al / Journal of the European Ceramic Society 20(2000)589-59 may cause an otherwise functional coating to appear tem thin foils were made of coated fiber cross-sec- non-functional. Knowledge of the conditions under tions as described elsewhere 35.36 Thin foils were which degradation occurs is therefore important for observed in either a JEOL 2000 FX operating at 200 interface evaluation, besides the obvious importance to kv, or in a Phillips CM 200 FEG operating at 200 k composite strength and failure mechanisms. 24-26 Isolation Energy dispersive spectroscopy(EDS)measurements of the causes of strength degradation may lead to were done in the Phillips CM 200 FEG with a 5 nm spot improved coating methods that minimize this degradation. size and a windowless detector This work describes characteristics of 3M Nextel 720 Coated filament tensile strengths and Weibull modulii fiber coated with monazite between 900 and 1300C. were measured using a 2.54 cm gauge length with 75 Characteristics of the 900oC coated fiber after heat- tests. For controls, tensile strengths were also measured treatment for up to 100 h at 1200oC are also described. for filaments that had been heat-treated without coat Changes in composition and microstructure of the ing, and for filaments that had been passed through the coating and fiber were observed by TEM. Precursor and fiber coater under conditions that mimicked a coating coating evolution were monitored by X-ray, TEM, and run, but without coating deposition. Further details of DTA/TGA. Precursor gas evolution was observed by tensile strength measurement are presented elsewhere. 3 DTA/TGA and mass spectrometry. Strength of coated and uncoated fibers were measured by single filament tensile tests. Changes in coating characteristics with 3. Results temperature were compared with changes in the coated filament tensile strength. The results are analyzed and 3. 1. X-ray, DTA/TGA, and mass spectrometry possible degradation mechanisms are discussed Precursor heat-treated for l h at 1200 C has monazite y peaks and small La3PO7 peaks, indicating a slight phosphate deficiency from the monazite stoichiometry (ig. 1). A weak exotherm at about 450C and a weak Lanthanum nitrate and phosphorous pentoxide were endotherm at about 950C were the only significant dissolved in dry ethanol with the appropriate stoichio- DTA features(Fig. 2). About 4.5% of the precursor metry to form 50 g/l of monazite. 6 These solutions had a mass was lost between 100 and 550 C. Above 550C density of 0.84 g/cm and a viscosity of 1. 39 mPa- S, as mass loss was less rapid, but an additional 1.5%was measured in a Brookfield programmable rheometer still lost between 550 and 1500C. This 1.5% loss cor- ( model DV-IID) at a shear rate of 1/300. The precursor responds to a volume of gas that is at least 10 to 50 was characterized after a I h heat- treatment at 1200c times larger than the volume of monazite it evolved by X-ray diffraction in a Rigaku Rotaflex Dif- from. A slight increase in mass loss at around 950C fractometer. Differential thermal analysis (DTA)and correlates with the temperature of the weak DTA thermogravimetric analysis (TGA) were done in a endotherm. By mass spectrometry, only HO, Nor Netzsch STA-409 at 10C/min up to 1500 C after the CO, and N2o or Co? were observed to evolve in sig precursor was dried for I h at 140 C. Mass spectro- nificant quantities above 600 C(Fig. 3). Lack of amu metry of gases evolved from the precursor was done in a resolution precluded distinguishing N2 from Co(amu Balzers QMs 420 at 5C/min up to 1050C. The mea- 28)or N20 from CO2(amu 44). Methane(CH4) and surement was done in argon so measurement overlap CH3 evolution at 500C roughly corresponds to the with atmospheric gases could be eliminated. A more detailed description of the mass spectrometry equip- ment is given elsewhere. 27 3M Nextel 720 alumina-mullite fibers 28-30 were con- tinuously coated with the monazite precursor in a ver- tical coater using hexadecane for immiscible liquid displacement. A coating speed of 1. 4 cm/s in an air atmosphere was used for all experiments. The furnace hot zone was about 8 cm in length, and total furnace length was 30 cm. The fiber coating apparatus and pro- cedures are described in more detail elsewhere 6.31-34 The fibers were desized in air at 1000C and 2.8 cm/s before coating. Coating runs were done at 900, 1000, 1 100, 1200, and 1300 C. Some fibers coated at 900C Fig. 1. X-ray diffraction pattern of monazite from precursor heat were then heat-treated at 1200%C for 0.2. 2. 10. or 100 h treated for I h at 1200C. All peaks correspond to monazite except in a furnace with MoSi, heating elements those highlighted in gray, which correspond to La,may cause an otherwise functional coating to appear non-functional. Knowledge of the conditions under which degradation occurs is therefore important for interface evaluation, besides the obvious importance to composite strength and failure mechanisms.24±26 Isolation of the causes of strength degradation may lead to improved coating methods that minimize this degradation. This work describes characteristics of 3M Nextel 720 ®ber coated with monazite between 900 and 1300C. Characteristics of the 900C coated ®ber after heat￾treatment for up to 100 h at 1200C are also described. Changes in composition and microstructure of the coating and ®ber were observed by TEM. Precursor and coating evolution were monitored by X-ray, TEM, and DTA/TGA. Precursor gas evolution was observed by DTA/TGA and mass spectrometry. Strength of coated and uncoated ®bers were measured by single ®lament tensile tests. Changes in coating characteristics with temperature were compared with changes in the coated ®lament tensile strength. The results are analyzed and possible degradation mechanisms are discussed. 2. Experiments Lanthanum nitrate and phosphorous pentoxide were dissolved in dry ethanol with the appropriate stoichio￾metry to form 50 g/l of monazite.6 These solutions had a density of 0.84 g/cm3 and a viscosity of 1.39 mPa.s, as measured in a Brook®eld programmable rheometer (model DV-III) at a shear rate of 1/300. The precursor was characterized after a 1 h heat-treatment at 1200C by X-ray di€raction in a Rigaku Rota¯ex Dif￾fractometer. Di€erential thermal analysis (DTA) and thermogravimetric analysis (TGA) were done in a Netzsch STA-409 at 10C/min up to 1500C after the precursor was dried for 1 h at 140C. Mass spectro￾metry of gases evolved from the precursor was done in a Balzers QMS 420 at 5C/min up to 1050C. The mea￾surement was done in argon so measurement overlap with atmospheric gases could be eliminated. A more detailed description of the mass spectrometry equip￾ment is given elsewhere.27 3M Nextel 720 alumina±mullite ®bers28±30 were con￾tinuously coated with the monazite precursor in a ver￾tical coater using hexadecane for immiscible liquid displacement. A coating speed of 1.4 cm/s in an air atmosphere was used for all experiments. The furnace hot zone was about 8 cm in length, and total furnace length was 30 cm. The ®ber coating apparatus and pro￾cedures are described in more detail elsewhere.6,31±34 The ®bers were desized in air at 1000C and 2.8 cm/s before coating. Coating runs were done at 900, 1000, 1100, 1200, and 1300C. Some ®bers coated at 900C were then heat-treated at 1200C for 0.2, 2, 10, or 100 h in a furnace with MoSi2 heating elements. TEM thin foils were made of coated ®ber cross-sec￾tions as described elsewhere.35,36 Thin foils were observed in either a JEOL 2000 FX operating at 200 kV, or in a Phillips CM 200 FEG operating at 200 kV. Energy dispersive spectroscopy (EDS) measurements were done in the Phillips CM 200 FEG with a 5 nm spot size and a windowless detector. Coated ®lament tensile strengths and Weibull modulii were measured using a 2.54 cm gauge length with 75 tests. For controls, tensile strengths were also measured for ®laments that had been heat-treated without coat￾ing, and for ®laments that had been passed through the ®ber coater under conditions that mimicked a coating run, but without coating deposition. Further details of tensile strength measurement are presented elsewhere.37 3. Results 3.1. X-ray, DTA/TGA, and mass spectrometry Precursor heat-treated for 1 h at 1200C has monazite X-ray peaks and small La3PO7 peaks, indicating a slight phosphate de®ciency from the monazite stoichiometry (Fig. 1). A weak exotherm at about 450C and a weak endotherm at about 950C were the only signi®cant DTA features (Fig. 2). About 4.5% of the precursor mass was lost between 100 and 550C. Above 550C mass loss was less rapid, but an additional 1.5% was still lost between 550 and 1500C. This 1.5% loss cor￾responds to a volume of gas that is at least 10 to 50 times larger than the volume of monazite it evolved from. A slight increase in mass loss at around 950C correlates with the temperature of the weak DTA endotherm. By mass spectrometry, only H2O, N2 or CO, and N2O or CO2 were observed to evolve in sig￾ni®cant quantities above 600C (Fig. 3). Lack of amu resolution precluded distinguishing N2 from CO (amu 28) or N2O from CO2 (amu 44). Methane (CH4) and CH3 evolution at 500C roughly corresponds to the Fig. 1. X-ray di€raction pattern of monazite from precursor heat￾treated for 1 h at 1200C. All peaks correspond to monazite except those highlighted in gray, which correspond to La3PO7. 590 R.S. Hay et al. / Journal of the European Ceramic Society 20 (2000) 589±597