A simple, compact, and efficient few-cycle laser source at a central wavelength of 1 µm is presented. The system is based on a high-energy femtosecond ytterbium-doped fiber amplifier delivering 130 fs, 250 µJ pulses at 200 kHz, corresponding to 1.5 GW of peak power and an average power of 50 W. The unprecedented short pulse duration at the output of this system is obtained by use of spectral intensity and phase shaping, allowing for both gain narrowing mitigation and the compensation of the nonlinear accumulated spectral phase. This laser source is followed by a single-stage of nonlinear compression in a xenon-filled capillary, allowing for the generation of 14 fs, 120 µJ pulses at 200 kHz resulting in 24 W of average power. High-harmonic generation driven by this type of source will trigger numerous new applications in the XUV range and attosecond science.
© 2017 Optical Society of America
Lasers sources emitting energetic femtosecond pulses have opened up the area of high field physics, with phenomena including high harmonic generation (HHG) in gases and solids. This electronic process results in the generation of optical attosecond pulse trains in the extreme ultraviolet (XUV) spectral range, allowing fundamental studies in very diverse fields such as physics, chemistry, material science and biology . The capabilities of HHG sources have expanded considerably with the ever increasing control of the XUV radiation properties, induced mainly by the increased control over the driving laser source. A striking example is the possibility to generate reproducible trains of attosecond pulses, or even reproducible isolated attosecond pulses, by stabilizing the carrier-envelope phase (CEP) of the driving few-optical-cycle ultrafast laser. Possible applications of XUV sources are also determined by more practical considerations such as available XUV photon flux, repetition rate, reliability and compactness of the overall system.
The vast majority of HHG sources are nowadays driven using Ti:Sapphire lasers that typically generate 30 fs pulses with few mJ energy at 1 kHz repetition rate and 800 nm wavelength, corresponding to average powers limited to around 10 W. Average powers can be increased to few tens of Watt by the use of cryogenic cooling, but this results in even more complex and costly sources. This limitation is mostly due to spectroscopic properties of Ti:Sapphire that imply a large quantum defect-related thermal load. Nonlinear compression of these sources leads to the few-cycle regime with pulse durations below 5 fs [2,3].
There have been intense research efforts aiming at designing HHG drivers with few-cycle pulses at high repetition rate and thus at higher average power levels. In particular, optical parametric chirped pulse amplification systems have been developed both to provide access to various wavelength ranges and potentially increase the average power available [4–6]. However, the efficiency of this type of architecture is limited to a maximum of 15%. Moreover, the average power scaling of these systems is not obvious due to thermal effects in the nonlinear crystals [7,8]. Another approach is to use femtosecond rare earth-doped fiber amplifiers that are capable of generating kW average powers . In this case, gain narrowing limits the pulse duration of high energy fiber CPA to above 250 fs, and accessing the few-cycle regime requires two stages of nonlinear compression . Nevertheless, the power scaling capability of this approach has already been validated at average powers up to 200 W , leading to high photon flux XUV sources at photon energies of 70 eV .
In this article, we present a few-cycle laser source based on the high-energy ytterbium-doped femtosecond fiber amplifier exhibiting the shortest pulse duration to our knowledge. Indeed, the use of a broadband seeder, together with spectral intensity and phase shaping allows for the generation of high temporal quality 130 fs, 250 µJ pulses at 200 kHz, corresponding to 1.5 GW of peak power and 50 W of average power. The unprecedented short pulse duration at this energy level allows compression down to the few-cycle regime using a single nonlinear compression stage. This makes Yb-doped fiber amplifier-based few cycle sources more robust, compact and efficient compared to sources requiring two stages of nonlinear compression. Spectral broadening in a Xenon-filled capillary results in the generation of 14 fs and 120 µJ pulses at 200 kHz, corresponding to 24 W of average power. This work represents a promising step towards turn-key compact HHG drivers at high repetition rate.
2. Experimental set-up and laser characterization
The experimental setup, shown in Fig. 1, fits in a footprint of 0.8 m × 1.5 m, and starts with a broadband seeder. The first element of this seeder is a home-made passively mode-locked directly diode pumped Ytterbium ultrafast oscillator delivering pulses at 40 MHz repetition rate. The pulses are compressed in time at the output of the oscillator and seeded in a parabolic amplifier made in house  to broaden the spectrum, resulting in a 60 nm bandwidth at 10 dB and supporting the generation of sub-60 fs pulses. The beam then goes through a stretcher unit that cuts the spectral width to 30 nm while ensuring a stretched pulsewidth after amplification larger than 1 ns. Pulses then pass through a commercially available spatial light modulator based pulse shaper to tailor the spectrum both in intensity and phase. The repetition rate of the oscillator is then reduced to 200 kHz by means of an acousto-optic modulator pulse picker. The output of the seeder is then coupled to a state of the art fiber chirped pulse amplifier based on a 1-m long rod-type fiber amplifier with a mode field diameter of 60 µm . It is pumped by a high power fiber coupled laser diode system emitting at 976 nm. At the output of the amplifier, the compressor unit is based on a 1750 lines/mm grating, with an overall efficiency of 85%. After a first alignment at high repetition rate and low energy level, the compressor is fixed and all remaining phase is then removed by sole action on the phase shaper upstream. To ensure an optimum stability of the system, we maximize the use of fiber pigtailed components that are either spliced or connected together.
The output pulse characteristics are optimized in two steps. First, the spectrum is shaped in intensity to maximize the bandwidth for an overall gain of more than 60 dB. Hence, strong attenuation is applied to the spectral components in the vicinity of 1030 nm, as shown in the inset of Fig. 2(a). As a result, the generated spectrum is about twice as broad as conventional fiber chirped pulse amplifiers, and exhibits an 25 nm full width at 10 dB spectral width at 250 µJ compressed energy. This spectrum corresponds to a Fourier-transform limited pulse exhibiting small side lobes, with a full width at half maximum (FWHM) duration of 120 fs, as shown in Fig. 2. The second optimization step consists in measuring the remaining spectral phase, by means of a home-built second-harmonic generation frequency-resolved optical gating (SHG-FROG) apparatus, and to subtract it using the phase shaper.
To the best of our knowledge, we achieve the shortest pulses from a high energy FCPA with a FWHM pulse duration of 130 fs at 250 µJ pulse energy and 200 kHz repetition rate. Previous research efforts to maintain the spectral content and thus the pulse duration close to or below 150 fs in fiber amplifiers were limited to the few µJ energy level [15,16].
Figure 2 also shows the pulse temporal profile which is compared to the Fourier transform-limited pulse. The SHG-FROG error is 4.4x10−3 over a grid size of 512 × 512 and retrieved results show excellent agreement to independently measured spectrum and autocorrelation. At this output energy, the estimated B-integral accumulated in the power amplifier is 5 rad. Accounting for the energy located outside the main pulse, the measured peak power is slightly in excess of 1.5 GW. Furthermore, this peak power value is verified by coupling a small fraction of the output beam in a singlemode fiber and comparing the SPM-broadened spectrum with simulations based on nonlinear Schrödinger equation.
The M2 parameter at the output of the fiber amplifier source is measured to be <1.3. The long-term power stability, an important parameter for most applications, was recorded over 70 hours and is shown on Fig. 3. As expected from an ultrafast fiber laser, the measured long term average power stability is excellent with a power stability of 0.12% RMS at >50 W for 70 hours without any active stabilization.
3. Nonlinear compression stage
To further decrease the pulse duration down to the few cycle regime, a hollow capillary nonlinear compression stage is added. The selected capillary has a core diameter of 250 µm and a length of 1 m, to keep the setup compact and the theoretical transmission over 80%. It is placed on a V-groove and sealed into a gas cell. Xenon is selected as nonlinear medium owing to its high nonlinear index among noble gases (n2 = 5.10−23 m2/W). Simple calculations indicate that intensities inside the capillary result in negligible ionization rates while allowing sufficient nonlinear broadening to access the few-cycle at pressures around 1 bar. The entrance window is AR coated at the laser wavelength while the output window is uncoated. Careful mode matching optimization is carried out to focus the laser beam at the input of the capillary and minimize coupling losses. A combination of half-wave plate and thin film polarizer allows the control of the input power going into the capillary. Another half-wave plate is used to control the polarization state inside the hollow core waveguide to optimize the output polarization purity. We measured the transmission of the output window and take it into account to infer the transmission of the capillary in the experiments. The maximum input energy is reduced to 225 µJ owing to several beam sampling for laser diagnostics and losses on the optical path to the capillary. Figure 4 shows the output power and the transmission of the capillary in vacuum and when filled with 1.5 bar of Xenon as a function of input pulse energy. In vacuum the transmission remains close to 65%. When filled with Xenon, the transmission start to decrease above 150 µJ of input energy, to reach 53% at 225 µJ input energy. The excess loss mechanism is here mainly attributed to the onset of propagation loss induced by self-focusing effects in the gas medium. To solve the issue, lowering the effective nonlinear index (gas pressure and/or nature) while increasing the length and core radius of the capillary could be done, at the expense of a less compact setup. At the output of the capillary, the beam is collimated with a + 300 mm lens and directed to a chirped mirror compressor. The M2 parameter is measured to be 1.15 × 1.18 at 150 µJ input energy. Figure 4(a) shows the far field beam profile at maximum energy / spectral broadening. Owing to the capillary geometry, spatial filtering occurs during the propagation and leads to nearly perfect output Gaussian beam. The polarization state is analyzed using a broadband polarization beam splitter cube and an achromatic half-waveplate, revealing a degree of linear polarization of > 98%. The compressor is composed of two pairs of commercial double-chirped mirrors exhibiting a mean group-delay dispersion of −120 fs2 from 650 to 1200 nm per pair of bounces optimized at the central wavelength of 800 nm.
The shortest pulse duration is obtained for 225 µJ of input energy, a static Xenon pressure of 1.1 bar and 7 pairs of bounces on the compressor mirrors i.e. ~-700 fs2 of GDD. In this configuration, the total transmission of the post-compression setup, including the gas-filled capillary module, ultra-broadband beam stirring mirrors, and compressor mirrors, is as high as >53% and the maximum output energy is therefore in excess of 120 µJ. Figure 5 shows the spectral and temporal characterization of these pulses. A home-made SHG-FROG, modified to support few-cycle pulse measurement, is employed: arms of the SHG-FROG are GDD compensated and a nonlinear crystal of 10 µm thickness is implemented. This apparatus allows a thorough characterization of the background that might appear in nonlinear compression setups, for instance due to the coherent pedestal of the initial laser pulse, or to coupling in higher-order modes of the capillary. Results are compared to the experimental spectrum, measured with a calibrated fiber-coupled optical spectrum analyzer (OSA). Figure 5(a) and 5(b) show the measured and retrieved FROG traces. The FROG error is 2.1x10−3 over a 1024 × 1024 grid size, highlighting the precision of reconstructed features in the pedestal of the few-cycle pulses. The FROG-reconstructed spectrum only lacks fidelity on the long wavelength side, which might be due to spectrometer calibration or geometric effects in the SHG-FROG crystal. The retrieved FWHM pulse duration is 14 fs FWHM which corresponds to four optical cycles at the driver central wavelength of 1030 nm. The achieved peak power taking into account the amount of energy located into the pedestal or in weak temporal replicas, as highlighted in the log-scale intensity plot of the inset of Fig. 5(a), is 5.8 GW, and the Fourier Transform-limited FWHM pulse duration is 12.6 fs.
To gain insight into the spectral broadening process, we perform numerical simulations of the propagation in the capillary and compression using a commercial software package that solves the 1D (z,t) nonlinear Schrödinger equation including losses, dispersion, self-phase modulation, and self-steepening . The FROG-measured laser intensity and phase profiles are used as the initial condition with an input energy of 225 µJ. The 250 µm capillary hollow core leads to a mode field diameter of 160 µm, while the 1.1 bar pressure of Xenon corresponds to a nonlinear optical index n2 = 5.5 × 10−23 m2/W. The simulation result is compared to the experiment in Fig. 5. The simulation allows to estimate the accumulated B-integral at 16 rad, with pulses that can be compressed down to 11 fs by adding a purely quadratic GDD of −230 fs2 resulting in a peak power of 6.2 GW. The discrepancy between the simulated GDD needed to compress the pulse and the experimentally determined one mostly comes from the output window of the post-compression module and collimation lens that add about 450 fs2 of positive GDD. The simulated spectral and temporal features of the compressed pulse are in good agreement with the experimental FROG result.
Finally, the broadest spectrum that could be generated experimentally is shown in Fig. 6. By increasing the pressure of Xenon inside the capillary to 2 bar, we are able to generate spectrum from 700 nm to 1400 nm. This results in a Fourier Transform limited pulse duration of 6.5 fs. Although this spectrum could not been compressed, since the current chirped mirror compressor does not support the required bandwidth, this result suggests that sub-10 fs pulse generation using a single nonlinear compression stage is feasible using similar optics than in [10,11].
To conclude, we demonstrate the nonlinear compression of a state-of-the art Yb-doped rod-type ultrafast amplifier producing pulses with 130 fs duration at a pulse energy level of 250 µJ resulting in 1.5 GW of peak power. There is a 40 times improvement in peak power in comparison with ~150 fs ultrafast fiber systems. This short pulse duration allows the generation of 120 µJ, 14 fs pulses at 200 kHz directly from a single compression stage based on a Xe-filled hollow-core dielectric waveguide. Future developments will focus on power and energy scaling at the output of the laser using spatial and temporal coherent combining techniques , which are compatible with nonlinear compression. The architecture featuring a single nonlinear compression stage should facilitate the implementation of CEP stabilization, as has already been demonstrated on a Ti:Sa amplifier , and contributes to make this sources simpler, more robust and compact, providing an ideal tool for a number of applications in attosecond science.
“Investissements d’Avenir” LabEx PALM (ANR-10-LABX-0039-PALM) project HILAC ; Agence Nationale de la Recherche project HELLIX (ANR-16-CE30-0027-01) ; Action de Soutien à la Technologie et à la Recherche en Essonne (Conseil Départemental de l’Essonne) project SOPHIE.
References and links
1. F. Calegari, G. Sansone, S. Stagira, C. Vozzi, and M. Nisoli, “Advances in attosecond physics,” J. Phys. B 49(6), 062001 (2016). [CrossRef]
2. S. Bohman, A. Suda, T. Kanai, S. Yamaguchi, and K. Midorikawa, “Generation of 5.0 fs, 5.0 mJ pulses at 1kHz using hollow-fiber pulse compression,” Opt. Lett. 35(11), 1887–1889 (2010). [CrossRef] [PubMed]
3. X. Chen, A. Jullien, A. Malvache, L. Canova, A. Borot, A. Trisorio, C. G. Durfee, and R. Lopez-Martens, “Generation of 4.3 fs, 1 mJ laser pulses via compression of circularly polarized pulses in a gas-filled hollow-core fiber,” Opt. Lett. 34(10), 1588–1590 (2009). [CrossRef] [PubMed]
4. A. Vaupel, N. Bodnar, B. Webb, L. Shah, and M. Richardson, “Concepts, performance review, and prospects of table-top, few-cycle optical parametric chirped-pulse amplification,” Opt. Eng. 53(5), 051507 (2013). [CrossRef]
5. H. Fattahi, H. G. Barros, M. Gorjan, T. Nubbemeyer, B. Alsaif, C. Y. Teisset, M. Schultze, S. Prinz, M. Haefner, M. Ueffing, A. Alismail, L. Vámos, A. Schwarz, O. Pronin, J. Brons, X. Tao Geng, G. Arisholm, M. Ciappina, V. S. Yakovlev, D. Kim, A. M. Azzeer, N. Karpowicz, D. Sutter, Z. Major, T. Metzger, and F. Krausz, “Third-generation femtosecond technology,” Optica 1(1), 45–63 (2014). [CrossRef]
6. M. Baudisch, B. Wolter, M. Pullen, M. Hemmer, and J. Biegert, “High power multi-color OPCPA source with simultaneous femtosecond deep-UV to mid-IR outputs,” Opt. Lett. 41(15), 3583–3586 (2016). [CrossRef] [PubMed]
7. M. Puppin, Y. Deng, O. Prochnow, J. Ahrens, T. Binhammer, U. Morgner, M. Krenz, M. Wolf, and R. Ernstorfer, “500 kHz OPCPA delivering tunable sub-20 fs pulses with 15 W average power based on an all-ytterbium laser,” Opt. Express 23(2), 1491–1497 (2015). [CrossRef] [PubMed]
8. J. Rothhardt, S. Demmler, S. Hädrich, J. Limpert, and A. Tünnermann, “Octave-spanning OPCPA system delivering CEP-stable few-cycle pulses and 22 W of average power at 1 MHz repetition rate,” Opt. Express 20(10), 10870–10878 (2012). [CrossRef] [PubMed]
9. M. Müller, M. Kienel, A. Klenke, T. Gottschall, E. Shestaev, M. Plötner, J. Limpert, and A. Tünnermann, “1 kW 1 mJ eight-channel ultrafast fiber laser,” Opt. Lett. 41(15), 3439–3442 (2016). [CrossRef] [PubMed]
10. J. Rothhardt, S. Hädrich, A. Klenke, S. Demmler, A. Hoffmann, T. Gotschall, T. Eidam, M. Krebs, J. Limpert, and A. Tünnermann, “53 W average power few-cycle fiber laser system generating soft x rays up to the water window,” Opt. Lett. 39(17), 5224–5227 (2014). [CrossRef] [PubMed]
11. S. Hädrich, M. Kienel, M. Müller, A. Klenke, J. Rothhardt, R. Klas, T. Gottschall, T. Eidam, A. Drozdy, P. Jójárt, Z. Várallyay, E. Cormier, K. Osvay, A. Tünnermann, and J. Limpert, “Energetic sub-2-cycle laser with 216 W average power,” Opt. Lett. 41(18), 4332–4335 (2016). [CrossRef] [PubMed]
12. J. Rothhardt, S. Hädrich, Y. Shamir, M. Tschnernajew, R. Klas, A. Hoffmann, G. K. Tadesse, A. Klenke, T. Gottschall, T. Eidam, J. Limpert, A. Tünnermann, R. Boll, C. Bomme, H. Dachraoui, B. Erk, M. Di Fraia, D. A. Horke, T. Kierspel, T. Mullins, A. Przystawik, E. Savelyev, J. Wiese, T. Laarmann, J. Küpper, and D. Rolles, “High-repetition-rate and high-photon-flux 70 eV high-harmonic source for coincidence ion imaging of gas-phase molecules,” Opt. Express 24(16), 18133–18147 (2016). [CrossRef] [PubMed]
13. Y. Zaouter, D. N. Papadopoulos, M. Hanna, F. Druon, E. Cormier, and P. Georges, “Third-order spectral phase compensation in parabolic pulse compression,” Opt. Express 15(15), 9372–9377 (2007). [CrossRef] [PubMed]
14. Y. Zaouter, F. Guichard, L. Daniault, M. Hanna, F. Morin, C. Hönninger, E. Mottay, F. Druon, and P. Georges, “Femtosecond fiber chirped- and divided-pulse amplification system,” Opt. Lett. 38(2), 106–108 (2013). [CrossRef] [PubMed]
15. M. Wunram, P. Storz, D. Brida, and A. Leitenstorfer, “Ultrastable fiber amplifier delivering 145-fs pulses with 6-μJ energy at 10-MHz repetition rate,” Opt. Lett. 40(5), 823–826 (2015). [CrossRef] [PubMed]
16. A. Fernández, K. Jespersen, L. Zhu, L. Grüner-Nielsen, A. Baltuška, A. Galvanauskas, and A. J. Verhoef, “High-fidelity, 160 fs, 5 μJ pulses from an integrated Yb-fiber laser system with a fiber stretcher matching a simple grating compressor,” Opt. Lett. 37(5), 927–929 (2012). [CrossRef] [PubMed]
17. Fiberdesk, http://www/fiberdesk.com
18. F. Guichard, Y. Zaouter, M. Hanna, K. L. Mai, F. Morin, C. Hönninger, E. Mottay, and P. Georges, “High-energy chirped- and divided-pulse Sagnac femtosecond fiber amplifier,” Opt. Lett. 40(1), 89–92 (2015). [CrossRef] [PubMed]
19. X. Chen, L. Canova, A. Malvache, A. Julien, R. Lopez-Martens, C. Durfee, D. Papadopoulos, and F. Druon, “1-mJ, sub-5-fs carrier-envelope phase-locked pulses,” Appl. Phys. B 99(1), 149–157 (2010). [CrossRef]