《复合材料 Composites》课程教学资源（学习资料）第五章 陶瓷基复合材料_j.1151-2916.1999.tb01944.x[1]

ournal JAm.C.soc,8214-82(1999 High-Temperature Oxidation of Boron Nitride ll, Boron Nitride Layers in Composites Nathan S Jacobson and Gregory N. Morscher NASA Lewis Research Center, Cleveland, Ohio 44135 Darren R. Bryant and Richard E Tressler The Pennsylvania State University, University Park, Pennsylvania 16802 The oxidation of BN composite interphases was examined SiC oxidation can occur at even lower temperatures, because with a series of model materials. Oxidation was examined of the boron interacting with the SiO, scale as it forms. The in both low-water-vapor(-20 ppm H2O/O2) environments SiO, and B2O3 react to form stable borosilicates, and there is at 900C and high-water-vapor(1% and 10%H2O/O2)en- SiO2-B2O3 eutectic at 372.C. Oxygen diffusivities are more vironments at 700 and 800C. The low-water-vapor case rapid in a borosilicate glass, and, hence, SiC oxidation is en was explored with layered BN/SiC materials. This case was hanced.5,6 These reactions lead to degradation of fiber prope dominated by borosilicate glass formation, and the 20 ppm ties as well as fiber/matrix bonding, which then degrades the water vapor gradually removed the boron from the glass, mechanical properties of the composite. Figure 1(a)illustrates er amou nt of SiO2 than would be expected oxidation of a SiC/SiC composite where borosilicate forms from simple SiC oxidation. Layered Sic/BN/SiC materials were also used to study low-water-vapor oxidation effects within the composite. The high-water-vapor case was ex- ed with SiC/BN/SiC minicomposites, and it was dom nated by volatilization of BN as HBO (8), H3 BO3 (8), and H3 BO(8). A model for recession of the BN fiber coating was developed based on the gas-phase diffusion of these species out of the annular region around the SiC fiber and concurrent sealing of this annular region by oxidation . Introduction B N NITRIDE(BN) is a promising interphase material for SiC-fiber-reinforced SiC-matrix composites. It allows the debonding and fiber-sliding requisite for strong and tough com osite behavior. There are several detailed microstructural studies of these as-processed composites. 1, 2 Ideally the micro- structure should consist of the SiC matrix, a BN fiber coating, and the Sic fiber 10 um The critical question is how this structure changes in service ith aggressive gases at elevated temperatures. With no matrix cracks, BN is stable in contact with SiC at high temperatures This has been confirmed with both thermodynamic calculations and experiments. 3 The only reaction noted in a low-oxygen- potential environment is the solubility of any free carbon in BN 3 However, in the case of a matrix crack, bn would be ex posed to high oxygen potentials. bn oxidizes to a liquid oxide B2O3) above-450°Cas 2BN+502(g)=B,O3()+N2(g) (1) At temperatures above -900C, the SiC oxidizes simulta- SiC +=O,(g)+SiO,(s)+co(g) T.A. Parthasarathy--contributing editor Fig. 1. Micrographs of composites and minicomposites with one image of a composite with CVD Bn deposited at <1000C, 100 h in oxygen at 816C(photo courtesy of D. Fox, NASA Lewis)and(b) 191031 Received May 8, 1997; approved September 24, 1998 secondary electron image of a minicomposite with CVD deposited at ican Ceramic Society -1000oC, 100 h in humid laboratory air at 50

High-Temperature Oxidation of Boron Nitride: II, Boron Nitride Layers in Composites Nathan S. Jacobson* and Gregory N. Morscher* NASA Lewis Research Center, Cleveland, Ohio 44135 Darren R. Bryant* and Richard E. Tressler* The Pennsylvania State University, University Park, Pennsylvania 16802 The oxidation of BN composite interphases was examined with a series of model materials. Oxidation was examined in both low-water-vapor (∼20 ppm H2O/O2) environments at 900°C and high-water-vapor (1% and 10% H2O/O2) en￾vironments at 700° and 800°C. The low-water-vapor case was explored with layered BN/SiC materials. This case was dominated by borosilicate glass formation, and the 20 ppm water vapor gradually removed the boron from the glass, leaving a larger amount of SiO2 than would be expected from simple SiC oxidation. Layered SiC/BN/SiC materials were also used to study low-water-vapor oxidation effects within the composite. The high-water-vapor case was ex￾plored with SiC/BN/SiC minicomposites, and it was domi￾nated by volatilization of BN as HBO2(g), H3BO3(g), and H3B3O6(g). A model for recession of the BN fiber coating was developed based on the gas-phase diffusion of these species out of the annular region around the SiC fiber and concurrent sealing of this annular region by oxidation. I. Introduction BORON NITRIDE (BN) is a promising interphase material for SiC-fiber-reinforced SiC-matrix composites. It allows the debonding and fiber-sliding requisite for strong and tough com￾posite behavior. There are several detailed microstructural studies of these as-processed composites.1,2 Ideally the micro￾structure should consist of the SiC matrix, a BN fiber coating, and the SiC fiber. The critical question is how this structure changes in service with aggressive gases at elevated temperatures. With no matrix cracks, BN is stable in contact with SiC at high temperatures. This has been confirmed with both thermodynamic calculations and experiments.1,3 The only reaction noted in a low-oxygen￾potential environment is the solubility of any free carbon in BN.3 However, in the case of a matrix crack, BN would be ex￾posed to high oxygen potentials. BN oxidizes to a liquid oxide (B2O3) above ∼450°C as 2BN + 3 2 O2~g! = B2O3~l! + N2~g! (1) At temperatures above ∼900°C, the SiC oxidizes simulta￾neously: SiC + 3 2 O2~g! + SiO2~s! + CO~g! (2) SiC oxidation can occur at even lower temperatures, because of the boron interacting with the SiO2 scale as it forms. The SiO2 and B2O3 react to form stable borosilicates, and there is SiO2–B2O3 eutectic at 372°C.4 Oxygen diffusivities are more rapid in a borosilicate glass, and, hence, SiC oxidation is en￾hanced.5,6 These reactions lead to degradation of fiber proper￾ties as well as fiber/matrix bonding, which then degrades the mechanical properties of the composite.7 Figure 1(a) illustrates oxidation of a SiC/SiC composite where borosilicate forms. T. A. Parthasarathy—contributing editor Manuscript No. 191031. Received May 8, 1997; approved September 24, 1998. *Member, American Ceramic Society. Fig. 1. Micrographs of composites and minicomposites with one edge ground off and exposed to oxidizing gases: (a) secondary electron image of a composite with CVD BN deposited at <1000°C, 100 h in oxygen at 816°C (photo courtesy of D. Fox, NASA Lewis) and (b) secondary electron image of a minicomposite with CVD deposited at ∼1000°C, 100 h in humid laboratory air at 500°C. J. Am. Ceram. Soc., 82 [6] 1473–82 (1999) Journal 1473

Journal of the American Ceramic SocietyJacobson et al. VoL. 82. No 6 One face has been ground to expose the fibers and BN coatings In addition, oxidation studies in low-water-vapor-containing or oxidation. After oxidation, a borosilicate ring forms around environments were done on SiC/BN/SiC layered structures each fiber These structures consisted of a CVd SiC coupon coated with In a companion paper, oxidation of monolithic BN was -10 um CVD BN, an intermediate layer of CVD SiC,-10 um explored. It was shown that oxidation rates were quite sensi- CVD BN, and finally an overcoat of -100 um CVD SiC. The ive to porosity, oxygen impurity levels, and crystallographic intermediate layer of CVD SiC was of varying thickness- rientation. Two types of chemically-vapor-deposited(CVD) om a -l um layer to -40 um mounds. As is shown, th BN were examined--one deposited at -1900 K and one de intermediate layer was quite useful in elucidating oxidation sited at -1400 K. bn deposited at the higher temperature behavior. For oxidation treatment, one face was gently polished had a higher density and a lower oxygen content and exhibited with nonaqueous solvents to expose the bn and BN/SiC inter- much better oxidation resistance than bn deposited at lower es ior oX idation, as shown in Fig. 1. Oxidation was per formed in the TGA, although weight changes were negligible It was also shown that the oxidation of bn was sensitive to The third system involved minicomposites. These were even small amounts of water vapor in the oxidizing gas stream single tows of -10 Hm unidirectional fibers that were first due to the formation of highly stable HBo,(g)si from the coated with -0. 5 um of Bn and then Sic to simulate a matrix reaction of water vapor and B,Oa(e). In the case of BN depos This is described in more detail in table i and Refs. 10 and 12 ited at-1900 K, the weight change curves exhibited paralinear One face of the minicomposites was ground in order to behavior due to the simultaneous weight gain from bn oxida- the fibers and fiber coatings for oxidation. Oxida tion and weight loss from B,O3(e) volatilization as HBO2 (g ments were done in a horizontal furnace with oxygen h Typical processing temperatures of BN interphases in SiC/ through water in order to create a controlled-water-vapor Sic composites are relatively low (1000C) in order to content stream achieve complete fiber coverage and uniform fiber-coatin thicknesses There are several oxidation studies of these com- (2) Postexposure Examination posites in the literature.9-I One oxidation study shows that a After oxidation exposure, the samples were examined with several techniques, depending on the exposure. The oxidized low-processing-temperature BN on a SiC fiber in an oxide BN/SiC samples were examined with both scanning electron omposite exhibits limited oxidation at 1100C. The SiC fiber xidizes in preference to the bn, which is consistent with cresco and Rutherford backscattering thermochemical models. However the situation changes in the copy(RBS). The RBS was done at SUNY-Albany. Helium presence of water vapor. Morscher et al. o have shown that a atoms with an incident energy of 2 Mev were between the beam and detector path was 14, and the angle low-processing-temperature BN fiber coating volatilizes readily in the presence of water vapor. This is illustrated in Fig between the sample normal and the detector was 7. a ste (b). Less volatilization occurs with BN materials processed at owers database allowed fitting the spectra to 0.01 Mev. the gh temperatures and/or doped with silicon. Brun and Luthra'1 standard"RUMP"analysis code was used 13 Estimated sensi- have recently examined oxidation of the BN phase in SiC/Sic tivity for boron was 0.5 at. % with an uncertainty of+25% nd of their composite and ex- The SiC/BN/SiC layered structures and minicomposites were examined by standard electron optical techniques-SEM pose it to an oxidizing environment containing a high concen- tration of water vapor. They observe BN oxidation to B2O3 and and electron microprobe analysis(EPMA). The Imens volatilization due to water vapor. Borosilicate glasses form, but were mounted and polished perpendicular to the In the case of the layered structure, this is shown in Fig. 2, in the boron appears to be leached out because of the presence of the case of the minicomposites. this was parallel to the fibers r. a series of model materials are examined to Polishing was done with nonaqueous solvents to preserve the study the oxidation of BN interphases in SiC/SiC composites water-soluble phases. In the SEM, energy dispersive sp We consider oxidation in both low-water-vapor environments opy(EDs)was used for elemental where borosilicate formation dominates, and high-water-vapor EPMA. wavelength di dispersive spectroscopy(WDS)was use nvironments where volatilization dominates to allow boron detection in the layered specimens alison.G bon maps were omitted due to embedding of polishing com- Il. Experimental Procedures ound in the soft bn phases. Figure 2 illustrates the locations (marked as regions a, b, and c) for each EPMA image and corresponding elemental maps Model Materials and Oxidation Exposures Three types of model materials were examined. Table I sum- IlL. Results and discussion marizes these model materials and their corresponding expo- sures. The oxidation experiments in low-water-vapor containing environments were done with samples consisting of 1)Oxidation in Low-Water-Vapor Environments--BN/ -2 um thick BN chemically-vapor-deposited on a CVD Sic SiC Model Compounds coupon. These were oxidized in a controlled-atmosphere ther- Consider first the experiments on BN films deposited on mogravimetric apparatus(TGA) CVD SiC. Two types of BN are used, as shown in Table I. The Table L. Model Compounds, Types of BN, and Reaction Temperatures for These Studies bn d Model material mperature°o temperatures(°C Oxidation environment Oxidation-low 2 Hm BN film on CVD <1000 water vapor 19003 Oxidation -low CVD SIC/BN on Cvd -1900° O2/20 ppm H,O Volatilization of Bn- Minicomposite--SIC -10001 O3/1%or10%H2O fibers coated with BI Advanced Ceramics Corp "BN deposited by 3M Cor St Paul, posited by Advanced Ceramics Corp, Lakewood, OH. SiC coupon from Morton Inc "SiC, BN deposited by

One face has been ground to expose the fibers and BN coatings for oxidation. After oxidation, a borosilicate ring forms around each fiber. In a companion paper,8 oxidation of monolithic BN was explored. It was shown that oxidation rates were quite sensi￾tive to porosity, oxygen impurity levels, and crystallographic orientation. Two types of chemically-vapor-deposited (CVD) BN were examined—one deposited at ∼1900 K and one de￾posited at ∼1400 K. BN deposited at the higher temperature had a higher density and a lower oxygen content and exhibited much better oxidation resistance than BN deposited at lower temperatures. It was also shown that the oxidation of BN8 was sensitive to even small amounts of water vapor in the oxidizing gas stream due to the formation of highly stable HBO2(g) species from the reaction of water vapor and B2O3(,). In the case of BN depos￾ited at ∼1900 K, the weight change curves exhibited paralinear behavior due to the simultaneous weight gain from BN oxida￾tion and weight loss from B2O3(,) volatilization as HBO2(g). Typical processing temperatures of BN interphases in SiC/ SiC composites are relatively low (∼1000°C) in order to achieve complete fiber coverage and uniform fiber-coating thicknesses. There are several oxidation studies of these com￾posites in the literature.9–11 One oxidation study9 shows that a low-processing-temperature BN on a SiC fiber in an oxide composite exhibits limited oxidation at 1100°C. The SiC fiber oxidizes in preference to the BN, which is consistent with thermochemical models. However, the situation changes in the presence of water vapor. Morscher et al.10 have shown that a low-processing-temperature BN fiber coating volatilizes readily in the presence of water vapor. This is illustrated in Fig. 1(b). Less volatilization occurs with BN materials processed at high temperatures and/or doped with silicon. Brun and Luthra11 have recently examined oxidation of the BN phase in SiC/SiC composites. They grind off an end of their composite and ex￾pose it to an oxidizing environment containing a high concen￾tration of water vapor. They observe BN oxidation to B2O3 and volatilization due to water vapor. Borosilicate glasses form, but the boron appears to be leached out because of the presence of water vapor. In this paper, a series of model materials are examined to study the oxidation of BN interphases in SiC/SiC composites. We consider oxidation in both low-water-vapor environments, where borosilicate formation dominates, and high-water-vapor environments, where volatilization dominates. II. Experimental Procedures (1) Model Materials and Oxidation Exposures Three types of model materials were examined. Table I sum￾marizes these model materials and their corresponding expo￾sures. The oxidation experiments in low-water-vapor￾containing environments were done with samples consisting of ∼2 mm thick BN chemically-vapor-deposited on a CVD SiC coupon. These were oxidized in a controlled-atmosphere ther￾mogravimetric apparatus (TGA). In addition, oxidation studies in low-water-vapor-containing environments were done on SiC/BN/SiC layered structures. These structures consisted of a CVD SiC coupon coated with ∼10 mm CVD BN, an intermediate layer of CVD SiC, ∼10 mm CVD BN, and finally an overcoat of ∼100 mm CVD SiC. The intermediate layer of CVD SiC was of varying thickness— from a ∼1 mm layer to ∼40 mm mounds. As is shown, this intermediate layer was quite useful in elucidating oxidation behavior. For oxidation treatment, one face was gently polished with nonaqueous solvents to expose the BN and BN/SiC inter￾faces for oxidation, as shown in Fig. 1. Oxidation was per￾formed in the TGA, although weight changes were negligible. The third system involved minicomposites. These were single tows of ∼10 mm unidirectional fibers that were first coated with ∼0.5 mm of BN and then SiC to simulate a matrix. This is described in more detail in Table I and Refs. 10 and 12. One face of the minicomposites was ground in order to expose the fibers and fiber coatings for oxidation. Oxidation treat￾ments were done in a horizontal furnace with oxygen bubbled through water in order to create a controlled-water-vapor￾content stream. (2) Postexposure Examination After oxidation exposure, the samples were examined with several techniques, depending on the exposure. The oxidized BN/SiC samples were examined with both scanning electron microscopy (SEM) and Rutherford backscattering spectros￾copy (RBS). The RBS was done at SUNY-Albany. Helium atoms with an incident energy of 2 MeV were used; the angle between the beam and detector path was 14°, and the angle between the sample normal and the detector was 7°. A stopping powers database allowed fitting the spectra to 0.01 MeV. The standard “RUMP” analysis code was used.13 Estimated sensi￾tivity for boron was 0.5 at.% with an uncertainty of ±25%. The SiC/BN/SiC layered structures and minicomposites were examined by standard electron optical techniques—SEM and electron microprobe analysis (EPMA). The specimens were mounted and polished perpendicular to the exposed face. In the case of the layered structure, this is shown in Fig. 2; in the case of the minicomposites, this was parallel to the fibers. Polishing was done with nonaqueous solvents to preserve the water-soluble phases. In the SEM, energy dispersive spectros￾copy (EDS) was used for elemental identification. In the EPMA, wavelength dispersive spectroscopy (WDS) was used to allow boron detection in the layered specimens. Elemental maps were done for boron, nitrogen, oxygen, and silicon. Car￾bon maps were omitted due to embedding of polishing com￾pound in the soft BN phases. Figure 2 illustrates the locations (marked as regions a, b, and c) for each EPMA image and corresponding elemental maps. III. Results and Discussion (1) Oxidation in Low-Water-Vapor Environments—BN/ SiC Model Compounds Consider first the experiments on BN films deposited on CVD SiC. Two types of BN are used, as shown in Table I. The Table I. Model Compounds, Types of BN, and Reaction Temperatures for These Studies Study Model material BN deposition temperature (°C) Oxidation temperatures (°C) Oxidation environment Oxidation—low 2 mm BN film on CVD <1000† 900 O2/20 ppm H2O water vapor SiC ∼1900‡ Oxidation—low CVD SiC/BN on CVD ∼1900¶ 900 O2/20 ppm H2O water vapor SiC coupon§ 1100 Volatilization of BN— Minicomposite—SiC ∼1000†† 700 O2/1% or 10% H2O high water vapor fibers coated with BN 800 in SiC matrix † BN deposited by General Electric Co., Schnectady, NY. ‡ BN deposited by Advanced Ceramics Corp., Lakewood, OH. § SiC coupon from Morton Inc. ¶ SiC, BN deposited by Advanced Ceramics Corp. ††BN deposited by 3M Corp., St. Paul, MN. 1474 Journal of the American Ceramic Society—Jacobson et al. Vol. 82, No. 6

une 1999 High-Temperature Oxidation of Boron Nitride: Boron Nitride Layers in Composites 1475 Grind one face BN 10 -100μmsic CVD o8 SiC/BN--2 hr -10μmBN SiC/BN--50 h Borosilicate glass 0.2 Oxidize 11 0005101620253035 Reglon a Distance from Sic/borosilicate interface, um Region b Fig. 4. Interpreted RBS data of CVD SiC with 2 um BN after oxi- Regis dation at 900]C for 2 h and 50 h. in a borosilicate liquid follows approximately ideal solution behavior and, thus, would have a high activity in this situa- tion. Note, the weight loss is not linear-apparently the diffu- sion of B,O3 out of the glass influences the kinetics. After 50 h, the RBS information(Fig. 4)shows little B,O, in the oxi- dation product, supporting this interpretation ount, grind and polish SEM examination of the surface after 50 h shows a smooth econd face to examine featureless glass. Figure 5 shows a cross section of this glass xtent of oxidation The average thickness of this product(-8 um)is substantially greater than that expected from oxidation of pure SiC in oxy Fig. 2. Preparation of layered samples. gen(<0.5 um). This indicates that boron-enhanced SiC oxi- dation is occurring simultaneously with the oxidation of BN B,O3 from the borosilicate. As discusse TGA results for the Bn deposited at <1000 C are shown in Fig. rates in SiO, and, hence, enhance oxidation rates. s ansport 3. Note, there is an initial rapid weight gain and then a slower Figure 6 illustrates the reaction kinetics for a high weight loss. After 2 h, a thick glassy borosilicate layer is ob temperature(1900C)CVD BN deposited on CVD SiC. The served on the coupon environment is the same as that for Fig. 3. The initial rapid In order to determine the relative amounts of B2O3 in the weight loss is likely an experimental artifact In the first 40 h, lass, RBS measurements are taken. Results shown in Fig 4 a weight gain occurs due to BN and SiC oxidation. This period indicate that, after 2 h, the oxidation product is a borosilicate of weight gain is also observed for the low-deposition liquid with a substantial quantity of B2O3 temperature BN(Fig. 3), but the rate for the latter is much After the initial weight gain, the specimen slowly loses faster because of the higher oxidation rates of low-deposition- weight. Thermochemically, this can only be due to volatilize temperature BN. Figure 6 also shows that, after h, a tion of B2O3 as HBO2(g) from reactions with residual water vapor. This is expected, because the chemical activity of B2O3 吕016 012 008 0510162025303540 Fig. 5. Cross section of CVD SiC with 2 um BN after oxidation at

TGA results for the BN deposited at <1000°C are shown in Fig. 3. Note, there is an initial rapid weight gain and then a slower weight loss. After 2 h, a thick glassy borosilicate layer is ob￾served on the coupon. In order to determine the relative amounts of B2O3 in the glass, RBS measurements are taken. Results shown in Fig. 4 indicate that, after 2 h, the oxidation product is a borosilicate liquid with a substantial quantity of B2O3. After the initial weight gain, the specimen slowly loses weight. Thermochemically, this can only be due to volatiliza￾tion of B2O3 as HBO2(g) from reactions with residual water vapor. This is expected, because the chemical activity of B2O3 in a borosilicate liquid follows approximately ideal solution behavior14 and, thus, would have a high activity in this situa￾tion. Note, the weight loss is not linear—apparently the diffu￾sion of B2O3 out of the glass influences the kinetics. After 50 h, the RBS information (Fig. 4) shows little B2O3 in the oxi￾dation product, supporting this interpretation. SEM examination of the surface after 50 h shows a smooth, featureless glass. Figure 5 shows a cross section of this glass. The average thickness of this product (∼8 mm) is substantially greater than that expected from oxidation of pure SiC in oxy￾gen (<0.5 mm).15 This indicates that boron-enhanced SiC oxi￾dation is occurring simultaneously with the oxidation of BN and vaporization of B2O3 from the borosilicate. As discussed in the introduction, small amounts of boron enhance transport rates in SiO2 and, hence, enhance oxidation rates.5 Figure 6 illustrates the reaction kinetics for a high￾temperature (∼1900°C) CVD BN deposited on CVD SiC. The environment is the same as that for Fig. 3. The initial rapid weight loss is likely an experimental artifact. In the first 40 h, a weight gain occurs due to BN and SiC oxidation. This period of weight gain is also observed for the low-deposition￾temperature BN (Fig. 3), but the rate for the latter is much faster because of the higher oxidation rates of low-deposition￾temperature BN.8 Figure 6 also shows that, after ∼60 h, a Fig. 2. Preparation of layered samples. Fig. 3. TGA kinetics of a CVD SiC coupon coated with 2 mm of low-deposition-temperature (<1000°C) CVD BN oxidized in O2/20 ppm H2O at 900°C. Fig. 4. Interpreted RBS data of CVD SiC with 2 mm BN after oxi￾dation at 900°C for 2 h and 50 h. Fig. 5. Cross section of CVD SiC with 2 mm BN after oxidation at 900°C for 50 h. June 1999 High-Temperature Oxidation of Boron Nitride: Boron Nitride Layers in Composites 1475

14 Journal of the American Ceramic SocietyJacobson et al. Vol. 82. No 6 In summary, at low oxygen contents, such as would be found deep within a composite, SiC is expected to oxidize and effe tively getter oxygen. At higher oxygen contents, SiC and BN ding to borosilicate glass formation As time progresses, residual water vapor volatilizes the B2O3 TIme, h (3) Oxidation in Higli-) Measurements of Recession Fig. 6. TGA kinetics of CvD Sic coated with 2 Figure 1(b)illustrates the effect of a high-water-vapor con- Dosition-temperature(1900C)CVD BN tent and lower-temperature gas-oxidizing gas stream. Under 70ppmH2Oat900°C. hese conditions. volatilization dominated. and the bn was om composite. This was most pro- nced for the bn deposited at lower t weight loss is observed, which is very likely due to water- cal parameter was the depth of recession. These recession dis- apor-enhanced boron volatilization from the borosilicate tances were measured with single -tow, BN-coated CVD glass. Thus, the two types of bn behave qualitatively ver minicomposites with a range of interphase thicknesses, as de different rates due to the different types of BN osses-but at scribed in the experimental section. After heat treatment les were An important implication for composites is that, as boron is determine the distance of bn recession due to volatilization. 19 emoved from the borosilicate liquid, the viscosity of the bo- Figure 9 shows an example of a polished longitudinal section rosilicate liquid increases. 6 This would be expected to occur of the minicamp described in Table I. The identity of each for composites where the BN-interphase oxidation product is phase is delineated in Fig. 9. EDS measurements are taken to xposed to the H,o-containing environment, i.e., with a matrix determine the depth of recession. The micrograph in Fig. 9 lustrates the transition from bn to oxidized bN. The unaf crack that extends to the BN coating. The liquid could even- fected BN interphase could be discerned from a stronger nI- tually freeze to form a solid glass as the Sio, content of the lass increases, leading to stress concentrations placed on the fibers and premature composite failure. This was believed to be observed for static and cyclic tensile stress experiments per- deposit type of BN contains ted. The oxidized BN interphase is a region of glass, EDS measurements show contains oxygen and limite formed at -900C in air after -2 h 12, nitrogen. Although too small for an accurate EDS measure ment, it is likely that this region is a borosilicate glass. Because (2) Oxidation in Low-Water-Vapor Environments- Layered Sic/BN/SiC Model Compounds rom bn to glass, the interfaces are not as clean as would be The layered compounds are closer to the actual desired for recession distance measurements. this leads to and are examined in the same low-water-vapor/oxyg scatter in the data ronment. Figure 2 illustrates the preparation of this sa Recession data are shown as the filled circles on Figs three regions at which the data in Fig. 7(a(c)are take (c).9 In each case, the actual coating thickness is listed indicate oxidation effects inside a composite the filled circle. There are several important points about Figure 7(a)is-4100 um into the sample data. Note, the time dependence is not linear, but rather reces- and the thin intermediate layer of SiC are qu There is sion rate decreases with time. In fact, it appears that often the a very limited amount of oxygen. Some of th hannels seal off, ending the recession processes. This is con- to be dissolved in the bn and some appears te Sic/bn sistent with the data of Brun and Luthra. I Recession is only measured at two temperatures, but there does not seem to be a Figure 7(b)is-2000 um into the sample. Although the Bn strong dependence on temperature. There is, however, a de- is intact, there is considerably more oxygen present. The thi pendence on water-vapor content in the gas stream. This sug- intermediate layer of SiC is almost entirely oxidized, although gests the importance of various H-B-0 gaseous species it appears to have some boron on it. Thermochemically, SiC should oxidize at a low Po,. Figure 8 shows the oxygen poter (4 Oxidation in High-Water-Vapor Environments- tials set by Sio,/SiC, SiO,/SiC, BN/B,O3, and BN/B,O3 Mechanism of recession CThe underline indicates unit activity. These are calculated We did a free-energy minimization calculation to deter from the janaf thermochemical datas as follows nine the vapor species in equilibrium with B2O3 and 1% and 10% H,O/O2. This calculation indicated HBO(g), H3 BO3(g), Si+O,= SiO (3) and H, B, O(g)are the dominant species. Henceforth, we shall SiC +O2= Sio2+ C refer to these with the general term H, B, o- g) The mechanism of recession is illustrated by Fig. 11. First, An analogous set of calculations is done for BN/B,O3. These assume there is a layer of B,O, on the BN. Then, the following calculations indicate that SiC oxidizes at a lower oxygen eps tential than BN, and, when the supply of oxygen is limited (1)H,O diffuses through the boundary layer. ffectively getters the oxygen so that bn does not oxidize. This (2)H2O diffuses through the annular-shaped channel consistent with the results and model of Sheldon et al. 9 fo (3) H,O reacts with the B,Oa(e) at the base of the channel BN-coated SiC fibers to form HBO(g), H3 BO3(g), and H3.(g) Figure 7(c) is taken adjacent to the exposed face of the (4)HBO2(g), H3BO3(g), and H B.(g) diffuse out of the sample, where the oxygen potential is high. Note, there is a channel micrograph--where the oxygen potential is highest--shows (5)HBO2(g), H3 BO3 (8), and H3 B3 O(g) diffuse out substantial amount of oxide formation. The right side of the ough the boundary layer 10 um of Sio2. As discussed in the previous section, this is (6) The cross section of the annular channel gradually de- far more than would be expected from simple oxidation of pure reases because of boron-enhanced oxidation of the walls SiC. However, this can be explained by boron-enhanced oxi- shown in Fig. 11 dation of SiC to borosilicate and subsequent volatilization of In order to describe this process, we need to determine which BO from the borosilicate. which leaves behind a thick sio )2 step is rate-controlling. It is well-known that boundary-layer control would lead to a linear dependence of recession distance

weight loss is observed, which is very likely due to water￾vapor-enhanced boron volatilization from the borosilicate glass. Thus, the two types of BN behave qualitatively very similar—showing both weight gains and weight losses—but at different rates due to the different types of BN. An important implication for composites is that, as boron is removed from the borosilicate liquid, the viscosity of the bo￾rosilicate liquid increases.16 This would be expected to occur for composites where the BN-interphase oxidation product is exposed to the H2O-containing environment, i.e., with a matrix crack that extends to the BN coating. The liquid could even￾tually freeze to form a solid glass as the SiO2 content of the glass increases, leading to stress concentrations placed on the fibers and premature composite failure. This was believed to be observed for static and cyclic tensile stress experiments per￾formed at ∼900°C in air after ∼2 h.12,17 (2) Oxidation in Low-Water-Vapor Environments— Layered SiC/BN/SiC Model Compounds The layered compounds are closer to the actual composite and are examined in the same low-water-vapor/oxygen envi￾ronment. Figure 2 illustrates the preparation of this sample and three regions at which the data in Fig. 7(a)–(c) are taken. These indicate oxidation effects inside a composite. Figure 7(a) is ∼4100 mm into the sample and shows the BN and the thin intermediate layer of SiC are quite intact. There is a very limited amount of oxygen. Some of the oxygen appears to be dissolved in the BN and some appears to be at the SiC/BN interfaces. Figure 7(b) is ∼2000 mm into the sample. Although the BN is intact, there is considerably more oxygen present. The thin intermediate layer of SiC is almost entirely oxidized, although it appears to have some boron on it. Thermochemically, SiC should oxidize at a low pO2 . Figure 8 shows the oxygen poten￾tials set by SiO2/SiC, SiO2/SiC, BN/B2O3, and BN/B2O3. (The underline indicates unit activity.) These are calculated from the JANAF thermochemical data18 as follows: Si + O2 4 SiO2 (3) SiC + O2 4 SiO2 + C (4) An analogous set of calculations is done for BN/B2O3. These calculations indicate that SiC oxidizes at a lower oxygen po￾tential than BN, and, when the supply of oxygen is limited, effectively getters the oxygen so that BN does not oxidize. This is consistent with the results and model of Sheldon et al.9 for BN-coated SiC fibers. Figure 7(c) is taken adjacent to the exposed face of the sample, where the oxygen potential is high. Note, there is a substantial amount of oxide formation. The right side of the micrograph—where the oxygen potential is highest—shows ∼10 mm of SiO2. As discussed in the previous section, this is far more than would be expected from simple oxidation of pure SiC. However, this can be explained by boron-enhanced oxi￾dation of SiC to borosilicate and subsequent volatilization of B2O3 from the borosilicate, which leaves behind a thick SiO2 layer. In summary, at low oxygen contents, such as would be found deep within a composite, SiC is expected to oxidize and effec￾tively getter oxygen. At higher oxygen contents, SiC and BN oxidize simultaneously, leading to borosilicate glass formation. As time progresses, residual water vapor volatilizes the B2O3 from this glass. (3) Oxidation in High-Water-Vapor Environments— Measurements of Recession Figure 1(b) illustrates the effect of a high-water-vapor con￾tent and lower-temperature gas-oxidizing gas stream. Under these conditions, volatilization dominated, and the BN was essentially removed from the composite. This was most pro￾nounced for the BN deposited at lower temperatures. The criti￾cal parameter was the depth of recession. These recession dis￾tances were measured with single-tow, BN-coated CVD minicomposites with a range of interphase thicknesses, as de￾scribed in the experimental section. After heat treatment, samples were mounted in epoxy and longitudinally polished to determine the distance of BN recession due to volatilization.19 Figure 9 shows an example of a polished longitudinal section of the minicomposite described in Table I. The identity of each phase is delineated in Fig. 9. EDS measurements are taken to determine the depth of recession. The micrograph in Fig. 9 illustrates the transition from BN to oxidized BN. The unaf￾fected BN interphase could be discerned from a stronger ni￾trogen peak, because this type of BN contains oxygen as￾deposited. The oxidized BN interphase is a region of glass, which EDS measurements show contains oxygen and limited nitrogen. Although too small for an accurate EDS measure￾ment, it is likely that this region is a borosilicate glass. Because of the fluidity of the glass products and the gradual transition from BN to glass, the interfaces are not as clean as would be desired for recession distance measurements. This leads to scatter in the data. Recession data are shown as the filled circles on Figs. 10(a)– (c).19 In each case, the actual coating thickness is listed above the filled circle. There are several important points about these data. Note, the time dependence is not linear, but rather reces￾sion rate decreases with time. In fact, it appears that often the channels seal off, ending the recession processes. This is con￾sistent with the data of Brun and Luthra.11 Recession is only measured at two temperatures, but there does not seem to be a strong dependence on temperature. There is, however, a de￾pendence on water-vapor content in the gas stream. This sug￾gests the importance of various H-B-O gaseous species. (4) Oxidation in High-Water-Vapor Environments— Mechanism of Recession We did a free-energy minimization20 calculation to deter￾mine the vapor species in equilibrium with B2O3 and 1% and 10% H2O/O2. This calculation indicated HBO2(g), H3BO3(g), and H3B3O6(g) are the dominant species. Henceforth, we shall refer to these with the general term HxByOz(g). The mechanism of recession is illustrated by Fig. 11. First, assume there is a layer of B2O3 on the BN. Then, the following steps occur: (1) H2O diffuses through the boundary layer. (2) H2O diffuses through the annular-shaped channel. (3) H2O reacts with the B2O3(,) at the base of the channel to form HBO2(g), H3BO3(g), and H3B3O6(g). (4) HBO2(g), H3BO3(g), and H3B3O6(g) diffuse out of the channel. (5) HBO2(g), H3BO3(g), and H3B3O6(g) diffuse out through the boundary layer. (6) The cross section of the annular channel gradually de￾creases because of boron-enhanced oxidation of the walls, as shown in Fig. 11. In order to describe this process, we need to determine which step is rate-controlling. It is well-known that boundary-layer control would lead to a linear dependence of recession distance Fig. 6. TGA kinetics of CVD SiC coupon coated with 2 mm of high-deposition-temperature (∼1900°C) CVD BN and oxidized in O2/ 70 ppm H2O at 900°C. 1476 Journal of the American Ceramic Society—Jacobson et al. Vol. 82, No. 6

une 1999 High-Temperature Oxidation of Boron Nitride: Boron Nitride Layers in Composites 1477 (a 10m Si map B map N map O map m园ismp N map o map Si map B map ma O map Fig. 7. SiC/BN/SiC oxidized 140 h at 900oC in dry O2-backscattered electron image and associated elemental maps: (a)4000 um into sample

Fig. 7. SiC/BN/SiC oxidized 140 h at 900°C in dry O2—backscattered electron image and associated elemental maps: (a) 4000 mm into sample, (b) 2000 mm into sample, and (c) ∼10 mm into sample. June 1999 High-Temperature Oxidation of Boron Nitride: Boron Nitride Layers in Composites 1477

Journal of the American Ceramic SocietyJacobson et al. Vol. 82. No 6 The factor of three in the last term is due to the three boron 120011001000900 atoms in H3 B3 O(g). The diffusivities in Eqs. (5H9)above are effective diffusivities, composed of a molecular and a Knudsen 10-12 component. The molecular diffusivity of the H, B, O(g)in the BN/B2037 annular channels can be estimated using the standard Chap- Shaded area--regio man-Enskog correlation: 25 of B203 stability 0.001853742 D molec 10-20 Sic/SiO2-7 (UoH-B. O).Bo(Mo,MH,B (10) Here T is the absolute temperature, P the total pressure, and M BN/B203 he molecular weight of the subscripted species. The parameter is the average collision diameter, which is equal to the average of the molecular diameters of o, and the H-B o (g) molecules. The molecular diameters for the h, B, o(g) molecules are taken from similar molecules.26 The collision 1032 integral, OLo, -HBO, is determi tables of average inter- molecular force meters 二时 ending collision inte- als. 25,26 Tables II(A)an the calculated mole 0.0006 0.0007 0.0008 0.0009 diffusivities for each of the H, B O(g)species. At hannel diameters of0.5 and 1.0 um, it is important to consider the knudsen diffusivity contributions. The Knudsen diffusion coefficient, D,, is given by Fig 8. Calculated oxygen potentials at BN/B,O3 and SiC/SiO,(Un- derline means the activity of that element is unity. (11) on time, which is not observed. Similarly, chemical-reaction control would lead to linear rates. It is likely that diffusion in It should be noted (Il) is for a cylind not al the pores with concurrent pore closure is rate controlling annulus. However In this This process of BN volatilization is analogous to carbon, T, and M as given above. The two diffusivities are combined meter with the same fibers:e 21-23 BN volatilization via B2O, to H, B, o (g)is some- ere are several treatments of this process in the lit what simpler, because the formation of the volatile species is a (12) one-step process, as opposed to carbon volatilization, which is Volatilization of BN occurs via the three vapor species dis- Tables II(A)and(B)list the Knudsen and effective diffusivities cussed above. First consider the flux due to each species: The complication is that these volatile species do not come HxByO-HrB, O- HrBy directly from the bn, but rather as shown in Fig. 11. So the reactions are This is simply Fick's first law, with the effective diffusivity BN +=O2(g)=B2O3()+n2(g) (13a) the pressure gradient through the pore VBN the molar vol- B2O3+H,O=2HBO(g) (13b) f Bn. and p apor pressure of H, B, o(g) Now we can convert Eq. (5)to a recession distance for B2O3+3H,0=2H, BO3(g) (13c) each H, B, O- specie dyH,B E=-J/H-B, O_ BN=BN/9 dp 3B2O3+3H2O=2H3B3O6g) (13d dt rt dyH-,o (6) Note that B,O is at less than unit activity. Recent data on the B,O3-SiO2 system indicate that the activities can be approxi We can approximate the pressure gradient as mated as ideal. However, we do not know the composition of the borosilicate melt, and, furthermore, it changes as the dPH,B, o, APH, B, 0: PH,, O reaction progresses. However, as an approximation, we shall assume the mole fraction of B, O3 to be 0.5 and take the activ ty of B,, to be 0.5. Table Ill lists the vapor pressures of Here PH, B o, is the pressure of the particular H, B,O-g)spe- HBO2 (g), H3 BO3(g), and H3 B3 O(g)in equilibrium with cies and yH. bo the recession distance due to this species. Now B2O(e)with an activity of 0.5. These pressures are used in the 2V PHBon Equation( 8)indicates that recession rates are parabolic with RI (8) time. However, our data suggest the annular channels seal as time progresses and limit the outward flux. This is probably The total recession rate of BN, which was measured, is simply due to oxidation of the annular channel walls the sum of the recession rates due to each H, B, o(g)species The growth of SiO, on SiC is best described by a linear- parabolic equation from Deal and Grove yg Rr(HBO2HBO2*DHaBO3H3BO3 x2+Ax= B(t+T) (14) 3B306 H3 B306 Here B is the parabolic rate constant, B/A the linear rate con-

on time, which is not observed. Similarly, chemical-reaction control would lead to linear rates. It is likely that diffusion in the pores with concurrent pore closure is rate controlling. This process of BN volatilization is analogous to carbon￾interphase burnout, observed in composites with carbon-coated fibers. There are several treatments of this process in the lit￾erature.21–23 BN volatilization via B2O3 to HxByOz(g) is some￾what simpler, because the formation of the volatile species is a one-step process, as opposed to carbon volatilization, which is a two-step process.21–23 Volatilization of BN occurs via the three vapor species dis￾cussed above. First consider the flux due to each species: JHxByOz = −DHxByOz dcHxByOz dyHxByOz = −DHxByOz RT dPHxByOz dyHxByOz (5) This is simply Fick’s first law, with the effective diffusivity DHxByOz calculated below. J is the flux of HxByOz(g), R the gas constant, y the recession distance, dPH xBy Oz /dyH xBy Oz the pressure gradient through the pore, VBN the molar vol￾ume of BN, and PHxByOz the vapor pressure of HxByOz(g). Now we can convert Eq. (5) to a recession distance for each HxByOz species:24 dyHxByOz dt = −JHxByOz VBN = VBNS DHxByOz RT dPHxByOz dyHxByOz D (6) We can approximate the pressure gradient as dPHxByOz dyHxByOz = DPHxByOz DyHxByOz ≈ PHxByOz yHxByOz (7) Here PHxByOz is the pressure of the particular HxByOz(g) spe￾cies and yHxByOz the recession distance due to this species. Now we can integrate Eq. (7): yH xByOz 2 = 2VBN~DHxByOz PHxByOz t! RT = kHxByOz t (8) The total recession rate of BN, which was measured, is simply the sum of the recession rates due to each HxByOz(g) species: yB = 2 VBNt RT ~DHBO2 PHBO2 + DH3BO3 PH3BO3 + 3DH3B3O6 PH3B3O6 ! (9) The factor of three in the last term is due to the three boron atoms in H3B3O6(g). The diffusivities in Eqs. (5)–(9) above are effective diffusivities, composed of a molecular and a Knudsen component. The molecular diffusivity of the HxByOy(g) in the annular channels can be estimated using the standard Chap￾man–Enskog correlation:25 Dmolec = 0.001853T1/2 P~sO2–HxByOz ! 2 VO2–HxByOz S 1 MO2 + 1 MHxByOz D (10) Here T is the absolute temperature, P the total pressure, and M the molecular weight of the subscripted species. The parameter sO2–HxByOz is the average collision diameter, which is equal to the average of the molecular diameters of O2 and the HxBy￾Oz(g) molecules. The molecular diameters for the HxByOz(g) molecules are taken from similar molecules.26 The collision integral, VO2–HBO2 , is determined from tables of average inter￾molecular force parameters and corresponding collision inte￾grals.25,26 Tables II(A) and (B) list the calculated molecular diffusivities for each of the important HxByOz(g) species. At channel diameters of 0.5 and 1.0 mm, it is important to consider the Knudsen diffusivity contributions. The Knudsen diffusion coefficient, Dk, is given by25 Dk = 2 3 d S 8RT pM D 1/2 (11) It should be noted that Eq. (11) is for a cylindrical pore, not an annulus. However, for this approximation, it is adequate. In this equation, d is the pore diameter with the same definitions for R, T, and M as given above. The two diffusivities are combined as22 1 DHxByOz = 1 Dk + 1 Dmolec (12) Tables II(A) and (B) list the Knudsen and effective diffusivities calculated for the experimental conditions. The complication is that these volatile species do not come directly from the BN, but rather via a borosilicate intermediate, as shown in Fig. 11. So the reactions are BN + 3 2 O2~g! = B2O3~l! + N2~g! (13a) B2O3 + H2O = 2HBO2~g! (13b) B2O3 + 3H2O = 2H3BO3~g! (13c) 3B2O3 + 3H2O = 2H3B3O6~g! (13d) Note that B2O3 is at less than unit activity. Recent data on the B2O3–SiO2 system14 indicate that the activities can be approxi￾mated as ideal. However, we do not know the composition of the borosilicate melt, and, furthermore, it changes as the reaction progresses. However, as an approximation, we shall assume the mole fraction of B2O3 to be 0.5 and take the activ￾ity of B2O3 to be 0.5. Table III lists the vapor pressures of HBO2(g), H3BO3(g), and H3B3O6(g) in equilibrium with B2O3(,) with an activity of 0.5. These pressures are used in the flux calculations. Equation (8) indicates that recession rates are parabolic with time. However, our data suggest the annular channels seal as time progresses and limit the outward flux. This is probably due to oxidation of the annular channel walls. The growth of SiO2 on SiC is best described by a linear– parabolic equation from Deal and Grove:27 x2 + Ax 4 B(t + t) (14) Here B is the parabolic rate constant, B/A the linear rate con￾Fig. 8. Calculated oxygen potentials at BN/B2O3 and SiC/SiO2. (Un￾derline means the activity of that element is unity.) 1478 Journal of the American Ceramic Society—Jacobson et al. Vol. 82, No. 6

une 1999 High-Temperature Oxidation of Boron Nitride: Boron Nitride Layers in Composites . Matri Matrix um Fig. 9. Polished longitudinal cross section and associated EDS data for minicomposite to show recession depth after exposure to 10% H,O/O, for 0 h at 700C. Transition from BN to a borosilicate glass is gradual. Also note that this type of as-deposited BN contains some dissolved oxygen. stant, t the time, and T a rrection for very thin oxide Here, d is the width of the annular region. For this approxima layers. For this approximat nore T. At these low tem- tion we have taken the oxidation rate of the fiber and the atures, the linear rate atrix to be equal. At 700 and 800oC, the rate constants SiC are extreme small. The 800%C value is fro ture,2 the 700oC value is extrapolated from this8 and four A(A-+4Bn other higher-temperature measurements. 5 All measurements (15) are in pure oxygen. The lowest linear and parabolic rate con- stants used in Eq(12)are given in Table IV As an approximation, the reduction in area of the annulus n the composite situation, water vapor in the gas stream27, fiber is give and residual boron are expected to enhance the oxidation of the channel walls. Because we do not know exactly how much he parabolic rate constant increases in this situation, we shall treat the parabolic rate constant, B, as an adjustable variable 2-2 and vary it from I to 5000 times its lowest value(Table IV until a good fit to the data is attained. The parameter A is not Note that is a function of time, because x is a function of time Table V(A)lists some calculated times to seal Here, Im is the radius of the matrix opening and rr the radius of nnular regions at 700C, and Table V(B) the fiber(Fig. 11). Also note that x in Eq. (14)is for SiO2 times for a 1 um annular region at 800C whereas x'in Eq. (14)is an x modified to reflect the molar ed oxidation is important in explaining the volume change on conversion from SiC to SiOz. This is sealing gradual cosy o to BN recession roughly a factor of two. Now we can simplify Eq(15)as of the annular channel effectively de- creases the diffusive r=r+d

stant, t the time, and t a time correction for very thin oxide layers. For this approximation, we ignore t. At these low tem￾peratures, the linear rate constant is important. Solving for x gives x = −A 2 + ~A2 + 4Bt! 1/2 2 (15) As an approximation, the reduction in area of the annulus around the fiber is given by f = ~rm − x8! 2 − ~rf + x8! 2 rm 2 − rf 2 (16) Note that f is a function of time, because x is a function of time. Here, rm is the radius of the matrix opening and rf the radius of the fiber (Fig. 11). Also note that x in Eq. (14) is for SiO2, whereas x8 in Eq. (14) is an x modified to reflect the molar volume change on conversion from SiC to SiO2. This is roughly a factor of two. Now we can simplify Eq. (15) as follows: rm = rf + d f = 1 − 2x8 d (17) Here, d is the width of the annular region. For this approxima￾tion, we have taken the oxidation rate of the fiber and the matrix to be equal. At 700° and 800°C, the rate constants for SiC are extremely small. The 800°C value is from the litera￾ture;28 the 700°C value is extrapolated from this28 and four other higher-temperature measurements.15 All measurements are in pure oxygen. The lowest linear and parabolic rate con￾stants used in Eq. (12) are given in Table IV. In the composite situation, water vapor in the gas stream27,29 and residual boron5 are expected to enhance the oxidation of the channel walls. Because we do not know exactly how much the parabolic rate constant increases in this situation, we shall treat the parabolic rate constant, B, as an adjustable variable and vary it from 1 to 5000 times its lowest value (Table IV) until a good fit to the data is attained. The parameter A is not varied. Table V(A) lists some calculated times to seal 1.0 and 0.5 mm annular regions at 700°C, and Table V(B) lists the sealing times for a 1 mm annular region at 800°C. Clearly, enhanced oxidation is important in explaining the observed sealing. This gradual closing of the annular channel effectively de￾creases the diffusive flux that leads to BN recession. dyHxByOz dt = − f JHxByOz VBN = f VBN S DHxByOz RT dPHxByOz dyHxByOz D (18) Fig. 9. Polished longitudinal cross section and associated EDS data for minicomposite to show recession depth after exposure to 10% H2O/O2 for 20 h at 700°C. Transition from BN to a borosilicate glass is gradual. Also note that this type of as-deposited BN contains some dissolved oxygen. June 1999 High-Temperature Oxidation of Boron Nitride: Boron Nitride Layers in Composites 1479

Journal of the American Ceramic Societyacobson et al. Vol. 82. No 6 3c0 23x103m2 55 53 HxByozlg 2 SiC Intact Fig. 11. Schematic of volatilization processes for exposed BN Table ll. Calculated Molecular, Knudsen, and Effective Diffusivities for Various HB, O(g)Species Knudsen diffusivity Effective diffusivity H 4.6 3.8 Time(hrs) 1200 cm/s), I H (c) 1000 H,BO3(g) 4.0 H3.(g) 2.5 0.64 Sealing K=3.7x 10"um/r Table Ill. Calculated Vapor Pressures above B,O3 with a(B2O3)=0.5 0.92 Pressure at700°cbar essure at8o0°cbar) 0% HOO, 1%H2O/O2 % HOO HBO 3.6×10-61.1×10-6 H3B03(g) 3.8×10 H3B3O6(g)2.8×10 88×10-6 This expression can be integrated to Time(hrs 2aND1BP1B2(,4_(F2+4B d 12dB ured interphase thickness shown. All BN interphase -1000oC:(a)-05 um BN interphase, 700C in 10%H,O/O2,(b)-1 The full expression for recession distance is then the sum of the BN interphase, 700C in 1% H2O/O2, and(c)-I um contributions from each of the three vapor species, analogous phase, 800oC in 1% H,O/O to Eq(9)

This expression can be integrated to yHxByOz 2 = 2VBN DHxByOz PHxByOz RT St + At d − ~A2 + 4Bt! 3/2 12dB D (19) The full expression for recession distance is then the sum of the contributions from each of the three vapor species, analogous to Eq. (9): Fig. 10. Results of pore recession calculations (–––) showing para￾bolic rate constants used and comparison to experiment (d) with mea￾sured interphase thickness shown. All BN interphases deposited at ∼1000°C: (a) ∼0.5 mm BN interphase, 700°C in 10% H2O/O2, (b) ∼1 mm BN interphase, 700°C in 1% H2O/O2, and (c) ∼1 mm BN inter￾phase, 800°C in 1% H2O/O2. Fig. 11. Schematic of volatilization processes for exposed BN. Table II. Calculated Molecular, Knudsen, and Effective Diffusivities for Various HxByOz(g) Species (A) 700°C Species Molecular diffusivity in O2 (cm2 /s) Knudsen diffusivity (cm2 /s) Effective diffusivity (cm2 /s) 0.5 mm 1 mm 0.5 mm 1 mm HBO2(g) 0.90 2.3 4.6 0.64 0.75 H3BO3(g) 0.80 1.9 3.8 0.56 0.66 H3B3O6(g) 0.72 1.3 2.6 0.46 0.56 (B) 800°C Species Molecular diffusivity in O2 (cm2 /s) Knudsen diffusivity (cm2 /s), 1 mm Effective diffusivity (cm2 /s), 1 mm HBO2(g) 1.1 4.8 0.90 H3BO3(g) 0.95 4.0 0.77 H3B3O6(g) 0.86 2.5 0.64 Table III. Calculated Vapor Pressures above B2O3 with a(B2O3) = 0.5 Species Pressure at 700°C (bar)† Pressure at 800°C (bar)† 10% H2O/O2 1% H2O/O2 1% H2O/O2 HBO2(g) 3.6 × 10−6 1.1 × 10−6 8.6 × 10−6 H3BO3(g) 1.2 × 10−4 3.8 × 10−6 3.0 × 10−6 H3B3O6(g) 2.8 × 10−4 8.8 × 10−6 5.7 × 10−6 † 1 bar 4 105 Pa. 1480 Journal of the American Ceramic Society—Jacobson et al. Vol. 82, No. 6

une 1999 High-Temperature Oxidation of Boron Nitride: Boron Nitride Layers in Composites Table IV. Lowest Possible Rate Constants for Oxidation of This leads to borosilicate glass formation. As time progresses, the small (-20 ppm) amounts of water vapor remove boron B(um/h) /A(um/h) from the glass, leading to a substantially larger amount of Sio 4.5×10-6 85×10-4 uld be expected from simple SiC oxidation. Experi- 3.7×10 4.2×10-3 ments on SiC/BN/SiC layered structures allow examination of low-water-vapor oxidation effects deep within the composite In such cases, the supply of oxygen is limited, and SiC oxidizes preferentially to BN. As oxygen potential increases, borosil- Table v. Calculated Times to Seal Annular Region cate formation becomes more important The second mechanism involves volatilization of bn by Rate constant high-water-vapor gas streams. This is due to the high stabil- the HBO(g), H3BO3(g), and H3,O(g) molecules 4.5×I 5.57×104h 2.21×105h olatilization leads to recession of fiber coatings in a composite material. This process is described with gas- 2.21×103h diffusion through a channel and concurrent channel 4.5 55.7h closure (B)800°C Late constant 1.0um Acknowledgments: We wish to thank Dr. David Harding, formerly of the RBS work. We also wish to thank J Smith, NYMA, NASA Lewis Gr for the electron microprobe work. We appreciate Dr. M. K. Brun of Gener 3.7×10-3 227h Electric for providing BN-coated SiC, and Drs. A. Moore and H. Sayir of 3.7×10 ounds. Helpful dis- ussions with Drs. K. L. Luthra and P. Meschter of General Electric Company and Professor B. Sheldon of Brown University are appreciated Refe =2102m:m+01+1s最如R和mHA and SIMS in At(42+4B32 K N. Lee and N S. Jacobson, "Chemical Stability of the Fiber Coating/ This full expression is used, with B as an adjustable parameter, -Based Ceramic Matrix Composites, "J. Am. Ceram. to calculate the dashed lines in Figs. 10(aHc)and compare R. S. Roth, J. R. Dennis, and H. F. McMurdie, Fig 6455 in Phase Diagrams them to the measured distances for Ceramists, Vol. VI. Edited by M. A. Clevinger and H. M. Ondik. American In Figs. 10(aHc), the dashed lines are from Eq (16), and the olid circles are the experimental measurements. The calcula- tions reproduce many of the experimental observations in this data set-the decrease in recession with time, the weak Sic Composite, "J.Am. Ceram. Soc, 81(112777-84(1998) temperature dependence, and strong PH,o dependence. At 700C with 10% H2O/O2 and 0.5 um BN(Fig. 10(a)). ment on the Subcritical Crack the largest recession is observed due to the 10% H,O. A 500x posite, m Eng. Sci. Proc., 13 [ 7-8] N.S. Jacobson, S. Farmer, A Moore, and H. Sayir, "High-Temperature enhanced oxidation rate leads to the observed annulus closure ation Behavior of Boron Nitride: I. Monolithic Boron Nitride. " J. Am. At 700C with 1% H,O/O, and I um BN(Fig. 10(b),reces- m.Soc,8212393-98(1999 sion is less due to the lower-water-vapor content. Now a 5000x B. W. Sheldon, E. Y. Sun, S.R. Nutt, and JJ. Brennan, "Oxidation of enhanced oxidation rate constant approximates channel clo- N-Coated SiC Fibers in Ceramic-Matrix Composites, J Am Ceram Soc., 79 sure, very likely due to larger annulus width. Finally at 800C N. Morscher, D. Bryant, and R, E. Tressler, "Environmental Durability little less than that at 700C, possibly due to more rapid char vith 1% H,O/O, and I um BN (Fig. 10(c), recession seems of Different BN Interphases( for SiC/SiC)in H,O Containing Atmospheres at nel sealing at this temperature. A 1000x enhanced oxidation appears to account for channel sealing. These large enhance 加 Ceramic Society, Cincinnati. H. a gth Annual Meeting of the A ments in oxidation rates are expected because of the presence ites Symposium, Paper No. SIII-027-97) of boron and water vapor. 27 2G. N. Morscher, " Tensile Stress-Rupture of SiC /SiCm Mi Although this model describes many of the basic observa- Not, so ng 2029 t2 (19,. terphases at Elevated Temperatures in Air, J.A. Ceram. cies out of the glass are an important step. We know oxidation Melts at 1475 K, "J. Am. Ceram. Soc., 76[11] 2809-12(1993 ISL. U. J. T Ogbuji and E J. Opila, "A C of the Oxidation Kinetics kinetics differ for BN deposited at different temperatures, and 30(1995) these kinetics should be included N. P. Bansal and R H. Doremus, Handbook of Glass Properties, pp. 242- 43. Academic Press, New York, 1986. IV. Conclusions perature on Failure for Precracked Hi-Nicalon/BN/CVD SiC Minicomposites in M.wbP18|519 C A. Davies, J.R. Downey Jr, D J. Frurip,R. A Mac. Donald, and A N. Syverud, JANAF Thermochemical Tables, 3rd ed; pp 24 iber coating in SiC/SiC composites with a series of model 270,633, 6732 American Chemical Society and American Physical Society opposites. Two important reaction processes have been ex- PD. R. Bryant, "Oxidation and volatiliz plored: (1)oxidation in low-water-vapor streams, which pro- motes borosilicate glass formation, and(2)oxidation in high- versity, Ur water-vapor streams, where the bN is volatilized. Experiments on a BN-coated SiC coupon indicate the BN readily oxidizes to B2O3 simultaneously with enhanced SiC oxidation to SiO2 个队:M段Sm lex Chemical Equilibria," Metall. B,21B,1013-23 zzi and R. Naslain, "Oxidation Mechanisms and Kinetics of lD

yB 2 = 2VBN ~DHBO2 PHBO2 +DH2BO3 PH3BO3 +DH3B3O6 PH3B3O6 ! RT × St + At d − ~A2 + 4Bt! 3/2 12dB D (20) This full expression is used, with B as an adjustable parameter, to calculate the dashed lines in Figs. 10(a)–(c) and compare them to the measured distances. In Figs. 10(a)–(c), the dashed lines are from Eq. (16), and the solid circles are the experimental measurements. The calcula￾tions reproduce many of the experimental observations in this data set—the decrease in recession with time, the weak temperature dependence, and strong pH2O dependence. At 700°C with 10% H2O/O2 and 0.5 mm BN (Fig. 10(a)), the largest recession is observed due to the 10% H2O. A 500× enhanced oxidation rate leads to the observed annulus closure. At 700°C with 1% H2O/O2 and 1 mm BN (Fig. 10(b)), reces￾sion is less due to the lower-water-vapor content. Now a 5000× enhanced oxidation rate constant approximates channel clo￾sure, very likely due to larger annulus width. Finally at 800°C with 1% H2O/O2 and 1 mm BN (Fig. 10(c)), recession seems a little less than that at 700°C, possibly due to more rapid chan￾nel sealing at this temperature. A 1000× enhanced oxidation appears to account for channel sealing. These large enhance￾ments in oxidation rates are expected because of the presence of boron5 and water vapor.27 Although this model describes many of the basic observa￾tions, many refinements could be made. Channel sealing is a critical issue and could be treated in more detail. The liquid borosilicate products are important, and diffusion of boron spe￾cies out of the glass are an important step. We know oxidation kinetics differ for BN deposited at different temperatures,8 and these kinetics should be included. IV. Conclusions We have examined the high-temperature oxidation of the BN fiber coating in SiC/SiC composites with a series of model composites. Two important reaction processes have been ex￾plored: (1) oxidation in low-water-vapor streams, which pro￾motes borosilicate glass formation, and (2) oxidation in high￾water-vapor streams, where the BN is volatilized. Experiments on a BN-coated SiC coupon indicate the BN readily oxidizes to B2O3 simultaneously with enhanced SiC oxidation to SiO2. This leads to borosilicate glass formation. As time progresses, the small (∼20 ppm) amounts of water vapor remove boron from the glass, leading to a substantially larger amount of SiO2 than would be expected from simple SiC oxidation. Experi￾ments on SiC/BN/SiC layered structures allow examination of low-water-vapor oxidation effects deep within the composite. In such cases, the supply of oxygen is limited, and SiC oxidizes preferentially to BN. As oxygen potential increases, borosili￾cate formation becomes more important. The second mechanism involves volatilization of BN by high-water-vapor gas streams. This is due to the high stabil￾ity of the HBO2(g), H3BO3(g), and H3B3O6(g) molecules. This volatilization leads to recession of fiber coatings in a model composite material. This process is described with gas￾phase diffusion through a channel and concurrent channel closure. Acknowledgments: We wish to thank Dr. David Harding, formerly of NYMA, NASA Lewis Group, and currently with SUNY Buffalo, for facilitating the RBS work. We also wish to thank J. Smith, NYMA, NASA Lewis Group, for the electron microprobe work. We appreciate Dr. M. K. Brun of General Electric for providing BN-coated SiC, and Drs. A. Moore and H. Sayir of Advanced Ceramic Corporation for providing model compounds. Helpful dis￾cussions with Drs. K. L. Luthra and P. Meschter of General Electric Company and Professor B. Sheldon of Brown University are appreciated. References 1 O. Dugne, S. Pouhet, A. Guette, R. Naslain, R. Fourmeaux, Y. Khin, J. Sevely, J. P. Rocher, and J. Cotteret, “Interface Characterization by TEM, AES, and SIMS in Tough SiC (ex-PCS) Fibre-SiC (CVI) Matrix Composites with a BN Interphase,” J. Mater. Sci., 28, 3409–22 (1993). 2 R. Naslain, “Fibre–Matrix Interphases and Interfaces in Ceramic Matrix Composites Processed by CVI,” Compos. Interfaces, 1 [3] 253–86 (1993). 3 K. N. Lee and N. S. Jacobson, “Chemical Stability of the Fiber Coating/ Matrix Interface in Silicon-Based Ceramic Matrix Composites,” J. Am. Ceram. Soc., 78 [3] 711–15 (1994). 4 R. S. Roth, J. R. Dennis, and H. F. McMurdie; Fig. 6455 in Phase Diagrams for Ceramists, Vol. VI. Edited by M. A. Clevinger and H. M. Ondik. American Ceramic Society, Columbus, OH, 1987. 5 J. Schlicting, “Oxygen Transport Through Glass Layers Formed by a Gel Process,” J. Non-Cryst. Solids, 63, 173–81 (1984). 6 L. U. J. T. Ogbuji, “A Pervasive Mode of Oxidative Degradation in a SiC– SiC Composite,” J. Am. Ceram. Soc., 81 [11] 2777–84 (1998). 7 C. H. Henager Jr. and R. H. Jones, “The Effects of an Aggressive Environ￾ment on the Subcritical Crack Growth of a Continuous-Fiber Ceramic Com￾posite,” Ceram. Eng. Sci. Proc., 13 [7–8] 411–19 (1992). 8 N. S. Jacobson, S. Farmer, A. Moore, and H. Sayir, “High-Temperature Oxidation Behavior of Boron Nitride: I, Monolithic Boron Nitride,” J. Am. Ceram. Soc., 82 [2] 393–98 (1999). 9 B. W. Sheldon, E. Y. Sun, S. R. Nutt, and J. J. Brennan, “Oxidation of BN-Coated SiC Fibers in Ceramic-Matrix Composites,” J. Am. Ceram. Soc., 79 [2] 539–43 (1996). 10G. N. Morscher, D. Bryant, and R. E. Tressler, “Environmental Durability of Different BN Interphases (for SiC/SiC) in H2O Containing Atmospheres at Intermediate Temperatures,” Ceram. Eng. Sci. Proc., 18 [3] 525–34 (1997). 11M. K. Brun and K. L. Luthra, “Ends-on Oxidation of BN Fiber Coatings in SiC–Si/SiC Composites”; presented at the 99th Annual Meeting of the Ameri￾can Ceramic Society, Cincinnati, OH, May 6, 1997 (Ceramic Matrix Compos￾ites Symposium, Paper No. SIII-027-97). 12G. N. Morscher, “Tensile Stress–Rupture of SiCf /SiCm Minicomposites with C and BN Interphases at Elevated Temperatures in Air,” J. Am. Ceram. Soc., 80 [8] 2029–42 (1997). 13L. R. Doolittle, “Algorithms for the Rapid Simulation of Rutherford Back￾scattering Spectra,” Nucl. Instrum. Methods Phys. Res., B9 [3] 334–51 (1985). 14M. Boike, K. Hilpert, and F. Muller, “Chemical Activities in B2O3–SiO2 Melts at 1475 K,” J. Am. Ceram. Soc., 76 [11] 2809–12 (1993). 15L. U. J. T. Ogbuji and E. J. Opila, “A Comparison of the Oxidation Kinetics of SiC and Si3N4,” J. Electrochem. Soc., 142 [3] 925–30 (1995). 16N. P. Bansal and R. H. Doremus, Handbook of Glass Properties; pp. 242– 43. Academic Press, New York, 1986. 17G. N. Morscher, “The Effect of Static and Cyclic Tensile Stress and Tem￾perature on Failure for Precracked Hi-Nicalon/BN/CVD SiC Minicomposites in Air,” Ceram. Eng. Sci. Proc., 18 [3] 737–45 (1997). 18M. W. Chase Jr., C. A. Davies, J. R. Downey Jr., D. J. Frurip, R. A. Mac￾Donald, and A. N. Syverud, JANAF Thermochemical Tables, 3rd ed; pp. 246, 270, 633, 1673. American Chemical Society and American Physical Society, New York, 1986. 19D. R. Bryant, “Oxidation and Volatilization of BN Interphases in SiC Fi￾ber-Reinforced SiC Matrix Composites”; M.S. Thesis. Pennsylvania State Uni￾versity, University Park, PA, 1997. 20G. Eriksson and K. Hack, “ChemSage—A Computer Program for the Cal￾culation of Complex Chemical Equilibria,” Metall. Trans. B, 21B, 1013–23 (1990). 21L. Filipuzzi and R. Naslain, “Oxidation Mechanisms and Kinetics of 1D￾Table IV. Lowest Possible Rate Constants for Oxidation of Pore Walls Temperature (°C) B (mm2 /h) B/A (mm/h) 700 4.5 × 10−6 8.5 × 10−4 800 3.7 × 10−5 4.2 × 10−3 Table V. Calculated Times to Seal Annular Region (A) 700°C Rate constant (mm2 /h) 0.5 mm 1.0 mm 4.5 × 10−6 5.57 × 104 h 2.21 × 105 h 4.5 × 10−5 5.57 × 103 h 2.21 × 104 h 4.5 × 10−4 557 h 2.21 × 103 h 4.5 × 10−3 55.7 h 221 h (B) 800°C Rate constant (mm2 /h) 1.0 mm 3.7 × 10−5 2.27 × 104 h 3.7 × 10−4 2.27 × 103 h 3.7 × 10−3 227 h 3.7 × 10−2 22.7 h June 1999 High-Temperature Oxidation of Boron Nitride: Boron Nitride Layers in Composites 1481

Journal of the American Ceramic Sociery-Jacobson et al. VoL. 82. No 6 SiC/C/SiC Composite Materials: IL, Modeling, "J. Am. Ceram. Soc., 77[2 2G. H. Geiger and D. R. Poirier, Transport Phenom A J. Eckel, J. D. Cawley, and T. A, Parthasarathy, "Oxidation Kinetics 26R. A Svehla, "Estimated Viscosities and Thermal Conductivi a Continuous Carbon Phase in a Nonreactive Matrix, "J. Am. Ceram. Soc., 78 at High Temperatures, "NASA Tech Rept. No R-132. NASA Le wis research 41972-80(1995) disch Jr, C. H. Henager Jr, G. D. Springer, and R, H. Jones, 27B. E Deal and A.S. Grove, "General Relationship for the Thermal Oxi- "Oxidation Kinetics of the Carbon-Interface in Nicalon -Fiber- Reinforced Sili- dation of Silicon, "J. Appl. Phys., 36 [ 12]3370-78(1963 posites.:”J.Am. Ceram.Soc.,80目56973(1997). 2C. E. Ramberg, G. Cruciani, K. E. Spear, R. E. Tressler, and C. F Rambe jacobson, Oxidation and Corrosion of Silicon-Based Ceram ogress in Ceramic Bo hallenge Toward the 21st Centar. Edited by T. Hirai, S. Hirano, and lly Vapor-Deposited Silicon Ceramic Society of Japan, To Carbide in Wet Oxygen, J. Am. Ce 7713]730-36(1994)

SiC/C/SiC Composite Materials: II, Modeling,” J. Am. Ceram. Soc., 77 [2] 467–80 (1994). 22A. J. Eckel, J. D. Cawley, and T. A. Parthasarathy, “Oxidation Kinetics of a Continuous Carbon Phase in a Nonreactive Matrix,” J. Am. Ceram. Soc., 78 [4] 972–80 (1995). 23C. F. Windisch Jr., C. H. Henager Jr., G. D. Springer, and R. H. Jones, “Oxidation Kinetics of the Carbon–Interface in Nicalon-Fiber-Reinforced Sili￾con Carbide Composites,” J. Am. Ceram. Soc., 80 [3] 569–73 (1997). 24N. S. Jacobson, “Oxidation and Corrosion of Silicon-Based Ceramics— Challenges and Critical Issues”; pp. 87–104 in Progress in Ceramic Basic Science—Challenge Toward the 21st Century. Edited by T. Hirai, S. Hirano, and Y. Takeda. Ceramic Society of Japan, Tokyo, Japan, 1996. 25G. H. Geiger and D. R. Poirier, Transport Phenomena in Metallurgy; p. 464. Addison-Wesley, Reading, MA, 1973. 26R. A. Svehla, “Estimated Viscosities and Thermal Conductivities of Gases at High Temperatures,” NASA Tech. Rept. No. R-132. NASA Lewis Research Center, Cleveland, OH, 1961. 27B. E. Deal and A. S. Grove, “General Relationship for the Thermal Oxi￾dation of Silicon,” J. Appl. Phys., 36 [12] 3370–78 (1965). 28C. E. Ramberg, G. Cruciani, K. E. Spear, R. E. Tressler, and C. F. Ramberg Jr., “Passive-Oxidation Kinetics of High-Purity Silicon Carbide from 800° to 1100°C,” J. Am. Ceram. Soc., 79 [11] 2897–911 (1996). 29E. J. Opila, “Oxidation Kinetics of Chemically Vapor-Deposited Silicon Carbide in Wet Oxygen,” J. Am. Ceram. Soc., 77 [3] 730–36 (1994). h 1482 Journal of the American Ceramic Society—Jacobson et al. Vol. 82, No. 6