Open Access
Add to BibSonomy
Abstract
Mode-locked fiber lasers that generate two frequency combs with different frequency intervals could find very important applications in low-complexity dual-comb metrology. We report a partially polarization-maintaining all-fiber dual-wavelength dual-comb mode locked laser. The polarization dependent loss of the wavelength division multiplexer combined with the polarization-maintaining fibers leads to the periodic Lyot filtering effect. Traditional single wavelength mode locking can be realized with the tunable central wavelength and spectral shape. By properly setting the state of the polarization controller, stable dual-wavelength asynchronized mode locked pulse trains with the repetition rate difference of hundreds of Hertz can be achieved. In this dual-wavelength mode locked fiber laser, the dispersion of the laser cavity leads to group velocity difference of the two mode locked pulse trains with different central wavelengths, which results in the generation of dual-comb output with the frequency difference of hundreds of Hertz. The central wavelengths of the dual-wavelength mode locked fiber laser are demonstrated to be tunable in a certain range as well. Meanwhile, a novel dual-wavelength soliton molecule mode locking is also demonstrated. Such compact all-fiber dual-wavelength dual-comb mode locked laser could find various practical applications that in need of dual-comb fiber laser system.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Mode-locked fiber lasers have found various applications in many fields as a kind of ultrafast laser source and optical frequency comb [1–5]. The frequency comb consists of numerous highly coherent spectral lines with equal space in the frequency domain, which can be used as a high performance “frequency ruler” in many applications. For instance, the frequency combs have revolutionized optical frequency metrology and have great potential in molecular spectroscopy. Mode locked lasers, as a high-performing optical frequency comb, have also been proved to be very powerful tools in molecular spectroscopy, optical clocks, precision frequency/time transfer, low phase noise microwave generation, astronomical spectrograph calibration and arbitrary optical waveform generation [6–13]. In some applications, for instance in the dual-comb spectroscopy system, two high mutual-coherence frequency combs with slightly different frequency space should be involved [14,15]. The frequency difference of the two frequency combs enables real-time temporal pulse walk-off between the two pulse trains, and then asynchronous cross-sampling between them [16,17]. In the traditional dual-comb system, two independent mode locked fiber lasers with slightly different repetition rates are used as the optical frequency comb sources. In order to maintain high mutual-coherence of the two independent mode-locked fiber lasers, a complicated servo system is usually used to stabilize the carrier-envelop-offset frequency of them [18–20]. However, the servo systems adopted in those schemes are overly complex and technically challenging that limit the applications of these dual-comb frequency systems. Therefore, developing a novel dual-comb system with simple configuration, low cost and high robustness is very important to realize practical application in the non-lab environment. Many efforts to simplify the dual-comb system have been made in the past few years [21]. Generating two mutual coherence mode locked pulse trains in one laser cavity seems to be very attractive for low cost and compact dual-comb system. In this case, no complicated servo systems are required. Although the manufacturing of traditional single-comb mode locked fiber laser is mature, it remains a great challenge to make a dual-comb mode locked fiber laser. In this dual-comb mode locked fiber laser, two mode locked pulse trains share the whole or partial laser cavity, which is the key to keep them with high mutual coherence and can cancel the common mode noise. Therefore, in this kind of dual-comb mode locked lasers, no complicated servo systems are required, which greatly simplifies the system and decreases the cost. Several schemes have been proposed to realize this dual-comb mode locking such as dual-wavelength mode locking [22], bidirectional mode locking [23,24] and polarization-multiplexed mode-locking [25]. An efficient method to obtain the dual comb laser is introducing the filter effect into the laser cavity. Therefore, two mode locked pulse trains with different central wavelengths can be generated [22,26–31]. Hébert et.al. proved a self-corrected chip-based dual-comb laser and demonstrated its high mutual coherence for dual-comb spectrometer [32]. Picometer-resolution dual-comb spectroscopy and comb-line resolved radio frequency spectrum have been demonstrated based on a free running dual-wavelength passively mode-locked fiber laser with high intrinsic mutual coherence [33,34]. Liu et.al. reported a versatile multi-wavelength ultrafast mode-locked fiber laser using three chirped fiber Bragg gratings with different central wavelengths [35]. Elaborately adjusting the loss and gain of the laser cavity to obtain dual-wavelength mode locking was demonstrated as well [34]. It’s found that, in the wavelength multiplexed dual-comb mode locked fiber lasers, the repetition rate difference of the two pulse trains is determined by the laser cavity group velocity dispersion [36,37]. Meanwhile, extremely high mutual coherence of the two frequency combs can be maintained, which makes these dual-comb lasers competitive for the emerging dual-comb-based high precision metrological applications [14,38]. Moreover, all polarization maintaining structure has been pursued to realize high stability of the dual-comb mode locked fiber laser. Earlier, various efforts have been made to get it possible [39,40]. Despite their all polarization maintaining configurations, most of these all polarization maintaining dual-comb mode locked fiber lasers contain some free-space optical components, which increases the complexity of the lasers and makes them hard to be adopted beyond research laboratories.
Here we report a tunable and switchable partially polarization-maintaining all-fiber dual-wavelength dual-comb mode locked laser. The linear all-fiber laser is mode-locked by a semiconductor absorber. Lyot filtering effect, which is caused by the polarization maintaining fibers and the polarization-dependent loss of the wavelength division multiplexer, leads to dual-wavelength mode locking. Due to the dispersion of the laser cavity, the two pulse trains with different central wavelengths are asynchronous in the time domain, forming a dual-comb with the frequency difference of hundreds of Hertz. Carefully changing the state of the PC in the laser cavity can realize the tunability of the central wavelengths of the two pulse trains in a certain range. By adjusting the state of the PC further, it is found that the all-fiber mode locked laser can operate in diverse regimes. For instance, it can operate in the traditional single wavelength mode locking regime with tunable central wavelength. Meanwhile, a novel dual-wavelength soliton molecule mode locking is demonstrated. Such compact all-fiber dual-wavelength dual-comb mode locked laser could find various applications that are in need of dual-comb laser system.
2. Experimental setup
The configuration of the mode locked fiber laser is shown in Fig. 1(a). The laser is based on an all-fiber linear cavity mode-locked with a semiconductor saturable absorber mirror (SESAM). The laser cavity consists of a SESAM, a 980/1550 wavelength division multiplexer (WDM), 80 cm polarization-maintaining Erbium-doped fiber (PM-EDF, Nufern, PM-ESF-7/125), a 10/90 optical coupler (OC), a polarization controller (PC) and a broadband optical reflector (OR). A 980 nm single mode laser diode with maximum pump power of 500 mW is used to pump the PM-EDF through the WDM. It is worth noting that in the Fig. 1(a), the components on the left of the broken line are polarization-maintaining and those on the right are non-polarization-maintaining. The total polarization fiber length is ~4.6 m and the non-polarization fiber length is ~0.9 m. Thus the total length of the lase cavity is ~5.5 m, corresponding to the fundamental repetition rate of ~18.8 MHz. The PM-EDF is spliced to the WDM with a relative angle of ~100, and the PM-EDF is spliced to the OC with a relative angle of 90°, i.e. the slow axis of the PM-EDF is aligned with the fast axis of the OC. One end of the laser cavity is butt-coupler to a commercial SESAM with modulation depth of 18%, saturation fluence of 50 μJ/cm2, and a relaxation time of 2 ps. The other end of the laser is connected to a broadband optical reflector with a single mode fiber pigtail. The dispersion of the PM-EDF and the WDM are ~16 ps/nm/km and ~-7.4 ps/nm/km, respectively. The dispersion of the SESAM is −900 fs2.Therefore, the average dispersion of the fiber is ~13 ps/nm/km, and the net cavity dispersion is ~-0.185 ps2 which means the mode locked laser operates in the anomalous regime. Two output ends of the 10/90 polarization-maintaining coupler are connected to optical isolators and monitored by an optical spectral analyzer (OSA, YOKOGAWA, AQ6375B), an oscilloscope (OSC) and a radio frequency (RF) analyzer.
Fig. 1 (a) Configuration of the all-fiber linear cavity mode locked laser based on SESAM. SESAM: semiconductor saturable absorber mirror; LD: laser diode; WDM: wavelength division multiplexer; PM-EDF: polarization-maintaining Erbium-doped fiber; OC: optical coupler; PC: polarization controller; OR: optical reflector; ISO: isolator. The components on the left of the broken line are polarization-maintaining and those on the right are non-polarization-maintaining. (b) Scheme of the setup to measure the transmittance of the all-fiber linear cavity. SC: supercontinuum; OSA: optical spectral analyzer.
In order to obtain dual-wavelength mode locking, specific intracavity optical filtering effect should be established. In this laser cavity, the polarization-dependent loss of the WDM combined with the polarization maintaining fibers could lead to the Lyot filtering effect. To confirm this, we tested the transmittance of laser cavity with a home-made supercontinuum source. The home-made supercontinuum source has stable and flat spectrum ranging from ~1300 nm to ~2200 nm. The scheme of the setup to measure the transmittance of the all-fiber linear cavity is shown in Fig. 1(b). In the measurement, the fiber pigtail 1 of the circulator was connected to the SC source and the fiber pigtail 3 of the circulator was connected to the OSA. In the first step, the fiber pigtail 2 of the circulator was connected to the OR with wide bandwidth and the spectrum was recorded in the OSA as the baseline. In the second step, we removed the SESAM of the fiber laser and removed the OR connected to the fiber pigtail 2 of the circulator as well. Then the fiber pigtail 2 of the circulator was connected to the laser cavity to replace the SESAM as shown in Fig. 1. Recording the spectra again and subtracting the baseline we can obtain the transmittance spectra of the laser cavity as shown in Fig. 2. Clear periodic transmittance peaks can be observed from 1560 nm to 1640 nm. From Fig. 2, it is clear that the adjustment in the state of the PC could lead to the change of wavelength spacing of the transmittance peaks and the modulation depth of the transmittance spectra. Meanwhile, adjusting the state of the PC could lead to the shift of the transmittance peaks as well. The transmittance curve shown in Fig. 2(a) indicates periodic transmittance peaks with wavelength spacing ~5.4 nm. By changing the state of the PC, the shape of transmittance curves can be changed gradually. From Fig. 2(a) to Fig. 2(f), it can be found that the shape of each transmittance peak transforms from n-shape to m-shape gradually and each peak splits into two eventually. In Fig. 2(f), periodic transmittance peaks with wavelength spacing ~2.7 nm was obtained, which is about half of that shown in Fig. 2(a). It should be pointed out that the modulation depth of the transmittance curves is tunable as well. It can be found that the transmittance curves increase shapely around ~1570 nm and the transmittance of the laser cavity is very low below 1560 nm. Obviously, it is due to the high absorption of the PM-EDF in the laser cavity. Although the SC source has very broad bandwidth, the Erbium-doped fiber has very high absorption in the range from ~1520 nm to ~1570 nm leading to the reducing visibility of the transmittance spectra below 1570 nm. Meanwhile, the transmittance curves decrease gradually in the long wavelength region, which can be attributed to the increasing loss of the fiber components in the laser cavity. All the optical components have the central operating wavelengths around 1550 nm. The transmissions of these optical components decrease gradually when they operate far from their central operating wavelengths. Therefore, the measured transmittance spectra decrease gradually above 1640 nm. Although the measured transmittance spectra indicate the periodic filtering effect of the laser cavity only in the range from ~1560 nm to ~1650 nm, it is reasonable to conclude that periodic filtering effect exists in the whole operating wavelength range of the mode locked fiber laser, i.e. from ~1520 nm to ~1580 nm.
Fig. 2 Six typical measured transmittance spectra of the laser cavity with different states of the PC.
Actually, the periodic filtering effect can be attributed to the Lyot filtering effect of the laser cavity. The polarization-maintaining WDM has a polarization-dependent loss which acts like a polarizer. Meanwhile, the polarization-maintaining fibers in the laser cavity lead to phase difference for the light along different principle axes. These factors result in the Lyot filtering effect in the laser cavity and the period of the transmittance peaks can be written as [41,42]:
Where is the central wavelength, Bavg is the average birefringence and Leff is the effective length. In this fiber laser, the birefringence of the PM-EDF and the pigtail fiber of OC are ~3.5 × 10−4 and ~3.7 × 10−4, respectively. The birefringence of the PM-EDF was given by the Nufern. The birefringence of the OC and the WDM was measured by the Sagnac interference method. Therefore, the Bavg × Leff = 2 × 3.7 × 10−4-0.8 × 3.5 × 10−4 and Δλ = 5.56 nm at 1600 nm which is in good agreement with the measured transmittance curves of the laser cavity shown in Fig. 2. Since the laser passes through the PM-EDF and OC twice per round, by properly setting the state of the PC, the value of Bavg × Leff can be 2 × (0.8 × 3.5 × 10−4-2 × 3.8 × 10−4) and Δλ = 2.78 nm.3. Results and discussion
3.1 Single wavelength mode-locking
Mode locking can be realized easily with the pump power of ~80 mW and the fiber laser tends to operate in the multi-pulse regime. Single pulse can be obtained by decreasing the pump power appropriately. By changing the state of the PC, the central wavelength and the shape of the spectra can be changed in a certain range. This is because changing the state of the PC leads to the changing of the transmittance curve as shown in Fig. 2. As the net cavity dispersion is negative, if the modulation depth of the Lyot filter is very weak, the mode locked laser tends to operate in the traditional soliton mode-locking regime. This can be confirmed by the experiment. With the pump power of ~70 mW, single wavelength mode-locking can self-start easily. Carefully adjusting the state of the PC, traditional soliton mode locking can be realized. Three typical output spectra of the mode locked fiber operating in the soliton mode locking regime are shown in Fig. 3(a). Obvious Kelly sidebands can be observed in these spectra. Both the central wavelength and the number of the sidebands are tunable by changing the state of the PC. The pulse spectrum, as the black line shows, has 2 order Kelly sidebands at both the short and long wavelength side. The central wavelength of the soliton locates at 1560.6 nm and the 3-dB spectral width is ~5.4 nm. The ± 1 order Kelly sidebands locates at 1551.1 nm and 1569.6 nm and the ± 2 order Kelly sidebands locates at 1546.5 nm and 1574.2 nm. Meanwhile, the pulse train was monitored by an oscilloscope (300 MHz) combined with a photodetector (Thorlabs, DET01CFC,1.2 GHz). The corresponding oscilloscope trace of the pulse train is shown in Fig. 3(b), in which stable pulse train with pulse spacing of ~53 ns can be observed. The RF spectra of the laser were measured by a RF spectrum analyzer (26.5 GHz) combined with a photodetector (Newport, 45 GHz). The corresponding RF spectra are show in Fig. 3(c)-(d). The fundamental RF peak locates at ~18.8 MHz coinciding with the measured pulse spacing and the high signal to noise ratio of >70 dB indicates the high stability of the mode locking. We believe that the slight noise shown in Fig. 3(c) mainly results from the relatively long relaxation time of the SESAM. Adjusting the state of the PC, the number of the Kelly sidebands can be decreased to 2 as the blue line shows. The ± 1 order Kelly sidebands locate at almost the same wavelengths. Meanwhile, the measured spectra show that the central wavelength of the soliton mode-locking laser is tunable. For example, the red line in Fig. 3(a) shows a mode locking spectrum with the central wavelength of ~1557.4 nm.
Fig. 3 (a) Three typical measured output spectra of output of fiber laser operating in the traditional soliton mode locking regime. Corresponding (b) oscilloscope trace and (c)-(d) RF spectra.
As we mentioned above, in the soliton mode locking regime, the output spectra of the laser are very broad which means the modulation depth of the transmittance curves is very weak. In this case, the soliton mode locking with a broadband spectrum can be realized and many Kelly sidebands appear symmetrically. With increasing in the modulation depth of the transmittance curves, because of the narrow period of the transmittance peaks and high loss at the two sides of the transmittance peak, the number of the Kelly sidebands would decrease. As expected, by carefully changing the state of the PC, soliton mode locking without Kelly sidebands can be obtained and the spectra are shown in Fig. 4(a). A transitional mode locking spectrum is shown by the red line in Fig. 4(a). As the red line shows, the central wavelength of the mode locking locates at 1560.3 nm with 3-dB bandwidth of ~1.9 nm. Only very weak ± 1 order Kelly sidebands located at 1551.1 nm and 1568.7 nm can be observed. The soliton mode locking without sidebands can be observed as the black and green lines show in Fig. 4(a). The spectral bandwidth can decrease further as the green line shows. The 3-dB spectral bandwidths of the soliton mode locking indicated by the green line is ~1.5 nm. Since the strong Kelly sidebands of traditional soliton mode locking fiber laser may result in some disadvantages such as high energy loss on output pulses, high noise and instability and limitation to the pulse duration [43]. This type of soliton mode locked fiber laser producing clean soliton without Kelly sidebands can improve the signal-to-noise ratio and stability of the soliton pulse laser.
Fig. 4 (a) Typical measured output spectra of the laser operating in the soliton mode locking regime without Kelly sidebands. (b) Typical measured output spectra of the laser operating in the soliton-similariton mode locking regime with narrow bandwidth.
By changing the state of the PC, the period of the transmittance peaks can be decreased further and mode locking with narrower spectral bandwidth can be realized. As it can be found in Fig. 2, the filter period can be reduced to about 2.7 nm. Therefore, the spectral bandwidth of the mode locking laser can be reduced further. In this case, the narrow bandwidth filter effect stemming from the Lyot filtering effect makes the laser mode locking transform from the traditional soliton mode locking to the soliton-similariton mode locking regime which shows parabola-like spectrum. As mentioned before, changing the state of the PC can also lead to the shift of the transmittance peaks and then realize the central wavelength tunability of the mode locked fiber laser. In the experiment, the central wavelength of the soliton-similariton mode locking laser is tunable as well. Typical mode locking spectra with different central wavelengths in the range from 1554 nm to 1562 nm are shown in Fig. 4(b).
3.2 Tunable dual-wavelength dual-comb mode-locking
As it has been demonstrated in Fig. 2 that periodic transmittance peaks can be observed due to the Lyot filtering effect, it is natural to think of realizing multi-wavelength mode locking with this laser. With higher pump power, it is easy to realize multi-pulse trains mode-locked with two different wavelengths. Slightly increasing the 980 nm pump power to ~100 mW and properly adjusting the PC, mode locking with two separated smooth spectral envelops can be observed on the OSA. In the time domain, the output of laser was monitored by an oscilloscope. When the dual-wavelength mode locking is realized, two asynchronous pulse trains can be observed on the oscilloscope. By decreasing the pump power to ~85 mW and properly setting the state of the PC, two asynchronous single-pulse trains with different central wavelengths can be realized. i.e, the dual-wavelength dual-comb mode locking. A typical output spectrum of the dual-wavelength mode-locked laser is shown in Fig. 5(a). The central wavelengths of the dual-wavelength mode locking laser locate at 1557.7 nm and 1562.7 nm, respectively. The 3-dB bandwidth of the two spectral envelopes are ~0.8 nm and ~1.1 nm, respectively. In this case, the spectral asymmetry of the mode locking spectra with different central wavelengths can be attributed to the slightly irregular transmittance curve.
Fig. 5 (a) Typical measured optical spectra of dual-wavelength mode-locking at 1557.7 nm and 1562.7 nm. (b) Corresponding RF spectrum around the first harmonic and (inset) local view of a subsidiary peak with 5 Hz RBW.
The measured RF spectrum of the dual-wavelength mode locking laser around the fundamental repetition rate is shown in Fig. 5(b). There are two main peaks located at 18801847 Hz and 18802086 Hz, which correspond to the fundamental repetition rate of the asynchronous mode locking pulse trains at 1562.7 nm and 1557.7 nm, respectively. The frequency difference of the two main peaks is 239 Hz, which indicates the repetition rate difference of the two asynchronous pulse trains. The two main RF peaks show signal-to-noise ratio of >70 dB, which proves the good stability of the dual-wavelength mode-locking. The mode locking can be maintained for long time. Besides the two main peaks, a series subsidiary peaks with uniform spacing of ~239 Hz can be observed on the both sides symmetrically. These subsidiary peaks can be attributed to the beat of the two combs which can prove the high coherence of them. The inset of Fig. 5(b) is the local view of one subsidiary peak. The narrow width of the subsidiary peak indicates the high coherence of the two combs.
The difference of the fundamental repetition rate of the mode locking pulse trains at 1557.7 nm and 1562.7 nm indicates that the two mode locked pulse trains with different central wavelengths are asynchronous in the time domain. This can be verified with results measured with the oscilloscope. Five typical screenshots of the oscilloscope at different time are shown in Fig. 6(a)-(e) and two evenly spaced pulse trains can be observed. For each pulse train, the pulse separation is about 53 ns which is good agreement with the fundamental repetition rate of ~18.8 MHz. Meanwhile, for the five screenshots at different time, it can be found that the time interval of the two pulse trains is time-dependent. In other words, the two pulse trains are asynchronous in the time domain and the temporal walk-off between the two pulse trains is verified. These results are consistent with the slight difference of the repetition rate of the pulse trains mode locked at different wavelengths. In the Fig. 6(e), the two asynchronous pulse trains almost overlap in the time domain. The interferogram of two pulse trains on the oscilloscope with the range of 20 ms is shown in Fig. 6(f). The beat note signal with an interval of ~4.2 ms can be observed, which corresponds to the repetition rate difference of ~239 Hz. This result confirms the dual-comb mode locking further.
Fig. 6 (a)-(e) Screenshots of the two asynchronous pulse trains. (f) Temporal interferogram of the two trains on the oscilloscope with the range of 20 ms.
The asynchronization of the two pulse trains located at different wavelengths can be attributed to the different group velocities of them. The difference of group velocities origins from the dispersion of the laser cavity. The fundamental repetition rate of the mode locked fiber laser is:
where c is the light velocity in the vacuum, n is the effective refractive index of the fiber and L is the length of the laser cavity. The time separation of the pulses mode locked with the fundamental repetition rate is:Therefore, the changing of time separation of the two mode locked pulse trains with different wavelengths in one cavity round trip can be calculated as:where is the wavelength difference of the two pulse trains and D is the average dispersion of the laser cavity. Therefore, the repetition rate difference can be calculated as [30,44]:In the experiment, the dual-wavelength mode locked laser has L≈11.0 m, which is double of the fiber length in the laser cavity, Davrage≈13 ps/nm/km and T≈53.2 ns. In the case that the wavelength difference is ~5.0 nm as seen from Fig. 5(a), the calculated fundamental repetition rate difference is ~251 Hz, which is in good agreement with the measured result shown in Fig. 5(b). The slight difference may originate from the inaccuracy estimation of the dispersion.
Keeping the pump power unchanged, tunability of the dual-wavelength mode locking can be realized by adjusting the state of the PC carefully. As it can be found from Fig. (2), the wavelengths of the transmittance peaks and the shapes of the transmittance curves are tunable by adjusting the state of the PC. Therefore, it’s expected that the central wavelengths and the shape of the spectra of the dual-wavelength mode locked laser are also tunable just as that of the aforementioned single wavelength mode locking. Figure 7 shows six typical measured spectra of the dual-wavelength mode-locking laser. It can be found that the central wavelengths of the dual-wavelength mode locked laser are tunable in a certain range. Meanwhile the wavelength separation varies very slightly with different dual-wavelength mode locking states. This can be attributed to the combined effect of the Lyot filtering effect, the loss and gain of the laser cavity. With the change of the central wavelengths of the dual-wavelength mode locking, the fundamental repetition rates of the two asynchronous pulse trains change correspondingly. In these dual-wavelength mode locked states, the dual-comb laser source is realized with the frequency difference of several hundreds of Hertz.
Fig. 7 Typical measured optical spectra of dual-wavelength mode-locking at different central wavelengths and spectral shapes.
In the experiment, a novel dual-wavelength soliton molecule mode locking is observed as well. When two solitons overlap in the time domain, they will interact with each other automatically. This soliton-soliton interaction leads to attractive and repulsive forces and dominates the formation of soliton molecules. The distance and the phase difference between the two pulses are defined by the energy and momentum balanced equation [45]. As a kind of bound-soliton, soliton molecule is stable and insensitive to small perturbation comparing with high-order soliton. It attracts a lot of interest and has great potential applications in high-speed fiber-optic communications and advanced time-resolved spectroscopy [46]. In the experiment, when we increase the pump power to ~100 mW and properly set the state of the polarizer controller, a novel dual-wavelength multi-pulse soliton molecule mode locking can be realized. Decreasing the pump power to ~85 mW, dual-wavelength soliton molecule mode locking was achieved. Figure 8(a) shows a dual-wavelength mode locking spectrum with two central wavelengths located at ~1557 nm and ~1562 nm. It’s worth noting that stable fringes with high contrast in the optical spectrum can be recognized, which is the typical characteristic of soliton molecule. In Fig. 8(a), the spectral envelope with the central wavelength of ~1562 nm is smooth without spectral fringes. It means the mode locking pulse with the central wavelength of 1562 nm is the traditional soliton. Meanwhile, the spectral envelope with the central wavelength of around 1557 nm has a spectral fringe with a period of 1.14 nm, which indicates that the pulse is a soliton molecule with pulse-separation of ~7.1 ps. The deep modulation on the soliton molecule spectrum indicates that the separation between the two entangled solitons in the molecule is very stable. In this case, the laser realizes dual-wavelength soliton molecule mode locking. Meanwhile, by changing the state of the PC, another dual-wavelength soliton molecule mode locking can be realized whose spectrum is shown in Fig. 8(b). In this case, the mode locked pulse with the central wavelength of ~1558 nm is the traditional single-soliton, while the spectral envelope located at ~1562.5 nm shows a spectral fringe with a period of 0.454 nm which means a soliton molecule with a pulse-separation of ~17.8 ps. The insets of the Fig. 8 show the corresponding autocorrelation traces of the soliton molecules at two different stages of the PC. These autocorrelation traces indicate that the corresponding dual-wavelength soliton molecules have the time delays of ~7.4 ps and ~17.6 ps, respectively. Both the measured autocorrelation traces are in good agreement with the measured spectral fringe periods. The little difference may result from the dispersion of the pigtail of the laser. Comparing with the previous mode locking, the dual-wavelength soliton molecule mode locking is more sensitive to the environment. Any small perturbation may lead to the soliton molecule mode locking transform into other mode locking state, even loss of the mode locking. Usually, the soliton molecule mode locking can be maintained about ten minutes. The spectral fringe period, i.e., the soliton separation of the soliton molecule, could change with changing of the state of the PC as well. This new-type dual-wavelength soliton molecule mode locking is reported for the first time as far as we know. In these dual-wavelength soliton molecule mode locking states, two asynchronous mode locked pulse trains can be observed similar to that shown in Fig. 5. This dual-wavelength soliton molecule mode locked laser showed RF spectrum similar to that in Fig. 5, which demonstrates that it’s a dual-wavelength dual-comb laser source.
Fig. 8 Typical measured optical spectra of dual-wavelength soliton molecule mode-locking. (Insets) The corresponding autocorrelation traces.
4. Conclusion
In conclusion, we report a partially polarization maintaining all-fiber mode locked laser based on semiconductor saturable absorber. Periodic Lyot filtering effect appears due to the polarization dependent loss of the polarization-maintaining WDM combined with the polarization maintaining fibers in the laser cavity. Single wavelength mode locking with different spectral shapes and central wavelengths were demonstrated. By properly setting the state of the PC in the laser cavity, we obtained stable and controllable dual-wavelength mode locking. Due to the dispersion of laser cavity, the mode locked pulses with different wavelengths had different group velocity forming a dual-frequency comb with frequency spacing difference of ~239 Hz. Meanwhile, similar to the single wavelength mode locking, the central wavelengths and the spectral shapes of the dual-wavelength mode locked laser were demonstrated to be tunable in certain range. A novel kind of dual-wavelength soliton molecule mode-locking was demonstrated as well. This dual-wavelength dual-comb mode locked fiber laser could find important applications and is expected to simplify dual-comb spectroscopy system greatly.
Funding
Japan Society for the Promotion of Science (JSPS) (15H02250, 17K18891, 18H01504).
Acknowledgments
This work was supported by the Japan Society for the Promotion of Science KAKENHI, grant No.15H02250, &17K18891 & 18H01504 and by JSPS and CNRS under the JSPS-CNRS joint research program.
References
1. M. Marangoni, A. Gambetta, C. Manzoni, V. Kumar, R. Ramponi, and G. Cerullo, “Fiber-format CARS spectroscopy by spectral compression of femtosecond pulses from a single laser oscillator,” Opt. Lett. 34(21), 3262–3264 (2009). [CrossRef] [PubMed]
2. G. Krauss, S. Lohss, T. Hanke, A. Sell, S. Eggert, R. Huber, and A. Leitenstorfer, “Synthesis of a single cycle of light with compact erbium-doped fibre technology,” Nat. Photonics 4(1), 33–36 (2010). [CrossRef]
3. A. Schneider, M. Stillhart, and P. Günter, “High efficiency generation and detection of terahertz pulses using laser pulses at telecommunication wavelengths,” Opt. Express 14(12), 5376–5384 (2006). [CrossRef] [PubMed]
4. X. Mao, D. Inoue, S. Kato, and M. Kagami, “Amplitude-modulated laser radar for range and speed measurement in car applications,” IEEE T INTELL TRANSP 13(1), 408–413 (2012). [CrossRef]
5. V. S. Pivtsov, B. N. Nyushkov, N. A. Koliada, S. A. Farnosov, and V. I. Denisov, “Development of a prototype compact fibre frequency synthesiser for mobile femtosecond optical clocks,” Quantum Electron. 44(6), 507–514 (2014). [CrossRef]
6. I. Coddington, W. C. Swann, and N. R. Newbury, “Coherent multiheterodyne spectroscopy using stabilized optical frequency combs,” Phys. Rev. Lett. 100(1), 013902 (2008). [CrossRef] [PubMed]
7. F. C. Cruz, D. L. Maser, T. Johnson, G. Ycas, A. Klose, F. R. Giorgetta, I. Coddington, and S. A. Diddams, “Mid-infrared optical frequency combs based on difference frequency generation for molecular spectroscopy,” Opt. Express 23(20), 26814–26824 (2015). [CrossRef] [PubMed]
8. G. Millot, S. Pitois, M. Yan, T. Hovhannisyan, A. Bendahmane, T. W. Hänsch, and N. Picqué, “Frequency-agile dual-comb spectroscopy,” Nat. Photonics 10(1), 27–30 (2016). [CrossRef]
9. J. Mandon, G. Guelachvili, and N. Picqué, “Fourier transform spectroscopy with a laser frequency comb,” Nat. Photonics 3(2), 99–102 (2009). [CrossRef]
10. G. B. Rieker, F. R. Giorgetta, W. C. Swann, J. Kofler, A. M. Zolot, L. C. Sinclair, E. Baumann, C. Cromer, G. Petron, C. Sweeney, P. P. Tans, I. Coddington, and N. R. Newbury, “Frequency-comb-based remote sensing of greenhouse gases over kilometer air paths,” Optica 1(5), 290–298 (2014). [CrossRef]
11. E. Baumann, F. R. Giorgetta, J. D. Deschênes, W. C. Swann, I. Coddington, and N. R. Newbury, “Comb-calibrated laser ranging for three-dimensional surface profiling with micrometer-level precision at a distance,” Opt. Express 22(21), 24914–24928 (2014). [CrossRef] [PubMed]
12. C. H. Li, A. J. Benedick, P. Fendel, A. G. Glenday, F. X. Kärtner, D. F. Phillips, D. Sasselov, A. Szentgyorgyi, and R. L. Walsworth, “A laser frequency comb that enables radial velocity measurements with a precision of 1 cm s(-1),” Nature 452(7187), 610–612 (2008). [CrossRef] [PubMed]
13. T. Udem, R. Holzwarth, and T. W. Hänsch, “Optical frequency metrology,” Nature 416(6877), 233–237 (2002). [CrossRef] [PubMed]
14. I. Coddington, N. Newbury, and W. Swann, “Dual-comb spectroscopy,” Optica 3(4), 414–426 (2016). [CrossRef]
15. B. Bernhardt, A. Ozawa, P. Jacquet, M. Jacquey, Y. Kobayashi, T. Udem, R. Holzwarth, G. Guelachvili, T. W. Hänsch, and N. Picqué, “Cavity-enhanced dual-comb spectroscopy,” Nat. Photonics 4(1), 55–57 (2010). [CrossRef]
16. A. Bartels, R. Cerna, C. Kistner, A. Thoma, F. Hudert, C. Janke, and T. Dekorsy, “Ultrafast time-domain spectroscopy based on high-speed asynchronous optical sampling,” Rev. Sci. Instrum. 78(3), 035107 (2007). [CrossRef] [PubMed]
17. I. Coddington, W. C. Swann, L. Nenadovic, and N. R. Newbury, “Rapid and precise absolute distance measurements at long range,” Nat. Photonics 3(6), 351–356 (2009). [CrossRef]
18. M. Cassinerio, A. Gambetta, N. Coluccelli, P. Laporta, and G. Galzerano, “Absolute dual-comb spectroscopy at 1.55 μ m by free-running Er: fiber lasers,” Appl. Phys. Lett. 104(23), 231102 (2014). [CrossRef]
19. G. Villares, A. Hugi, S. Blaser, and J. Faist, “Dual-comb spectroscopy based on quantum-cascade-laser frequency combs,” Nat. Commun. 5(1), 5192 (2014). [CrossRef] [PubMed]
20. S. Schiller, “Spectrometry with frequency combs,” Opt. Lett. 27(9), 766–768 (2002). [CrossRef] [PubMed]
21. L. C. Sinclair, I. Coddington, W. C. Swann, G. B. Rieker, A. Hati, K. Iwakuni, and N. R. Newbury, “Operation of an optically coherent frequency comb outside the metrology lab,” Opt. Express 22(6), 6996–7006 (2014). [CrossRef] [PubMed]
22. 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]
23. S. Mehravar, R. A. Norwood, N. Peyghambarian, and K. Kieu, “Real-time dual-comb spectroscopy with a free-running bidirectionally mode-locked fiber laser,” Appl. Phys. Lett. 108(23), 231104 (2016). [CrossRef]
24. T. Ideguchi, T. Nakamura, Y. Kobayashi, and K. Goda, “Kerr-lens mode-locked bidirectional dual-comb ring laser for broadband dual-comb spectroscopy,” Optica 3(7), 748–753 (2016). [CrossRef]
25. X. Zhao, T. Li, Y. Liu, Q. Li, and Z. Zheng, “Polarization-multiplexed, dual-comb all-fiber mode-locked laser,” PHOTONICS RES 6(9), 853–857 (2018). [CrossRef]
26. L. Yun, X. Liu, and D. Mao, “Observation of dual-wavelength dissipative solitons in a figure-eight erbium-doped fiber laser,” Opt. Express 20(19), 20992–20997 (2012). [CrossRef] [PubMed]
27. Z. X. Zhang, Z. W. Xu, and L. Zhang, “Tunable and switchable dual-wavelength dissipative soliton generation in an all-normal-dispersion Yb-doped fiber laser with birefringence fiber filter,” Opt. Express 20(24), 26736–26742 (2012). [CrossRef] [PubMed]
28. 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]
29. 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]
30. R. Li, H. Shi, H. Tian, Y. Li, B. Liu, Y. Song, and M. Hu, “All-polarization-maintaining dual-wavelength mode-locked fiber laser based on Sagnac loop filter,” Opt. Express 26(22), 28302–28311 (2018). [CrossRef] [PubMed]
31. G. Hu, Y. Pan, X. Zhao, S. Yin, M. Zhang, and Z. Zheng, “Asynchronous and synchronous dual-wavelength pulse generation in a passively mode-locked fiber laser with a mode-locker,” Opt. Lett. 42(23), 4942–4945 (2017). [CrossRef] [PubMed]
32. N. B. Hébert, J. Genest, J. D. Deschênes, H. Bergeron, G. Y. Chen, C. Khurmi, and D. G. Lancaster, “Self-corrected chip-based dual-comb spectrometer,” Opt. Express 25(7), 8168–8179 (2017). [CrossRef] [PubMed]
33. X. Zhao, G. Hu, B. Zhao, C. Li, Y. Pan, Y. Liu, T. Yasui, and Z. Zheng, “Picometer-resolution dual-comb spectroscopy with a free-running fiber laser,” Opt. Express 24(19), 21833–21845 (2016). [CrossRef] [PubMed]
34. R. Liao, Y. Song, W. Liu, H. Shi, L. Chai, and M. Hu, “Dual-comb spectroscopy with a single free-running thulium-doped fiber laser,” Opt. Express 26(8), 11046–11054 (2018). [CrossRef] [PubMed]
35. X. Liu, D. Han, Z. Sun, C. Zeng, H. Lu, D. Mao, Y. Cui, and F. Wang, “Versatile multi-wavelength ultrafast fiber laser mode-locked by carbon nanotubes,” Sci. Rep. 3(3), 2718 (2013). [CrossRef] [PubMed]
36. K. Kieu and M. Mansuripur, “All-fiber bidirectional passively mode-locked ring laser,” Opt. Lett. 33(1), 64–66 (2008). [CrossRef] [PubMed]
37. X. Zhao, Z. Zheng, Y. Liu, G. Hu, and J. Liu, “Dual-wavelength, bidirectional single-wall carbon nanotube mode-locked fiber laser,” IEEE PHOTONIC TECH L 26(17), 1722–1725 (2014). [CrossRef]
38. I. Coddington, W. C. Swann, L. Nenadovic, and N. R. Newbury, “Rapid and precise absolute distance measurements at long range,” Nat. Photonics 3(6), 351–356 (2009). [CrossRef]
39. Y. Nakajima, Y. Hata, and K. Minoshima, “All-Polarization-Maintaining Dual-Comb Fiber Laser with Nonlinear Amplifying Loop Mirror,” in CLEO: Science and Innovations(OSA, 2018), pp. 1–2.
40. R. Wang, X. Zhao, W. Bai, J. Chen, Y. Pan, and Z. Zheng, “Polarization-maintaining, dual-wavelength, dual-comb mode-locked fiber laser,” in CLEO: Applications and Technology(Optical Society of America, 2018), pp. h2A-h139A.
41. K. Ozgören and F. O. Ilday, “All-fiber all-normal dispersion laser with a fiber-based Lyot filter,” Opt. Lett. 35(8), 1296–1298 (2010). [CrossRef] [PubMed]
42. P. S. Liang, Z. X. Zhang, Q. Q. Kuang, and M. H. Sang, “All-fiber birefringent filter with fine tunability and changeable spacing,” Laser Phys. 19(11), 2124–2128 (2009). [CrossRef]
43. M. L. Dennis and I. N. Duling, “Experimental study of sideband generation in femtosecond fiber lasers,” IEEE J QUANTUM ELECT 30(6), 1469–1477 (1994). [CrossRef]
44. X. Yao, “Generation of bidirectional stretched pulses in a nanotube-mode-locked fiber laser,” Appl. Opt. 53(1), 27–31 (2014). [CrossRef] [PubMed]
45. N. N. Akhmediev, A. Ankiewicz, and J. M. Soto-Crespo, “Stable soliton pairs in optical transmittance lines and fiber lasers,” J. Opt. Soc. Am. B 15(2), 515–523 (1998). [CrossRef]
46. U. Keller, “Recent developments in compact ultrafast lasers,” Nature 424(6950), 831–838 (2003). [CrossRef] [PubMed]
