrossRef] [PubMed]
5. V. Fomin, M. Abramov, A. Ferin, A. Abramov, D. Mochalov, N. Platonov, and V. Gapontsev, “10kW single mode fiber laser,” SyTu-1.3, Symposium on High-Power Fiber Lasers, 14th International Conference, Laser Optics 2010, St. Petersburg, Fla. (2010).
6. S. J. McNaught, J. E. Rothenberg, P. A. Thielen, M. G. Wickham, M. E. Weber, and G. D. Goodno, “Coherent combining of a 1.26-kW fiber amplifier,” in Advanced Solid-State Photonics, OSA Technical Digest (CD) (Optical Society of America, 2010), paper AMA2.
7. C. Wirth, O. Schmidt, I. Tsybin, T. Schreiber, R. Eberhardt, J. Limpert, and A. Tünnermann, “1kW narrow-linewidth fiber amplifier for spectral beam combining,” in Advanced Solid-State Photonics, OSA Technical Digest (CD) (Optical Society of America, 2008), paper WA6.
9. J. Limpert, O. Schmidt, J. Rothhardt, F. Röser, T. Schreiber, A. Tünnermann, S. Ermeneux, P. Yvernault, and F. Salin, “Extended single-mode photonic crystal fiber lasers,” Opt. Express 14(7), 2715–2720 (2006). [CrossRef] [PubMed]
10. C. Wirth, O. Schmidt, I. Tsybin, T. Schreiber, T. Peschel, F. Brückner, T. Clausnitzer, J. Limpert, R. Eberhardt, A. Tünnermann, M. Gowin, E. ten Have, K. Ludewigt, and M. Jung, “2 kW incoherent beam combining of four narrow-linewidth photonic crystal fiber amplifiers,” Opt. Express 17(3), 1178–1183 (2009). [CrossRef] [PubMed]
12. T. Schreiber, C. Wirth, O. Schmidt, T. V. Andersen, I. Tsybin, S. Böhme, T. Peschel, F. Brückner, T. Clausnitzer, F. Röser, R. Eberhardt, J. Limpert, and A. Tünnermann, “Incoherent beam combining of continuous-wave and pulsed Yb-doped fiber amplifiers,” IEEE J. Sel. Top. Quantum Electron. 15(2), 354–360 (2009). [CrossRef]
13. M. D. Mermelstein, “SBS threshold measurements and acoustic beam propagation modeling in guiding and anti-guiding single mode optical fibers,” Opt. Express 17(18), 16225–16237 (2009). [CrossRef] [PubMed]
14. A. Kobyakov, M. Sauer, and D. Chowdhury, “Stimulated Brillouin scattering in optical fibers,” Adv. Opt. Photon. 2(1), 1–59 (2010). [CrossRef]
15. Y. Liu, Z. Lv, Y. Dong, and Q. Li, “Research on stimulated Brillouin scattering suppression based on multi-frequency phase modulation,” Chin. Opt. Lett. 7(1), 29–31 (2009). [CrossRef]
17. P. Wang, and W. A. Clarkson, “Tunable Yb-doped fiber amplified sontaneous emission source,” in Conference on Lasers and Electro-Optics/International Quantum Electronics Conference, OSA Technical Digest (CD) (Optical Society of America, 2009), paper CFM6.
18. P. Zhou, X. Wang, Y. Ma, K. Han, and Z. Liu, “Active phase locking of fiber amplifiers with 180 GHz ultrabroad linewidth,” Prog. Electromagn. Res. Lett. 17, 145–152 (2010). [CrossRef]
19. S. J. Augst, J. K. Ranka, T. Y. Fan, and A. Sanchez, “Beam combining of ytterbium fiber amplifiers (Invited),” J. Opt. Soc. Am. B 24(8), 1707–1715 (2007). [CrossRef]
20. C. Jauregui, T. Eidam, J. Limpert, and A. Tünnermann, “The impact of modal interference on the beam quality of high-power fiber amplifiers,” Opt. Express 19(4), 3258–3271 (2011). [CrossRef] [PubMed]
21. S. D. Jackson, “Direct evidence for laser re-absorption as initial cause for self-pulsing in three-level fiber lasers,” Electron. Lett. 38(25), 1640–1642 (2002). [CrossRef]
22. B. N. Upadhyaya, A. Kuruvilla, U. Chakravarty, M. R. Shenoy, K. Thyagarajan, and S. M. Oak, “Effect of laser linewidth and fiber length on self-pulsing dynamics and output stabilization of single-mode Yb-doped double-clad fiber laser,” Appl. Opt. 49(12), 2316–2325 (2010). [CrossRef] [PubMed]
23. J. Rothhardt, S. Hädrich, T. Gottschall, J. Limpert, A. Tünnermann, M. Rothhardt, M. Becker, S. Brückner, and H. Bartelt, “Generation of flattop pump pulses for OPCPA by coherent pulse stacking with fiber Bragg gratings,” Opt. Express 17(18), 16332–16341 (2009). [CrossRef] [PubMed]
25. G. P. Agrawal, Nonlinear Fiber Optics, 4th ed. (Academic Press, 2007).