Electrochemical preparation of porous semiconductors: From phenomenology to understanding

Many semiconductors, besides silicon, can be made porous by anodization. Theories for pore formation in an electrode must explain how current flow is favoured at the bottom of the pores. Existing theories for porous silicon include `chemical' theories generally based on some kind of autocatalytic mechanism, `physical' theories focussing on hole access to the surface, and numerical simulations often based on random-walk approaches. The fact that many semiconductors can be made porous suggests that the mechanisms leading to porous silicon formation have to do more with the semiconducting character of the electrode, than with specific features of silicon electrochemistry, which tends to favour the `physical' theories. However, the problem of the initial stages of pore growth may be addressed through a linear stability analysis of the interface, a framework which can incorporate physical as well as chemical aspects. Application to the growth of porous silicon from p-Si allows one to understand the observed change in the distribution of structure sizes as the layer thickens, and the dependence of these sizes on the resistivity of the starting material. This approach seems promising for reaching a more general understanding of the mechanism of pore formation in other semiconductors.