It is important to identify the detailed structure of large defects in ceramics to improve ceramic processing, because large defect govern the strength of ceramics . Clearly, their formation mechanism must be explicitly identified to produce better ceramics. Full characterization of defects, especially of large ones, should be the starting point. However, this has been extremely difficult except for that directly responsible for specific fracture, i.e., the fracture origin. Other defects are hidden under the surface and can be detected only when they are exposed to the external surface accidentally. Even in that case, the apparent characteristics of the defect may be very different from the reality; very often only its small part is exposed to the surface. The progress of processing of ceramics has been retarded significantly by the difficulty in characterization of large defects. Clearly, a new method is needed to characterize large defects in ceramics.
Identification of the location of large defects is the most difficult part in the characterization. Optical microscopy provides a promising way to identify and study large defects in ceramics [2-9]. In that method, a thin specimen (typically <200 µm) is cut from the ceramics and both faces are polished to make a transparent specimen. Subsequently, the bulk of specimen is examined with a transmission optical microscope to find and examine large defects. The method shows very high sensitivity in identifying rare features in the microstructure. In the present method, the optical microscopy is used in combination with scanning electron microscopy (SEM) to examine the detailed structure of large defect. In the SEM examination, a large defect is exposed to the external surface by polishing.
In this paper, details of the procedure and some examples of large defects in microstructure found in alumina ceramics, for which the fracture pattern is unclear and even the examination of fracture origin has been difficult with the conventional fractography. The method is applicable for a variety of ceramics, since many of them are essentially transparent.
Commercial alumina granules (TM-DS-6, Taimei Chemicals Co. Ltd., Japan) were used as starting materials in the present study. Green compacts were prepared by die pressing at 20 MPa, followed by cold isostatic pressing at 200 MPa. Green compacts were sintered at 1350ºC for 2 h. Sintered bodies have 99.5% relative density and the size about 20x55x4 mm3. For the examination of large defects, the sintered body was cut and ground with a surface grinding machine. A diamond cutter with grit size of 200 mesh (average grain size: 80 µm) and a grinding wheel of 400 mesh (average grain size: 40 µm) were used to prepare thin specimens (about 0.5 mm thick). Both surfaces of specimens were carefully polished to a thickness of 150 µm with diamond slurry using a precision polishing machine (EJ-380IN, Engis Japan Co., Japan). A transmission optical microscope (BX-51, Olympus Co., Japan) was used for the bulk examination of the specimen. The depth of a given large defect was determined by multiplying the traveling distance of specimen stage of the microscope for focusing them by the refractive index n of material (n=1.76 for alumina). Then, the specimen was polished down to the location of the target defect to expose it on the surface. The surface was coated with sputtered gold film and the defects were characterized in detail by SEM. Three dimensional structure of the large defect was characterized by successive SEM examinations for the same defect after every 10 µm polishing.
Figure 1 shows the SEM micrograph of the alumina granules used in this study. Each granule contains a dimple, suggesting that the granule be prepared with a well- dispersed slurry and be fairly hard.
Figure 1. Scanning electron micrograph of alumina granules.
Figure 2 shows micrographs taken at the same location with the reflection and the transmission modes. A surprising difference is noted in these micrographs for the dark features, which correspond to the defects. In the transmission mode, a few very large defects are noted, together with many small ones. This shows the true structure of this ceramics. The transmission mode of observation reveals all defects located at various depths in the bulk of this specimen. In the reflection mode, the size and number of defects appears much smaller. Without the above transmission micrograph, one may conclude that this is a good ceramic, as has been often the case in this kind of study.
Figure 2. Internal structure of sintered body observed with reflection mode and transmission mode. (In same location).
Figure 3 shows optical and SEM micrographs for the same large defect. The photomicrograph of large defect was taken in the bulk of ceramics. It shows a defect having a shape somewhat similar to a dimple of granule. The SEM micrograph was taken after the specimen was polished to expose the defect on the polished surface. The structure is very different from what is imagined from the photomicrograph. It consists of cracks rather than simple pores. The difference of the shape observed with these characterization methods clearly reflects the characteristics of transmission and surface analysis. The transmission mode of observation shows the shadow of the defect.
Figure 3. A large defect in sintered body observed with optical microscopy and scanning electron microscopy.
Figure 4 shows another example of large defect. This defect appeared to have near spherical shape in the photomicrograph. The SEM micrograph shows again that this defect also has a complicated shape and that the structure is crack-like rather than a simple pore. The image appears little fuzzy in the photomicrograph, since the defect was located at a deep position.
Figure 4. Another example of large defect in sintered body observed with optical microscopy and scanning electron microscopy.
Figure 5 shows successive images for the defect shown in Figure 3 taken after every 10 µm polishing. The structure is very interesting. It appears in Figure 5 (a) that sharp cracks penetrate deep into the bulk. However this is not true. These cracks mostly vanished and porous region is noted at 10 µm below. The location and morphology of porous region changed continuously with depth. Part of them remained even 50 µm below. It indicates directly that this large defect has volumetric shape at least 70 µm wide and over 50 µm deep.
(a) 0 µm, (b) 10 µm, (c) 20 µm, (d) 30 µm, (e) 40 µm, (f) 50 µm
Figure 5. 3D images of the large defect shown in Figure 3.
We succeed to develop the three dimensional characterization method for large defect in ceramics. Examination of many large defects showed that a majority of large defect appeared to have round shape in optical micrographs. Detailed SEM examination, however, showed that many of them have crack like shape rather than simple pore.
Large defect noted in the sintered body were clearly formed from the inhomogeneous packing structure of granules and/or powder particles in the starting green compact. Preservation and development of pores during sintering have been discussed frequently. Lange reported that green compact consist of small amount of dense region such as aggregate and loosely packed region. In dense region, densification occurs rapidly than in loosely packed region, then void space generate between them . Zheng et al.  and Shinohara et al.  reported that pore coalescence is considered responsible for the pore stability during sintering. Uematsu et al evaluated directly for morphological change of large processing pore in densification process . They reported that large pores in green compact persist and grew during sintering. In the final stage of densification, there are thermodynamic and kinetic arguments. Kingery and Francois reported that pore grew if the pore coordination number is larger than certain critical value which depends on the dihedral angle . Zhao and Harmer examined densification of large pore in alumina . They found that large pores maintained their size or slightly grew during sintering.
Clearly, the present method is a powerful tool for the critical examination of ceramics processing. The micrographs show that the large defects originate from the dimple structure of granules or the non-uniform packing of granules. In this particular system, an intensive research should be needed in slurry characteristics for preparing soft granules without a dimple. Preparation of granules with better flowing should also be examined.
We succeeded to fully characterize large defects in sintered alumina with a new method. Location of large defects was determined from transmission mode photomicrograph, and then they were fully characterized with SEM following polishing. Inhomogeneous packing structure of the granules in a green compact is clearly responsible for the formation of large defects in ceramics.
This study was supported by the 21st century of center of excellence (COE) program in Nagaoka University of Technology.
References and Notes
1. 1. F. F. Lange, “Powder Processing Science and Technology for Increased Reliability”, J. Am. Ceram. Soc., 72 (1989)3-15
2. 2. N. Shinohara, M. Okumiya, T. Hotta, K. Nakahira, M. Naito and K. Uematsu “Morphological Changes in Process- Related Large Pores of Granular Compacted and Sintered Alumina,” J. Am. Ceram. Soc., 83 (2000) 1633-1640
3. 3. K. Uematsu, J.-Y. Kim, Z. Kato, N. Uchida and K. Saito “Direct Observation Method for Internal Structure of Ceramic Green Body -Alumina Green Body as an Example-” J. Ceram. Soc. Japan, 98 (1990)515-516
4. 4. K. Uematsu, M. Miyashita, J.-Y. Kim and N. Uchida, ”Direct Study of the Behavior of Flaw-Forming Defect in Sintering”, J. Am. Ceram. Soc., 75 (1992) 1016-1018
5. 5. Y. Zhang, N. Uchida, K. Uematsu, T. Hotta, K. Nakahira and M.Naito, “A Direct Examination of the Fracture Strength-Defect Size Relationship in Alumina Ceramics”, Key Eng. Mater. (1999) 269-274
6. 6. Y. Zhang, M. Inoue, N. Uchida and K. Uematsu, “Characterization of Processing Pores and Their Relevance to the Strength in Alumina Ceramics”, J. Mater. Res., 14 (1999) 3370-3374
7. 7. J.-Y. Kim, M. Miyashita, N. Uchida and Keizo Uematsu, “Change of Internal/Pore Structure in Alumina Green Body during Initial Sintering” J. Mater. Sci. 27, (1992) 6609-6614
8. 8. Y. Takao, T. Hotta, K. Nakahira, M. Naito, N. Shinohara, M. Okumiya and K. Uematsu, “Processing Defects and Their Relevance to Strength in Alumina Ceramics Made by Slip Casting” J. Eur. Ceram. Soc., 20 (2000) 389-395
9. 9. N. Shinohara, M. Okumiya, T. Hotta, K. Nakahira, M. Naito and K. Uematsu, “Variation of the Microstructure and Fracture Strength of Cold Isostatically Pressed Alumina Ceramics with the Alteration of Dewaxing Procedures,” J. Eur. Ceram. Soc., 20 (2000) 843-849
10. 10. F. F. Lange, “Sinterability of Agglomerated Powders” J. Am. Ceram. Soc., 67 (1984) 83-89
11. 11.J. Zheng and J. S. Reed, “The Different Roles of Forming and Sintering on Densification of Powder Compacts” Am. Ceram. Soc. Bull., 71 (1992) 1410-1416
12. 12. W. D. Kingery and B. Francois, “The Sintering of Crystalline Oxides, I. Interactions between Grain Boundaries and Pores”, in Sintering and Related Phenomena, Gordon Breach, New York, (1967) pp.471-496
13. 13. J. Zhao and M. P. Harmer, “Effect of Pore Distribution on Microstructure Development: Ⅱ, First- and Second- Generation Pores,” J. Am. Ceram. Soc., 75 (1988) 530-539