Erbium-doped YAG lasers that emit infrared (IR) light at 3 µm have become accepted for a variety of medical applications, including dermatology, stomatology, and ophthalmology. Wave lengths at about 3 µm are attractive because human tissue absorbs most of the energy at these wavelengths, making it possible to ablate both hard and soft tissue. A drawback to these systems, however, has been the awkwardness of transmitting the 3-µm energy from the source laser to the tissue surface. Of the several types of fibers that can transmit such long wavelengths—including those based on zirconium fluoride, sapphire, and hollow fibers—all are stiff and must be protected, adding thickness to the endoscope or handpiece.
An alternative system for producing 3-µm energy has been demonstrated by Rudolf Steiner at the University of Ulm (Germany), H.-J. Pohl at VITCOM (Jena, Germany), and I. A. Mironov at the Vavilov State Optical Institute (St. Petersburg, Russia).1 Instead of transmitting the energy from an external 3-µm laser through a fiber to the tissue, the researchers developed a tiny, optically pumped laser at the distal end of a fiber that converts light at near-IR wavelengths to 2.9-µm light very near the tissue.
The fiber carries the more easily transmitted 1.12-µm energy from a Nd:YAG pump laser to the 2.9-µm laser. According to the researchers, the converter [laser] unit can be miniaturized to dimensions of 2 mm in diameter by 20 mm in length—no longer than a fiber tip and small enough to be introduced into an endoscope.
Various pumping schemes
Instead of using an Er:YAG laser to generate the 3-µm light, Steiner`s group uses an end-pumped holmium-doped barium ytterbium fluoride (BaYb2F8) crystal to convert the near-IR light into the mid-IR range at 2.9 µm. This rare-earth crystal has been the subject of work at Vavilov for several years.2
The researchers wrestled with various pumping schemes. It is possible to use a Nd:YLF laser at 1047 nm to pump the ytterbium ions, which then transfer energy to the holmium ions. However, the re searchers found a more-efficient method based on direct pumping of the holmium ions; pumping with a Nd:YAG laser at 1120 nm directly excites these ions into the upper laser level. The upper lifetime is 2.5 ms, and the threshold energy absorbed is 0.2 J/cm2.
With this scheme, the efficiency is sensitive to the lifetime of the lower level, but not very sensitive to the holmium and ytterbium ion concentrations or to the pump power density. Efficiencies of up to 25% have been obtained, which means that a 2-J pump pulse can be converted into a 2.9-µm output pulse of 500 mJ. A miniaturized version (1.6-mm diameter, 13 mm long) demonstrated usable powers of 200 mJ.
Another advantage of this laser crystal is that the temperature dependence is low. Output energy remains constant from room temperature to nearly 100°C. This stability allows air cooling to be used for most applications, which simplifies the design of the laser. Thermal lensing is also mild in the laser, so the output divergence changes little with increased temperature. The thermal limit of the laser is dictated by the amount of heat the crystal can withstand without damage. A 20-mm-long crystal had a maximum output power of 20 W.
The researchers also optimized the fiber. A quartz fiber, free of water, with a core diameter of 440 µm was used with a thin coating of tin that allowed a bending radius of 4 cm. A complete experimental system of pump laser, fiber, and converter laser demonstrated a total conversion efficiency of 23%—almost three times the overall efficiency of Er:YAG laser systems and considerably less expensive, in addition to offering the user a more flexible fiber.
One of the technical difficulties in developing this laser design is making good mirrors for the converter laser. The coatings are difficult to produce because each mirror must have a high transmission at one wavelength and high reflection at the other. The group intends, therefore, to directly coat the ends of the laser crystal, which should simplify assembling the device.
REFERENCES
1. R. Steiner et al., Proc. SPIE 3199, 215 (1998).
2. B. Antipenko and A. Egorov, Proc. SPIE 2041, 429 (1994).