A wavelength-spacing controllable, dual-wavelength synchronously mode locked Er:fiber laser oscillator based on dual-branch nonlinear polarization rotation (NPR) technique was presented. The center wavelengths were at 1542 nm and 1561 nm, which had pulse durations of 1.38 ps and 1.70 ps, respectively. Experimentally, the synchronous mode locking was achieved by precisely adjusting the cavity length of one branch. A tolerance in the cavity length mismatch of 0.46 mm for synchronous mode locking was demonstrated. The frequency difference of the two pulse trains was measured to be less than 1 mHz. Additionally, this synchronously mode locked dual-wavelength laser had a wavelength tunable range of about 5.6 nm, and a controllable wavelength spacing from 10.5 nm to 28.2 nm, corresponding to a tunable frequency difference from 1.32 THz to 3.26 THz. To the best of our knowledge, this is the first demonstration of synchronously mode locked dual-wavelength output directly from a Er:fiber laser oscillator, using dual-branch NPR technique.
© 2016 Optical Society of America
Dual-wavelength synchronously mode locked lasers and dual-wavelength synchronized two lasers have been widely used in terahertz (THz) beating generation , coherent anti-Stokes Raman scattering (CARS) microscopy , Stimulated Raman scattering (SRS) microscopy , sum-frequency generation (SFG) , difference-frequency generation (DFG) in mid-infrared (MIR) , and coincidence single-photon frequency upconversion .
To date, dual-wavelength synchronously mode locked lasers have been generated in different kinds of bulk solid state laser oscillators using different rare earth ion doped crystals, including Ti:sapphire crystal based 0.8 μm laser , Nd-doped  or Yb-doped  crystal based 1 μm laser, and Tm-doped crystal based 2 μm laser . They are either based on employing double slit  or benefitting from the modulation of the semiconductor saturable absorber mirror (SESAM) [8–10]. The two wavelengths lasing are from the same gain medium, which makes the wavelength spacing small enough for THz wave generation using DFG scheme. However, it is not easy to tune the wavelength spacing in these lasers, because of the well-defined energy-level in the gain medium. Alternatively, one can also actively or passively synchronize two different independent mode locked lasers, which could be two bulk solid state lasers [4, 5, 11], one bulk solid state laser and one fiber laser [12, 13], and two fiber lasers [6, 14, 15]. Usually, because the two involved lasers are based on different gain mediums, the wavelengths spacing could be much larger to a few hundreds of nanometers (264 nm in [4, 5], 430 nm in , 520 nm in , 300 nm in , ~200 nm in , ~500 nm in , and 380 nm in ), which makes them suitable for MIR wave generation using DFG scheme.
It is well known that fiber laser is much more compact and cost-effective than its bulk solid state laser counterpart. However, dual-wavelength synchronously mode locked fiber lasers are seldom reported. So far, although there already have been many reports on mode locked dual-wavelength generation directly from Nd-doped , Er-doped [17, 18], Bi-doped , Yb-doped [20, 21], and Tm-doped  fiber lasers, the involved two wavelengths are not synchronous but with a very small frequency difference, which makes them unsuitable for applications that require tight timing synchronization such as DFG. It was from 500 Hz to several megahertz in , 470 Hz in , 108 Hz in , 4.1 kHz in , and 6 kHz in . If the used measurement device doesn’t have high enough resolution, the small frequency difference cannot be resolved. Actually, this small frequency difference in mode locked dual-wavelength fiber lasers is from the non-perfect dispersion compensation inside the cavity. According to , the frequency difference of the two wavelengths could be expressed as Δf = c2·D·dλ/[n2·(L + L·D·dλ·c/n)], where c is the speed of light in vacuum, D is the average dispersion parameter in the cavity, dλ is wavelength spacing of the propagated two wavelengths, n is the refractive index of the fiber, and L is the cavity length. It tells us that, if L is the same for the two involved different wavelengths, the only possibility to synchronize them is to make the dispersion inside the cavity be exactly zero, which is almost impossible in practical conditions. Another easier way to make Δf be zero is to adjust the cavity lengths for different wavelengths to compensate for the optical path difference (OPD) that is from the refractive index difference of the two wavelengths, which is what we used in the experiments reported in this contribution.
As far as we know, there were two papers that did report synchronous mode locked dual-wavelength fiber lasers [23, 24]. In , a piece of Yb-doped multicore fibers, where different cores amplified different spectra bands, was used to achieve synchronous dual-wavelength ultrashort pulses with a wavelength spacing of 2.56 nm. And SESAM was used. However, the multicore fiber is not commercially available yet. In , black phosphorus (BP) was used to get a tri-wavelength synchronously mode-locked operation in Er:fiber laser. The author claimed that the laser was synchronously mode-locked, but there was no direct measurement of the repetition rate. However, inconspicuous modulations were observed on the autocorrelation trace, with a beat frequency of 0.06 THz, by which the author believed that the three wavelengths were in a synchronous mode-locked state. Therefore, in summary, there is still no report on synchronously mode-locked dual-wavelength fiber lasers based on NPR technique, which is different from the above mentioned real saturable absorber device like SESAM and BP.
In this contribution, a dual-branch NPR based Er:fiber laser oscillator, which could deliver wavelength-switchable, -tunable, and synchronously mode-locked dual-wavelength laser pulses at 1.5 μm band, was demonstrated. Cross phase modulation (XPM) effect, was employed to obtain the synchronous dual-wavelength running. By that, two pulse trains, propagating at different group velocities initially, will trap each other. At the same time, the center wavelengths of each pulse will be shifted to make the group velocity same, while the spectrum doesn’t have to be overlapped . In our case, two ring cavities sharing a piece of common Er-doped gain fiber were configured by introducing a 50:50 fiber coupler. Dual-wavelength mode locking could be easily achieved once the two lasers are synchronized. The tolerated cavity length difference was measured to be 0.46 mm. Within that region, the frequency difference of the two wavelengths was measured to be less than 1 mHz. The switch between synchronous dual-wavelength mode locking and single wavelength mode locking can be realized by rotating a half-waveplate (HWP) inside the laser cavity. In addition, this synchronous dual-wavelength mode locked laser has a wavelength-tunable range of about 5.6 nm. The wavelength spacing of two wavelengths can be varied from ~10 nm to ~26 nm, which corresponds to a frequency difference from 1.32 THz to 3.26 THz. All of these performances ensure the oscillator’s diversity. To the best of our knowledge, this is the first demonstration of synchronously mode locked dual-wavelength output directly from a Er:fiber laser oscillator, using dual-branch NPR technique.
2. Experimental setup
Figure 1 shows the experimental setup of the wavelength-switchable and -tunable dual-wavelength mode locked Er:fiber laser oscillator. It was composed of two laser cavities, which shared a piece of common gain fiber and an ejection port at polarization beam splitter PBS1. A piece of 75 cm long Er-doped single mode fiber (Thorlabs ER80-8/125) was used as the gain medium, which was pumped by a wavelength stabilized 976 nm single mode fiber coupled laser diode (LD) with a maximum output power of 450 mW. The pump light from the LD was coupled into the cavity by a wavelength-division multiplexer (WDM). Polarization control was achieved through the use of the bulk zero-order waveplates. An isolator (ISO) was used to enforce unidirectional operation. A diffraction grating with a groove density of 600 lines/mm was used to disperse the spectrum spatially. A HWP W1 and a PBS2 were used to divide the beam to two parts and control the power sent to collimator 2 (Col2) and collimator 3 (Col3). The effectively received spectrum bandwidth by Col2 or Col3 was determined by the distance between the collimator and the grating, while the center wavelength was determined by the transverse position (x-position) of the collimator. Col3 was mounted on a translation stage, and could be moved longitudinally (z-position), in order to precisely adjust the cavity length. Although this would influence the bandwidth that could be accepted by Col3, remarkable bandwidth reduction is not expected because of the fact that the movement of Col3 along the z-direction (~0.5 mm) would be very small compared with the distance between grating and Col3 (~120 mm). A 50:50 fiber coupler was used to combine these two branches. The output was taken directly from the NPE ejection port of PBS1. The total dispersion of the laser cavity is calculated as −0.11 ps2, while the dispersion of the shared cavity is calculated as −0.084 ps2.
3. Experimental results
3.1 Switchable single wavelength mode locking
We firstly demonstrated the switchable single wavelength mode locking operation. Two cases are shown as examples here in Fig. 2. Lasing at other wavelengths could also be obtained by changing the x-positions of Col2 and Col3. In Fig. 2(a), continuous wave (CW) laser emission at two wavelengths of 1541 nm and 1555.4 nm is shown. By rotating HWP W1, we could change the operation condition from dual-wavelength CW oscillation to single wavelength mode locked lasing. At a pump power of 350 mW, the output powers were measured to be 52.3 mW, 63.6 mW, and 63.4 mW, when the oscillator was in dual-wavelength CW operation, mode locked at 1541 nm, and mode locked at 1555.4 nm. In Fig. 2(b), switchable single wavelength mode locking operation at 1542.5 nm and 1561.2 nm is also shown, for which the experimental setup was the same as before, except for that the x-position of the two collimators were changed. As we mentioned above, these collimators worked as slits. Therefore, different wavelengths enter the collimators when the collimators are moved transversely.
3.2 Switchable and synchronous dual-wavelength mode locking
In order to obtain a switchable and synchronous dual-wavelength mode locked laser, the optical cavity lengths for both branches should be kept the same. In our case, the position of Col2 was fixed. By axially translating Col3, we could adjust the cavity length difference very precisely. The repetition rates of laser pulses from both branches were monitored using frequency counters (Agilent 53230A). Sharing the same gain fiber, within which XPM happens, helps synchronizing these two mode locked lasers [11, 12]. When the repetition rates of two branches were close enough to each other, one wavelength started to be captured by the other, then the frequency remained equal to each other. At a pump power of around 350 mW, synchronous dual-wavelength mode locking was achieved, and the output power was measured to be 43.3 mW. Figure 3(a) shows the spectrum of the dual-wavelength mode locked laser. Two side bands at 1525.6 nm and 1570.8 nm appeared, which is attributed to the four wave mixing (FWM) effect. The switchability of working condition could be realized by rotating HWP W1. Figure 3(b) shows the mode locked spectrum at 1540 nm and 1555 nm respectively. The output power is measured to be 60.5 mW at 1540 nm and 63.9 mW at 1555 nm respectively.
The bandwidths of the mode locked spectrum are both around 3 nm in FWHM. The beams were dispersed outside the cavity using a reflective-type grating and guided onto an infrared detection card. As shown in Fig. 4, two laser spots are clearly observed about half meter away from the grating. From Fig. 4, we could tell that there is essentially no wavelength overlap.
Figure 5 shows another case of dual-wavelength mode locking at around 1542.5 nm and 1561.3 nm. The change of peak wavelength was caused by the x-position’s change of collimators Col2 and Col3. From Fig. 5, it is clear that by rotating HWP W1, the output power ratio of each wavelength could be controlled, which could be used to manage the modulation depth of the generated THz beat. As the output power ratio is equal to unity, a high quality quasi-periodic optically beat pulse train could be generated with the modulation depth of 100% .
The pulse duration was measured by an intensity auto-correlator (Alnair labs HAC-200). The results are shown in Fig. 6. The pulse durations were measured to be 1.38 ps and 1.70 ps, when the laser was mode locked at 1542.5 nm and 1561.3 nm respectively. The time bandwidth product is calculated as about 0.522 and 0.628 respectively, which could be attributed to the not optimized GDD in the laser cavity. Figure 6(c) shows the autocorrelation trace of the dual-wavelength synchronously mode locked laser. As expected, clear temporal modulation resulted from the beating of the two wavelengths appears. The beating has a period of 0.45 ps, which corresponds to 2.2-THz separation frequency domain. The clear fringes guarantee the tight synchronization with the timing jitter of shorter than the optical cycle. We numerically simulated the beating between these two pulses with different center wavelengths. The interference pattern could be calculated by , where I1 and I2 are the intensity of each pulses, is the frequency difference of the two pulses. The modulation depth is dependent on the intensity ratio and the pulse width ratio of the two pulses. The closer the two wavelengths’ intensity and pulse width are, the deeper the modulation is. We used the measured pulse shape function to simulate the beating pattern, and the simulated result accords well with the measured one.
The mode locked laser beam was directed to a photodiode, whose output was recorded by an oscilloscope (Agilent DSO1024A). The pulse train is shown in Fig. 7(a), and the repetition rate was measured to be 39.1 MHz. The mismatch tolerance of the cavity length between two branches is an important indicator to show how stable the synchronization is. Experimentally, we measured the synchronization range for synchronous dual-wavelength mode locking by translating the collimator Col3. A frequency counter (Agilent 53230A) was used to measure the pulse repetition rates for both wavelengths. The Col3′ z-positions dependent frequency differences were calculated by subtracting these two frequencies, and the results are shown in Fig. 7(b). The laser can maintain self-started synchronous dual-wavelength mode locking from z = 7.48 mm to z = 7.94 mm at the maximum 450 mW pump power, corresponding to a tolerance in the cavity length mismatch of 0.46 mm. We noticed that as long as the laser was frequency locked, no matter how Col3 was changed, the repetition rate of 1561 nm pulse train, equals to that of the 1543 nm pulse train within an accuracy of 10−3 Hz, which is from a frequency counter’s resolution. However when the cavity length mismatch exceeded the locking limit, the synchronous dual-wavelength running ceased and only one wavelength could be mode locked. To further increase the synchronization range, introducing dispersion compensation fiber into the cavity or increasing the fiber length to strengthen the nonlinear effect, could be considered in the future.
In order to further investigate how the mode locking evolved, the laser spectra were taken and recorded at different positions of Col3. The experimental results are shown in Fig. 8. It is found that when Col3 was moving away from W6, the peak of shorter wavelength was shifted from 1540.7 nm to 1545.7 nm. While the peak of longer wavelength was shifted from 1559.4 nm to 1563.2 nm. When two physical cavity length difference changes, the center wavelength separation shifts to compensate for the round-trip time difference. This long tolerance of 0.46 mm is coming from large group delay dispersions in one-round trip due to the long fiber length. When the cavity length mismatch exceeded the locking limit, the synchronous dual-wavelength running ceased and only one wavelength could be mode locked. That is because, when 1561 nm and 1543 nm pulse overlap temporally, the spectra of both wavelengths shifted owing to the strong XPM effect. In this way, the pulses’ group velocity could be changed accordingly, driving the repetition rate to match with each other to keep synchronous dual-wavelength running condition. The CW breakthrough shown in Fig. 8 is due to the high pump power of 450 mW, which was needed to maintain the whole range dual-wavelength mode locking.
3.3 Tunable dual-wavelength mode locking
The tunablity of this dual-wavelength includes two aspects. Firstly, the peak wavelength of dual-wavelength can be tuned by changing the tilting angle of the grating. Dual-wavelength mode locking can be maintained for about 5.6 nm without touching any components inside the cavity other than the grating, as shown in Fig. 9. The mode locking will be lost if we further tune the grating. The CW breakthrough is due to the high pump power at 450 mW. Hence logarithmic coordinate is used in Fig. 9.
Secondly, from previous experiments shown in Fig. 2, we expect that the spacing of the two wavelength peaks could be tuned by moving the collimators Col2 and Col3 transversely. In order to verify our expectation, we kept the position of Col3 unchanged, while moving the Col2 transversely. In this case, the longer wavelength could keep at around 1559 nm, when the shorter wavelength peak changed from 1548.4 nm to 1533.1 nm. The separation spacing of two wavelengths peak could be easily varied from 10.5 nm to around 28.2 nm. The experimental results are shown in Fig. 10. It is obvious that the smaller the shorter wavelength is, the wider the obtained lasing bandwidth is. This is because the diffraction angle is smaller for shorter wavelength. From the grating equation, we have dλ/dβ = acosβ/m, where λ is the incident laser wavelength, β is the diffraction angle, a is the blaze spacing of the grating and m is the diffraction order. From this equation, we can conclude that smaller β results in larger dλ/dβ, which means broader spectrum is received for a same slit, i.e. broader lasing bandwidth could be achieved. Additionally, the broader shorter wavelength will result in broader longer wavelength correspondingly due to the XPM effect. From Fig. 10, we also find that there are some small modulation peaks of the spectra in the case of maximal separation. We think it is caused by slightly SPM of the laser pulse. Larger wavelength spacing results in the larger net cavity length mismatch. Only the spectrum shape can be changed by the laser itself to keep the synchronization. The larger spacing causes the spectrum wider to get shorter pulse duration for stronger SPM, which leads the multi-peak structure in the spectrum.
4. Conclusion and discussion
In this paper, we reported a novel Er:fiber laser system which can provide a switchable and tunable synchronous dual-wavelength mode locked laser with collinear output. Once the synchronization is realized, the laser working condition could be switched from dual-wavelength mode locking to single wavelength mode locking by simply rotating the HWP W1 inside the cavity. Coupling achieved by XPM in the fiber laser results in a large cavity mismatch tolerance of 0.46 mm for our laser system. The pulse durations of two wavelengths are measured to be less than 2 ps. The achieved pulse duration is limited by the accepted spectrum from Col2 and Col3. The pulse duration could be reduced if the distance from the grating to collimators is shortened by utilizing smaller mechanical mounts. Furthermore, this synchronous dual-wavelength laser has a wavelength-tunable range of about 5.6 nm and a variable wavelength spacing from 10.5 nm to 28.2 nm. The relatively poor tunability compared with single wavelength running situation is due to the XPM changing when the wavelengths changes. The tuning range could be enlarged by compensating for the XPM by translating Col2 and Col3 longitudinally in real time. The upper limit of the wavelength spacing is determined by the transverse position of Col2 and Col3, if enough pump power could be offered. The wavelength spacing could be enlarged by transversing Col2 to one side of the spectrum while putting Col3 to the other side of the spectrum. The largest wavelength spacing should be able to be comparable with the lasing wavelength range of the laser. The lower limit of the wavelength spacing is due to the bandwidth of the two wavelengths. As long as there is no spectrum overlapping of the two lasing wavelengths, the synchronous dual-wavelength running should work. We could also increase the whole laser efficiency and the tuning range by changing the 50:50 coupler from an off the shelf 2 × 2 one to a customized 2 × 1 one. In terms of long term stability, although we didn’t measure it experimentally, we observed at least for two hours passive synchronization without any box cover. We believe that longer synchronization could be obtained with a cover and better temperature control of the cavity. This novel laser design could be used in other rare earth doped fibers and working in the other wavelengths. It will be a promising instrument for many applications.
National Natural Science Foundation of China (Grant No. 61605133). Sichuan Province International Cooperation Research Program, China (2016 HH0033). Photon Frontier Network Program and Photon and Quantum Basic Research Coordinated Development Program from MEXT, Japan.
References and links
1. Y. J. Huang, Y. S. Tzeng, C. Y. Tang, S. Y. Chiang, H. C. Liang, and Y. F. Chen, “Efficient high-power terahertz beating in a dual-wavelength synchronously mode-locked laser with dual gain media,” Opt. Lett. 39(6), 1477–1480 (2014). [CrossRef] [PubMed]
2. D. J. Jones, E. O. Potma, J. X. Cheng, B. Burfeindt, Y. Pang, J. Ye, and X. S. Xie, “Synchronization of two passively mode-locked, picosecond lasers within 20 fs for coherent anti-Stokes Raman scattering microscopy,” Rev. Sci. Instrum. 73(8), 2843–2848 (2002). [CrossRef]
3. C. W. Freudiger, W. Yang, G. R. Holtom, N. Peyghambarian, X. S. Xie, and K. Q. Kieu, “Stimulated Raman scattering microscopy with a robust fibre laser source,” Nat. Photonics 8(2), 153–159 (2014). [CrossRef] [PubMed]
4. H. W. Xuan, N. Wang, Y. D. Zhang, Z. H. Wang, and Z. Y. Wei, “A Tunable ultrafast source by Sum-Frequency generation between two actively synchronized ultrafast lasers,” Chin. Phys. Lett. 8(6), 064207 (2012). [CrossRef]
5. H. W. Xuan, Y. Zou, S. Wang, H. Han, Z. H. Wang, and Z. Y. Wei, “Generation of ultrafast mid-infrared laser by DFG between two actively synchronized picosecond lasers in a MgO:PPLN crystal,” Appl. Phys. B 108(3), 571–575 (2012). [CrossRef]
6. K. Huang, X. R. Gu, H. F. Pan, E. Wu, and H. P. Zeng, “Synchronized fiber lasers for efficient coincidence single-photon frequency upconversion,” IEEE J. Sel. Top. Quantum Electron. 18(2), 562–566 (2012). [CrossRef]
8. G. Q. Xie, D. Y. Tang, H. Luo, H. J. Zhang, H. H. Yu, J. Y. Wang, X. T. Tao, M. H. Jiang, and L. J. Qian, “Dual-wavelength synchronously mode-locked Nd:CNGG laser,” Opt. Lett. 33(16), 1872–1874 (2008). [CrossRef] [PubMed]
9. W. Tian, Z. Wang, J. Liu, J. Zhu, L. Zheng, X. Xu, J. Xu, and Z. Wei, “Dissipative soliton and synchronously dual-wavelength mode-locking Yb:YSO lasers,” Opt. Express 23(7), 8731–8739 (2015). [CrossRef] [PubMed]
10. L. C. Kong, Z. P. Qin, G. Q. Xie, X. D. Xu, J. Xu, P. Yuan, and L. J. Qian, “Dual-wavelength synchronous operation of a mode-locked 2-μm Tm:CaYAlO4 laser,” Opt. Lett. 40(3), 356–358 (2015). [CrossRef] [PubMed]
11. Z. Wei, Y. Kobayashi, Z. Zhang, and K. Torizuka, “Generation of two-color femtosecond pulses by self-synchronizing Ti:sapphire and Cr:forsterite lasers,” Opt. Lett. 26(22), 1806–1808 (2001). [CrossRef] [PubMed]
12. D. Yoshitomi, Y. Kobayashi, M. Kakehata, H. Takada, K. Torizuka, T. Onuma, H. Yokoi, T. Sekiguchi, and S. Nakamura, “Ultralow-jitter passive timing stabilization of a mode-locked Er-doped fiber laser by injection of an optical pulse train,” Opt. Lett. 31(22), 3243–3245 (2006). [CrossRef] [PubMed]
13. D. Yoshitomi, X. Zhou, Y. Kobayashi, H. Takada, and K. Torizuka, “Long-term stable passive synchronization of 50 µJ femtosecond Yb-doped fiber chirped-pulse amplifier with a mode-locked Ti:sapphire laser,” Opt. Express 18(25), 26027–26036 (2010). [CrossRef] [PubMed]
14. M. Zhang, E. J. R. Kelleher, A. S. Pozharov, E. D. Obraztsova, S. V. Popov, and J. R. Taylor, “Passive synchronization of all-fiber lasers through a common saturable absorber,” Opt. Lett. 36(20), 3984–3986 (2011). [CrossRef] [PubMed]
15. J. Sotor, G. Sobon, J. Tarka, I. Pasternak, A. Krajewska, W. Strupinski, and K. M. Abramski, “Passive synchronization of erbium and thulium doped fiber mode-locked lasers enhanced by common graphene saturable absorber,” Opt. Express 22(5), 5536–5543 (2014). [CrossRef] [PubMed]
17. X. Zhao, Z. Zheng, L. Liu, Y. Liu, Y. Jiang, X. Yang, and J. Zhu, “Switchable, dual-wavelength passively mode-locked ultrafast fiber laser based on a single-wall carbon nanotube modelocker and intracavity loss tuning,” Opt. Express 19(2), 1168–1173 (2011). [CrossRef] [PubMed]
18. D. Mao and H. Lu, “Formation and evolution of passively mode-locked fiber soliton lasers operating in a dual-wavelength regime,” J. Opt. Soc. Am. B 29(10), 2819–2826 (2012). [CrossRef]
19. A. P. Luo, Z. C. Luo, W. C. Xu, V. Dvoyrin, V. Mashinsky, and E. Dianov, “Tunable and switchable dual-wavelength passively mode-locked Bi-doped all-fiber ring laser based on nonlinear polarization rotation,” Laser Phys. Lett. 8(8), 601–605 (2011). [CrossRef]
20. H. Lin, C. Guo, S. Ruan, J. Yang, D. Ouyang, Y. Wu, and L. Wen, “Tunable and switchable dual-wavelength dissipative soliton operation of a weak-birefringence all-normal-dispersion Yb-doped fiber laser,” IEEE Photonics J. 5(5), 1501807 (2013). [CrossRef]
21. X. L. Liu, H. S. Wang, Y. S. Wang, Z. J. Yan, and L. Zhang, “Single-polarization, dual-wavelength mode-locked Yb-doped fiber laser by a 45°-tilted fiber grating,” Laser Phys. Lett. 12(6), 065102 (2015). [CrossRef]
22. Z. Yan, X. Li, Y. Tang, P. P. Shum, X. Yu, Y. Zhang, and Q. J. Wang, “Tunable and switchable dual-wavelength Tm-doped mode-locked fiber laser by nonlinear polarization evolution,” Opt. Express 23(4), 4369–4376 (2015). [CrossRef] [PubMed]
23. P. Rigaud, V. Kermene, Ch. Simos, A. Desfarges-Berthelemot, G. Bouwmans, L. Bigot, A. Hideur, and A. Barthelemy, “Dual-wavelength synchronous ultrashort pulses from a mode-locked Yb-doped multicore fiber laser with spatially dispersed gain,” Opt. Express 23(19), 25308–25315 (2015). [CrossRef] [PubMed]
24. R. W. Zhao, J. Li, B. T. Zhang, X. W. Li, X. C. Su, Y. R. Wang, F. Lou, H. K. Zhang, and J. L. He, “Triwavelength synchronously mode-locked fiber laser based on few-layered black phosphorus,” Appl. Phys. Express 9(9), 092701 (2016). [CrossRef]
25. Y. J. Huang, Y. S. Tzeng, C. Y. Tang, and Y. F. Chen, “Efficient dual-wavelength synchronously mode-locked picosecond laser operating on the 4F3/2-4F11/2 Transition with compactly combined dual gain media,” IEEE J. Sel. Top. Quantum Electron. 21(1), 56–62 (2015). [CrossRef]