Oxidation behavior of pressureless liquid-phase-sintered α-SiC in ambient air at elevated temperatures

Abstract

The long-duration oxidation behavior of a pressureless liquid-phase-sintered (LPS) α-SiC with 10 vol% Y₃Al₅O₁₂ additives was studied by furnace oxidation tests in ambient air at 1100 to 1450 °C. The oxidation of this LPS SiC ceramic was found to be passive throughout these temperatures due to the formation of oxide scales, with a change in the oxidation behavior occurring at 1350 °C. It was also found that the oxidation behavior is very complex, exhibiting two distinct stages at all temperatures: (i) initial nonparabolic oxidation, where the rate-limiting mechanism is the outward diffusion of Y³⁺ and Al³⁺ cations from the secondary intergranular phase into the oxide scale with the activation energy of the oxidation being 504 ± 32 kJ/mol, followed by (ii) parabolic oxidation below 1350 °C, where the rate-determining mechanism is the inward diffusion of oxygen through the oxide scale with the activation energy being 310 ± 47 kJ/mol, or paralinear oxidation at and above 1350 °C, where oxidation is controlled by some mixed reaction/diffusion process. The existence of two oxidation regimes reflects the progressive crystallization of the oxide scale during the oxidation. Finally, guidelines are provided for the design and fabrication of low-cost, highly oxidation-resistant LPS SiC or other LPS nonoxide ceramics.

Appendix

The aim of this appendix is to describe the procedure followed to model the two-stage oxidation curves. First, the oxidizing time at which the functional form in the (Δm)² − t plots changed (i.e., the transition time or t₀) was evaluated for all oxidizing temperatures. This was done by calculating the moment at which the numerical derivative of the (Δm)² − t curves in Fig. 9 becomes constant or almost constant. (t₀ was found to be close to 100 h below 1350 °C, and then decreased up to 10 h at 1450 °C). Second, the oxidation kinetics for t > t₀ was described by the parabolic rate law [i.e., Eq. (3) ] below 1350 °C and by the paralinear rate law [i.e., Eq. (9)] at and above 1350 °C. To this end, the free parameters of the parabolic (k⁰_p and b⁰) and paralinear (k^/_p, k_l, and b) models were refined by nonlinear least squares fitting until the best agreement between the calculated and observed oxidation curves over the t₀–500 h range was achieved. The best values of k⁰_p and b⁰, or of k^/_p, k_l, and b, were thus obtained for the oxidation curves. Third, continuity of the oxidation curve and of its derivate at t₀ was imposed, which allowed the following relations between the parameters of the arctan and parabolic/paralinear models to be derived:

$${k_{\rm{p}}} = {\left( {k_{\rm{p}}^0} \right)^2}{{{t_0}} \mathord{\left/ {\vphantom {{{t_0}} {\left[ {{f^2}\left( {{t_0}k_{\rm{p}}^0 + {b_0}} \right)} \right],}}} \right. \kern-\nulldelimiterspace} {\left[ {{f^2}\left( {{t_0}k_{\rm{p}}^0 + {b_0}} \right)} \right],}}$$

((A1))

$${k_{\rm{p}}} = {\left( {\sqrt {k_{\rm{p}}^\prime } - 2{k_1}\sqrt {{t_0}} } \right)^{{2 \mathord{\left/ {\vphantom {2 {{f^2}}}} \right. \kern-\nulldelimiterspace} {{f^2}}}}},$$

((A2))

Fourth, Eqs. (A1) and (A2) were substituted individually into Eq. (4), which allowed the arctan rate law to be rewritten in the forms (in the absence of oxidation at t = 0):

((A3))

below 1350 °C, and:

((A4))

at and above 1350 °C. Fifth, the free parameters of the arctan model (β and f, because t₀, k⁰_p, and b⁰ or t₀, k^/_p, k_l, and b were determined previously) were refined by nonlinear least squares fitting until the best agreement between the calculated and observed oxidation curves over the 0–t₀ range was achieved. The best values of β and f were thus obtained for each oxidation curve. Last, the rate constants k_p below and at and above 1350 °C were determined with the inputs of the β, f, t₀, k⁰_p, and b⁰ or of β, f, t₀, k^/_p, k_l, and b into Eqs. (A1) and (A2), respectively.