I have some problems when simulating the surface depletion in an epitaxial GaAs layer with ATLAS. According to the abrupt depletion aproximation, the depletion region extends to a depth of sqrt(2/e x epsilon x Vbs/(Nd-Na)). For a unintentionally doped layer of Nd=1E15 cm-3, a depletion depth of about 1µm is expected assuming Vbs=0.6 V.

I have tried to simulate this in ATLAS by using a insulator/semiconductor system: SiO2 as an upper layer and an n-doped GaAs (1E15 cm-3) below. For SiO2, I defined the energy gap of 9eV and affinity of 0.95eV and included the flag semicond to treat the oxide as a semiconductor. The simulation results show an extremely thin space charge region of few nanometer in clear contradiction with the theory. I would appreciate any suggestions.

Best wishes, Thank you in advance.