Basic studies of the corrosion of SiC by silicate melts

Abstract

Three different commercially available SiC materials (one SiSiC and two sintered SSiC) and single crystals of α-SiC were corroded at 1000 °C by silicate melts based on Na₂O-2SiO₂ with different additives of Al₂O₃, CaO, Fe₂O₃ and TiO₂. During the corrosion experiments the samples were completely immersed in the melts. Parts of the samples were subsequently leached in H F for weight loss determination and surface change examinations, and other parts of them were prepared as polished cross-sections for SEM-EDXS investigations of SiC-melt interfaces. Corrosion starts by dissolution of a silicon dioxide layer in the silicate melt followed by oxidation of the free silicon phase, and silicon carbide phase, and grain boundary attack. Corrosive attack and corrosion mechanisms are dependent on the composition and structure of the silicon carbide material. An influence of Fe, Al, Ca and Ti addition on the extent of corrosion was observed. In the vicinity of SiC-melt interfaces SiO₂-rich melt regions were found. The unusually high corrosion rate of SiC samples in the case of Fe₂O₃ addition to sodium disilicate was concluded to be the consequence of Fe²⁺-Fe³⁺ transition. Some defects in materials which were introduced during manufacturing or machining corroded much more severely than the defect-free material.