Open Access
Add to BibSonomyAbstract
We have demonstrated a tunable single frequency source of continuous-wave (CW) coherent ultraviolet (UV) radiation at λ3ω=326 nm. Laser light of a tunable diode laser at λω=977 nm was split and injected into two independent fiber amplifiers yielding 1 W and 0.4 W, respectively. The 1 W branch was resonantly frequency doubled, resulting in 120 mW of useful power at λ2ω=488 nm. The third harmonic was generated by summation of the second branch of λω and λ2ω which were enhanced by a doubly resonant cavity. This light source has TEM00 character and can be continuously tuned over more than 18 GHz. It is of interest for efficient laser cooling of In and potentially other applications.
©2008 Optical Society of America
1. Introduction
Laser cooling is a well established technique for controlling the motion of neutral atoms. While it is in widespread use for the generation of e.g. cold alkali gases for fundamental research, it is also of interest for applications, e.g. atomic nanofabrication (ANF) [1]. Technologically relevant materials such as indium and other elements of the third group of the periodic table and others frequently offer only short wavelength UV optical transitions suitable for laser cooling, i.e. exhibiting a closed strong dipole transition allowing a high rate of absorption-emission cycles. The applicability of these methods thus depends very much on the availability of robust and not too expensive sources of coherent, intense light at short wavelengths. In our laboratory we are experimenting with a laser cooled In atomic beam. With such a beam the methods of ANF could for example allow the generation of a fully 3D structured (In,Al)As crystal with periodically modulated In concentration.
For laser manipulation of In atoms the 5P 3/2-5D 5/2 states offer an excellent cycling transition for UV light at 325.6 nm. While at present no direct single frequency laser exists not to speak of tunable laser light sources at this wavelength, it can be generated straightforwardly by frequency doubling a 651 nm dye laser. For applications, however, robust solid state laser sources are much preferred. The recent and very rapid evolution of fiber lasers and fiber amplifiers suggests their application as coherent UV light sources by upconversion with appropriate nonlinear processes. Fiber lasers and amplifiers are compact and economic, and their light offers excellent diffraction limited beam quality. They are thus ideally suited to replace Ti:sapphire lasers for efficient UV generation [2, 3, 4, 5, 6] as has been shown for second and fourth harmonic generation in [7, 8, 9] already. The adoption of upconversion schemes from infrared to UV frequencies with fiber devices renders applications possible which up to now were impaired by the complex and costly operation of Ti:sapphire system, e. g. atom lithography with 3rd group elements.
In this paper we describe a fiber-based CW coherent UV light source which is based on generating the third harmonic of 977 nm light. The upconverted light at 326 nm is suitable for laser cooling of In atoms and other applications. A seed laser at the fundamental wavelength λω is amplified by two independent fiber amplifiers. Then, one of the infra-red light beams at λω is frequency doubled in an external cavity (EC). Subsequently, the λω light from the second fiber amplifier and the second-harmonic (SH) light at λ2ω are coupled into a doubly resonant cavity in order to generate third-harmonic light at λ3ω.
2. Two fiber amplifiers seeded by a single ECDL
The schematic of our experimental setup is shown in Fig. 1. As a seed source we employ a home-built external cavity diode laser (ECDL) emitting λω at 977 nm. Two 60 dB optical isolators protect the ECDL from backreflections. A 920 nm, 40W multi-mode diode laser module (Fianium PUMA-920-40) serves as a pump source. As a gain medium for the amplification of the seed power, an Yb-doped double clad fiber (YDCF) is selected. Because of a large absorption peak of Yb ions at ~977 nm, a special fiber (Crystal Fibre DC-27-6-Yb) with a small inner cladding is required in order to reduce the threshold pump power and suppress unwanted emission at ~1040 nm [10]. The diameters of the core and inner cladding of the fiber used in this experiment are 6 and 27 µm, respectively. The numerical aperture (NA) of the inner cladding is 0.55. The fibers are angle cleaved (~8°) in order to suppress back-reflection induced lasing. The measured pump absorption is 1.1 dB/m. Two independent fiber amplifiers are used rather than just splitting the output beam from one amplifier in order to obtain efficient operation and to escape the photo darkening effect caused primarily by a high population inversion which is required in 3 level Yb-doped fiber light sources at ~977 nm [11]. We split the seed and pump beams into two beams each and form two seed-pump beam pairs by combining them with a dichroic mirror. The pairs are then coupled to the YDCFs by a 3.1 mm focal length and 0.68 NA collimating lens. At each output port of the FAs, the pump light is eliminated by an interference filter.
Fig. 1. Experimental setup. Our system uses a diode laser system (ECDL), two fiber amplifiers (FA) and two external cavities (EC) for upconversion, λω→{λω,λ2ω}→λ3ω. ECDL, external cavity diode laser; EC, external cavity; YDCF, Yb-doped double clad fiber; HC, Hänsch-Couillaud method; OI, optical isolator; DM, dichroic mirror; M, mirror; CL, cylindrical lens; ML, mode-matching lens; IF, interference filter; PZT, piezo-electric transducer; BS, beam splitter; P, polarizer; QP, quartz plate.
Figure 2 shows the experimental and theoretical output power from both FAs as a function of the launched pump power. The lengths of the first (FA1, YDCF1) and the second (FA2, YDCF2) fibers are 1.2 m and 0.8 m, respectively. With 14 W pump power launched into each FA, we obtain 1.26 W (0.63 W) at λω from FA1 (FA2) using 20 mW (15 mW) seed power. Solid lines are the calculated output powers based on a rate equation model [12], and experiment and theory are in good agreement. The corresponding gains are 18 dB for FA1 and 16 dB for FA2. The amplified spontaneous emission (ASE) is suppressed by more than 33 dB below the carrier as measured with an optical spectrum analyzer. Although the YDCF is non-polarization preserving, we have small polarization fluctuations only. We further improve the polarization stability by introducing a polarizer at the expense of slow amplitude fluctuations at the few percent level.
3. Second harmonic generation in an external cavity
The output of FA1 is frequency-doubled in an external cavity [2, 3] as shown in the upper right corner of Fig. 1. We employ a Brewster-cut type-I phase-matched LiB3O5 (LBO) crystal with length 10 mm. The reflectivity of the input mirror M1 for λω is 98.3%, and the ouput mirror M4 is dichromatically coated to provide high reflectivity for λω and 95% transmission at λ2ω. The radii of curvature of M3 and M4 were chosen to be -50 mm, and the distance between the two curved mirrors is optimized to maximize the Boyd-Kleinman factor [13]. A small mirror M2 is mounted on a piezoelectric transducer (PZT) to lock the cavity length to λω by the Hänsch-Couillaud method [14]. The angle of incidence of M3 and M4 is chosen to compensate the astigmatism caused by the crystal Brewster surfaces [15]. The waists of the eigen mode at the center of the LBO crystal are calculated to be 36 µm in the tangential and 24 µm in the sagittal plane. The blue radiaton at λ2ω emitted from the cavity is mode-matched to EC2 by a spherical and two cylindrical lenses.
Fig. 2. Experimental output powers of FA1 (rectangles) and FA2 (circles) at 977 nm as a function of the launched pump power at ~920 nm. The solid lines show the theoretical output powers. Suppression of ASE is more than 33 dB below the carriers as measured with an optical spectrum analyzer.
Figure 3 shows the experimental and theoretical SH power at λ2ω as a function of the incident fundamental power. The maximum SH power measured behind the mirror M4 is 120 mW at 915 mW fundamental power. The circulating power of 87W is estimated by measuring the infrared power leaking from M3. A single-pass conversion efficiency of 2×10-5 W-1 is deduced corresponding to 46% of the calculated value. The difference between the measured and calculated single-pass conversion efficiency may be attributed to the deviation of the effective nonlinear coefficient from the specified value used for the calculation. Also insufficient knowledge about focusing parameters may play a role. The linear loss of 0.44% per round trip in the cavity is inferred from a measurement of the reflected fundamental power yielding an estimated mode-matching efficiency of 74% [6]. The overall conversion efficiency is 13%, and internal conversion efficiency with respect to the incoupled fundamental power is 27%. The solid line in Fig. 3 is a theoretical curve based on the measured parameters. The measured and the theoretical second harmonic power are in good agreement.
4. Third harmonic generation in a doubly resonant cavity
Third harmonic generation is based on the frequency upconversion of λ2ω light from the EC1 and another λω beam from FA2 [4, 5]. The blue light beam at λ2ω from the first enhancement cavity (EC1) and the second infrared light beam at λω from FA2 are combined by a dichroic mirror and then coupled to the second bow-tie shaped cavity through mode matching lenses as shown at the bottom right corner of Fig. 1. Here a 10-mm-long type-I phase-matched β-BaB2O4 (BBO) crystal was chosen, since it has high transmission at UV wavelengths. Both of the facets of the BBO crystal are antireflection coated for λω, λ2ω and λ3ω light. The input coupler M5 has 98.5% reflectivity at λω and 97% reflectivity at λ2ω. A small mirror M6 is mounted on a low-voltage PZT to keep the cavity resonant. The output coupler M8 has high reflectivity for λω, λ2ω, and 87% transmission for λ3ω. The radii of curvature of the curved mirrors M7 and M8 are -75 mm. The distance between the mirrors is optimized for maximum conversion efficiency. The angle of incidence of the second cavity is 12°. The waist is calculated to be 43 µm for λω and 24 µm for λ2ω light.
Fig. 3. Experimental (rectangles) and theoretical (solid line) second harmonic power at 488 nm as a function of the input power at 977 nm.
To maintain the second cavity simultaneously resonant with the λω and λ2ω wavelengths, two steps are used [5]. The length of EC2 is locked to the λω wavelength using the Hänsch-Couillaud method [14]. A small optical path length difference (OPD) between λω and λ2ω caused by dispersion in the cavity is compensated by slightly rotating two quartz plates mounted at Brewster’s angle. The thickness of each plate is 2 mm and tuning the angle by ±0.2° can cover 250 nm OPD corresponding to ~ the half wavelength of λ2ω. The quartz plates are mounted on counter rotating galvo motors and dithered at 270 Hz. Phase sensitive detection of the blue light at λ2ω leaking from M7 allows us to derive an error signal for compensating cavity dispersion, so that EC2 is resonant for the blue wavelength λ2ω, too.
The experimental values of third harmonic power at λ3ω as a function of the input power at λω are shown in Fig. 4. The maximum third harmonic power at λ3ω measured behind the output coupler M8 is 12 mW for 405 mW at λω and 120 mW at λ2ω, respectively. The measured circulating powers are 26 W for λω and 2.4 W for λ2ω. The estimated single-pass conversion efficiency is 2.2×10-4 W-1, which is 88% of the calculated value. By measuring the rejected power of the blue and infrared beam from EC2, the linear losses are estimated to be 1.23% for λω and 2% for λ2ω. The mode-matching efficiencies are calculated to be 80% for λω and 49% for λ2ω beam. The overall conversion efficiency is 2.3%, and the internal efficiency with respect to the coupled fundamental waves is 3.6%. The result of a theoretical calculation [16] of the UV power based on the measured parameters is shown in the solid line in Fig. 4, and the theory agrees well with the experimental result. The mode-hop free tuning range of UV light is more than 18 GHz.
Fig. 4. Experimental (rectangles) and theoretical (solid line) third harmonic power at 325 nm as a function of the input power at 977 nm. Blue light power at 488 nm is fixed at 120 mW.
5. Conclusion
In summary, we have developed a tunable single-mode UV source at λ3ω. Two FAs using a single seed laser were built to generate 1 W (FA1) and 0.4 W (FA2) fundamental power for nonlinear frequency conversion. The output of FA1 was frequency doubled in a cavity to generate 120 mW blue light at λ2ω. To generate UV light at λ3ω through third harmonic generation, the output of FA2 and the second harmonic light beam at λ2ω were coupled into the second cavity which was doubly resonant with both λω and λ2ω wavelengths. We have obtained 12 mW UV light at λ3ω and tuned the UV laser frequency continuously more than 18 GHz.
Realistic improvements of our system include a longer fiber for FA2 (1 W instead of 0.4 W) and a reduction of intracavity loss in EC1 (400mWinstead of 120mW at λ2ω). Then we should approach 100 mW output power at λ3ω=326 nm. Already with the current setup a UV light source suitable for laser cooling of indium with the 5P 3/2-5D 5/2 closed transition is available.
Acknowledgments
This work was supported by the NANOCOLD project of the European Commission. We thank P. Georges for fruitful discussions on high power fundamental sources at 977 nm.
References and links
1. D. Meschede and H. Metcalf, “Atomic nanofabrication: atomic deposition and lithography by laser and magnetic forces,” J. Phys. D 36, R17–R38 (2003). [CrossRef]
2. E. S. Polzik and H. J. Kimble, “Frequency doubling with KNbO3 in an external cavity,” Opt. Lett. 16, 1400–14021991). [CrossRef] [PubMed]
3. S. Bourzeix, M. D. Plimmer, F. Nez, L. Julien, and F. Biraben, “Efficient frequency doubling of a continuous wavetitanium:sapphire laser in an external enhancement cavity,” Opt. Commun. 99, 89–94 (1993). [CrossRef]
4. S. Sayama and M. Ohtsu, “Tunable UV CW generation by frequency tripling of a Ti:sapphire laser,” Opt. Commun. 137, 295–298 (1997). [CrossRef]
5. J. Mes, E. J. van Duijin, R. Zinkstok, S. Witte, and W. Hogervorst, “Third-harmonic generation of a continuous-wave Ti:Sapphire laser in external resonant cavities,” Appl. Phys. Lett. 82, 4423–4425 (2003). [CrossRef]
6. H. Kumagai, K. Midorikawa, T. Iwane, and M. Obara, “Efficient sum-frequency generation of continuouswave single-frequency coherent light at 252 nm with dual wavelength enhancement,” Opt. Lett. 28, 1969–1971 (2003). [CrossRef] [PubMed]
7. A. Friedenauer, F. Markert, H. Schmitz, L. Petersen, S. Kahra, M. Herrmann, Th. Udem, T. W. Hänsch, and T. Schätz, “High power all solid state laser system near 280 nm,” Appl. Phys. B. 84, 371–373 (2006). [CrossRef]
8. P. Herskind, J. Lindballe, C. Clausen, J. L. Sorensen, and M. Drewsen, “Second-harmonic generation of light at 544 and 272 nm from an ytterbium-doped distributed-feedback fiber laser,” Opt. Lett. 32, 268–270 (2007). [CrossRef] [PubMed]
9. T. Südmeyer, Y. Imai, H. Masuda, N. Eguchi, M. Saito, and S. Kubota, “Efficient 2nd and 4th harmonic generation of a single-frequency, continuous-wave fiber amplifier,” Opt. Express 16, 1546–1551 (2008). [CrossRef] [PubMed]
10. J. Nilsson, J. D. Minelly, R. Paschotta, A. C. Tropper, and D. C. Hanna, “Ring-doped cladding-pumped single-mode three-level fiber laser,” Opt. Lett.. 23, 355–357 (1997). [CrossRef]
11. S. Jetschke, S. Unger, U. Röpke, and J. Kirchhof, “Photodarkening in Yb doped fibers: experimental evidence of equilibrium state depending on the pump power,” Opt. Express 15, 14838–14843 (2007). [CrossRef] [PubMed]
12. R. Paschotta, J. Nillson, A. C. Tropper, and D. C. Hanna, “Ytterbium-doped fiber amplifiers,” IEEE J. Quantum Electron. 33, 1049–1056 (1997). [CrossRef]
13. G. D. Boyd and D. A. Kleinman, “Parametric interaction of focused Gaussian light beams,” J. Appl. Phys. 39, 3597–3639 (1968). [CrossRef]
14. T.W. Hänsch and B. Couillaud, “Laser frequency stabilization by polarization spectroscopy of a reflecting reference cavity,” Opt. Commun. 35, 441–444 (1980). [CrossRef]
15. J. Hald, “Second harmonic generation in an external ring cavity with a Brewster-cut nonlinear crystal: theoretical considerations,” Opt. Commun. 197, 169–173 (2001). [CrossRef]
16. Y. Kaneda and S. Kubota, “Theoretical treatment, simulation, and experiments of doubly resonant sum-frequency mixing in an external resonator,” Appl. Opt. 36, 7766–7775 (1997). [CrossRef]
