Ytterbium-Phosphate Fiber Laser Reaches New Power

Breck Hitz

Conventional silica-fiber lasers are encountering several roadblocks — stimulated Brillouin scattering and photodarkening are among the most serious — as investigators push their single-mode power upward to the kilowatt level. Phosphate fibers may provide a detour around both issues, and researchers at Stanford University in California and at NP Photonics Inc. in Tucson, Ariz., recently demonstrated what they believe is the first cladding-pumped, ytterbium-phosphate fiber laser, which generates what they also believe is the highest power reported from any phosphate fiber laser.

Figure 1. The researchers measured the absorption and emission spectra in a 400-μm-thick sample of phosphate glass whose chemical composition was identical to the core’s. An off-center core and a small airhole diametrically opposite the core broke the inner cladding’s circular symmetry (inset). Reprinted with permission of Optics Letters.

One approach to limiting the stimulated Brillouin scattering in any fiber laser is to shorten the fiber, thereby increasing the threshold for the nonlinear effect. But shortening the fiber reduces the number of lasing ions available, unless the doping level is increased to compensate for the shorter length. The catch is that the rare-earth dopants have limited solubility in silica, and maximum doping levels are reached relatively quickly.

The solubility of rare-earths in phosphate fibers, on the other hand, is an order of magnitude greater than in silica fibers, leading one to hope that a performance improvement by a factor of 10 might be possible in these fibers. Another hopeful possibility is that photodarkening might be a less-detrimental effect in phosphate fibers.

Figure 2. A dichroic mirror separated the fiber laser’s output power from the incoming pump power.

Pursuing these possibilities, the researchers fabricated dual-clad phosphate fibers with 12 percent ytterbium doping by weight (1.42 × 10²¹ Yb³⁺/cm³) in a 10-μm-diameter circular core, surrounded by a 240-μm-diameter inner cladding. The 10-μm core was single-mode for wavelengths longer than 914 nm. To break the circular symmetry of the inner cladding and maximize the spatial overlap between pump modes and the core, the researchers offset the core from the fiber center by 15 μm and placed a small airhole, also offset by 15 μm, on the opposite side of the center (Figure 1).

They placed an 84.6-cm length of the fiber in a resonator formed by a butt-coupled maximum reflector mirror at one end and the Fresnel reflection from the fiber facet at the other. A 940-nm diode laser launched up to 80 W of pump power into the fiber through a dichroic mirror that directed the laser’s output to a power meter (Figure 2). The laser generated a maximum of 19.6 W in single mode, in excellent agreement with an advanced simulation code developed at Stanford (Figure 3).

Figure 3. The experimental data (squares) fit the theoretical prediction (solid line) very well.

Except for the ends of the fiber, which were held in aluminum mounts, the fiber lay uncooled on the laboratory table. The scientists observed a decrease in output power after the laser had operated for about 30 s, indicating thermal degradation.

They estimate that, when the laser was operating at maximum output, about 60 of the 80 W of launched pump power was absorbed in the fiber, half of that resulting from simple propagation loss. They calculate that a reduction in the passive loss, from the current 3 dB/m to 0.3 dB/m, would double the slope efficiency of the laser. They also believe that shifting the pump wavelength from 940 to 975 nm, where ytterbium absorption is stronger, would significantly increase the laser’s performance.

Optics Letters, Nov. 15, 2006, pp. 3255-3257.