Open Access
Add to BibSonomyAbstract
Nanosecond-laser induced damage of RbTiOPO4 crystals (RTP) has been studied at 1064 nm as a function of propagation direction and polarization orientation. A significant difference in the Laser Induced Damage Threshold (LIDT) was observed for x-cut and y-cut crystals in Pockels cell configuration, where the light propagation direction is along the x and y axes of the crystal respectively. In Pockels cell configuration the polarization is oriented at 45° with respect to the z-axis of the crystal. Experiments with the polarization oriented parallel to the principal axes of the crystal pointed out the importance of the polarization direction for the LIDT whereas the propagation direction did not significantly influence the LIDT. Comparison of the experimental data with a simple model reveals the influence of frequency doubling on the LIDT in Pockels cell configuration. In the case of the y-cut Pockels cell, the generation of frequency doubled light causes an LIDT below the LIDT of x and z-polarized light at the fundamental wavelength.
©2007 Optical Society of America
1. Introduction
Laser induced damage in nonlinear optical crystals is often a limitation for the design of compact and high power laser systems. Frequency converters and Pockels cells both make use of the same type of crystalline materials. For the Pockels cells the important material parameters are the electro-optical constants defining the switching voltage of the cell and the dynamic properties of the crystal limiting the maximum driving frequency. RbTiOPO4 (RTP) is a non-hygroscopic material with high electro-optic constants [1, 2] good extinction ratio and no piezoelectric ringing [2]. For these reasons it is often chosen for high repetition rate Pockels cells or for pulse picking from high repetition rate pulse trains. Like KTP, RTP is an orthorhombic crystal, point group mm2, space group Pna21 [1, 3].
Laser damage testing of nonlinear optical crystals is more demanding than testing isotropic materials. Depending on the propagation direction, the polarization and the divergence of the test beam, we need to consider walk-off and nonlinear frequency conversion effects. Additionally, depending on the type and the size of the crystals, inhomogeneities generated during crystal growth may influence the measured Laser Induced Damage Thresholds (LIDT) [4, 5].
Only few systematic literature data on the anisotropy of the laser induced damage threshold is available except for KDP [6, 7, 4] and KTP [8–10]. Especially KDP attracted a lot of attention in the last years since it has been chosen for the frequency doubling and tripling stages of the National Ignition Facility (USA) and the Laser Mega Joule (France). Yoshida et al. reported different LIDT for different propagation directions and the same polarization when testing KDP at 1064 nm or 532 nm wavelength [4]. Concerning KTP most of the work concerns the grey-tracking phenomenon. Hu et al. [8] observed an important dependence of the grey-tracking sensitivity on the polarization direction during continuous wave frequency doubling of the 1064 nm wavelength. Further, Yoshida et al. reported a significant anisotropy during pulsed LIDT testing of KTP in a 1-on-1 mode [10]. They however measured efficient intensity dependent backscattering in their setup giving rise to questions whether the measured anisotropy has to be attributed to the resistance of the material or to the anisotropy of stimulated Raman scattering in the crystal.
In this contribution we discuss the LIDT anisotropy observed in RTP, and more precisely the damage thresholds of x-cut and y-cut RTP Pockels cells.
2. Experimental
2.1. Samples
In the following we name x, y and z the principal axes of the crystal. The crystals are used at normal incidence and the crystal is named x-cut if the light propagation direction is along the x-axis of the crystal.
We used commercial-grade non-assembled RTP crystals for Pockels cells from two providers: Provider A (Cristal Laser SA, France) and provider B (Raicol Crystals Ltd, Israel). The crystals from provider A were grown using a flux method. All crystals were cut along the principal axes and the surfaces used by the laser beam were anti-reflection coated for normal incidence at 1064 nm. All crystals were 10 mm long. The data shown in section 3.1 has been acquired on six y-cut crystals of dimensions 4×4 mm2 (provider A) and three x-cut crystals (provider B) (4×4 mm2) as well as one 20×20 mm2 x-cut crystal (provider A).
Samples for the experiments on the type of the measured anisotropy were also 10 mm long and cut along the principal axes, but had no AR coating. The data shown in section 3.2 were acquired on three crystals of dimensions 10×10 mm2 (provider A), one for each propagation direction.
2.2. Laser damage measurements
The laser damage measurements have been carried out according to an S-on-1 method [11] employing a Q-switched Nd:YAG laser at its fundamental wavelength (Quantel Ultra GRM). In order to ensure the practical significance of the LIDT-test we used the same linear polarization as for Pockels cell operation and a test beam that was parallel within the whole sample length. The coefficient of determination (R2) of a two-dimensional Gaussian fit on the measured beam profiles was higher than 0.95 in the used part of the focal region. The beam diameter varied by less than 10% within the whole sample length. The used beam diameter at the waist was 75±2 μm and the pulse duration of the laser was 6±1 ns.
Fig. 1. Schematic of the LIDT test setup: The filter F blocks 1064 nm and 532 nm light in order to protect the CCD camera. NG is a neutral grey filter and M a macroscope.
A schematic of the setup is shown in Fig. 1 and a detailed description of the measurement procedure is given in [12]. The test beam passes an automated variable attenuator, a small fraction of the beam is reflected onto the power meter (D) and the transmitted beam is focused by an f = 150 mm plano-convex lens (L) into the centre of the sample. Damage detection is carried out by in situ digital image processing of scattering images acquired by a CCD-camera. For this purpose a fibered halogen lamp illuminates the sample. Light that is scattered by laser-induced damages is then imaged with high depth-of-field onto the CCD camera using a macroscope (M). This allows us to detect damages in the whole depth of the sample. The camera is protected from direct and frequency doubled laser light by two blocking filters (F) for 1064 nm and 532 nm wavelengths respectively. Additionally, manual analysis of the sample under a microscope with differential interference contrast was carried out for control purposes. By this means we could not detect any other damaged sites than those already detected by in situ image processing.
As additional information to the standard S-on-1 test procedure we also saved for each site the received number of shots. Thus it is possible to extract the 1-on-1 damage curves from the measurement data [13]. The pulse repetition rate of the laser was set to 10 Hz for all experiments and the maximum number of pulses was S = 200. The number of sites that have been tested for one fluence ranges between 2 and 30 depending on the measurements. As a consequence the precision of the experimental data differs for each data point as indicated by the error bars on the graphs. The confidence level for all error bars is 68.3%.
3. Results and discussion
3.1. Measurements in Pockels cell configuration
For this study numerous measurements have been carried out on many individual crystals supplied by two different providers (see section 2.1). We also tested crystals originating from two different growth runs of one provider. All laser damage probability data in Pockels cell configuration was highly repeatable, and in consequence the curves presented here are averaged over data acquired on several crystals.
In RTP Pockels cells, the light may propagate parallel to the x-axis (x-cut crystals) or parallel to the y-axis (y-cut crystals). In both cases, the switching voltage is applied in the ±z direction of the crystal and the incoming linear polarization is oriented at 45° with respect to the z-axis.
Upon testing with a parallel beam, bulk damages are created in a large range of depths in the crystal. Sometimes several damages are created at different depths when testing one site (Fig. 2(a)). Typically, the damage morphology is described by micro-cracks along the principal axes (Fig. 2(b)). For the y-cut RTP no surface damage was observed. In the case of x-cut RTP, due to its higher bulk-threshold, we observed about half of the damages on the surfaces.
Fig. 2. Microscope images of damaged RTP crystals. The left image shows two laser induced damages that were created on a single test site. The right image shows two laser induced damages observed along the optical axis.
Figure 3 shows the damage probability curves for both x-cut and y-cut RTP tested at 1064 nm. Here both, bulk damage and surface damage, are considered in order to present a damage probability approaching the risk of operation of an RTP Pockels cell. A significant anisotropy of the LIDT is observed with a 200-on-1 damage threshold of about 6 J/cm2 for the y-cut crystals and 12 J/cm2 for the x-cut crystals. The two curves measured in x-cut RTP originating from two different providers are identical within the error.
In contrast to the findings of Yoshida et al., who measured in KTP up to 60% coupling loss close to the LIDT [10], we did not find any evidence for nonlinear scattering effects or other loss mechanisms in our experiment. The transmission of the crystal was close to 100% for all fluences up to damage creation.
Fig. 3. Damage probability curves in 200-on-1 mode considering bulk and surface damage. Both graphs show tests in Pockels cell configuration (polarization at 45° to z-axis). Plot points with error bars represent data resulting from measurements on different crystals.
Figure 4 shows 1-on-1 damage probability curves for the same configuration, but considering only damage in the bulk of the crystal. The estimated damage thresholds, that can be compared to the measurements in section 3.2 are 6.5 J/cm2 and 16.5 J/cm2 for y-cut and x-cut crystals respectively.
Fig. 4. Damage probability curves in 1-on-1 mode and Pockels cell configuration. Only bulk damage is considered. The dashed lines are guides to the eye.
The fact that x-cut crystals and y-cut crystals differ in their damage probability curves is not really surprising as RTP is a biaxial crystal. The extend of the anisotropy in LIDT is nevertheless surprising as the physical properties in x and y directions are quite similar and only the z-direction properties are significantly different. For example the refractive indices at 1064 nm are nx = 1.7672,ny = 1.7760 and nz = 1.8573 [14].
In fact, inside the crystal only the eigenmodes can propagate. Walk-off for both eigenwaves is zero for our test conditions [15]. For example, testing of an x-cut crystal in Pockels cell configuration involves two waves both propagating in x-direction and being y-polarized and z-polarized respectively. The experiments described in the next paragraph have been carried out in order to separate the influences of propagation direction and polarization on the observed LIDT anisotropy.
3.2. Measurements for polarizations aligned with the principal axes
Damage probability curves have been measured for several combinations of polarization direction and propagation direction both of them being aligned with one of the principal axes of the crystal. For these experiments we used uncoated crystals and only volume damage was considered for the 1-on-1 damage probability curves shown in Figs. 5 and 6.
Fig. 5. Damage probability curves in 1-on-1 mode for fixed polarization and different propagation directions. The curves for different propagation directions are acquired using separate crystals. The dashed lines are guides to the eye.
Figure 5 shows two viewgraphs each containing two test results obtained for different propagation directions at a given polarization of the test beam. The damage probability does not depend significantly on the propagation direction.
Both viewgraphs of Fig. 6 contain two test results obtained for different polarization directions and fixed propagation direction. The data show that the damage probability depends on the polarization direction and that the LIDT is higher for z-polarized light than for x or y-polarized light. The significantly higher damage threshold of RTP for z-polarized light recalls the much lower sensitivity of KTP to grey-tracking for this polarization reported by Hu et al. [8]. The same kind of defect stabilization by a nearby Rb-vacancy as discussed by Hu et al. has been shown by Jiang et al. [16], and hopping of the Rb-ions would destabilize the colour centres as in the case of KTP. Hopping of the Rb-ions under the influence of an electromagnetic field is however reduced, compared to KTP, as reflected by the ionic conductivity value which is about one hundred times lower for RTP than for KTP [17, 2].
Fig. 6. Damage probability curves in 1-on-1 mode for a given propagation direction and different polarizations. Both curves in each viewgraph are acquired on the same crystal. The dashed lines are guides to the eye.
In summary, we conclude from Figs. 5 and 6 that the laser induced damage threshold of RTP is sensitive to the polarization direction and not to the propagation direction. This is a difference to the situation in KDP where the propagation direction seems to be the most important parameter for testing with parallel beams [7, 4].
3.3. Discussion
In order to compare the damage threshold values reported in section 3.1 and section 3.2 we need to calculate a damage threshold for w-polarized light Tw from the known damage thresholds for u and v-polarized light Tu and Tv. For this purpose we assume that the observed damage is initiated by a critical force acting on the crystal lattice and being caused by the electrical field of the laser light. The critical electrical field Ec ∝ √T can then be considered as a measure for the resistance of the material for light of a certain polarization. The resistance for w-polarized light, E c,w, can be calculated using a linear model: The polarization vector w is normalized and its representation in terms of the normalized polarization vectors u and v is w = wu u + wv v. Each component wu and wv is then multiplied by the corresponding resistance and E c,w is the norm of the vector E c,w = E c,u w u u + E c,v w v v.
For example considering an x-cut crystal and using u = y-axis and v = z-axis we find for the Pockels cell configutation (w = (u+v)/√2):
and
Estimating from Fig. 5b Ty = 10.5 J/cm2 and from Fig. 6b Tz = 22 J/cm2, we would expect from this model a threshold of T 45° = 16.25 J/cm2 for the x-cut Pockels cell and a very similar value for the y-cut Pockels cell. The measured 1-on-1 damage threshold for the x-cut Pockels cell, estimated to 16.5 J/cm2 from Fig. 4, is very close to the calculated value.
For the y-cut Pockels cell however the measured 1-on-1 damage threshold is much smaller than the expected value. It is in particular smaller than the damage threshold of purely x-polarized light. The most probable explanation for the low threshold of the y-cut Pockels cell is the presence of frequency doubling. In spite of the small conversion efficiency, the amount of frequency doubled light during Pockels cell operation is not negligible and the 532 nm blocking filter in front of the camera is absolutely necessary in order to avoid CCD-damage.
Using the enhanced Sellmeier formulas in reference [14] we can calculate the phase mismatch for frequency doubling in x-cut and y-cut Pockels cells. In fact the phase mismatch is smaller for the y-cut crystals than for the x-cut crystals. Finally this calculation yields that the maximum frequency doubled intensity is about 5.8 times higher in a y-cut crystal than in an x-cut crystal [18]. Favre et al. showed that the LIDT of KTP decreases by a factor of ten in presence of both 1064 nm and 532 nm light compared to the LIDT of one of the wavelengths alone [19]. Our findings suggest that a similar mechanism may be important in RTP.
In conclusion the y-cut Pockels cell does not suffer from the LIDT anisotropy at 1064 nm, but more probably from the higher amount of second harmonic generation and a cooperative damage process similar to the one observed in KTP.
4. Summary and conclusions
We studied the laser induced damage threshold of RbTiOPO4 (RTP) crystals at 1064 nm. A strong anisotropy between x-cut crystals and y-cut crystals was found when testing the crystals in Pockels cell configuration, i.e. using a polarization at 45° with respect to the z-axis. We measured a 200-on-1 laser damage threshold of 12 J/cm2 for x-cut RTP-Pockels cells, whereas the 200-on-1 LIDT value for y-cut Pockels cells was only 6 J/cm2. The anisotropy in the LIDT is also visible in the 1-on-1 laser damage curves where only bulk damage was considered. In order to understand this observation, we carried out laser damage measurements with the polarization of the test beam aligned with the principal axes of the crystal and propagating the beam along one of the principal axes. The results showed that the propagation direction has no significant direct influence on the LIDT. The anisotropy of the LIDT with respect to the polarization of the test beam was confirmed. The threshold difference between x-polarized and y-polarized light however is small compared to the threshold difference between one of these polarizations and z-polarized light.
A simple model based on an anisotropic sensitivity of the material to laser induced damage allowed us to calculate the LIDT value of the x-cut Pockels cell. Neglecting second harmonic generation in Pockels cell configuration the model cannot explain the low LIDT measured for the y-cut Pockels cell. In particular the LIDT of the y-cut Pockels cell is smaller than the LIDT of the “weakest” involved polarization, i.e. x-polarization.
Including the promotion of laser damage by 532 nm light, we can explain qualitatively the lower threshold for y-cut Pockels cells. In this way the propagation direction influences indirectly, by the means of SHG, the LIDT of the RTP crystal.
Acknowledgments
The authors would like to thank the Direction Générale de l’Armement (DGA) for financial support, as well as Thales Laser and the Centre National d’Etudes Spatiales (CNES) for the good collaboration.
References and links
1. K. Zhang and X. Wang, “Structure sensitive properties of KTP-type crystals,” Chin. Sci. Bull. 46, 2028–2036 (2001). [CrossRef]
2. H. Albrecht, C. Bonnin, Y. Gromfeld, and M. A. Herrmann, “Characterization of RbTiOPO4 crystal for nonlinear and electro-optic applications,” in Solid state lasers XV, H. J. Hoffman and R. K. Shori, eds., Proc. SPIE 6100, 61001F (2006). [CrossRef]
3. J. D. Bierlein and H. Vanherzeele, “Potassium titanyl phosphate: properties and new applications,” J. Opt. Soc. Am. B 6, 622–633 (1989). [CrossRef]
4. H. Yoshida, T. Jitsuno, H. Fujita, M. Nakatsuka, M. Yoshimura, T. Sasaki, and K. Yoshida, “Investigation of bulk laser damage in KDP crystal as function of laser irradiation direction, polarization, and wavelength,” Appl. Phys. B 70, 195–201 (2000). [CrossRef]
5. Y. Furukawa, S. A. Markgraf, M. Sato, H. Yoshida, T. Sasaki, H. Fujita, and T. Yamanaka, “Investigation of the bulk laser damage of lithium triborate, LiB3O5, single crystals,” Appl. Phys. Lett. 65, 1480–1482 (1994). [CrossRef]
6. M. Pommiès, D. Damiani, B. Bertussi, J. Capoulade, H. Piombini, J. Natoli, and H. Mathis, “Detection and characterization of absorption heterogeneities in KH2PO4 crystals,” Opt. Commun. 267, 154–161 (2006). [CrossRef]
7. A. K. Burnham, M. Runkel, M. D. Feit, A. M. Rubenchik, R. L. Floyd, T. A. Land, W. J. Siekhaus, and R. A. Hawley-Fedder, “Laser-induced damage in deuterated potassium dihydrogen phosphate,” Appl. Opt. 42, 5483–5495 (2003). [CrossRef] [PubMed]
8. X. B. Hu, J. Y. Wang, H. J. Zhang, H. D. Jiang, H. Liu, X. D. Mu, and Y. J. Ding, “Dependence of photochromic damage on polarization in KTiOPO4 crystals,” J. Cryst. Growth 247, 137–140 (2003). [CrossRef]
9. H. Li, F. Zhou, X. Zhang, and W. Ji, “Bound electronic Kerr effect and self-focusing induced damage in second-harmonic-generation crystals,” Opt. Commun. 144, 75–81 (1997). [CrossRef]
10. H. Yoshida, T. Jitsuno, H. Fujita, M. Nakatsuka, T. Kamimura, M. Yoshimura, T. Sasaki, A. Miyamoto, and K. Yoshida, “Laser-induced damage in nonlinear crystals on irradiation direction and polarization,” in Laser-Induced Damage in Optical Materials, G. J. Exarhos, A. H. Guenther, M. R. Kozlowski, K. L. Lewis, and M. J. Soileau, eds., Proc. SPIE 3902, 418–422 (1999). [CrossRef]
11. Determination of laser-damage threshold of optical surfaces Part 2: S-on-1 test,” (2001).
12. L. Gallais and J. Y. Natoli, “Optimized metrology for laser-damage measurement: application to multiparameter study,” Appl. Opt. 42(6), 960–971 (2003). [CrossRef] [PubMed]
13. Determination of laser-damage threshold of optical surfaces Part 1: 1-on-1 test,” (2000).
14. Y. Guillien, B. Ménaert, J. P. Fève, P. Segonds, J. Douady, B. Boulanger, and O. Pacaud, “Crystal growth and refined Sellmeier equations over the complete transparency range of RbTiOPO4,” Opt. Mater. 22, 155–162 (2003). [CrossRef]
15. B. E. A. Saleh and M. C. Teich, “Optics of anisotropic media,” in Fundamentals of Photonics (Wiley-Interscience, New York, 1991), pp. 210–223.
16. Y. Jiang, L. E. Halliburton, M. Roth, M. Tseitlin, and N. Angert, “Hyperfine structure associated with the dominant radiation-induced trapped hole center in RbTiOPO4 crystals,” Phys. Status Solidi B 242, 2489–2496 (2005). [CrossRef]
17. Q. Jiang, A. Lovejoy, P. A. Thomas, K. B. Hutton, and R. C. C. Ward, “Ferroelectricity, conductivity, domain structure and poling conditions of rubidium titanyl phosphate,” J. Phys. D 33, 2831–2836 (2000). [CrossRef]
18. W. Lauterborn, T. Kurz, and M. Wiesenfeldt, “Nonlinear Optics,” in Coherent Optics: Fundamentals and Applications (Springer, 1999), pp. 239–265.
19. S. Favre, T. C. Sidler, and R. P. Salathe, “High-power long-pulse second harmonic generation and optical damage with free-running Nd:YAG laser,” IEEE J. Quantum Electron. 39, 733–740 (2003). [CrossRef]
