Infrared Microscopy for Examination of Structure in Spray-Dried Granules and Compacts Keizo Uematsu,* ,† Nozomu Uchida, Zenji Kato, and Satoshi Tanaka Department of Chemistry, Nagaoka University of Technology, Nagaoka, Niigata 940–2188, Japan Tadashi Hotta and Makio Naito* Japan Fine Ceramics Center, Mutsuno, Nagoya 456–8587, Japan Novel infrared microscopy has been developed to improve the liquid-immersion method, which is a technique that has been developed by the authors to study the packing structures of powder particles in ceramic green bodies. This paper demon- strates the high potential of the novel technique by presenting clear structures of Si 3 N 4 powder granules and their compacts. I. Introduction I DEALLY, the green-body structure in the processing of ceramics must be fully understood, because it governs the microstructure (and, thus, the properties) of ceramics. 1,2 In reality, however, this understanding has been extremely difficult, mainly because of the limited capability of current characterization tools. The authors have developed a process called liquid-immersion microscopy (LIM) to fulfil this need. 3,4 This procedure uses the application of an immersion liquid to make a green body transparent. This transparency is achieved via the suppression of reflected light at the particle/liquid interface through a good matching of the refractive indexes for relevant phases. The green-body structure subsequently is examined via optical microscopy in the transmis- sion mode. This observation mode is very important for evaluating rare features such as exceptionally detrimental defects in the green body. One of these features could be directly responsible for the formation of a fracture origin in the sintered body. The high capability of LIM has been demonstrated repeatedly in many papers. 5–7 However, LIM techniques must be improved. There have been two limitations: (i) the necessity of thin specimens for examina- tion, and (ii) the availability of an immersion liquid. The maximum thickness of a specimen for examination has been 0.5 mm in the optimum case; 8 generally, the value has been less than a few tenths of a millimeter. The second limitation is much more important than the former. Immersion liquids are not available for many important ceramic materials, because the refractive indexes of such ceramics often are too high for a liquid to match at room temperature. For example, an earlier work showed that the liquid necessary to match a Si 3 N 4 green compact with a refractive index (n) of 2.05 was highly toxic and unstable. 5 These limitations should be removed using a light of long wavelength  (e.g., infrared (IR) light) for observation. In fact, the first limitation can be successfully removed when IR microscopy is used. Another study has shown that the maximum specimen thickness for examination increases to 2 mm in an alumina green body. 9 The same microscopy method also should remove the necessity of using a toxic immersion liquid for observation. This paper reports the successful observation of the internal structure in a Si 3 N 4 green body, using a nontoxic liquid. II. Experimental Procedure The raw materials used in this study were similar to those used in a previous study. 10 Namely, Si 3 N 4 (SN-E10, UBE Industries, Yamaguchi, Japan), 5 mass% of alumina (Al 2 O 3 ) (AKP-30, Sumitomo Chemical, Tokyo, Japan), and 5 mass% of yttria (Y 2 O 3 ) (NRN, Daiichi Kigenso Kagaku Kogyo, Osaka, Japan) were used as sintering aids; a binder and a dispersant also were used. The components were made into a slurry via ball milling for 24 h, and the granules were prepared through the spray-drying process, as presented in the previous study. 10 A compact was formed by die pressing at a pressure of 19.6 MPa and heated at a temperature of 1300°C for 2 h, for binder removal and also to increase its strength up to that needed for handling. A thin section for observation (0.15 mm) was prepared by hand grinding a small piece of the specimen with fine sandpaper (No. 2400). The immersion liquid used in this study was a saturated solution of sulfur in methylene iodide (n = 1.79). A few drops of the immersion liquid were placed at one side of the specimen, to be absorbed without trapped air. The commercial IR microscope that was used for the obser- vation (Model BX50IR, Olympus, Tokyo, Japan) was equipped with an IR camera (Model C2741, Hamamatsu Electronics, Hamamatsu, Japan). The maximum operating wavelength of the camera was 1.8 m. A filter was used to exclude light with  1.3 m. The structure of the specimen also was examined via scanning electron microscopy (SEM) (Model JSM4330LV, JEOL, Tokyo, Japan). Gold film was sputtered onto the sample surface. III. Results and Discussion Figure 1 shows an SEM micrograph of the Si 3 N 4 powder granules. Other than a few granules, these particles generally are almost spherical; some have a dimple. This shape often is observed in granules that have been made from a well-dispersed slurry. 5 However, the micrograph does not show whether or not all the granules contain a dimple. Figure 2 shows an IR photomicrograph of Si 3 N 4 powder granules taken with the nontoxic immersion liquid (n = 1.79). This method gave much-more-detailed and accurate structural informa- tion than did SEM analysis. Detailed internal structure is clearly visible for all granules. The granules are irregularly shaped and J. J. Lannutti—contributing editor Manuscript No. 188611. Received April 21, 2000; approved September 15, 2000. Supported by NEDO, through the International Joint Research Grant Program, and by a Grain-in-Aid for Scientific Research from the Ministry of Education, Culture and Sport. *Member, American Ceramic Society. † Author to whom correspondence should be addressed. 254 journal J. Am. Ceram. Soc., 84 [1] 254 –56 (2001)