GROWTH OF SEMICONDUCTOR HETEROSTRUCTURES ON PATTERNED SUBSTRATES - DEFECT REDUCTION AND NANOSTRUCTURES

The current status of the growth of semiconductor single and multilayered structures on non-planar patterned substrates (NPPS) is reviewed with particular focus on the use of NPPS for (1) defect reduction in the growth of lattice mismatched systems and (2) realization of 2- and 3-dimensionally confined structures via a one-step growth process. In the first category a comparison of the currently-employed theoretical framework and experimental findings reveals serious discrepancies whose origin, it is argued, lies in the inadequacy of the underlying conceptual models and not merely in the limitations of the calculated expressions. It is further argued that an atomistic understanding of strained growth on unpatterned and patterned substrates alike can be gained only by recognizing the intimate connection between the kinetics of growth and of defect formation. The role of intra- and inter-facet migration in impacting the growth on NPPS in general, and in impacting defect formation in strained growth in particular, is evaluated. It is shown that existing observations point to a critical role being played by the relative orientation of the surface dangling orbitals and the ledge directions defining the mesa shapes. This is shown to define the notion of a substrate encoded size-reducing epitaxy (SESRE) that can be exploited to prepare, in-situ and via a one step growth on appropriately chosen mesas, nanoscale laterally-confined templates that serve as ''substrates'' for (a) subsequent growth of strained structures with significantly reduced defect density in strained systems and (b) realization of nanostructures. In the second category, after reviewing the current status of nanostructure fabrication via growth on NPPS, attention is focused on the use of SESRE as a specific and more versatile approach to the realization of 2D and 3D confined structures. Examples from the recent first realization of 3D confined structures via growth on NPPS are cited while also noting the serious technological issues relating to sufficient control on the as-patterned size and growth conditions that impact realization of uniform 2D arrays. The review concludes with some comments on the future directions of investigations.