Abstract
The electrochemical oxidation of porous silicon layers formed on lightly p‐doped substrates in aqueous electrolytes has been studied in detail. Two different oxidation regimes are observed, depending on the presence of a native oxide on the silicon pore walls. When anodic oxidation is performed on samples previously dried in air, a thin oxide layer is already present and the oxidation regime is determined by the transport of the oxidizing species across the oxide. In this regime, the anodic polarization accumulates the majority charge carriers in the whole depth of the pore walls, and oxidation proceeds homogeneously throughout the whole porous layer. If an experimental procedure which prevents native oxide formation is used, the oxidation process is governed by hole supply at the silicon‐electrolyte interface, and preferential oxidation of the bottom of the porous layer is obtained at low anodic current densities. The electrochemical oxidation stops before complete oxidation because the electrical contact between the bulk and the partially oxidized porous layer is cut by the formation of a continuous oxide at the bottom of the porous layer. The oxidized thickness obtained before this break increases with the current density, and the whole thickness can be oxidized if the oxidation current density is greater than the value which corresponds to hole accumulation in the whole depth. However, homogeneous oxidation of the layer never corresponds to complete oxidation, and at the end of the anodization process about 40% of silicon atoms in the porous layer are unoxidized, and form tiny silicon microcrystallites distributed throughout the oxidized porous film.