The high-power all-fiber superfluorescent source operating near 980 nm is studied experimentally and numerically. In experiment, an all-fiber superfluorescent source operating near 980 nm is fabricated with the distributed side-coupled cladding-pumping (DSCCP) Yb-doped fiber (YDF). By optimizing the active fiber and angle-cleaving of the output port, a recorded 17.1-W output power and 14.6% slope efficiency of 980-nm ASE are obtained. No parasitic laser oscillation is observed at the maximum output power. The power scalability of the source is also numerically investigated. A simple but effective method is present to numerically determine the threshold of parasitic laser oscillation. It is found that the output power can be scaled up to 50 W and 100 W with the optical feedback of each output port suppressed to 1.2 × 10−6 and 7 × 10−7, respectively. It is also revealed that coupling coefficient should be larger than 6 to realize more than 50% slope efficiency. These results provide significant guidance for understanding and designing the high-power superfluorescent Yb-doped source (SYFS) operating near 980 nm and other sorts of three-level fiber sources.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
The study on the Yb-doped fiber source operating near 980 nm (or 980-nm Yb-doped fiber source) has been started in 1980s [1, 2]. Nowadays, it becomes more and more attractive because it has many applications as the pump source of Yb- or Er-doped fiber lasers and can also be an alternative of blue source by frequency conversion [3–10]. In essence, the 980-nm emission of Yb-doped fiber involves the three-level transition of Yb-ion, which brings two challenges. The first one is more-than-50% population inversion needed for 980-nm lasing or amplification, which induces the high pump threshold [11, 12]. The second one is its competition with the 1030-nm amplified spontaneous emission (ASE) involving the quasi-four-level transition of Yb-ion that needs only less-than-5% population inversion [11, 12]. Then, how to suppress the 1030-nm ASE is of great importance for the 980-nm Yb-doped fiber source.
One way for suppressing the 1030-nm ASE is introducing the wavelength-sensitive component such as the cavity mirrors, fiber Bragg gratings (FBG) or seed light, and thus, the 980-nm fiber laser was focused by a number of studies. The early 980-nm fiber laser is mainly fabricated with the single-cladding single-mode fiber and core-pumping scheme, and its output power is limited to an order of mW because of the low pump power coupled into the active core [1, 2]. Later, the double-cladding fiber (DCF) is introduced into the 980-nm fiber laser in order to increase the pump power, and the output power of 980-nm fiber laser is greatly scaled up [13–20]. In 2008, 94-W output power was obtained at 977 nm with the DC rod-type photonic crystal fiber (PCF), although the spatial-light coupling configuration was used in experiment [15, 16]. Later, the all-fiber 980-nm laser was also demonstrated with the DC Yb-doped fiber and multi-Watt output power was obtained [17–20].
Besides the 980-nm fiber laser, the 980-nm superfluorescent Yb-doped fiber source (SYFS) is another 980-nm Yb-doped fiber source which is realized by the ASE and owns characteristics such as low coherence and high temporal stability . However, the current studies on the 980-nm SYFS are very limited. Frankly, compared with the 980-nm fiber laser, the SYFS suffers two more challenges. The first one is the absence of the wavelength-sensitive components, which makes the suppression of 1030-nm ASE more difficult. The second one is the parasitic laser oscillation, which is mainly induced by the optical feedback of output port and limits the power scaling of SYFS. Besides, the high-power ASE produced in the active core may be destructive to the fiber components spliced to the active core. Therefore, it is not so easy to realize the high-power all-fiber SYFS operating near 980 nm. Currently, the maximum output power realized by an all-fiber 980-nm SYFS was about 8 W which was limited by the parasitic laser oscillation . Then, how about the power scalability of the SYFS is still not so clear and needs to be investigated.
In this paper, a more-than 10-W all-fiber 980-nm SYFS is demonstrated in experiment with the DSCCP Yb-doped fiber. Compared with the conventional DCF, the DSCCP fiber is more beneficial to the SYFS in protecting the pump sources and minimizing the optical feedback induced by the splicing poins or core distortion [22-23]. By optimizing the active fiber length and angle-cleaving of the output ports, a recorded 17.1-W output power is obtained with the slope efficiency about 14.6%. The 3-dB bandwidth is about 3.5 nm with the central wavelength about 977.3 nm. Then, by presenting the numerical criterion of the threshold of parasitic laser oscillation, the scalability of the 980-nm SYFS is also numerically studied with the rate-equation model. The effects of the optical feedback of output port and the coupling of pump light on the SYFS are discussed.
2. Experimental setup and results
The experimental setup used here (see Fig. 1) is similar to the one used in . The distributed side-coupled cladding-pumping (DSCCP) YDF is used as the active fiber, and its cross section is given as the inset of Fig. 1. The DSCCP YDF consists of a signal fiber and a pump fiber. Core 1 gives the core of the pump fiber, and Core 3 and Cladding 2 are the Yb-doped core and inner-cladding of the signal fiber. Cladding 4 acts both the cladding of pump fiber and the outer-cladding of signal fiber. Core 1 and Cladding 2 are optically contacted but physically separated from each other. The pump light is injected into Core 1, and coupled gradually into Cladding 2 by means of the evanescent wave coupling and activate the Yb-ion doped in Core 3 where the ASE will be induced. The fiber is also known as the GTWave fiber, multi clad fiber or multi-element first cladding fiber. Here, it is named as the DSCCP fiber because of its unique way of pump light propagation.
The DSCCP fiber used here is also the same as that in , and its parameters are given in Table 1. Compared with , there are mainly two improvement induced in our experiment. The first one is lengthening the DSCCP fiber to 0.49 m in order to improve the pump absorption and slope efficiency. The second one is weakening the parasitic laser oscillation (i.e., the limitation of power scaling of SFYS in ) by increasing the cleaved angle of output port from 8° to about 12°. The single-cladding 60-μm-core-diameter passive fiber is connected to each port of signal fiber to filter out the residual pump light in the signal fiber. A dichroic mirror is used to separate the 980-nm and 1030-nm ASE and its reflectivity is larger than 98% around 980 nm and lower than 2% around 1030 nm. The bi-directional pumping scheme is also used in our experiment.
The output powers are shown in Fig. 2. Figure 2(a) gives the output powers of 980-nm ASE, which illuminates that the output power and pertinent slope efficiency are obviously improved compared with the results given in  (i.e., the 8.38-W combined output power and 11.7% slope efficiency). It can be seen that the slope efficiency is increased to about 14. 59%, and with the 204.6-W pump power, the backward and forward output powers are 8.43W and 8.67W, respectively, corresponding to a combined output power of 17.1W. Furthermore, Fig. 3 also shows that no parasitic laser oscillation is present even in the spectrum at the maximum output power, which means that the threshold of parasitic laser oscillation be obviously increased with the 12° angle-cleaving of output ports. Then, the output power of the SYFS is not limited by the parasitic laser oscillation any more, and can be further scaled up by increasing the pump power.
However, Fig. 2(a) also shows that the increment of output power of 980-nm ASE is not so linear and rolls over to some extent with the increment of pump power. Besides, the 1030-nm ASE power is also decreased correspondingly (see Fig. 2(b)). We also find that the output power reduction corresponding to the rolling over cannot be recovered by restarting the SFS even with the low pump power, which means the permanent loss is induced in the Yb-doped fiber. Therefore, we deduce that such rolling over be mainly induced by the photo-darkening in the active fiber. Actually, the similar results were also reported in  which showed that the output power in 980-nm can decrease sharply in minute level because of photo-darkening. Considering that several minutes are need to complete one measurement at each pump power (including the output powers and spectra), the photo-darkening will become more and more serious with the measurement time increment. As a result, the rolling over is present in Fig. 2a. Our results illuminate again that the photo-darkening should also be considered in the 980-nm SYFS. One way to weakening the photo-darkening is co-doping Ce in the active core . From Fig. 2(a), it can also be found that the pump threshold of the 980-nm SYFS is about 65 W.
Figure 2(b) gives the output powers of the 1030-nm ASE. It can be found that the output power of the 1030-nm ASE reaches its maximum value (about 7 W) at the 83.33-W pump power, and then, decreases monotonously with further increment of pump power, which means that the increment of pump power is beneficial to the suppression of 1030-nm ASE . Compared with the pertinent results in , it can be found that the 1030-nm ASE is also increased to some extent, which is the cost of lengthening the active fiber . In spite of that, the total power of 1030-nm ASE is only about 3.5 W at the maximum output power, and correspondingly, more-than 13.7-dB peak suppression may be realized [see Fig. 3(a)], which can be further improved by increasing the pump power . Figure 3(b) gives the zoom-in of 980-nm ASE spectra. It can be seen that the 980-nm ASE spans from 974 to 982 nm with central wavelength about 977.3 nm, and its 3-dB bandwidth is about 3.5 nm.
3. The numerical study on the power scalability of 980-nm SYFS
Although the threshold of parasitic laser oscillation has been elevated in our experiment, its value that gives the power limitation of SYFS was not measured because of the limited pump power. Then, the numerical study will be given in this section to answer this question. Here, the threshold is defined as the output power of 980-nm ASE corresponding to the presence of parasitic laser oscillation, because such a definition can give a clear illumination about the power limitation of SYFS. The numeric model used here is the widely-used steady rate-equation model which can be given as follows [25–28]:26]:
The model can be solved with the boundary conditions given as:
Firstly, the model is used to study the experimental SYFS in Fig. 1. With the parameter values given in Table 2, the output power and pertinent spectrum are calculated and given in Fig. 4. It can be found that the numerical results agree well with the experimental results except the slope efficiency and the suppression of parasitic 1030-nm ASE. The numerical slope efficiency is a little higher than the experimental one because the photo-darkening is not taken into account in the numerical calculations. By comparing Fig. 4(b) with Fig. 3(b), it can be found that although the numerical results of 1030-nm ASE suppression is similar to the experimental one around the 60-W pump power, the numerical one increases to around 20 dB while the experimental one only increases to around 10 dB with the pump power increasing to 100 W. It means that the numerical prediction of 1030-nm ASE is weaker to some extent than the experimental one. The reason for the better 1030-nm ASE suppression may be that the absorption around 980 nm is a little lower in the numerical calculation.
Next, we would like to study the threshold of parasitic laser oscillation which gives the limitation of power scaling of SYFS. However, we find it is not so easy to numerically determine the threshold of parasitic laser oscillation of 980-nm SYFS from the spectral variation. The reason is that the peak of 980-nm ASE is so narrow that there is no sharp transition to mark the present of parasitic laser oscillation [see Fig. 4(c)].
In order to determine numerically the threshold, we define a parameter GR as:Fig. 5(b) also illuminates that GR is smaller than 1 at some wavelengths around the wavelength of parasitic laser oscillation (i.e., two digs in the inset of Fig. 5(b)) because no enough gain can be obtained at these wavelengths in the gain competition. Here, the total gain of ASE (i.e., the sum of gains at all the wavelengths) should be utilized as the convergence parameter in the numerical calculation. Besides, because the parasitic laser oscillation is not a steady state, the spectrum of parasitic laser oscillation obtained by such a calculation is not real. Thus, the GR method can just give an effective method to predict whether the parasitic laser oscillation is present or not, but cannot be used to study the dynamics behavior of parasitic laser oscillation in the SFS.
Figure 5 gives the output spectra and pertinent parameter GR corresponding to the various values of R. It can be found that the peak of 980-nm ASE become narrow when the R become large, but it is difficult to judge whether the parasitic laser oscillation happened. It can also be seen that there is a peak with a value larger than 1 in the spectrum when the R is large enough, which means the total gain at the peak wavelength is larger than the loss and the parasitic laser oscillation will be induced. Besides, there are always some digs around the peak which should be induced by the gain competition. Such a peak is absent when R is small enough that the parasitic will not appear [see Fig. 5(b)]. It means that the peak value of GR can be utilized to determine whether the parasitic laser oscillation is present or not.
The variation of the peak value of GR with the R is calculated and given in Fig. 6. It can be seen that when R is larger than 2.1 × 10−6, the peak value calculated by the method in this paper begins to increase sharply with R, which implies the presence of parasitic laser oscillation. With the 2.1 × 10−6 optical feedback, the output power of 980-nm ASE is about 25 W with the 0.5-m active fiber, we can conclude that, the threshold of parasitic laser oscillation should be 25 W. Besides, we also find that when the R is too large, the parasitic laser oscillation may be present around 1030 nm rather than around 980 nm [see Fig. 5(a)]. It is reasonable because the optical feedback is not wavelength-sensitive, and then, when the feedback of 1030-nm ASE is large enough, it will become more dominant in the gain competition with the 980-nm ASE  and induce the parasitic laser oscillation preferentially. It is found that the parasitic laser oscillation will be present around 1030 nm when R is larger than 7 × 10−6 with the 0.5-m active fiber length. It can also be found that the value of R is increased to 3 × 10−5 when the 1030-nm parasitic laser oscillation is present with the 0.4-m active fiber length. It means that one way to suppress the 1030-nm parasitic laser oscillation is shortening the active fiber length which is beneficial to the 980-nm gain .
Then, with this method, we can get the relationship between the threshold and R, and pertinent results are given in Fig. 7. It can be found that the R should be suppressed to lower-than 1.2 × 10−6 to obtain the 50-W 980-nm ASE and the value should be decreased to 7 × 10−7 to obtain the 100-W 980-nm ASE. In addition, it should be note that a recorded 17.1-W 980nm ASE output power are obtained in experiment and the 17W threshold of 980-nm parasitic laser oscillation correspond to the reflectivity of 3 × 10−6. As the parasitic laser oscillation is not present in experiment, the estimated reflectivity of the experiment should be lower than 3 × 10−6. It can also be found that the threshold of 980-nm parasitic laser oscillation does not change too much with the same R, which means that L only has a negligible effect on the threshold of 980-nm parasitic laser oscillation.
Besides the limitation of parasitic laser oscillation, another problem that needs to be considered is how to elevate the output efficiency of the 980-nm SYFS. Here, it should be noted that in the DSCCP fiber, the pump absorption is not only determined by the active core, but also determined by the coupling between the pump core and inner-cladding of signal fiber. The effect of the pump absorption of active core has been discussed in former studies such as , which will affect the optimum length of active fiber but not affect too much the scaling of 980-nm output power. Thus, in the following part, we would like to analyze the effect of the coupling coefficient k (see Eqs. (3) and (4)) on the efficiency of the SYFS.
The variations of the output power of 980-nm ASE with various coupling coefficient k are given in Fig. 8(a). It can be found that with the increment of k, the slope efficiency is obviously elevated and the pump threshold of 980-nm ASE is obviously reduced. The variation of slope efficiency with the coupling coefficient k is also given in Fig. 8(b). It can be found that the slope efficiency increases monotonously with the increasement of k. It is also implied that k should be larger than 6 to realize more than 50% slope efficiency. Figure 8(b) also gives the thresholds of parasitic laser oscillation corresponding to various coupling coefficient k. It can be found that the threshold will not change too much with k, which means that k only has a negligible effect on the power scalability of the 980-nm SYFS.
The high-power 980-nm SYFS is investigated experimentally and numerically in this paper. In experiment, an all-fiber 980-nm SYFS is fabricated with the DCSSP YDF. By optimizing the angle-cleaving of output port and active fiber length, a recorded 17.1-W output power and 14.6% slope efficiency of 980-nm ASE are obtained and no parasitic laser oscillation is observed at the maximum output power.
In order to reveal the power scalability of the SYFS, the numerical study is carried out. A simple but effective method is present to numerically determine the threshold of parasitic laser oscillation. It is found that the optical feedback should be lower than 1.2 × 10−6 and 7 × 10−7 to obtain the 50-W and 100-W 980-nm ASE, respectively. It is also found that the 1030-nm parasitic laser oscillation can be present preferentially when R is large enough, which can be suppressed by shortening the length of active fiber. Besides, it is also revealed that the coupling coefficient k of pump light is important for elevating the output efficiency of SYFS, although it only has a negligible effect on the power scalability of the source. It is implied that coupling coefficient should be larger than 6 to realize more than 50% slope efficiency. The coupling coefficient can be elevated by minimizing the interval between the pump core and inner-cladding (see Fig. 1) which can be realized by optimizing the fabrication process of DSCCP fiber. We believe that the pertinent results are of great significant and can provide important guidance for understanding and designing the 980-nm SYFS and other sorts of three-level fiber source.
National Natural Science Foundation of China (NSFC) (61405249).
J. Cao would like to thank Dr. Xiaolin Wang for their useful help.
References and links
1. D. C. Hanna, I. R. Perry, R. G. Smart, P. J. Suni, J. E. Townsend, and A. C. Tropper, “Efficient superfluorescent emission at 974 nm and 1040 nm from an Yb-doped fiber,” Opt. Commun. 72(3), 230–234 (1989). [CrossRef]
2. J. R. Armitage, R. Wyatt, B. J. Ainslie, and S. P. Craig-Ryan, “Highly efficient 980 nm operation of a Yb/sup 3+/−doped silica fibre laser,” Electron. Lett. 25(5), 298–299 (1989). [CrossRef]
3. D. J. Richardson, J. Nilsson, and W. A. Clarkson, “High power fiber lasers: current status and future perspectives,” J. Opt. Soc. Am. B 27(11), 63–92 (2010). [CrossRef]
4. A. S. Kurkov, “Oscillation spectral range of Yb-doped fiber lasers,” Laser Phys. Lett. 4(2), 93–102 (2007). [CrossRef]
5. J. Cao, S. Guo, X. Xu, J. Chen, and Q. Lu, “Investigation on power scalability of diffraction-limited Yb-doped fiber lasers,” IEEE J. Sel. Top. Quantum Electron. 20(5), 373–383 (2014). [CrossRef]
6. D. B. S. Soh, C. Codemard, and S. Wang, “A 980-nm Yb-doped Fiber MOPA Source and Its Frequency Doubling,” IEEE Photonics Technol. Lett. 16(4), 1032–1034 (2004). [CrossRef]
7. G. Machinet, J. Lhermite, and E. Cormier, “40 W picosecond fiber laser at 976 nm,” inCLEO:2011- Laser Applications to Photonic Applications, OSA Technical Digest (CD) (Optical Society of America, 2011), paper CMS2.
8. M. Laroche, C. Bartolacci, B. Cadier, H. Gilles, S. Girard, L. Lablonde, and T. Robin, “Generation of 520 mW pulsed blue light by frequency doubling of an all-fiberized 978 nm Yb-doped fiber laser source,” Opt. Lett. 36(19), 3909–3911 (2011). [CrossRef] [PubMed]
9. P. Li, G. Zhong, Z. Liu, J. Chi, X. Zhang, C. Yang, Z. Zhao, Y. Li, X. Wang, H. Zhao, and D. Jiang, “980nm Yb-doped double-clad photonic crystal fiber amplifier and its frequency doubling,” Opt. Laser Technol. 44(7), 2202–2205 (2012). [CrossRef]
10. J. He, S. Du, Z. Wang, Z. Wang, J. Zhou, and Q. Lou, “Linearly-polarized short-pulse AOM Q-switched 978 nm photonic crystal fiber laser,” Opt. Express 21(24), 29249–29254 (2013). [CrossRef] [PubMed]
12. Y. Liu, J. Cao, H. Xiao, S. Guo, L. Si, and L. Huang, “Study on the output properties of fiber lasers operating near 980 nm,” J. Opt. Soc. Am. B 30(2), 266–274 (2013). [CrossRef]
13. K. H. Ylä-Jarkko, R. Selvas, D. B. S. Son, J. K. Sahu, C. A. Codemard, J. Nilsson, S. A. Alam, and A. B. Grudinin, “A 3.5 W 977 nm Cladding-pumped Jacketed Air-Clad Ytterbium-Doped Fiber Laser,” in Advanced Solid-State Photonics, J. Zayhowski, ed., Vol. 83 of OSA Trends in Optics and Photonics (Optical Society of America, 2003), paper 103.
14. D. B. S. Soh, C. Codemard, J. K. Sahu, J. Nilsson, V. Philippov, C. Alegria, and Y. Jeong, “A 4.3 W 977 nm ytterbium-doped jacketed-air-clad fiber amplifier,” in Advanced Solid-State Photonics, OSA Technical Digest (Optical Society of America, 2004), paper MA3.
16. J. Boullet, Y. Zaouter, R. Desmarchelier, M. Cazaux, F. Salin, J. Saby, R. Bello-Doua, and E. Cormier, “High power ytterbium-doped rod-type three-level photonic crystal fiber laser,” Opt. Express 16(22), 17891–17902 (2008). [CrossRef] [PubMed]
17. M. Laroche, C. Bartolacci, G. Hervé, G. Sylvain, T. Robin, and B. Cadier, “All-fiber Yb-doped CW and pulsed laser sources operating near 980nm,” in Advances in Optical Materials, OSA Technical Digest (CD) (Optical Society of America, 2011), paper ATuB9.
18. R. Wang, Y. Liu, J. Cao, S. Guo, L. Si, and J. Chen, “Experimental study on the all-fiberized continuous-wave ytterbium-doped laser operating near 980 nm,” Appl. Opt. 52(24), 5920–5924 (2013). [CrossRef] [PubMed]
19. Y. Yu, Y. An, J. Cao, S. Guo, and X. Xu, “Experimental Study on All-Fiberized Continuous-Wave Yb-Doped Fiber Amplifier Operating Near 980 nm,” IEEE Photonics Technol. Lett. 28(4), 398–401 (2016). [CrossRef]
20. M. Leich, M. Jäger, S. Grimm, D. Hoh, S. Jetschke, M. Becker, A. Hartung, and H. Bartelt, “Tapered large-core 976 nm Yb-doped fiber laser with 10 W output power,” Laser Phys. Lett. 11(4), 045102 (2014). [CrossRef]
21. Y. Ren, J. Cao, H. Ying, H. Chen, Z. Pan, S. Du, J. Chen, and C. Zhang, “8-W all-fiber superfluorescent source operating near 980 nm,” in Laser Congress 2017 (ASSL, LAC), OSA Technical Digest (online) (Optical Society of America, 2017), paper JTh2A.22. [CrossRef]
22. Y. An, Y. Yu, J. Cao, and J. Chen, “Power scalability of a single-stage Yb-doped superfluorescent fiber source,” Laser Phys. Lett. 13(2), 1612–2011 (2016). [CrossRef]
23. Y. An, J. Cao, Z. Huang, S. Guo, X. Xu, and J. Chen, “High-power all-fiberized superfluorescent source with distributed side-coupled cladding-pumped fiber,” Appl. Opt. 53(36), 8564–8570 (2014). [CrossRef] [PubMed]
24. P. Jelger, M. Engholm, L. Norin, and F. Laurell, “Degradation-resistant lasing at 980 nm in a Yb/Ce/Al-doped silica fiber,” J. Opt. Soc. Am. B 27(2), 338–342 (2010). [CrossRef]
25. Z. Huang, J. Cao, S. Guo, J. Chen, and X. Xu, “Comparison of fiber lasers based on distributed side-coupled cladding-pumped fibers and double-cladding fibers,” Appl. Opt. 53(10), 2187–2195 (2014). [CrossRef] [PubMed]
26. C. Henry, “Theory of spontaneous emission noise in open resonators and its application to lasers and optical amplifiers,” J. Lightwave Technol. 4(3), 288–297 (1986). [CrossRef]
27. H. M. Pask, R. J. Carman, D. C. Hanna, A. C. Tropper, C. J. Mackechnie, P. R. Barber, and J. M. Dawes, “Ytterbium-doped silica fibre lasers: versatile sources for the 1-1.2µm region,” IEEE J. Sel. Top. Quantum Electron. 1(1), 2–13 (1995). [CrossRef]
28. Z. Jiang and J. R. Marciante, “Impact of transverse spatial-hole burning on beam quality in large-mode-area Yb-doped fibers,” J. Opt. Soc. Am. B 25(2), 247–254 (2008). [CrossRef]