Singly resonant optical parametric oscillator for mid infrared

Transcription

1 Singly resonant optical parametric oscillator for mid infrared S Das, S Gangopadhyay, C Ghosh and G C Bhar, Laser Laboratory, Physics Department Burdwan University, Burdwan , India FAX: ; dgp_buphygcb@sancharnet.in ABSTRACT Tunable mid-infrared radiation by singly resonant optical parametric oscillation based on KTA crystal pumped by Gaussian shape beam from Q-switched Nd:YAG laser has been demonstrated. Threshold energy of oscillation and single pass conversion efficiency of incident pump energy to infrared wavelength has also been measured. Keywords: Nonlinear optics, optical frequency conversion, optical parametric oscillator, singly resonant cavity, threshold of oscillation, conversion efficiency, mid-infrared radiation, 1. INTRODUCTION Frequency conversion by nonlinear optical frequency mixing technique is a useful technique to get tunable radiation in the spectral range where laser source at all not available or laser performs poorly. There are several methods like second harmonic generation (SHG), sum frequency mixing (SFM), difference frequency mixing (DFM) and optical parametric oscillation (OPO). Out of different techniques, the broad tuning ranges and efficient power conversion characteristic makes OPO s are very attractive in many applications requiring wide tunability and high peak power. By optical parametric oscillation one can easily generate coherent radiation continuously from UV to infrared and even THz domain. In the infrared they play an important role in the molecular finger print region of the spectrum 2-20 µm where molecular species have their fundamental absorption features and where we lack broadly tunable laser source. Due to the robustness and compactness of OPOs, make them attractive for different applications like laser spectroscopy (linear and nonlinear), remote chemical sensing, trace gas and vapor detection, biology, medicine and in laser radar. Since the first demonstration of OPO in 1965 using LiNbO 3 crystal 1 there has been a lot of improvement in this front due to advent of novel nonlinear material having wide transparency range with very low optical losses, high laser damage threshold and lot of improvement in optical coating technology. Periodically poled material 2 has revolutionized the nonlinear frequency processes by enabling devices with collinear propagation of the interacting beams, possibility to use the highest nonlinearity of the material and can overcome the limitation on tunability imposed by phase matching in birefringent material, by engineering suitable quasi phase-matched grating periods. In this paper we report the generation of tunable radiation in 3-5 µm range using KTA crystal by singly resonant optical parametric oscillation pumped by 10ns pulsed 1064 nm radiation from a Q-switched Nd:YAG laser. KTA has the potentiality for tuning to wavelengths inaccessible with KTP. KTA can be tuned almost over its whole transparency range (KTA has long wavelength cut-off down to 5.3 µm, inset of fig.2) covering the important 4-5 µm region, which is not accessible with KTP. Again KTA possesses slightly higher nonlinearity than KTP EXPERIMENT In our experiment we use a 41 o XZ cut (ϕ=90 o ) KTA crystal of 15mm long having 10x10 mm 2 cross-section. The input face of the crystal is anti-reflection (AR) 1064 nm (i.e. pump beam) and AR 1.3 to 1.7 µm (i.e. the signal

2 beam) and the output face is AR 1.3 to 1.7 µm. We use a hemispherical cavity configuration formed with a planoconcave input mirror having 5 m radius of curvature while the output mirror is a plane parallel. The input face of the input mirror (substrate BK7) is AR 1064 nm while the curved face is high reflection (HR) µm and high transmission (HT) nm. The face toward the cavity, of the output mirror (substrate CaF 2 ) is HR 1064 nm and 1.3 to 1.7 µm and HT 3-6 µm. The other surface is not coated. The cavity is pumped by 1064 nm radiation having pulse length of 10 ns from a Q-switched Nd:YAG laser. The pump beam shape is Gaussian in nature having divergence <1 mrad and 3.5 mm in diameter. The input and output mirrors are placed in a holder having both tilt motions. In addition the output mirror also has translation motion along the cavity axis controlled by a micrometer screw to facilitate the change in cavity length. The crystal is placed on a precision circular table having least count of 0.04 o, capable of rotating in the horizontal plane. The pump beam is horizontally polarized as such and made it vertically polarized by a 90 o polarisation rotator to satisfy the o-eo interaction. The generated infrared radiation (i.e. idler beam) is detected with a liquid nitrogen (LN 2 ) cooled MCT detector operated without preamplifier. Though the output mirror has high reflectivity (~ 99.5%) at the unconverted pump beam, in spite of that the detector sensed scattered pump beam radiation from different optics in the system. To avoid this we use an uncoated Ge filter to block the detection of the unwanted pump beam radiation by the MCT detector and it is found that a 3 mm thick Ge filter completely blocks the residual/scattered pump beam. We also use a quartz beam splitter (uncoated) to monitor the pump beam energy simultaneously and also a Faraday isolator to prevent any reflected pump beam entering back into the pump laser cavity. The electrical signal from the detector is displayed on a 100MHz-storage oscilloscope. At first the cavity length is kept at 20 mm and later on the cavity length is increased to 30 mm by steps through the micrometer screw attached with the output mirror and the corresponding threshold of oscillation energies are measured. The input pump beam as well as output infrared energies are measured with a energy meter of M/S Scientech Inc. 3. RESULT AND DISCUSSION 4. By rotating the crystal in the horizontal plane the phase-matched situation is achieved and by rotating the crystal in clockwise & anti-clockwise direction the phase-matched peak is ascertained. The phase matched angular tuning characteristics for OPO in this crystal in XZ plane for o-eo interaction is shown in fig. 1. The smooth curves (1), (2) & (3) in fig.1 are theoretically predicted phase matching angle for OPO in the XZ plane of KTA crystal pumped by µm radiation obtain from the Sellmeier coefficients given by Kato et al 4, Fanimore et al 5 and Feve et al 6 respectively while the dots ( ) represent our measured phase matched angles internal to the crystal. It is seen from this fig. that our measured phase matched angles follow closely the theoretical prediction obtain from the Sellmeier coefficients of Feve et al 6. Maximum error in our measurement is of ±0.04 o. We detect infrared radiation from about 4.0 µm to 5.2 µm. From theoretical calculation it is seen that for a particular phase matched angle there are possibility to get two infrared radiations simultaneously. But we detect idler radiation from 4.0 to 5.2 µm by blocking the radiation in the 3 µm spectral range with a polysteren sheet. The infrared radiation is checked with a monochromator of M/S Spex We measure the threshold of oscillation energy for this OPO at first keeping the cavity length at 20 mm at different infrared wavelengths and then also by increasing the cavity length by steps to 30 mm. These are shown in figs 2. Here the smooth curves are theoretical prediction calculated by using the analytical expression for threshold of energy fluence for a singly resonant OPO of Brosnan and Byer s equation 7,8 1.8τ 25L 1 2 J th = + 2α l + ln + ln2 (i) Kg s L 2 eff cτ R Where τ is the full width at half maxima (FWHM) of the pump pulse, l is the length of the crystal, α is the loss coefficient of the crystal, L is the optical length of the OPO cavity, g s is the mode coupling coefficient, R is the overall reflectivity of the end mirrors and L eff is the effective parametric gain length. The physical path length of the cavity is appropriately converted to optical path length of the cavity by considering the crystal. The gain coefficient K is given by the expression

3 K = 2ω s ω i d 2 eff / n s n i n p ε 0 c 2.(ii) where ω s and ω i are the signal and idler frequencies, d 2 eff is the effective nonlinear coefficient of the crystal, n s, n i, and n p are the refractive indices at signal, idler and pump frequencies respectively and ε 0 is the permeability of free space. The energy fluence J th is appropriately converted to energy multiplying it Generated Wavelength (µ m ) (2) (1) (3) Phase Matching Angle (deg.) Fig.1: Illustration of phase matched angular tuning characteristics for KTA optical parametric oscillator in XZ-plane i.e. ϕ = 90 o for o-eo interaction. The smooth curves (1), (2) & (3) are theoretical prediction obtained from the Sellmeier coefficients given by Kato et al 4, Fanimore et al 5 and Feve et al 6 respectively. The dots ( ) represent our measured phase matching angles internal to the crystal. with the area of the incident pump beam. The pump beam size is determined by taking the burn pattern of the pump beam on a photographic plate near the input mirror of the OPO cavity and diameter of the same is measured with a traveling microscope. In fig.2 the curves (1), (2), (3), (4) and (5) respectively represents the threshold of oscillation energy for the cavity length 20mm, 22mm, 25mm, 28 mm & 30 mm while the corresponding measured values are represented respectively by the symbol ( ) ( ) (? ) ( ) & (? ). It can be seen from fig 2. that the threshold of oscillation energy follows the transmission (inset of the fig.2) characteristics of the crystal. It is also found the measured threshold value is slightly higher than theoretically predicted value. This is due to the losses of incident pump beam energy (though it is small) occur at the input mirror face as well as at the input crystal face in spite of anti-reflection coating (>99%). In addition we observe the infrared radiation by LN 2 cooled MCT detector using uncoated Ge filter. There is a lot of Fresnel reflection losses occurs of the infrared radiation from the two surfaces of the uncoated Ge filter. Again the rear face of the mirror (substrate CaF 2 ) is not AR coated at the infrared radiation and hence some reflection loss takes place from this face also. Threshold of oscillation energy is measured very carefully and the maximum error in our measurement may be about 1-2 mj. We have also measured the threshold of oscillation energy at different cavity length which is represented in fig.3 keeping the infrared wavelength fixed at 4.9 µm. Here the smooth curve is theoretical prediction obtained from the expression (i) at stated above while the dots ( ) represents the measured value. Here also the reason of difference between the theoretical value and the measured values are the same as explained earlier.

4 1 0 0 Measured Threshold Energy (mj) Transmission (%) W a v e l e n g t h ( µ m ) (5) ( 4 ) (2) ( 1 ) ( 3 ) Idler wavelength( µ m ) Measured Threshold Energy (mj) Idler Wavelength = 4.9 µ m Cavity Length (mm) Fig.2: Illustration of threshold of oscillation energy for KTA (crystal length 15 mm) singly resonant optical parametric oscillator in XZ plane (o-eo interaction) with infrared wavelengths at different cavity length. Here the smooth curves (1), (2), (3), (4) & (5) are theoretically calculated using the expression (i) as described in text for the cavity length of 20 mm, 22 mm, 25 mm, 28 mm & 30 mm respectively while the corresponding measured values are represented by the symbols ( ) ( ) (? ), ( ) and (? ) respectively Fig.3: Illustration of threshold of oscillation energy for KTA (crystal length 15 mm) singly resonant optical parametric oscillator in XZ plane (o-eo interaction) with cavity length measured keeping the infrared wavelength fixed at 4.9 µm. The smooth curve here is theoretically calculated using the expression (i) as described in the text while different dots ( ) represent the measured values We also measure the generated infrared beam energy at 4.9 µm with increase in input pump beam energy. As expected it is observed the increase in infrared energy with increase in input pump beam energy. We do not increase the pump beam energy too much to avoid any damage on the crystal coating. The maximum conversion efficiency of energy from input pump beam energy to the infrared beam energy we measured at 4.9µm is about 1.2% keeping the cavity length 20 mm. We achieved about 700 µj energy applying about 60 mj energy of µm radiation from Nd:YAG laser. 4. CONCLUSION In conclusion we have demonstrated and studied the operation characteristics of KTA optical parametric oscillator in singly resonant mode using a Q-switched Gaussian shape pump beam with the capability of tuning in 4-5 µm region of infrared having finger prints of important atmospheric constituents.

5 ACKNOWLEDGEMENT The authors are grateful to the Board of Research in Nuclear Sciences (BRNS), DAE, Government. of India for partial financial support. REFERENCES 1. Tunable coherent parametric oscillation in LiNbO 3 at optical frequencies Giordmaine J A and R C Miller, Phys. Rev. Lett. 14, 973 (1965) 2. Quasi-phase-matched second harmonic generation:tuning and tolerances Fejer M M, Magel Jundt G A, D H and Byer R L, IEEE J Quan. Electron. 28, 2631 (1992) 3 Optical parametric oscillation with KTiOAsO 4 Powers P E, Ramakrishna S and Tang C L, Opt. Lett. 18, 1171 (1993) o phase-matched mid-infrared parametric oscillation in undoped KTiOAsO 4 Kato K, Umemura N and Tanaka E, Jpn. J Appl. Phys. 36, L403 (1997) 5. Infrared corrected sellmeier coefficients for potassium titanyl arsenate Fanimore D L, Schepler K L, Ramachandran B and McPherson S R, J Opt. Soc. Am. 12, 794 (1995) 6. Phase matching measurements and sellmeier equations over the complete transparency range of KTiOAsO 4, RbTiOAsO 4 and CsTiOAsO 4 Feve J P, Boulanger B, Pacaud O, Rousseau I, Menaert B, Marnier G, Villeval P, Bonnin C, Loiacono G M and Loiacono D N, J. Opt. Soc. Am B 17, 775 (2000) 7. Optical parametric oscillator treshold and linewidth studies Brosnan S J and Byer R L IEEE J Quan. Electron. 15, 415 (1979) 8. Optimum pump-pulse duration for optical parametric oscillator Bapna R C, Dasgupta K and Nair L G, Opt. & Laser Tech. 29, 349 (1997)