1970 Joumal of the American Ceramic Society-Boakye et al. Vol 87. No. 10 100 ZNS 60 ZES-C ZNS-C ZNS.C ZNS ES-C 20040060080010001200 40060080010001200 Temperature(C) Fig 3. DTA/TGA of ZNS, ZNS-C, ZES, and ZES-C precursors species at 1000oC. The ZNS-C precursor had -1% weight gain above refractory VO, phase should have formed, which would have been 900C, which possibly was from formation and subsequent oxidation ineffective as a flux of trace amounts of carbides. SiO, may have locally sea Unlike vanadium-doped ZNS, lithium-doped ZNS formed Zr- rom oxidation at -900-1200oC and, consequently, promoted local SiO4 in argon at 1200C(Fig. 2(d). However, in argon, lithium- carbothermal reduction of oxides to carbides. It is surprising that the doped ZNS-C formed only a trace of ZrSiO4 at 1200oC (Fig. 2(m)) TGAs of the ZNS and ZNS-C precursors were dissimilar at high and formed no ZrSiO4 at 1300C (Fig. 2(1)). Lithium also emperatures. The only difference between the two precursors was the enhanced ZrsiO4 formation by a liquid-phase flux mechanism, but PA added to ZNS-C, which should have burned off above 600oC. lithium formed only one stable oxide, which was unaffected by an When Sio, was absent from the precursors, as for ZN-C and ZP-C, argon atmosphere. Unlike the V2Os-doped precursors, the Li,o the weight loss was similar to that for the ZNS-C precursor below flux should have been effective; therefore, other explanations 550C; above 550C, weight loss was similar to that for the ZNs lack of ZrSiO4 formation must be sought. The slightly higher precursor(Fig. 4), as was expected formation rate at the lower temperature is difficult to explai Comparison of the weight loss from the ZNS and ZNS-C kinetically and suggests a thermodynamic cause precursors showed that the ZNS-C precursor yielded --55 wt%(36 The stability of ZrsiO4 in the presence of carbon was tested Dl%) ZrSiO and 45 wt%(64 vol%)carbon (4) ZrSio, Stability In the absence of vanadium or lithium dopa ZrSiO4(s)+C(s)-ZrO(s)+ Sio(g)+Co(g) conversion to Zrsio4 occurred at 1400.C for ZNS pre ZrSiO4 was stable in argon, but, in the presence of carbon, it Vanadium-doped ZNS recurse ely reacted to air at 21100C(Fig. 2(b)), but, in argon, the precursors did not partially decomposed to m-Zro2(Figs. 2(u)and (v). This result react until 1400C. Vanadium has been suggested to function as a uid-Phase flux through formation of melted V2Os,but, in argon and particularly in the presence of carbon, the much more (a) epoxy 100 Vextel20 10 (c) Nextel 20 2( nm1 Temperature℃ Fig. 5. TEM images of ZNS-C-derived coatings on Nextel 720 Coating was done at 1000.C in argon. Coatings are an intimate mechanical Fig 4. TGA of ZNS, ZNS-C, ZN-C, and ZP-C precursors. Weight loss is mixture of I-ZrO,, SiO2, and carbon (a)Coating pliability is evident on normalized to final solids yield at 1500C, and it is plotted logarithmically cracks fo filaments when they are close together. degradation occurs. Circled points show weight loss for precursors above during TEM specimen preparation. (c)Continuous coating coverage is 1000C, and they are used in Fig. 14 achieved. althou are sometimes thinspecies at 1000°C. The ZNS-C precursor had
1% weight gain above 900°C, which possibly was from formation and subsequent oxidation of trace amounts of carbides. SiO2 may have locally sealed carbon from oxidation at
900°–1200°C and, consequently, promoted local carbothermal reduction of oxides to carbides. It is surprising that the TGAs of the ZNS and ZNS-C precursors were dissimilar at high temperatures. The only difference between the two precursors was the PA added to ZNS-C, which should have burned off above 600°C. When SiO2 was absent from the precursors, as for ZN-C and ZP-C, the weight loss was similar to that for the ZNS-C precursor below 550°C; above 550°C, weight loss was similar to that for the ZNS precursor (Fig. 4), as was expected. Comparison of the weight loss from the ZNS and ZNS-C precursors showed that the ZNS-C precursor yielded
55 wt% (36 vol%) ZrSiO4 and 45 wt% (64 vol%) carbon. (4) ZrSiO4 Stability In the absence of vanadium or lithium dopants, complete conversion to ZrSiO4 occurred at 1400°C for ZNS precursors.31,41 Vanadium-doped ZNS precursors completely reacted to ZrSiO4 in air at
1100°C (Fig. 2(b)), but, in argon, the precursors did not react until 1400°C. Vanadium has been suggested to function as a liquid-phase flux through formation of melted V2O5, 69 but, in argon and particularly in the presence of carbon, the much more refractory VO2 phase should have formed, which would have been ineffective as a flux. Unlike vanadium-doped ZNS, lithium-doped ZNS formed Zr￾SiO4 in argon at 1200°C (Fig. 2(d)). However, in argon, lithium￾doped ZNS-C formed only a trace of ZrSiO4 at 1200°C (Fig. 2(m)) and formed no ZrSiO4 at 1300°C (Fig. 2(l)). Lithium also enhanced ZrSiO4 formation by a liquid-phase flux mechanism, but lithium formed only one stable oxide, which was unaffected by an argon atmosphere. Unlike the V2O5-doped precursors, the Li2O flux should have been effective; therefore, other explanations for lack of ZrSiO4 formation must be sought. The slightly higher formation rate at the lower temperature is difficult to explain kinetically and suggests a thermodynamic cause. The stability of ZrSiO4 in the presence of carbon was tested. ZrSiO4 and 46:54 ZrSiO4:carbon (PA) mixtures were heat-treated in argon at 1300°C for 2 h. The ZrSiO4 decomposition reaction is ZrSiO4(s) C(s) 3 ZrO2(s)  SiO(g)  CO(g) (2) ZrSiO4 was stable in argon, but, in the presence of carbon, it partially decomposed to m-ZrO2 (Figs. 2(u) and (v)). This result Fig. 3. DTA/TGA of ZNS, ZNS-C, ZES, and ZES-C precursors. Fig. 4. TGA of ZNS, ZNS-C, ZN-C, and ZP-C precursors. Weight loss is normalized to final solids yield at 1500°C, and it is plotted logarithmically to emphasize weight loss at high temperatures, where fiber strength degradation occurs. Circled points show weight loss for precursors above 1000°C, and they are used in Fig. 14. Fig. 5. TEM images of ZNS-C-derived coatings on NextelTM 720. Coating was done at 1000°C in argon. Coatings are an intimate mechanical mixture of t-ZrO2, SiO2, and carbon. (a) Coating pliability is evident on cracks formed during thin-section preparation. (b) Coatings bridge adjacent filaments when they are close together, and sometimes crack or debond during TEM specimen preparation. (c) Continuous coating coverage is achieved, although coatings are sometimes thin. 1970 Journal of the American Ceramic Society—Boakye et al. Vol. 87, No. 10