2718 Journal of the American Ceramic Society--Bitterlich and Heinrich No.10 Ta 300p Fig 9. Polished stack laminated with composition CerO showing defects in the interlayer(scanning electron microscopy, SEM) On laminating the tapes with a pre-ceramic polymer, the process is greatly improved compared with the precursor-free suspension. Therefore, a good joining of the tapes is obtained However, the residues of the precursor are still present after Interna pyrolysis. The amount of sintering additives in the interlayer is lower than in the tape material because the pastes consist of the precursor and a fixed mixture of silicon nitride powder and sin- tering additives. Carbon is present as one of the precursor prod ucts, which reacts with the glassy phase by decreasing its amount and developing gases. Furthermore, the crystal size of the silicon nitride from the precursor is different from that of the Fig. 7. Fracture surface of tapes laminated with Cer30 after pyrolysis silicon nitride powder .Thus, in the interlayer, the starting (scanning electron microscopy, SEM conditions for the liquid sintering process are not the same as in the material of the tapes, which will infuence the sintering be- havior and the development of the microstructure. Neverthe- interlayers. The hindered sintering of the tape material results in interlayers can hardly be distinguished from the tape material the residual porosity, which can be seen in the microstructure. by SEM at high magnification(Fig. 10(b)and(d). The joining Lamination with a paste without ceramic precursors ("Cer between interlayer and tape material seems to be excellent. Fur 0")has been carried out for the purpose of comparison. After thermore, the microstructure, i. e. grain size distribution, mor- pressureless lamination and evaporation of the solvent, the me- phology, and amount of grain-boundary phase, is qualitatively bonds between the tapes and the interlayer, which only consists on the development of the microstructure is weak, so that no of the dry ceramic powder. The poor joining is the origin for significant dif can be detected by SEM at high magnifi many defects present after sintering at the interfaces between the cations(Fig. 10(b) and(d). At lower magnifications(Fig. 10(a) interlayer and the tape material ( Fig 9). In this photo, the equal and (c), it can be seen that the tape material contains more distance of the larger pores indicate that during screen printing, pores than the interlayer-especially for high precursor contents some of the meshes were plugged up, indicating the worse (Fig. 10(c)). This difference in porosity and density is correlated processing behavior of the paste""Cer 0 to the density in the pyrolyzed status and the constrained sin- tering of the tape material and varies with the amount of pre- cursor in the laminating paste. (A Strength: The strength measurements show a signif- icant difference depending on the direction of testing, i. e in the directions parallel and perpendicular to the stacking direction Stacks made by compression, for example, have a characteristic strength of 638 MPa(parallel) and 490 MPa(perpendicular). respectively, and a Weibull modulus of 13(parallel) and 6(per- pendicular)(Fig. 11(a)). Such anisotropy in strength was also found for stack y pressureless lamination with pastes: the strength decrease from 617 to 427 MPa and so does the weibull modulus. from 12 to 5, when the tensile stress is applied perpendicular to the The observed anisotropy cannot be attributed to the influence Polished cross-section of a stack prepared under non-optimized of the pre-ceramic polymer because it is present in samples itions, showing cracks in the tape material; arrows indicate the po- by both processes. To prove the assumption that the st sition of the interlayers. degradation of about 25% might be attributed to the orientationinterlayers. The hindered sintering of the tape material results in the residual porosity, which can be seen in the microstructure. Lamination with a paste without ceramic precursors (‘‘Cer 0’’) has been carried out for the purpose of comparison. After pressureless lamination and evaporation of the solvent, the me￾chanical strength is quite low because of the lack of any joining bonds between the tapes and the interlayer, which only consists of the dry ceramic powder. The poor joining is the origin for many defects present after sintering at the interfaces between the interlayer and the tape material (Fig. 9). In this photo, the equal distance of the larger pores indicate that during screen printing, some of the meshes were plugged up, indicating the worse processing behavior of the paste ‘‘Cer 0’’. On laminating the tapes with a pre-ceramic polymer, the process is greatly improved compared with the precursor-free suspension. Therefore, a good joining of the tapes is obtained. However, the residues of the precursor are still present after pyrolysis. The amount of sintering additives in the interlayer is lower than in the tape material because the pastes consist of the precursor and a fixed mixture of silicon nitride powder and sin￾tering additives. Carbon is present as one of the precursor prod￾ucts, which reacts with the glassy phase by decreasing its amount and developing gases.19 Furthermore, the crystal size of the silicon nitride from the precursor is different from that of the silicon nitride powder.20 Thus, in the interlayer, the starting conditions for the liquid sintering process are not the same as in the material of the tapes, which will influence the sintering be￾havior and the development of the microstructure.21 Neverthe￾less, after sintering of such laminated tapes, the previous interlayers can hardly be distinguished from the tape material by SEM at high magnification (Fig. 10(b) and (d)). The joining between interlayer and tape material seems to be excellent. Fur￾thermore, the microstructure, i.e. grain size distribution, mor￾phology, and amount of grain-boundary phase, is qualitatively the same in both regions. Obviously, the effect of the precursor on the development of the microstructure is weak, so that no significant difference can be detected by SEM at high magnifi- cations (Fig. 10(b) and (d)). At lower magnifications (Fig. 10(a) and (c)), it can be seen that the tape material contains more pores than the interlayer—especially for high precursor contents (Fig. 10(c)). This difference in porosity and density is correlated to the density in the pyrolyzed status and the constrained sin￾tering of the tape material and varies with the amount of pre￾cursor in the laminating paste. (3) Properties (A) Strength: The strength measurements show a signif￾icant difference depending on the direction of testing, i.e. in the directions parallel and perpendicular to the stacking direction. Stacks made by compression, for example, have a characteristic strength of 638 MPa (parallel) and 490 MPa (perpendicular), respectively, and a Weibull modulus of 13 (parallel) and 6 (per￾pendicular) (Fig. 11(a)). Such anisotropy in strength was also found for stacks made by pressureless lamination with pastes: the strength decreases from 617 to 427 MPa and so does the Weibull modulus, from 12 to 5, when the tensile stress is applied perpendicular to the stacking direction instead of parallel (Fig. 11(b)). The observed anisotropy cannot be attributed to the influence of the pre-ceramic polymer because it is present in samples made by both processes. To prove the assumption that the strength degradation of about 25% might be attributed to the orientation Fig. 7. Fracture surface of tapes laminated with Cer30 after pyrolysis (scanning electron microscopy, SEM). Fig. 8. Polished cross-section of a stack prepared under non-optimized conditions, showing cracks in the tape material; arrows indicate the po￾sition of the interlayers. Fig. 9. Polished stack laminated with composition Cer0 (no precursor) showing defects in the interlayer (scanning electron microscopy, SEM). 2718 Journal of the American Ceramic Society—Bitterlich and Heinrich Vol. 88, No. 10