1. Introduction

Since the first demonstration of lasing from Dy²⁺:CaF₂ ceramics, much effort has been made to develop high power solid state lasers based on transparent ceramics [1–8] to replace melt grown single crystals which often suffer from drawbacks such as segregation of the dopant from the host, optical inhomogeneity caused by stress during crystal growth and high cost and low productivity due to high temperature processing. Polycrystalline ceramics are advantageous over single crystals in many ways. The process is simple, cost effective, and typically carried out at lower temperatures. More importantly, much higher doping concentrations in ceramics can be obtained without phase segregation as is often observed in single crystals [9]. Currently, rare earth doped yttrium aluminum garnet (YAG) such as Nd:YAG and Yb:YAG is the most extensively studied and widely used ceramics for high power laser material [2, 10]. However, YAG is not a suitable host material for high-power laser operation systems due to its relatively low thermal conductivity and high thermal expansion. The sesquioxides such as Sc₂O₃, Y₂O₃, and Lu₂O₃ are very promising host materials for high-power laser applications, mainly due to their high thermal conductivity and the high absorption and emission cross-sections of trivalent rare-earth ions in these materials [11]. Among them, Lu₂O₃ stands out as the best host materials especially for Yb doped high power ceramic laser systems. Since the lutetium and ytterbium ions have very similar ionic radii and bonding forces, the ytterbium ion can easily replace a lutetium ion upon doping without the overall thermal conductivity being affected. In fact, we have recently reported a record high lasing performance from highly Yb-doped Lu₂O₃ ceramics [7, 8, 12]. We demonstrated 74% slope efficiency from a 10% Yb doped Lu₂O₃ transparent ceramic. We also reported that the chemical, morphological, and phase quality of the powder is the key factor determining the quality and the performance of the laser oscillation from the resulting ceramics.

Since there is no commercially available high quality powder for transparent ceramic laser applications, various processes, including co-precipitation, combustion, flame spray, and sol-gel synthesis, have been proposed and used to produce high purity and fine sesquioxide nano-powders suitable for solid state laser applications [13–16]. The main objective of these studies was to obtain agglomeration-free and pure (chemically, morphologically, and phase) nano-powders to minimize absorption and scattering loss mainly caused by presence of transition metals, grain boundary phases, residual pores and secondary phases, that prohibit an efficient laser oscillation from the non-uniform or translucent ceramic laser gain medium. For example, the co-precipitation of the precursors using inorganic salts in a base condition and a subsequent calcination is the most convenient and cost effective technique, and is suitable for mass production of homogeneous powders. In our recent study on the synthesis of sesquioxide powders by the co-precipitation method, we employed a similar method to synthesize high purity Yb³⁺ doped Lu₂O₃ or Y₂O₃ nano-powder to fabricate highly transparent ceramics to demonstrate record high lasing efficiency [7, 8, 12, 17]. However, the powder produced by this wet process sometimes includes large (>100μm) and hard agglomerates, and attempts to fabricate optical quality ceramics using these powders often result in a transparent ceramics containing large grains and numerous structural and morphological defects. Agglomeration by an aqueous based co-precipitation method is believed to be caused by the strong intra- and/or inter-molecular hydrogen bonding between hydroxide or hydroxynitrate precursors and water molecules [7]. Small and extremely polar water molecules attract the precursors, causing them to pack close upon drying and to agglomerate during both the drying and calcination processes. Such agglomerated powders typically break into smaller particles by either wet or dry ball-milling. This method is very convenient and works very well; however, ball milling causes another unwanted issue where the milled particles now contain a small fraction of abraded balls as a contaminant. This will result as a defect in the ceramics and makes the resulting ceramics translucent or opaque. An alternate way to break up the particles with less chance of cross contamination is jet milling [18–20]. Jet milling differs from other milling procedures in that the material is not ground against a hard surface. Instead, a gaseous medium is introduced and the grinding takes place by colliding particles. The gas may convey the feed material at high velocity in opposing streams or it may move the material around the periphery of the grinding and classifying chamber. The high turbulence causes the particles and feed materials to collide and grind upon themselves by abrasion. Jet milling employs compressed air or gas to produce particles less than one micron. Precisely aligned jets create a vortex. Material is fed into this vortex along an engineered tangent circle and accelerates. High-speed rotation subjects the material to particle-on-particle impact, creating increasingly smaller particles. While centrifugal force drives large particles toward the perimeter, fine particles move toward the center where they exit through the vortex finder.

In this study, we employed a jet mill as a tool to break the agglomerates in the powders synthesized by wet chemistry methods to produce a uniform and high quality transparent ceramic. Figures 1(a) and 1(b) show the commercial jet mill system used for this work (2” Sturtevant Micronizer^®, Hanover, MA.) and the schematic diagram of how the jet mill works, respectively. In this paper, we report a convenient and efficient jet mill method to produce homogenous ceramic powders suitable to fabricate highly transparent ceramics for high power laser application.

 

Fig. 1 A photo of the 2” Sturtevant Micronizer, Hanover, MA, used in this study and a schematic diagram showing how the jet mill works. Photos were used in courtesy and with permission from Sturtevant Inc. http://www.sturtevantinc.com/products/micronizer.

Download Full Size | PDF

2. Experimental procedure

Lu₂O₃ powders doped with Yb³⁺ in various doping concentration were synthesized by the co-precipitation method. Details of powder synthesis have been published elsewhere [7]. In brief, highly pure Lu and Yb precursor crystals were first obtained by recrystallization to obtain a highly purified nitrate mixture. The mixed crystal was dissolved in de-ionized H₂O and was added dropwise slowly into a warm H₂O/ammonium hydroxide solution under vigorous stirring. A white precipitate started to form and was washed with de-ionized water and acetone. The wet precursor powder was dried and calcined to obtain Yb-doped Lu₂O₃ powder. For jet milling experiments, a commercial 2” Sturtevant Micronizer^® was used. Typically, 15~30g of as-produced powder was fed through a vibratory V-shape trough. Feed rate was varied from 0.25 lb/hr to 1.5 lb/hr. Grind pressure of 105 Psi and feed pressure of 80 Psi were used for all experiment. All experiments used a single 2” x 14” collection bag. The powder was pre-screened at 16 mesh to remove oversized material before jet mill. Table 1 summarizes the experimental conditions used for jet milling.

Table 1. Experimental parameters used for jet mill 10% Yb Lu₂O₃ powders

View Table | View all tables in this article

For ceramic fabrication, the jet milled powder was mixed with a sintering aid, placed in a graphite-foil lined graphite die, and hot pressed at 1500-1700°C for 2-6 hours at a pressure of 50 MPa. Samples were then hot isostatically pressed at 1600-1800°C in argon at 200 MPa for 2 hours and optically polished.

The purity of the powder and ceramic was characterized using glow discharge mass spectroscopy (GDMS). GMDS is an analytical technique capable of providing trace-level elemental quantification for a wide range of solid and powder materials. It has an excellent detection limit (ppm to ppb), accommodating conducting, semi-conducting and insulating samples. The median and mean particle size and distribution of the powder were measured by laser diffraction/scattering using a Horiba LA-950 system. Before laser diffraction/scattering was performed, the samples were sonicated until the point where the particle size distributions did not change with further sonication. The laser diffraction/scattering was then performed immediately. Powder surface area was measured using microporosimetry (BET method, Micromeritics ASAP 2020). Transmission measurements on polished ceramics were performed using UV–Vis and Fourier transform infrared (FTIR) spectrometers.

3. Results and discussion

Jet milling was chosen as the best route for potentially breaking up agglomerates from Yb doped Lu₂O₃ powder synthesized in our laboratory by the wet/aqueous method. In this process, the agglomerate break-up is enabled by forcing the powder to collide with itself at high velocity using high pressure gas. The powder underwent jet milling under various conditions and the milled powders were characterized and compared with as-produced powder prior to de-agglomerations via jet milling. Figure 2 shows the particle size distribution of the powders obtained before and after jet mill. Feed rate was varied from 0.25 lb/hr to 1.5 lb/hr and the agglomerate size of the resulting powder was measured. It is shown that jet mill is very effective to break the larger powder into a much smaller sizes. Median particle size decreased from 8.74 μm to 1.38 μm after 10 minutes jet milling at 0.25 lb/hr feed rate. The particle size continues to decrease as the feed rate increases and reached a minimum particle size of 1.06 μm at the feed rate 0.75 lb/hr. It shows that particle-on-particle impact reaches maximum at this feed ratio creating smaller particles. Higher feed ratio greater than 0.75 lb/hr was not effective to further break the particles into smaller sizes and rather showed a slight increase in the particle sizes which might be caused by re-aggregation of the fine particles. The specific surface area was measured and shown a similar trend as the particle sizes. It reached the maximum value of 6.39 g/m² at 0.75 lb/hr feed rate which is a good indication of smaller particle size. In Table 2, mean diameter, median diameter, and specific surface area of the particles milled at various feed rate are summarized.

 

Fig. 2 Particle size analysis results obtained from 10% Yb:Lu₂O₃ powder synthesized by wet chemistry jet milled at various feed rate.

Download Full Size | PDF

Table 2. Mean diameter, median diameter, and specific surface area of the jet milled powders obtained at various feed rate

View Table | View all tables in this article

The purity of the jet milled powder was also evaluated since one of the biggest concerns using jet milling is a potential contamination during the milling process caused by the abrasion and collision with the liner materials which will, in turn, deteriorate the quality of the transparent ceramics. A tungsten carbide liner and accessories were chosen and used in this study since it is the hardest and the most abrasion-resistant material available for this application. Table 3 summarizes the GDMS chemical analysis results obtained from the powders before and after jet milling. We collected the sacrificial powder as well as the as-produced and final jet milled powder and compared the changes in the chemical impurity. The sacrificial milling was carried out with the purpose of coating and conditioning the surface of the liner with the powders. As seen in Table 3, the powder from sacrificial milling contained various impurities such as Al, Co, Mn as well as W, mostly from the liner materials and from unknown sources (Al and Mn). This is a good indication that grinding takes place by colliding particles with each other as well as against the liner material during the jet milling process. Surprisingly, there was no noticeable increase in impurities picked up during the jet milling process from the powder jet milled after the initial sacrificial milling. Unexpected decrease in the content of some elements such as Si and Ca from final jet milled powder compared to the as-produced powder was also observed. The detailed chemical analysis of the powder is beyond the scope of this article and will be reported elsewhere. Figure 3 shows photos of optically polished ceramics after hot pressing/HIPing with powders that were (a) as-produced and (b) final jet milled after the sacrificial run, respectively. More homogeneous and uniform ceramic was obtained from the final jet milled powder compared with the one made from as-produced powder. The splotches and defects seen from the ceramics, fabricated using as-produced powder which are believed to be caused by the presence of the hard agglomerates in the powders obtained by wet synthesis, are not observed in the transparent ceramic fabricated using jet milled powder.

Table 3. Chemical analysis results obtained from the powders before and after jet milling. Elements greater than 1ppm by wt. are shown.

View Table | View all tables in this article

 

Fig. 3 Photos of transparent ceramics fabricated using as-produced (a) and jet milled Yb:Lu₂O₃ powders synthesized by a wet co-precipitation method.

Download Full Size | PDF

Figure 4 shows the optical transmission plot of the optically polished ceramics fabricated from the jet milled 10% Yb-doped Lu₂O₃ powder. A theoretical transmission curve (solid line) calculated using the refractive index measured by VUV-VASE and IR-VASE spectroscopic ellipsometers (J.A. Woollam Company) is also shown. As seen in Fig. 4, transmission of the Yb:Lu₂O₃ is very close to the theoretical limit which is a good indication of the excellent quality of the transparent ceramic. We believe this is due to high chemical purity of the synthesized powder as well as uniform morphology of the jet milled powder in which optical loss due to absorption or scattering caused by hard agglomerates is minimized in the final ceramics.

 

Fig. 4 Transmission plot of the optically polished ceramic fabricated from the synthesized 10%Yb:Lu₂O₃ powder after jet milling. Thickness of the corresponding ceramic is 2.1mm. A theoretical transmission of Lu₂O₃ (solid line) is also shown for comparison.

Download Full Size | PDF

4. Conclusion

This study demonstrated that jet milling is a very effective technique to break up large agglomerates without cross-contamination of co-precipitated/calcinded Yb:Lu₂O₃ ceramic powder. Homogeneous, uniform, and highly transparent ceramic was obtained using the final jet milled powder compared to the one made from as-produced powder where some defects and splotches were typically observed. The sesquioxide Yb³⁺:Lu₂O₃ ceramic successfully developed in this study is a very promising lasing material for high-power laser applications, mainly due to its high thermal conductivity, and the high absorption and emission cross-sections of trivalent rare-earth ions in these materials. The experimental transmission determined for the Yb³⁺:Lu₂O₃ ceramics obtained from the jet milled powder was very close to the theoretical limit, demonstrating the excellent quality of this transparent ceramic.