Ultraviolet lasers have many important applications in biohazard detection and environmental sensing, yet very few stable and robust ultraviolet lasers exist. The third and fourth harmonics of neodymium lasers (355 and 266 nm) are fairly easy to generate, but there is a need for reliable sources at intermediate wavelengths. Recently, scientists at Macquarie University in Sydney, Australia, demonstrated a Raman laser that can be tuned to generate one of more than seven wavelengths between the second and third harmonics of neodymium.The scientists have shown how Raman media can be used to enable wavelength switching of a laser between several visible wavelengths (see “Raman Lasers Offer Power and Wavelength Versatility,” Photonics Spectra, July 2006, page 52). They have now extended that work into the ultraviolet. They pumped their Raman laser, which used KGd(WO4)2 (KGW) as the Raman material, with 532-nm pulses from a Q-switched, frequency-doubled Nd:YAG laser supplied by Lumonics. An intracavity ß-barium borate (BBO) crystal generated the ultraviolet output by sum-frequency and second-harmonic generation of the pump radiation and/or the first four Stokes lines generated by the Raman laser (Figure 1).Figure 1. Scientists generated eight ultraviolet wavelengths by mixing various pairs of Stokes and/or pump photons in a ß = barium borate (BBO) crystal. Images reprinted with permission of Optics Letters.A dichroic mirror coupled the ultraviolet radiation out of the resonator and prevented it from reaching the KGW crystal, which does not transmit in the ultraviolet. The scientists rotated the BBO crystal to phase-match the mixing pairs of fundamental and Stokes beams inside the Raman laser resonator. After rotating the BBO, they readjusted the mirror behind it to maximize the power generated. Including the second harmonic of the 532-nm fundamental, they generated ultraviolet outputs at eight wavelengths, each at a different orientation of the BBO crystal (Figure 2).Figure 2. Energies generated at each of the eight ultraviolet wavelengths are shown as a function of pump energy. The lines are added as an aid to the eye, but to avoid clutter no line is drawn for the 311-nm data (solid squares).The data in Figure 2 were generated when the pump laser operated at 10 Hz, producing 10-ns-long green pulses. The laser’s ultraviolet output can be explained qualitatively by considering the intracavity dynamics. The pump’s second harmonic, at 266 nm, increased rapidly with rising pump energy, but then saturated as the intracavity fundamental was held relatively constant by Stokes generation.Likewise, the ultraviolet outputs involving the first and second Stokes lines (i.e., at wavelengths between 272 and 295 nm) were generated with decreasing efficiency at pump energies above ∼4 mJ, because intracavity power at the first and second Stokes wavelengths increased less rapidly as the third and fourth Stokes lines reached threshold. But the efficiency decrease for ultraviolet output involving the third and fourth Stokes was less noticeable, because the higher Stokes orders never reached threshold.To demonstrate the usefulness of the Raman technique to generate significant average power in the ultraviolet, the scientists pumped the Raman laser with another 532-nm laser, this one operating at 5 kHz. They generated as much as 47 mW at 279.4 nm, and 10 or more milliwatts at each of the ultraviolet lines below 300 nm.The scientists believe that there are many improvements possible before the Raman ultraviolet laser reaches its full potential. They utilized only one Raman shift of KGW (at 901 cm–1), but the material has another strong shift at 768 cm–1. And, of course, other ultraviolet lines could be generated with other Raman crystals. Their laser’s performance at wavelengths longer than 300 nm was limited chiefly by low mirror reflectivity at those wavelengths, and it could be improved with better optics.Moreover, they made little attempt to optimize beam sizes in the laser resonator during these initial experiments. As a result, conversion efficiencies were several percent at best, but they believe that, by improving intracavity focusing and the overall resonator design, they could boost the pump-to-ultraviolet efficiencies to higher than 10 percent.Optics Letters, April 1, 2007, pp. 814-816.