We report a continuous-wave singly resonant optical parametric oscillator pumped by a widely tunable titanium-doped sapphire ring laser. It produces up to 0.8 W of mid-infrared power. The wavelength can be tuned in a few seconds from 2.5 to 3.5 µm or from 3.4 to 4.4 µm and scanned up to 40 GHz without mode-hops by only changing the pump beam wavelength. Spectroscopic capability is demonstrated by measuring parts of the photoacoustic absorption spectrum of NH3 near 3196 cm−1.
© 2010 OSA
A continuous-wave (cw) optical parametric oscillator (OPO)  produces coherent light through frequency down-conversion. The highest power levels and best conversion efficiencies are reached in a singly resonant OPO (SRO) . Such a device is useful for high-resolution spectroscopy in the mid-infrared, where the availability of lasers is limited. In a cw SRO a pump beam generates two new beams: a near-infrared signal beam that resonates inside the OPO cavity and a mid-infrared idler beam that immediately exits the device and is available for applications. The total energy of the two newly created photons equals to the energy of the corresponding pump photon. Therefore, to adjust the frequency of the idler beam, it is sufficient to change the frequency of either the signal or the pump beam.
Previously, either tuning range and/or rate have limited the usefulness of a cw SRO. While scanning with a tunable pump source, fast mode-hop free modulation may be achieved but only for a narrow range. A cw SRO can be pumped by a tunable source such as a multiwatt amplified fiber or semiconductor laser, but their mode-hop free tuning ranges are at best a couple of hundred GHz [3,4]. Wide tuning involves altering the frequency of the resonating signal beam via mode-hops. The extremes of the tuning range are determined by the poling period structure and the optical properties of the nonlinear crystal where down-conversion occurs . The poling period can be altered for even wider tuning range. This is achieved by changing the temperature of the nonlinear crystal or mechanically moving the crystal so that the beams pass through a section with a new poling period. In practice, both methods provide an extensive tuning range, but are slow when compared to adjustments in the pump laser frequency causing little or no changes in the OPO cavity.
In this work, we report a cw SRO system that is pumped by a single-mode cw titanium-doped sapphire (Ti:Al2O3) ring laser. Earlier, such OPO devices have been multi-resonant [5,6], which increases complexity. As far as we know, our setup is the first truly singly resonant, continuous-wave SRO pumped by a widely tunable Ti:Al2O3 laser. The SRO can be operated as a passive system except for an accurate temperature stabilization of the nonlinear crystal. It has an extensive tuning range of about either 2.5 to 3.5 µm or 3.4 to 4.4 µm in the idler wavelength depending on the poling period of the crystal and the pump laser tuning range used. Each tuning range is fully accessible by changing the pump beam wavelength without changing the poling period, temperature of the crystal, or the dimensions of the OPO cavity. In the longer idler wavelength region, it is possible to tune across a large idler beam wavelength range with minimal changes in the signal beam wavelength. The tuning rate of the system is limited by the tuning rate of the pump source. The system uses a ring-cavity. Etalons can be employed inside it to improve stability allowing the pump laser frequency to be scanned up to its maximum of 40 GHz without mode-hops in the resonating signal beam. This results in the same idler beam scanning range with high resolution.
2. Experimental setup
The optical layout of the single-mode Ti:Al2O3 laser pumped cw SRO system is schematically shown in Fig. 1 . The Ti:Al2O3 ring laser (Coherent MBR-PS) is pumped by a cw intra-cavity frequency doubled Nd:YVO4 laser (Coherent Verdi V18) with the maximum optical output power of 18.5 W. The Ti:Al2O3 laser pumps the SRO and its wavelength is tuned in our experiments in the range of about 775 to 860 nm. The maximum output power of the Ti:Al2O3 laser is up to 6 W depending on the operating wavelength.
We tested standing-wave cavity geometries with two and four mirrors and observed that the 2-mirror linear cavity was susceptible to thermal and photorefractive effects and instabilities due to them. A 4-mirror cavity can be designed more freely for example to minimize changes in the size of the resonating beam waist due to small variations in the distances between the cavity mirrors. Ring cavity geometries provide more stable operation and lower oscillation threshold than the standing-wave cavities. Therefore, our SRO consists of a symmetric, 4-mirror ring cavity in a bow-tie geometry. A 50 mm long MgO-doped, periodically poled LiNbO3 (MgO:PPLN, HC Photonics) crystal is centered at the focal point of the pump beam. The facets of the crystal are perpendicularly cut and anti-reflection coated for all three wavelengths. The reflectance is R < 0.6% or better for the pump beam and from 0.6 to about 3% for the signal beam depending on its wavelength. The crystal contains nine parallel beam paths with different poling periods ranging from 21.25 to 23.25 µm in steps of 0.25 µm. The poling period can be changed by mechanically translating the crystal across the beam path. Thermoelectric elements control the temperature of the crystal. The maximum temperature is limited to about 120 °C. The pump beam is focused into the crystal using a plano-convex lens with a focal length of 150 mm. All of the cavity mirrors are coated to highly reflect the signal beam with R > 99.5%. The concave cavity mirrors transmit the idler and pump beams with transmission T > 95%. The beams exit the SRO system through the second concave cavity mirror and are collimated using a calcium fluoride lens.
The dimensions of the oscillator cavity are such that the lowest order transversal mode of the resonator matches the focusing parameter ξ = 2.0 of the pump beam . Two concave mirrors (QTF, the radius of curvature 100 mm) are at the distance d1 apart and symmetrically placed around the nonlinear crystal. The other two cavity mirrors (Thorlabs BB1-E03) are flat and also symmetrically placed so that a secondary focus is between the flat mirrors. The signal beam path length from the concave mirror to the secondary focus is d2. Various values of d1 and d2 were tried in the experiments depending on the details in optimizing the cavity geometry. For example, the best single-mode stability of the SRO was obtained by selecting d1 = 138 mm and d2 = 260 mm, which corresponds to the center of the cavity stability region and the free spectral range (FSR) of about 420 MHz. However, such dimensions typically lead to bistable behavior of the idler power versus the pump power due to thermal lensing. The effect of thermal lensing can be minimized by selecting d1 = 153 mm and d2 = 218 mm with FSR of about 460 MHz, closer to the cold-cavity stability limit . However, the cavity becomes susceptible to mechanical vibrations and acoustic noise, increasing the probability of mode-hops. Therefore, we mostly used the dimensions corresponding to the cavity stability center in order to ensure stable single mode operation of the SRO.
The wavelength of the pump beam was monitored during coarse tuning by coupling part of the residual pump beam into a wavemeter (Burleigh WA-2000). The wavelength of the idler beam was monitored using another wavemeter (Exfo WA-1500) and a spectrum analyzer (Exfo WA-650) with a resolution of about 50 MHz. The optical power of the pump and idler beams was measured using two different power meters before (Coherent Labmaster E) and after (Newport 1918-C) the SRO system.
3. Results and discussion
The power of the idler beam before the collimating lens as a function of the power of the pump beam and the pump depletion are shown in Figs. 2 a ) and b), respectively. The cavity alignment was optimized to account for the thermal lensing effect at high pump powers. The MgO:PPLN crystal was stabilized at the temperature of 57.4 °C and the poling period of 21.75 µm was used. The pump beam wavelength was set at 791.62 nm. The idler wavelength varied between 3131 and 3139 nm during the measurement due to mode-hops triggered by the changes in the pump power. The lowest SRO oscillation threshold reached was about 1.5 W. The maximum idler beam power and pump depletion were 0.8 W and over 0.8, respectively.
The MgO:PPLN poling period and temperature were fixed to 21.25 µm and 47 °C, respectively, in order to chart the oscillation threshold as a function of wavelength tuning. The results are in Fig. 3 . The measurement was repeated using different poling periods and temperatures to verify that the shape of the dependence of the oscillator threshold on signal beam wavelength is approximately constant. The output idler power varies following the same trend as the oscillation threshold plotted in Fig. 3 because of wavelength-dependent intracavity losses. The maximum idler power is close to 0.8 W (Fig. 2a) but, for example, can be limited to well below 100 mW at the peak of high oscillation threshold near 1110 nm in signal wavelength. It may be possible to expand the overall tuning range of the SRO by using coatings that induce low losses for a wider range of signal beam wavelengths.
Figure 4 (a) shows the measured idler beam wavelength as a function of the pump beam wavelength using the nine different poling periods of the MgO:PPLN crystal. The corresponding signal beam wavelength of the SRO as a function of the pump beam wavelength is in Fig. 4(b). No changes in the SRO cavity are needed within a poling period. The coarse wavelength adjustment takes just a few seconds, which is the time it takes to manually tune the pump laser. The tuning range has two separate regions of operation: short and long idler wavelengths.
The OPO may have to be empirically optimized for solid tuning in one of the two regions of operation by adjusting the mirror distances d1 and d2 by a few millimeters or so. This allows the signal beam with short or long wavelength to better match the focused pump beam in the nonlinear crystal. It was possible to operate both in short and long idler wavelength regions using an intermediate poling period of 22.50 or 22.75 µm. This gives a total of about 1.5 µm tuning range in the idler beam using just a single poling period, provided that changes in the cavity dimensions are made.
The widest possible tuning range of the SRO within either region of operation is available using a single poling period of the MgO:PPLN crystal, which eliminates the need of mechanical adjustments in the oscillator cavity. The idler wavelength ranging from 2.5 to 3.5 µm or from 3.4 to 4.4 µm is accessible by tuning just the pump beam wavelength when using the poling periods of, e.g., 21.25 and 23.00 µm, respectively. At the pump wavelengths above 840 nm, the coatings of the mirrors and crystal facets partially reflect the pump beam, which leads to weakly pump-resonant operation and instabilities.
The dependence of the idler beam frequency on the MgO:PPLN crystal temperature was found to be nearly linear. The wavelength of the pump beam was fixed at 789.85 nm and the poling period was 21.25 µm, which resulted in an idler beam tuning range of about 2.55 to 2.90 µm with a change of over 100 °C in temperature. Temperature tuning is possible but slow and limited in range compared to pump beam wavelength tuning.
The mode-hop behavior of the SRO was monitored while tuning the Ti:Al2O3 pump laser so that the idler beam wavelength changed in the short wavelength region from about 2.5 to 3.5 µm. The measurement was repeated using different poling periods and temperatures of the MgO:PPLN crystal with similar results. The magnitudes of the mode-hops vary but on average these are of order of about 100 GHz in the idler beam frequency and occur typically about every 20-30 GHz in the pump beam frequency if no etalons are used. It is interesting to note that temperature tuning leads to approximately normally distributed mode-hops that are random and can be along or against the tuning direction. On the other hand, pump beam tuning results in mode-hops that are more constant in magnitude and almost always towards the same direction.
Fine-tuning of the cw SRO system by scanning the Ti:Al2O3 laser up to 40 GHz was tested at different configurations. Best results were obtained by using an etalon at the secondary focus of the signal beam inside the SRO cavity to prevent mode-hops of the signal beam during the scan. The etalon acts as a passive element and needs no adjustments during the scan. A typical repetitive tuning pattern observed in the idler beam while repeatedly scanning the pump beam is shown in Fig. 5 . The scanning speed of the system depends on the pump laser properties and can be up to several GHz per second, but scan time of 2 minutes was used because the wavemeter can only take about one measurement per second. The temperature of the MgO:PPLN crystal was stabilized to 60.7 °C and the poling period was 21.25 µm. A 0.4 mm uncoated YAG etalon with FSR of about 375 GHz was used. The pump beam wavelength was near 785 nm.
We measured part of the absorption spectrum of ammonia, NH3, using photoacoustic spectroscopy (PAS) in order to verify the capability of our SRO for real spectroscopic measurements. The sample was at room temperature and 1 mbar pressure. The laser beam passed once through a static longitudinal resonant photoacoustic cell. The cell contains a copper waveguide, whose length is 18 cm and a microphone with an internal FET amplifier. The photoacoustic signal was detected using a lock-in amplifier. The power of the idler beam was about 100 mW. The Ti:Al2O3 pump beam frequency was scanned for about 40 GHz. A typical spectrum near 3196 cm−1 is shown in Fig. 6 . We were also able to detect a spectrum of NH3 near 2381 cm−1 by operating the SRO in the long idler wavelength region.
We have demonstrated, to our knowledge, the first truly singly resonant, continuous-wave SRO pumped by a widely tunable Ti:Al2O3 laser. It produces up to 0.8 W of single-mode idler beam power in the mid-infrared region. The SRO shows wide tuning range with two distinctive idler beam wavelength regions: about 2.5 − 3.5 µm and 3.4 − 4.4 µm are available by changing just the pump beam wavelength without touching the optimized SRO setup. The mode-hop behavior of the SRO during pump tuning is different from that during crystal temperature tuning. Using an etalon inside the oscillator, mode-hop-free scanning is achieved for several tens of GHz by fine-tuning the pump beam frequency. The suitability of the system for spectroscopy is demonstrated by measuring parts of the absorption spectrum of NH3. The oscillation threshold of the system is below 2 W and the required pump tuning range to achieve an extensive range of idler beam wavelengths is below 10 nm.
We acknowledge the Academy of Finland for financial support (# 129673 and # 132881).
References and links
1. L. E. Myers, R. C. Eckardt, M. M. Fejer, R. L. Byer, W. R. Bosenberg, and J. W. Pierce, “Quasi-phase-matched optical parametric oscillators in bulk periodically poled LiNbO3,” J. Opt. Soc. Am. B 12(11), 2102–2116 (1995). [CrossRef]
2. W. R. Bosenberg, A. Drobshoff, J. I. Alexander, L. E. Myers, and R. L. Byer, “93% pump depletion, 3.5-W continuous-wave, singly resonant optical parametric oscillator,” Opt. Lett. 21(17), 1336–1338 (1996). [CrossRef] [PubMed]
3. I. Lindsay, B. Adhimoolam, P. Groß, M. Klein, and K. Boller, “110GHz rapid, continuous tuning from an optical parametric oscillator pumped by a fiber-amplified DBR diode laser,” Opt. Express 13(4), 1234–1239 (2005). [CrossRef] [PubMed]
5. A. Rihan, E. Andrieux, T. Zanon-Willette, S. Briaudeau, M. Himbert, and J.-J. Zondy, “A pump-resonant signal-resonant optical parametric oscillator for spectroscopic breath analysis,” Appl. Phys. B DOI: 10.1007/S00340-010-3996-8, (2010).
6. G. A. Turnbull, D. McGloin, I. D. Lindsay, M. Ebrahimzadeh, and M. H. Dunn, “Extended mode-hop-free tuning by use of a dual-cavity, pump-enhanced optical parametric oscillator,” Opt. Lett. 25(5), 341–343 (2000). [CrossRef]
7. G. D. Boyd and D. A. Kleinman, “Parametric Interaction of Focused Gaussian Light Beams,” J. Appl. Phys. 39(8), 3597–3639 (1968). [CrossRef]
9. G. Guelachvili, and K. N. Rao, Handbook of infrared standards II (Academic Press, 1993), pp. 152–153.