Solar energy technologies involve components and devices whose operation depend critically on various interfaces. To obtain optimum performance from these devices it is essential to be able to analyze in some detail the characteristics of these interfaces. In principle, there now exist enough experimental techniques to determine essentially every parameter that might need to be known about the surface of a solid or the S-S interface. In practice, we are clearly not yet at that stage. The reason is not for lack of experimental techniques or theoretical approaches but primarily because of the limitations of these approaches. The variety of techniques at present available indicates that improvement will probably come from alterations in existing techniques rather than in the development of new ones. For characterizing solar interfaces, in situ measurements are critically necessary and, because very few probes are available for making relevant measurements, considerable effort should be devoted to developing in situ probes. The most popular tool for determining the elemental composition at S-G and S-S interfaces is AES in which the electrons resulting from radiotionless de-excitation of atoms are energy analyzed. Almost all peaks in an Auger spectrum can be unambiguously identified. However, there are various limitations. The sensitivity level under good normal operating conditions is usually in the 0.1%–1% range of the surface (1 nm) region. Insulating surfaces often become charged, causing severe and irreproducible shifts in the spectrum of detected elements. Many materials are subject to electron beam effects which can cause irreversible rapid change in the surface being studied. For example, the preferential removal of oxygen is observed from many metal oxides, even from apparently very stable materials such as alumina. Another useful approach is to use an ion beam as a probe for surface elemental composition, leading to the techniques of SIMS and ISS. The major limitation of ion probes is the destruction of the substrate, either for the analysis (SIMS) or because of ion bombardment (ISS). The chemical nature of the various elements is also of interest, e.g. whether the detected silicon signal is derived from elemental silicon, SiO, SiO 2, part of another silicon compound or some combination of these. For some materials, AES can answer some of these questions, but not in every case. In principle, XPS (also called ESCA) can answer these questions, but in practice the valence state may not be identifiable. In addition to knowledge about the two-dimensional surface lateral spatial distribution of components at the interface, precise knowledge of where individual surface atoms are located relative to each other, i.e. specific surface structure, might be required. Answers to the latter question might result from using LEED, EXAFS, EELS etc. However, to make these experimental techniques useful requires the addition of considerable non-trivial theoretical calculations. The questions of crystallinity can usually be answered by LEED. Unless detailed current-voltage curves are obtained and combined with elaborate theoretical analysis, only limited structural information can be gleaned from the data. The morphology of the surface is of interest. Here, SEM is clearly the popular technique being used. It can provide considerable information up to the limit of resolution of the instrument, which is usually about 10 nm under good operating conditions. However, care must be taken in the interpretation of SEM images or photographs to avoid reaching false conclusions. In all the approaches mentioned above, the power density effects of the beam probes must always be taken into account and can often be a severe limiting factor. Specific details about limitations of the popular techniques as well as observations about other techniques are considered.
Read full abstract