We reported a high average power and energy microsecond pulse erbium-doped fluoride fiber MOPA system centered at 2786.8 nm. The master oscillator was a passively Q-switched erbium-doped fluoride fiber laser based on SESAM in a linear cavity. Then a one-stage erbium-doped fluoride fiber amplifier was used to boost its average output power to 4.2 W and pulse energy to 58.87 μJ. The pulse duration and repetition rate were 2.29 µs and 71.73 kHz, respectively. To the best of our knowledge, the achieved average output power and pulse energy are the recorded levels for the passively Q-switched fiber lasers at 3 μm wavelength region.
© 2016 Optical Society of America
Mid-infrared (Mid-IR) laser sources emitting at the spectral range of 2~5 µm have attracted tremendous attention owing to their widespread applications in laser surgery, gas monitoring, remote sensing, special material processing, missile countermeasures, nonlinear mid-infrared photonics . In comparison to other lasers such as solid state laser, quantum cascade laser, difference frequency generation, sum frequency generation, optical parametric oscillator, etc., fiber laser is a more outstanding platform because of its excellent beam quality, good heat dissipation, high optical conversion efficiency, and compact structure . Until now, different rare earth ions such as erbium [3–10], holmium [11–14], and dysprosium [15–17] have been doped into fluoride fiber, respectively to demonstrate Mid-IR emissions in the 3 µm or even longer wavelength regime. Recently, V. Fortin et al. have constructed an all-fiber erbium-doped fluoride fiber laser at 2.94 µm using a pair of fiber Bragg gratings. Maximum output power of 30.5 W was obtained marking the current record in this wavelength region . Very recently, O. Henderson-Sapiret al. have extended the wavelength of erbium-doped fluoride fiber to 3.78 μm by dual wavelength pumping . It is also the longest wavelength achieved from fiber lasers at room temperature.
Compared to continuous wave (CW) regime, pulsed fiber lasers in the mid-IR wavelength region especially with high average output power and pulse energy possessed larger potential in laser surgery, special material processing, infrared countermeasures, etc. Currently, two technical methods (i.e., gain switching [18–20] and Q-switching [21–32]) have been employed. Recently, M. Gorjan et al. have presented a gain-switched erbium-doped fluoride fiber laser at 2.8 µm using an active pulsed diode pump system. Maximum average output power of 2 W and peak power of 68 W (about 20.4 µJ) were obtained with the repetition rate of 100 kHz and pulse duration of 300 ns . In contrast to gain-switching requiring high energy pulse pump, Q-switching especially saturable absorbers (SA) induced passive Q-switching is much more feasible in applications since its simpleness, compactness, and low cost . Until now, different SAs including semiconductor saturable absorber mirror (SESAM) [21–23], Fe2+:ZnSe crystal [24–26], and some broadband two dimension (2D) materials (e.g., grapheme [27,28], topological insulator (TI) [29,30], and black phosphorus (BP) [31,32]) have been used to Q-switch holmium- or erbium-doped fluoride fiber lasers in the 3 µm wavelength region. Recently, T. Zhang et al. used the Fe2+:ZnSe crystal to demonstrate a passively Q-switched erbium-doped fiber laser, the maximum average output power of 1.01 W and pulse energy of 11.37 μJ were obtained which were the current highest levels of SA directly Q-switched fiber lasers in this wavelength region . However, over-bleaching of SA at a high intensity would extinguish pulses thus impeding their average power and energy enhancement. Master Oscillator Power Amplifier (MOPA), as a feasible way to significantly improve pulse power and energy, has been widely applied in fiber laser systems at 1 µm [34–36], 1.55 µm [37–39] and 2 µm [40–42]. Recently, G. W. Zhu et al. employed a passively Q-switched erbium-doped fluoride fiber laser based on graphene as the MO to present a one-stage erbium-doped fluoride fiber MOPA at 2.8 µm. The maximum average output power of 1 W and pulse energy of 24 µJ. Furthermore, the theoretical simulations also predicted the possibility of 300 µJ pulse energy with further increased pump power .
In this paper, we demonstrated the high average output power and energy pulse generation from a one-stage fluoride fiber amplifier at 2.8 µm seeded by a passively Q-switched erbium-doped fluoride fiber laser based on SESAM. The relationships between MO and amplifier in terms of output performances e.g., pulse duration, repetition rate, average output power, pulse energy, efficiency and spectrum were all investigated. The achieved maximum average power of 4.2 W and pulse energy of 58.87 μJ were the current highest levels of passively Q-switched fiber laser at 3 μm wavelength region.
2. Experimental setup and results
The experimental setup of high average output power and pulse energy erbium-doped fluoride fiber MOPA system is shown in Fig. 1. It mainly consists of a microsecond MO and a one-stage amplifier.
2.1 Master oscillator
The MO is a passively Q-switched erbium-doped fluoride fiber laser using SESAM as shown in Fig. 1. Its pump laser from a 976 nm laser diode (LD) (Lumics) was collimated using an aspheric lens (Thorlabs, 90%T@976nm) with a focal length of 20 mm and a numerical aperture (NA) of 0.52 and then launched into the inner cladding of the gain fiber via a ZnSe objective lens (Innovation Photonics, 90%T@976nm, 85%T@~2.8µm) with a focal length of 6 mm. Here the objective lens also acted as the collimator of the outputted ~2.8 µm laser. A dichroic mirror with a transmission of 94% at 976 nm and a reflectivity of 95% at ~2.8 μm was placed between the two lenses at an angle of 45° with respect to the pump beam to direct the generated pulse laser. The gain fiber was a piece of double-cladding erbium-doped fluoride fiber (LeVerre Fluoré, France). Its core has a diameter of 18 µm and a NA of 0.12. The attenuation loss at ~2.8 µm is less than 100 dB∕km, as specified by the manufacturer. The double-D-shaped inner cladding has the short/long diameters of 250/270 μm with a NA of 0.4. The erbium ions dopant concentration is 70 000 ppm thus 5.6 m fiber length can provide above 95% pump absorption efficiency. The highly erbium-doped fluoride fiber was selected as a result of the following two reasons: (1) its pump source can employ the commercial high power 976 nm laser diode; (2) highly dopants can yield a high conversion efficiency as a result of efficient energy transfer upconversion (ETU) process. The fiber end close to the pump laser held by a commercial fiber holder with V-groove (Thorlabs) was perpendicularly cleaved and acted as the output coupler with the help of ~4% Fresnel reflection. The other end of the fiber was cleaved at an angle of 8° to reduce its residual feedback and then butted against a commercial SESAM (BATOP, SAM-3000-33-10ps-x) mounted by a 3D adjuster. Here, the selection of SESAM was mainly ascribed to its mature fabrication process and high stability which were helpful for generating long-term stable Q-switching.
Figure 2 shows the output performances of the MO i.e., the passively Q-switched erbium-doped fluoride fiber laser based on SESAM. When the launched pump power was increased to 0.36 W, CW laser was firstly generated. Once it reached 0.42 W, unstable Q-switching was observed and then Q-switching became stable as the increased launched pump power to 0.45 W. This stable Q-switching state can be maintained until the maximum launched pump power of 1.27 W. The corresponding temporal pulse evolutions at a scanning range of 600 μs, 130 μs, and 14 μs are shown in Fig. 2(a) and its insets, respectively. The repetition rate of 71.43 kHz and pulse duration of 1.7 μs were achieved. The amplitude fluctuation within ± 3% indicated stable Q-switching. Figure 2(b) shows the optical and RF (inset) spectra at this pump level. The center wavelength and FWHM were 2786.8 nm and 1.3 nm, respectively. The peak frequency of 71.43 kHz matched well with the Q-switched repetition rate. The signal-to-noise ratio of 37.6 dB was also located at the typical range of passive Q-switching [21,22,25–32]. Figure 2(c) shows the pulse duration and repetition rate as a function of the launched pump power. It is observed that the repetition rate increases almost linearly from 32.68 kHz to 71.43 kHz while the pulse duration decreases from 3.16 µs to 1.7 µs with the increased launched pump power from 0.45 W to 1.27 W. Note that both of them are the typical features of passive Q-switching. Figure 2(d) shows the average output power and pulse energy as a function of the launched pump power. It is seen that both the average output power and pulse energy increase almost linearly with increasing the launched pump power. At the maximum launched pump power of 1.27 W, maximum average output power of 0.252W at a slope efficiency of 25.1% and pulse energy of 3.52 µJ were achieved. Once the launched pump power was beyond this level, Q-switching became unstable caused by the heat accumulation in the SESAM. It should be also noted that the average output power and pulse energy in this case referred to the deduced values with respect to the fiber end according to the measured values.
The one-stage amplifier mainly consisting of a high power 976 nm pigtailed LD (DILAS) with a core diameter of 200 µm and a NA of 0.22 and a segment of double-cladding erbium-doped fluoride fiber is shown in Fig. 1. The pump laser from the LD was collimated using an aspheric lens (90%T@976 nm) with a focal length of 20 mm and a NA of 0.543, then it was launched into the inner cladding of the erbium-doped fluoride fiber via a CaF2plano-convex lens (92%T@976nm, 93%T@~2.8 µm) with a focal length of 20 mm. Here, the CaF2plano-convex lens was also used to collimate the outputted ~2.8 μm laser from the fiber core. A dichroic mirror with a transmission of 94% at 976 nm and reflectivity of 95% at~2.8 μm was placed between the two lenses at an angle of 45° with respect to the pump beam to direct the generated laser. A 3 μm bandpass filter (Thorlabs, FB3000-500) placed along the output beam was used to remove the residual pump light. In this part, the erbium-doped fluoride fiber was same as that used in the MO. The length was selected to be 4.3m which can provide above 90% pump absorption efficiency. The fiber end close to the pump was cleaved at an angle of 10° to reduce the residual reflection. In contrast, the fiber end close to the MO was perpendicularly cleaved in order to make the seed coupling easier. The Q-switched pulses generated from the MO passed through a free-space polarization dependent isolator (PDI) (FastPulse Technology, Inc.) and then was launched into the core of the double-cladding erbium-doped fluoride fiber using a CaF2 plano-convex lens with a transmission of 72%.The launching efficiency was measured to be 52.8% using the cutback method. Note that the PDI has a low transmission of 33% with respect to our randomly polarized Q-switched pulses. Here our selection of PDI was due to lack of polarization independent isolator in the lab. Therefore the total launching efficiency from the fiber pump end in the MO to the fiber end close to the MO was calculated to be 10.1% (i.e., 85% × 95% × 33% × 72% × 52.8%).
Based on this arrangement, the output performances of the one-stage amplifier at different launched MO powers were investigated. The relative experimental results in this case are all labeled as Case 1. Figure 3 shows the temporal pulse trains and the corresponding single pulse waveforms of the amplified Q-switched pulses at different launched pump powers of 0.29 W, 4.46 W, and 10.69 W when the launched MO power was fixed at the maximum 25.5 mW. The low amplitude fluctuations within ± 5% indicated that amplified Q-switching was quite stable. Their repetition rates were all around 71 kHz. The corresponding pulse durations were 1.28 μs, 2.17 μs, and 2.18 μs, respectively as shown in Fig. 3(d). The fluctuations of the pulse durations and repetition rates were less than 4% and 3% of their RMSs, respectively indicating their high stability. Note that the pulse duration of 1.28 μs at the low launched pump power of 0.29 W was smaller than that of the MO (i.e., 1.7 µs). This is because only the peak region of the pulse is amplified due to lack of populations on the 4I11/2 level at the low pump level therefore the pulse is narrowed. With the increased launched pump power, the whole pulse is amplified thus yielding almost unchanged pulse duration. Once the launched pump power was beyond10.69 W, however, the amplified Q-switching became unstable with sharply degraded amplitudes and severe timing jitter and then changed into CW emission state with the further increased launched pump power to 10.8 W. To explain this phenomenon, their spectra were measured as shown in Fig. 4. It is found that the spectra are almost same centered at 2786.8 nm when the launched pump power is below 10.69 W. The calculated normalized frequency of 2.43 slightly larger than 2.405 indicated the output was close to the single mode. No obvious ASE pedestals in the spectra and the calculated ASE suppression of >27 dB indicated quite low ASE components involved. When the launched pump power was increased to 10.8 W, the spectrum was blue-shifted with the center wavelength of 2785.1 nm. At this time, if the MO was switched off, the spectrum almost kept unchanged. It suggested that the CW oscillation occurred in the amplifier whose feedbacks were from the residual reflections of perpendicularly cleaved fiber end, and two CaF2 plano-convex lenses. The phenomenon can be explained by the fact that with the increased launched pump power, more MO photons are required to consume populations on the 4I11/2 level, otherwise, the CW oscillation will defeat pulse amplification. At the other launched MO powers, stable Q-switching amplification can be still observed with the similar phenomenon. Figure 5(a) shows the repetition rate and pulse duration as a function of the launched pump power at different launched MO powers. It is observed that the repetition rates are almost unchanged at the levels around 71 kHz, 52 kHz, and 32 kHz with respect to the launched MO power of 25.5 mW, 12.6 mW, and5.2 mW, respectively with the increased launched pump power during amplifying similar to previous report .Their significantly slight fluctuations which can be ignored in this case were mainly resulted from the pump power fluctuation of the MO, heat disturbance in the amplifier, environment, etc. But the pulse durations firstly increased and then tended to be unchanged. Besides, it is also found that the Q-switching amplification was terminated earlier for the smaller launched MO power which agrees with our previous explanation. Finally, the pulse durations of 2.14 µs, 2.6 µs, and 4.11 µs were obtained at the maximum launched pump power of 10.69 W, 7.91 W, 4.67 W with respect to the launched MO power of 25.5 mW, 12.6 mW, 5.2 mW, respectively. Moreover, the amplified average output powers and pulse energies were also measured with the varied launched pump power at different launched MO powers as shown in Fig. 5(b). It is observed that all the average output powers increase almost linearly with the increased launched pump power. The power amplification efficiency increased from 25.93% to 30.54% with the increased launched MO power from 5.2 mW to 25.5 mW similar to the previous report . Similarly, the pulse energies also increased linearly with the launched pump power. On contrary, the energy amplification efficiency decreased with the increased launched MO power as a result of the increased repetition rate. Finally, the maximum average output power of 3.2 W and pulse energy of 44.72 μJ were achieved at the maximum launched pump power of 10.69 W and MO power of 25.5 mW.
In the above experiment, pulse amplifying was finally terminated by CW oscillation in the amplifier. Besides, it is also found that higher launched MO power can lead to higher pump power supporting pulse amplifying. For our system where the maximum launched MO power was fixed, the simplest way to boost the average output power and pulse energy is to enhance CW oscillation threshold by reducing the feedback. Then the fiber end in the amplifier close to the MO was cleaved at an angle of 10° to reduce the residual reflection. In comparison to perpendicularly cleaved fiber end, however, it is more difficult for angle-cleaved fiber end to realize high-efficiency coupling. In order to obtain the same launched MO power as Case 1, the positions of angle-cleaved fiber end and CaF2 plano-convex lens were carefully optimized. The relative experimental results in this case are labeled as Case 2.
In this system, stable Q-switching can be achieved from the amplifier as well. Figure 6 shows the pulse trains and single pulse waveforms at the launched pump power 0.78 W, 5.87 W, 14.03 W when the launched MO power was fixed at 25.5 mW. Their repetition rates were all same as those of the MO similar to that in Case 1. The corresponding pulse durations were 1.46 μs, 2.31 μs, and 2.28 μs, respectively. Here the pulse duration and repetition rate were still stable like Case 1. Besides, it is observed that the pulse duration of 1.46 μs at the low launched pump power of 0.78 W was smaller than that of MO (i.e., 1.7 µs) which has been explained in Case 1. When the launched pump power was beyond 14.03 W, the unstable Q-switching appeared and then CW oscillation happened with further increasing launched pump power to 14.2 W. The spectra at the launched pump power of 0.78 W, 5.87 W, 14.03 W, and 14.2 W are shown in Fig. 7. It is seen that the center wavelength was blue-shifted from 2786.8 nm to 2785.4 nm while the FWHM was slightly broadened from 1.3 nm to 1.44 nm similar to the phenomena observed in Case 1. Figure 8(a) shows the repetition rate and pulse duration as a function of the launched pump power at different launched MO powers. Different from Case 1 as shown in Fig. 5(a), the CW oscillation thresholds were all enhanced compared to those at the same launched MO powers. It was caused by the reduced feedback as a result of angle-cleaving the other fiber end. However, the residual reflections from two CaF2 plano-convex lenses which cannot be ignored gave rise to the CW oscillation at high pump power. If the MO was switched off, the maximum output power of the ASE in the amplifier was measured to be only 78.2 mW which was also limited by the CW oscillation. Finally, the pulse duration of 2.29 µs, 2.82 µs, 4.4 µs and repetition rate of 71.73 kHz, 52.87 kHz, 32.96 kHz were obtained at the launched MO power of 25.5 mW, 12.6 mW, 5.2 mW, respectively. Figure 8(b) shows the amplified average output powers and pulse energies with the increased launched pump power at different launched MO powers. The same phenomena were observed as those displayed in Fig. 5(b). Compared to Case 1, the higher maximum average output power of 4.2 W at a power amplification efficiency of 30.79% and higher pulse energy of 58.87 μJ were achieved at the maximum launched pump power of 14.03 W and MO power of 25.5 mW. These were the current highest levels of passively Q-switched fiber laser systems in 3 μm wavelength region.
In our experiment, although the highest average output power of 4.2 W and pulse energy of 58.87 μJ were achieved, there were still some room for further improvement for the current system. Specifically, they can be divided into the following two aspects: (1) improving the MO power launched into the gain fiber of the amplifier; (2) reducing the residual reflections in the amplifier.
For the first scheme, both MO power and launching efficiency from the MO to the gain fiber of the amplifier were responsible. In our case, the maximum MO power was mainly limited by the performance of the SESAM. At a high pump level, large amount of heat accumulation in it made Q-switching unstable. Thus active or passive cooling units (e.g., liquid-nitrogen cooling, air cooling, water cooling, metal cooling, etc.) can be designed and imposed on it to accelerate heat dissipation. Besides, Fe2+:ZnSe crystal able to work well at high power can be also considered albeit with more complex alignment. Expect the management of the passive modulator, it is also critical to optimize the total launching efficiency from the MO to the gain fiber of the amplifier to improve the launched MO power. In our experiment, only about 12.5% of the laser from the MO was launched into the erbium-doped fluoride fiber in the amplifier. The 33% low transmission of the PDI played a dominant negative role due to lack of high transmission polarization-independent isolator in lab. Moreover, the transmission of the CaF2 plano-convex lens with a focal length of 40 mm was measured to be just 72% using a home-made 2.8μm erbium-doped fluoride fiber laser. It was abnormally lower than the value (>90%) provided by the manufacturer due to some defects existed which led to a relatively low launching efficiency of 52.8%. For the second scheme, the most available method was to anti-coat the optical components (i.e., CaF2 plano-convex lens, fiber end, etc.) in the amplifier thus reducing their residual reflections. Besides, the all-fiber arrangement by connecting the MO with the amplifier was also an ideal approach in the future.
Based on the above schemes, significant improvement of the output without serious pulse shape distortion was reasonably predicted by increasing the pump power, since the saturation energy of our used erbium-doped fluoride fiber at 2786.8 nm was calculated to be ~145 μJ . However, higher average output power and pulse energy should resort to multi-stage amplifying arrangement by exploiting larger-mode-area erbium-doped fluoride fiber.
In this paper, we demonstrated high average power and energy microsecond pulse generation centered at 2786.8 nm from a one-stage erbium-doped fluoride fiber MOPA system. The MO was a SESAM passively Q-switched erbium-doped fluoride fiber laser producing the shortest pulse duration of 1.7 μs and largest repetition rate of 71.43 kHz. It can generate the maximum average output power of 251.6 mW and pulse energy of 3.52 μJ. The amplifier included a piece of erbium-doped fluoride fiber pumped at 976 nm. The experimental results suggested that higher launched MO power and lower residual reflections in the amplifier were helpful for pulse amplification. Besides, the repetition rate was almost unchanged during amplifying, but the pulse duration was slightly larger than that of MO. Consequently, at the maximum launched MO power of 25.5 mW and pump power of 14.03 W, maximum average output power of 4.2 W at a power amplification efficiency of 30.79% and pulse energy of 58.87 μJ were achieved marking the current highest levels of passively Q-switched fiber laser systems in the 3 μm wavelength region. The repetition rate and pulse energy were 71.73 kHz and 2.29μs, respectively. In the next step, multi-stage amplifying was planned to further improve the average output power and pulse energy. Besides, more efforts on amplifying mode-locked ultrafast pulses would be also deployed in the near future.
This work was supported by National Natural Science Foundation of China (61377042, 61435003, 61505024, 61421002); Open Fund of State Key Laboratory of Advanced Optical Communication Systems and Networks (2015GZKF004); Open Fund of Key Laboratory of Specialty Fiber Optics and Optical Access Networks, Shanghai University (SKLSFO2014-07); Open Fund of Science and Technology on Solid-State Laser Laboratory (H04010501W2015000604); Science and Technology Innovation Young Talent Project of Sichuan Province (2016RZ0073)
References and links
1. D. Pile and N. Horiuchi, “RPCh.Won, andO Graydon, “Extending opportunities,” Nat. Photonics 6(7), 407 (2012). [CrossRef]
2. S. D. Jackson, “Towards high-power mid-infrared emission from a fibre laser,” Nat. Photonics 6(7), 423–431 (2012). [CrossRef]
6. J. F. Li, L. L. Wang, H. Y. Luo, J. T. Xie, and Y. Liu, “High power cascaded erbium doped fluoride fiber laser at room temperature,” IEEE Photonics Technol. Lett. 28(6), 673–676 (2016). [CrossRef]
7. H. Toebben, “Room temperature cw fibre laser at 3.5µm in Er3+-doped ZBLAN glass,” IEEE Electron. Lett. 28(14), 1361–1362 (1992). [CrossRef]
12. S. Crawford, D. D. Hudson, and S. D. Jackson, “High-power broadly tunable 3-μm fiber laser for the measurement of optical fiber loss,” IEEE Photonics J. 7(3), 150239 (2015). [CrossRef]
13. C. Carbonnier, H. Tobben, and U. B. Unrau, “Room temperature CW fibre laser at 3.22 μm,” IEEE Electron. Lett. 34(9), 893–894 (2002). [CrossRef]
14. J. Schneider, C. Carbonnier, and U. B. Unrau, “Characterization of a Ho(3+)-doped fluoride fiber laser with a 3.9- μm emission wavelength,” Appl. Opt. 36(33), 8595–8600 (1997). [CrossRef] [PubMed]
15. Y. H. Tsang, A. E. El-Taher, T. A. King, and S. D. Jackson, “Efficient 2.96 μm dysprosium-doped fluoride fibre laser pumped with a Nd:YAG laser operating at 1.3 μm,” Opt. Express 14(2), 678–685 (2006). [CrossRef] [PubMed]
16. S. D. Jackson, “Continuous wave 2.9μm dysprosium-doped fluoride fiber laser,” Appl. Phys. Lett. 83(7), 1316–1318 (2003). [CrossRef]
18. B. C. Dickinson, P. S. Golding, M. Pollnau, T. A. King, and S. D. Jackson, “Investigation of a 791-nm pulsed-pumped 2.7-μm Er-doped ZBLAN fibre laser,” Opt. Commun. 191(3–6), 315–321 (2001). [CrossRef]
20. Y. L. Shen, K. Huang, S. Q. Zhou, K. P. Luan, L. Yu, A. Q. Yi, G. B. Feng, and X. S. Ye, “Gain-switched 2.8 μm Er3+-doped double-clad ZBLAN fiber laser,” Proc. SPIE 9543, 95431E (2015).
21. J. F. Li, H. Y. Luo, Y. L. He, Y. Liu, L. Zhang, K. M. Zhou, A. G. Rozhin, and S. K. Turistyn, “Semiconductor saturable absorber mirror passively Q-switched 2.97 m fluoride fiber laser,” Laser Phys. Lett. 11(6), 065102 (2014). [CrossRef]
22. J. Li, H. Luo, L. Wang, Y. Liu, Z. Yan, K. Zhou, L. Zhang, and S. K. Turistsyn, “Mid-infrared passively switched pulsed dual wavelength Ho(3+)-doped fluoride fiber laser at 3 μm and 2 μm,” Sci. Rep. 5, 10770 (2015). [CrossRef] [PubMed]
23. Y. Shen, Y. Wang, K. Luan, K. Huang, M. Tao, H. Chen, A. Yi, G. Feng, and J. Si, “Watt-level passively Q-switched heavily Er(3+)-doped ZBLAN fiber laser with a semiconductor saturable absorber mirror,” Sci. Rep. 6, 26659 (2016). [CrossRef] [PubMed]
24. C. Wei, X. Zhu, R. A. Norwood, and N. Peyghambarian, “Passively Q-switched 2.8-μm nanosecond fiber laser,” IEEE Photonics Technol. Lett. 24(19), 1741–1744 (2012). [CrossRef]
25. J. Li, H. Luo, L. Wang, B. Zhai, H. Li, and Y. Liu, “Tunable Fe(2+):ZnSe passively Q-switched Ho3+-doped ZBLAN fiber laser around 3 μm,” Opt. Express 23(17), 22362–22370 (2015). [CrossRef] [PubMed]
26. T. Zhang, G. Feng, H. Zhang, S. Ning, B. Lan, and S. Zhou, “Compact watt-level passively Q-switchedZrF4-BaF2-LaF3-AIF3-NaFfiber laser at 2.8 μm using Fe2+:ZnSesaturable absorber mirror,” Opt. Eng. 55(8), 086106 (2016). [CrossRef]
27. C. Wei, X. Zhu, F. Wang, Y. Xu, K. Balakrishnan, F. Song, R. A. Norwood, and N. Peyghambarian, “Graphene Q-switched 2.78 μm Er3+-doped fluoride fiber laser,” Opt. Lett. 38(17), 3233–3236 (2013). [CrossRef] [PubMed]
28. G. Zhu, X. Zhu, K. Balakrishnan, R. A. Norwood, and N. Peyghambarian, “Fe2+:ZnSe and graphene Q-switched singly Ho3+-doped ZBLAN fiber lasers at 3 μm,” Opt. Mater. Express 3(9), 1365–1377 (2013). [CrossRef]
29. J. Li, H. Luo, L. Wang, C. Zhao, H. Zhang, H. Li, and Y. Liu, “3-μm Mid-infrared pulse generation using topological insulator as the saturable absorber,” Opt. Lett. 40(15), 3659–3662 (2015). [CrossRef] [PubMed]
30. P. Tang, M. Wu, Q. Wang, L. Miao, B. Huang, J. Liu, C. Zhao, and S. Wen, “2.8 μm pulsed Er3+: ZBLAN fiber laser modulated by topologicalinsulator,” IEEE Photonics Technol. Lett. 28(14), 1573–1576 (2016). [CrossRef]
31. Z. Qin, G. Xie, H. Zhang, C. Zhao, P. Yuan, S. Wen, and L. Qian, “Black phosphorus as saturable absorber for the Q-switched Er:ZBLAN fiber laser at 2.8 μm,” Opt. Express 23(19), 24713–24718 (2015). [CrossRef] [PubMed]
32. J. Li, H. Luo, B Zhai, R Lu, Z Guo, H Zhang, and Y Liu,“Black phosphorus: a two-dimensionsaturable absorption material formid-infrared Q-switched and modelockedfiber lasers,” Sci. Rep. 6, 30316 (2016). [PubMed]
33. Z Liand and H. Y Luo, “Recent progress on passively switched mid-infrared fiber lasers at 3 µm,” J. Electron. Sci. Technol. 13(4), 305–314 (2015).
34. J. Limpert, A. Liem, T. Gabler, H. Zellmer, A. Tünnermann, S. Unger, S. Jetschke, and H. R. Müller, “High-average-power picosecond Yb-doped fiber amplifier,” Opt. Lett. 26(23), 1849–1851 (2001). [CrossRef] [PubMed]
35. K. K. Chen, J. H. V. Price, S. U. Alam, J. R. Hayes, D. Lin, A. Malinowski, and D. J. Richardson, “Polarisation maintaining 100W Yb-fiber MOPA producing microJ pulses tunable in duration from 1 to 21 ps,” Opt. Express 18(14), 14385–14394 (2010). [CrossRef] [PubMed]
36. L. Zhang, Y. G. Wang, H. J. Yu, W. Sun, Y. Y. Yang, Z. H. Han, Y. Qu, W. Hou, J. M. Li, X. C. Lin, and Y. Tsang, “20 W high-power picosecond single-walled carbon nanotube based MOPA laser system,” J. Lightwave Technol. 30(16), 2713–2717 (2012). [CrossRef]
39. M. Michalska and J. Swiderski, “Highly efficient, kW peak power, 1.55 µm all-fiber MOPA system with a diffraction-limited laser output beam,” Appl. Phys. B 117(3), 841–846 (2014). [CrossRef]
40. X. Wang, X. Jin, P. Zhou, X. Wang, H. Xiao, and Z. Liu, “All-fiber-integrated narrowband nanosecond pulsed Tm-doped fiber MOPA,” IEEE Photonics Technol. Lett. 27(14), 1473–1476 (2015). [CrossRef]
41. X. Wang, X. Jin, P. Zhou, X. Wang, H. Xiao, and Z. Liu, “105 W ultra-narrowband nanosecond pulsed laser at 2 μm based on monolithic Tm-doped fiber MOPA,” Opt. Express 23(4), 4233–4241 (2015). [CrossRef] [PubMed]
42. J. Liu, C. Liu, H. Shi, and P. Wang, “High-power linearly-polarized picosecond thulium-doped all-fiber master-oscillator power-amplifier,” Opt. Express 24(13), 15005–15011 (2016). [CrossRef] [PubMed]
43. G. Zhu, X. Zhu, R. A. Norwood, and N. Peyghambarian, “Experimental and numerical investigationson Q-switched laser-seeded fiberMOPA at 2.8 μm,” J. Lightwave Technol. 32(23), 3951–3955 (2014). [CrossRef]
44. A. Oladeji, A. Phillips, S. Lamrini, K. Scholle, P. Fuhrberg, A. B. Seddon, T. M. Benson, and S. Sujecki, “Design of erbium doped double clad ZBLAN fibre laser,” J. Phys. 619(1), 012044 (2015).