An average-power-scalable, two-stage optical parametric chirped pulse amplifier is presented providing 90-μJ signal pulses at 1.55 μm and 45-μJ idler pulses at 3.1 μm at a repetition rate of 100 kHz. The signal pulses were recompressible to within a few percent of their ~50-fs Fourier limit in anti-reflection coated fused silica at negligible losses. The overall energy conversion efficiency from the 1030-nm pump to the recompressed signal reached 19%, significantly reducing the cost per watt of pump power compared to similar systems. The two-stage source will serve as the front-end of a three-stage system permitting the development of novel experimental strategies towards laser-based imaging of molecular structures and chemical reactivity.
© 2015 Optical Society of America
The emerging research field of dynamic imaging of molecules via time-resolved photoelectron holography , laser-induced electron diffraction  and high-order harmonic spectroscopy  calls for the development of high repetition rate (i.e. >> 10 kHz) optical driver pulses at wavelengths significantly beyond the limit of Ti:sapphire oscillator-amplifier systems (i.e. >> 1 μm). The use of longer ionizing wavelengths in these imaging methods permits (i) the generation of high ponderomotive and recollision photoelectron energies and (ii) strong field ionization in the tunneling regime, both at moderate laser intensities . The resulting experimental conditions lead to low target ionization fractions and facilitate the interpretation of experiments through the applicability of simplifying assumptions, such as the single-active electron approximation. High repetition rate of the optical driver source is another essential requirement, when electron-ion coincidence detection is used, necessitating event rates << 1/pulse for unambiguous identification of coincident electrons and ions .
Higher ponderomotive energies are also linked to shorter wavelengths in high-order harmonic generation (HHG) potentially reaching down into the water window (i.e. 2.3-4.4 nm, corresponding to 280-530-eV photon energies). It was shown that the unfavorable scaling of the single atom response in HHG with laser wavelength can be counteracted by appropriate phase-matching . However, mJ-level driver pulse energies are needed in the short-wavelength IR (SWIR, 1.4-3 μm) to reach soft X-ray photon energies in the water window . Such pulse energies are still out of reach of current state-of-the-art high repetition rate ultrafast coherent sources. Increasing the pulse energies to the mJ scale would permit the introduction of table-top ultrafast transient absorption techniques in the water window and a two-order-of-magnitude increase in the photon flux of isolated attosecond soft X-ray pulses using multi-cycle SWIR and mid-wavelength IR (MIR, 3-8 μm) driver pulses .
Optical parametric chirped pulse amplification (OPCPA) combined with diode-pumped, picosecond, Yb-laser amplifier pump sources near 1 μm is an average-power-scalable approach for generating the required driver pulses for the above applications in the SWIR and the blue edge of the MIR (i.e. 3-4.5 μm). Due to the relatively low overall efficiency and the resulting high cost, only a very limited number of high-repetition rate SWIR and MIR OPCPAs capable of driving strong-field experiments exist. Previously reported systems include an ultra-broadband OPCPA providing 50-kHz, 22-μJ, 44-fs pulses at 3.4 μm  and an OPCPA with an output of 160-kHz, 20-μJ, 55-fs, carrier-envelope-stable pulses at 3.1 μm . The parametric amplifiers in these sources are based exclusively on MgO-doped periodically and aperiodically poled lithium niobate (LN), which exhibits an effective nonlinear coefficient far superior to those of bulk birefringent crystals applicable in the same wavelength range. Nevertheless, thermally induced photorefractive effects were shown to lead to beam distortions in LN limiting the scalability of the average power [11,12]. Recently, a two-stage ultrashort-pulse OPCPA based on beta barium borate (BBO) has been reported with pulse energies up to 75 μJ at 1.5 μm and 100 kHz . However, significant thermal load on the OPCPA stages was observed due to idler absorption complicating further power scaling.
Above a certain level of average power, the performance of the nominally transparent materials used in OPCPAs is also known to deteriorate . In this work, we explore the suitability of potassium titanyl arsenate (KTA) for average power scaling in the sub-100-fs domain. Compared to LN, KTA has been proven to be a more robust and resistant alternative, albeit at the cost of a much lower effective nonlinearity [12,15,16]. KTA is comparable with BBO in terms of effective nonlinearity, but superior regarding transparency in the SWIR and the MIR: BBO starts to absorb for wavelengths longer than about 2.1 μm, while the absorption of KTA is negligible up to ~3.6 μm. Here, we present an ultrashort-pulse, two-stage OPCPA based on a combination of a MgO-doped periodically poled LN (PPLN) pre-amplifier and a KTA booster amplifier with an unprecedentedly high pulse energy at a repetition rate of 100 kHz and a recompressed 1.55-μm signal pulse duration of ~50 fs at a negligible thermal load on the booster stage. Furthermore, the overall 1030-nm-pump-to-signal energy conversion efficiency of our front-end is much higher than in competing systems relying either on poled LN or BBO. The OPCPA described here will serve as the front-end of a dual-beam, three-stage OPCPA system providing optically synchronized pulses at 1.55 and 3.1 μm with hitherto inaccessible average powers and peak intensities paving the way for imaging molecular structures and chemical reactivity with attosecond time- and picometer spatial resolution.
2. Experimental setup and results
2.1 Pump chirped pulse amplifier chain
The scheme of the front-end is shown in Fig. 1. The pump amplifier chain relies on the chirped pulse amplification (CPA) concept and is based on Yb-doped active media with a broad-enough spectral bandwidth at 1030 nm to support ≤1-ps pulses. The short pump pulses necessitate passive (optical) synchronizaton between pump and seed pulses in the OPCPA. The seed pulses for both the pump CPA and the OPCPA are provided by a commercial multi-branch 80-MHz Er-fiber oscillator amplifier system (Toptica AG). A portion of the Er-oscillator output is amplified in a separate Er-fiber combined with a highly nonlinear fiber (HNLF) to generate a supercontinuum with the dispersive wave shifted to 1030 nm for seeding the pump amplifier chain.
The stretcher consists of a nonlinearly chirped fiber Bragg grating (CFBG) and a fiber circulator (Teraxion, Inc.) with dispersion coefficients of −99.4 ps/nm and −1.24 ps/nm2 at 1030 nm that were chosen to keep the B-integral low in the amplifier chain and to compensate the third order group delay dispersion (GDD) of the grating-based compressor. Approximately 240 μW at 80 MHz (i.e. 3 pJ) with a spectral full width at half maximum (FWHM) of 8.8 nm are available at the output port of the stretcher unit.
The output of the stretcher seeds a multi-stage Yb-fiber amplifier (Active Fiber Systems GmbH) containing an acousto-optic pulse picker to select a 100-kHz pulse train out of the 80-MHz input pulse train. The pulses are amplified to an average power of 10 mW (i.e. 100 nJ) at a spectral FWHM of 8.2 nm centered at 1030 nm.
The output pulses from the Yb-fiber amplifier are sent through a fiber-integrated isolator with its output fiber pigtail plugged into an Yb:YAG amplifier system (Amphos GmbH) consisting of a pre-amplifier module and a final Innoslab booster module with an output power of 440 W (i.e. 4.4 mJ) and a spectral FWHM reduced below 1.5 nm due to gain narrowing in the Yb:YAG stages. The M2 values in the horizontal and vertical plane are 1.2 and 1.1, respectively. The beam is astigmatic to pre-compensate the astigmatism added by the subsequent grating compressor.
The compressor (Amphos GmbH) is based on 1740-line/mm transmission gratings used in the Littrow configuration and provides an output power of 280 W. The M2 values in the horizontal and vertical plane are 1.5 and 1.1, respectively. The transmission gratings in the compressor show signs of thermal deformations leading to a power dependent change in the output beam parameters. The astigmatic deformations are pre-compensated before the compressor in order to obtain negligible astigmatism in the output beam, when the full 280-W power is transmitted through the compressor. The recompressed pulses were characterized using second-harmonic frequency resolved optical gating (SH-FROG) to cross check possible distortions in the spectral phase caused by the CFBG, which turned out to be negligible, cf. Figure 2. The retrieved pulse duration was 1.1 ps. A few percent of the energy is contained in a satellite pulse ~9 ps behind the main pulse. The origin of the satellite pulse was found to be the CFBG unit. As shown in the inset of Fig. 2(a), the near field beam profile at the output of the compressor is slightly elliptical and shows characteristic side-lobes due to diffraction in the Innoslab cavity, containing a few percent of the total power. Most of this diffraction tail is removed by a water-cooled aperture at the compressor output.
2.2 Optical parametric chirped pulse amplifier chain
The 80-MHz seed pulses of the OPCPA are provided by the recompressed output of the master unit of the multi-branch Er-fiber laser system. The master unit contains a HNLF to reduce the Fourier limit of the output pulse duration to slightly below 30 fs. The highly structured spectrum is centered at 1.56 μm and stretches from 1.4 to 1.65 µm. In addition to the ~30-fs main pulse, there is a weak, several-100-fs-long pedestal characteristic of fiber-based, spectrally broadened systems.
Dispersion management of the signal and idler pulses is based on the combination of a pulse shaper and stretching/compression in bulk materials. The approximately an order of magnitude shorter pump pulse duration in our system compared e.g. with [9,10] facilitates high fidelity and high throughput signal/idler pulse recompression in relatively short blocks of bulk material. The GDD of the signal wave in the OPCPA stages is chosen to be positive. The 1.56-μm seed of our OPCPA chain remains the seed wave until the planned third (last) stage, where the MIR idler beam will also be extracted. Due to chirp transfer from the signal to the idler in the parametric amplification process, the idler pulses from the third stage will have opposite GDD and will be recompressed separately in anti-reflection (AR) coated bulk materials with positive group velocity dispersion at 3.1 μm, such as Si.
In the OPCPA front-end, the free-space output pulses of the fiber laser module pass through a commercial 640-pixel spatial-light-modulator-based pulse shaper (Biophotonic Solutions, Inc.) and are subsequently stretched in a 74-mm long, AR coated block of N-SF57. For the first OPCPA stage, which is pumped at moderate average power levels, we chose a 2-mm-long, AR-coated, 5%-MgO-doped, fanout PPLN crystal (HC Photonics Corp.) used in collinear geometry. Seeded by 1-nJ stretched seed pulses and pumped by 4-W average power (i.e. 40 µJ/pulse) at a peak intensity of 40 GW/cm2, the first stage yields a gain of 2800 and an output signal pulse energy of 2.8 μJ. The corresponding pump-to-signal energy conversion efficiency is 7%. The idler beam at 3.1 μm generated in the first stage is filtered out by dichroic optics. There is only an almost negligible spectral narrowing in the first stage as shown by Fig. 3 and the Fourier limit for the pulse duration at 1.55 μm is 35 fs. The signal pulses are then further stretched by an additional 74-mm long AR-coated N-SF57 block to reduce the pump to signal pulse duration ratio and therefore increase the conversion efficiency in the second OPCPA stage.
The second OPCPA stage is a 4-mm-thick, AR-coated KTA crystal (Cristal Laser S.A.) used in noncollinear, type-II arrangement and pumped by an average power of 43 W (i.e. 430 µJ/pulse) at a peak intensity of 80 GW/cm2. The gain in the second OPCPA stage is in excess of 30 leading to 9 W of output signal average power (i.e. 90 μJ/pulse). The overall 1030-nm pump-to-signal energy conversion efficiency is 19%, which exceeds the corresponding conversion efficiencies in competing systems by a large margin [9,10,13]. For example, an efficiency of 8% in the case of the 7.5-W, 1.5-μm, 100-kHz signal beam was obtained in . The AR coatings showed no sign of damage even after extended use. Due to the noncollinear geometry, the 3.1-μm idler beam generated in the second stage is angularly dispersed and propagates into a direction different from the direction of the signal beam. The measured pulse energy at 3.1 μm is 45 μJ (i.e. 50% of the signal pulse energy), as expected in the high parametric gain limit from the conservation of energy, the ratio of photon energies hνidler/ hνsignal = 1.55/3.1, and negligible losses at 3.1 μm. The resulting pump-to-idler energy conversion efficiency is 9%. As a comparison, the corresponding overall pump-to-idler conversion efficiency achieved in (i) three poled LN OPCPA stages was 4.6% in  and the efficiency in (ii) four poled LN OPCPA stages was 4-4.5% in . Thus, in spite of its inferior nonlinear coefficient, KTA is far superior to LN in terms of achievable conversion efficiency.
The power stability of the amplified signal pulse train after warm-up is 0.9% rms measured for a 20-s period. Over longer periods, the power slowly decreases due to the drifting arrival time of pump and seed pulses. The issue of temporal drifts and timing fluctuations in ps or sub-ps pulse pumped OPCPAs is well known [17,18]. In order to maintain stable operation and signal/idler pulse parameters, we are currently implementing an active timing stabilization scheme.
The superfluorescence background in the output with both OPCPA stages pumped was found to be negligible even with the seed blocked in front of the whole amplifier chain.The low level of superfluorescence is a result of the high seed pulse energy, the relatively large seed to pump duration ratio, and the relatively low gain in the first and second OPCPA stages. As a comparison, the seed pulse energy was only 92.5 pJ in  and 10 pJ in .
The amplified signal pulses were upcollimated and sent through 31 cm of AR-coated fused silica at negligible losses, which cancelled most of the GDD introduced by the N-SF57 stretcher blocks. Recompression was achieved in two steps. The first step involved automatic pulse compression of the seed pulses transmitted through the unpumped OPCPA chain using MIIPS® technology (Multiphoton Intrapulse Interference Phase Scan) through the use of the commercial software (MIIPS 2.0) of the pulse shaper. Performing MIIPS compression with unpumped OPCPA stages was necessary, because the standard phase scan procedure results in a varying group delay of the signal pulses, which would lead to a varying temporal offset between pump and signal in the OPCPA stage and an interference with the MIIPS algorithm. In the second step, both OPCPA stages were pumped and the polynomial dispersion coefficients were manually tuned using the MIIPS 2.0 software in order to minimize the pulse duration retrieved by SH-FROG. The measured and retrieved FROG traces are shown in Fig. 4, while the retrieved temporal and spectral properties of the recompressed pulses are shown in Fig. 5. The measured pulse duration at FWHM intensity was 52 fs, which was within a few percent of the transform limited value. As shown in Fig. 6, the agreement between the measured and retrieved spectra is excellent. The near field signal beam profile measured after the second stage is smooth and slightly elliptical, cf. inset in Fig. 5(a).
We have demonstrated an OPCPA front-end providing record-high pulse energies in the short-wave infrared at a repetition rate of 100 kHz and a pulse duration of ~50 fs. In contrast to existing systems relying mainly on PPLN, our system is based on a more robust nonlinear crystal in the booster stage enabling higher conversion efficiency with only two stages. The combination of higher seed pulse energy, shorter pump pulse duration, and a more resistant nonlinear crystal helps to suppress beam distortions and optical damage and thus enables a reduction of the number of OPCPA stages  or to avoid the use of sophisticated apodized aperiodically poled LN , while still maintaining a large parametric gain bandwidth. The described front-end system will seed a collinear OPCPA stage providing ~0.5-mJ signal pulses at 1.55 μm and > 200-µJ idler pulses at 3.1 μm.
This research has been funded by the Leibniz-Gemeinschaft grant no. SAW-2012-MBI-2.
References and links
1. Y. Huismans, A. Rouzée, A. Gijsbertsen, J. H. Jungmann, A. S. Smolkowska, P. S. W. M. Logman, F. Lépine, C. Cauchy, S. Zamith, T. Marchenko, J. M. Bakker, G. Berden, B. Redlich, A. F. G. van der Meer, H. G. Muller, W. Vermin, K. J. Schafer, M. Spanner, M. Yu. Ivanov, O. Smirnova, D. Bauer, S. V. Popruzhenko, and M. J. J. Vrakking, “Time-resolved holography with photoelectrons,” Science 331(6013), 61–64 (2011). [CrossRef] [PubMed]
2. M. G. Pullen, B. Wolter, A.-T. Le, M. Baudisch, M. Hemmer, A. Senftleben, C. D. Schröter, J. Ullrich, R. Moshammer, C. D. Lin, and J. Biegert, “Imaging an aligned polyatomic molecule with laser-induced electron diffraction,” Nat. Commun. 6, 7262 (2015). [CrossRef] [PubMed]
3. M. Negro, M. Devetta, D. Faccialá, S. De Silvestri, C. Vozzi, and S. Stagira, “High-order harmonic spectroscopy for molecular imaging of polyatomic molecules,” Faraday Discuss. 171, 133–143 (2014). [CrossRef] [PubMed]
4. B. Wolter, M. G. Pullen, M. Baudisch, M. Sclafani, M. Hemmer, A. Senftleben, C. D. Schröter, J. Ullrich, R. Moshammer, and J. Biegert, “Strong-field physics with mid-IR fields,” Phys. Rev. X 5(2), 021034 (2015). [CrossRef]
5. Z. Ansari, M. Böttcher, B. Manschwetus, H. Rottke, W. Sandner, A. Verhoef, M. Lezius, G. G. Paulus, A. Saenz, and D. B. Milošević, “Interference in strong-field ionization of a two-centre atomic system,” New J. Phys. 10(9), 093027 (2008). [CrossRef]
6. T. Popmintchev, M.-C. Chen, O. Cohen, M. E. Grisham, J. J. Rocca, M. M. Murnane, and H. C. Kapteyn, “Extended phase matching of high harmonics driven by mid-infrared light,” Opt. Lett. 33(18), 2128–2130 (2008). [CrossRef] [PubMed]
7. S. L. Cousin, F. Silva, S. Teichmann, M. Hemmer, B. Buades, and J. Biegert, “High-flux table-top soft x-ray source driven by sub-2-cycle, CEP stable, 1.85-μm 1-kHz pulses for carbon K-edge spectroscopy,” Opt. Lett. 39(18), 5383–5386 (2014). [CrossRef] [PubMed]
8. M.-C. Chen, C. Mancuso, C. Hernández-García, F. Dollar, B. Galloway, D. Popmintchev, P.-C. Huang, B. Walker, L. Plaja, A. A. Jaroń-Becker, A. Becker, M. M. Murnane, H. C. Kapteyn, and T. Popmintchev, “Generation of bright isolated attosecond soft X-ray pulses driven by multicycle midinfrared lasers,” Proc. Natl. Acad. Sci. U.S.A. 111(23), E2361–E2367 (2014). [CrossRef] [PubMed]
10. M. Baudisch, H. Pires, H. Ishizuki, T. Taira, M. Hemmer, and J. Biegert, “Sub-4-optical-cycle, 340 MW peak power, high stability mid-IR source at 160 kHz,” J. Opt. 17(9), 094002 (2015). [CrossRef]
11. J. R. Schwesyg, M. Falk, C. R. Phillips, D. H. Jundt, K. Buse, and M. M. Fejer, “Pyroelectrically induced photorefractive damage in magnesium-doped lithium niobate crystals,” J. Opt. Soc. Am. B 28(8), 1973–1987 (2011). [CrossRef]
12. M. Baudisch, M. Hemmer, H. Pires, and J. Biegert, “Performance of MgO:PPLN, KTA, and KNbO₃ for mid-wave infrared broadband parametric amplification at high average power,” Opt. Lett. 39(20), 5802–5805 (2014). [CrossRef] [PubMed]
13. Y. Shamir, S. Hädrich, S. Demmler, M. Tschernajew, J. Limpert, and A. Tünnermann, “Short-IR GW peak power OPCPA system with record average power at 100 kHz for high-field physics,” in Advanced Solid State Lasers 2015, OSA Technical Digest Series (online) (Optical Society of America, 2015), paper AW3A.3.
14. J. Rothhardt, S. Demmler, S. Hädrich, T. Peschel, J. Limpert, and A. Tünnermann, “Thermal effects in high average power optical parametric amplifiers,” Opt. Lett. 38(5), 763–765 (2013). [CrossRef] [PubMed]
15. D. Kraemer, M. L. Cowan, R. Hua, K. Franjic, and R. J. D. Miller, “High-power femtosecond infrared laser source based on noncollinear optical parametric chirped pulse amplification,” J. Opt. Soc. Am. B 24(4), 813–818 (2007). [CrossRef]
16. G. Andriukaitis, T. Balčiūnas, S. Ališauskas, A. Pugžlys, A. Baltuška, T. Popmintchev, M.-C. Chen, M. M. Murnane, and H. C. Kapteyn, “90 GW peak power few-cycle mid-infrared pulses from an optical parametric amplifier,” Opt. Lett. 36(15), 2755–2757 (2011). [CrossRef] [PubMed]
17. A. Schwarz, M. Ueffing, Y. Deng, X. Gu, H. Fattahi, T. Metzger, M. Ossiander, F. Krausz, and R. Kienberger, “Active stabilization for optically synchronized optical parametric chirped pulse amplification,” Opt. Express 20(5), 5557–5565 (2012). [CrossRef] [PubMed]
18. S. Hädrich, J. Rothhardt, M. Krebs, S. Demmler, J. Limpert, and A. Tünnermann, “Improving carrier-envelope phase stability in optical parametric chirped-pulse amplifiers by control of timing jitter,” Opt. Lett. 37(23), 4910–4912 (2012). [CrossRef] [PubMed]