Activation of dopants in SiC: theoretical study

The electrical activation and solubility of dopants are the limiting factors that hamper the manufacturing of a highly doped SiC. In nitrogen-doped SiC a complete electrical activation was achieved \footnote{M.~Laube {\em et~al\/}, J.~Appl.~Phys.~{\bf 91}, 549 (2002).}\textsuperscript {,} \footnote{D.~Schulz {\em et~al\/}, Mat. Sci. Forum {\bf 338-342}, 87 (2000).} only for impurity concentration below $2-5\times 10^{19}$cm$^{-3}$. Yet for phosphorus a full activation was obtained \footnotemark[1] up to $10^{20}$cm$^{-3}$. We study the different activation behavior of these dopants theoretically by an {\em ab initio\/} DFT approach. We find that phosphorus mainly substitutes for silicon, whereas nitrogen is incorporated exclusively into the carbon sites. In a thermodynamic equilibrium the activation of both donors is not limited by the self-compensation. Phosphorus is found to be fully activated until the onset of precipitation. In contrast, nitrogen preferentially incorporates in the neutral nitrogen-vacancy complexes at concentrations above $2\times10^{19}$cm$^{-3}$. This leads to the nitrogen passivation, in agreement with the experimental findings \textsuperscript {1,2}.