Optical parametric generation at extremely low pump irradiance in a long periodically poled Lithium Niobate
Shy Accoa, Pinhas Blaua, Shaul Pearla, Ady Arieb
bDepartment of Electrical Engineering - Physical Electronics, Faculty of Engineering, Tel Aviv
ABSTRACT
Optical parametric generator (OPG) is a very attractive optical down-conversion configuration since it is a single pass process and no cavity mirror’s alignment is required. Thus the system configuration is much more simple and robust. Traditionally, OPG processes were demonstrated using a pump source with a pulse length of the order of picoseconds or less. This is because GW/cm2 order of magnitude pump irradiance was required to excite an OPG process, and such irradiance in nanosecond long pulses commonly damages the non-linear crystal. The introduction of periodically poled crystals with high non-linear coefficients has significantly lowered the threshold for parametric processes. This progress in non-linear crystals enables exciting OPG processes at less than 100MW/cm2 irradiance, using nanoseconds long pulses from Q-switched lasers. We present an OPG with a threshold of less than 10 MW/cm2 using an 80 mm long Periodically Poled Lithium Niobate (PPLN) non-linear crystal. High signal conversion efficiency and high power were obtained at 25 nanosecond pulse length, 10 kHz repetition rate pumping without damaging the crystal. Theoretical approaches for explaining this OPG regime are discussed.
Keywords: Optical parametric generation, Parametric noise, Periodically poled crystals. 1. INTRODUCTION
The need for tunable IR sources and the lack of laser gain media in part of this wavelength range turn parametric conversion to an attractive candidate for such sources. Optical parametric oscillators (OPO) and amplifiers (OPA) and OPG configurations are used in order to convert the wavelength of a given laser to another desirable wavelength.
The commonly used nonlinear optical material for frequency conversion in the IR wavelength range is Lithium Niobate. Periodically poling of Lithium Niobate (PPLN) provides nonlinear optical crystals with noncritical phasematching and exceptionally high effective nonlinear coefficient (17 pm/V).
The availability of long crystal with low loss enables extremely high parametric gain which permit low-threshold devices, including long pulses (dozen of nsec) and cw operation.
The main disadvantage of the OPG is the high pump irradiance required in order to get efficient frequency conversion. In this case the pump level will ultimatly be limited by crystal damage threshold.
Recently, MgO doped lithium niobate crystals became available. These exhibit the same non linear optics performance as PPLN but with higher damage threshold.
In this paper we report on experimental study of OPG in long periodically poled MgO:LiNbO3 crystals. We focused on studying the behavior of OPG configuration for small gain-length product (g0L≈10) in three different crystal lengths 35, 50, and 80mm.
2. EXPERIMENTAL SETUP
In the experiments describe here we investigated the generated signal power in an OPG configuration at three different periodically poled MgO:LiNbO3 crystal lengths. The PPLN crystals (HC Photonics Corp) had a grating period of 28.8µm that is phase-matched for 1.064µm pump to 1.45µm signal and 4µm idler conversion. A schematic drawing of the experimental setup is shown in figure 1.
We used Nd:YVO4 Q-switched laser, with 25ns pulse length, 10KHz repetition rate. The laser is linearly polarized and its beam quality is M2<1.1.
The pump power was controlled by a λ/2 waveplate and a polarizer. The pump beam was focused into the PPLN crystal to a beam waist of 195µm radius in the center of the crystal. The crystal facets were polished at 5° with respect to the wall domain, to prevent oscillation induced by surface reflection. Both facets of the PPLN crystal are antireflection coated at the signal, idler and pump wavelength in order to ensure single-pass OPG process. The crystal was placed in an oven which allowed varying and controlling the temperature of the crystal within the range of 90°-180°C. The crystal was heated in order to avoid photorefractive damage and to satisfy the phase-matching condition at the required wavelengths.
To measure the output power of the signal and the idler we inserted a spectral filters into the OPG output beam path.
Fig. 1. Experimental setup for PPLN OPG.
3. EXPERIMENTAL RESULTS
Figures 2-4 show the dependence of the measured signal power on average pump power. This dependency is shown for the three different crystals length (35mm, 50mm and 80mm).
Figure 2 shows the measured signal output power in a 35mm long MgO:LiNbO3 crystal versus pump power. As can be seen at the beginning of the parametric process the signal output power grows moderatlly and when the pump power was increased the signal grows exponetially. In the 35 mm long crystal the avarage minimal signal power that was measured is 143µwatt at an avarage pump power of 4.5watt. At this power level the pump irradiance was 30 Mwatt/cm2.
Fig. 2. Measured and calculated average OPG output power versus pump power for 35mm long PP-MgO:LiNbO3 crystal.
Fig. 3. Measured and calculated average OPG output power versus pump power for 50mm long PP-MgO:LiNbO3 crystal.
Figure 3 shows the measured signal output power in a 50mm long MgO:LiNbO3 crystal versus pump power. In the 50 mm long crystal the avarage minimal signal power that was measured is 55µwatt at an avarage pump power of 2.5 watt. At this point the pump irradiance was 16.74 Mwatt/cm2.
Fig. 4. Measured and calculated average OPG output power versus pump power for 80mm long PP-MgO:LiNbO3 crystal.
Figure 4 shows the measured signal output power in a 80mm long MgO:LiNbO3 crystal versus pump power. In the 80 mm long crystal the avarage minimal signal power that was measured is 178µwatt at an avarage pump power of 1.1watt. At this point the pump irradiance was 7.36 Mwatt/cm2., which is the lowest pump irradiance that the OPG was excited in our experiments.
4. BRIEF THEORY
A quantum mechanical model of an OPG was first studied by Louisel et al. [3]. Since then, several theoretical approaches to describe parametric noise were developed [4], [5].
OPG refers to a nonlinear process in which a pump wave at frequency ωp propagating through a nonlinear crystal interacts with the parametric noise and generates two waves at frequency ωs (signal) and ωi (idler) that satisfying the phase matching condition.
Since the signal generation in an OPG is initiating from the quantum noise, the treatment at the beginning of the parametric process is done by quantum mechanical theory and cannot be describe by the classical optical parametric equations. The classical equations require assuming a nonzero signal or idler at the crystal entrance.
Quantum mechanical calculations of the OPG process in this configuration will be presented elsewhere [6].
Figure 5 presents the measured and calculated signal output power of the three different MgO:LiNbO3 crystal lengths. Each crystal has different minimum energy that must be accumulated before the OPG is excited. As can be seen, the theoretical model describes well this threshold effect.
Fig. 5. Measured and calculated average OPG output power for 35mm, 50mm, 80mm long PP-MgO:LiNbO3 crystals.
Figure 6 shows the OPG measured and calculated signal output in 80mm long MgO:LiNbO3 crystal at higher pump power. When the pump power increases (above g0L≈10) a discrepancy from the theoretical model is observed. Such discrepancy appears also at shorter crystal in higher pump power.
Figure 7 shows the spectra of 35mm long PP-MgO:LiNbO3 OPG, measured at different pump power levels. At a pump power of 5 Watts, the spectral width is 2.3nm, while at a higher pump power of 8 Watts, the spectral width broadens to 3.6nm.
Fig. 6. Average OPG output power, versus pump power experimentally measured and predicted by the model, for a 80mm
Measured beam spectraat pump power of 5watt
Measured beam spectraat pump power of 6watt
Measured beam spectraat pump power of 7watt
Fig. 7. Spectrum of the signal wave of 35mm long PP-MgO:LiNbO3 crystal OPG measured at different pump power levels.
The broadening of the signal spectrum as the pump power is increased can be attributed to the amplification of non-collinear components of the signal beams.
Measured beam spectra atpump pow er of 5.8w att
Measured beam spectra atpump pow er of 5.2w att
Fig. 8. Spectrum of the signal wave of 50mm long PP-MgO:LiNbO3 crystal OPG measured at different pump power levels.
Figure 8 shows the spectra of a 50mm long PP-MgO:LiNbO3 OPG, measured at different pump power levels. At a pump power of 4.8 Watts, the spectral width is 5nm, while at a higher pump power of 5.8 Watts, the spectral width broadens to 6.1nm.
In this case too, the broadening of the signal spectrum as the pump power is increased can be attributed to the amplification of non-collinear components of the signal beams.
5. CONCLUSION
We have demonstrated a low threshold, high repetition rate (10KHz) PP-MgO:LiNbO3 OPG. The OPG was pumped by a long (25nsec) pulse Q–switch Nd:YVO4 laser at 1.064µm.
We have studied the behavior of OPG configurations with three different crystal lengths and showed experimentally that, long crystals have a lower pump threshold than short crystal. It was shown that each crystal length has a different minimum energy that must be accumulated before the OPG is excited.
We have studied the dependency of signal spectra on pump power in two OPG crystal lengths. The spectral width of the generated OPG radiation increased at high pump power, due to amplification of non-collinear components of the generated beam.
The theoretical, quantum mechanical model was found to be accurate only in the low signal generation efficiency regime, wherein the pump depletion effect is negligible along the crystal and constant pump field can be assumed.
We are currently investigating possible explanations for this discrepancy. Apparently this requires full numerical simulation of the OPG process.
REFERENCES
1. Y. Guan, J. W. Haus, P. Powers, "Broadband and off-axis optical parametric generation in periodically poled
LiNbO3 ", J. Opt. Soc. Am. B. 21, pp. 1225-1233 (2004).
2. R. L. Byer, S. E. Harris, "Power and bandwidth of spontaneous parametric emission", Phys. Rev.168, pp. 1064-
3. W. H. Louisell, A. Yariv, A. E. Siegman, "Quantum fluctuation and noise in parametric processes Ι", Phys. Rev.124, pp. 1646-1654 (1961).
4. D. A. Kleinman, "Theory of optical parametric noise", Phys. Rev.174, pp. 1027-1040 (1968) 5. W.G. Wagner, R. W. Hellwarth, "Quantum noise in a parametric amplifier with lossy modes", Phys. Rev.133, pp.
6. S. Acco, P. Blau, S. Pearl, & A. Arie, "The onset of optical parametric generation in long periodically-poled
Current Perspective Coronary Heart Disease: Reducing The Risk A Worldwide View Gerd Assmann, FRCP; Rafael Carmena, MD; Paul Cullen, FRCPI; Jean-Charles Fruchart, PhD;Fabrizio Jossa, MD; Barry Lewis, FRCP; Mario Mancini, MD; Rodolfo Paoletti, MD;for the International Task Force for the Prevention of Coronary Heart DiseaseWorldwide, cardiovascular diseases are now the most risk increases
Dean Lillard, Cornell University and DIW You can't always get what you want: observations on self-reported satisfaction, consumption, and underlying utility Although social science surveys have long asked respondents to report levels of life satisfaction or happiness, economists have only begun to use those data in the last ten years or so. Other social scientists often joke that, rat