.S. Hay et al. / Journal of the European Ceramic Society 20(2000)589-59 with even less porosity. The most obvious features after 3.4. Tensile strength 100 h at 1200oC were spheroidization of coatings less than 50 nm thick(Fig. 7), roughening of the monazite Coated filament tensile strength decreased from 2 coating-fiber interface( Figs. 6-8), and facetting of alu- GPa28 with both increased deposition temperature and mina grains in the fiber and near the coating along basal with time at 1200 C after deposition at 900 C ( Table 1 planes(Figs. 6-8). Lanthanum segregation at facetted Fig. 10). Filaments coated at 1300C had average tensile alumina-mullite interfaces near the coating was mea- strengths of 1. 21 GPa. Filaments heat-treated for 100 h sured by analytical TEM(Fig 8). Phosphorous segre- at 1200 C after coating at 900oC had only 0.83 GPa gation was not found and lanthanum was not found in average tensile strength. The actual average may have the interior of alumina grains. Wetting thin films were been slightly lower, because some filaments were too not observed along monazite-alumina, monazite-mul- weak to mount in grips for tensile testing, and their lite, or alumina-mullite interphase boundaries, or along contribution to the average would certainly have made the corresponding triple junctions(Fig. 9) it lower. Filaments coated at 1200C were slightly stronger than those coated at 1100 C, as were filaments 900℃C heat-treated for 2 h at 1200oC compared to 12 min at 1200.C. These observations were anomalous to the overall trend; it is possible that they are related to tran- sient flaw healing, but more data is required to sub stantiate this Control experiments suggest that much of the strength drop may be unrelated to the presence of a coating. Filaments heat-treated for 100 h at 1200C without a coating had only 1.04 GPa tensile strength (Table 1, Fig. 10). Similarly, filaments passed through the coating furnace at 1300C and 1200oC had only 1.5 GPa tensile strength, although filaments passed through water, water/HNO3, ethanol, and water/ octanol were stronger (Table I, Fig. 10). The strength drop between fibers coated at 1000 and 1100C was steeper than others, and may correlate with the slight 1300℃C Fig. 5. TEM pictures of coating deposited at(a)900oC,(b)I (c)1100 C, (d)1300C Simulated and observed diffraction patt 50 nm 900C coating and observed diffraction pattern from a 1300C are to the rig 200nm 2000mm Fig. 6. TEM pictures of coat posited at 900C and later heat- treated at 1200C for 2 h or 10 Fig. 7. TEM pictures of spheroidized coatings.with even less porosity. The most obvious features after 100 h at 1200C were spheroidization of coatings less than 50 nm thick (Fig. 7), roughening of the monazite coating±®ber interface (Figs. 6±8), and facetting of alu￾mina grains in the ®ber and near the coating along basal planes (Figs. 6±8). Lanthanum segregation at facetted alumina±mullite interfaces near the coating was mea￾sured by analytical TEM (Fig. 8). Phosphorous segre￾gation was not found, and lanthanum was not found in the interior of alumina grains. Wetting thin ®lms were not observed along monazite±alumina, monazite±mul￾lite, or alumina±mullite interphase boundaries, or along the corresponding triple junctions (Fig. 9). 3.4. Tensile strength Coated ®lament tensile strength decreased from 2 GPa28 with both increased deposition temperature and with time at 1200C after deposition at 900C (Table 1, Fig. 10). Filaments coated at 1300C had average tensile strengths of 1.21 GPa. Filaments heat-treated for 100 h at 1200C after coating at 900C had only 0.83 GPa average tensile strength. The actual average may have been slightly lower, because some ®laments were too weak to mount in grips for tensile testing, and their contribution to the average would certainly have made it lower. Filaments coated at 1200C were slightly stronger than those coated at 1100C, as were ®laments heat-treated for 2 h at 1200C compared to 12 min at 1200C. These observations were anomalous to the overall trend; it is possible that they are related to tran￾sient ¯aw healing, but more data is required to sub￾stantiate this. Control experiments suggest that much of the strength drop may be unrelated to the presence of a coating. Filaments heat-treated for 100 h at 1200C without a coating had only 1.04 GPa tensile strength (Table 1, Fig. 10). Similarly, ®laments passed through the coating furnace at 1300C and 1200C had only 1.5 GPa tensile strength, although ®laments passed through water, water/HNO3, ethanol, and water/1- octanol were stronger (Table 1, Fig. 10). The strength drop between ®bers coated at 1000 and 1100C was steeper than others, and may correlate with the slight Fig. 5. TEM pictures of coating deposited at (a) 900C, (b) 1000C, (c) 1100C, (d) 1300C. Simulated and observed di€raction pattern of 900C coating and observed di€raction pattern from a 1300C coating are to the right. Fig. 6. TEM pictures of coating deposited at 900C and later heat￾treated at 1200C for 2 h or 100 h. Fig. 7. TEM pictures of spheroidized coatings. 592 R.S. Hay et al. / Journal of the European Ceramic Society 20 (2000) 589±597