Polytype formation and transformation during the reaction-bonding of silicon carbide

Abstract

This paper explores the polytype distribution and polytypic transformations occurring during the reaction-bonding of a commercial silicon carbide, REFEL. The initially formed polytype is found to be almost exclusively β-SiC, which either deposits epitaxially on the original α-SiC grits or as fine material, in situ in the silicon. The exothermic nature of the reaction causes the epitaxially deposited material to undergo the β→α solid state transformation, usually going almost to completion in a few minutes. Two types of epitaxial material are distinguished ; type I has its stacking sequence directly seeded by the original grain, whilst type II has no direct contact with the original α stack, but can be seeded by 'sideways' nucleation. The fine (type III) material is never found to transform, either because of nucleation problems or because it is formed at a lower temperature. The kinetics β→α transformation indicate a peak local temperature of ~ 2000°C and the rapidity of the transformation is sufficient to rule out any large-scale diffusion. Other microstructural features are noted, including the unexpected presence of some 2H-SiC in certain 'nodules' in the compact, associated with high impurity levels and unusual reaction conditions. The observed behaviour emphasizes the importance of local temperature, impurities and nucleation effects in determining polytype stability. Further, the preferential initial growth of β-SiC confirms it as the high-symmetry, rapidly-forming polytype.

Layered growth modelling of epitaxial growth processes for SiC polytypes

“…Our results show that the 6H-SiC phase is the ground state phase with the lowest total energy. This is in agreement with experiments [2,33,46], which indicate that 6H-SiC is the stable phase at high temperatures and probably 4H-SiC is stable at low temperatures. Polytypic transformation of 4H-SiC into 6H-SiC has been observed experimentally [47,48].…”

“…These results seem to suggest that the 3C-SiC phase is never the stable structure at any temperature. However experiments [2,[33][34][35][36] found that the 3C-SiC phase does prefer to grow epitaxially. A good model that Heine et al [12] have introduced can explain the experimental phenomenon.…”