Corrosion process valence band

Specific features of corrosion processes at semiconductors (as against to metals) are caused by the fact that charge carriers of both signs, namely conduction band electrons and valence band holes, take part in charge exchange between a solid and a solution. Therefore, the condition of Eq. (43) is insufficient, so account should be made of charge balance for each type of the carriers because equilibrium between the bands, which is established via generation-recombination processes, may not be reached. 

In several cases it has been found that the oxidation of the redox system occurs entirely via hole transfer directly from the valence band to the reduced form of the couple. Then both processes, oxidation of the redox system and corrosion, proceed independently. This is usually not visible from measurements with an n-type electrode, because the photocurrent is entirely determined by the light intensity. As already mentioned above, p-type electrodes are more suitable, because the current is determined by majority carrier transfer (reaction rate Vf, in Fig. 21). From the thermodynamic point of view, the oxidation of Cu at GaAs is an interesting case. The corresponding current-potential curves are given in Fig. 22 [93]. The corrosion current is not changed upon addition of Cu, i.e. corrosion and redox process are completely independent. In this case, the kinetics of the direct hole transfer is obviously very fast, i.e. the redox current is considerably larger than the corrosion current. Both processes occur indepen-... 

Illumination of a semiconductor under open-circuit conditions in an etching (oxidizing) solution gives rise to corrosion even in darkness. In the simplest case where the cathodic partial reaction of a corrosion process proceeds exclusively through the conduction band and the anodic one through the valence band, the corrosion rate for specimens of any conductivity type is limited by the minority-carrier supply to the surface and is therefore low in darkness. Illumination accelerates corrosion processes. Comparison with the case considered above shows that here the chemical polarization of the semiconductor by an oxidizer introduced into the solution acts as anodic polarization. 

For the cathodic reaction of the corrosion, there are two different charge-transfer processes. One involves holes in the valence band as a cathodic hole-injecting reaction and the other involves electrons in the conduction band as a cathodic electron-emitting reaction ... 

---