Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

High power ultra-widely tuneable femtosecond pulses from a non-collinear optical parametric oscillator (NOPO)

Open Access Open Access

Abstract

We present an ultra-widely tunable non-collinear optical parametric oscillator with an average output power of more than 3 W and a repetition frequency of 34 MHz. The system is pumped by the second harmonic of a femtosecond Yb:KLu(WO4)2 thin-disk laser oscillator. The wavelength of the signal pulse can be rapidly tuned over a wide range from the visible to the NIR just by scanning the resonator length.

©2012 Optical Society of America

1. Introduction

Within the last decades femtosecond optical parametric oscillators (OPOs) have been well studied to generate pulses with narrow bandwidth and a wide tunability range [14], as well as broadband ultra-short pulses down to 15 fs [5]. Commercial Ti:sapphire lasers have been widely used for pumping these systems as they are able to provide up to 3.5 W of pump power in combination with a short pump pulse duration in the 200 fs range. However, the lack of power scalability of these pump lasers is the main limitation of the available OPO output power which is typically in the lower mW-range. As a result, the application of OPO’s is mainly restricted to wavelength regions where no other laser source is available. Recently, novel pump concepts relying on pump sources such as fiber amplifiers and high power bulk-oscillators have been demonstrated to overcome this limitation, leading to much higher output powers of short and tunable pulses in the NIR-spectral range [6,7]. All these systems are designed for wavelengths above 1 µm in a collinear quasi-phase-matching geometry (QPM). The drawback of these systems is in the narrow phase matched spectral bandwidth for the long crystals in combination with the narrow band pump source which limits both, the shortest available pulse durations but also the tunability without changing the QPM poling period of the crystal [8].

In this work, we present for the first time to our knowledge an optical parametric oscillator based on non-collinear critical phase matching geometry (NOPO) which is pumped by a high power frequency doubled Yb:KLu(WO4)2 thin disk laser oscillator to realize high average output powers as well as signal wavelengths spanning from the visible to the near infrared. Ultra broadband bandwidth capabilities of the signal-resonant oscillator can be achieved following the well established concepts of octave spanning Ti:sapphire oscillators [9, 10]. Due to the broadband dispersion management as well as the broadband phase matching, we are able to tune the signal-wavelength within a wide range from 650 nm to 1200 nm without any crystal movement. This opens up the possibility of rapid tuning across the whole spectral range by simply changing the resonator length.

2. Experimental setup

The signal-resonant non-collinear optical parametric oscillator (NOPO) is based on an ultra-broadband oscillator (shown in Fig. 1 ) with 15 bounces on double chirped mirrors (DCM), supporting a spectral range from 600 to 1200 nm [11]. By varying the thickness of an intracavity broadband AR-coated glass block and a pair of BaF2 wedges (W1/W2) or LaK8 prisms in Brewster configuration, the total internal dispersion of the NOPO can be chosen to be slightly negative, zero, or positive. With a piezo stepper stage for the end mirror, the resonator length of 4.4 m can be tuned with sub-wavelength step size. The 2 mm long β-Barium-Borate crystal (BBO), responsible for the parametric gain, is pumped at 515 nm in non-collinear phase matching geometry by the second harmonic of our home-built Yb:KLu(WO4)2 thin-disk laser oscillator, delivering 22 W of average output power at 1030 nm [12], which is actively stabilized by an 2.4 GHz rf signal generator to a fixed repetition rate of 34.21 MHz. With the Fourier limited 500 fs pulses and pulse energies of 0.6 µJ, the second harmonic is generated in a 1.6 mm long LBO crystal to more than 13 W of green power which corresponds to an efficiency of 60%.

 figure: Fig. 1

Fig. 1 Schematic of NOPO-setup. LBO: SHG crystal (1.6 mm), DM: dichroic mirror (HR 1030, AR 515 nm), BD: beam dump, PM: pump mirror, L: lens (f = 60 mm), BBO: nonlinear crystal (2 mm), DCM: double-chirped mirror pairs, FS: AR-coated Fused Silica (10 mm / 16 mm / 29 mm opt. length), W1/W2: wedge pair (BaF2) or prism pair (LaK8), OC: output coupler (T = 15%)

Download Full Size | PDF

The green pump beam is focused by a lens (f = 60 mm) to match the focal size of the NOPO resonator mode inside the BBO crystal cut for Type I phase matching in Brewsters angle for the ordinary polarization direction for signal and idler (the pump suffers 21.6% reflection losses at the crystal surface). The “magic angle” of 2.4° between pump beam and resonator mode is chosen in non-collinear pointing vector walk-off compensation geometry where the ultra-broadband phase-matching from 650 nm to 1.2 µm enables broadband operation as well as wideband tunability without changing phase matching angles.

3. Experimental results

In a first step, the NOPO is characterized in the positive dispersion regime at different center wavelengths in terms of output power and stability. Figure 2 (a) shows the measured power slopes behind the optimum output coupling of 15%. For a center wavelength of 790 nm a maximum signal output power of more than 3 W (90 nJ) and a slope efficiency of 36.5% could be achieved (black curve). This represents a total photon efficiency of 45% and a spatially chirped idler output power of 1.6 W at 1480 nm to the side of the crystal. Tuning the signal output wavelength to 690 nm just by varying the resonator length, the slope efficiency drops to 31.2% (red curve) which is mainly caused by the spectral characteristics of the output coupling mirror.

 figure: Fig. 2

Fig. 2 (a) OPO signal output power vs. pump power (corrected for the 21.6% reflection losses) for two different signal wavelengths. (b) Picked continuously tunable signal spectra by varying the resonator length. The inset shows the calculated Fourier limit and optimized average output power, respectively.

Download Full Size | PDF

These data have been obtained at a moderate average round trip dispersion of approximately 200 fs2 from 10 mm of glass in the resonator. Due to the unavoidable dispersion ripples of the DCM’s, the dispersion curve can reach values close to zero at certain wavelengths. This can lead to pulsing instabilities due to the missing spectral filter (see also section 4). To obtain a continuously clean pulsing behavior over a wide tuning range, the total internal positive dispersion is increased to approximately 800 fs2 by using approximately 16 mm of fused silica inside the resonator. The clean narrow-band pulses in this dispersion regime can be tuned continuously in the range between 680 and 850 nm. The spectrum can be tuned further up to 950 nm, where we observe unstable pulsing (see below). In Fig. 2 (b) five selected output spectra are plotted on a log scale. The legend reveals the maximum signal output power for each resonator length as well as the Fourier limited pulse duration. (Note, that this maximum power is achieved after a slight realignment of one axis of the end mirror. The output power without any alignment with a higher internal dispersion is shown in Fig. 4(b) .)

 figure: Fig. 4

Fig. 4 Tunability of the signal wavelength by changing the OPO resonator length without any realignment.(a) Signal spectra for different amounts of internal net dispersion. The black curves reveal the calculated internal group delay (1) −800 fs2 no glass, (2) 800 fs2 FS, (3) 2000 fs2 FS, (4) 2000 fs2 LaK8, (5) 3500 fs2 LaK8, (6) 5000 fs2 LaK8 - approximate values at 800 nm. (b) Signal spectra of curve (4) measured by a CCD based spectrometer (grey) and a scanning spectrometer (green) and the corresponding signal output power (red curve).

Download Full Size | PDF

Beside the spectral characteristics of the output coupler, which is responsible for the lower limit of the tuning range at 680 nm, the drop in output power is mainly caused by the phase-matching conditions which were aligned for maximum output power at 800 nm by choosing the optimum walk-off compensation geometry. The limitation for stable tuning above 850 nm is mainly caused by the decreasing positive internal dispersion for higher wavelengths and the parasitic frequency doubling of the signal and idler wave [13]. With a flattened intracavity dispersion curve (employing e.g. a glass block with higher zero dispersion wavelength, e.g. SF59), and by using the non-collinear tangential phase matching geometry [14], a broader tunability with almost constant output power should be feasible.

The chosen amount of dispersion inside the oscillator leads to spectral bandwidths of about 11 nm capable of generating transform-limited pulse durations between 65 and 85 fs. The pulses have been compressed to less than 90 fs by a SF10 prism sequence with an apex distance optimized for the respective center wavelength. The deviation to the Fourier limit as well as the need for the adaptation of the apex distance for different center wavelengths is mainly attributed to the large amount of third order dispersion from the compressor itself. By using e.g. a grism compressor, peak powers up to the MW range are anticipated across the whole tuning range without compressor adaptation.

In order to verify stable pulsing over the whole tuning range, measurements of the radio-frequency spectrum up to 25 GHz have been performed. In Fig. 3 (a) the radio frequency analysis of the fundamental repetition rate of the output pulse train is plotted for the operation with 690 nm (red curve) and 780 nm (black curve) center wavelength. The inset shows a zoom with 1 Hz resolution. More than 80 dB of noise suppression verify the stable operation. On the right hand side in Fig. 3 the long-term-stability during one hour is plotted. In addition to the pump, also the length of the NOPO was actively stabilized to a fixed wavelength with the Piezo stepper. The slightly decreasing output power from 3.3 to 3.1 W (red curve) is caused by long-term power drift of the pump laser. Note, that independent from the NOPO detuning the repetition rate always matches the pump. An almost constant RF signal has shown the perfect synchronization of the NOPO repetition rate to the pump laser.

 figure: Fig. 3

Fig. 3 (a) Radio-frequency analysis of the fundamental repetition frequency for two different signal wavelengths with 10 Hz resolution. The inset shows the zoom for 780 nm with 1 Hz resolution. (b) Signal output power (red). The pump thin disk laser is running with actively controlled repetition frequency. The drop in the signal output power is related to the pump power.

Download Full Size | PDF

4. Influence of intra-cavity dispersion

The results presented in the last section have been obtained with an intra-cavity dispersion of approximately 800 fs2 which corresponds to a group delay of 2.4 fs/nm or 2.8 nm wavelength shift per 1 µm resonator length detuning respectively at 800 nm center wavelength. In Fig. 4 (a) a complete map of measured output spectra in dependence of the resonator length detuning is depicted for negative internal dispersion of −800 fs2 without additional dispersion (beside the BBO and the DCM’s), 800 fs2 and for 2000 fs2 by using the two blocks of fused silica, and 2000 fs2, 3500 fs2, and 5000 fs2 of positive dispersion by using the LaK8 prisms. The respective spectral power density on a logarithmic scale is indicated in the color map and shows an almost constant spectral bandwidth and continuous tuning of the central wavelength over a range from 650 up to 850 nm. The calculated intracavity group delay is plotted for comparison as solid curves. The small residual ripples are caused by the higher order dispersion-structure of the double chirped mirror pairs [15].

For the steepest curves (1, 2) with the lowest dispersion spectral splitting can be observed for signal-wavelengths above 850 nm. This is attributed to the decreasing dispersion of fused silica at longer wavelengths together with the dispersion-structure of the DCM’s which leads to zero points and unstable pulsing for certain wavelengths above 850 nm. In contrast, with the larger dispersion, a clean tuning behavior up to 1200 nm can be observed, as the relative contribution of the DCM’s decreases.

The relative signal output power of the curve (4) is shown in Fig. 4 (b). The grey shaded spectra are cross sections from the left figure with spectral widths around 20 nm. Because of the lower sensitivity of the used CCD based spectrometer at longer wavelengths, spectra measured by a scanning spectrometer are added for wavelengths above 1 µm (green shaded spectra). As already mentioned, the tuning is achieved by just moving the end mirror. Due to the instantaneous response time of the gain medium, this allows for high speed sweeping of the central wavelength over an very large wavelength range with high output power and sub-100 fs pulses. As there is no need for an adaptation of phase matching angle, the speed of tuning is just limited by the Piezo driven end mirror. Preliminary measurements with standard components easily allowed for high speed sweeping of 1 kHz.

5. Conclusion and outlook

We introduced the NOPO, a very stable, ultra wide and rapidly tunable non-collinear optical parametric oscillator with average powers of more than 3 W and a tunability range between 650 nm and 1200 nm. During tuning, the NOPO repetition rate stays fixed to the repetition rate of the pump laser, independently from the NOPO resonator length. With a stabilized pump laser system, the spectrum can be tuned rapidly with an excellent noise behavior. This allows for high speed pump probe experiments, making the NOPO a powerful tool for coherent spectroscopy with very high spectral power density. Furthermore, the concept features power scalability just by increasing the pump power and adapting the spot sizes inside the BBO.

Acknowledgments

The authors thank VENTEON Femtosecond Laser Technologies GmbH for supporting the development and assembling of the oscillator and Dr. Daniel Steingrube for the helpful support with the evaluation software. This work was funded by Deutsche Forschungsgemeinschaft within the Cluster of Excellence QUEST, Centre for Quantum Engineering and Space-Time Research.

References

1. D. T. Reid, G. T. Kennedy, A. Miller, W. Sibbett, and M. Ebrahimzadeh, “Widely tunable, near- to mid-infrared femtosecond and picosecond optical parametric oscillators using periodically poled LiNbO3 and RbTiOAsO4,” IEEE J. Sel. Top. Quantum Electron. 4(2), 238–248 (1998). [CrossRef]  

2. K. V. Bhupathiraju, A. D. Seymour, and F. Ganikhanov, “Femtosecond optical parametric oscillator based on periodically poled stoichiometric LiTaO3 crystal,” Opt. Lett. 34(14), 2093–2095 (2009). [CrossRef]   [PubMed]  

3. J. D. Rowley, S. Yang, and F. Ganikhanov, “Power and tuning characteristics of a broadly tunable femtosecond optical parametric oscillator based on periodically poled stoichiometric lithium tantalate,” J. Opt. Soc. Am. B 28(5), 1026–1036 (2011). [CrossRef]  

4. M. Ghotbi, A. Esteban-Martin, and M. Ebrahim-Zadeh, “Tunable, high-repetition-rate, femtosecond pulse generation in the ultraviolet,” Opt. Lett. 33(4), 345–347 (2008). [CrossRef]   [PubMed]  

5. G. M. Gale, M. Cavallari, and F. Hache, “Femtosecond visible optical parametric oscillator,” J. Opt. Soc. Am. B 15(2), 702–714 (1998). [CrossRef]  

6. F. Adler, K. C. Cossel, M. J. Thorpe, I. Hartl, M. E. Fermann, and J. Ye, “Phase-stabilized, 1.5 W frequency comb at 2.8-4.8 microm,” Opt. Lett. 34(9), 1330–1332 (2009). [CrossRef]   [PubMed]  

7. T. P. Lamour, L. Kornaszewski, J. H. Sun, and D. T. Reid, “Yb:fiber-laser-pumped high-energy picosecond optical parametric oscillator,” Opt. Express 17(16), 14229–14234 (2009). [CrossRef]   [PubMed]  

8. R. Hegenbarth, A. Steinmann, G. Tóth, J. Hebling, and H. Giessen, “Two-color femtosecond optical parametric oscillator with 1.7 W output pumped by a 7.4 W Yb:KGW laser,” J. Opt. Soc. Am. B 28(5), 1344–1352 (2011). [CrossRef]  

9. S. Rausch, T. Binhammer, A. Harth, J. Kim, R. Ell, F. X. Kärtner, and U. Morgner, “Controlled waveforms on the single-cycle scale from a femtosecond oscillator,” Opt. Express 16(13), 9739–9745 (2008). [CrossRef]   [PubMed]  

10. T. Binhammer, S. Rausch, M. Jackstadt, G. Palmer, and U. Morgner, “Phase-stable Ti:sapphire oscillator quasi-synchronously pumped by a thin-disk laser,” Appl. Phys. B 100(1), 219–223 (2010). [CrossRef]  

11. F. X. Kärtner, U. Morgner, R. Ell, T. Schibli, J. G. Fujimoto, E. P. Ippen, V. Scheuer, G. Angelow, and T. Tschudi, “Ultrabroadband double-chirped mirror pairs for generation of octave spectra,” J. Opt. Soc. Am. B 18(6), 882–885 (2001). [CrossRef]  

12. G. Palmer, M. Schultze, M. Siegel, M. Emons, U. Bünting, and U. Morgner, “Passively mode-locked Yb:KLu(WO4)2 thin-disk oscillator operated in the positive and negative dispersion regime,” Opt. Lett. 33(14), 1608–1610 (2008). [CrossRef]   [PubMed]  

13. J. Bromage, J. Rothhardt, S. Hädrich, C. Dorrer, C. Jocher, S. Demmler, J. Limpert, A. Tünnermann, and J. D. Zuegel, “Analysis and suppression of parasitic processes in noncollinear optical parametric amplifiers,” Opt. Express 19(18), 16797–16808 (2011). [CrossRef]   [PubMed]  

14. A. L. Oien, I. T. McKinnie, P. Jain, N. A. Russell, D. M. Warrington, and L. A. W. Gloster, “Efficient, low-threshold collinear and noncollinear β-barium borate optical parametric oscillators,” Opt. Lett. 22(12), 859–861 (1997). [CrossRef]   [PubMed]  

15. U. Morgner, F. X. Kärtner, S. H. Cho, Y. Chen, H. A. Haus, J. G. Fujimoto, E. P. Ippen, V. Scheuer, G. Angelow, and T. Tschudi, “Sub-two-cycle pulses from a Kerr-lens mode-locked Ti:sapphire laser,” Opt. Lett. 24(6), 411–413 (1999). [CrossRef]   [PubMed]  

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (4)

Fig. 1
Fig. 1 Schematic of NOPO-setup. LBO: SHG crystal (1.6 mm), DM: dichroic mirror (HR 1030, AR 515 nm), BD: beam dump, PM: pump mirror, L: lens (f = 60 mm), BBO: nonlinear crystal (2 mm), DCM: double-chirped mirror pairs, FS: AR-coated Fused Silica (10 mm / 16 mm / 29 mm opt. length), W1/W2: wedge pair (BaF2) or prism pair (LaK8), OC: output coupler (T = 15%)
Fig. 2
Fig. 2 (a) OPO signal output power vs. pump power (corrected for the 21.6% reflection losses) for two different signal wavelengths. (b) Picked continuously tunable signal spectra by varying the resonator length. The inset shows the calculated Fourier limit and optimized average output power, respectively.
Fig. 4
Fig. 4 Tunability of the signal wavelength by changing the OPO resonator length without any realignment.(a) Signal spectra for different amounts of internal net dispersion. The black curves reveal the calculated internal group delay (1) −800 fs2 no glass, (2) 800 fs2 FS, (3) 2000 fs2 FS, (4) 2000 fs2 LaK8, (5) 3500 fs2 LaK8, (6) 5000 fs2 LaK8 - approximate values at 800 nm. (b) Signal spectra of curve (4) measured by a CCD based spectrometer (grey) and a scanning spectrometer (green) and the corresponding signal output power (red curve).
Fig. 3
Fig. 3 (a) Radio-frequency analysis of the fundamental repetition frequency for two different signal wavelengths with 10 Hz resolution. The inset shows the zoom for 780 nm with 1 Hz resolution. (b) Signal output power (red). The pump thin disk laser is running with actively controlled repetition frequency. The drop in the signal output power is related to the pump power.
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.