We demonstrate optical parametric chirped pulse amplification (OPCPA) at 1-kHz repetition rate in periodically poled stoichiometric LiTaO3 (PPSLT) with 1 mol % MgO doping. Diode pumping was used both for the fiber laser generating the femtosecond seed pulses and the nanosecond laser/amplifier employed as a pump source. The high gain (≈63 dB) and large bandwidth (20 nm) obtained for single-stage non-degenerate OPCPA operation provide a compact and efficient solution for amplification of ultrashort pulses near 1.57 µm. Stretched femtosecond pulses could be amplified up to 39.5 µJ in a 7-mm long, 2-mm thick sample of PPSLT. The pulse duration of the amplified signal pulses (FWHM) after recompression amounted to 315 fs.
©2004 Optical Society of America
High-energy, high-intensity ultrashort laser pulses gain in importance in many areas including ultrafast spectroscopy, surgery, imaging and material processing. The combination of the chirped pulse amplification (CPA) concept with optical parametric amplification (OPA), originally demonstrated in , offers a simple and attractive alternative for amplification of ultrashort pulses to high energies. Using this so-called optical parametric chirped pulse amplification (OPCPA) technique, the high pulse energy from powerful nano- or picosecond lasers/amplifiers can be efficiently transferred to temporally stretched femtosecond pulses through an instantaneous optical parametric conversion process in a nonlinear medium. Compared to the widespread conventional regenerative/multi-pass femtosecond laser amplifier systems, OPCPA exhibits several advantages. These include large gain bandwidth, high gain achievable without the use of gated electro-optic modulators and multi-pass amplification, lack of cumulative spectral narrowing, low nonlinear phase distortion and B-integral accumulation, high contrast ratio, low heat deposition and larger wavelength flexibility due to the availability of different high-quality nonlinear crystals. The critical factor for the realization of efficient femtosecond OPCPA is related to the availability of ultrafast (femtosecond) seed oscillators and powerful pump sources delivering pulses with duration in the order of the stretched seed pulse length (0.5–5 ns). Conventional actively Q-switched lasers, especially those operating at kHz-repetition rates, do not satisfy this requirement which limits the conversion efficiency and can lead to thermal problems.
A recently published paper  gives an excellent overview of the progress in OPCPA in the last decade. In short, three basic trends can be observed. The first one, directed towards extremely high-power pulses (see Table 1 in ) had the objective of achieving TW peak powers at low repetition rates (<10 Hz). Such schemes were operated collinearly and at degeneracy, and were pumped by the second harmonic of Nd-doped YAG, YLF and glass lasers (amplifiers) at 532 or 527 nm, respectively. Birefringent crystals of LBO, BBO and KDP, having highest damage thresholds, were used in the parametric amplifier stages. The highest output power achieved so far amounts to 16.7 TW corresponding to 120-fs long pulses of 2 J energy at 1064 nm . The second trend, towards ultimately short and tunable pulses, is related to the concept of noncollinear parametric amplification and pumping with shorter, typically 150 fs long, pulses. These systems, based exclusively on BBO, are extensively discussed in  in relation with Table 2 and culminated in the generation of laser pulses as short as 4 fs at 600 nm. The third trend is related to the development of more compact and reliable systems operating at kHz repetition rates having both power and pulse duration at the intermediate level. Such systems were designed as all-diode-pumped and/or fiber based alternatives of commercially available Ti:sapphire based regenerative amplifiers but operate at different wavelengths. Periodically poled (PP) ferroelectric crystals for quasi-phase-matching (QPM) seem predestined for application in these systems since on the one hand they ensure through their superior effective nonlinearity substantial improvement of the parametric gain as compared to conventional birefringent crystals and on the other hand their limited thickness sets an upper limit for the pulse energy in order to avoid optical damage. The first application of PP LiNbO3 (PPLN) in a quasi-degenerate OPCPA scheme resulted in 0.68 ps long pulses of 0.6 mJ energy after compression, at 1.56 µm and repetition rate of 10 Hz . The same material was used later in a non-degenerate scheme producing, again at 1.56 µm, pulses of 1.6 ps duration and 1.08 mJ energy at 1 kHz [5–6]. We were the first to apply the more resistant to damage PP KTiOPO4 (PPKTP) in a non-degenerate OPCPA scheme  achieving a gain of 55 dB at the signal wavelength. The recompressed pulses at 1.57 µm had a FWHM of 320 fs. Subsequently PPKTP has been used as a preamplifier stage for a high-power OPCPA system  and also in a double stage OPCPA scheme .
Recently a new material very suitable for electric field poling, stoichiometric LiTaO3 or SLT, emerged . Compared to LiNbO3 or congruent LiTaO3 (CLT), SLT exhibits about an order of magnitude lower coercive field which is comparable or even lower than that of KTP . This is essential for the fabrication of thick QPM devices for high-power applications. Additional advantages of SLT include high damage threshold, low thermo-optic coefficients and wide transmission [11, 12]. The nonlinear coefficients of SLT are larger than those of CLT . From the d33 value of 13.8 pm/V measured for CLT at 1064 nm by the Maker fringe technique , it can be concluded that d33(SLT)≈1.5…2 d33(KTP). Recent measurements based on phase-matched second-order nonlinear processes in PPSLT  and PPKTP  indicated effective nonlinearities of 10 pm/V and 8 pm/V, respectively, or deff(PPSLT)=1.25 deff(PPKTP). In this work we present what we believe to be the first application of PPSLT in a 1 kHz OPCPA scheme. The single-stage amplification leading to signal pulse energy of 39.5 µJ corresponds to a parametric gain of 1.93×106 which is roughly 6 times higher than previously achieved with PPKTP . Due to this high gain, PPSLT crystals can be used for compact high-power ultrashort pulse amplifier systems in the near infrared.
2. Experimental setup
The PPSLT samples used in the present experiment were uncoated and had an aperture of 2.0 mm (thickness) ×4.8 mm (width). Samples with 7 and 10-mm length and QPM periods of 30.7 and 31.0 µm were available. They were used at room temperature. The temperature dependence of the QPM period for two relevant seed wavelengths and pumping at 1064 nm can be seen in Fig. 1. The calculation is based on the temperature-dependent Sellmeier dispersion relations from .
The SLT was grown by the double-crucible Czochralski method with 1.0 mol % MgO-doping . Owing to the low coercive field a lot of random micro-domains were formed in the as-grown SLT wafer. These initial micro-domains could become nucleation sites when electric field is applied for polarization reversal. To avoid this, a cleaning process was performed by applying a higher electric field (~2.5 kV/mm) than the coercive field prior to the actual poling. Periodically patterned photoresist covered by a metal film was formed then on the +Z surface of a 2-mm thick SLT wafer of 2” diameter. A single-pulse electric field of 2.1 kV/mm for 1.8 s was subsequently applied for the poling. Previous studies revealed a smooth profile of the domain boundaries for PPSLT with the same thickness .
The experimental scheme of the single stage PPSLT-based OPCPA (Fig. 2) is similar to that previously described in . As a seed source, a diode-pumped passively mode-locked Er3+-all-fiber oscillator operating near 1.58 µm, with a spectral bandwidth of ≈80 nm (FWHM), was used. It delivered 55-fs long output pulses with 14-mW average power at 56 MHz. The pulses from the oscillator were expanded to 250 ps in an all-reflective single grating (600 l/mm) stretcher containing an f=30” parabolic mirror. Spectral clipping in one of the stretcher retroreflectors reduced the useful bandwidth down to ≈50 nm. This, however, played no role in the present study since the bandwidth of the pulses at the OPCPA output was determined by the narrower gain bandwidth of the PPSLT samples. Depending on the alignment, different spectral portions of the oscillator spectrum could be selected for seeding of the OPCPA at the signal wavelength.
Pump pulses at 1064 nm were available from a commercial diode-pumped 1.5-mJ Nd:YAG amplifier (High Q-Laser) seeded by an actively Q-switched 1-ns Nd:YVO4 microlaser (ECR Corp.) at 1 kHz. The output beam of the amplifier had an M2-parameter of <1.2. Both the Pockels cell of the Nd:YAG regenerative amplifier and the Q-switched microlaser were synchronously triggered by a signal obtained from the Er-fiber laser using a frequency divider and a delay generator. The average power used for seeding at the signal wavelength amounted to 1.5 mW which corresponds to a single pulse energy of 27 pJ. The seed and the pump pulses were collinearly combined by a suitable dichroic mirror prior to entrance into the PPSLT. The pump beam diameter (2w) in the position of the PPSLT samples was 0.5 mm.
The energy of the amplified signal pulses was measured by a calibrated pyroelectric detector after elimination of undesired light such as high-order phase-matched second harmonic of the pump pulses and spontaneous parametric fluorescence in the visible spectral range. The amplified signal pulse spectrum was recorded by a 0.5-m monochromator and the pulse duration, after recompression, was estimated by autocorrelation measurements using second harmonic generation in a 1-mm thick type-I BBO crystal.
3. Results and discussion
Prior to the OPCPA experiment, two 7 and 10 mm long PPSLT dummy samples from the same wafer were tested with intensive pump pulses to estimate the damage thresholds. Photorefraction and green induced infrared absorption are known as damage mechanisms for LiTaO3. It is also known that MgO-doping of photorefractive crystals like LiTaO3 and LiNbO3 in general increases their photorefractive damage resistivity by several orders of magnitude . With the 1-ns long pulses at 1064 nm we did not observe any damage in the 7 mm long SLT sample up to peak-on-axis pump intensities of 750 MW/cm2. However, for the 10-mm long sample, at a level of ≈500 MW/cm2, we observed damage of the rear surface in the presence of green light produced by non-phase-matched SHG of the pump pulses.
In Fig. 3 we show the dependence of the energy of the amplified signal pulse on the peak on-axis pump intensity for the 7 and 10-mm long PPSLT samples with the grating period of 30.7 µm. Signal pulse energies of 39.5 µJ and 45 µJ were obtained for an incident pump energy of 600 µJ with these two crystals, respectively. We established that the limiting factor for the conversion efficiency is set by the low (27 pJ) seed pulse energy which is not sufficient for suppression of the unseeded spontaneous parametric fluorescence if larger beam sizes were used. Unseeded parametric fluorescence lead to deterioration of the spectral quality of the amplified pulse and did not allow to properly recompress it. This undesired effect was much stronger in the longer PPSLT crystal. That is why for optimization and further characterization of the spectral and temporal properties of the amplified pulses we used only the 7-mm long PPSLT samples.
At the maximum pump intensity applied (see Fig. 3) the unwanted parametric fluorescence in the case of the 7-mm long PPSLT sample of 30.7 µm period could be minimized to less than 30%, measured with seed interrupted, by careful optimization of overlap and beam sizes of the pump and the seed pulses. The amplified signal pulse energy produced in the seeded case corresponded to a parametric gain of 1.93×106 taking into account the Fresnel losses at the uncoated crystal faces. This is, to our knowledge, the highest parametric gain achieved with single-stage femtosecond OPCPA up to now. Direct comparison with simple analytical expressions for the parametric gain is not possible because of the spatial and temporal dependence of the gain saturation and the limited gain bandwidth, but calculations for the small signal and plane-wave case neglecting the difference in the group velocities indicate that the d33 coefficient of PPSLT should be larger than the value derived from SHG measurements .
The recorded signal spectrum exhibited a well-defined shape, and the spectral bandwidth agreed well with the theoretically calculated spectral gain bandwidth of 20 nm for a 7-mm long PPSLT sample. Fig. 4 shows the amplified signal spectra with the two 7-mm long PPSLT samples of different periods recorded at a peak on-axis pump intensity of 610 W/cm2. As can be seen, given the broad seed spectrum, the signal pulses could be amplified in 30.7 µm as well as in 31.0 µm QPM-period PPSLT and the output spectra were centered near 1.56 µm and 1.59 µm, respectively. In terms of output energy, we measured slightly higher (by 5%) values with the 30.7 µm PPSLT. This result could be expected from Fig. 1 since the period of 30.7 µm is obviously better suited for QPM.
In addition to the amplified signal pulses, the present non-degenerate OPCPA scheme generated simultaneously idler pulses near 3.34 µm with an energy as high as 18.5 µJ. The total amplified energy (signal+idler) amounted to 58 µJ, which corresponds to an overall internal conversion efficiency of 14.6 %. The conversion efficiency in the present experiment is increased roughly 4 times compared to the previously demonstrated single-stage OPCPA with PPKTP . The increase is attributed to the higher effective nonlinearity of PPSLT. Comparing now with a PPKTP sample of exactly the same length (7 mm) under identical conditions we verified that the actual improvement is 2–3 times depending on the saturation level. The transmission of the compressor containing an analogous grating amounted to 65 %.
Deconvolution of the autocorrelation trace in Fig. 5, assuming Gaussian pulse shape, leads to a pulse length (FWHM) of 315 fs, which corresponds to a pulsewidth-bandwidth product of 0.77. The imperfect recompression, related to higher order dispersion, and the deviation from the Fourier limit are caused mainly by the contribution of not fully suppressed unseeded spontaneous parametric fluorescence.
In conclusion, compact 1-kHz femtosecond OPCPA was demonstrated with PPSLT for the first time. It permitted the highest gain to be achieved by single-pass femtosecond OPCPA. The output energy could be further increased by applying antireflection coatings to the PPSLT crystal for the pump, signal and idler wavelengths. Non-degenerate OPCPA with PPSLT in the present scheme offers the attractive feature of simultaneous availability of idler pulses in the mid-IR. In the present experiment it was the low seed energy that required relatively tight focusing of both the seed and pump beams using only a part of the available PPSLT thickness and pump energy. The increase of the seed energy turns out to be essential for the suppression of spontaneous parametric fluorescence in order to ensure simultaneously high conversion efficiency and proper recompression to the initial pulse duration using only one amplification stage. Work is in progress to solve this problem by increasing the power of the Er-fiber oscillator. This will allow to fully utilize the large aperture size of the available PPSLT for the generation of yet higher output energies using the total available (1.5 mJ) pump power and/or a second amplification stage.
This work was supported in part by the Ministry of Science and Technology of Korea through the Strategic National R & D program (Grant No. M1-0330000001-04G0900-00112).
1. A. Dubieties, G. Jonusauskas, and A. Piskarskas, “Powerful femtosecond pulse generation by chirped and stretched pulse parametric amplification in BBO crystal,” Opt. Commun. 88, 437–440 (1992). [CrossRef]
2. R. Butkus, R. Danielius, A. Dubietis, A. Piskarskas, and A. Stabinis, “Progress in chirped pulse optical parametric amplifiers,” Appl. Phys. B 79, 693–700 (2004). [CrossRef]
3. L. H. Lin, Z. Z. Xu, X. D. Yang, R. X. Li, H. H. Lu, W. Y. Wang, Y. X. Leng, Z. Q. Zhang, Y. H. Jiang, S. Q. Jin, D. J. Ying, and W. Q. Zhang, “Recent progress in table-top multiterawatt laser systems at SIOM,” CLEO/Pacific Rim 2003. The 5th Pacific Rim Conference on Lasers and Electro-Optics, Proceedings Vol. I, Taipei, Taiwan, Dec. 15–19, 2003, p. 356.
4. A. Galvanauskas, A. Hariharan, D. Harter, M. A. Arbore, and M. M. Fejer, “High-energy femtosecond pulse amplification in a quasi-phase-matched parametric amplifier,” Opt. Lett. 23, 210–212 (1998). [CrossRef]
5. A. Galvanauskas, A. Hariharan, F. Raksi, K. K. Wong, D. Harter, G. Imeshev, and M. M. Fejer, “Generation of diffraction-limited femtosecond beams using spatially-multimode nanosecond pump sources in parametric chirped pulse amplification systems,” in Conference on Lasers and Electro Optics, Paper CThB4, Technical Digest (Optical Society of America, Washington, D.C., 2000), pp. 394–395.
6. A. Galvanauskas, A. Hariharan, and D. Harter, “Diode pumped parametric chirped pulse amplification system with mJ output energies,” in Trends in Optics and Photonics Vol. 43, Twelfth International Conference on Ultrafast Phenomena, Paper WE6-1, Technical Digest (Optical Society of America, Washington, DC, 2000), pp. 617–619.
7. F. Rotermund, V. Petrov, F. Noack, V. Pasiskevicius, J. Hellström, F. Laurell, H. Hundertmark, P. Adel, and C. Fallnich, “Compact all-diode-pumped femtosecond laser source based on chirped pulse optical parametric amplification in periodically poled KTiOPO4,” Electron. Lett. 38, 561–563 (2002). [CrossRef]
8. I. Jovanovic, J. R. Schmidt, and C. A. Ebbers, “Optical parametric chirped-pulse amplification in periodically poled KTiOPO4 at 1053 nm,” Appl. Phys. Lett. 83, 4125–4127 (2003). [CrossRef]
9. V. Petrov, F. Noack, F. Rotermund, V. Pasiskevicius, A. Fragemann, F. Laurell, H. Hundertmark, P. Adel, and C. Fallnich, “Eficient all-diode-pumped double stage femtosecond optical parametric chirped pulse amplification at 1-kHz with periodically poled KTiOPO4,” Jpn. J. Appl. Phys. 42, L1327–L1329 (2003). [CrossRef]
10. T. Hatanaka, K. Nakamura, T. Taniuchi, H. Ito, Y. Furukawa, and K. Kitamura, “Quasi-phase-matched optical parametric oscillation with periodically poled stoichiometric LiTaO3,” Opt. Lett. 25, 651–653 (2000). [CrossRef]
11. K. Kitamura, Y. Furukawa, K. Niwa, V. Gopalan, and T. E. Mitchell, “Crystal growth and low coercive field 180° domain switching characteristics of stoichiometric LiTaO3,” Appl. Phys. Lett. 73, 3073–3075 (1998). [CrossRef]
12. Y. Furukawa, K. Kitamura, E. Suzuki, and K. Niva, “Stoichiometric LiTaO3 single crystal growth by double crucible Czochralski method using automatic powder supply system,” J. Cryst. Growth 197, 889–895 (1999). [CrossRef]
13. I. Shoji, T. Kondo, A. Kitamoto, M. Shirane, and R. Ito, “Absolute scale of second-order nonlinear-optical coefficients,” J. Opt. Soc. Am. B 14, 2268–2294 (1997). [CrossRef]
14. N. E. Yu, S. Kurimura, Y. Nomura, and K. Kitamura, “Stable high-power green light generation with thermally conductive periodically poled stoichiometric lithium tantalate,” Jpn. J. Appl. Phys. 43, L1265–L1267 (2004). [CrossRef]
15. J. Hellström, V. Pasiskevicius, H. Karlsson, and F. Laurell, “High-power optical parametric oscillation in large-aperture periodically poled KTiOPO4,” Opt. Lett. 25, 174–176 (2000). [CrossRef]
16. A. Bruner, D. Eger, M. B. Oron, P. Blau, M. Katz, and S. Ruschin, “Temperature-dependent Sellmeier equation for the refractive index of stoichiometric lithium tantalate,” Opt. Lett. 28, 194–196 (2003). [CrossRef] [PubMed]
17. N. E. Yu, S. Kurimura, Y. Nomura, M. Nakamura, K. Kitamura, J. Sakuma, Y. Otani, and A. Shiratori, “Periodically poled near-stoichiometric lithium tantalate for optical parametric oscillation,” Appl. Phys. Lett. 84, 1662–1664 (2004). [CrossRef]