New Crystalline Laser Generates 1 Watt at an Eye-Safe 1.5 μm
Er,Yb:YAl3(BO3)4 laser is several times more powerful than previous lasers.
Because the human cornea does not transmit wavelengths longer than 1.5 μm, lasers in this spectral region are less dangerous than shorter-wavelength lasers and thus are designated “eyesafe.” Such lasers are desirable in any application where there is a possibility that the beam could encounter human eyeballs; for example, aerial navigation, free-space telecommunications, environmental sensing and medicine. The rare-earth element erbium has a good laser transition at 1.5 μm but is a poor absorber at the ∼1-μm wavelength of common pump diodes. For that reason, most eye-safe lasers today are co-doped with both erbium and ytterbium. The ytterbium ion absorbs efficiently at the convenient laser diode wavelength of ∼975 nm and transfers energy to the erbium ion, where a population inversion occurs (Figure 1).
Figure 1. In an erbium-ytterbium co-doped laser, an ytterbium ion absorbs a pump photon and transfers the energy to an erbium ion. The erbium ion subsequently can be stimulated to emit a laser photon on the 4I13/2 - 4I15/2 transition.
As a practical matter, most eye-safe lasers use a glass host for the ytterbium and erbium ions, but the low thermal conductivity of glass inhibits its continuous-wave operation and limits average power to a few hundred milliwatts at most. Higher powers induce severe thermal focusing and, very quickly, catastrophic physical damage to the glass itself. Crystal hosts have better thermal conductivity, and although they can produce a continuous wave output easily, experiments to date have failed to produce a crystalline Er,Yb laser capable of more than 250 mW of power on average. Now, however, scientists at Belarus National Technical University in Minsk and at Moscow State University have demonstrated an Er,Yb:YAl3(BO3)4 laser generating 1 W of output power at 1555 nm.
Studies done previously of this laser crystal at other institutions have produced initial spectroscopic data and even quasi-continuous wave lasing, but the current work included a careful spectroscopic analysis of the material as well as the high-power lasing result. Spectroscopically, the scientists in Belarus and in Russia found that the energy transfer efficiency from ytterbium to erbium was ~88 percent and that the spontaneous lifetime of the 4I11/2 level in erbium was short — 80 ns — leading to efficient population of the upper laser level.
Figure 2. A pair of lenses (represented by double-ended arrows here) focused the pump light to a 110-μm spot at the end mirror (M1), where it closely matched the waist of the laser mode. (LD = laser diode.) Reprinted with permission of Optics Letters.
For their lasing experiments, the scientists used a folded resonator with two flat end mirrors and a folding mirror with a 100-mm radius of curvature (Figure 2). They pumped longitudinally a homegrown, 1.5-mm-long slab of Er,Yb:YAl3(BO3)4 with a 7-W, fiber-coupled diode laser at 976 nm. The laser crystal was doped with 11 atomic percent ytterbium and 1.5 atomic percent erbium, and a pair of 80-mm lenses focused the pump light to a 110-μm spot to match the laser mode inside the crystal.
Figure 3. With a 2 percent output coupler, both the slope efficiency and the wavelength shifted when the absorbed pump power reached about 2.5 W. (Toc = output coupler transmission, Pabs = absorbed pump power, Pout = output power.) Reprinted with permission of Optics Letters.
The input-output data differed with various output couplers (Figure 3). With a 2 percent coupler, the wavelength switched from 1602 to 1555 nm as the absorbed power reached 2.5 W and remained at that value as the absorbed power increased to 4 W and the output reached 1 W. The reason for the wavelength switch — and for the wavelengths generated with different output couplers — is not clear, but it probably is not particularly surprising, either. The scientists point out that intracavity thermal and other passive losses vary with output coupling and circulating power, and with several peaks in the observed stimulated-emission cross section, it is easy to imagine that the greatest net gain may shift with operating conditions.
Optics Letters, Nov. 15, 2007, pp. 3233-3235.
MORE FROM PHOTONICS MEDIA