A long-term stable picosecond dissipative soliton (DS) is achieved for the first time using nonlinear polarization evolution. The environmental stabilization is performed by a Faraday mirror, which can cancel environmentally induced changes in the birefringence of the fiber. The laser cavity with all-polarization-maintaining fiber components generates DS pulses with 2.9 nJ single pulse energy and 5.9 ps pulse width. The output power test over 2 hours shows the excellent mode-locking stability of this design.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
In the past decade, dissipative soliton (DS) fiber lasers have been extensively studied owing to their high pulse energy . Despite much progress, the development of stable, reliable and cost-effective DS fiber lasers for practical uses is still a research subject of high interest. In order to achieve such goal, DS fiber lasers with all-fiber and all-polarization-maintaining (PM) structure are considered to be a practicable solution. Various mode-locking saturable absorbers (SA) have been studied to achieve the stable DS operation. Materials like semiconductor saturable absorber mirror , graphene  and carbon nanotubes  offer easy self-starting and environmentally stable mode-locking with simple oscillator design. However, DS fiber lasers based on such materials suffer from low energy damage threshold and limited lifetime. Compared to the real saturable absorber materials, additive pulse mode-locking (APM), for example, nonlinear optical loop mirror  or nonlinear amplifying loop mirror , can be implemented in the all-PM fiber cavities to produce ultrafast pulses with outstanding stability, long-term reliability and high pulse energy. However, the designs of such ultrafast fiber lasers are relatively complicated and the self-starting issue is also very challenging for the figure-of-eight fiber lasers .
Nonlinear polarization evolution (NPE) is another APM technology to produce DS pulses with high pulse energy . However, due to its working mechanism, NPE is rarely considered to be a potential candidate to achieve long-term stability. Despite the difficulties, many efforts have been made to realize NPE based ultrafast fiber laser with either all-fiber or all-PM structure. Mortag et al. reported an all-fiber NPE cavity, which can produce pulses with 3.6 nJ pulse energy and 76 fs recompressed pulse width. The laser cavity was constructed entirely from standard single-mode (SM) fibers . Kharenko et al. demonstrated another all-fiber NPE cavity, which can produce 50 nJ pulses compressible down to 250 fs. The laser cavity was built from a hybrid SM–PM configuration with 10 μm core-size large-mode-area fibers . As for the all-PM configuration, Wang et al. proposed a cross-splicing method to compensate the influence of fiber birefringence in a NPE mode-locked PM fiber ring laser. The produced DS had pulse energy of 2.1 nJ and 11.7 ps pulse width . Although the laser cavity was constructed with all-PM fibers and some bulk optical components, the environmental stability of this laser was not reported.
Some attempts were also made to combine both all-fiber and all-PM structure for NPE mode-locking. Szczepanek et al. reported for the first time a femtosecond all-PM fiber oscillator mode-locked by a NPE process in standard PM fibers . The key part of the laser cavity (NPE SA section) contained three pieces of PM fiber formed by a sequence of the special angle splicing. The laser can generate DS pulses at a 20.54 MHz repetition rate with the dechirped pulse duration around 150 fs and 0.85 nJ pulse energy. In order to compensate the group velocity mismatch (GVM) between two orthogonal polarizations, the length of the two fibers spliced with a 90° rotation must be equal within a small fraction of the fiber's beat length (in the level of sub-millimeter). Therefore, precise control of splicing length is very essential yet difficult for this laser design. The environmental stability of this laser was not reported. In order to get long-term stable pulses, special temperature control may be necessary for those PM fibers with angle splicing. Another approach to achieve all-PM fiber NPE mode-locking was demonstrated by Nielsen . The laser's environmental stabilization is achieved by a Faraday mirror (FM). Unlike the structure reported by Szczepanek, the 90° polarization rotation was performed by the FM and thus the difference between the two orthogonal polarizations was intrinsically the same. Pulse duration of 5.6 ps is obtained at a repetition rate of 5.96MHz and at an average output power of 8mW. The long-term stability of the laser was tested over a period of 200 h and no drop of the mode-locking operation was observed. However, the laser worked in noise-like mode-locked regime; no accomplishment of stable DS pulses was yet reported by this laser configuration.
In this paper, an environmentally stable DS fiber laser based on NPE technology is demonstrated. The laser cavity constructed by all-PM fiber components generates an environmentally stable, linearly polarized laser pulses. Self-starting and stable DS mode-locked operation is achieved with 2.9 nJ pulse energy and 5.9 ps pulse width. The output power test over 2 h shows the excellent mode-locking stability of this design, which, to the best of our knowledge, is the first picosecond DS demonstration of all-PM fiber NPE with over 2 h stability.
2. Experiment setup
The schematic diagram of the proposed NPE mode-locked fiber laser is illustrated in Fig. 1. The ring cavity is composed of entirely PM fibers (PM980, Nufern) and PM fiber-pigtailed components. The gain is provided by a 30cm PM ytterbium-doped fiber (PM-YDF 5/130-VIII, Nufern), which is pumped by a single-mode 976 nm laser diode (LD) through a filter wavelength-division multiplexer (WDM). A spectral bandpass filter centered at 1033 nm with an 8 nm 3dB-bandwidth is used to cut the spectrum of the chirped pulses and promote stable DS mode-locking in the all-normal dispersion regime. Approximately 30% of the beam is extracted from the cavity through a 30:70 PM fiber coupler placed after the PM gain fiber. A 1030 nm PM circulator with its fast axis blocked is placed in the cavity for three purposes. The first purpose is to act as an isolator to ensure unidirectional light propagation for the ring cavity. The second is to serve as a polarizer in the NPE structure, so that the peak of the pulse passes through the circulator while the wings of the pulse are extinguished, thus achieving pulse shortening. The last purpose is to implement the FM into the ring cavity. The FM induces 90° polarization change and reflects the beam back, thus it can cancel the environmentally induced changes in the birefringence of the fibers and compensate the GVM between two orthogonal polarizations. The splice between the port#2 of the PM circulator and the FM is made with an angle of 25° between the axes of the two fibers in order to switch the polarization state away from linear. A piece of 11 m PM fiber is placed right after the angled splice point to sever as a Kerr medium. The total cavity length is 30 m which corresponds to 0.8 ps2 of the net cavity dispersion at the central wavelength of 1033 nm.
In the experiment, the output power was measured with a power meter (Thorlabs, S145C). The optical spectra of the mode-locked laser were recorded by an optical spectrum analyzer (ANDO, model: AQ6317) having a resolution of 0.02 nm. The temporal property was measured with a 2.5 GHz oscilloscope (Keysight, DSO-S 254A) through a 1 GHz InGaAs detector. The RF spectra were recorded by a RF spectrum analyzer (Agilent, Model: N9320B) and the autocorrelation traces of the pulses were measured with a commercial autocorrelator (Femtochrome, Model: FR 103-XL).
3. Experiment results and discussion
Self-starting stable DS operation was obtained when the pump power was adjusted between 75 to 200 mW. The output power increases almost linearly with the pump power, as shown in the black curve of Fig. 2(a), which gives a slope efficiency of 12%. The maximum average output power of 19.2 mW was obtained at a pump power of 200 mW. The pulse repetition rate was measured to be at 6.7 MHz which gave the pulse energy of 2.9 nJ. When the pump power was adjust between 300 to 500 mW, the laser generated noise-like (NL) pulses with the same repetition rate. The slope efficiency of the NL pulses is also 12% with a maximum output power of 49.2 mW, corresponding to a 7.3 nJ NL pulse energy. In the pump power between 200 to 300 mW, the laser's working region was not stable, switching constantly between DS and NL. Limited by the maximum pump power of the LD, the performance of the mode-locked fiber laser was not tested beyond the pump power of 500 mW.
The spectra of the mode-locked laser under the DS and NL working regions are presented in Fig. 2(b). Steep rising and falling edges could be observed in the spectrum under the pump power of 200 mW, which is attributed to the DS feature (black curve of Fig. 2(b)). d Typical self-phase modulation (SPM) feature was also observed on the short wavelength side of the spectrum. Such SPM sidebands are induced due to the high peak power of the DS pulses. The spectrum of the NL pulses was recorded under the pump power of 500 mW. The NL spectrum has a characteristic smooth triangle-shaped spectrum peaked at 1033.7 nm and with a bandwidth of 3.3 nm, as shown in the red curve of Fig. 2(b). All the spectra here have been taken with a resolution of 0.02 nm (5.69 GHz); therefore, the absence of spectral modulation excludes the possibility of double pulses operation in the oscillator with a separation smaller than ~177 ps. Such conclusion could be further confirmed by the pulse's autocorrelation trace test, which will be presented in the following.
The RF spectra for both DS and NL pulses are measured at a resolution of 10 Hz for the fundamental frequency, as shown in Fig. 3. For the DS pulses [Fig. 3(a)], its fundamental frequency spectrum with a high signal-to-noise ratio (SNR) of 60 dB indicates high pulse-to-pulse stability. For the NL pulses [Fig. 3(b)], not only the SNR is 10 dB lower than that of the DS pulses, two side peaks also emerge, indicating the existence of amplitude modulation. The harmonic RF spectra tested by the span of 500 MHz are present in the inset of Fig. 3. No sinusoidal modulation was observed, which ensures single pulse operation in the oscillator with a separation more than 2 ns.
Figures 4(a) and 4(b) shows the DS and NL pulses' autocorrelation traces obtained directly at the laser output when the pump power is 200 and 500 mW, respectively. For the DS pulses, the positively chirped output pulses have a near-perfect Sech-function shape with a pulse duration of 5.9 ps (autocorrelation width of 9 ps). Thus the corresponding peak power of the 2.9 nJ DS pulses is 490 W. The transform-limited pulse width is calculated to be 374 fs, according to the 3 nm spectral bandwidth [Fig. 2(b)], which means the DS pulses are highly chirped. In contrast, NL pulses show a narrow sharp peak in the center of a broad pulse. Such feature is the characteristics of NL pulses. The corresponding pulse trains measured with a photodetector (1 GHz bandwidth) are also shown in Fig. 4. They both exhibit a pulse spacing of 148 ns, which matches well with the pulse round trip time of a 30 m long cavity length.
The power stability test of the DS pulses over 2 hours is shown in Fig. 4(c). A mean output power of 19.2 mW with ∼0.03 mW rms instabilities over 2 h represents a 0.1% relative rms noise with respect to the average signal power, indicating the good power and mode-locking stability of the laser. Benefiting from the all-fiber and PM configuration, this NPE mode-locking operation could be maintained by tapping the fiber and by slightly changing the environmental temperature.
4. Theoretical analysis
In order to obtain deeper insight into the working principle of the NPE SA section of our laser, making a theoretical analysis is necessary. Although NPE has already been well-studied theoretically , a simplified equivalent model for our NPE SA section is proposed as following. The light coming out from the port #2 of the PM circulator is linearly polarized. After going through the angled splicing point, the light is divided into two orthogonal beams, which travel by the fast and slow axis of the 11 m PM fiber. The intensity ratio between the two orthogonal beams is determined by the splicing angle. Therefore, the 11m PM fiber together with the FM can be equivalently considered as a nonlinear fiber loop mirror with two orthogonal beams going through opposite direction, as illustrated in Fig. 5(a). The FM can be considered as a π phase shifter insider the loop, which imposes higher loss for CW lasing. The splitting ratio α between light of fast and slow axis is defined by the splicing angle θ:
As for a nonlinear fiber loop mirror, assuming the input power is |EIN|2, the power of the reflected light is given by :
Figure 5(b) shows the calculated reflection through the NPE device as a function of normalized input power for three different splicing angles. Theoretically, the best reflection ratio (100%) is supposed to occur for a splicing angle closest to 45°. The highest reflection ratios for 20°, 25° and 30° are 41%, 59% and 75%, respectively, which means the NPE section with smaller splicing angle suffers from higher loss. However, the NPE section with large splicing angle suffers from the problem of high switching energy. Therefore, one needs to choose proper splicing angle to generate DS pulses for different cavity parameters. To illustrate the dependence of this angle on mode-locking mechanism, a numerical model based on the cubic Ginzburg–Landau equation for DS mode-locked fiber laser  is solved with simulation parameter close to our experimental settings. Figures 5(c) and 5(d) present the DS spectra with three different splicing angles experimentally and numerically. The similar spectral width and SPM spectral ripple shape indicate a good agreement between the experiment and numerical simulation. The result shows that DS pulses can be generated under a range of splicing angles; just like DS pulses can be generated under a range of splitting ratio for NOLM or NALM based mode-locked fiber lasers .
In summary, we have demonstrated an all-PM-fiber environmentally-stable DS fiber laser incorporating with a novel NPE structure. A FM was engaged inside the NPE section to cancel the environmentally induced changes in the birefringence of the fibers and compensate the GVM between two orthogonal polarizations. At the pump power of 200 mW, 19.2 mW DS pulses were ejected from the cavity, corresponding to a pulse energy of 2.9 nJ. Pulse width of 5.9 ps was measured by autocorrelation trace test. NL pulses could also be generated under higher pump power, with highest pulse energy up to 7.3 nJ. The output power test over 2 h shows the excellent mode-locking stability of this design, which, to the best of our knowledge, is the first picosecond DS demonstration of all-PM-fiber NPE with over 2 h stability. We believe that the laser design presented here has a lot of potentials and could be a good ultrafast laser source for applications including micromachining, nonlinear imaging and spectroscopy.
National Natural Science Foundation of China (61505229, 61378026 and 61575210); Youth Innovation Promotion Association, CAS.
References and links
1. P. Grelu and N. Akhmediev, “Dissipative solitons for mode-locked lasers,” Nat. Photonics 6(2), 84–92 (2012). [CrossRef]
2. L. Zhang, J. Zhou, Z. Wang, X. Gu, and Y. Feng, “SESAM Mode-Locked, Environmentally Stable, and Compact Dissipative Soliton Fiber Laser,” IEEE Photonics Technol. Lett. 26(13), 1314–1316 (2014). [CrossRef]
3. H. Zhang, D. Tang, R. J. Knize, L. Zhao, Q. Bao, and K. P. Loh, “Graphene mode locked, wavelength-tunable, dissipative soliton fiber laser,” Appl. Phys. Lett. 96(11), 111112 (2010). [CrossRef]
4. J. H. Im, S. Y. Choi, F. Rotermund, and D. I. Yeom, “All-fiber Er-doped dissipative soliton laser based on evanescent field interaction with carbon nanotube saturable absorber,” Opt. Express 18(21), 22141–22146 (2010). [CrossRef] [PubMed]
5. J. Szczepanek, T. M. Kardaś, M. Michalska, C. Radzewicz, and Y. Stepanenko, “Simple all-PM-fiber laser mode-locked with a nonlinear loop mirror,” Opt. Lett. 40(15), 3500–3503 (2015). [CrossRef] [PubMed]
6. J. Zhou and X. Gu, “32-nJ 615-fs Stable Dissipative Soliton Ring Cavity Fiber Laser With Raman Scattering,” IEEE Photonics Technol. Lett. 28(4), 453–456 (2016). [CrossRef]
7. W. Hänsel, H. Hoogland, M. Giunta, S. Schmid, T. Steinmetz, R. Doubek, P. Mayer, S. Dobner, C. Cleff, M. Fischer, and R. Holzwarth, “All polarization-maintaining fiber laser architecture for robust femtosecond pulse generation,” Appl. Phys. B 123(1), 41 (2017). [CrossRef]
8. D. S. Kharenko, O. V. Shtyrina, I. A. Yarutkina, E. V. Podivilov, M. P. Fedoruk, and S. A. Babin, “Generation and scaling of highly-chirped dissipative solitons in an Yb-doped fiber laser,” Laser Phys. Lett. 9(9), 662–668 (2012). [CrossRef]
9. D. Mortag, D. Wandt, U. Morgner, D. Kracht, and J. Neumann, “Sub-80-fs pulses from an all-fiber-integrated dissipative-soliton laser at 1 µm,” Opt. Express 19(2), 546–551 (2011). [CrossRef] [PubMed]
10. D. S. Kharenko, V. A. Gonta, and S. A. Babin, “50 nJ 250 fs all-fibre Raman-free dissipative soliton oscillator,” Laser Phys. Lett. 13(2), 025107 (2016). [CrossRef]
11. Y. Wang, L. Zhang, Z. Zhuo, and S. Guo, “Cross-splicing method for compensating fiber birefringence in polarization-maintaining fiber ring laser mode locked by nonlinear polarization evolution,” Appl. Opt. 55(21), 5766–5770 (2016). [CrossRef] [PubMed]
12. J. Szczepanek, T. M. Kardaś, C. Radzewicz, and Y. Stepanenko, “Ultrafast laser mode-locked using nonlinear polarization evolution in polarization maintaining fibers,” Opt. Lett. 42(3), 575–578 (2017). [CrossRef] [PubMed]
14. A. Komarov, K. Komarov, D. Meshcheriakov, F. Amrani, and F. Sanchez, “Polarization dynamics in nonlinear anisotropic fibers,” Phys. Rev. A 82(1), 013813 (2010). [CrossRef]
16. L. Zhang, A. R. El-Damak, Y. Feng, and X. Gu, “Experimental and numerical studies of mode-locked fiber laser with large normal and anomalous dispersion,” Opt. Express 21(10), 12014–12021 (2013). [CrossRef] [PubMed]