Open Access
Add to BibSonomy
Abstract
We demonstrate a gain-switched fiber laser, yielding a maximum average power of 1.04 W at 3.46 μm, which is the current record of a pulsed rare-earth-doped fiber laser at the wavelength beyond 3 μm, to our knowledge. The corresponding pulse energy is 10.4 μJ with a repetition rate of 100 kHz. A dual-wavelength pumping scheme consisting of a home-made 1950 nm passively Q-switched fiber laser system with a μs-scale pulse width. A 976 nm continuous wave laser diode was used to gain-switch a double-cladding Er-doped ZBLAN fiber laser cavity. Possible laser-quenching behavior during a single-pump pulse was circumvented for the moderate pump peak power and relatively large-pump pulse width. Synchronous gain-switched pulses were achieved with a tunable repetition rate at a wide range of 55~120 kHz, which is the highest gain-switching repetition rate at this band and only limited by our pulsed-pump source. Moreover, the significance of pump pulse width for repetition rate improvement is also discussed. These results provide an available way to produce high-power pulses at the mid-infrared range of 3~5 μm.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Mid-infrared lasers emitting in the region of 3~5 μm have attracted large amount of attention owing to their widespread applications in spectroscopy, infrared countermeasures, gas sensing and monitoring, etc [1–4]. Particularly, 3.5 μm band greatly overlaps with the absorption lines of the C-H bone, which is present in some molecules (e.g., methane, ethane, propane, etc.) [5] and polymer materials [6]. Thus this kind of laser could serve the fields of remote sensing and special material processing well. Recently, O. Henderson-Sapir et al. proposed a novel dual-wavelength pumping (DWP) scheme (985 nm + 1973 nm) based on Er-doped ZBLAN fiber, in which the ions on the long lifetime 4I11/2 level could be recycled by 1973 nm pump, to achieve hundreds of mW 3.4~3.6 μm (4F9/2→4I9/2) continuous wave (CW) laser at room temperature, for the first time [7]. This work provides a good opportunity for efficient lasing at this band while opens the door of developing fiber lasers beyond 3 μm. Since then, many experimental and theoretical efforts have been made in order to better understand the dynamics of this laser system hence improving the output performance [8–12]. Until now, CW laser power has been scaled to 5.6 W in an all-fiber architecture [8] while the maximum tuning wavelength range has been up to 450 nm (3.33~3.78 μm) [9]. After that, it has been also pushed towards pulsed version owing to the more versatile applications [13–15]. N. Bawden et al. presented the first Q-switched Er-doped ZBLAN fiber laser around 3.5 μm based on an acousto-optic modulator (AOM), giving a maximum average power of ~350 mW (120 kHz) [13] and a maximum peak power of 14.5 W (15 kHz) [14]. Therein, further power scaling was inhibited by the damaged pump fiber end with the enhanced laser peak power. Then Z. P. Qin et al. used black phosphorus (BP) as the saturable absorber (SA) to demonstrate passive Q-switching and mode-locking in an Er-doped ZBLAN fiber laser at this band, for the first time [15]. The achieved maximum average power of 120 mW (64.3 kHz) under Q-switching regime was limited by the heat damage of the BP. Compared to Q-switching, gain-switching, resorting to pulse pumping, has been regarded as a compact and simple method to produce short pulse and widely used at shorter 3 μm band, since no additional intra-cavity modulators are needed [16–20]. At 2.8 μm (4I11/2→4I13/2), the record average power of 11.2 W has been achieved from an all-fiber gain-switched Er-doped ZBLAN fiber laser [20]. Recently, F. Jobin et al. used an amplified actively Q-switched laser source with a pulse width of several tens of ns at 1976 nm combined with a 976 nm CW laser diode (LD) as the pump to build up a gain-switched DWP Er-doped ZBLAN fiber laser [21]. Stable pulses at 3.552 μm were obtained within a narrow repetition rate range of 15~20 kHz, outside which pulses would become significantly unstable. Consequently, a quite high peak power of 204 W (15 kHz) was obtained while the maximum average power was ~114 mW (20 kHz) that was 2 orders lower than the record they achieved at 2.8 μm [20]. Although further power scaling at 20 kHz in this case was only limited by the 1976 nm pump power, the possible laser quenching behavior observed at lower repetition rates resulted from the excited state absorption (ESA) transition at 1976 nm between 4F9/2 and 4F7/2 levels has been regarded as a potential limit factor. Besides, high pump peak power is also a considerable factor inhibiting the scaling of gain-switching power [21]. Very recently, the numerical model of a cascade-gain-switched DWP Er-doped ZBLAN fiber laser at this band (1550 nm→1970 nm→3.5 μm) was built up by J. L. Yang et al. to explore its dynamics. The influence of pump peak power, pulse width, repetition rate, etc., on gain-switching output performance was investigated. An upper limit of the pump repetition rate of around 100 kHz was estimated at the 976 nm CW pump power of 5 W [22].
In this work, we demonstrate a 1.04 W gain-switched DWP Er-doped ZBLAN fiber laser at 3.46 μm, which is the first watt-level pulsed rare-earth-doped fiber laser beyond 3 μm. To circumvent the possible laser quenching behavior during a single pump pulse and the damage of the fiber end caused by high peak power, a passively Q-switched fiber laser system at 1950 nm with a μs-scale width and a moderate peak power was selected as the pump combined with a 976 nm CW LD. Its flexibly adjustable pump repetition rate at the range of 55~120 kHz is beneficial to experimentally identify the upper limit of pump repetition rate. Finally, synchronous gain-switching with a maximum repetition rate of 120 kHz was obtained, only limited by the 1950 nm pulsed pump source. The influence of pump repetition rate on laser performance was also investigated. Moreover, the significance of pump pulse width for ~3.5 μm gain-switching formation is discussed.
2. Experimental setup
The schematic of our constructed gain-switched DWP fiber laser at ~3.5 μm is shown in Fig. 1 which is a typical double-end pumped linear cavity. The inset of Fig. 1 shows the principle of the DWP scheme in our case where ions on the ground state 4I15/2 level are excited to the 4I11/2 level by 976 nm CW laser first and subsequently periodically to the 4F9/2 level by 1950 nm pulse laser. Ultimately, ~3.5 μm gain-switched pulses are produced from the transition 4F9/2→4I9/2. The gain fiber is a piece of 3.2 m length commercial double-cladding Er-doped ZBLAN with a dopant concentration of 1.5 mol.% (Le Verre Fluoré, France). The core has a diameter of 16.5 μm with a numerical aperture (NA) of 0.15 and the inner cladding has a diameter of 260 μm with two parallel flats separated by 240 μm. Its outer cladding is a low-index fluoroacrylate polymer with a diameter of 290 μm. Both ends of the fiber were perpendicularly cleaved and closed by two different specifically coated dichroic mirrors (DMs) (i.e., DM2 and DM3) as the cavity feedbacks. The DM2 has a high transmittance (HT) of >99%(>90%) at 976 nm (2.8 μm) and a reflectance of 60% around 3.5 μm acting as the output coupler, and the DM3 has a HT of 94%(>90%) at 1950 nm (2.8 μm) and a high reflectance (HR) of >99% around 3.5 μm. The 976 nm CW laser was from a 30 W LD (BWT, China) with a fiber pigtail having a core diameter of 105 nm and a NA of 0.22. It was launched into the inner cladding of the gain fiber after collimating using an uncoated asphericlense with a focal length of 10.5 mm (ACL1210U, Thorlabs) and focusing using a mid-infrared AR-coated ZnSe objective lense (LFO-5-12, Innovation Photonics) with a focal length of 12 mm. Although the ZnSe objective lense has a moderate transmittance of 69% at 976 nm, its transmittance around 3.5 μm is up to 93% (provided by the manufacturer). The launching efficiency was roughly measured to be 80% using a piece of 10 cm length same ZBLAN fiber where the absorption could be almost neglected considering its low absorption coefficient. Another specifically coated DM1 having a HT of >99% at 976 nm and a high reflectance (HR) of >99% around 3.5 μm was placed between them at an angle of 45° with respect to the 976 nm laser to steer the generated ~3.5 μm laser. The 1950 nm pulses came from a home-made Tm-doped fiber based MOPA system consisting of a Bi2O2Se passively Q-switched Tm-doped fiber laser seed and a two-stage Tm-doped fiber based amplifier pumped by high power 793 nm LDs. It had a tunable repetition rate and pulse duration of 55~120 kHz and 1.58~2.21 μs, respectively. Note that 1950 nm wavelength is still located within the strong absorption band of the ESA 4I11/2→4F9/2, although its absorption cross-section has an ~20% reduction compared to its peak at 1976 nm. The 1950 nm pulses outputted from the MOPA system through a SMF28e fiber pigtail were launched into the core of the other end of the gain fiber using a pair of same ZnSe objective lenses (LFO-5-12, Innovation Photonics) with focal lengths of 12 mm, which have a HT of 90% at 1950 nm. The total launching efficiency was measured to be 82% using the same method as before, of which about 93% propagated in the core. This was estimated according to the ratio of the residual 1950 nm laser when the 976 nm pump was off and adjusted to a high enough power, respectively. Accordingly, the final core launching efficiency was calculated to be ~76% ( = 82%*93%). Note that the propagation losses of 1950 nm laser in the core and inner cladding were regarded as the same in this estimation. In fact, the actual core launching efficiency should be slightly larger than ~76% because the 1950 nm laser experiences a larger propagation loss in the inner cladding than the core. The seed pulses were monitored using a (HgCd)Te photodetector (PCI-2TE-12, Poland) with a rise time of ≤2 ps connecting to one channel (CH1) of a 500 MHz bandwidth digital oscilloscope as the trigger signal. Note that the 1950 nm pulse width was almost unchanged after amplifying except slightly higher pulse leading edge. The output ~3.5 μm laser was monitored using an InAs detector (Judson J12D) with a response time of 2 ns connected to another channel (CH2) of the oscilloscope after purifying using a bandpass filter (FB3500-500, Thorlabs) which has a transmittance of 77% around 3.5 μm (provided by the manufacturer). A radio frequency (RF) spectrum analyzer (AV4033A) with a scanning range of 30 Hz-18 GHz was used to record the RF spectrum. A monochromator with a minimum scanning resolution of 0.1 nm (Princeton instrument Acton SP2300) was used to measure the optical spectrum.
Fig. 1 Experimental setup of the gain-switched DWP double-cladding Er-doped ZBLAN fiber laser (DM, dichroic mirror; HT, high transmittance; HR, high reflectance). The inset shows the simplified energy diagram with relevant pump and laser transitions.
3. Results
3.1 Output depending on 1950 nm pump power (pulse energy)
We know that passively Q-switched repetition rate is positively related to its pump power. In our seed, although the pulse repetition rate could be tuned at the range of 55~120 kHz, for the repetition rates beyond 100 kHz, Q-switched pulses were a bit unstable due to the excessive heat accumulation in the Bi2O2Se SA. Thus, the seed repetition rate was set at 100 kHz while the launched 976 nm pump power was optimized to be 2.3 W which could just maximize ~3.5 μm output power at the launched 1950 nm pump power (pulse energy) of 6.1 W (61 μJ).
When the launched 1950 nm pump power was increased slightly beyond the laser threshold of 1.97 W (19.7 μJ), low-amplitude pulsing matching with the pump was observed. Further increasing the pump to 3.40 W (34 μJ), stable gain-switching at 50 kHz half of the pump repetition rate was achieved, which could maintain until the pump power (pulse energy) of 3.48 W (34.8 μJ) as shown in Fig. 2(a). Figure 2(b) shows the zoomed single pulse waveforms where there was a μs-scale delay between two closest pump and laser pulses in a period and the laser pulse width was 2.51 μs. Beyond this pump level, the laser went into a chaotic state until the pump power (pulse energy) of 3.58 W (35.8 μJ), at which it recovered to stable gain-switching but with a repetition rate of 100 kHz. It could last until the pump power (pulse energy) of 6.10 W (61 μJ) as shown in Fig. 2(c), giving a decreased pulse width of 1.61 μs extracted from Fig. 2(d). If increasing the pump power again, the second relaxation spike appeared. This temporal evolution process was similar to that observed in gain-switched Er- and Ho-doped ZBLAN fiber lasers at ~3 μm [17,19].
Fig. 2 Temporal behaviors at the launched 1950 nm pump power (pulse energy) of (a), (b) 3.48 W (34.8 μJ) (100 kHz) and (c), (d) 6.10 W (61 μJ) (100 kHz). (P976nm = 2.3 W)
Figure 3(a) shows the ~3.5 μm output average power as a function of its launched 1950 nm pump power. Due to shorter pulse pump wavelength and lower output cavity feedback, our pump threshold (i.e., 1.97 W) was larger than the ~1.5 W in a CW case where the same gain fiber was used [23]. It is seen that the output average power increases almost linearly with the launched 1950 nm pump power at a slope efficiency of 24.6%. No laser quenching observed at this pump range implied that our μs-scale pump pulse was not strong enough so that the 4I11/2 level was completely depopulated during a single pump pulse [20]. At the pump power of 6.1 W, the maximum average power of 1.04 W was obtained, which was the record of a pulsed rare-earth-doped fiber laser beyond 3 μm. Note that the average power here referred to that after the DM2 and was corrected by a factor of 0.716 (0.93*0.77) based on the data recorded by the powermeter. The transmittances of 93% for the ZnSe objective lens and 77% for the bandpass filter provided by the manufacturers (available online) were also checked in our experiment. The long-term power stability under gain-switching was monitored within 30 min as shown in Fig. 3(b). The power fluctuation of 2.49% indicated the high stability. Figure 3(c) shows the corresponding optical spectrum with three peaks locating at 3457.5 nm, 3465 nm, and 3469.5 nm, respectively. The signal-to-noise ratio (SNR) was >40 dB with a RBW of 300 Hz as shown in the inset of Fig. 3(c). Figure 3(d) shows the laser pulse width, time delay, pulse energy, and peak power with respect to the pump power. For the 50 kHz gain-switched pulses, the laser system underwent standard gain-switching behavior with the increased pump power (pulse energy), yielding a simultaneous reduction in laser pulse width and time delay and an increase in pulse energy and peak power. The similar evolution trends were observed for the 100 kHz gain-switched pulses. All of them are typical gain-switched features [16–20]. At the pump power (pulse energy) of 6.1 W (E = 61 μJ), the shortest pulse width of 1.61 μs and time delay of 1.81 μs were gotten. The maximum pulse energy and peak power were 10.4 μJ and 6.46 W, respectively.
Fig. 3 (a) Output average power versus launched 1950 nm pump power, (b) average power stability within 30 min, (c) optical and RF spectra at the launched 1950 nm pump power of 6.1 W, (d) laser pulse width, time delay, pulse energy, and peak power versus launched 1950 nm pump power. (P976nm = 2.3 W)
3.2 Output depending on 1950 nm pump repetition rate
In Section 3.1, the influence of 1950 nm pump power (or pulse energy) on laser performance was studied by varying its peak power. Generally, the peak power should be fixed if revealing the influence of 1950 nm pump repetition rate [16,20]. However, in this way, we found it was difficult to isolate it from the varied pulse energy in our case, since the pump pulse width was always related to the repetition rate due to passive Q-switching nature. Thus, the investigation was performed at a fixed pump pulse energy of 50 μJ. It was still reasonable because several hundreds of ns variations for μs-scale pump pulse width slightly affected laser performance for a fixed pump pulse energy [19,22]. Note that gain-switching here referred to the state with the same repetition rate as its pump.
Figure 4 shows the evolutions of seed and laser pulse widths, average power, pulse energy, and peak power with the varied pump repetition rate at a fixed pump pulse energy of 50 μJ. It is observed that the seed pulse width decreases first as typical Q-switching behavior and then increase with the increased repetition rate. This opposite evolution phenomenon was caused by the excessive heat induced performance change of the Bi2O2Se SA and observed in our recently reported passively Q-switched fiber laser based on multi-lawyer antimonene [24]. Over this repetition rate range, the laser pulse width decreased while pulse energy increased. This could be explained by the fact that with the increased pump repetition rate, more residual population were left on the 4F9/2 level before the arrival of the next pump pulse, thus more population were accumulated after it came, resulting in larger pulse energy and shorter pulse width. This behavior has been observed in gain-switched Er-doped ZBLAN fiber lasers at 2.8 µm [17,20,25]. Moreover, the decreased seed pulse width with the repetition rate significantly contributed to the decreased laser pulse width as well due to faster population buildup. Over the same range, both the average and peak power increased. At the maximum repetition rate of 120 kHz, the pulse width, pulse energy, average power, and peak power were 1.77 μs, 8.33 μJ, 1 W, and 4.71 W, respectively. Figure 4(b) shows the pulse trains at different pump repetitions rates of 55 kHz, 90 kHz, and 120 kHz recorded at the maximum pump powers before the appearance of the second relaxation spike. The maximum pulse repetition rate of 120 kHz has exceeded the theoretically predicted limit of 100 kHz at the 976 nm pump power of 5 W [22] and was only limited by our used seed repetition rate. Compared to the previous report on active Q-switching at this waveband in a similar DWP system which had a tunable repetition rate range of 5~100 kHz [14], although the tunable range of 55~120 kHz in our case was narrower, further extension was expected by scaling the pump repetition rate range since the gain-switched pulses were still quite stable especially at the low repetition rate edge.
Fig. 4 (a) Seed and laser pulse widths, average power, pulse energy, and peak power versus pump repetition rate (E1950 nm = 50 μJ), (b) pulse trains at different pump repetition rates of 55 kHz, 90 kHz, and 120 kHz.
4. Discussion
In order to experimentally identify the upper limit of repetition rate, a NPR passively mode-locked Tm-doped fiber laser at 1960 nm under dissipative soliton resonance (DSR) condition was constructed to replace the Q-switched seed. It had a similar structure to that in [26] and could produce repetition rate of even MHz by varying the SMF28e fiber length in the cavity. First, the pump repetition rate was set at 206 kHz with a maximum pulse width of 30 ns. We found no gain-switched pulses were built up at the whole available pulse pump power range. Then the repetition rate was further decreased to 116 kHz which enabled gain-switching for the case of Q-switching. Consequently, gain-switching was not yet achieved as shown in Fig. 5. The phenomenon suggested that gain-switching formation was closely related to the pump pulse width. Particularly, larger pump pulse width may contribute to higher upper limit of repetition rate, which should be correlated with the laser quenching behavior. A possible explanation is that under narrow pulse pumping, it is easier for the ions excited from 4I11/2 to 4F9/2 level to participate in the ESA between 4F9/2 and 4F7/2 levels, thus more time was needed for these ions to go back to the 4I11/2 level corresponding to lower allowable pump repetition rate [22]. Although gain-switching has been achieved in a DWP Er-doped ZBLAN fiber laser pumped by Q-switched pulses with a width of 40~50 ns [21], it could only operate at a lower repetition rate of 15~20 kHz, matching with the explanation proposed above. This indicated that the buildup of gain-switching in our case was expected if further reducing the repetition rate of DSR pulses. Unfortunately, it was difficult to further reduce the repetition rate due to the appearance of harmonic regime under stronger intra-cavity nonlinearity. Compared to that (i.e., 5 m) [21], although the used gain fiber in our case was shorter, it had a 1.5 times larger dopant concentration which could provide almost same gain. Therefore, the cavity length may be not a significant factor determining whether gain-switching could be formed since the limit induced by ESA was directly related to the pump pulses and ions on the 4I11/2 level. Numerical simulations are still needed to confirm this hypothesis in the future. To improve the gain-switching repetition rate, a modification of the passively Q-switched seed is needed by reducing the modulation depth of the SA to improve its repetition rate while adjusting the cavity length to control the pulse width. A thorough study on the relationship between pump pulse width and limit of repetition rate is also necessary based on a modified numerical model according to the experimental results. In our case, further power scaling was prevented by the appearance of second relaxation spike, which is a common ending of gain-switching with the increased pump. Generally, high energy and repetition rate narrow pump pulses are favorable for high power owing to fast pulse buildup, but not for this gain-switched DWP system due to the negative ESA between 4F9/2 and 4F7/2 levels. Thus, numerical optimization is demanded for further scaling both average and peak power.
Fig. 5 (a) DSR pulse waveform and (b) pulse trains with a repetition rate of 116 kHz at the launched 1960 nm pump power of 6.73 W.
5. Conclusion
In summary, we demonstrate a gain-switched Er-doped ZBLAN fiber laser at 3.46 μm where a DWP scheme consisting of a 1950 nm home-made Bi2O2Se passively Q-switched fiber laser system and a 976 nm CW LD was used. The achieved maximum average power of 1.04 W (100kHz) at a slope efficiency of 24.6% with respect to the launched 1950 nm pulse laser was the record of a pulsed rare-earth-doped fiber laser beyond 3 μm, to the best of our knwoledge. No laser quenching behavior was observed. The corresponding pusle energy, peak power, and pusle duration were 10.4 μJ, 1.61 μs, and 6.46 W, respectively. Synchronous gain-switched pulses with a tunable repetition rate at the range of 55~120 kHz were obtained by controlling the 1950 nm pulsed laser. The maximum repetition rate of 120 kHz was also the highest from gain-switching at this band and only limited by the 1950 nm pulsed pump source. At a fixed pump pulse energy of 50 μJ, the relationship between gain-switching output performance and pump repetition rate was studied. The results implied that higher repetition rate was benifical for narrower pulse width, higher pulse energy and peak power. Besides, it was found that the narrower pump pulse may allow a lower limited pump repetition rate for the DWP system due to the possible laser quenching behvior during a single pump pusle. In the future, a more thorough numericial model would be built up to reaveal the relationship between them while optimizing the pump parameters for further power scaling.
Funding
National Natural Science Foundation of China (NSFC) (Grant No. 61435003, 61722503, and 61421002), Open Fund of Science and Technology on Solid-State Laser Laboratory, Fundamental Research Funds for the Central Universities (Grant No. ZYGX2016J068), International Scientific Cooperation Project of Sichuan Province (Grant No. 2017HH0046), Joint Fund of Ministry of Education for Equipment Pre-research (Grant No. 6141A02033411).
Acknowledgments
H. Y. Luo, J. Yang, and F. Liu contributed equally to this work. H. Y. Luo and J. F. Li designed the experiment, processed the data, and then prepared the manuscript. H. Y. Luo and J. Yang built up the 3.5 μm gain-switched fiber laser. F. Liu built up the DSR seed and two-stage amplifier. Z. Hu built up the passively Q-switched seed. X. Yao and F. Yan helped H. Y. Luo to finish the measurement. J. F. Li, H. L. Peng and F. Ouellette discussed the results with H. Y. Luo. Y. L. supervised the project.
References
1. B. G. Lee, M. A. Belkin, R. Audet, J. MacArthur, L. Diehl, C. Pflügl, F. Capasso, D. C. Oakley, D. Chapman, A. Napoleone, D. Bour, S. Corzine, G. Höfler, and J. Faist, “Widely tunable single-mode quantum cascade laser source for mid-infrared spectroscopy,” Appl. Phys. Lett. 91(23), 231101 (2007). [CrossRef]
2. H. H. P. Th. Bekman, J. C. van den Heuvel, F. J. M. van Putten, and H. M. A. Schleijpen, “Development of a mid-infrared laser for study of infrared countermeasures techniques,” Proc. SPIE 5615, 27–38 (2004). [CrossRef]
3. D. Halmer, S. Thelen, P. Hering, and M. Mürtz, “Online monitoring of ethane traces in exhaled breath with a difference frequency generation spectrometer,” Appl. Phys. B 85(2–3), 437–443 (2006). [CrossRef]
4. J. J. Scherer, J. B. Paul, H. J. Jost, and M. L. Fischer, “Mid-IR difference frequency laser-based sensors for ambient CH4, CO, and N2O monitoring,” Appl. Phys. B 110(2), 271–277 (2013). [CrossRef]
5. B. M. Walsh, H. R. Lee, and N. P. Barnes, “Mid infrared lasers for remote sensing applications,” J. Lumin. 169, 400–405 (2016). [CrossRef]
6. C. Frayssinous, V. Fortin, J.-P. Bérubé, A. Fraser, and R. Vallée, “Resonant polymer ablation using a compact 3.44 μm fiber laser,” J. Mater. Process. Technol. 252, 813–820 (2018). [CrossRef]
7. O. Henderson-Sapir, J. Munch, and D. J. Ottaway, “Mid-infrared fiber lasers at and beyond 3.5 μm using dual-wavelength pumping,” Opt. Lett. 39(3), 493–496 (2014). [CrossRef] [PubMed]
8. F. Maes, V. Fortin, M. Bernier, and R. Vallée, “5.6 W monolithic fiber laser at 3.55 μm,” Opt. Lett. 42(11), 2054–2057 (2017). [CrossRef] [PubMed]
9. O. Henderson-Sapir, S. D. Jackson, and D. J. Ottaway, “Versatile and widely tunable mid-infrared erbium doped ZBLAN fiber laser,” Opt. Lett. 41(7), 1676–1679 (2016). [CrossRef] [PubMed]
10. O. Henderson-Sapir, J. Munch, and D. J. Ottaway, “New energy-transfer upconversion process in Er3+:ZBLAN mid-infrared fiber lasers,” Opt. Express 24(7), 6869–6883 (2016). [CrossRef] [PubMed]
11. A. Malouf, O. Henderson-Sapir, M. Gorjan, and D. J. Ottaway, “Numerical modeling of 3.5 μm dual wavelength pumped erbium-doped mid-infrared fiber lasers,” IEEE J. Quantum Electron. 52(11), 1600412 (2017).
12. F. Maes, V. Fortin, M. Bernier, and R. Vallée, “Quenching of 3.4 μm dual-wavelength pumped Erbium doped fiber lasers,” IEEE J. Quantum Electron. 53(2), 1 (2017). [CrossRef]
13. N. Bawden, H. Matsukuma, O. Henderson-Sapir, E. Klantsataya, S. Tokita, and D. J. Ottaway, “Q-switched dual-wavelength pumped 3.5-μm erbium-doped mid-infrared fiber laser,” Proc. SPIE 10512, 1 (2018).
14. N. Bawden, H. Matsukuma, O. Henderson-Sapir, E. Klantsataya, S. Tokita, and D. J. Ottaway, “Actively Q-switched dual-wavelength pumped Er3+ :ZBLAN fiber laser at 3.47 µm,” Opt. Lett. 43(11), 2724–2727 (2018). [CrossRef] [PubMed]
15. Z. Qin, T. Hai, G. Xie, J. Ma, P. Yuan, L. Qian, L. Li, L. Zhao, and D. Shen, “Black phosphorus Q-switched and mode-locked mid-infrared Er:ZBLAN fiber laser at 3.5 μm wavelength,” Opt. Express 26(7), 8224–8231 (2018). [CrossRef] [PubMed]
16. M. Gorjan, R. Petkovšek, M. Marinček, and M. Čopič, “High-power pulsed diode-pumped Er:ZBLAN fiber laser,” Opt. Lett. 36(10), 1923–1925 (2011). [CrossRef] [PubMed]
17. C. Wei, H. Luo, H. Shi, Y. Lyu, H. Zhang, and Y. Liu, “Widely wavelength tunable gain-switched Er^3+-doped ZBLAN fiber laser around 28 μm,” Opt. Express 25(8), 8816–8827 (2017). [CrossRef] [PubMed]
18. H. Luo, J. Li, C. Zhu, X. Lai, Y. Hai, and Y. Liu, “Cascaded gain-switching in the mid-infrared region,” Sci. Rep. 7(1), 16891 (2017). [CrossRef] [PubMed]
19. H. Luo, J. Li, Y. Hai, X. Lai, and Y. Liu, “State-switchable and wavelength-tunable gain-switched mid-infrared fiber laser in the wavelength region around 2.94 μm,” Opt. Express 26(1), 63–79 (2018). [CrossRef] [PubMed]
20. P. Paradis, V. Fortin, Y. O. Aydin, R. Vallée, and M. Bernier, “10 W-level gain-switched all-fiber laser at 2.8 μm,” Opt. Lett. 43(13), 3196–3199 (2018). [CrossRef] [PubMed]
21. F. Jobin, V. Fortin, F. Maes, M. Bernier, and R. Vallée, “Gain-switched fiber laser at 3.55 μm,” Opt. Lett. 43(8), 1770–1773 (2018). [CrossRef] [PubMed]
22. J. L. Yang, H. Z. Zhong, S. Y. Zhang, Y. L. Tang, and D. Y. Fan, “Cascade-gain-switching for generating 3.5-μm nanosecond pulses from monolithic fiber lasers,” IEEE Photonics J. 10(5), 1 (2018). [CrossRef]
23. Z. Q. Zhipeng Qin, G. X. Guoqiang Xie, J. M. Jingui Ma, P. Y. Peng Yuan, and L. Q. Liejia Qian, “Mid-infrared Er:ZBLAN fiber laser reaching 3.68 μm wavelength,” Chin. Opt. Lett. 15(11), 111402 (2017). [CrossRef]
24. H. Y. Luo, X. L. Tian, Y. Gao, R. F. Wei, J. F. Li, J. R. Qiu, and Y. Liu, “Antimonene: a long-term stable two-dimensional saturable absorption material under ambient conditions for the mid-infrared spectral region,” Photon. Res. 6(9), 900–907 (2018). [CrossRef]
25. 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).
26. X. Wu, D. Y. Tang, H. Zhang, and L. M. Zhao, “Dissipative soliton resonance in an all-normal-dispersion erbium-doped fiber laser,” Opt. Express 17(7), 5580–5584 (2009). [CrossRef] [PubMed]
