Solid-State Reactive Sintering of Transparent Polycrystalline Nd:YAG Ceramics Sang-Ho Lee,* ,w Sujarinee Kochawattana,* and Gary L. Messing* , ** Department of Materials Science and Engineering and Materials Research Institute, The Pennsylvania State University, University Park, Pennsylvania 16802 John Q. Dumm* II–VI Incorporated, Saxonburg, Pennsylvania 16056 Gregory Quarles and Vida Castillo VLOC Inc., Subsidiary of II–VI Inc., New Port Richey, Florida 34655 Transparent polycrystalline Nd:YAG ceramics were fabricated by solid-state reactive sintering a mixture of commercial Al 2 O 3 , Y 2 O 3 , and Nd 2 O 3 powders. The powders were mixed in meth- anol and doped with 0.5 wt% tetraethoxysilane (TEOS), dried, and pressed. Pressed samples were sintered from 17001 to 18501C in vacuum without calcination. Transparent fully dense samples with average grain sizes of B50 lm were obtained at 18001C for all Nd 2 O 3 levels studied (0, 1, 3, and 5 at.%). The sintering temperature was little affected by Nd concentration, but SiO 2 doping lowered the sintering temperature by B1001C. Abnormal grain growth was frequently observed in samples sin- tered at 18501C. The Nd concentration was determined by energy-dispersive spectroscopy to be uniform throughout the samples. The in-line transmittance was 480% in the 350–900 nm range regardless of the Nd concentration. The best 1 at.% Nd:YAG ceramics (2 mm thick) achieved 84% transmittance, which is equivalent to 0.9 at.% Nd:YAG single crystals grown by the Czochralski method. I. Introduction S INGLE-crystal Nd:YAG, which is commercially produced by the Czochralski method, has a number of limitations. The Nd concentration that affects the performance in laser applica- tions, is limited to 0.2–1.4 at.% as a result of the segregation distribution coefficient. 1 The Nd concentration in the grown crystal varies in the axial direction, which results in a highly strained core region and other symmetrical inhomogeneities. 2 These defects result in optical birefringence and wave front dis- tortion. To circumvent these defects, parts for laser applications are drilled from the crystal to avoid the core area and thus only part of the single crystal is used. Greskovich et al. 3,4 were the first to produce a ceramic laser host using 89Y 2 O 3 –10ThO 2 –1Nd 2 O 3 (mole%). But, due to the excessive scattering by residual pores less than 0.5 mm in size, the laser efficiency was unsatisfactory. DeWith and VanDijk 5 and Sekita et al. 6 reported efforts to fabricate transparent YAG ce- ramics for optical applications, but their ceramics were translu- cent and of insufficient quality to achieve the optical properties of single-crystal Nd:YAG. Ikesue et al. 7,8 first demonstrated the possibility of fabricating transparent Nd:YAG ceramics of suf- ficient quality for solid-state lasers with reasonable efficiency. More recently, a number of studies have shown that transparent polycrystalline Nd:YAG is equivalent or better than single crys- tals grown by the Czochralski method. 9–11 In the polycrystalline Nd:YAG ceramics, the doping concentration can be increased to as much as 9 at.%, and a homogeneous distribution can be achieved throughout the specimen. In addition, the specimen size can be significantly increased while the largest boule grown by the Czochralski method is B100 mm in diameter. Transpar- ent ceramics can be produced by starting with phase-pure YAG powders, or by the reactive sintering approach in which a mix- ture of metal oxides is sintered. The purpose of this paper is to report how various aspects of the process affect fabrication of transparent polycrystalline Nd:YAG (0–5 at.% Nd) ceramics by solid-state reactive sinte- ring. This general process is similar to that reported by Ikesue et al. 7,8 but the purpose of this study was to make the process as simple as possible. In this study, spray drying was not used, and the starting powders are commercially available Al 2 O 3 ,Y 2 O 3 , and Nd 2 O 3 powders without pretreatment. One of the goals of the research program was to develop a low-cost process that can be readily adapted to a manufacturing environment. II. Experimental Procedure Submicron a-Al 2 O 3 powder (AKP50, 499.99%, 0.1–0.3 mm, Sumitomo, Japan) and coarse Y 2 O 3 (499.999%, 2–4 mm, NYC Co., Tokyo, Japan), and Nd 2 O 3 (499.99%, 1–2 mm, NYC Co.) powders were used. Tetraethoxysilane (TEOS, 99.9999%, Alfa, Ward Hill, MA) of 0.5 wt% (0.14 wt% of SiO 2 ) was used as a sintering aid in all samples reported in this paper. Powder batch- es of 0, 1, 3, and 5 at.% Nd:YAG were ball milled in a poly- ethylene bottle for 16 h with 5 mm diameter Al 2 O 3 balls in methanol. After ball milling, the slurry was dried, uniaxially pressed into 12.7 mm diameter pellets at 20 MPa, and cold iso- statically pressed at 200 MPa. The initial densification behavior of the pellet was examined with a thermal mechanical analyzer (TMA) (Thermo-Mechanical Analyzer, TMA 50, Shimadzu, Kyoto, Japan) by heating to 14001C at 51C/min. The phase transformation, which occurs in the range of 11001–15001C, was observed by X-ray diffraction (XRD; PAD V, Scintag, Cuper- tino, CA). Sintering was conducted at 17001–18501C for up to 16 h in a tungsten mesh-heated vacuum furnace (M60, Centorr Vacuum 1945 J ournal J. Am. Ceram. Soc., 89 [6] 1945–1950 (2006) DOI: 10.1111/j.1551-2916.2006.01051.x r 2006 The American Ceramic Society D. Green—contributing editor Presented at the 9th International Ceramic Processing Science Symposium, Coral Springs, FL, Jan. 8–11, 2006. This work was supported by VLOC Inc. through funds received from VLOC’s prime contract #N66001-00-C-6008. *Member, American Ceramic Society. w Author to whom correspondence should be addressed. e-mail: sangho@psu.edu **Fellow, American Ceramic Society. Manuscript No. 21385. Received January 15, 2006; approved February 20, 2006.