Subject semiconductor surfaces

It should be mentioned that as well as for metals the passivation of semiconductors (particularly on Si, GaAs, InP) is also a subject of intense investigation. However, the goal is mostly not the suppression of corrosion but either the fonnation of a dielectric layer that can be exploited for devices (MIS stmctures) or the minimization of interface states (dangling bonds) on the semiconductor surface [63, 64]. 

Thus, it may be seen that, by reducing the particle radius, it is possible to obtain systems where transit from the particle interior to the surface occurs more rapidly than recombination, implying that quantum efficiencies for photoredox reaction of near unity are feasible. However, the achieving of such high quantum efficiencies depends very much upon the rapid removal of one or both types of charge carrier upon their arrival at the semiconductor surface, underlining the importance of the interfacial charge-transfer kinetics. This is the subject of the next section. 

The intensity of work on semiconductor surfaces appears to have diminished somewhat in the last few years but this is certainly not an indication that all the pertinent problems have been solved, or that the subject has become less significant. 

The most commonly used direct method for determination of surface stress is the bending beam method in which a lever bends when subjected to a change in surface stress in one of its faces in order to minimize its stored strain energy. The relationship between the deflection of a cantilever and the different stresses in its smfaces was first determined by Stoney [25] and in surface science it has been used to measure the stress changes associated with the reconstructions of semiconductor surfaces [26 - 27]. 

Whilst in a qualitative sense this theory has a certain basic validity, it does not provide very real physical insight into the electronic and structural effects now known to be associated with adsorption on semiconductors. In the following sections, we shall attempt to show how the subject has been developed from the much greater understanding of semiconductor surfaces which is now available. 

Photocurrent generation is one of the most interesting direct applications of photosynthetic studies. The adsorption of sensitizers onto semiconductor surfaces has been found to be an efficient way to generate photocurrents and has been studied extensively. Ruthenium bipyridyl complexes, in particular, have been the focus of recent research [137-139]. In these cases, only the first layer of molecules, which is in direct contact with the surface, is active. A highly porous semiconductor material was therefore employed to compensate for the low level of absorption of the single molecular layer. Other varieties of chromophores, semiconductor materials, and electron carriers for totally solid systems have been the subjects of extensive studies. The present... 

The Si(111) surface is probably the most studied semiconductor surface. Yet, the details of the atomic and electronic structure are still considered open subjects. Experimental interest remains high because it is possible to cleave Si in vacuum and produce clean surfaces which can be studied with a host of techniques. Theoretically, this surface is considered to be the prototype semiconductor surface. 

It is evident from this review that TMC surfaces are interesting subjects of both experimental and theoretical investigations. However, at this point, information on TMC surfaces is limited compared with the substantial accumulation of data on elemental metal and semiconductor surfaces. There remain many unsolved elemental problems for the TMC surfaces. For example, the surface states on the group V TMC(lll) have substantial dispersion, whereas those the group IV TMC(lll) have little dispersion, which cannot be explained by a simple rigid band model. The TMC crystals inevitably include vacancies on the carbon site, which should have considerable effects on the surface electronic structure, although systematic studies of this point have not yet been performed. Parallel extensive studies, both theoretical and experimental, are needed for further progress in the surface chemistry and physics of TMCs. 

It should be noted immediately that not all the frequencies absorbed by a semiconductor are photocatalytically active, but only those that are also photoelectrically active, i.e., that cause an internal photoelectric effect in the semiconductor. Note further that the sign and magnitude of the photo-catalytic effect depend on the past history of the specimen exposed to illumination i.e., they depend on the external influences to which the specimen in question was subjected in the course of the whole of its life, and also on the conditions of the experiment (temperature, intensity of illumination, etc.). For example, by introducing into the semiconductor an impurity of any concentration or by adsorbing foreign gases on its surface it is possible to render its catalytic activity more or less sensitive to illumination. 

The difference between a photoconductive detector and a photodiode detector lies in the presence of a thin p-doped layer at the surface of the detector element, above the bulk n-type semiconductor. Holes accumulate in the p-layer, and electrons in the n-type bulk, so between the two there is a region with a reduced number density of carriers, known as the depletion layer. The important effect of this is that electron-hole pairs, generated by photon absorption within this depletion layer, are subjected to an internal electric field (without the application of an external bias voltage) and are automatically swept to the p and n regions, and... 

---