Letter
Vol. 40, No. 22 / November 15 2015 / Optics Letters
5311
Kilowatt-level fiber amplifier with spectralbroadening-free property, seeded by a random fiber laser XUEYUAN DU, HANWEI ZHANG, PENGFEI MA, HU XIAO, XIAOLIN WANG, PU ZHOU,*
AND
ZEJIN LIU
College of Optoelectronic Science and Engineering, National University of Defense Technology, Changsha 410073, China *Corresponding author: [email protected] Received 7 July 2015; revised 30 August 2015; accepted 18 September 2015; posted 25 September 2015 (Doc. ID 245442); published 9 November 2015
In this Letter, we demonstrate a kilowatt (kW) level highpower fiber laser amplifier with a clear sign of spectralbroadening-free property. The high-power fiber lasing is realized by employing a master oscillator power-amplifier (MOPA) configuration, seeded by a temporally stable random fiber laser (RFL) that utilizes Raman amplification and random distributed feedback from a long passive fiber. The output power reaches 1.03 kW with a 1070 nm wavelength and an optical-to-optical efficiency of 74.6%. Despite the typical nonlinear spectral broadening in most traditional MOPA systems, the output spectral linewidth is well maintained during the whole high-power amplification process. The suppressed linewidth broadening in the spectral domain during high-power amplification is significant for further power scaling, spectral beam combination, and other applications that require narrow-linewidth highpower lasing. © 2015 Optical Society of America OCIS codes: (060.2320) Fiber optics amplifiers and oscillators; (140.3490) Lasers, distributed-feedback; (290.5870) Scattering, Rayleigh; (290.5910) Scattering, stimulated Raman. http://dx.doi.org/10.1364/OL.40.005311
Fiber lasers and amplifiers have shown great industrial application potential and contain prominent properties such as high conversion efficiency, outstanding beam quality, stable compact structure, and convenient heat management [1–3]. Thanks to the introduction of a Ytterbium-doped fiber amplifier (YDFA) based on master oscillator power–amplifier (MOPA) structural design, the power scalability of continuous wave (CW) fiber amplifier has raised significantly to multi-kW level [4,5]. Considering the high-power intensity and relatively long fiber length, the nonlinearity effect has clear influence on highpower fiber laser performance, one of which is spectral broadening [3,6]. Previous theoretical analysis on the spectral broadening has revealed the correlation between spectral broadening and nonlinearity, considering the effects of four-wave mixing (FWM) [7], self-phase modulation (SPM) [8], or groupvelocity dispersion (GVD) [9], and some theoretical analyses 0146-9592/15/225311-04$15/0$15.00 © 2015 Optical Society of America
on the spectral evolution of CW YDFA, considering the nonlinear effect has also been carried out [10]. In practical conditions where the length of the active fiber in a high-power YDFA is relatively short (several meters), the phase shift caused by dispersion can be neglected. The major phase shift leading to the spectral broadening is induced by the nonlinear effects. The previously reported models derived from the generalized nonlinear Schrödinger (NLS) equation have considered the self-phase modulation (SPM) effect as the main reason for nonlinear spectral broadening involving some weak FWM interactions [11,12]. However, conventional MOPA YDFAs usually use a typical Yb-doped fiber oscillator as the seed source, which suffers self-pulsation effect with temporal instability and intensity fluctuations [13–15]. The intensity fluctuations in the oscillator-based lasers cause higher peak powers compared with temporally stable seed sources, which results in strong nonlinear effect under high pump powers. Due to the close connection between the nonlinear effect and the laser’s temporal instability, the spectral broadening is expected to be controlled by eliminating the temporal fluctuations. Single-frequency MOPA systems based on the single-frequency seed source also show good temporal stability and the preservation of the spectral properties [16–18]. However, further power scalability of the single-frequency MOPAs is mainly restricted by the stimulated Brillouin scattering (SBS) effect, and power output exceeding KW level has not yet been reported. Thus, we intend to build a temporally stable seed source to suppress the nonlinear spectral broadening during high-power amplification and to optimize the output property of MOPA system. Random fiber laser (RFL) has drawn great attention in recent years [19], and researches, including power scaling, temporal stability, spectral evolution, intensity-noise transfer, and application potentials have been widely investigated [20–27]. As RFL utilizes random distributed feedback (RDFB) via Rayleigh scattering in long passive fibers, one remarkable feature is that the increase of pump power well above threshold leads to stabilized radiation both in the temporal and spectral domains [19]. In addition, during the relative intensity-noise (RIN) transfer in RFL configuration, pump noise can be minimized with the increase of pump power or using a longer
5312
Vol. 40, No. 22 / November 15 2015 / Optics Letters
passive fiber [25,28]. Therefore, a temporally stable seed laser source based on RFL configuration can be a practical solution to overcome the temporal instability of conventional seed source. In this Letter, a temporally stable 1070 nm random fiber laser pumped by 1018 nm Ytterbium-doped fiber laser (YFL) with steady 10 W RFL power output is for the first time to the best of our knowledge prepared as the seed source, followed by kW level high-power amplification based on MOPA configuration. The specified 1.06–1.09 μm wavelength range for a MOPA system ensures efficient power amplification according to the gain property of Yb-ion. Thus, we choose a 1070 nm wavelength RFL to be the seed source applicable for power amplification, and the pump laser should be correspondingly near 1018 m wavelength referring to the Raman shift value. However, most of the previously reported RFLs failed to cover this wavelength range due to the small net gain of Yb-doped fiber near 1018 nm wavelength [29], and the devastating self-oscillation or spurious oscillation caused by the ASE effect [30,31]. First, the 1018 nm YFL which served as the pump laser of the 1070 nm RFL has to be optimized under tremendous work considering influential factors such as the effect of ASE feedback, the fiber type and fiber length of Yb-doped fiber, reflectivity of the fiber Bragg gratings (FBGs), and the pumping direction. [30]. We build an YFL at 1018 nm with stable highpower output under optimized structure design. Based on that, the 1070 nm random fiber laser pumped by 1018 nm YFL can be achieved by utilizing stimulated Raman scattering (SRS) and RDFB from passive fiber. The experimental setup of the 1070 nm RFL seed laser illustrates a forward-pumped half-opened cavity structure in Fig. 1. The pump laser is a 1018 nm YFL, and a 1018 nm bandpass filter (BPF) is included in the structure as a precaution against backward feedback around ASE wavelength (1030 nm), which often results in spurious oscillation. By using a 1018/1070 nm wavelength division multiplexer (WDM), the pump laser is injected into the cavity via the WDM. The common port of the WDM is spliced into a 0.5 km HI1060 optical fiber, which provides both Raman gain and random distributed feedback via Rayleigh scattering. Intracavity light is reflected back by a 1070 nm high-reflectivity fiber Bragg grating (HR-FBG) at the 1070 nm port of the WDM, which helps form a half-opened RFL structure [20]. At the output end of the RFL seed laser,
Fig. 1. Experimental setup of the RFL seed and the amplifier. BPF, bandpass filter; WDM, wavelength division multiplexer; P 4 , extra port of WDM; RDFB, random distributed feedback; HR-FBG, highreflectivity fiber Bragg grating; LD, 976 nm laser diode; PM, power meter; PD, photodetector; OSC, oscilloscope; OSA, optical spectrum analyzer.
Letter the 1070/1120 nm WDM functions as a spectral filter to eliminate the second-order Raman Stokes wave components. All the free ends in the RFL structure are angle cleaved to avoid any Fresnel reflection. Some essential factors, including power efficiency, spectral purity, and temporal stability, are considered during the preparation of the 1070 nm RFL seed. For traditional random fiber lasers, the passive fiber used to provide feedback and gain is usually several or several tens of kilometers, which causes relatively low higher-order Raman Stokes threshold and limits the power scalability of RFL [32]. In this case, the short-cavity structure using a 0.5 km passive fiber instead of a longer one enables sufficient power output of first-order Stokes wave at 1070 nm, but also obviously increases the random lasing threshold [22,33,34]. However, the half-opened cavity structural design by introducing a HR-FBG from one end can decrease the RFL lasing threshold to approximately half, which helps overcome the obstacle of high lasing threshold in shortcavity RFL [20]. In addition, another advantage of utilizing short-cavity structural design is that, by decreasing the passive fiber length, the optical conversion efficiency becomes higher and even approaches the quantum limit due to the reduction of the pump and Stokes wave attenuation [32,33]. With the above structural optimization, the achieved 1070 nm RFL seed can finally reach 10 W level power output before the notable emergence of second-order Raman Stokes spectral constituents. The spectral feature of the RFL seed is measured through an optical spectrum analyzer and shown in Fig. 2(a). By numerical integrating, the 1070 nm component is calculated to contain more than 99.9% of the total power, which guarantees the spectral purity of the seed laser. Figure 2(b) illustrates the temporal characteristic monitored by a 180 M Hz high-speed photodetector and an oscilloscope with 1 GHz bandwidth. According to the measured temporal behavior and, correspondingly, its radiofrequency (RF) spectrum, there is no longitudinal mode beating signal corresponding to the cavity length, and the RFL operates in temporally stable CW regime without pulsation. In a conventional fiber laser based on fiber oscillator structure, there exists oscillating peaks corresponding to the cavity length which are easily observed in the frequency domains. The temporally modulated radiation with peak power of several or several tens of times higher than the measured average power causes strong nonlinear effects. To suppress the spectral broadening, the RFL seed source should be well prepared ensuring that the frequency spectrum is stable and smooth, as illustrated in Fig. 2(b). The fiber amplifier system depicted in Fig. 1 is a typical MOPA regime of two stages: a master oscillator (MO) formed by a random fiber laser working as the seed source and a power
Fig. 2. (a) Measured spectrum of 1070 nm RFL seed. (b) Measured time domain of 1070 nm RFL seed (inner) and its corresponding RF spectrum.
Letter amplifier (PA). In the amplifier stage, six 976 nm laser diodes (LDs) with maximal output power of 250 W are combined by a 6 1 × 1 signal/pump combiner and injected into a piece of gain fiber. The gain fiber used in the amplifier is a 3 m long 25/250 μm (diameter of core/inner cladding) Yb-doped double-clad fiber (YDF) with ∼20 dB total cladding absorption at 976 nm. After the YDF, a cladding light stripper is installed to dump the unabsorbed pump light, as well as the cladding modes. Efficient heat dissipation has been made by placing both the combiner and the gain fiber on a metal cooling plate. At the output end of the fiber amplifier, the matching passive fiber of a homemade endcap with anti-reflection film coating is spliced to the pump dump to avoid unexpected end reflection. The power scaling characteristic of the spectral-broadeningfree fiber amplifier output with the incorporated pump power is shown in Fig. 3. A maximal 1030 W output power is obtained when 1381.4 W pump power is launched into the amplifier and the optical efficiency is 74.6%. The output laser power grows almost linearly with increasing pump power, and the fitted slope efficiency is 79.7%. The sustained linearly increasing trend shows higher laser output potential, and no obvious power-limiting phenomena have been observed during the power scaling process. The obtained output power in this Letter is mainly limited by the available pump brightness and the power endurance of the combiner. Figure 4(a) depicts the output spectra of the RFL seed source and the laser output at maximal power, and Fig. 4(b) shows the evolutions of detailed spectra at different output powers. The optical spectrum analyzer (OSA) used for laser spectra measurement has a minimum resolution of 0.02 nm. The measured spectrum shows an interesting result that, even at kW level laser output, the spectral linewidth is well maintained to about the same as the seed source [around 1 nm of full width at half-maximum (FWHM)], which is totally different from other conventional MOPA systems. Usually in other high-power Yb-doped fiber amplifiers based on MOPA configuration, the spectral broadening can be observed with a linewidth several times larger than the seed source [4,5,10,35]. Figure 4(c) refers to the measured 3 (FWHM) and 10 dB spectral linewidth at different output power levels which helps discuss the spectral linewidth evolution. The linewidth values show the spectral-broadening-free property during the kW level amplification process, and we consider that the absence of spectral broadening is a result of a temporally stable seed source with suppressed fluctuations. The fact of no spectral broadening by utilizing RFL as the seed source in high-power MOPA system has proved that our conception on stabilizing spectral evolution with temporally controlled seed source is
Fig. 3. Power scaling of the output laser with the incorporated pump power in the power amplifier.
Vol. 40, No. 22 / November 15 2015 / Optics Letters
5313
Fig. 4. (a) Spectra of the seed source and the amplifier at 1.03 kW maximal output power. (b) Detailed spectra showing spectral evolution at different output powers. (c) Change of 3 dB (FWHM) and 10 dB linewidth at different output powers.
experimentally applicable. In addition, we believe that such a spectral-broadening-free characteristic under high-power amplification is significant for further application that requires narrow-linewidth high-power lasing. Usually in practical applications, the output beam quality (M 2 factor) of the MOPA structure is also a significant aspect that should be considered. After proper collimation and attenuation on the output laser, the beam quality is measured using a four-sigma method with the help of M 2 -200-FW (Spiricon Corporation). As the measured M 2 factor of the amplifier output is ∼1.2 when only the seed laser is injected into the amplifier, the 9 W seed output can be considered in fundamental mode operation, as shown in Fig. 5(a). However, the 25/250 μm (diameter of core/inner cladding) fiber used in the amplifier can support several transverse modes which may possibly reduce the beam quality during power amplification. The measured beam quality values at 40 W laser output are, respectively, M 2x ∼ 1.37 and M 2y ∼ 1.39, and shown in Fig. 5(b). By further increasing the power output, a clear sign of beam quality degeneration can be observed when the output power exceeds 700 W. In addition, the M 2 factors are correspondingly M 2x ∼ 1.9 and M 2y ∼ 2.0 when the system operates at ∼1 kW power output; see Fig. 5(c). The beam degeneration is mainly caused by the onset of mode instability (MI) which has been frequently observed in high-power fiber amplifiers based on large mode area (LMA) fibers [2,36–40]. The existence of the MI can be further convinced by detecting the far-field beam profile and the temporal characteristics of the laser output. During the real-time detection on the beam profile, the beam spot keeps fluctuating and switching between the fundamental LP01 mode and the higher-order LP11 mode. The temporal profile of the laser operating at kW power output was also measured by a 180 MHz high-speed detector and an oscilloscope with 1 GHz bandwidth. The recorded temporal profile exhibits oscillations with frequencies on the order of several kHz, which is another way of proving the onset of mode instability [39]. As mentioned in previous work [41], the existence of MI in high-power LMA fiber amplifiers does not restrict further power increase, and the output power still grows linearly with the increase of pump power. During the further linear growth of the output power with the onset of MI, the beam quality
5314
Vol. 40, No. 22 / November 15 2015 / Optics Letters
Fig. 5. Beam quality of (a) 9 W seed laser, (b) 400 W laser power output, and (c) ∼1 kW laser power output. Inset: the corresponding beam profile. The change of beam profile recorded when the output power gradually increases from 350 to 630 W (see Visualization 1).
degenerates into a relatively stabilized degree, and the M 2 value of ∼2 is asymptotically approached at the highest output power in this case, which properly corresponds with previous research [41]. We believe that, in future work, such beam quality degeneration can be avoided by applying proper methods such as fiber coiling and pump wavelength shifting to mitigate MI in LMA fiber amplifiers [39,42]. In conclusion, we demonstrate a novel high-power fiber laser based on MOPA configuration which is seeded by random fiber laser. Due to the temporal stability of the RFL seed, up to kW level power output has been achieved with spectral-broadening-free property which is quite unique for high-power MOPA systems. To the best of our knowledge, this is also the first reported high-power MOPA using RFL as the seed source. We believe our result which shows suppressed linewidth broadening in spectral domain under high-power amplification is significant for further power scaling, spectral beam combination, and other applications.
Funding. National Natural Science Foundation of China (NSFC) (61322505). REFERENCES 1. D. J. Richardson, J. Nilsson, and W. A. Clarkson, J. Opt. Soc. Am. B 27, B63 (2010). 2. C. Jauregui, J. Limpert, and A. Tünnermann, Nat. Photonics 7, 861 (2013). 3. M. N. Zervas and C. A. Codemard, IEEE J. Sel. Top. Quantum Electron. 20, 219 (2014). 4. Q. Fang, W. Shi, Y. Qin, X. Meng, and Q. Zhang, Laser Phys. Lett. 11, 105102 (2014). 5. H. Yu, H. Zhang, H. Lv, X. Wang, J. Leng, H. Xiao, S. Guo, P. Zhou, X. Xu, and J. Chen, Appl. Opt. 54, 4556 (2015).
Letter 6. G. P. Agrawal, Nonlinear Fiber Optics (Academic, 2007). 7. S. A. Babin, D. V. Churkin, A. E. Ismagulov, S. I. Kablukov, and E. V. Podivilov, Opt. Lett. 31, 3007 (2006). 8. S. I. Kablukov, E. A. Zlobina, E. V. Podivilov, and S. A. Babin, Opt. Lett. 37, 2508 (2012). 9. D. B. S. Soh, J. P. Koplow, S. W. Moore, K. L. Schroder, and W. L. Hsu, Opt. Express 18, 22393 (2010). 10. W. Liu, H. Xiao, X. Wang, and P. Zhou, Proc. SPIE 8904, 89040Q (2013). 11. A. G. Kuznetsov, E. V. Podivilov, and S. A. Babin, J. Opt. Soc. Am. B 29, 1231 (2012). 12. J. T. Manassah, Opt. Lett. 15, 329 (1990). 13. S. K. Turitsyn, A. E. Bednyakova, M. P. Fedoruk, A. I. Latkin, A. A. Fotiadi, A. S. Kurkov, and E. Sholokhov, Opt. Express 19, 8394 (2011). 14. A. E. Bednyakova, O. A. Gorbunov, M. O. Politko, S. I. Kablukov, S. V. Smirnov, D. V. Churkin, M. P. Fedoruk, and S. A. Babin, Opt. Express 21, 8177 (2013). 15. B. N. Upadhyayaa, U. Chakravartya, A. Kuruvillaa, S. M. Oaka, M. R. Shenoyb, and K. Thyagarajanb, Opt. Commun. 283, 2206 (2010). 16. C. Zeringue, C. Vergien, and I. Dajani, Opt. Lett. 36, 618 (2011). 17. Y. Jeong, J. Nilsson, J. K. Sahu, D. B. S. Soh, C. Alegria, P. Dupriez, C. A. Codemard, D. N. Payne, R. Horley, L. M. B. Hickey, L. Wanzcyk, C. E. Chryssou, J. A. Alvarez-Chavez, and P. W. Turner, Opt. Lett. 30, 459 (2005). 18. T. Theeg, H. Sayinc, J. Neumann, and D. Kracht, IEEE Photon. Technol. Lett. 24, 1864 (2012). 19. S. Turitsyn, S. Babin, A. E. El-Taher, P. Harper, D. V. Churkin, S. I. Kablukov, J. D. Ania-Castañón, V. Karalekas, and E. V. Podivilov, Nat. Photonics 4, 231 (2010). 20. W. L. Zhang, Y. J. Rao, J. M. Zhu, Z. X. Yang, Z. N. Wang, and X. H. Jia, Opt. Express 20, 14400 (2012). 21. D. V. Churkin, S. A. Babin, A. E. El-Taher, P. Harper, S. I. Kablukov, V. Karalekas, J. D. Ania-Castanon, E. V. Podivilov, and S. K. Turitsyn, Phys. Rev. A 82, 33828 (2010). 22. H. Zhang, P. Zhou, H. Xiao, and X. Xu, Laser Phys. Lett. 11, 75104 (2014). 23. C. Huang, X. Dong, S. Zhang, N. Zhang, and P. P. Shum, IEEE Photon. Technol. Lett. 26, 1287 (2014). 24. S. Sugavanam, Z. Yan, V. Kamynin, A. S. Kurkov, L. Zhang, and D. V. Churkin, Opt. Express 22, 2839 (2014). 25. J. Nuno, M. Alcon-Camas, and J. D. Ania-Castanon, Opt. Express 20, 27376 (2012). 26. M. Peng, X. Bao, and L. Chen, Opt. Lett. 38, 1866 (2013). 27. X. Du, H. Zhang, X. Wang, X. Wang, P. Zhou, and Z. Liu, Laser Phys. Lett. 12, 45106 (2015). 28. S. K. Turitsyn, S. A. Babin, D. V. Churkin, I. D. Vatnik, M. Nikulin, and E. V. Podivilov, Phys. Rep. 542, 133 (2014). 29. A. S. Kurkov, Laser Phys. Lett. 4, 93 (2007). 30. H. Xiao, P. Zhou, X. Wang, S. Guo, and X. Xu, IEEE Photon. Technol. Lett. 24, 1088 (2012). 31. H. Xiao, P. Zhou, X. L. Wang, S. F. Guo, and X. J. Xu, Laser Phys. Lett. 9, 748 (2012). 32. H. Zhang, P. Zhou, X. Wang, X. Du, H. Xiao, and X. Xu, Opt. Express 23, 17138 (2015). 33. Z. Wang, H. Wu, M. Fan, L. Zhang, Y. Rao, W. Zhang, and X. Jia, IEEE J. Sel. Top. Quantum Electron. 21, 900506 (2015). 34. I. D. Vatnik, D. V. Churkin, E. V. Podivilov, and S. A. Babin, Laser Phys. Lett. 11, 75101 (2014). 35. L. Huang, P. Ma, R. Tao, C. Shi, X. Wang, and P. Zhou, Appl. Opt. 54, 2880 (2015). 36. B. Samson, A. Carter, and K. Tankala, Nat. Photonics 5, 466 (2011). 37. L. Dong, Opt. Express 21, 2642 (2013). 38. C. Jauregui, H. Otto, F. Stutzki, F. Jansen, J. Limpert, and A. Tünnermann, Opt. Express 21, 19375 (2013). 39. R. Tao, P. Ma, X. Wang, P. Zhou, and Z. Liu, J. Opt. 17, 45504 (2015). 40. A. V. Smith and J. J. Smith, Opt. Express 19, 10180 (2011). 41. H. Otto, F. Stutzki, N. Modsching, C. Jauregui, J. Limpert, and A. Tünnermann, Opt. Lett. 39, 6446 (2014). 42. R. Tao, P. Ma, X. Wang, P. Zhou, and Z. Liu, Photonics Res. 3, 86 (2015).
Vol. 40, No. 22 / November 15 2015 / Optics Letters
5311
Kilowatt-level fiber amplifier with spectralbroadening-free property, seeded by a random fiber laser XUEYUAN DU, HANWEI ZHANG, PENGFEI MA, HU XIAO, XIAOLIN WANG, PU ZHOU,*
AND
ZEJIN LIU
College of Optoelectronic Science and Engineering, National University of Defense Technology, Changsha 410073, China *Corresponding author: [email protected] Received 7 July 2015; revised 30 August 2015; accepted 18 September 2015; posted 25 September 2015 (Doc. ID 245442); published 9 November 2015
In this Letter, we demonstrate a kilowatt (kW) level highpower fiber laser amplifier with a clear sign of spectralbroadening-free property. The high-power fiber lasing is realized by employing a master oscillator power-amplifier (MOPA) configuration, seeded by a temporally stable random fiber laser (RFL) that utilizes Raman amplification and random distributed feedback from a long passive fiber. The output power reaches 1.03 kW with a 1070 nm wavelength and an optical-to-optical efficiency of 74.6%. Despite the typical nonlinear spectral broadening in most traditional MOPA systems, the output spectral linewidth is well maintained during the whole high-power amplification process. The suppressed linewidth broadening in the spectral domain during high-power amplification is significant for further power scaling, spectral beam combination, and other applications that require narrow-linewidth highpower lasing. © 2015 Optical Society of America OCIS codes: (060.2320) Fiber optics amplifiers and oscillators; (140.3490) Lasers, distributed-feedback; (290.5870) Scattering, Rayleigh; (290.5910) Scattering, stimulated Raman. http://dx.doi.org/10.1364/OL.40.005311
Fiber lasers and amplifiers have shown great industrial application potential and contain prominent properties such as high conversion efficiency, outstanding beam quality, stable compact structure, and convenient heat management [1–3]. Thanks to the introduction of a Ytterbium-doped fiber amplifier (YDFA) based on master oscillator power–amplifier (MOPA) structural design, the power scalability of continuous wave (CW) fiber amplifier has raised significantly to multi-kW level [4,5]. Considering the high-power intensity and relatively long fiber length, the nonlinearity effect has clear influence on highpower fiber laser performance, one of which is spectral broadening [3,6]. Previous theoretical analysis on the spectral broadening has revealed the correlation between spectral broadening and nonlinearity, considering the effects of four-wave mixing (FWM) [7], self-phase modulation (SPM) [8], or groupvelocity dispersion (GVD) [9], and some theoretical analyses 0146-9592/15/225311-04$15/0$15.00 © 2015 Optical Society of America
on the spectral evolution of CW YDFA, considering the nonlinear effect has also been carried out [10]. In practical conditions where the length of the active fiber in a high-power YDFA is relatively short (several meters), the phase shift caused by dispersion can be neglected. The major phase shift leading to the spectral broadening is induced by the nonlinear effects. The previously reported models derived from the generalized nonlinear Schrödinger (NLS) equation have considered the self-phase modulation (SPM) effect as the main reason for nonlinear spectral broadening involving some weak FWM interactions [11,12]. However, conventional MOPA YDFAs usually use a typical Yb-doped fiber oscillator as the seed source, which suffers self-pulsation effect with temporal instability and intensity fluctuations [13–15]. The intensity fluctuations in the oscillator-based lasers cause higher peak powers compared with temporally stable seed sources, which results in strong nonlinear effect under high pump powers. Due to the close connection between the nonlinear effect and the laser’s temporal instability, the spectral broadening is expected to be controlled by eliminating the temporal fluctuations. Single-frequency MOPA systems based on the single-frequency seed source also show good temporal stability and the preservation of the spectral properties [16–18]. However, further power scalability of the single-frequency MOPAs is mainly restricted by the stimulated Brillouin scattering (SBS) effect, and power output exceeding KW level has not yet been reported. Thus, we intend to build a temporally stable seed source to suppress the nonlinear spectral broadening during high-power amplification and to optimize the output property of MOPA system. Random fiber laser (RFL) has drawn great attention in recent years [19], and researches, including power scaling, temporal stability, spectral evolution, intensity-noise transfer, and application potentials have been widely investigated [20–27]. As RFL utilizes random distributed feedback (RDFB) via Rayleigh scattering in long passive fibers, one remarkable feature is that the increase of pump power well above threshold leads to stabilized radiation both in the temporal and spectral domains [19]. In addition, during the relative intensity-noise (RIN) transfer in RFL configuration, pump noise can be minimized with the increase of pump power or using a longer
5312
Vol. 40, No. 22 / November 15 2015 / Optics Letters
passive fiber [25,28]. Therefore, a temporally stable seed laser source based on RFL configuration can be a practical solution to overcome the temporal instability of conventional seed source. In this Letter, a temporally stable 1070 nm random fiber laser pumped by 1018 nm Ytterbium-doped fiber laser (YFL) with steady 10 W RFL power output is for the first time to the best of our knowledge prepared as the seed source, followed by kW level high-power amplification based on MOPA configuration. The specified 1.06–1.09 μm wavelength range for a MOPA system ensures efficient power amplification according to the gain property of Yb-ion. Thus, we choose a 1070 nm wavelength RFL to be the seed source applicable for power amplification, and the pump laser should be correspondingly near 1018 m wavelength referring to the Raman shift value. However, most of the previously reported RFLs failed to cover this wavelength range due to the small net gain of Yb-doped fiber near 1018 nm wavelength [29], and the devastating self-oscillation or spurious oscillation caused by the ASE effect [30,31]. First, the 1018 nm YFL which served as the pump laser of the 1070 nm RFL has to be optimized under tremendous work considering influential factors such as the effect of ASE feedback, the fiber type and fiber length of Yb-doped fiber, reflectivity of the fiber Bragg gratings (FBGs), and the pumping direction. [30]. We build an YFL at 1018 nm with stable highpower output under optimized structure design. Based on that, the 1070 nm random fiber laser pumped by 1018 nm YFL can be achieved by utilizing stimulated Raman scattering (SRS) and RDFB from passive fiber. The experimental setup of the 1070 nm RFL seed laser illustrates a forward-pumped half-opened cavity structure in Fig. 1. The pump laser is a 1018 nm YFL, and a 1018 nm bandpass filter (BPF) is included in the structure as a precaution against backward feedback around ASE wavelength (1030 nm), which often results in spurious oscillation. By using a 1018/1070 nm wavelength division multiplexer (WDM), the pump laser is injected into the cavity via the WDM. The common port of the WDM is spliced into a 0.5 km HI1060 optical fiber, which provides both Raman gain and random distributed feedback via Rayleigh scattering. Intracavity light is reflected back by a 1070 nm high-reflectivity fiber Bragg grating (HR-FBG) at the 1070 nm port of the WDM, which helps form a half-opened RFL structure [20]. At the output end of the RFL seed laser,
Fig. 1. Experimental setup of the RFL seed and the amplifier. BPF, bandpass filter; WDM, wavelength division multiplexer; P 4 , extra port of WDM; RDFB, random distributed feedback; HR-FBG, highreflectivity fiber Bragg grating; LD, 976 nm laser diode; PM, power meter; PD, photodetector; OSC, oscilloscope; OSA, optical spectrum analyzer.
Letter the 1070/1120 nm WDM functions as a spectral filter to eliminate the second-order Raman Stokes wave components. All the free ends in the RFL structure are angle cleaved to avoid any Fresnel reflection. Some essential factors, including power efficiency, spectral purity, and temporal stability, are considered during the preparation of the 1070 nm RFL seed. For traditional random fiber lasers, the passive fiber used to provide feedback and gain is usually several or several tens of kilometers, which causes relatively low higher-order Raman Stokes threshold and limits the power scalability of RFL [32]. In this case, the short-cavity structure using a 0.5 km passive fiber instead of a longer one enables sufficient power output of first-order Stokes wave at 1070 nm, but also obviously increases the random lasing threshold [22,33,34]. However, the half-opened cavity structural design by introducing a HR-FBG from one end can decrease the RFL lasing threshold to approximately half, which helps overcome the obstacle of high lasing threshold in shortcavity RFL [20]. In addition, another advantage of utilizing short-cavity structural design is that, by decreasing the passive fiber length, the optical conversion efficiency becomes higher and even approaches the quantum limit due to the reduction of the pump and Stokes wave attenuation [32,33]. With the above structural optimization, the achieved 1070 nm RFL seed can finally reach 10 W level power output before the notable emergence of second-order Raman Stokes spectral constituents. The spectral feature of the RFL seed is measured through an optical spectrum analyzer and shown in Fig. 2(a). By numerical integrating, the 1070 nm component is calculated to contain more than 99.9% of the total power, which guarantees the spectral purity of the seed laser. Figure 2(b) illustrates the temporal characteristic monitored by a 180 M Hz high-speed photodetector and an oscilloscope with 1 GHz bandwidth. According to the measured temporal behavior and, correspondingly, its radiofrequency (RF) spectrum, there is no longitudinal mode beating signal corresponding to the cavity length, and the RFL operates in temporally stable CW regime without pulsation. In a conventional fiber laser based on fiber oscillator structure, there exists oscillating peaks corresponding to the cavity length which are easily observed in the frequency domains. The temporally modulated radiation with peak power of several or several tens of times higher than the measured average power causes strong nonlinear effects. To suppress the spectral broadening, the RFL seed source should be well prepared ensuring that the frequency spectrum is stable and smooth, as illustrated in Fig. 2(b). The fiber amplifier system depicted in Fig. 1 is a typical MOPA regime of two stages: a master oscillator (MO) formed by a random fiber laser working as the seed source and a power
Fig. 2. (a) Measured spectrum of 1070 nm RFL seed. (b) Measured time domain of 1070 nm RFL seed (inner) and its corresponding RF spectrum.
Letter amplifier (PA). In the amplifier stage, six 976 nm laser diodes (LDs) with maximal output power of 250 W are combined by a 6 1 × 1 signal/pump combiner and injected into a piece of gain fiber. The gain fiber used in the amplifier is a 3 m long 25/250 μm (diameter of core/inner cladding) Yb-doped double-clad fiber (YDF) with ∼20 dB total cladding absorption at 976 nm. After the YDF, a cladding light stripper is installed to dump the unabsorbed pump light, as well as the cladding modes. Efficient heat dissipation has been made by placing both the combiner and the gain fiber on a metal cooling plate. At the output end of the fiber amplifier, the matching passive fiber of a homemade endcap with anti-reflection film coating is spliced to the pump dump to avoid unexpected end reflection. The power scaling characteristic of the spectral-broadeningfree fiber amplifier output with the incorporated pump power is shown in Fig. 3. A maximal 1030 W output power is obtained when 1381.4 W pump power is launched into the amplifier and the optical efficiency is 74.6%. The output laser power grows almost linearly with increasing pump power, and the fitted slope efficiency is 79.7%. The sustained linearly increasing trend shows higher laser output potential, and no obvious power-limiting phenomena have been observed during the power scaling process. The obtained output power in this Letter is mainly limited by the available pump brightness and the power endurance of the combiner. Figure 4(a) depicts the output spectra of the RFL seed source and the laser output at maximal power, and Fig. 4(b) shows the evolutions of detailed spectra at different output powers. The optical spectrum analyzer (OSA) used for laser spectra measurement has a minimum resolution of 0.02 nm. The measured spectrum shows an interesting result that, even at kW level laser output, the spectral linewidth is well maintained to about the same as the seed source [around 1 nm of full width at half-maximum (FWHM)], which is totally different from other conventional MOPA systems. Usually in other high-power Yb-doped fiber amplifiers based on MOPA configuration, the spectral broadening can be observed with a linewidth several times larger than the seed source [4,5,10,35]. Figure 4(c) refers to the measured 3 (FWHM) and 10 dB spectral linewidth at different output power levels which helps discuss the spectral linewidth evolution. The linewidth values show the spectral-broadening-free property during the kW level amplification process, and we consider that the absence of spectral broadening is a result of a temporally stable seed source with suppressed fluctuations. The fact of no spectral broadening by utilizing RFL as the seed source in high-power MOPA system has proved that our conception on stabilizing spectral evolution with temporally controlled seed source is
Fig. 3. Power scaling of the output laser with the incorporated pump power in the power amplifier.
Vol. 40, No. 22 / November 15 2015 / Optics Letters
5313
Fig. 4. (a) Spectra of the seed source and the amplifier at 1.03 kW maximal output power. (b) Detailed spectra showing spectral evolution at different output powers. (c) Change of 3 dB (FWHM) and 10 dB linewidth at different output powers.
experimentally applicable. In addition, we believe that such a spectral-broadening-free characteristic under high-power amplification is significant for further application that requires narrow-linewidth high-power lasing. Usually in practical applications, the output beam quality (M 2 factor) of the MOPA structure is also a significant aspect that should be considered. After proper collimation and attenuation on the output laser, the beam quality is measured using a four-sigma method with the help of M 2 -200-FW (Spiricon Corporation). As the measured M 2 factor of the amplifier output is ∼1.2 when only the seed laser is injected into the amplifier, the 9 W seed output can be considered in fundamental mode operation, as shown in Fig. 5(a). However, the 25/250 μm (diameter of core/inner cladding) fiber used in the amplifier can support several transverse modes which may possibly reduce the beam quality during power amplification. The measured beam quality values at 40 W laser output are, respectively, M 2x ∼ 1.37 and M 2y ∼ 1.39, and shown in Fig. 5(b). By further increasing the power output, a clear sign of beam quality degeneration can be observed when the output power exceeds 700 W. In addition, the M 2 factors are correspondingly M 2x ∼ 1.9 and M 2y ∼ 2.0 when the system operates at ∼1 kW power output; see Fig. 5(c). The beam degeneration is mainly caused by the onset of mode instability (MI) which has been frequently observed in high-power fiber amplifiers based on large mode area (LMA) fibers [2,36–40]. The existence of the MI can be further convinced by detecting the far-field beam profile and the temporal characteristics of the laser output. During the real-time detection on the beam profile, the beam spot keeps fluctuating and switching between the fundamental LP01 mode and the higher-order LP11 mode. The temporal profile of the laser operating at kW power output was also measured by a 180 MHz high-speed detector and an oscilloscope with 1 GHz bandwidth. The recorded temporal profile exhibits oscillations with frequencies on the order of several kHz, which is another way of proving the onset of mode instability [39]. As mentioned in previous work [41], the existence of MI in high-power LMA fiber amplifiers does not restrict further power increase, and the output power still grows linearly with the increase of pump power. During the further linear growth of the output power with the onset of MI, the beam quality
5314
Vol. 40, No. 22 / November 15 2015 / Optics Letters
Fig. 5. Beam quality of (a) 9 W seed laser, (b) 400 W laser power output, and (c) ∼1 kW laser power output. Inset: the corresponding beam profile. The change of beam profile recorded when the output power gradually increases from 350 to 630 W (see Visualization 1).
degenerates into a relatively stabilized degree, and the M 2 value of ∼2 is asymptotically approached at the highest output power in this case, which properly corresponds with previous research [41]. We believe that, in future work, such beam quality degeneration can be avoided by applying proper methods such as fiber coiling and pump wavelength shifting to mitigate MI in LMA fiber amplifiers [39,42]. In conclusion, we demonstrate a novel high-power fiber laser based on MOPA configuration which is seeded by random fiber laser. Due to the temporal stability of the RFL seed, up to kW level power output has been achieved with spectral-broadening-free property which is quite unique for high-power MOPA systems. To the best of our knowledge, this is also the first reported high-power MOPA using RFL as the seed source. We believe our result which shows suppressed linewidth broadening in spectral domain under high-power amplification is significant for further power scaling, spectral beam combination, and other applications.
Funding. National Natural Science Foundation of China (NSFC) (61322505). REFERENCES 1. D. J. Richardson, J. Nilsson, and W. A. Clarkson, J. Opt. Soc. Am. B 27, B63 (2010). 2. C. Jauregui, J. Limpert, and A. Tünnermann, Nat. Photonics 7, 861 (2013). 3. M. N. Zervas and C. A. Codemard, IEEE J. Sel. Top. Quantum Electron. 20, 219 (2014). 4. Q. Fang, W. Shi, Y. Qin, X. Meng, and Q. Zhang, Laser Phys. Lett. 11, 105102 (2014). 5. H. Yu, H. Zhang, H. Lv, X. Wang, J. Leng, H. Xiao, S. Guo, P. Zhou, X. Xu, and J. Chen, Appl. Opt. 54, 4556 (2015).
Letter 6. G. P. Agrawal, Nonlinear Fiber Optics (Academic, 2007). 7. S. A. Babin, D. V. Churkin, A. E. Ismagulov, S. I. Kablukov, and E. V. Podivilov, Opt. Lett. 31, 3007 (2006). 8. S. I. Kablukov, E. A. Zlobina, E. V. Podivilov, and S. A. Babin, Opt. Lett. 37, 2508 (2012). 9. D. B. S. Soh, J. P. Koplow, S. W. Moore, K. L. Schroder, and W. L. Hsu, Opt. Express 18, 22393 (2010). 10. W. Liu, H. Xiao, X. Wang, and P. Zhou, Proc. SPIE 8904, 89040Q (2013). 11. A. G. Kuznetsov, E. V. Podivilov, and S. A. Babin, J. Opt. Soc. Am. B 29, 1231 (2012). 12. J. T. Manassah, Opt. Lett. 15, 329 (1990). 13. S. K. Turitsyn, A. E. Bednyakova, M. P. Fedoruk, A. I. Latkin, A. A. Fotiadi, A. S. Kurkov, and E. Sholokhov, Opt. Express 19, 8394 (2011). 14. A. E. Bednyakova, O. A. Gorbunov, M. O. Politko, S. I. Kablukov, S. V. Smirnov, D. V. Churkin, M. P. Fedoruk, and S. A. Babin, Opt. Express 21, 8177 (2013). 15. B. N. Upadhyayaa, U. Chakravartya, A. Kuruvillaa, S. M. Oaka, M. R. Shenoyb, and K. Thyagarajanb, Opt. Commun. 283, 2206 (2010). 16. C. Zeringue, C. Vergien, and I. Dajani, Opt. Lett. 36, 618 (2011). 17. Y. Jeong, J. Nilsson, J. K. Sahu, D. B. S. Soh, C. Alegria, P. Dupriez, C. A. Codemard, D. N. Payne, R. Horley, L. M. B. Hickey, L. Wanzcyk, C. E. Chryssou, J. A. Alvarez-Chavez, and P. W. Turner, Opt. Lett. 30, 459 (2005). 18. T. Theeg, H. Sayinc, J. Neumann, and D. Kracht, IEEE Photon. Technol. Lett. 24, 1864 (2012). 19. S. Turitsyn, S. Babin, A. E. El-Taher, P. Harper, D. V. Churkin, S. I. Kablukov, J. D. Ania-Castañón, V. Karalekas, and E. V. Podivilov, Nat. Photonics 4, 231 (2010). 20. W. L. Zhang, Y. J. Rao, J. M. Zhu, Z. X. Yang, Z. N. Wang, and X. H. Jia, Opt. Express 20, 14400 (2012). 21. D. V. Churkin, S. A. Babin, A. E. El-Taher, P. Harper, S. I. Kablukov, V. Karalekas, J. D. Ania-Castanon, E. V. Podivilov, and S. K. Turitsyn, Phys. Rev. A 82, 33828 (2010). 22. H. Zhang, P. Zhou, H. Xiao, and X. Xu, Laser Phys. Lett. 11, 75104 (2014). 23. C. Huang, X. Dong, S. Zhang, N. Zhang, and P. P. Shum, IEEE Photon. Technol. Lett. 26, 1287 (2014). 24. S. Sugavanam, Z. Yan, V. Kamynin, A. S. Kurkov, L. Zhang, and D. V. Churkin, Opt. Express 22, 2839 (2014). 25. J. Nuno, M. Alcon-Camas, and J. D. Ania-Castanon, Opt. Express 20, 27376 (2012). 26. M. Peng, X. Bao, and L. Chen, Opt. Lett. 38, 1866 (2013). 27. X. Du, H. Zhang, X. Wang, X. Wang, P. Zhou, and Z. Liu, Laser Phys. Lett. 12, 45106 (2015). 28. S. K. Turitsyn, S. A. Babin, D. V. Churkin, I. D. Vatnik, M. Nikulin, and E. V. Podivilov, Phys. Rep. 542, 133 (2014). 29. A. S. Kurkov, Laser Phys. Lett. 4, 93 (2007). 30. H. Xiao, P. Zhou, X. Wang, S. Guo, and X. Xu, IEEE Photon. Technol. Lett. 24, 1088 (2012). 31. H. Xiao, P. Zhou, X. L. Wang, S. F. Guo, and X. J. Xu, Laser Phys. Lett. 9, 748 (2012). 32. H. Zhang, P. Zhou, X. Wang, X. Du, H. Xiao, and X. Xu, Opt. Express 23, 17138 (2015). 33. Z. Wang, H. Wu, M. Fan, L. Zhang, Y. Rao, W. Zhang, and X. Jia, IEEE J. Sel. Top. Quantum Electron. 21, 900506 (2015). 34. I. D. Vatnik, D. V. Churkin, E. V. Podivilov, and S. A. Babin, Laser Phys. Lett. 11, 75101 (2014). 35. L. Huang, P. Ma, R. Tao, C. Shi, X. Wang, and P. Zhou, Appl. Opt. 54, 2880 (2015). 36. B. Samson, A. Carter, and K. Tankala, Nat. Photonics 5, 466 (2011). 37. L. Dong, Opt. Express 21, 2642 (2013). 38. C. Jauregui, H. Otto, F. Stutzki, F. Jansen, J. Limpert, and A. Tünnermann, Opt. Express 21, 19375 (2013). 39. R. Tao, P. Ma, X. Wang, P. Zhou, and Z. Liu, J. Opt. 17, 45504 (2015). 40. A. V. Smith and J. J. Smith, Opt. Express 19, 10180 (2011). 41. H. Otto, F. Stutzki, N. Modsching, C. Jauregui, J. Limpert, and A. Tünnermann, Opt. Lett. 39, 6446 (2014). 42. R. Tao, P. Ma, X. Wang, P. Zhou, and Z. Liu, Photonics Res. 3, 86 (2015).
