November 200 Communications of the American Ceramic Sociery 1963 Table L. Experimental Analysis of AlPO, after Three Different heat treatments Al(at%) P(at%) P: AIP Standard AlPO4(Aldrich, 99.99%) 21 1600°c(10h,air→1400°C/(100h) 21 r-treated AIPOA 1800C/(5 h), air-treated AlPO 21.56 24.30 Flexural strengths were measured by three-point bend testing with a screw-driven machine(Model 4502, Instron Corp, Canton, MA). The flexural strength and work of fracture values were determined after testing three to five samples. The specimen size mm, and the crosshead speed was 0. 1 mm/min 2 Theta II. Results and discussion e Kyoritsu KM 101 mullite had a particle size and specific vol% mullite and 50% AlPO4, after sintering at 1600oC for 10 h. So Fig. 1. X-ray diffraction profiles for a crushed pellet composed surface area of 0. 8 um and 8.5 m/g, respectively. The crystalline AlPO4 powder had particle sizes of 10 and 0.9 um, before and after 1-h attrition milling, respectively. The amorphous and crys- lline form of AlPO, powder had specific surface areas of 136 and 87 m-/g, respectively, after 1-h attrition milling. Because of the Mullite concern for decomposition of AlPO4 at high temperature, ICP analyses were done for three different kinds of AlPO,, and the results are summarized in Table I. The three samples were()a standard AlPO,(Aldrich, 99.99%),(ii)AlPO, which was heat- treated at 1600C/(10 h)in air, followed by an anneal at 1400oC/ (100 h)in air; and (iii) AlPO4 which was heat-treated at 1800C/(5 h)in air. If the reaction 2AIPO4-AL2O3+ P2Os occurs and P2O5 AIPOA evaporates at high temperature, the relative amounts of Al in the sample will increase. Table I compares the atom percent of Al and P for the three different AlPO specimens. This table indicates that there were almost no differences in the relative amounts of the two elements in the three different aluminum phosphates. One is led to the conclusion that aluminum phosphate is chemically stable at high temperatures. Table II summarizes the tape-casting formulations for the mullite and AlPOA. The solid loading was 25.1 vol% for both of 10 microns the powders In the case of the AlPO, tapes, 30% extra solvent was The results(Fig. 1) indicate that mullite and aluminum phosphate sintered at 1600C for 10 h are completely compatible with each other, without formation of any third phas Figure 2 is a scanning electron micrograph(SEM) of the 3A1,03 2SiO,/AlPO a laminated composite sintered at 1600C for porous, weak AlPO4 interphase and had a tortuous crack path in 10 h. The mullite layer is dense and the AIPOa layer is clearly the composite. These observations imply that AlPO4 can function porous. A single-phase AlPO a pellet sintered at 1600 C/(10 h) had as a porous and weak interphase material in a laminated composite a measured density of 1.56 g/cm, which corresponded to 61% of causing crack deflection to occur, and hence increase the overall theoretical densi ity, and a three-point bending strength of 1. 5 MPa toughness of a composite The low sinterability of the AlPO4 is consistent with the literature The overall thermal expansion coefficients of mullite and AlPO Mullite sintered under the same condition had a density and are 5.3 X 10-/K and 2. x 10-6/K, respectively. 2 Some residual three-point bending strength of 3. 11 gcm'(98% of theoretical tress may exist at the interfaces between the matrix and porous density)and 308 MPa, respectively. Figure 3 shows typical crack AlPO4 interphase due to thermal expansion coefficients mismatch. deflection in the mullite-aluminum phosphate laminate system However, we have fabricated numerous mullite and AlPO4 laminated after three-point bend testing. The crack is well-deflected along the samples and could find no line-broadening or peak shifts Table Il. Tape Casting Formulations for Mullite and AlPO Solvent Plasticizers MEK Dispersant Powder Extra additions Comments 25.1 576 13 5.7 4.7 5.6 30 vol% solvent Too high viscosity hyl alcohol USP, AAPER ALCOL and -21A, Witco); PVB= poly(vinyl butyral)( Butvar B90, Solutia), PG d polyethylene glycol) ooN F. Fcc Gonad c, Union Carbide dP ie dibutyl (99%, Aldrich Chemical).Flexural strengths were measured by three-point bend testing with a screw-driven machine (Model 4502, Instron Corp., Canton, MA). The flexural strength and work of fracture values were determined after testing three to five samples. The specimen size was 3 mm (H)  4 mm (W)  40 mm (L), the supporting span was 30 mm, and the crosshead speed was 0.1 mm/min. III. Results and Discussion The Kyoritsu KM 101 mullite had a particle size and specific surface area of 0.8 m and 8.5 m2 /g, respectively. The crystalline AlPO4 powder had particle sizes of 10 and 0.9 m, before and after 1-h attrition milling, respectively. The amorphous and crys￾talline form of AlPO4 powder had specific surface areas of 136 and 87 m2 /g, respectively, after 1-h attrition milling. Because of the concern for decomposition of AlPO4 at high temperature, ICP analyses were done for three different kinds of AlPO4, and the results are summarized in Table I. The three samples were (i) a standard AlPO4 (Aldrich, 99.99%); (ii) AlPO4 which was heat￾treated at 1600°C/(10 h) in air, followed by an anneal at 1400°C/ (100 h) in air; and (iii) AlPO4 which was heat-treated at 1800°C/(5 h) in air. If the reaction 2AlPO4 3 Al2O3 P2O5 occurs and P2O5 evaporates at high temperature, the relative amounts of Al in the sample will increase. Table I compares the atom percent of Al and P for the three different AlPO4 specimens. This table indicates that there were almost no differences in the relative amounts of the two elements in the three different aluminum phosphates. One is led to the conclusion that aluminum phosphate is chemically stable at high temperatures. Table II summarizes the tape-casting formulations for the mullite and AlPO4. The solid loading was 25.1 vol% for both of the powders. In the case of the AlPO4 tapes, 30% extra solvent was added because of its high viscosity. Chemical compatibility between AlPO4 and mullite was studied by X-ray diffractometry. The results (Fig. 1) indicate that mullite and aluminum phosphate are completely compatible with each other, without formation of any third phase. Figure 2 is a scanning electron micrograph (SEM) of the 3Al2O3
2SiO2/AlPO4 laminated composite sintered at 1600°C for 10 h. The mullite layer is dense and the AlPO4 layer is clearly porous. A single-phase AlPO4 pellet sintered at 1600°C/(10 h) had a measured density of 1.56 g/cm3 , which corresponded to 61% of theoretical density, and a three-point bending strength of 1.5 MPa. The low sinterability of the AlPO4 is consistent with the literature.4 Mullite sintered under the same condition had a density and three-point bending strength of 3.11 g/cm3 (98% of theoretical density) and 308 MPa, respectively. Figure 3 shows typical crack deflection in the mullite–aluminum phosphate laminate system after three-point bend testing. The crack is well-deflected along the porous, weak AlPO4 interphase and had a tortuous crack path in the composite. These observations imply that AlPO4 can function as a porous and weak interphase material in a laminated composite, causing crack deflection to occur, and hence increase the overall toughness of a composite. The overall thermal expansion coefficients of mullite and AlPO4 are 5.3  106 /K and 2.3  106 /K, respectively.12 Some residual stress may exist at the interfaces between the matrix and porous AlPO4 interphase due to thermal expansion coefficients mismatch. However, we have fabricated numerous mullite and AlPO4 laminated samples and could find no line-broadening or peak shifts. Fig. 1. X-ray diffraction profiles for a crushed pellet composed of 50 vol% mullite and 50% AlPO4, after sintering at 1600°C for 10 h. Fig. 2. SEM micrograph of a 3Al2O3
2SiO2/AlPO4 laminated composite sintered at 1600°C for 10 h. Table I. Experimental Analysis of AlPO4 after Three Different Heat Treatments Al (at.%) P (at.%) Standard AlPO4 (Aldrich, 99.99%) 21.41 24.77 1600°C/(10 h), air 3 1400°C/(100 h), air-treated AlPO4 21.56 25.03 1800°C/(5 h), air-treated AlPO4 21.56 24.30 Table II. Tape Casting Formulations for Mullite and AlPO4 † Powder Solvent‡ Dispersant‡ (PS) Binder‡ (PVG) Plasticizer‡ Extra additions Comments Eth (60%) MEK (40%) PG DP Mullite 25.1 57.6 1.3 5.7 4.7 5.6 – – AlPO4 25.1 57.6 1.3 5.7 4.7 5.6 30 vol% solvent Too high viscosity † All ingredients are in volume percent. ‡ Eth  ethanol (ethyl alcohol USP, AAPER ALCOL and Chemical); MEK  methyl ethyl ketone (99.8%, Fisher Scientific); PS  phosphate ester (Emphos PS-21A, Witco); PVB  poly(vinyl butyral) (Butvar B90, Solutia); PG  poly(ethylene glycol) (300NF, FCC Grade, Union Carbide); DP  dibutyl phthalate (99%, Aldrich Chemical). November 2003 Communications of the American Ceramic Society 1963