Processing and Characterization of Polycrystalline YAG (Yttrium Aluminum Garnet) Core-Clad Fibers

New techniques could make polycrystalline YAG suitable for use in fiber lasers.

YAG-based fiber lasers could offer efficient operation at power levels beyond those achievable in current state-of-the-art silica-based fiber lasers if losses can be minimized. To address this, researchers have investigated creating both single-crystal and polycrystalline YAG fibers. Among the cases reported is the preparation of single-crystal YAG fibers using laser heated pedestal growth (LHPG), which resulted in fiber diameters of 400 μm and optical losses around 1–2 dB/m in the 1–3 μm wavelength range. Single-crystal YAG fibers with diameters of ~ 30 μm have even been reported.

SEM micrographs of cross-sectioned YAG fiber (a) and cross-sectioned foil (b) showing second-phase inclusions. EDS spectra (c) collected via TEM on matrix YAG and second-phase inclusions numbered in (b). SEM micrograph (d) of a polished section of YAG fiber without second-phase inclusions.

Although single-crystal YAG fibers are now available, robust cladding pro - cesses have yet to be developed for them. While a number of cladding methods were investigated, they appear to be in their initial stages and lack the maturity required for the envisioned laser application. Indeed, glass can be used as a cladding on single-crystal fibers, but its thermal conductivity is about ten times lower than that of crystalline YAG. Therefore, it can only be used for characterization purposes as it negates the motivation to use YAG as a fiber media and is therefore not practical for actual applications.

Applying ceramic coatings to fully dense fiber is also problematic due to constrained sintering conditions which result in cracks forming in the cladding. On the other hand, polycrystalline YAG fibers can be prepared with conventional ceramic processing, which involves a variety of processing steps including the formation of so-called “green fiber”, the creation of binder-removed fiber, and finally fully dense fiber.

Optical propagation losses were measured by injecting a 1480-nm fiber-coupled laser into polycrystalline YAG fibers that were cladded with 3 μm of Schott SF57 glass. The loss of these fibers, measured to be ≥ 50 1/m, needs to be significantly lowered to facilitate practical applications. It was subsequently discovered that these fibers all contained second-phase inclusions which likely lead to excess scattering loss. It was determined the cladding thickness was too thin to optimally confine the light within the fibers, and the refractive index difference between the YAG core and the glass cladding was too large at the characterization wavelength used. As a result of these findings, work focused on removing the second-phase inclusions and on developing a process to apply an undoped YAG cladding to a doped poly-crystalline YAG fiber in order to optimize the refractive index difference between core and cladding.

Yb-doped YAG powders were ball-milled and classified to remove agglomerates and contamination from the milling media. Binder and plasticizer were mixed with the Yb-doped YAG powder and the mixture was extruded through a 50-μm custom-made nozzle at pressures 3000–5000 psi. The fibers were then dried at room temperature for ≥8 hours before any future process steps were taken.

An un-doped YAG slurry was subsequently prepared and the Yb-doped fibers were dip coated in this slurry multiple times to form the YAG cladding. After each dip coating, the coated fiber was dried at 120°C. Organics in the cladded green fiber were removed during slow ramping up to 600°C and the sintering was carried out under a variety of conditions.

To further investigate the composition of these fibers, they were mounted, polished, and coated with carbon as a conductive layer for scanning electron microscopy (SEM, Quanta ESEM, FEI). Then the back-scattered electron mode of the SEM was used to analyze fibers for the presence of second-phase inclusions. In addition, focused ion beam (FIB, Nova, FEI) was used to prepare foils for energy dispersive x-ray spectroscopy (EDS) analysis via TEM (CM 200, Philips) to perform chemical analysis on the second-phase inclusions and YAG matrix. Optical scattering losses of these fibers were also measured.

While earlier fibers guided light, the best loss coefficient was measured to be ~50 1/m. During the analysis of these fibers, however, it was found that they contained second-phase inclusions [as shown in Figure (a)]. This implies the loss of the fibers would improve if the second-phase inclusions could be removed. The EDS spectra shown in Figure (c) make it clear that the second-phase inclusions are Si rich and Al deficient. While it is still unclear what causes the formation of the second-phase inclusions, they are believed to increase the scattering losses of these fibers. After a number of experiments where the sintering conditions were changed, conditions were found that allowed the creation of polycrystalline YAG fibers without inclusions; the microstructure of such a fiber is shown in Figure (d).

This work was done by Nicholas Usechak, Hyun Jun Kim, Santeri Potticary, Matthew O'Malley and GE Aviation’s Geoff Fair for the Air Force Research Laboratory. AFRL-0246



This Brief includes a Technical Support Package (TSP).
Document cover
PROCESSING AND CHARACTERIZATION OF POLYCRYSTALLINE YAG (YTTRIUM ALUMINUM GARNET) CORE-CLAD FIBERS -POSTPRINT

(reference AFRL-0246) is currently available for download from the TSP library.

Don't have an account? Sign up here.