We report long-term stable passive synchronization of a femtosecond Yb-doped fiber chirped-pulse amplifier (CPA) with a mode-locked Ti:sapphire laser for pump-seed synchronization of an optical parametric chirped-pulse amplification (OPCPA) system. The fiber CPA system delivers pulses with a wavelength of 1035 nm, energy of 50 µJ, and duration of 690 fs at a repetition rate of 0.4 MHz. The seed fiber oscillator is passively synchronized with a mode-locked Ti:sapphire laser by injection of the Ti:sapphire laser pulses into the cavity of the fiber oscillator. The second harmonic (SH) output with a wavelength of 518 nm, energy of 18 µJ, and duration of 1.2 ps was prepared for the OPCPA pump. The measured timing jitter between the pump (fiber SH) and the seed (Ti:sapphire) was 42 ± 14 fs, while the jitter between two oscillator outputs was 1.4 ± 0.5 fs. The robust synchronization technique allows long-term stable operation over 8 h.
©2010 Optical Society of America
Long-term stable synchronization of two-color ultrashort pulses is an essential technology for a wide variety of applications, including ultrafast pump-probe spectroscopy , optical parametric chirped-pulse amplification (OPCPA) , coherent pulse synthesis [3–5], and optical clockwork [6–8]. OPCPA generally requires robust synchronization between high-energy “pump” pulses and much weaker “seed” pulses. Typically, the synchronization between two-color pulses is followed by amplification of one of the pulses as a pump. To date, most OPCPA experiments have utilized active synchronization [9,10], direct seeding of the spectral portion of broadband seed [11–13], or nonlinear frequency conversion in microstructure fibers [14,15]. The active synchronization requires complicated electronic feedback circuits. The latter two passive schemes are free from complicated feedback loops. However, direct seeding requires a broadband seed spectrum including the gain region of the amplifier. The nonlinear frequency conversion process is sensitive to the input energy coupled to the micro-core of the microstructure fiber, which would prevent robust long-term operation. Furthermore, due to its low conversion efficiency, an extra preamplifier is required to boost the pulse energy to effectively seed the amplifier.
An alternative approach involves the passive synchronization of two mode-locked lasers by injection of one laser pulse into another laser cavity [8,16,17]. For example, when the “master” laser pulse is injected into the cavity of the “slave” fiber laser, the difference in round-trip group delay between the two lasers is automatically compensated for by the spectral shift induced by nonlinear cross-phase modulation of two copropagating pulses. This scheme has proven to have ultralow jitter down to a few femtoseconds . As the cross-phase modulation is wavelength-independent and the negative feedback feature of this scheme assures relative insensitiveness to injection power fluctuation, this scheme is a promising long-term stable method of synchronization. With recent progress in Yb-doped fiber laser technology providing a route to high average power and high repetition rate pulsed sources [18,19], Hao et al. reported a 262 W Yb-doped fiber amplifier system synchronized to a few-cycle Ti:sapphire laser . More recently, Yan et al. reported a 131 W Yb-doped fiber amplifier system passively synchronized with a Ti:sapphire laser by the injection method . However, there have been few reports regarding pump-seed synchronization by the injection scheme, and its timing jitter characteristics have yet to be investigated. To our knowledge, there have been no previous reports regarding passive injection-based synchronization of a fiber chirped-pulse amplifier (CPA) system delivering femtosecond and microjoule pulses.
Here, we present the first report of long-term stable passive synchronization of an Yb-doped fiber CPA system in the microjoule energy and femtosecond duration range with a mode-locked Ti:sapphire laser by the injection scheme, to our knowledge. The fiber CPA system delivers pulses with pulse energy greater than 50 μJ and a duration of 690 fs at a repetition rate of 0.4 MHz. The second harmonic (SH) output with a pulse energy of 18 μJ and a duration of 1.2 ps was prepared for the OPCPA pump. The long-term stable synchronized operation over 8 h was confirmed. For verification of the synchronization performance, a simple OPCPA experiment was also performed. The pump-seed jitter characteristics were estimated from the fluctuation of the OPCPA idler. In addition to the general purpose of OPCPA, we have a special application of the synchronized pump source to amplify Fourier-synthesized arbitrary waveforms by employing multicolor OPCPA. The Ti:sapphire laser used in this experiment is the pump source of the subharmonic optical parametric oscillator (OPO) to generate multicolor synchronized and phase-locked pulses for Fourier synthesis . Future extension of the presented system to multicolor OPCPA will lead to various applications, such as coherent control of electron wave packets in atoms or condensed matter in the attosecond regime .
2. Yb-doped fiber CPA system
The experimental setup is shown schematically in Fig. 1 . Passive synchronization between the Yb-doped fiber oscillator and Ti:sapphire laser is achieved by injection of the partial output of the Ti:sapphire laser into the fiber laser cavity. The Yb-doped fiber CPA system amplifies the oscillator output and its frequency-doubled output is used as a pump for the OPCPA.The synchronized seed provided by the Ti:sapphire laser is amplified in the OPCPA system comprised of a grating stretcher and an SF57 bulk compressor.
Figure 2 shows the experimental setup of the Yb-doped fiber CPA system consisting of a mode-locked Yb-doped fiber oscillator, a grating stretcher, two preamplifiers, an acousto-optic modulator (AOM) based pulse selector, a main amplifier, and a grating compressor. The system starts with a home-made mode-locked Yb-doped fiber oscillator operating at a repetition rate of 40 MHz. The cavity design is similar to that described previously . The fiber section includes an Yb-doped gain fiber 30 cm in length and a single-mode fiber (SMF) 4 m in length, which was lengthened for efficient nonlinear interaction of copropagating pulses for passive synchronization. The resulting reduction of repetition rate helps reduce the power loss and improve the contrast ratio in subsequent pulse picking. A pigtailed pump diode delivering an average power of 300 mW at 976 nm is coupled to the cavity with a wavelength-division multiplexer (WDM). A pair of gratings with a groove density of 600 grooves/mm are used for intracavity dispersion compensation. The grating separation was set to 8 cm to provide a negative net dispersion of ~(−2.7 ± 0.2) × 104 fs2 for the spectral shift induced by the cross-phase modulation to compensate for the group delay mismatch to enable self-synchronization. As a result of stable mode-locking based on nonlinear polarization evolution, the laser emits pulses with a duration of 700 fs and 30 mW average power with pump power of 120 mW. With higher pump power, multipulsing or harmonic mode-locking is observed. The output pulse is then stretched to about 1 ns with a grating stretcher consisting of a transmission grating with a groove density of 1740 grooves/mm and a gold-coated concave mirror with a radius of curvature of 1.5 m. The finite size of the mirrors clips the dispersed spectrum into a bandwidth of about 20 nm. The output spectrum was adjusted to match the gain region of the amplifier. The 1st preamplifier is comprised of an Yb-doped single-mode polarization-maintaining double-clad fiber 4 m in length with a core diameter of 10 μm and a pump clad diameter of 125 μm. To precompensate for the power loss in pulse picking, the amplifier amplifies the stretched pulse to an average power of about 1 W with a backward pump of 5 W at 976 nm from a 200-μm core fiber-coupled laser diode. After reduction of the pulse repetition rate to 0.4 MHz with an AOM-based pulse selector, the output of about 2 mW is seeded into the 2nd preamplifier comprised of a large-mode-area Yb-doped single-modepolarizing double-clad photonic crystal fiber 1.2 m in length with a core diameter of 40 μm and a pump clad diameter of 200 μm. The amplifier is backwardly pumped with light at a wavelength of 976 nm and average power of 22 W from a 400-μm core fiber-coupled laser diode, and amplifies the pulses to an average power of about 3 W, corresponding to a pulse energy of 7.5 μJ. The low extraction efficiency is due to the unsaturated amplification because of insufficient seed power. As the main amplifier, to minimize nonlinear effects in amplification, we used an ultra-large-mode-area single-mode polarization-maintaining rod-type photonic crystal fiber 80 cm in length with a core diameter of 70 μm and a pump clad diameter of 200 μm. The main amplifier is backwardly pumped with a 400-μm core fiber-coupled laser diode emitting up to 75 W at a wavelength of 976 nm. A pair of transmission gratings with a groove density of 1740 grooves/mm are used to recompress the amplified pulses with an efficiency of 62%.
Figure 3 shows the output power characteristics of the main amplifier. At the maximum pump power of 73 W, a compressed output with an average power of 21.3 W was obtained, which corresponds to a pulse energy of 53.3 μJ. The optical-to-optical slope efficiency (including pump coupling loss) was 30% after compression, resulting in an efficiency of 48% before compression. As shown in Fig. 4(a) , the spectrum of the 50-μJ output has a full width at half maximum (FWHM) bandwidth of 2.9 nm, indicating remarkable spectral narrowing due to gain shaping and reabsorption at short wavelengths. The compressed pulse shape was measured by the method of second harmonic generation frequency resolved optical gating (SHG-FROG) with the principal component generalized projection (PCGP) algorithm . Figure 4(c) and (d) shows the measured and the retrieved FROG traces, respectively. Figure 4(b) shows the retrieved temporal shape of the 50-μJ pulse (rms error 2.1%), revealing a pulse duration of 690 fs (FWHM). In comparison with Fourier-transformed pulse duration of 612 fs, which is also shown in Fig. 4(b), the recompressed pulse is nearly transform-limited. Figure 4(e) represents the comparison of the measured temporal shapes of the 50-μJ and the 5-μJ pulses when the pump power of the main amplifier was changed. The suppression of the observed pedestal with lower pulse energy indicates that it is attributable to nonlinear phase distortion in the main amplifier. To confirm this, we performed numerical simulation to investigate nonlinear effects on the pulse shape distortion by solving the nonlinear Schrödinger equationFigure 4(f) shows the simulated temporal shapes of the 50-μJ (nonlinear phase φNL ~5 rad) and the 5-μJ (φNL ~0.5 rad) pulses in comparison with the result without self phase modulation (SPM)(φNL = 0). It is clearly seen that the pedestal appears with increase of the SPM effect. The estimated energy fraction included in the main peak of the 50-μJ pulse is 80%, corresponding to a peak power of 62 MW.
To prepare the pump pulse for OPCPA, the compressed pulse is focused onto a crystal of LiB3O5 (LBO) 2 mm thick with a lens (focal length 200 mm) for SHG. The maximum SHG output power of 7.25 W at a wavelength of 518 nm was obtained from the fundamental power of 21.3 W with a conversion efficiency of 34%, resulting in a corresponding pulse energy of 18.1 μJ. Figure 5(a) shows the SHG spectrum with a bandwidth of 0.88 nm (FWHM). The measured autocorrelation shows a pulse duration of 1.2 ps with the assumption of sech2 pulse shape [Fig. 5(b)].
3. Long-term stable passive synchronization with Ti:sapphire laser
A home-made Ti:sapphire oscillator is mode-locked at a repetition rate of 80 MHz using a cw pump with average power of 8 W at a wavelength of 532 nm from a frequency-doubled diode-pumped Nd:YVO4 laser (Coherent Verdi). As shown in Fig. 1, the output beam of a mode-locked Ti:sapphire laser is split into two branches with a beam splitter. One branch (200 mW) is injected into the ring cavity of the fiber oscillator for synchronization by transmitting the multilayer-coated mirror, as shown in Fig. 2, while the other branch is used as the seed of OPCPA. Before injection, the beam divergence is adjusted with a telescope to obtain a high coupling efficiency, and the polarization is also optimized with a half wave plate for robust synchronization. When the pump power is greater than about 150 mW, the fiber oscillator tends to be harmonically mode-locked (80 MHz repetition), but a stable fundamental mode-locking (40 MHz repetition) can be maintained with a pump power of 120 mW. Figure 6 shows the long-term characteristics of synchronized 1-s gated repetition frequencies of two lasers counted with two frequency counters. Long-term stable synchronization over 8 h was confirmed. The zoomed traces in the inset show that the two lasers synchronize well in a short-term range. The timing jitter between the two oscillator outputs was measured from the fluctuation of sum frequency generation (SFG). The two outputs are combined collinearly with a dichroic mirror and focused to a β-BaB2O4 (BBO) crystal 1 mm thick to obtain sum frequency generation. The timing jitter was estimated from the SFG fluctuation where the relative delay was tuned at half maximum of the cross-correlation. The measurement error was estimated from the amplitude noise, which was measured with the SFG fluctuation at the peak of correlation. The Fourier-transformed spectrum of the fluctuation was recorded with a vector signal analyzer (Agilent 89441A). Figure 7 shows the power spectral density (PSD) and its rms integration of the timing jitter in comparison with the amplitude noise and the noise floor of the detection system. The results indicated an integrated rms timing jitter of 1.4 ± 0.5 fs over a Fourier frequency range from 1 Hz to 0.2 MHz.
4. OPCPA (verification of synchronization performance)
For verification of the synchronization performance, we performed a simple OPCPA experiment. The experimental setup is also shown in Fig. 1. The well-known concept of downchirped pulse amplification (DPA)  was employed to match the durations of seed and pump pulses with little power loss in the bulk compressor. A slightly chirped seed pulse with a duration of 130 fs from the Ti:sapphire laser is stretched to 500 fs by providing negative chirp using a pair of reflection gratings with a groove density of 830 grooves/mm. Although the stretched pulse duration is not perfectly identical to the pump pulse duration, the chirp was determined by the dispersion of an available SF57 bulk compressor 10 cm in length. The seed and pump pulses are focused with separate lenses (focal length 300 mm) and combined collinearly with a dichroic mirror. A BBO crystal 5 mm thick is placed at the focus to amplify the signal. The amplified signal is separated and recompressed with the bulk compressor. A single-pass gain of about 330 was obtained with amplified output energy of 125 nJ for the input energy of 0.375 nJ. The recompressed pulse duration was 130 fs (FWHM).
5. Timing jitter analysis
The timing jitter between synchronized pump and seed pulses was measured from the fluctuation of idler output. Similarly as mentioned above, the timing jitter was estimated at half maximum of the cross-correlation, whereas the measurement error was evaluated from the amplitude noise effect measured from the fluctuation at the peak of correlation. Figure 8 shows the PSD and the integration of timing jitter. The integration of the timing jitter over a Fourier frequency range from 1 Hz to 0.2 MHz (Nyquist frequency) is 42 ± 14 fs, which is much higher than that between oscillators. To investigate the origin of the accumulated jitter, we also measured the timing jitter at the output of each preamplifier. Since the output pulse is highly chirped with the stretcher, the pulse was recompressed by a grating compressor prior to the jitter measurement to keep sufficient resolution. The measured jitter PSD profiles at two preamplifier outputs are shown in Fig. 9 . The integrated values over a frequency range from 1 Hz to 0.2 MHz at the 1st and the 2nd preamplifier outputs are 17 ± 12 fs and 28 ± 14 fs, respectively. Figure 10 summarizes the timing jitter values at each point as a function of optical path length from the oscillator. The result indicates that the accumulated timing jitter is approximately proportional to the optical path length, which suggests that the origin of the timing jitter is the optical path fluctuation by mechanical vibration, temperature variation, and/or air turbulence etc. This speculation is also supported by the experimental result that the reduction of amplified power had little influence on the timing jitter. Although the influence of the stretcher-compressor pair on the jitter was difficult to measure because of too low signal, the effect could be included in the optical path length fluctuation, since the optical delay of the pulse in the stretcher and the compressor is determined by the optical path length . Despite this jitter accumulation, the total jitter is less than a tenth of the stretched pulse durations and the effect on the amplitude instability would be resolved for longer stretching within the limitation of the practical size of bulk compressors.
In conclusion, we demonstrated long-term stable passive synchronization of an Yb-doped fiber CPA system with microjoule energy and femtosecond duration with a mode-locked Ti:sapphire laser by the injection scheme. The fiber CPA system developed here delivers pulses with a pulse energy higher than 50 μJ and a duration of 690 fs at a repetition rate of 0.4 MHz. The SH output with a pulse energy of 18 μJ and a duration of 1.2 ps was prepared for the pump of OPCPA. Robust synchronization between the Ti:sapphire laser and Yb-doped fiber laser was demonstrated with a timing jitter of 1.4 ± 0.5 fs by injection of Ti:sapphire laser pulses into the fiber ring cavity. The long-term stable synchronized operation over 8 h was also confirmed. To verify the synchronization, we performed a simple OPCPA experiment using two synchronized pulse sources. The pump-seed jitter estimated from the fluctuation of idler revealed a timing jitter of 42 ± 14 fs. We investigated jitter characteristics in the fiber amplifier stages, and the results indicate that the optical path fluctuation is the main cause of the accumulated jitter. The long-term stable pump-seed synchronization of the fiber CPA system with the aid of the injection-based passive scheme is promising for practical OPCPA systems.
References and links
1. A. H. Zewail, “Femtochemistry: atomic-scale dynamics of the chemical bond,” J. Phys. Chem. A 104(24), 5660–5694 (2000). [CrossRef]
2. A. Dubietis, G. Jonušauskas, and A. Piskarskas, “Powerful femtosecond pulse generation by chirped and stretched pulse parametric amplification in BBP crystal,” Opt. Commun. 88(4-6), 437–440 (1992). [CrossRef]
3. R. K. Shelton, L.-S. Ma, H. C. Kapteyn, M. M. Murnane, J. L. Hall, and J. Ye, “Phase-coherent optical pulse synthesis from separate femtosecond lasers,” Science 293(5533), 1286–1289 (2001). [CrossRef] [PubMed]
4. T. R. Schibli, J. Kim, O. Kuzucu, J. T. Gopinath, S. N. Tandon, G. S. Petrich, L. A. Kolodziejski, J. G. Fujimoto, E. P. Ippen, and F. X. Kaertner, “Attosecond active synchronization of passively mode-locked lasers by balanced cross correlation,” Opt. Lett. 28(11), 947–949 (2003). [CrossRef] [PubMed]
5. D. Yoshitomi, Y. Kobayashi, and K. Torizuka, “Characterization of Fourier-synthesized optical waveforms from optically phase-locked femtosecond multicolor pulses,” Opt. Lett. 33(24), 2925–2927 (2008). [CrossRef] [PubMed]
6. D. J. Jones, K. W. Holman, M. Notcutt, J. Ye, J. Chandalia, L. A. Jiang, E. P. Ippen, and H. Yokoyama, “Ultralow-jitter, 1550-nm mode-locked semiconductor laser synchronized to a visible optical frequency standard,” Opt. Lett. 28(10), 813–815 (2003). [CrossRef] [PubMed]
7. A. Bartels, N. R. Newbury, I. Thomann, L. Hollberg, and S. A. Diddams, “Broadband phase-coherent optical frequency synthesis with actively linked Ti:sapphire and Cr:forsterite femtosecond lasers,” Opt. Lett. 29(4), 403–405 (2004). [CrossRef] [PubMed]
8. D. Yoshitomi, Y. Kobayashi, M. Kakehata, H. Takada, K. Torizuka, T. Onuma, H. Yokoi, T. Sekiguchi, and S. Nakamura, “Ultralow-jitter passive timing stabilization of a mode-locked Er-doped fiber laser by injection of an optical pulse train,” Opt. Lett. 31(22), 3243–3245 (2006). [CrossRef] [PubMed]
9. N. Ishii, L. Turi, V. S. Yakovlev, T. Fuji, F. Krausz, A. Baltuška, R. Butkus, G. Veitas, V. Smilgevicius, R. Danielius, and A. Piskarskas, “Multimillijoule chirped parametric amplification of few-cycle pulses,” Opt. Lett. 30(5), 567–569 (2005). [CrossRef] [PubMed]
10. S. Witte, R. Zinkstok, W. Hogervorst, and K. S. E. Eikema, “Generation of few-cycle terawatt light pulses using optical parametric chirped pulse amplification,” Opt. Express 13(13), 4903–4908 (2005). [CrossRef] [PubMed]
11. C. P. Hauri, P. Schlup, G. Arisholm, J. Biegert, and U. Keller, “Phase-preserving chirped-pulse optical parametric amplification to 17.3 fs directly from a Ti:sapphire oscillator,” Opt. Lett. 29(12), 1369–1371 (2004). [CrossRef] [PubMed]
12. S. Adachi, H. Ishii, T. Kanai, N. Ishii, A. Kosuge, and S. Watanabe, “1.5 mJ, 6.4 fs parametric chirped-pulse amplification system at 1 kHz,” Opt. Lett. 32(17), 2487–2489 (2007). [CrossRef] [PubMed]
13. M. Schultze, T. Binhammer, A. Steinmann, G. Palmer, M. Emons, and U. Morgner, “Few-cycle OPCPA system at 143 kHz with more than 1 microJ of pulse energy,” Opt. Express 18(3), 2836–2841 (2010). [CrossRef] [PubMed]
14. C. Y. Teisset, N. Ishii, T. Fuji, T. Metzger, S. Köhler, R. Holzwarth, A. Baltuška, A. M. Zheltikov, and F. Krausz, “Soliton-based pump-seed synchronization for few-cycle OPCPA,” Opt. Express 13(17), 6550–6557 (2005). [CrossRef] [PubMed]
15. J. Rothhardt, S. Hädrich, D. N. Schimpf, J. Limpert, and A. Tünnermann, “High repetition rate fiber amplifier pumped sub-20 fs optical parametric amplifier,” Opt. Express 15(25), 16729–16736 (2007). [CrossRef] [PubMed]
17. C. Zhou, Y. Cai, L. Ren, P. Li, S. Cao, L. Chen, M. Zhang, and Z. Zhang, “Passive synchronization of femtosecond Er- and Yb-fiber lasers by injection locking,” Appl. Phys. B 97(2), 445–449 (2009). [CrossRef]
18. Y. Zaouter, J. Boullet, E. Mottay, and E. Cormier, “Transform-limited 100 microJ, 340 MW pulses from a nonlinear-fiber chirped-pulse amplifier using a mismatched grating stretcher-compressor,” Opt. Lett. 33(13), 1527–1529 (2008). [CrossRef] [PubMed]
19. T. Eidam, S. Hanf, E. Seise, T. V. Andersen, T. Gabler, C. Wirth, T. Schreiber, J. Limpert, and A. Tünnermann, “Femtosecond fiber CPA system emitting 830 W average output power,” Opt. Lett. 35(2), 94–96 (2010). [CrossRef] [PubMed]
21. M. Yan, W. Li, Q. Hao, Y. Li, K. Yang, H. Zhou, and H. Zeng, “High-power nanosecond ytterbium-doped fiber laser passively synchronized with a femtosecond Ti:sapphire laser,” Opt. Lett. 34(21), 3331–3333 (2009). [CrossRef] [PubMed]
22. Y. Kobayashi, H. Takada, M. Kakehata, and K. Torizuka, “Phase-coherent multicolor femtosecond pulse generation,” Appl. Phys. Lett. 83(5), 839–841 (2003). [CrossRef]
23. T. Nakajima, “Effects of the carrier-envelope phase on atomic ionization by the pulse train in the multiphoton-ionization regime,” Phys. Rev. A 79(4), 043414 (2009). [CrossRef]
25. D. J. Kane, “Principal components generalized projections: a review,” J. Opt. Soc. Am. B 25(6), A120 (2008). [CrossRef]
26. D. M. Gaudiosi, A. L. Lytle, P. Kohl, M. M. Murnane, H. C. Kapteyn, and S. Backus, “11-W average power Ti:sapphire amplifier system using downchirped pulse amplification,” Opt. Lett. 29(22), 2665–2667 (2004). [CrossRef] [PubMed]
27. S. D. Brorson and H. A. Haus, “Diffraction gratings and geometrical optics,” J. Opt. Soc. Am. B 5(2), 247–248 (1988). [CrossRef]