Epitaxy, LEED observations

The growth of one crystal upon the surface of another is epitaxy. In the broadest sense of this term, epitaxy includes all of adsorption, corrosion and growth of thin films. When a thin film of deposit is laid down, it can be expected that its structure may differ radically from the bulk structure of a thick deposit. Structure of a deposit containing less than one monolayer can be as different from that of a thin film which precedes bulk growth, as the thin film structure may be from that of a heavy deposit. Some remarkable LEED observations of this kind have been made that are still only poorly understood. 

It is quite probable that the patterns that order only at elevated temperatures are due to a mixed layer of oxygen and the metal in question. Ducros and Merrill (134) have suggested that the complex patterns observed on Pt(U0) can be attributed to an epitaxial layer of Pt02 with the (100) plane of the oxide parallel to the Pt(110) surface. Conrad et al. (130) explained the LEED patterns obtained by exposure of Pd( 111) to NO at elevated temperatures by a PdO epitaxial layer with the PdO (100) surface parallel to the substrate. Many of the other patterns observed cannot be explained by simple models and, at the present state of development, LEED intensity calculations are not able to deal with such complicated unit cells so that the... 

In the monolayer range, a (Nxl) superstructure (with N = 5-6) has been observed by LEED, and grazing X-ray diffraction [56]. Based on theoretical works, it can be assumed that the Pd adatoms tend to remain close to the epitaxial hollow sites with the formation of a vacancy each N atoms [57]. The observation of a N order superstmeture along the [001] direction indicates an interaction between Pd rows and/or vacancies. A schematic representation of the proposed structure, for N = 6, is reported in Fig. 10. This (Nxl) stmeture is retained up to a coverage of about 2 ML. 

Study of epitaxial growth by the LEED method is often quite clear-cut and uncomplicated by formation of unexpected structures. One simply observes development of a characteristic crystal plane of the deposit bulk structure, and this is usually quite obvious from simple inspection of the pattern. Orientation of the film on the substrate is usually easily evident also. In some cases, the overlayer is coherent with the substrate as a coincidence lattice producing complicated LEED patterns. Yet it is usually relatively straightforward to decipher such patterns, because spacings in the overlayer structure are often easily assigned from known X-ray spacings of the substance being deposited [see, for example, Bauer (95)]. 

The problem is complicated by the circumstance that depending on temperature and pre-treatment of the substrate, the surface may be reconstructed in different ways, i.e., different superstructures are observable by RHEED [e.g. on SiC(OOOl) (3 x 3), (2 X 2), and (V3 x V3)R30°], which may influence the nucleation of the epitaxial layer. Recently, such superstructures on SiC surfaces have been analyzed by LEED and modeled [202, 203]. 

The monolithic heterostructures with 8 and 10 layers were formed on Si(100) substrates. In 8-layer structure the distance between layers was 100 nm and a growth was controlled by LEED. The surface of 8-layer sample is smooth (orms= 0.28 nm), and defects are absent. On each growth stage a bright 2x1 LEED pattern was obtained. So, an epitaxial growth was observed on every stage of the silicon deposition. p-FeSi2NC s (with Moure pictures) andy-FeSi, NC s (without Moure picture) have been observed by HRTEM. 

---