Bonding at the $\mathrm{SiC}\mathrm{\text{−}}{\mathrm{SiO}}_2$ Interface and the Effects of Nitrogen and Hydrogen

Unlike the $\mathrm{Si}\mathrm{\text{−}}{\mathrm{SiO}}_2$ interface, the $\mathrm{SiC}\mathrm{\text{−}}{\mathrm{SiO}}_2$ interface has large defect densities. Though nitridation has been shown to reduce the defect density, the effect of H remains an open issue. Here we combine experimental data and the results of first-principles calculations to demonstrate that a Si-C-O bonded interlayer with correlated threefold-coordinated C atoms accounts for the observed defect states, for passivation by N and atomic H, and for the nature of residual defects.

Figure 1

$\mathrm{SiC}\mathrm{\text{−}}{\mathrm{SiO}}_2$ interface-state density profiles in the upper half of the 4H-SiC band-gap for the different interface treatments. The horizontal axis, $E_C-E$ denotes the energy from the CBE. The discontinuities and slope changes at $\sim 0.6\mathrm{eV}$ are a reflection of the temperature dependence of electron emission from interface traps. The error bars on the points amount are roughly a factor 2 in the $D_{\mathrm{it}}$ values.

Figure 2

Schematics of (a) the lowest-energy configuration: C takes the place of O and O is pushed out resulting in a C-O bond. (b) A slight-higher-energy configuration: C shares the O site resulting in a Si-O-C-Si bridge.

Figure 3

(a) Schematic of 2 C atoms on the SiC side of the interface (a C-C dumbell replaces a single Si atom). (b) The same defect passivated by 2 H atoms (each C forms a C-H bond).