Abstract

We review theoretical interpretations of Schottky barriers and Fermi-level pinning, which result when metals and other chemical species are deposited on semiconductor surfaces. Experiments indicate that these two phenomena are closely connected, so a theory of Schottky barriers must also explain Fermi-level pinning for submonolayer coverages of both metallic and nonmetallic species. Proposed mechanisms include the following: (a) Intrinsic surface states. For GaAs and several other materials, there are no intrinsic surface states within the band gap; GaP, e.g., does have surface states in the gap, but they are not at the correct energy to explain Schottky barrier formation. (b) M e t a l-induced gap states. These states, which require a thick metal overlayer, cannot explain Fermi-level pinning at submonolayer metallic coverages. They also cannot explain why a single semiconductor (n-type InP) exhibits two distinct Schottky barrier heights. Furthermore, they cannot explain why the Schottky barrier persists when there is an oxide layer between semiconductor and metal. Metal-induced states can in principle give rise to Schottky barriers at defect-free interfaces, but they fail to explain much of the existing experimental data for III–V semiconductors and Si. (c) The classic Schottky model. This model is not in agreement with experiment for III–V and Group IV semiconductors, but does appear to account for the measurements involving nonreactive metals on GaSe−a layered material expected to be relatively free of defects. (d) The Spicer defect model. This phenomenological model, now supported by microscopic theoretical studies, appears to account for many of the observations regarding Schottky barrier and Fermi-level pinning. We review our theoretical investigations within the framework of the defect model, which provide a satisfactory explanation of the principal observations for both III–V and Group IV semiconductors. We conclude that the levels responsible for Schottky barriers and Fermi-level pinning arise from two sources: (1) bulk-derived deep levels (e.g., the deep donor level for the antisite defect AsGa, which persists when this defect is present at the surface, but which is shifted in energy), and (2) dangling-bond deep levels (which are also shifted in energy according to the environment of the dangling bond). Most of the observed Schottky barriers—for both III–V and Group IV semiconductors—are attributed to dangling bonds.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call