We report on stable passively harmonic mode-locking dissipative pulses with high repetition rate and narrow bandwidth in 2µm Tm: CaYAlO4 laser. At the large intracavity intensity, the laser generated 1st-order to 5th-order passively harmonic solitons with fundamental repetition rate of ~198 MHz and 5th-order repetition rate up to 0.98 GHz, which was mainly caused by the peak power clamp effect. The solitons yielded a tunable central wavelength from 1940nm to 1950 nm, and a narrow optical spectrum bandwidth of 60 pm without any active optical filter. At low intracavity intensity, the laser operated on the typical SESAM-guided mode-locking mode, with the scaled output average power up to 1.15 W. To our knowledge, this is the first observation of passively harmonic mode locking in 2µm solid laser system, and the first Watt-level output average power in Tm: CYA mode locking laser.
© 2017 Optical Society of America
Ultra-short pulsed lasers operating at ~2 μm have attracted considerable attention due to their novelty applications in advanced polymers processing , Three-Dimensional (3-D) Laser Printing , novel optical communication , and water window attosecond pulses generation . Especially, the mode locked lasers with high repetition rate(~GHz) have promising potential in high-speed optical communication, clocking, spectroscopy, precision metrology , and high efficiency MIR-SC generation . Based on the larger nonlinearity caused by the single mode wave guide structure of fiber, the well-known conventional solitons, dissipative solitons, and other non-linear pulses were successfully obtained. With the energy quantisation effect, the stable passively harmonic mode-locking (HML) was obtained in fiber lasers, generating very high repetition rate [7,8]. Typical HML of bound solitons in Er fiber laser was reported, with 9 order harmonic of repetition rate up to 271.6 MHz .Yin reported the first Bi2Te3 nanosheets mode locked 2 μm HML in Tm fiber laser, with 10 order harmonic of repetition rate up to 215 MHz . Wang reported the first ns-levle 2 μm HML in a very long fiber cavity (~440m), with repetition rate of 2.7 MHz at 6-order soliton . However, the repetition rate and output average power is largely limited by the strong non-linear in fibers.
Due to development of the high brightness pumps, some interesting non-linear dynamics in pulse shaping such as the conventional solitons, dissipative solitons and four wave mixed (FWM) effect can be observed. The sech-shaped solitons with the typical kelly sideband in Cr2+:ZnSe solid laser was first observed by Sennaroglu . Recently, Tang reported a novelty dissipative soliton of a high brightness diode pumped Yb:NaY(WO4)2 solid-state laser with pulse duration of 54 fs . In the 2 μm wavelength region, the conventional mode locked lasers with many Tm/Ho-doped crystals or ceramics, such as the Tm:CALGO crystal , Tm:KLu(WO4)2 crystal , Tm, Ho:KY(WO4)2 crystal , Tm:Lu2O3 ceramic  and Ho:YAG ceramic , have been widely investigated, showing the operation of fundamental mode locking with limited the repetition rate of < 200 MHz. Recently, the structure disorder Tm:CaYAlO4 (Tm: CYA) crystal has attracted increasing interest as an excellent 2 μm laser crystal, due to the large gain emission cross-section, considerable cross-relaxation, and the very broad gain linewidth (~200 nm) which has the potential to support few-cycle short pulses [19,20]. Wang firstly reported the synchronous fundamental mode locking operation with dual-wavelength in Tm: CYA solid laser . Compared with the low fundamental repetition rate(<50MHz) of fiber lasers with long cavity [7–11], the shorter cavity length of solid laser is easy to obtain the higher fundamental repetition rate. Lin demonstrated a harmonic mode locking and multiple pulsing in a Kerr-lens mode-locked Ti:sapphire laser at 800 nm . Zen reported rational harmonic mode-locking based on the Yb: Gd2SiO5 crystal with broad emission bandwidth, generating 200-MHz harmonics and pulse bunches at 1064 nm . However, to the best of our knowledge, the stable passively harmonic mode locking in the Mid-Infrared conventional bulk lasers has never been reported.
In this letter, for the first time, a stable passively harmonic mode-locking dissipative pulses with SESAMs in 2 µm solid-state laser is reported. At the large intracavity intensity, the laser generated 1st order to 5th order passively harmonic solitons with fundamental repetition rate of ~198 MHz and 5th-order repetition rate up to ~0.98 GHz, which is the highest repetition rate in 2µm solid lasers. With the optimized output ratio of coupler, the stable dissipative pulses mode locking with a maximum output average power of 1.15 W was obtained. It is found that the peak-power clamp and gain spectrum filtering contributed to the HML pulse shaping.
2. Experiment setup
The laser is schematically shown in Fig. 1(a). The short Z-type laser cavity was constructed by four mirrors. The pump was a fiber coupled laser diode with 105 μm core diameter and 0.22 NA, at 790 nm. The pump beam was collimated and focused into the gain crystal by two coupling convex lenses with the same focal length. The focused pump spot radii was about 50 μm in the gain crystal. For the linear polarization output, the a-cut Tm:CYA crystal with anti-reflectively coated from 1900 nm to 2100 nm was employed with a length of 6.1 mm, a cross-section of 3 × 3 mm, and a Tm doping of 4 at %. The focused laser beam waist radii inside the gain crystal was 80 × 100 μm in sagittal and tangential planes respectively, which matched well with the pump light spot. The SESAMs (BATOP GmBH) used is designed to operate at 1800-2100 nm with a modulation depth of 1.2%, non-saturable absorption of 0.8%, a relaxation time of 10 ps, and a saturation fluence of 70 μJ/cm2.The cavity mode was additionally focused by a concave mirror with cavity mode diameter of about 120 μm on SAM. The low output ratio mirror of 1.5% was used to increase the intracavity intensity. The total optical length is short as ~75.5 cm, corresponding to the fundamental repetition rate of ~198 MHz.
An InAs photo detector with a rise time of 28 ps was connected to a 63 GHz digital oscilloscope which was employed to record pulse waveforms. An Optical Spectrum Analyzer (Yokogawa AQ6375 1.2~2.4 μm) with 0.05 nm resolution was used to monitor the output optical spectrum. The radio frequency (RF) spectrum was measured by a 10 GHz spectrum analyzer (Agilent E447A) with minimum resolution bandwidth (RBW) of 3 Hz. An autocorrelator (Pulsecheck, APE) was employed to record the pulse width.
3. Results and discussion
As typical solid lasers mode-locked by SESAMs, the laser operation went through instable Q-switched mode locking and cw mode-locking working regimes with the increasing of pump power. At the pump power of 2.8 W, a stable regular mode-locking pulse train was observed on the oscilloscope as shown in Fig. 1(b). The repetition rate of pulse train recorded by the oscilloscope was 198.4MHz, which well matches with the cavity roundtrip time of ~5ns, indicating one pulse in one roundtrip (Fig. 2(a)). When interrupting the cavity with a block, the stable pulse train can return without any additional adjustment, indicating a self-started and stable mode locking. Further increasing the pump power, the pulse train was stable and the output power increased as well. Continuously increasing the pump power to 4 W, the pulse split into two identical pulses and always appear in the middle, but the generated pulses showed slight oscillation in time, indicating a large timing jitter. Furtherly increasing the pump power, the position of splitting-pulse became relatively static and pulses are equally spaced, with exact doubled repetition rate of 397.8 MHz, output average power of 120 mW and energy of 0.6 nJ. Different from the conventional multi-pulses , our pulses were equally spaced along in the cavity and they shared the same peak intensity and pulse width (Fig. 2(b)). When the pump power was continuously increased from 4 W to 8.8 W, stable pulse trains of the three-to-five times of fundamental repetition rate were observed as well (Figs. 2(c)-2(e)). The peak intensity decreased and the pulse width recorded by photo detector also show an increasing tendency. Once it is obtained, the pulses kept stable without moving and can kept for a long time. This indicated that the laser operated on the harmonica mode locking.
Figure 3 shows the output average power with different orders. With the increasing of pump power, the output average power of HML increased from 48 mW to 380 mW at the maximum available pump power, with the slop efficiency of 5.9%. It is seen that with the increase of the pump power, the HML order reveals a stepwise progression, while the output power rises in a linear fashion. With the increased the pump power, the intracavity pulse energy has an obvious clamp tendency of ~40 nJ, due to the formation of harmonic mode locking. With the larger more powerful pumps, higher orders HML with higher repetition rate (even few GHz repetition rate) and larger output average power can be obtained.
To confirm the stability of our HML pulse train, the radio frequency (RF) spectrums were shown in Fig. 4. As can be seen, a string of regular comb-shaped peaks located on the spectrum, indicating that the HML repetition rates are exactly the 2-5 time of fundamental peak of 198.4MHz (Fig. 4(a)). At the fifth-order HML, the repetition rate was up to 984.8MHz. No frequency sidebands were observed between the RF adjacent peaks, indicating the high stability without Q-switching interruption. Once the HML was obtained, it is self-started and can stay long time. At the RBW of 100 Hz, the typical RFs of the fundamental, 3rd-order and 5th-order have a measured signal-to-noise (SNR) of 70 dB, 62 dB and 60 dB, respectively (Fig. 4(b)). The corresponding frequency width of the noise component (FWNC) was measured to be 3.5 kHz, 4.1kHz,and 10.5 kHz, respectively. Following the Ref , the pulse-to-pulse energy fluctuations were estimated to be 1.9 × 10−3, 3.6 × 10−3 and 1 × 10−2.The relative timing jitter was estimated to be 3 × 10−4, 5.7 × 10−4 and 1.6 × 10−3. Considering the cavity period of ~5 ns, the corresponding frequency jitters of 1st-orde, 3rd-order and 5th-order HML are ~1.5 ps, ~2.9 ps and ~8 ps. We can see that the timing jitters of fundament and the 3rd-order HML are very low, indicating that they are stable. The timing jitter of the 5th-order was relatively large, which was caused by instable pulse splitting limited by our maximum available pump power. With more powerful pump sources, we believe the 5th-order can be more stable and higher orders HML with higher repetition rate (even few GHz repetition rate) can be expected .
Figure 5 shows the optical spectrum of different HML orders. The central wavelength of fundamental HML was 1948.4 nm with only one peak in the large scale, indicating no other CW or Q-switching interruption (Fig. 5(a)). The spectrum bandwidth was narrow as ~62 pm. The spectrum shape was well fitted by the Gauss function, indicating the Gaussian-shaped and also the miss of well-known kelly sideband, which is different from typical conventional solitons of sech2-shaped spectrum in fiber lasers. Typical 3rd-order HML spectrum was shown in Fig. 5(b), with the similar spectrum bandwidth to fundamental. The spectrum also shows a gauss-shape and no modulation sidebands caused by the interaction of splitting pulses located on the central peak, which is also different from the HML bound solitons in fiber lasers [7,9]. Other orders of HML solitons share the similar spectrum shape and narrow spectrum width of <0.11 nm (Fig. 5(c)). But the central wavelengths with different orders are different, slightly shifting to the short wavelengths by increasing the pump power. The stable fundamental mode locking at other central wavelength of 1941.7 nm, 1944.5nm and 1950.8 nm can also be obtained by adjusting the diffraction loss of cavity mirrors or SESAMs. This indicates that the mode locking can be obtained at multiple wavelengths, it may be due to the longitudinal modes competition at large pump power. In our cavity, although no active optical filter was inserted, all the spectrum widths were strongly limited as ~0.1nm, indicating a spectrum filtering effect .
The HML spectrum width of ~70 pm predicted the transformed limitation pulse duration of ~80 ps (assuming Gaussian fitting), which is close to the pulse duration of 85 ps measured by the high-speed detector (Fig. 1(c)). For accurate pulse duration, we failed to obtain the autocorrelation trace due to the very low pulse energy (0.6 nJ) and low responsibility of APE autocorrelator in 2 µm.
In order to increase the output average power, the large output couple ratio mirrors were selected. At the output couple ratio of 20%, no stable mode locking was obtained except Q-switching due to the very low intracavity intensity. At the output couple ratio of 10%, stable mode locking can be obtained and output power was greatly improved. Figure 6(a) shows the pulse characteristics at the high output ratio. The threshold of mode locking was 4.1W, which is much higher than the HML solitons. The output power of CW mode locking has a linear scale with the increasing pump power, with the slope efficiency up to 15%. At the pump power of 7.2 W, the pulses split, resulted in stable 2nd-order HML with the output power of 800 mW, corresponding to the intracavity pulse energy of nearly 40 nJ before splitting, indicating an obvious intracavity pulse energy clamp. At the maximum available pump power of 9.1 W, the stable 2nd-order harmonic mode locking still maintained with the output power up to 1.15 W, without the interruption of Q-switching. They also showed the shifted central wavelength, Gaussian spectrum shape and narrow bandwidth of 0.13 nm (Fig. 6(Inset)) which were very similar to the HML at 1.5% output ratio.
Figure 6(b) shows the stable fundamental mode locking pulse autocorrelation trace at the output power of ~700 mW. The single peak of autocorrelation trace confirmed a single mode-locking pulse operation. The pulse duration was estimated to be ~53 ps by the Gaussian-shaped hypothesis. Considering the spectrum bandwidth of 0.13 nm, the time-bandwidth product (TBP) was estimated to be ~0.55, which is slightly larger than the transformed 0.441, indicating a slight chirping in the pulse. In other wavelengths, pulse durations ranging from 50 to 80 ps were recorded with single pulse operation. The corresponding SNR of >74 dB was obtained indicating a very stable mode locking, the M-square was measured to be ~1.2 and the polarization degree was ~23 dB measured by the Glan-Taylor prism.
For the formation of HML, we find that the peak power clamp effect, finite gain bandwidth and the self-organizing effect of split pulses in the cavity contributed a lot. Due to the same intracavity pulse energy (before splitting) of ~40 nJ at the output ratio of 1.5% and 10%, the corresponding clamp peak powers was estimated to be ~500 W and ~800 W, considering the typical pulse duration of ~85 ps and ~53 ps. The lower clamp peak power at the 1.5% coupler was due to the longer duration caused by the stronger spectrum limitation. This indicates that long-duration pulse of narrow spectrum is easier to split than the short-duration pulse with wide spectrum. That is to say, the strong spectrum limitation decreased the threshold of the pulse-splitting, which is a typical characteristic of dissipative fiber solitons . This indicates the peak power clamp effect and the spectrum filtering effect in the cavity is the leading factor of pulse splitting. The peak power clamp in our cavity may be caused by the two-photon absorption of the fast SESAMs we employed, which can result in pulse splitting . In solid-state lasers, the refractive index change caused by the acoustic effect(created by the electrostriction in fiber) for a short-length crystal(6-mm Tm:CYA in our experiment)is too small to pull the pulses into equal spacing, and the transient gain dynamics was recognized to be the main reason for pulse re-organization in solid-state lasers [22,28]. When each pulse passes through the gain medium, the drift of pulse is proportional to the value of the differential gain. The gain dynamics will continue to push each of the split pulses until the time spacing between adjacent pulses become equal. In our experiment, the recovery time of Tm:CYA about 6.3 ms  is much longer than cavity round trip time of 5 ns and the depleted gain is much smaller than the total net gain. After repeating this process in the cavity, two pulses will become equal spacing in one round trip time. Similar analysis can also be applied for multiple pulses to result in higher-order HML. More details still need further exploration.
The long pulse duration of sub-100 ps is mainly due to the spectrum limitation effects, which may be caused by the strong intracavity absorption of H2O and CO2 molecules around 1.95 µm [14,29]. In our setup, all the optical elements including SESAM have a large and smooth transmission bandwidth (1900-2100 nm), so the active spectrum filtering of cavity elements can be ruled out. The obtained mode locked central wavelengths of 1941.7 nm, 1944.5nm and 1950.8 nm in our experiments are in the typical atmosphere absorption band, and the emission wavelengths locating at the transmission window indicated the atmosphere absorption modulated the mode locking spectrum (Fig. 5(c)). The strong absorption split the emission spectrum of Tm: CYA and resulted in multiple narrow peaks on the gain spectrums, which largely limited the mode locking spectrum bandwidth. Actually, similar results in Ref [15,21]. also have been reported with long pulse duration and narrow spectrum. So the strong intracavity absorption of atmosphere played an important role in the narrow bandwidth and long pulse duration of our mode locked pulses. With the proper nitrogen purging in the laser cavity or tuning the emission wavelength away from the atmosphere absorption band, the broad spectrum and shorter pulse duration can be expected [14,18].
In conclusion, we have demonstrated a stable passively HML pulses in 2 µm solid laser system for the first time. It has a high repetition rate of ~1GHz and the narrowest spectrum of ~60 pm without any active spectrum filter. By optimizing the intracavity intensity, the laser operated on the stable HML of high output of 1.15 W, which is the highest output power in Tm: CYA crystal lasers. This should be widely applied in 2 µm optical communication, optical clock, remote sensing, and light detection systems.
Priority Academic Program Development of Jiangsu Higher Education Institutions (No.15KJB510009); Natural Science Foundation of Jiangsu Province, China (No. BK20160221).
References and links
1. K. Sugioka and Y. Cheng, “Ultrafast lasers—reliable tools for advanced materials processing,” Light Sci. Appl. 3(4), e149 (2014). [CrossRef]
2. H. Chen, K. Guo, H. Yang, D. Wu, and F. Yuan, “Thoracic Pedicle Screw Placement Guide Plate Produced by Three-Dimensional (3-D) Laser Printing,” Med. Sci. Monit. 22, 1682–1686 (2016). [CrossRef] [PubMed]
3. Z. Li, A. M. Heidt, N. Simakov, Y. Jung, J. M. O. Daniel, S. U. Alam, and D. J. Richardson, “Diode-pumped wideband thulium-doped fiber amplifiers for optical communications in the 1800 - 2050 nm window,” Opt. Express 21(22), 26450–26455 (2013). [CrossRef] [PubMed]
4. T. Popmintchev, M. C. Chen, D. Popmintchev, P. Arpin, S. Brown, S. Alisauskas, G. Andriukaitis, T. Balciunas, O. D. Mücke, A. Pugzlys, A. Baltuska, B. Shim, S. E. Schrauth, A. Gaeta, C. Hernández-García, L. Plaja, A. Becker, A. Jaron-Becker, M. M. Murnane, and H. C. Kapteyn, “Bright coherent ultrahigh harmonics in the keV x-ray regime from mid-infrared femtosecond lasers,” Science 336(6086), 1287–1291 (2012). [CrossRef] [PubMed]
5. G. Agrawal, Non-linear fiber optics. (New York: Academic, 2001).
6. K. Liu, J. Liu, H. Shi, F. Tan, and P. Wang, “High power mid-infrared supercontinuum generation in a single-mode ZBLAN fiber with up to 21.8 W average output power,” Opt. Express 22(20), 24384–24391 (2014). [CrossRef] [PubMed]
7. D. Tang, L. Zhao, B. Zhao, and A. Liu, “Mechanism of multisoliton formation and soliton energy quantization in passively mode-locked fiber lasers,” Phys. Rev. A 72(4), 043816 (2005). [CrossRef]
8. J. Koo, J. Park, J. Lee, Y. M. Jhon, and J. H. Lee, “Femtosecond harmonic mode-locking of a fiber laser at 3.27 GHz using a bulk-like, MoSe2-based saturable absorber,” Opt. Express 24(10), 10575–10589 (2016). [CrossRef] [PubMed]
10. K. Yin, B. Zhang, L. Li, T. Jiang, X. Zhou, and J. Hou, “Soliton mode-locked fiber laser based on topological insulator Bi2Te3 nanosheets at 2 μm,” Photon. Res. 3(3), 72–76 (2015). [CrossRef]
11. X. Wang, P. Zhou, X. Wang, H. Xiao, and Z. Liu, “Pulse bundles and passive harmonic mode-locked pulses in Tm-doped fiber laser based on nonlinear polarization rotation,” Opt. Express 22(5), 6147–6153 (2014). [CrossRef] [PubMed]
14. Y. Wang, G. Xie, X. Xu, J. Di, Z. Qin, S. Suomalainen, M. Guina, A. Härkönen, A. Agnesi, U. Griebner, X. Mateos, P. Loiko, and V. Petrov, “SESAM mode-locked Tm:CALGO laser at 2 µm,” Opt. Mater. Express 6(1), 131–136 (2016). [CrossRef]
15. W. B. Cho, A. Schmidt, J. H. Yim, S. Y. Choi, S. Lee, F. Rotermund, U. Griebner, G. Steinmeyer, V. Petrov, X. Mateos, M. C. Pujol, J. J. Carvajal, M. Aguiló, and F. Díaz, “Passive mode-locking of a Tm-doped bulk laser near 2 microm using a carbon nanotube saturable absorber,” Opt. Express 17(13), 11007–11012 (2009). [CrossRef] [PubMed]
16. A. A. Lagatsky, F. Fusari, S. Calvez, J. A. Gupta, V. E. Kisel, N. V. Kuleshov, C. T. A. Brown, M. D. Dawson, and W. Sibbett, “Passive mode locking of a Tm,Ho:KY(WO4)2 laser around 2 μm,” Opt. Lett. 34(17), 2587–2589 (2009). [CrossRef] [PubMed]
17. A. A. Lagatsky, Z. Sun, T. S. Kulmala, R. S. Sundaram, S. Milana, F. Torrisi, O. L. Antipov, Y. Lee, J. H. Ahn, C. T. A. Brown, W. Sibbett, and A. C. Ferrari, “2μm solid-state laser mode-locked by single-layer graphene,” Appl. Phys. Lett. 102(1), 013113 (2013). [CrossRef]
18. Y. Wang, R. Lan, X. Mateos, J. Li, C. Hu, C. Li, S. Suomalainen, A. Härkönen, M. Guina, V. Petrov, and U. Griebner, “Broadly tunable mode-locked Ho:YAG ceramic laser around 2.1 µm,” Opt. Express 24(16), 18003–18012 (2016). [CrossRef] [PubMed]
19. R. Moncorgé, N. Garnier, P. Kerbrat, C. Wyon, and C. Borel, “Spectroscopic investigation and two-micron laser of Tm3+:CaYAlO4, single crystals,” Opt. Commun. 141(1-2), 29–34 (1997). [CrossRef]
20. J. Q. Di, D. H. Zhou, X. D. Xu, C. T. Xia, W. Gao, H. R. Zheng, Q. L. Sai, F. Mou, and J. Xu, “Spectroscopic properties of Tm,Ho:CaYAlO4 single crystals,” Cryst. Res. Technol. 49(7), 446–451 (2014). [CrossRef]
21. 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]
22. J. H. Lin, W. F. Hsieha, and H. H. Wu, “Harmonic mode locking and multiple pulsing in a soft-aperture Kerr-lens mode-locked Ti:sapphire laser,” Opt. Commun. 212(1-3), 149–158 (2002). [CrossRef]
23. Q. Hao, W. X. Li, E. Wu, and H. P. Zeng, “Diode-Pumped Rational Harmonic Mode-Locked Yb:GSO Laser,” IEEE J. Quantum Electron. 45(1), 86–89 (2009). [CrossRef]
24. D. von der Linde, “Characterization of the noise in continuously operating mode-locked lasers,” Appl. Phys. B 39(4), 201–217 (1986). [CrossRef]
27. E. R. Thoen, E. M. Koontz, M. Joschko, P. Langlois, T. R. Schibli, F. X. Kärtner, E. P. Ippen, and L. A. Kolodziejski, “Two-photon absorption in semiconductor saturable absorber mirrors,” Appl. Phys. Lett. 74(26), 3927–3929 (1999). [CrossRef]
28. J. N. Kutz, B. C. Collings, K. Bergman, and W. H. Knox, “Stabilized Pulse Spacing in Soliton Lasers Due to Gain Depletion and Recovery,” IEEE J. Quantum Electron. 34(9), 1749–1757 (1998). [CrossRef]
29. A. Stark, L. Correia, M. Teichmann, S. Salewski, C. Larsen, V. M. Baev, and P. E. Toschek, “Intracavity absorption spectroscopy with thulium-doped fibre laser,” Opt. Commun. 215(1-3), 113–123 (2003). [CrossRef]