We report on the first demonstration, to the best of our knowledge, of sub-100 fs pulses directly from the diode-pumped mode-locked Yb:KGW bulk oscillators operated at a low repetition rate. The 36 MHz oscillator delivered 78 fs pulses with pulse energy of 50 nJ and peak power of 0.65 MW. The cavity was extended by inserting a 1:1 imaging telescope, allowing 85 fs pulses to be generated at a repetition rate of 18 MHz. The pulse energy up to 83 nJ was reached, corresponding to a peak power as high as 1 MW. Sub-100 fs regime was achieved by dual action of the Kerr-lens and saturable absorber (KLAS) mode locking.
© 2014 Optical Society of America
Over the past decade there has been significant effort devoted to the development of ultrashort pulse lasers with high peak powers . Of particular interest there were laser oscillators that could deliver Megawatt peak powers with sub-100 fs pulses [2, 3]. While this regime was reached by Ti:Sapphire lasers almost two decades ago owing to their ability to produce sub-10 fs pulses , it remained very challenging for the crystalline Yb-ion based diode-pumped lasers because of their relatively narrow gain bandwidths. On the other hand, performance of a broadband gain medium like Yb:glass was limited by its poor thermal properties. This situation changed only very recently with the development of a broadband Yb:CALGO laser crystal with greatly improved thermal conductivity , which was used with a semiconductor saturable absorber mirror (SESAM) mode locking to produce 94 fs pulses with 1.6 MW of peak power and 62 fs pulses with 1.1 MW from bulk  and thin-disk  oscillator configurations, respectively.
In this work we present an alternative approach to reach such peak power regime. It is based on dual action of Kerr-lens and saturable absorber (KLAS) mode locking which was implemented in an extended cavity bulk Yb:KGW laser and resulted in generation of 85 fs pulses with 1.0 MW of peak power. A combination of KLAS mode locking with a high gain commercially available Yb:KGW laser crystal enabled us to achieve this performance with more than 2 times lower pump power when compared to the previous reports [6, 7]. At the same time, to the best of our knowledge, this is also the first demonstration of sub-100 fs pulses directly from the extended cavity Yb-ion based oscillators.
The extended cavity oscillators offer straightforward approach towards the scaling of the peak power by lowering the repetition rate (typically below 30-40 MHz). This can be simply implemented by incorporating pure optical components such as intracavity Herriott cells and 1:1 imaging telescopes [8–11]. Nonetheless, the increased influence of the nonlinear effects usually results in a trade-off between the pulse duration and the peak power (or pulse energy). So far, pulses with 2.3 MW peak power (1 μJ energy) have been delivered from a 10 MHz Yb:KYW bulk oscillator . The pulse duration, however, was limited to 430 fs. Similarly, pulses with 1.3 MW peak power and 323 fs duration were generated from a 23.7 MHz Yb:CALGO bulk oscillator . On the other hand, short pulses with 145 fs duration but at a reduced peak power of 0.16 MW (24 nJ energy) were demonstrated directly from a 27 MHz Yb:CALGO bulk oscillator . Recently we demonstrated 67 fs pulses with 3 W of average output power at a repetition rate of 77 MHz from an Yb:KGW bulk oscillator . The enhanced performance of this laser resulted from the KLAS mode locking approach which combined the fast saturable absorber-like action of Kerr-lens mode locking (KLM) with the self-starting operation of a SESAM. Therefore, this laser system forms an excellent platform to explore its potential in scaling of the peak power through the pulse energy at lower repetition rates. We report on the generation of sub-100 fs pulses with MW peak power level directly from the long cavity Yb:KGW bulk oscillators operated at either 36 MHz or 18 MHz.
2. Experimental setup
The experiments were carried out with a 1.5 at.%-doped Yb:KGW crystal slab (Eksma). The 5-mm-long and 1.2-mm-thick slab was antireflection coated and cut for beam propagation along the Ng-axis. The excitation was provided by a fiber-coupled laser diode with 100 µm fiber core and 0.22 NA, which offered an unpolarized pump mode with a maximum power of 30 W at 980 nm. The pump mode was launched into the crystal through a pair of achromatic doublets, forming a mode waist of 300 μm at the center of the crystal. The pump absorption of 50-60% was measured under non-lasing conditions and depended on the pump power level. The laser cavity setup is schematically depicted in Fig. 1. Three Gires–Tournois interferometer (GTI) mirrors (Layertec GmbH) with different negative group velocity dispersions (GVD) of −250 fs2, −550 fs2 and −1300 fs2 were combined, providing a variable round trip dispersion ranging from −3400 fs2 to −5200 fs2. All high reflection (HR) coated cavity mirrors were designed to exhibit low group velocity dispersion (GVD) at the laser wavelength (Laseroptik GmbH). A SESAM with 2% modulation depth and 70 μJ/cm2 fluence saturation was used to assist in the initiation of the mode locking. The SESAM was placed on a micrometer driven translation stage. Adjusting the distance between the SESAM and R3 mirror (see Fig. 1) allowed for continuous control of the cavity mode size inside the crystal.
3. Results and discussion
To optimize the KLM as well as the spatial matching between the pump and cavity modes for efficient lasing, we estimated the strength of the thermal and Kerr lenses, which are the two main factors that influence the focusing of a cavity mode inside the crystal. The strength of the thermal lens was determined experimentally by measuring the beam size variation of the cavity mode away from the output coupler and then modeling it with the ABCD matrix analysis. Due to the anisotropic thermal properties of the Yb:KGW crystals, the thermal lens was estimated to be 10-15 diopters at the maximum pump power, depending on the measured directions (along the Nm-axis or Np-axis). The relatively large cavity mode size inside the Kerr medium resulted in a lowered strength of the Kerr lens, which can be estimated through the following equation :16], the crystal length Lc = 5 mm, the pulse peak intensity Ip = 15 GW/cm2, and the cavity mode size w = 150 μm, the strength of the Kerr lens was calculated to be 27 diopters. Since the Kerr lens is about twice stronger than the thermal lens, this illustrates that it is an important factor in consideration of the soft gain aperturing and mode matching.
The initially designed cavity (without intracavity telescopes) had a round-trip length of 8 m, leading to a repetition rate of 36 MHz. At low pump power levels, the cavity performed with degraded lasing efficiency because the pump power dependent thermal lensing was not strong enough to focus the cavity mode for efficient mode matching in the crystal. The mode locking was initiated by the SESAM and the Kerr lensing was optimized at the maximum pump power level, where the thermal lens was treated as a parasitic factor that exhibited a constant value. In the optimum KLAS mode locking regime, pulses with 78 fs duration (assuming a sech2 temporal shape) and 1.8 W of average output power were directly generated from the oscillator. The autocorrelation trace of the pulses was acquired using a commercial long range (200 ps) autocorrelator (Femtochrome, FR-103XL) and is shown in Fig. 2(a). To ensure single pulse mode-locked operation the longer time ranges were monitored with a fast oscilloscope with photodiode that had combined resolution of ~100 ps.
Considering the repetition rate of 36 MHz, the pulse energy of 50 nJ and peak power of 0.65 MW were obtained. The transmission of the output coupler was 10%, which indicated 0.5 μJ of intracavity pulse energy circulated in the oscillator. For optimizing the GVD at such an energy level, two GTI mirrors with −250 fs2 and −550 fs2 dispersions were placed for two bounces in each single trip, and an HR mirror with close to zero dispersion was placed instead of the third GTI mirror, allowing for a total round trip dispersion of −3200 fs2 to be compensated. The corresponding emission spectrum was centered at 1032 nm with a FWHM of 15 nm [Fig. 2(b)], yielding a time-bandwidth product of 0.32. The laser’s output radiation was polarized along the Nm-axis.
The laser optimized for the KLAS mode locking could be operated at higher intracavity pulse energy levels at the cost of pulse duration as more GVD was introduced. Figure 3 shows the output power and the pulse duration at different compensated GVD levels. It should be noted that all of these results were obtained at the highest pump power of 30 W. As the round trip negative dispersion of 5200 fs2 was added into the cavity, we obtained pulses with an average output power of 2.2 W, which corresponds to a pulse energy of 60 nJ. The pulse duration at this level of energy was 100 fs, resulting in a slightly reduced peak power of 0.61 MW.
In order to understand the effect of Kerr lensing on stabilization of the single-pulsed mode locking, we studied the mode locking regimes with different cavity mode sizes inside the gain medium by translating the SESAM. As the distance between the SESAM and the concave mirror R3 was increased, the cavity mode size inside the crystal was decreased, allowing for a better spatial mode matching and therefore increased output power. From the point of view of self-amplitude modulation (SAM), however, the decrease of the cavity mode size reduced its gain aperturing effect, therefore resulting in lower effective modulation depth. As a consequence, the laser started suffering from the multiple-pulse instability. At the same time, the spectral bandwidth experienced a significant narrowing. For example, as the SESAM was translated around 4 mm away from the optimum position for Kerr lensing, the double-pulsed mode locking regime was observed. The output power increased to 2.5 W (which corresponds to 70 nJ of pulse energy), and the emission spectrum was narrowed to 13 nm. Further translation of the SESAM (about 4 mm) led to the increase of the output power to 3.3 W (91 nJ of pulse energy). In this case the laser operated in multiple-pulse regime with 7 nm of spectral bandwidth. At this point, the mode locking was believed to be purely dominated by the SESAM, since the cavity mode size was very close to the pump mode size when 10 diopters of thermal lensing was taken into account. As the pump power was decreased, single pulses could be generated without broadening of the spectrum. Similar to our previous work , no pure KLM was achieved when the SESAM was replaced with an HR mirror. This supports the fact that in our case mode locking in the sub-100 fs regime was achieved and sustained only by the dual action of the SESAM and the KLM which we described earlier as KLAS mode locking. We believe that in our previous work  KLAS was also the main mode locking mechanism.
In order to reduce the repetition rate, a 4 m long 1:1 imaging telescope was inserted into the cavity. As a result the cavity was lengthened to 16 m, which allowed generation of a pulse train at a repetition rate of 18 MHz. Since more mirrors were added, the optimized value of negative dispersion was increased to 3700 fs2. At the maximum pump power, the oscillator optimized for KLAS mode locking generated 80 fs pulses at an average output power of 1 W. The corresponding pulse energy was 55 nJ and the intracavity pulse energy was 0.55 μJ, which was close to the one obtained with the 36 MHz oscillator. To extract more energy from the cavity, we used an output coupler with higher transmission of 15%. The average output power and the pulse energy were enhanced to 1.5 W and 83 nJ, respectively. The obtained pulses showed a slightly longer duration of 85 fs [Fig. 4(a)]. The 14 nm of spectral bandwidth (see Fig. 4) resulted in almost transform-limited time-bandwidth product of 0.33. At this point, the corresponding peak power as high as 1 MW was reached. The output beam shape was circular with beam quality factor M2 measured to be better than 1.1 in both directions.
As a comparison, we summarized the previously reported laser performance from the extended-cavity Yb-ion bulk laser oscillators in Fig. 5 [6, 10–13, 17–25]. These laser oscillators were usually operated at repetition rates ranging from 10 MHz to 40 MHz. However, pulses with sub-100 fs duration have never been delivered directly from the oscillator. The Yb:CALGO crystals with broad spectral bandwidth seem quite promising to reach this goal, allowing the generation of pulses with 145 fs duration directly from the oscillator and 93 fs duration after external compression . Nonetheless, such short pulse duration was obtained at the cost of much lower pulse energy and peak power, which were limited to 17 nJ and 180 kW. In this work, sub-100 fs pulses were generated directly from the oscillator. Moreover, the peak power increased by more than 5 times.
Another comparison can be made with the thin-disk oscillators which, in many cases, were operated at similar repetition rates with high peak powers. In most cases, however, the pulses delivered from such lasers suffered from the long duration of >200 fs. Moreover, despite the higher peak power that can be reached from the thin-disk oscillators, much higher pump power is needed. For example, producing 1 MW of peak power required only 30 W of pump power in the extended cavity Yb:KGW bulk oscillator presented in this work, whereas about 140 W of pump power was needed for a Yb:KLuW thin-disk oscillator .
Considering only sub-100 fs regime with Megawatt peak powers (regardless of the repetition rate), recently Yb:CALGO bulk and thin-disk oscillators generated 94 fs pulses with 1.6 MW of peak power  and 62 fs pulses with 1.1 MW , respectively. The former oscillator required >62 W and the later >71 W of pump. As can be seen, our approach allows us to reach similar operating regime with more than twice lower amount of pump power which can be explained in part by the high gain of the used Yb:KGW crystal. In this case it also can be instructive to define a simple figure of merit, peak-to-pump power efficiency, which shows how many kWs of peak power can be generated per each Watt of the used pump power and therefore has the units of kW/W. This figure of merit is more universal than the optical-to-optical efficiency because it also takes into account the duration of the generated pulses. The Yb:CALGO bulk  and thin-disk  oscillators have 26.4 kW/W and 15.3 kW/W peak-to-pump power efficiencies, respectively, as compared to the higher efficiency of 33.3 kW/W obtained in this work. Taking into account low pump absorption in the used Yb:KGW crystal, the use of polarized pump in our case can improve the peak-to-pump power efficiency even further.
We believe that optimization of the pulse duration and power scaling of the broadband Yb-ion based oscillators can result in multi-MW peak power laser systems in the sub-100 fs regime. An alternative approach to reach such performance could be chirped pulse mode locking which was already successfully implemented with Ti:sapphire lasers .
In conclusion, we have demonstrated the generation of high peak power sub-100 fs pulses from the extended cavity Yb:KGW bulk oscillators. Pulses with 78 fs duration and 0.65 MW peak power were delivered at a repetition rate of 36 MHz. At a lower repetition rate of 18 MHz, 85 fs pulses with peak power as high as 1 MW were generated. To the best of our knowledge, this is the first demonstration of sub-100 fs pulses from the extended cavity Yb-ion based oscillators.
This research was supported by the Natural Sciences and Engineering Research Council of Canada, the Western Economic Diversification Canada and the University of Manitoba.
References and links
2. T. Südmeyer, S. V. Marchese, S. Hashimoto, C. R. E. Baer, G. Gingras, B. Witzel, and U. Keller, “Femtosecond laser oscillators for high-field science,” Nat. Photonics 2(10), 599–604 (2008). [CrossRef]
3. T. Südmeyer, C. Kränkel, C. R. E. Baer, O. H. Heckl, C. J. Saraceno, M. Golling, R. Peters, K. Petermann, G. Huber, and U. Keller, “High-power ultrafast thin disk laser oscillators and their potential for sub-100-femtosecond pulse operation,” Appl. Phys. B 97(2), 281–295 (2009). [CrossRef]
4. L. Xu, G. Tempea, A. Poppe, M. Lenzner, Ch. Spielmann, F. Krausz, A. Stingl, and K. Ferencz, “High-power sub-10-fs Ti:sapphire oscillators,” Appl. Phys. B 65(2), 151–159 (1997). [CrossRef]
6. A. Greborio, A. Guandalini, and J. Aus der Au, “Sub-100 fs pulses with 12.5 W from Yb:CALGO based oscillators,” Solid State Lasers XXI: Technology and Devices, in SPIE Photonics West 2012, paper 8235–11.
7. A. Diebold, F. Emaury, C. Schriber, M. Golling, C. J. Saraceno, T. Südmeyer, and U. Keller, “SESAM mode-locked Yb:CaGdAlO4 thin disk laser with 62 fs pulse generation,” Opt. Lett. 38(19), 3842–3845 (2013). [CrossRef] [PubMed]
8. R. Herriott, H. Kogelnik, and R. Kompfner, “Off-axis paths in spherical mirror interferometers,” Appl. Opt. 3(4), 523–526 (1964). [CrossRef]
10. A. Major, V. Barzda, P. A. E. Piunno, S. Musikhin, and U. J. Krull, “An extended cavity diode-pumped femtosecond Yb:KGW laser for applications in optical DNA sensor technology based on fluorescence lifetime measurements,” Opt. Express 14(12), 5285–5294 (2006). [CrossRef] [PubMed]
11. A. Major, R. Cisek, D. Sandkuijl, and V. Barzda, “Femtosecond Yb:KGd(WO4)2 laser with >100 nJ of pulse energy,” Laser Phys. Lett. 6(4), 272–274 (2009). [CrossRef]
12. E. P. Mottay and C. Hönninger, “MicroJoule level diode-pumped femtosecond oscillator,” Commercial and biomedical applications of ultrafast lasers VII, Proc. SPIE 6460, San Jose, USA, 64600I (2007).
13. D. N. Papadopoulos, F. Druon, J. Boudeile, I. Martial, M. Hanna, P. Georges, P. O. Petit, P. Goldner, and B. Viana, “Low-repetition-rate femtosecond operation in extended-cavity mode-locked Yb:CALGO laser,” Opt. Lett. 34(2), 196–198 (2009). [CrossRef] [PubMed]
16. A. Major, I. Nikolakakos, J. S. Aitchison, A. I. Ferguson, N. Langford, and P. W. E. Smith, “Characterization of the nonlinear refractive index of the laser crystal Yb:KGd(WO4)2,” Appl. Phys. B 77(4), 433–436 (2003). [CrossRef]
17. C. Hoenninger, A. Courjaud, P. Rigail, E. Mottay, M. Delaigue, N. Dguil-Robin, J. Limpert, I. Manek-Hoenninger, and F. Salin, “0.5µJ diode pumped femtosecond laser oscillator at 9 MHz,” in Advanced Solid-State Photonics, Vienna 2008, ME2.
18. K. E. Sheetz, E. E. Hoover, R. Carriles, D. Kleinfeld, and J. A. Squier, “Advancing multifocal nonlinear microscopy: development and application of a novel multibeam Yb:KGd(WO4)2 oscillator,” Opt. Express 16(22), 17574–17584 (2008). [CrossRef] [PubMed]
20. G. Palmer, M. Schultze, M. Siegel, M. Emons, U. Bünting, and U. Morgner, “Passively mode-locked Yb:KLu(WO4)2 thin-disk oscillator operated in the positive and negative dispersion regime,” Opt. Lett. 33(14), 1608–1610 (2008). [CrossRef] [PubMed]
21. F. Brunner, T. Südmeyer, E. Innerhofer, F. Morier-Genoud, R. Paschotta, V. E. Kisel, V. G. Shcherbitsky, N. V. Kuleshov, J. Gao, K. Contag, A. Giesen, and U. Keller, “240-fs pulses with 22-W average power from a mode-locked thin-disk Yb:KY(WO4)2 laser,” Opt. Lett. 27(13), 1162–1164 (2002). [CrossRef] [PubMed]
23. O. H. Heckl, C. Kränkel, C. R. E. Baer, C. J. Saraceno, T. Südmeyer, K. Petermann, G. Huber, and U. Keller, “Continuous-wave and modelocked Yb:YCOB thin disk laser: first demonstration and future prospects,” Opt. Express 18(18), 19201–19208 (2010). [CrossRef] [PubMed]
24. O. Pronin, J. Brons, C. Grasse, V. Pervak, G. Boehm, M.-C. Amann, V. L. Kalashnikov, A. Apolonski, and F. Krausz, “High-power 200 fs Kerr-lens mode-locked Yb:YAG thin-disk oscillator,” Opt. Lett. 36(24), 4746–4748 (2011). [CrossRef] [PubMed]
25. S. Ricaud, A. Jaffres, K. Wentsch, A. Suganuma, B. Viana, P. Loiseau, B. Weichelt, M. Abdou-Ahmed, A. Voss, T. Graf, D. Rytz, C. Hönninger, E. Mottay, P. Georges, and F. Druon, “Femtosecond Yb:CaGdAlO4 thin-disk oscillator,” Opt. Lett. 37(19), 3984–3986 (2012). [CrossRef] [PubMed]
26. S. Naumov, A. Fernandez, R. Graf, P. Dombi, F. Krausz, and A. Apolonski, “Approaching the microjoule frontier with femtosecond laser oscillators,” New J. Phys. 7, 216 (2005). [CrossRef]