We report the generation of nanoseconds radiation at 177.3 nm with a maximum average power of 34.7 mW by second harmonic generation (SHG) in a 2.06-mm thick KBe2BO3F2 (KBBF) crystal pumped with a homemade 4.2 W nanoseconds Nd: YAG laser at 355 nm operating at 10 KHz and 49 ns, which corresponds to an energy efficiency of ~0.826%. To our knowledge, it is the highest power generated at 177.3 nm. The dependence of phase matching angle of KBBF on temperature is presented for the first time. We also present the details on the measures for stable operation of a 4 mW nanosecond output at 177.3 nm with a lower pumping power in a thinner 1.37-mm KBBF crystal, and the stable output power is improved by about 20 times compared with previous results.
©2009 Optical Society of America
Vacuum ultraviolet (VUV) laser sources play an important role in VUV semiconductor photolithography, micro-machining, high-resolution photoelectron spectroscopy and photochemical synthesis [1,2]. New phenomena are expected to be observed with VUV laser used in Raman spectrometers, and it will promote the researches in the fields of photo-electronic materials and catalysis, especially in the study of structure characterization of catalyst. So far, the uniaxial crystal KBe2BO3F2 (KBBF) has been successfully applied in VUV generations by sixth harmonic generation of Nd: YAG at 1064 nm  or fourth harmonic generation of Ti: sapphire lasers from 700 to 800 nm . However, most of work reported so far was either pumped by ps lasers [5–8] or by fs lasers  and the bandwidth is rather broad. Nanosecond VUV generation was also previously reported by frequency-doubling of a Ti: sapphire laser in KBBF and the output was able to cover from 175 to 210 nm, but the output power around 177 nm was in the order of 0.1 mW or less . No work is reported in generating radiation at 177.3 nm by frequency doubling of the third harmonics of Nd: YAG nanosecond laser at 355 nm with high output power and narrow bandwidth. The generation of nanosecond VUV by frequency doubling in KBBF crystal is difficult due to the facts that the pumping intensity of a nanosecond laser is rather low and the thickness of KBBF is generally very thin (1-2 mm) because of its layered structure. It prevents the SHG from generating effectively. However, the generation of nanosecond VUV is very useful since the bandwidth of nanosecond radiation can be very narrow and it can be used in high-resolution spectroscopy. So far the energy resolution of vacuum ultraviolet laser-based angle-resolved photoemission is only 0.36 meV due to the limited bandwidth of the ps-VUV source . With a nanosecond VUV, the energy resolution can be improved by 1-2 orders of magnitude. In our work, the bandwidth of the nanosecond pumping laser at 355nm was measured to be less than 1 cm−1, which corresponds to ~0.1 meV. The bandwidth can be reduced to 0.003 cm−1 by inserting an intra-cavity etalon in the oscillator when it is needed and the corresponding bandwidth can be as narrow as ~0.0004 meV.
In this paper, we report the generation of nanosecond VUV radiation at 177.3 nm by second harmonic generation of a 4.2 W nanosecond frequency-tripled Nd: YAG laser in two KBBF crystals, whose thickness are 2.06 mm and 1.37 mm, respectively. The effect of the crystal thickness on the conversion efficiency was experimentally investigated. Due to the considerable absorption of 177.3 nm radiation by the KBBF crystal, it may heat up the crystal and destroy the phase-matching condition, the instability of the high output 177.3 nm radiation has become an issue, especially when a 2.06-mm thick crystal is used. We investigated the measures for improving the stability of the output at 177.3 nm, including using a thinner crystal, pumping it at a lower power density and cooling the KBBF crystal, while maintaining a considerably high output power.
2. The experimental details
2.1 The laser system and vacuum chamber
The experimental setup for generating VUV radiation at 177.3 nm was basically the same as that in the earlier report  except that it was pumped by a home-made high-output and high beam-quality nanosecond laser system. The laser system together with the vacuum chamber for nanosecond VUV generation is shown in Fig. 1 .
Experimentally, we adopt two identical side-pumped laser modules (LM), which is different from general end pumping configuration, in the resonators to acquire high out power with high beam quality. A Nd:YAG rod with size of Ø2 × 76mm is used in each laser module to obtain near diffraction limited mode. A 90° rotator is inserted between two laser modules to compensate the thermal birefringence, which makes the stable regions of different polarization overlap each other and reduces the unstable region. Symmetrical plane-plane resonators are adopted to achieve maximum stable region. An acousto-optic modulator with high diffraction is used to generate a repetition rate of 10 kHz. Fundamental pumping laser with high beam quality is critical to attain high conversion efficiency from the fundamental to the second harmonics. To acquire high beam quality, the operation of laser is designed at the critical unstable point of the resonators. When it works at 10 kHz repetition rate, the output 1064nm laser power is 30 W, and the measured M2 factors in x and y directions are both 1.48. A type I non-critically phase matched LBO crystal (θ = 90○,ϕ = 0○) with the size of 4 × 4 × 40 mm is used to generate the second harmonic wave. And a type II phase matching LBO crystal (θ = 45.1○, ϕ = 90○) with size of 4 × 4 × 30mm is used to mix the second harmonics and the fundamental to achieve the third harmonic generation (THG). Because of temperature sensitivity of the nonlinear frequency conversion, it is necessary to control the temperature of LBO crystal precisely to achieve highly stable output of THG at 355nm. In the experiment, the LBO is placed in the oven, whose temperature can be controlled precisely and the variation of the oven’s temperature is less than ± 0.05○C. A prism is used to split 1064nm, 532nm and 355nm lasers. An oscilloscope Tektronix DPO4104 and a fast photo-diode are used to measure the repetition rate and the pulse width of the laser at 355nm. The output at 355nm was used to pump a KBBF crystal after being focused onto a KBBF-prism coupling device (KBBF-PCD) by a lens with f = 500 mm. The beam waist of the pumping ordinary-wave at 355nm beam is located at the center of KBBF and the spot size is 284µm (HWHM). The measured repetition rate and pulse width of 355nm laser are 10 kHz and 49 ns, respectively, and the maximum output at 355 nm is 4.74 W, which corresponds to a pumping intensity of 15.2 MW/cm2 at the beam waist. As a comparison, the output power of commercially available 10-kHz, nanosecond product at 355nm is usually less than 2 W, which results in a much lower output power at 177.3nm compared with that of using a more than 4-W pumping laser.
2.2The prism-coupling device (PCD)
Because of the layer-structure of the KBBF crystal , it cannot grow thick enough for cutting at appropriate angles to satisfy the need of applications, a Prism-Coupling Device (PCD), in which the KBBF crystal is optically contacted with the prisms, is adopted and it solves the problem of the phase-matching in the KBBF. The structure of PCD is shown in Fig. 2 and the corresponding details can be obtained in the previous publications [4–9].
There were two types of PCD used in our experiment, the prism angle θ of them are both at 68.6оso that the angle of incidence in the crystal satisfies the phase matching condition for SHG of the pump at 355 nm. In the earlier design, two identical CaF2 prisms are used in the PCD for the input and the output coupling. Although CaF2 material is generally considered to have a relatively good moisture resistivity, however, it was found experimentally that at a high pumping intensity of UV the surface of CaF2 prism could be damaged after several months and no damage is found on the surface of fused silica prism with the KBBF-PCD purged by nitrogen when it is pumped by the laser, but leave it un-purged when it is not under working condition. Such a phenomenon was tentatively attributed to the slight hygroscopicity of CaF2, which may absorb a layer of moisture at the surface and subsequently absorb the UV pump beam through nonlinear optical processes, especially when it is pumped by a high-repetition nanosecond UV laser, whose damage threshold is much lower than that of picoseconds and femtosecond pumping. In addition, CaF2 is not hard enough to be polished as fine as the fused silica which may further reduce the damage threshold. The fused silica, on the other hand, has a very high laser damage threshold for the UV at 355 nm and, therefore, it is more suitable to be used as input coupling prism than CaF2. The CaF2 output coupler can also find surface optical damage after long term operation pumped by the nanosecond UV laser. The detailed mechanism for the optical damage is still under study. However, considering the fact that the absorption in fused silica is ~7% higher than the absorption in CaF2 at 177 nm (the data are deduced from the transmission of ~90% for the fused silica and ~97% for CaF2 in 5-mm-thick samples after taking into account of Fresnel loss  and the fact that size of the prism used in PCD is ~9x23x9 mm3, and taking into account the expensive VUV photons generated by KBBF, the fused silica is generally not recommended for the output coupler. Therefore, in the newly designed PCD, a fused silica prism is used for input coupling and while a CaF2 prism is still used for the output coupling, as shown in Fig. 2(b).
In our experiment, two KBBF crystals with different thickness were used. The first KBBF crystal has a thickness of 2.06 mm and both the input and output coupling prisms for KBBF-PCD are made of CaF2, while the second crystal is 1.37-mm thick and it employs a fused silica prism for input coupling in the KBBF-PCD. As mentioned earlier, the fused silica prism allows long-term pumping of high power laser at 355 nm without causing optical damage. The generated VUV output at 177.3 nm was measured by VUV power meter (LP-3A, Physcience Opto-Electronics).
3. Experimental results and discussions
3.1 Dependence of the 177.3 nm output on pumping power and crystal thickness
Since the efficiency of SHG depends heavily on the thickness of crystal and the pumping intensity of the pump, we first measured the dependence of the output 177.3 nm on the pumping power at 355 nm with two KBBF crystals. The results are shown in Fig. 3 . It was found that with the 2.06-mm thick KBBF crystal the output could be as high as 34.7 mW when the pumping power was increased to 4.2 W, which corresponds to a pumping intensity of 13.5 MW/cm2, and the energy conversion efficiency is 0.826%, while the corresponding output power at 177.3 nm was about 14.1 mW from the 1.37-mm thick KBBF crystal at the same pumping intensity and the corresponding conversion efficiency is 0.336%. From Fig. 3, one can see that based on the trend of VUV output power versus pumping power a higher output at 177.3 nm is expected when the pumping power is increased. It is also clear that at the pumping power available in our experiment, the thicker KBBF generates 177.3 nm much more effectively than the thinner crystal, indicating that the pumping intensity is far from saturation and a higher efficiency is expected by using a higher pumping intensity as long as the crystal is not damaged by the pumping laser.
The 34.7 mW output power at 177.3 nm obtained from the 2.06-mm thick crystal is the highest average output at 177.3 nm ever reported. The high output can be attributed to the following factors: (a) high pumping power, (b) good beam quality of the pump beam, and (c) thick (2.06 mm) KBBF crystal used. However, this high power output was very instable and it decreases gradually as the pumping continues. By adjusting the angle of KBBF crystal, the output at 177.3 nm can be increased again and reached a maximum when the phase-matching condition was satisfied, however, it cannot be hold at the highest output.
3.2 Heat problem due to absorption of the generated VUV at 177.3 nm and the UV pump at 355 nm
The high output at more 30 mW was hard to maintain and it dropped significantly as the pumping continued. It was found later that at a pumping power of more than 4 W at 355 nm for a period of time, the high output cannot be recovered completely by readjusting the orientation of the crystal, indicating the formation of color center in the crystal due probably to the absorption of strong VUV at 177.3 nm and two-photon absorption of the pumping laser at 355 nm under such strong pumping.
With a lower pumping intensity, the output was reduced to a lower level. With the pumping power keeping at a low level and when the output can be decreased due to the phase mismatch, but it can always be recovered by readjusting the angle of the KBBF crystal properly. However, it was very difficult to maintain stable operation for the 2.06-mm crystal. The reduction of the output is mainly attributed to the evaluated temperature of the crystal through absorbing the VUV radiation generated by the KBBF crystal or two-photon absorption of the pump as can be seen from the absorption curve in Fig. 4 . In addition, the prism since CaF2 also absorbs the radiation at 177.3 nm. The absorption resulted in the heating of KBBF- PCD and destroyed the phase-matching condition.
The measured absorption coefficient of KBBF as a function of wavelength is shown in Fig. 4, from which we can get that the absorption coefficient of KBBF is 1.25 cm−1 at 177.3nm. The absorption produces heat in the crystal and in the coupling prisms. The heat could be accumulated in the KBBF-PCD when it is pumped by the laser with a high repetition rate (10-KHz in our case) and high average power (> 4 W) for a period of time, and thus elevates the temperature of the crystal. The elevated temperature of the crystal would break the phase-matching condition since the phase-matching angle is temperature dependent and thus reduce the output of the VUV at 177.3 nm. Therefore, the high output from 2.06-mm thick KBBF is very hard to maintain.
In addition, the absorption of the generated VUV and two-photon absorption of the high-power pumping laser in the output coupling prism also results in building up of heat in the PCD and elevates the operation temperature of the KBBF crystal. Therefore, heat management is critical for high-power long-term stable operation of the system.
In order to quantify the dependence of output power on the angle deviation from phase matching angle, the output power of 177.3 nm is measured when the input angle in KBBF crystal is changed, the measured results are shown in Fig. 5 . The asymmetry in the figure remains to be understood. We can get from Fig. 5 that the permitted angle tolerance (FWHM) is about 0.08 degree, which is very tight. Therefore, a small angle deviation from phase matching angle may result in significant decrease of the output power at 177.3 nm.
In order to obtain quantitative relationship between the phase-angle and the temperature of the crystal, the phase matching angle of KBBF crystal was measured at various temperatures. Direct measurement of crystal’s temperature was hard to perform and we measured the temperature of the metal holder for the KBBF PCD, which gave the relative temperature variation of the KBBF crystal. In the experiment, the temperature of metal holder is changed from 15 °C to 30 °C. To ensure the accuracy of measurement, the phase matching angle was measured one hour after the metal holder’s temperature of KBBF-PCD reached the set temperature. The phase matching angle variation versus KBBF-PCD temperature variation is shown in Fig. 6 . It was found from Fig. 6 that the variation of phase matching angle was about 0.05° when the temperature of KBBF-PCD changed from 15 °C to 30 °C.
3.3 Stabilization of output at 177.3 nm
The instability of the output was attributed to poor heat management in the early stage of the experimental setup where no cooling measure was taken. In the case of low pumping intensity or thin crystal, the generated VUV radiation is relatively low, for example, less than 1 mW, and the heat problem is less significant. The phase-matching angle of the crystal could be maintained by adjusting the orientation of the crystal properly and no cooling is needed. We mainly concentrated on stabilization of a medium output, for example, a few mW, which is high enough for most of applications.
Since KBBF, CaF2 and fused silica have relatively low thermal conductivity, the thermal conductivities of KBBF, CaF2 and fused silica are ~2.5 W / (m · K), 9.7 W / (m · K) and 1.38 W / (m · K), respectively. The heat produced by the absorption of VUV by the crystal and the prisms cannot be dissipated effectively, especially when high power VUV is generated. The accumulated heat in the PCD then elevated the temperature of KBBF crystal and broke the phase matching condition and thus reduced the output radiation of 177.3 nm. In the case of high pumping power and thick crystal, the generated VUV radiation was high and the heating became significant, and, therefore, it was very difficult to maintain long-term stable operation. It is clear that in order to maintain a stable output at 177.3 nm, it is necessary to maintain the temperature of KBBF-PCD at constant and some measures need to be taken to dissipate the generated heat and to stabilize the operation temperature of KBBF-PCD. To solve the problem, both water-cooling to the crystal holder and direct nitrogen-purging to the crystal were adopted simultaneously to maintain the condition of phase matching.
In order to achieve long-term stable operation, a fused silica prism was used for input coupling in the KBBF-PCD so that it allowed long term pumping of high-power laser at 355nm without optical damage on the surface. A 1.37-mm thick KBBF was used in the KBBF-PCD and the pumping power of laser was reduced to 2.3 W to reduce the output power at 177.3 nm to about 4 mW, as can be seen in Fig. 3. In addition, cooling of KBBF was introduced by using a metal holder, and the KBBF-PCD was purged by N2 flow, which is also helpful in taking the heat away from the crystal. With an average pumping power of 2.3 W, corresponding to a pumping intensity of 7.4 MW/cm2, it took about 45 minutes for the PCD to reach thermal equilibrium and for the output at 177.3 nm to become stable.
To see the stability of the output at 177.3 nm after taking the above-mentioned measures to manage the heat, the output of 177.3 nm radiation generated from a thin 1.37-mm thick KBBF crystal and the pumping power at 355 nm were monitored simultaneously. The variation of temperature is controlled within 0.5оC. After KBBF-PCD reaching the thermal equilibrium, the powers of the pump at 355 nm and the radiation at 177.3 nm were measured every 5 minutes for a period of two hours. The data are presented in Fig. 7 . It is seen that the fluctuation at 177.3 nm follows the fluctuation of the pumping at 355 nm well and the fluctuations of the output at 177.3 nm was found to be ± 3.4% in 2 hours at an average power of 4 mW. From Fig. 7 it is clear that with the cooling measures worked well and it can provide stable output around 4 mW at 177.3 nm. Compared with the 0.2 mW output published previously , the stable output was improved by about 20 times. In the experiments, higher output power of more than 4mW is also tested, and it cannot be as stable as that at 4-mW or lower. We are still trying to work out the cooling condition for producing higher output using the 1.37-mm crystal.
With the 2.06-mm KBBF crystal, the output at 177.3 nm can be much higher than that generated from the 1.37-mm crystal. However, it was very difficult to be stabilized at this high output. We are still working on figuring out a more effective way to stabilize the highest output.
In summary, with a 2.06-mm thick KBBF crystal, a maximum output of 34.7 mW at 177.3 nm has been achieved when it is pumped with a 4.2 W nanosecond laser at 355nm operated at repetition rate of 10 kHz and pulse width of 49 ns. This is the highest power at 177.3 nm reported so far and it corresponds to an energy conversion efficiency of 0.826%. The variation of phase matching angle of KBBF at 177.3nm is 0.05° when its temperature changed from 15 °C to 30 °C, while the temperature bandwidth for SHG at 355 nm in KBBF is found to be about 0.08 °C. By using a 1.37-mm thick crystal, cooling the crystal properly, adopting a fused silica prism for input coupling in the PCD, and pumping the crystal with a lower average laser power of 2.3 W at 355 nm, the output at 177.3 nm can be stabilized at an average output of 4 mW with a fluctuation of ± 3.4% in 2 hours.
This work was supported by the State Key Program for Basic Research of China (Grant No. 2004CB619006 and 2004CB619006), the National High Technology Research and Development Program (No.2006AA030104), the National Natural Science Foundation of China (Grant No. 50590404), and the Knowledge Innovation Program of Chinese Academy of Sciences (Grant No. KJCX2.Y200420 and KJCX2.YW.H03).
Dr. Jing-Yuan Zhang thanks the support of Georgia Southern University during his sabbatical leave.
+ Permanent address: Physics Department, Georgia Southern University, Statesboro, GA 30460, USA
References and links
1. T. Kiss, T. Shimojima, F. Kanetaka, K. Kanai, T. Yokoya, S. Shin, Y. Onuki, T. Togashi, C. Q. Zhang, C. T. Chen, and S. Watanabe, “Ultrahigh-resolution photoemission spectroscopy of superconductors using a VUV laser,” J. Electron Spectrosc. Relat. Phenom. 144–147, 953–956 (2005). [CrossRef]
2. T. Suganuma, H. Kubo, O. Wakabayashi, H. Mizoguchi, K. Nakao, Y. Nabekawa, T. Togashi, and S. Watanabe, “157-nm coherent light source as an inspection tool for F(2) laser lithography,” Opt. Lett. 27(1), 46–48 (2002). [CrossRef]
3. C. Chen, Z. Xu, D. Deng, J.-y. Zhang, G. K. Wong, B. Wu, N. Ye, and D. Tang, “The vacuum ultraviolet phase-matching characteristics of nonlinear optical KBe2BO3F2 crystal,” Appl. Phys. Lett. 68(21), 2930–2932 (1996). [CrossRef]
4. C. Chen, J. Lu, T. Togashi, T. Suganuma, T. Sekikawa, S. Watanabe, Z. Xu, and J. Wang, “Second-harmonic generation from a KBe(2) BO(3)F(2) crystal in the deep ultraviolet,” Opt. Lett. 27(8), 637–639 (2002). [CrossRef]
5. G. Wang, X. Wang, Y. Zhou, Y. Chen, C. Li, Y. Zhu, Z. Xu, C. Chen, C Li, Y Zhu, Z Xu, and C Chen, “12.95 mW sixth harmonic generation with KBe2BO3F2 crystal,” Appl. Phys. B 91(1), 95–97 (2008). [CrossRef]
6. G. Liu, G. Wang, Y. Zhu, H. Zhang, G. Zhang, X. Wang, Y. Zhou, W. Zhang, H. Liu, L. Zhao, J. Meng, X. Dong, C. Chen, Z. Xu, and X. J. Zhou, “Development of a vacuum ultraviolet laser-based angle-resolved photoemission system with a superhigh energy resolution better than 1 meV,” Rev. Sci. Instrum. 79(2), 023105 (2008). [CrossRef] [PubMed]
7. C.-T. Chen, “J. H. LÜ, G. L. Wang, Z. Y. Xu, J. Y. Wang, C. Q. Zhang, Y. G. Liu, “Deep ultraviolet Harmonic Generation with KBe2BO3F2 Crystal,” Chin. Phys. Lett. 18(8), 1081 (2001). [CrossRef]
8. T. Togashi, T. Kanai, T. Sekikawa, S. Watanabe, C. Chen, C. Zhang, Z. Xu, and J. Wang, “Generation of vacuum-ultraviolet light by an optically contacted, prism-coupled KBe2BO3F2 crystal,” Opt. Lett. 28(4), 254–256 (2003). [CrossRef] [PubMed]
9. H. Zhang, G. Wang, L. Guo, A. Bo, D. Cui, Z. Xu, R. Li, Y. Zhu, X. Wang, and C. Chen, “175 to 210 nm widely tunable deep-ultraviolet light generation based on KBBF crystal,” Appl. Phys. (Berl.) 93, 323–326 (2008).
10. L. Mei, C. He, and C. Chen, “Flux growth of KBe2BO3F2,” J. Cryst. Growth 132(3-4), 609–610 (1993). [CrossRef]
11. Experimental data provided by Acton Optics & Coatings, 2009.
12. Y. Zhou, G. L. Wang, C. M. Li, Q. J. Peng, D. F. Cui, Z. Y. Xu, X. Y. Wang, Y. Zhu, C. T. Chen, G. D. Liu, X. L. Dong, and X. J. Zhou, “Sixth Harmonic of A Nd:YVO4 Laser Generation In KBBF for ARPES,” Chin. Phys. Lett. 25(3), 963–965 (2008). [CrossRef]