Open Access
Add to BibSonomyAbstract
We present a high peak power optical parametric chirped pulse amplifier (OPCPA) seeded by a cavity dumped Ti:Sapphire oscillator. A frequency doubled high power Ytterbium-doped fiber amplifier is pumping the device. Temporal synchronization of the pump pulses is done via soliton generation in a highly nonlinear photonic crystal fiber. This soliton is fiber amplified and spectrally filtered in several fiber amplifiers. A simple birefringent pulse shaper generates a flat-top temporal pump pulse profile. Direct amplification of these pulses in large mode area fibers without using a stretcher and compressor provides significantly reduced complexity. For the first time to our knowledge broadband amplification around 800 nm central wavelength is demonstrated in BIB3O6 (BIBO) crystals. The stretched Ti:Sapphire oscillator pulses are amplified up to a pulse energy of 25 μJ. Recompression with a grating compressor yields 50.7 fs pulses with 16.2 μJ pulse energy.
©2009 Optical Society of America
1. Introduction
Today high peak power ultrashort pulses can routinely be generated by Ti: Sapphire based laser systems and have found numerous applications during the last decade especially in fundamental science. However, many applications, for example the generation of high harmonics (HHG), are characterized by low conversion efficiency [1]. Application of coherent soft x - rays to a lensless imaging techniques has been experimentally demonstrated [2], but also suffers from long detector integration times which are currently of the order of hours. An increased laser repetition rate is desirable, but limited due to thermal lensing in the Ti:Sapphire crystal [3]. Cryogenically cooling of the Ti:Sapphire crystal helped to increase the average powers up to 40 W [4], but not further until now. Furthermore the cooling technique is complex and the amplifiers require high power pump lasers with excellent beam quality which are still under development.
In contrast, Yb-based fiber lasers which emit at the 1 μm wavelength region provide enhanced efficiency and can be pumped by low brightness laser diodes. Additionally, Yb-doped fibers are proven to handle high average powers due to reduced thermo-optical distortions. Femtosecond fiber amplifiers based on chirped-pulse amplification with output pulse energies of up to 1 mJ [5] and average powers up to 325 W [6] have been reported recently. Although the small amplification bandwidth of Yb-doped fibers does not allow for high energy sub-100 fs pulse generation [7] they are suitable high power pump lasers for optical parametric amplifiers (OPA). It is well known that OP As, due to their large gain bandwidth in non-collinear geometry can generate ultrashort pulses containing only a few optical cycles [8]. The high gain provided by only a few millimeters of crystal makes nonlinear phase distortions negligible. Energy is conserved in the amplification process and today’s crystal quality provides low residual absorption leading to excellent average power handling capability.
The combination of fiber amplifiers with an OPA for high repetition rate ultrashort pulse generation has been demonstrated and reported in several publications. Sub 20 fs pulses have been generated in the NIR [9], VIS [10] and UV [11], at repetition rates of up to 2 MHz. However the achieved pulse energies were below 1 μJ in these experiments. 52 fs pulses with a pulse energy of 41 μJ and a peak power as high as 500 MW have been generated recently with a degenerated OPA operated at a repetition rate as high as 97 kHz [12]. Even shorter 29 fs pulses with a peak power as high as 2 GW have been generated at a repetition rate of 30 kHz [13]. All of these systems were pumped by a fiber chirped-pulse amplification system which are rather complex. In particular, to achieve the high OPA pump pulse energies of 1 mJ [5] a large stretcher and compressor unit is needed in the pump laser to prevent pulses from nonlinear distortion in the amplifier fibers. In addition, to match the broadband seed pulse duration to those of the pump pulses a second stretcher and a final compressor for ultrashort pulse generation is typically needed.
In this contribution we report on a different approach which generates high energy pump pulses without the need of a stretcher and compressor. A schematic overview of the setup is shown in Fig. 1. The wavelength shifted pulses from a cavity dumped Ti:Sapphire oscillator, delivered by a highly nonlinear photonic crystal fiber, are fiber amplified and spectrally narrowed by fiber Bragg gratings. The bandwidth is set to 15 pm resulting in 100 ps pulses. Additionally, we use birefringent Yttrium Vanadate (YVO) crystals to generate a flat-top like pulse shape. Further amplification up to 400 μJ of pulse energy takes place in two Yb-doped large-mode area photonic crystal fibers. The pulses are frequency doubled in a 10 mm long critically phase matched LiB3O5 (LBO) and then used to pump a two stage OPA based on BIB3O6 (BIBO) crystals. The stretched seed pulses, delivered by the cavity dumped output of the Ti:Sapphire oscillator followed by a conventional grating stretcher, are amplified up to a pulse energy of 25 μJ. Recompression with a grating pair leads to 16.2 μJ pulses with a pulse duration of 50.7 fs.
Fig. 1. Schematic overview of the experimental setup including the cavity dumped Ti:Sapphire oscillator, the highly nonlinear photonic crystal fiber (HNL PCF), the fiber amplifier chain, second harmonic generation (SHG), two optical parametric amplifiers (BIBO crystals) and the grating stretcher and compressor.
2. Generation of synchronized flat-top pump pulses at 1030 nm wavelength
Due to the lack of energy storage, parametric amplifiers require strict temporal synchronization of pump and signal pulse. In principle, it is possible to seed a short pulse Yb-doped fiber amplifier directly with an octave-spanning Ti:Sapphire oscillator. However the pulse energy in the Yb gain region around 1030 nm is normally less than a picojoule. An enhanced frequency conversion process is needed to successfully counteract upcoming amplified spontaneous emission (ASE) in the fiber amplifiers. Extensive spectral broadening and the generation of solitons in the anomalous dispersion wavelength range can be observed in small core photonic crystal fibers with enhanced nonlinearity and tailored dispersion. Once a soliton is generated its center wavelength can be shifted by the soliton self-frequency shift. The generation of temporal synchronized radiation at 1030 nm central wavelength has already been demonstrated [9]. An overview about the experimental setup of the fiber based pump pulse generation and amplification is shown in Fig. 2.
In order to generate the synchronized pump pulses we dechirp the 80 MHz repetition rate pulses delivered by the Ti:Sapphire oscillator with a pair of chirped mirrors and couple them into a 40cm long photonic crystal fiber (Crystal Fibre, PCF NL-3.7-975). The launched pulses have a pulse energy of 0.8 nJ and a pulse duration of less than 20 fs. The resulting spectrum contains approximately 10 pJ in the Yb gain region and is shown in Fig. 3 (red) together with the spectrum of the pulses delivered by the Ti:Sapphire oscillator. The wavelength shifted pulses are amplified in several fiber amplifier stages. At first the pulses are amplified in two step index single mode preamplifiers operated in double pass configuration. The spectral bandwidth is reduced by a fiber Bragg grating at the end of each double pass amplifier. Both amplifiers consist of a circulator, 50 cm active fiber (Nufern PM-YSF-HI), a fiber Bragg grating, providing feedback for the second amplification pass, and a fiber coupled single mode pump diode, which is coupled to the active fiber by a wavelength division multiplexer (WDM). All fiber components of the two preamplifiers are polarization maintaining and spliced together resulting in a stable alignment free setup.
Fig. 2. Setup of the fiber amplifiers serving as pump source in the experiment (PCF - photonic crystal fiber, FBG - fiber Bragg grating (Δλ = 1 nm / 15 pm), POL - polarizer, HWP - half wave plate, OI - optical isolator, AOM - acousto- optical modulator)
The first preamplifier is operated at 200 mW pump power resulting in 3 mW of output power. The spectral bandwidth is reduced by a fiber Bragg grating with a spectral bandwidth of 1.0 nm and a center wavelength of 1030.0 nm. A fiber coupled isolator protects the amplifier from feedback. The pump power of the second preamplifier is set to 130 mW resulting in 10 mW output power. The second fiber Bragg grating was designed to have a spectral bandwidth of 15 pm at 1030.1 nm central wavelength. The spectral bandwidth at the output of the second preamplifier is measured to be below 50 pm which is the resolution limit of the spectrometer. Measurement of the pulse duration with a fast photodiode resulted in a pulse duration of 100 ps. The measured spectra of the Yb amplifier seed (red), after the first preamplifier (black) and after the second preamplifier (blue) are shown in Fig. 4. The corresponding spectral bandwidths are 35 nm (seed), 1 nm (first amplifier) and smaller than 50 pm (second amplifier).
Fig. 3. Measured spectrum of the pulses coupled to the HN PCF (black) and output spectrum including the red shifted spectral components in the Yb gain region (red)
Fig. 4. Measured spectra of the Yb amplifier seed (red), after the first preamplifier (black) and after the second preamplifier (blue)
To generate a flat-top like temporal pulse shape, which is advantageous for the efficiency of the OPCPA [14], the pulses pass through seven birefringent Yttrium Vanadate (YVO) crystals with a total length of 210 mm. Both optical surfaces of the YVO crystals were antireflection-coated with a residual reflectivity of less than 0.1 %. The polarization of the optical pulses is set 45 degrees with respect to the optical axis of the crystal. So the same amount of optical power is polarized ordinary and extraordinary. Both pulse replica experience a temporal delay while propagating through the crystal. The calculated delay per crystal length at 1030 nm wavelength is 0,748 ps/mm, giving a 158 ps delay in our setup. This temporal delay is found to be an optimum between long pulse duration and smooth flat-top pulse profile. Interference of both replica is observed at a polarizer which is orientated 45 degrees to the optical axis and located at the crystals output. In the experiment we control the phase difference of both pulses by controlling the crystal temperature and their relative amplitude with a half wave plate placed in front of the crystals. The typical flat-top like pulse shape with a measured pulse duration of 210 ps is shown in Fig. 5 (red).
These pulses are coupled into a 1.5 m long double clad Yb doped photonic crystal fiber which has a mode field diameter of 33 μm at 1030 nm wavelength. The 976 nm radiation of a 200 μm fiber coupled laser diode is imaged to the 170 μm diameter pump core by a telescope. This fiber amplifier is used in a double pass configuration while a Tellurium Dioxide (TeO2) acousto-optical modulator is used to reduce the repetition rate to 80 kHz after the first pass of the amplifier fiber. An electronic delay generator is used to select the right pulses out of the 80 MHz pulse train for temporal overlap in the OPA later. In the second amplification pass the pulse energy is increased to 1.25 μJ which corresponds to an average power of 100 mW.
The power amplifier consists of a 1.2 m long rod type photonic crystal fiber [15] with an active core diameter of 80 μm. With a 976 nm pump power of 80 W, delivered by a 400 μm diameter fiber (NA=0.22), and launched into the 200 μm diameter pump core an average output power of 32 W is achieved. The measured output spectrum is shown in Fig. 6. Owing to the large mode field diameter (71 μm) of the rod type fiber, spectral broadening due to self phase modulation is low, no stimulated Raman scattering is observed and the spectral bandwidth is increased to only 0.2 nm (FWHM) which makes the pulses suitable for efficient second harmonic generation.
Fig. 5. Temporal pulse shape of the amplifier infrared pulses (red) and the second harmonic pulses (green)
The total amount of energy in ASE and unwanted pre- and post-pulses is measured to be below 5 %, which corresponds to a pulse energy of 380 μJ and a peak power of 1.8 MW. The degree of polarization is slightly reduced in the power amplifier, resulting in about 95 % of the pulse energy can be found in the linear polarization state necessary for the second harmonic generation (SHG). When focusing into a 10 mm long critically phase matched LBO crystal the generation of 16 W of 515 nm pulses is observed. This corresponds to a SH pulse energy of 200 μJ and a conversion efficiency of 53 % with respect to the components which are correctly polarized. The temporal pulse shape of the second harmonic pulses is shown in Fig. 5 (green). The measured pulse duration of the SH pulse is 180 ps. Note that the amplification in the power amplifier prefers the leading edge of the pulse caused by the high extracted energy per pulse. However, experimentally this effect is compensated by slightly changing the relative amplitude of the two pulse replica in the pulse shaper by turning the input polarization with a half wave plate.
4. Optical parametric amplifier design
Several requirements have to be fulfilled by the nonlinear crystal. First of all, it is supposed to support broadband phase-matching located around 800 nm. Furthermore, the required non-collinear angle and the walk off angle should be small or either cancel out to avoid gain limitation caused by the transversal separation. The crystals transparency range should also cover the spectral range of the idler. Lastly, a high nonlinear coefficient and a high damage threshold are required. The most promising candidates are several crystals belonging to the borate group (BBO, LBO, BIBO).
Table. 1. Characteristics of the nonlinear crystals BBO, LBO and BIBO for broadband type I interaction 800 nm + 1446 nm = 515 nm
Table 1 shows the main characteristics of the crystals. The data are taken from the manufacturer and from the freely available software SNLO [16]. Due to the relatively low peak power of our pump pulse a comparison of these crystals in terms of achievable gain per amplification stage is necessary. Firstly, the damage threshold of the nonlinear crystal and especially its coating sets a limit to the pump intensity you can apply to the crystal. Secondly, the angular acceptance of the nonlinear interaction limits the useful crystal length for a given pump focus diameter. The acceptance angle is given by θ=θ*·1 where θ* is the angular acceptance of the crystal and l is the crystal length. For a beam, with the beam quality factor M2 and the central wavelength λ, the focus diameter is given by d=(2·M2-λ)/(π·θ), where θ is the angle of convergence. Using the undepleted pump approximation, for perfect phase matching and high gain (Δk=0, Γ>>1) the parametric gain G is in first order given by G=1/4·exp(2Γ1), where Γ is the nonlinear gain coefficient and L is the crystal length. The nonlinear gain coefficient Γ is expressed as
where deff is the effective nonlinear coefficient, nI, nS and nP are the refractive indices of idler, signal and pump, while λI, λS and λP are the corresponding wavelengths and IP is the pump intensity. Finally, the parametric gain is given by the following equation
It has to be stressed that within this simple approximation for a given pump wavelength λ the maximum parametric gain is only given by the pump pulse peak power PP and the two material constants of the crystal y and θ*. For a given signal wavelength one can compare the achievable gain of different crystal materials.
A third factor, which limits the maximum gain is transversal separation of pump and signal beam. In a non-collinear geometry the signal wave can propagate with an angle of Ω ± α relative to the optical axis, where Ω is the phase matching angle and α is the non-collinear angle (angle between the k-vectors of pump and signal). The separation angle φ between the Poynting-vectors of signal and pump in the crystal is then given by the sum or the difference of the walk-off angle δ (angle between k-vector and Poynting-vector of the extraordinary polarized wave) and non-collinear angle α. The latter case is typically chosen to maximize the interaction length of pump and signal. With a given beam diameter d the interaction length is the limited to l=tan(φ)·d. Within the approximations mentioned before one can calculate a maximum gain for the nonlinear crystal which is limited by the separation angle φ and can be written as
The important crystal parameters are shown in Table 2 together with the calculated gain values. For the calculations a pump wavelength of 515 nm, a pulse peak power of 1 MW and M2=1.3 is assumed. The last row of the table shows the maximum gain which can be achieved with the different crystal materials. This value is given by the minimum of the calculated gain limited by the angular acceptance (fourth row) and the calculated gain limited by the separation angle φ (fifth row). Note that the angular acceptance is limiting the gain in BBO to approximately 250 and the gain in the LBO crystal is limited by the given separation angle to about 160. Applying these crystals in the given case would not only result in limited gain but also in strongly elliptical beam profiles. Fortunately, BIBO crystals have superior properties due to a large angular acceptance combined with a high nonlinear parameter and are therefore the right choice for our setup.
Table 2. Nonlinear parameter, angular acceptance and resulting maximum gain of BBO, LBO and BIBO for type I interaction 800 nm + 1446 nm = 515 nm.
5. Optical parametric chirped pulse amplifier
The experimental setup of the OPCPA is shown in Fig. 7. The seed pulses, with 8 nJ pulse energy, are delivered by the cavity dumped output of the Ti:Sapphire oscillator. A grating stretcher is used to stretch the pulses typically to 120 ps at 60 nm spectral bandwidth. Its throughput efficiency is measured to be 50 %. The parametric amplifiers consist of a 10 mm and a 6 mm long BIBO crystal both cut at θ=169° and φ=90°. The two optical surfaces are protected by P - coating and the output surface is wedged a 1° to avoid multiple internal reflections. The internal angle between pump and signal is chosen to be 3.9° to provide broadband amplification. Temporal overlap of pump and signal is provided by means of an optical delay stage. The first parametric amplifier is driven by 200 μJ pump pulses and delivers 2.5 μJ of output pulses when a pump intensity of 1.0 GW/cm2 is applied (focal spot diameter 350 μm). This corresponds to a gain factor of 625. Due to the very low pump depletion the spatial beam profile of the remaining pump is still sufficient for pumping the second OPA. The pump beam is refocused to a diameter of 370 μm at the second crystal giving a pump intensity of 0.93 GW/cm2. The signal diameter is chosen to be 450 μm which is slightly larger than the pump. The second crystal boosts the signal pulse energy to 25 μJ which corresponds to a gain factor of 10 and a pump to signal conversion of 13 %. The amplified beam is collimated to a beam diameter of 5 mm to prevent the compressor gratings from damage. The far field profile of the amplified beam is shown in Fig. 8.
The spectra measured after the grating compressor are shown in Fig. 9. The black curve is the spectrum of the seed, which is taken at high sensitivity of the spectrometer. The amplified spectra with the first amplifier pumped (blue) and both amplifiers pumped (red) were measured with lower spectrometer sensitivity. Note that you therefore cannot compare the spectral intensity of seed and amplified spectra.
Fig. 9. Measured spectra at the compressor output: Seed spectrum (black), amplified spectrum with one (blue) and both (red) OPAs pumped.
Fig. 10. Autocorrelation trace (green), FT (black dashed) and 4th order Dispersion corrected FT (blue dashed) of the measured spectrum.
The amplified spectral bandwidth is as large as 49 nm limited by the ratio of signal to pump pulse duration. A shorter signal pulse or a longer pump pulse would lead to an increased spectral bandwidth. However in our experiment the minimum signal pulse duration is given by the stretcher and compressor design and the pump pulse duration is mainly set by the last fiber Bragg grating in the second fiber preamplifier and the birefringent pulse shaper which is optimized for a smooth flat-top pulse shape.
Recompression to a measured autocorrelation width of 93 fs is achieved by optimizing the grating distance and the grating angle in the compressor to compensate for second and third order dispersion. A deconvolution factor of 1.83 is found by stretching the Fourier transform (FT) of the measured spectrum with fourth order dispersion. With this the pulse duration can be calculated to 50.7 fs. The measured autocorrelation trace (green), the FT of the measured spectrum (black dashed) and the 4th order dispersion corrected Fourier transform (blue dashed) are shown in Fig. 10. The Fourier limited pulse duration is as short as 26.2 fs. In fact the compressed pulse in our experiment is nearly a factor of two longer and has a reasonable amount of pre- or post-pulses. This is mainly caused by the surface quality of the optics in the stretcher and compressor which has to be improved in future. The compressed pulse energy is measured to be 16.2 μJ, corresponding to a compressor throughput of 65 %. The pulse peak power is estimated to 180 MW with the 4th order dispersion corrected FT.
6. Conclusion and outlook
We demonstrated a new approach of an optical parametric chirped pulse amplifier system driven by a frequency doubled high power fiber amplifier system delivering flat-top like 180 ps long pump pulses with a pulse energy of 200 μJ at 515 nm wavelength. Both the fiber amplifier and the OPA were seeded by the same Ti:Sapphire oscillator. The system delivers 16.2 μJ pulses at a repetition rate as high as 80 kHz. For the first time to our knowledge broadband amplification around 800 nm central wavelength was observed in BIBO crystals with a non-collinear angle of 3.9°. The amplified spectral bandwidth was as large as 49 nm. The amplified pulses could be recompressed to a pulse duration of 50.7 fs, limited by the surface quality of the mirrors in the stretcher and compressor. The pulse peak power is estimated to 180 MW.
Future work will be focused on the generation of a longer 1030 nm pulses allowing for a higher pulse energy of the pump laser. A new Öffner type stretcher design with high surface quality mirrors is in preparation to eliminate the phase distortions on the recompressed pulses. We are convinced that sub 30 fs pulses with a pulse peak power of more than 1 GW can be achieved with these improvements. Further scaling in terms of pulse energy seems in reach if a multiple pump scheme with several fiber amplifiers is applied [17] leading to a unique high power ultrashort pulse source.
Acknowledgements
This work has been partly supported by the German Federal Ministry of Education and Research (BMBF) with project 03ZIK455 ‘onCOOPtics’.
References and links
1. M. Ferray, A. L’Huillier, X. F. Li, L. A. Lompre, G. Mainfray, and C. Manus, “Multiple harmonic conversion of 1064 nm radiation in rare gases,” J. Phys. B 21, L31 (1988). [CrossRef]
2. R. Sandberg, A. Paul, D. Raymondson, S. Hädrich, D. Gaudiosi, J. Holtsnider, R. Tobey, O. Cohen, M. Murnane, H. Kapteyn, C. Song, J.i Miao, Y. Liu, and F. Salmassi, “Lensless diffractive imaging using tabletop coherent high-harmonic soft-X-ray beams,” Phys. Rev. Lett. 99, 098103 (2007). [CrossRef] [PubMed]
3. S. Backus, C. Durfee, M.M. Murnane, and H.C. Kapteyn. “High power ultrafast lasers,” Rev. Sci. Instrum. 69, 1207–1223 (1998). [CrossRef]
4. I. Matsushima, H. Yashiro, and T. Tomie, “10 kHz 40 W Ti:sapphire regenerative ring amplifier,” Opt. Lett. 31, 2066–2068 (2006).http://www.opticsinfobase.org/abstract.cfm?URI=ol-31-13-2066 [CrossRef] [PubMed]
5. F. Röser, T. Eidam, J. Rothhardt, O. Schmidt, D. N. Schimpf, J. Limpert, and A. Tünnermann, “Millijoule pulse energy high repetition rate femtosecond fiber chirped-pulse amplification system,” Opt. Lett. 32, 3495–3497 (2007). http://www.opticsinfobase.org/abstract.cfm?URI=ol-32-24-3495 [CrossRef] [PubMed]
6. T. Eidam, S. Hädrich, F. Röser, E. Seise, T. Gottschall, J. Rothhardt, T. Schreiber, J. Limpert, and A. Tünnermann, “325W Average Power Fiber CPA System delivering sub-400fs pulses", IEEE J. Sel. Top. Quantum Electron. (to be published).
7. R. Paschotta, J. Nilsson, A. Tropper, and D. Hanna, “Ytterbium-doped fiber amplifiers,” IEEE J. Quantum Electron. 33, 1049–1056 (1997). [CrossRef]
8. 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, 2487–2489 (2007) http://www.opticsinfobase.org/abstract.cfm?URI=ol-32-17-2487 [CrossRef] [PubMed]
9. 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, 16729–16736 (2007). http://www.opticsinfobase.org/abstract.cfm?URI=oe-15-25-16729 [CrossRef] [PubMed]
10. C. Schriever, S. Lochbrunner, P. Krok, and E. Riedle, “Tunable pulses from below 300 to 970 nm with durations down to 14 fs based on a 2 MHz ytterbium-doped fiber system,” Opt. Lett. 33, 192–194 (2008). http://www.opticsinfobase.org/abstract.cfm?URI=ol-33-2-192 [CrossRef] [PubMed]
11. C. Homann, C. Schriever, P. Baum, and E. Riedle, “Octave wide tunable UV-pumped NOPA: pulses down to 20 fs at 0.5 MHz repetition rate,” Opt. Express 16, 5746–5756 (2008). http://www.opticsinfobase.org/abstract.cfm?URI=oe-16-8-5746 [CrossRef] [PubMed]
12. J. Rothhardt, S. Hädrich, F. Röser, J. limpert, and A. Tünnermann, “500 MW peak power degenerated optical parametric amplifier delivering 52 fs pulses at 97 kHz repetition rate,” Opt. Express 16, 8981–8988 (2008). http://www.opticsinfobase.ore/abstract.cfm?URI=oe-16-12-8981 [CrossRef] [PubMed]
13. S. Hädrich, J. Rothhardt, F. Röser, T. Gottschall, J. Limpert, and A. Tünnermann, “Degenerate optical parametric amplifier delivering sub 30 fs pulses with 2GW peak power,” Opt. Express 16, 19812–19820 (2008). http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-16-24-19812 [CrossRef] [PubMed]
14. J.A. Fülop, ZS. Major, B. Horvát, F. Tavella, A. Baltuška, and F. Krausz, “Shaping of picosecond pulses for pumping optical parametric amplification,” Appl. Phys. B 87, 79–84 (2007). [CrossRef]
15. J. Limpert, O. Schmidt, J. Rothhardt, F. Röser, T. Schreiber, A. Tünnermann, S. Ermeneux, P. Yvernault, and F. Salin, “Extended single-mode photonic crystal fiber laser,” Opt. Express 14, 2715–2720 (2006). [CrossRef] [PubMed]
16. www.sandia.gov/imrl/X1118/xxtal.htm.
17. A. Dubietis, R. Danielius, G. Tamošauskas, and A. Piskarskas, “Combining effect in a multiple-beam-pumped optical parametric amplifier,” J. Opt. Soc. Am. B 15,1135–1139 (1998). http://www.opticsinfobase.org/josab/abstract.cfm?URI=josab-15-3-1135 [CrossRef]
