Oxidation of SiC powder by high-temperature, high-pressure H2O

Oxidation of SiC powder by high-temperature, high-pressure H2O Masahiro Yoshimura, Jun-ichiro Kase, and Shigeyuki Somiya Research Laboratory of Engineering Materials and Department of Materials Science and Engineering, Tokyo Institute of Technology, 4259 Nagatsuta, Midori, Yokohama, 227 Japan (Received 10 October 1985; accepted 6 January 1986) The reaction between SiC powder and H 2 O has been studied at 400°-800 °C under 10 and 100 MPa. Silicon carbide reacted with H 2 O to yield amorphous SiO2 and CH 4 by the reaction SiC + 2H 2 O^SiO 2 + CH 4 above 500 °C. Cristobalite and tridymite crystallized from amorphous silica after the almost complete oxidation of SiC above 700 °C. The oxidation rate, as calculated from the weight gain, increased with temperature and pressure. The Arrhenius plotting of the reaction rate based on a Jander-type model gave apparent activation energies of 167-194 kJ/mol. Contrasted with oxidation in oxidative atmosphere, oxidation in H 2 O is characterized by the diffusion of H 2 O and CH 4 in an amorphous silica layer where the diffusion seemed to be rate determining. Present results suggest that the oxidation of SiC includes the diffusion process of H 2 O in silica layers when atmospheres contain water vapor.

I. INTRODUCTION Silicon carbide, as well as silicon nitride, has attracted great attention in recent years because of its potential application as one of high-temperature structural ceramics. Oxidation resistance is one of the most important properties in these applications. As for silicon carbide powders, the oxidation behavior, investigated for many samples in various oxidizing atmospheres, indicates a parabolic kinetics.1"7 The reaction is believed to be as follows: SiC

CO2

Suzuki3 and Jorgensen etal.1 studied the effects of water vapor on the oxidation of silicon carbide powders. Singhal8 also examined the effects of water vapor on the oxidation of hot-pressed silicon carbide. They argued that the oxidation rate was accelerated when water vapor was added to oxygen. However, the reaction of SiC by only H 2 O has not been studied yet. We have studied the reactions between SiC powder and H 2 O under hydrothermal conditions where almost only high-temperature, high-pressure H 2 O exists. Furthermore, we will discuss the effect of H 2 O on the oxidation of SiC in comparison with a similar study on Si3N4 reported previously.9

400° to 800 °C, and time for 0 to 72 h. After the treatment the pressure vessel was quenched, then the sample tube was dried and weighed to calculate the oxidation ratio of SiC. Solid products were examined by x-ray diffractometry (XRD) and scanning electron microscopy (SEM). The oxidation ratio was also checked by XRD through the decrease of the SiC phase. Gaseous products in the vessel were analyzed by a mass spectrometer. III. RESULTS Under 100 MPa H 2 O, the oxidation of SiC was not observed at 400 °C but proceeded above 500 °C as a function of time (Fig. 1). Weight gain and appearance of a halo in x-ray diffraction indicated that amorphous silica was formed by the oxidation of SiC as the first step at every temperature above 500 °C. Cristobalite and tridymite crystallized above 700 °C after 8 h where SiC was almost completely changed into amorphous silica. Scanning electron micrographs showed that the surfaces of the SiC particles were covered with amorphous

II. EXPERIMENTAL PROCEDURE The starting material was SiC powder (Starck A10, average grain size 0.65 /um, 94% a form). The sample powder was placed in an Au tube (2.7 mm i.d., 0.15 mm thick, and 35 mm long) of which the top was flattened and bent by pressing to prevent the escape of solid materials. The tube was heated in a test-tube-type pressure vessel under pressures of 10 or 100 MPa of redistilled water. In the treatment the temperature was varied from 100

J. Mater. Res. 1 (1), Jan/Feb 1986

3

8

1(00•c

24 (h)

Tine

700 *C

500 •c

•

600•c

o 0

800-C

Sol id 11ne

Equation (2)

Crystalline S1O2

FIG. 1. Isothermal oxidation of SiC powder under 100 MPa H 2 O.

0003-6951 /86/010100-04$01.75

© 1986 Materials Research Society

Yoshimura, Kase, and Somiya: Oxidation of SiC powder

01 3

24

Time

700 "C

500-C

72 ( h )

S o l i d l i n e : Equation (2)

300'C

600-C o

C r y s t a l l i n e SIO2

FIG. 2. Isothermal oxidation of SiC powder under 10 MPa H 2 O.

silica by the oxidation and these silica grains grew and sintered during further hydrothermal treatments. Under 10 MPa H 2 O, the oxidation behavior was similar to that under 100 MPa H 2 O but the oxidation rate was slower, as shown in Fig. 2. Amorphous silica was produced at every temperature and then crystallized to form cristobalite at 800 °C after 72 h. The oxidation ratio was approximately represented by a Jander-type equation,9 as illustrated with solid lines in Figs. 1 and 2: [l-(l-a)1/3]2-[l_(l_a0)1/3]2 = ^ ,

(2)

where t is the time, a is the oxidation ratio at t h, a0 is the oxidation ratio at 0 h, and K is the rate constant. Using this equation we can calculate the rate constants at every temperature. Arrhenius plots of the rate constants under both pressures (Fig. 3) suggest that H 2 O pressure does not change the reaction mechanism but accelerates the reaction rate. Calculated apparent activation energies are 167 kJ/mol under 100 MPa H 2 O and 194 k J /

mol under 10 MPa H 2 O for the oxidation of SiC powder. These values are somewhat smaller than the values reported for the oxidation of SiC powders in dry air or oxygen (210-330 kJ/mol1>4>5). It should be also noted that the oxidation of SiC powder proceeds in H 2 O at temperatures as low as 500 °C in contrast to oxidation above 900 °C in air or oxygen. These differences suggest differing diffusion species in the silica layer between the oxidation in air or oxygen and in pressurized H 2 O. Collected gases from the vessel contained a large amount of CH 4 and a small amount of CO 2 . The formation of CH 4 indicates that the oxidation reaction is SiC + 2H 2 O-*SiO 2 + CH 4

(3)

This reaction is completely different from Eq. (1) where oxygen reacts with SiC to produce CO 2 as a product. Singhal8 had expected the reaction to proceed as in Eq. (3) in water vapor, but he failed to detect the production of CH 4 . Another possible reaction in pressurized H 2 O is SiC + 4H 2 O-+SiO 2 + CO 2 + 4H 2 .

(4)

The Gibbs free energies of reactions (3) and (4) are calculated assuming the ideal gas laws to apply for CH 4 , CO2, and H 2 as follows: AG(3) = G° iO2 + G£H 4

2£ ^ H 2 O ) , AG(4) = G°Si

(5)

5RTln(P/0.1) (6)

where Tis the requisite temperature (K), Pis the requisite pressure (MPa), R is the gas constant, G° is the Gibbs free energy of/ under 0.1 MPa at T (Ref. 10), and /go dGHzO is the difference of Gibbs free energy of H 2 O between 0.1 MPa and pressure P MPa at temperature T (calculated from Ref. 11). Figure 4 gives the tempera-

700

o.o -1.0

100 MPa 167 kJ/mole

-2.0

-3.0

-4.0

10 MPa 194 kJ/mole -5.0

-6.0

0.9

1.0

l.i

1.2

200

1.3

400

600

800

1000(°C)

Temperature

FIG. 3. Arrhenius plot of the hydrothermal oxidation isotherms of SiC powder. (Jander-type model.)

FIG. 4. Gibbs free energies of the reactions to produce CH 4 or CO 2 in high-temperature, high-pressure H 2 O.

J. Mater. Res., Vol. 1, No. 1, Jan/Feb 1986

101

Yoshimura, Kase, and Somiya: Oxidation of SiC powder

H20

FIG. 5. Reaction model of hydrothermal oxidation of SiC and Si3N4 powder.

SiC

ture dependence of AG for various pressures and shows that SiC is a sufficiently strong reducing agent to make CH 4 rather than CO 2 as the stable carbon species in the hydrothermal capsule. Under 0.1 MPa the reaction which makes CO 2 is more stable than CH 4 above 600 °C. These calculations are completely compatible with Singhal's result. We propose here a reaction model for SiC oxidation in H 2 O [Fig. 5 (a) ]. After the amorphous silica layer is formed on the surface of SiC particles, H 2 O and CH 4 diffuse from outside and from inside of the silica layer, respectively. A similar model, Fig. 5(b), was reported9 for the oxidation of Si3N4 powder in pressurized H 2 O where the apparent activation energies were 70-80 k J / mol (under 100 MPa at 200°-400 °C and under 10 MPa at 200°-300°C) and 130 kJ/mol (under 10 MPa at 700°-800 °C). The former value, which is similar to the activation energy of H 2 O diffusion in amorphous silica [60-105 kJ/mol (Ref. 12,13)] would indicate that the oxidation of Si3N4 was controlled by H 2 O diffusion rather than by O 2 diffusion in the silica layers.9 The H 2 O diffusion in the silica layer should be the same in the oxidation both of SiC and of Si3N4. The difference between the oxidation of SiC and Si3N4 in pressurized H 2 O, therefore, would be due to the difference of produced gases CH 4 and NH 3 . The former is a nonpolar gas while the latter is a polar gas. It is probable that nonpolar species are difficult to diffuse in amorphous silica phases. Thus the oxidation rate of SiC in H 2 O seems to be controlled by the diffusion of CH 4 in the amorphous silica layer or separate diffusion of carbon and hydrogen if CH 4 has dissociated in the amorphous silica.

vapor and in the presence of oxygen, but water vapor might promote the transition of amorphous silica to cristobalite, which might change diffusion paths from a bulk diffusion to a grain-boundary diffusion. Calculated Gibbs free energies of reaction by oxygen (1) are - 1175 and - 1016kJ/molunder0.1 MPa of oxygen pressure at 25° and 1000 °C, respectively. Therefore even if water vapor had been added to oxygen, oxygen must be a stronger oxidizing agent than water vapor. According to the present investigation it is probable that water vapor added to oxygen may cause the following two-step reactions in the oxidation of SiC: SiC

+ CH 4 ,

CH 4

2H 2 O

(3') (7)

or SiC + 4H,O-^SiCs + CO, + 4H, 4H 2 + 2O 2 -*4H 2 O,

(4') (8)

rather than the one-step reaction in Eq. (1): SiC

CO 2

Although reaction (4) is more stable than (3) at high temperature under atmospheric pressure, there is a possibility that Eq. (3) occurs as the one of the processes of the reactions. Since H 2 O diffuses faster than oxygen in the silica layer, the overall reaction of (3) plus (7) or (4) plus (8) would proceed faster than the reaction (1). This is one of the reasons for acceleration of the oxidation of SiC in atmospheres containing H 2 O.

IV. DISCUSSION

REFERENCES

The effects of water vapor on the oxidation of SiC have been discussed in previous studies. Suzuki3 proposed that water vapor might accelerate the diffusion of oxygen, carbon monoxide, or carbon dioxide through the silica layer. Jorgensen et al? considered that the diffusing species would be the same in the presence of water

'T. Nakatogawa, J. Chem. Soc. Jpn. Ind. Chem. Sect. 57,348 (1954). Suzuki, Yogyo Kyokai Shi 65, 88 (1957). 3 H. Suzuki, Yogyo Kyokai Shi 67, 157 (1959). 4 G. Ervin, J. Am. Ceram. Soc. 41, 347 (1958). 5 R. F. Adamsky, J. Phys. Chem. 63, 305 (1959). 6 P. J. Jorgensen, M. E. Wadsworth, and I. B. Cutler, J. Am. Ceram. Soc. 42, 613 (1959).

102

2 H.


Yoshimura, Kase, and SSmiya: Oxidation of SiC powder

7 P.

J. Jorgensen, M. E. Wadsworth, and I. B. Cutler, J. Am. Ceram. Soc. 44, 258 (1961). SS. C. Singhal, J. Am. Ceram. Soc. 59, 81 (1976). 9 M. Yoshimura, J. Kase, and S. Somiya, Yogyo Kyokai Shi 94, 129 (1986). 10 D. R. Stull, and H. Prophet, JANAF Thermochemical Tables, NSRDS-NBS37 (NationalBureauofStandards, Washington, D.C,

1971), 2nd ed. H. Keenan, F. G. Keyes, P. G. Hill, and J. G. Moore, Steam Tables: Thermodynamic Properties of Water Including Vapor, Liquid, and Solid Phases (International system of units-SI) (Wiley, New York, 1978). nI. Burn and J. P. Roberts, Phys. Chem. Glasses 11, 106 (1970). 13 S. White, Nature (London) Phys. Sci. 230, 192 (1971). U J.


103