23_temporal Temporal Coherence Behavior 95 Temporal Coherence Behavior of a Nd:YAG Pumped Waveguide Raman Shifter Marilou M. Cadatal*, Ma. Leilani Y. Torres, and Wilson O. Garcia National Institute of Physics, University of the Philippines, Diliman, Quezon City 1101 E-mail: marilou.cadatal@up.edu.ph ABSTRACT Science Diliman (July–December 2004) 16:2, 95–100 *Corresponding author We study the behavior and report the control of the temporal coherence length zc of a 355/532 nm Nd:YAG pumped H2 Raman shifter with and without capillary waveguide (CWG). Depending on the application of the Raman-shifter light source, zc could be tuned rapidly or slowly by varying the H2 pressure P or the input power Pin. A more dynamic zc behavior is observed for a waveguide Raman shifter. INTRODUCTION Stimulated Raman scattering (SRS) in H2 is an efficient technique of extending the tuning range of lasers into the vacuum ultraviolet and far infrared (Mannik & Brown, 1986; Berry et al., 1982; Torres et al., 2003; Palero et al., 2001; Palero et al., 2000; Papayanis et al., 1998). It is simple, economical, robust, and capable of high conversion efficiency (Torres et al., 2003; Palero et al., 2001; Palero et al., 2000). Recently, the threshold energy required for SRS has been lowered and the conversion efficiency was improved by using a capillary waveguide (CWG) (Mannik & Brown, 1986; Berry et al., 1982; Torres et al., 2003). On the other hand, laser sources with controllable temporal coherence properties are valuable in spectroscopy and optical microscopy. With its highly directional beam outputs, these sources with low coherence can be efficient illuminators in conventional optical microscopes for generating images with acceptable levels of optical noise caused by speckles, edge ringing and shifting, and point diffractions (Garcia et al., 2001). Conversely, sources with longer coherence lengths are useful in interferometric applications. This work analyzes the temporal coherence properties of a H2 Raman shifter with and without a CWG and demonstrates the control of its associated temporal coherence length using the H2 pressure (P) and input power (Pin). EXPERIMENTAL SETUP The schematic diagram of the experimental setup is shown in Fig. 1. The fundamental 1064 nm output of a Q-switched linearly polarized neodynium-doped yttrium-aluminum-garnet (Nd:YAG) laser (Spectra Physics GCR-230) is converted to 532 nm (second harmonic) and 355 nm (third harmonic) via KDP crystals. The Nd:YAG is operating at 10 Hz repetition rate with a pulsewidth of 5–6 ns (FWHM). The telescope expands the beam before it passes through the aperture, A. Lens L1 (f = 290 mm) focuses the beam onto the entrance of a 50-cm-long capillary tube (CWG) placed inside the Raman shifter. The capillary tube has a 1 mm bore diameter. H2 is used as the Raman-active medium since it has the largest Raman frequency shift Cadatal, Torres, and Garcia 96 (4155 cm–1). Lens L2 collimates the Raman lines to the Pellin-Broca (PB) prism, which disperses the Raman output into the Rayleigh (R), Stokes (S), and anti-Stokes (AS) components. The time-averaged optical spectra S(l) are obtained using a grating spectrometer (SPEX Model 1000M, 3 s integration time, and 0.001 nm resolution). The power of each Raman line is measured using a power meter. This is used to normalize the spectral lines obtained from the spectrometer. DETERMINATION OF Z C Since the Raman shifter is a pulsed light source, its corresponding temporal coherence length zc is extremely difficult to measure directly with a scanning Michelson interferometer. A better technique is to employ a nonscanning multichannel Fourier transform interferometer in which the interferogram points are formed simultaneously in space, instead of through time. This requires a very fast detector array with a large number of pixels and a uniform response over a wide spectral range for a uniform sampling of the interferogram. But such a photodiode array is expensive (Garcia et al., 2001). The associated zc of S(l) is determined from the following procedure proposed by Garcia et al. (2001): Step 1: The sampled spectrum, {S(lm)}, is converted into its equivalent k representation, {S(km)}, where k = 1/l; m = 1,…,M; and kM (cm –1) = 50,000. The 355/532 nm R line is included in {S(lm)}. Step 2: An even (2M+1) data sequence, {S(n)}, is derived from {S(km)}, where S(n) = S(km) = S(–n), and n = –M,–M+1,…,M. Step 3: The inverse Fourier transform of S(n); F– 1[{S(n)}] = I(z) is then obtained. Here, z is a distance variable in the range –MkM 9.4 mW, a stable zc minimum is achieved at around 150 psi. At lower Pin (7.5 and 5.6 mW), the minimum zc value is observed to shift to higher P. Rapid control of zc is possible for P < 150 psi, while a slower variation of zc is possible for P > 150 psi. The zc starts to saturate at P > 300 psi for Pin > 9.4 mW. However, zc is not observed to saturate for Pin < 9.4 mW since at lower Pin, the R line dominates the spectrum. Without the CWG [Fig. 3(b)], zc exhibits a less pronounced inverted humplike behavior with the stable zc minimum shifting to a higher P (300 psi) for Pin > 9.4 mW. For Pin < 9.4 mW, the minimum zc value is again observed to shift to higher P. At Pin = 9.4 mW, zc ranges from 2.99 to 4.68 mm, while at Pin = 26.4 mW, zc could be tuned from 2.74 to 4.68 mm. Fig. 2. On the left are the S(k) for (i) P=150 psi and (ii) 600 psi (a) with and (b) without CWG. On the right are the corresponding interferograms of S(k). N or m al iz ed i nt en si ty ( a. u. ) Wave number (x103 cm–1) Path length (µµµµµm) N or m al iz ed i nt en si ty ( a. u. ) Fig. 3. zc measurements as a function of H2 pressure for Pin = 5.6, 7.5, 9.4, 13.3, 18.5, and 26.4 mW (a) with and (b) without CWG. Pressure (psi) C o h er en t le n g th ( µµµµ µ m ) 50 250 450 650 1.5 2.5 3.5 4.5 5.5 Pressure (psi) C o h er en t le n g th ( µµµµ µ m ) 50 250 450 650 1.5 2.5 3.5 4.5 5.5 (a) (b) Cadatal, Torres, and Garcia 98 Compared with that with a CWG, a slower control of zc is possible for P < 300 psi before reaching its minimum. For P > 300 psi, the value of zc starts to saturate. Moreover, a lower zc minimum is obtained in the presence of a CWG since more lines are produced as shown in Fig. 2. The Stokes intensity is a function of Pin and the H2 P as described by (Torres et al., 2003; Palero et al., 2000) (2) where IP and Is(0) are the pump laser intensity and the initial Stokes intensity, respectively, l is the interaction length between the Raman-active medium, and gR is the steady-state Raman gain coefficient. The gR is proportional to the Raman-active gas P (Torres et al., 2003; Palero et al., 2000). Without the CWG (lower l), a higher P is needed to achieve the same Is(l) at a constant IP. However, an analysis of gR as a function of P shows that gR saturates as P is increased (Palero et al., 2001). Hence, even if the P is increased at a particular Pin, the number of Stokes lines is not increased. This leads to the saturation of the zc value at higher P. Figures 4(a) and 4(b) present the dependence of zc with Pin with and without a CWG. In the presence of a CWG [Fig. 4(a)], the zc value is robust against variations of Pin > 10.1 mW for a particular P. For instance, at 300 psi, zc is maintained at about 3.21 mm. For Pin < 10.1 mW, zc can be controlled by varying the Pin at a given P or by varying the P at a given Pin. Without the CWG, higher zc values are obtained since lesser Raman lines are produced. At 100 psi, zc is maintained at 4.69 mm since the spectrum is still dominated by the R line. At P < 150 psi (with CWG) and P < 200 psi (without CWG), zc decreases with increasing Pin since at low gas P, increasing Pin leads to the generation of more Raman lines. However, this does not work under high pressures where increasing Pin only results in S2 (with CWG) and S1 (without CWG) dominating over the Raman lines (Fig. 2). ZC BEHAVIOR FOR THE 532 NM PUMP The wavelengths of the generated Raman lines in H2 are summarized in Table 2. C o h er en t le n g th ( µµµµ µ m ) (b) C o h er en t le n g th ( µµµµ µ m ) (a) Input power (mW) 5 10 15 20 25 1.5 2.5 3.5 4.5 5.5 Input power (mW) 5 10 15 20 25 1.5 2.5 3.5 4.5 5.5 Fig. 4. zc measurements as a function of Pin for H2 pressure = 100, 150, 200, 300, 450, and 600 psi (a) with and (b) without CWG. Table 2. Generated Raman lines (lpump= 532 nm). Raman line Wavelength (nm) Raman line Wavelength (nm) S1 AS1 AS2 683.0 435.7 368.9 AS3 AS4 319.9 282.3 Figure 5 shows the dependence of zc on H2 P for different Pin (a) with and (b) without a CWG. As in the 355 nm pump wavelength, a rapid control of zc is possible at certain P while a slow variation is possible for other P. A noticeable shift in the minimum zc value Temporal Coherence Behavior 99 Fig. 5. zc measurements as a function of H2 pressure for Pin =15.4, 28.3, 32, 42.2, and 46.46 mW (a) with and (b) without CWG. Pressure (psi) C o h er en ce l en g th ( µµµµ µ m ) 0 200 400 600 2.0 3.0 4.0 5.0 6.0 (b) Pressure (psi) C o h er en ce l en g th ( µµµµ µ m ) 0 200 400 600 2.0 3.0 4.0 5.0 6.0 (a) 7.0 7.0 P in ( m W ) Pressure (psi) 60 0 40 0 20 0 10 0 80 60 42.2 32.0 28.3 21.0 15.4 46.46 (a) P in ( m W ) Pressure (psi) 60 0 40 0 20 0 10 0 80 60 42.2 32.0 28.3 21.0 15.4 46.46 (b) Fig. 6. Bandwidth as a function of H2 pressure for Pin =15.4, 21, 28.3, 32, 42.2, and 46.46 mW (a) with and (b) without CWG. could be observed for both waveguide and conventional Raman shifters. The shift is due to the change in the spectral bandwidth of the Raman shifter as a function of P (Fig. 6). As Pin decreases, more SRS lines are generated at higher P, thereby broadening the spectral bandwidth of the Raman shifter at higher P. Consequently, as Pin decreases, zc minimum shifts to a higher P. While the shifting is observed for all Pin with a 532 nm pump wavelength, the minimum zc is observed to shift to higher pressures only at low Pin for a 355 nm pump (Fig. 3). This is because the cumulative gain coefficient is inversely proportional to the pump and Stokes wavelengths (Palero et al., 2001) so that higher pump wavelengths lead to lower Stokes conversion efficiencies and higher threshold energies. The dependence of zc as a function of Pin is presented in Fig. 7. For the waveguide Raman shifter operating at 200 psi, zc decreases for Pin < 21 mW due to the generation of Raman lines. However, further increase in Pin only results in S1 dominating over the other Raman lines. Hence, zc starts to increase for Pin > 21 mW. At higher P (> 200 psi), zc increases as Pin is increased since an increase in Pin only results to S1 dominating the spectrum. Similarly, without the CWG, zc decreases at Pin where Raman lines are generated. As S1 dominates over the spectrum, zc increases. CONCLUSION The zc of a waveguide Raman shifter can be varied over a wide range via an appropriate selection of H2 P and Pin. For both 355 and 532 nm pumped waveguide Raman shifter, zc exhibited an inverted hump behavior. For the 355 nm pumped waveguide Raman shifter, a stable zc minimum is obtained at around 150 psi for Pin > 9.4 mW. If pumped with 532 nm, zc minimum shifts to higher pressures for decreasing Pin. Lower zc values could be obtained by a waveguide Raman shifter. Cadatal, Torres, and Garcia 100 Fig. 7. zc measurements as a function of Pin for H2 pressure =200, 300, 400, 500, and 600 psi (a) with and (b) without CWG. C oh er en ce l en gt h ( µµµµ µ m ) (a) Input power (mW) 10 20 30 40 50 2.5 3.0 3.5 4.0 4.5 5.0 C oh er en t le ng th ( µµµµ µ m ) (b) Input power (mW) 10 20 30 40 50 2.5 3.0 3.5 4.0 4.5 5.0 ACKNOWLEDGMENTS The authors gratefully acknowledge the Instrumentation Physics Laboratory of the National Institute of Physics for lending the detector used in this work and the Philippine Department of Science and Technology (DOST) through the Engineering and Science Education Project (ESEP) for the equipment grant. This project is supported by the University of the Philippines and the Philippine Council for Advanced Science and Technology Research and Development (PCASTRD). REFERENCES Berry, A.J., D.C. Hanna, & D.B. Hearn, 1982. Low threshold operation of a waveguide H2 Raman laser. Opt. Commun. 43: 229–232. Born, M. & E. Wolf, 1999. Principles of optics, 7th ed. Cambridge University Press, Cambridge. Daza, M.R., A. Tarun, K. Fujita, & C. Saloma, 1999. Temporal coherence behavior of a semiconductor laser under strong optical feedback. Opt. Commun. 161: 123–131. Garcia, W., J. Palero, & C. Saloma, 2001. Temporal coherence control of Nd:YAG pumped Raman shifter. Opt. Commun. 197: 109–114. Mannik, L. & S.K. Brown, 1986. Tunable infrared generation using third Stokes output from a waveguide Raman shifter. Opt. Commun. 57: 360–364. Palero, J.A., J.O.S. Amistoso, M.F. Baclayon, & W.O. Garcia, 2000. Generation of UV, VIS, and NIR laser light by stimulated Raman scattering in hydrogen with a pulsed 355 nm Nd:YAG laser. Proceedings of the 18th SPP Philippine Physics Congress. Palero, J.A., R.S. Ibarreta, & W.O. Garcia, 2001. Frequency conversion of a 532 nm Nd:YAG laser using a hydrogen Raman shifter. Proceedings of the 19th SPP Philippine Physics Congress. Papayanis, A.D., G.N. Trikrikas, & A.A. Serafetinides, 1998. Generation of UV and VIS laser light by stimulated Raman scattering in H2, D2, and H2/He using a pulsed Nd:YAG laser at 355 nm. Appl. Phys. B. 67: 563–568. Torres, M.L., M. Cadatal, & W. Garcia, 2003. Stimulated Raman scattering in a 532 nm Nd:YAG laser pump hydrogen Raman shifter with a capillary waveguide. Proceedings of the 21st SPP Philippine Physics Congress.