STM investigations of dislocation formation in heteroepitaxy

Semiconductor heterostructures are of great importance for microelectronic devices as well as for fundamental research on the physics low dimensional systems. While most of the early investigations were restricted to materials with well matched lattice constants, lattice-mismatched heterostructures have attracted tremendous interest because the much larger choice of materials allows much more freedom in design and fabrication of such structures. The existence of a lattice-mismatch between the epitaxial layers, however, poses a number of serious problems for heteroepitaxial growth. Although the layers initially assume the in-plane lattice constant of the substrate, as the thickness increases they become mechanically unstable against lattice relaxation, marked by the onset of misfit dislocation formation at the layer/substrate interface. While the absolute mechanical stability of 2D overlayers is well described by the Matthews and Blakeslee model, many experiments have found strong deviations from equilibrium theories. This is because epitaxial growth is usually carried out far from thermodynamic equilibrium, where the actual course of strain relaxation is determined by the dislocations kinetics rather than by the energetics. These kinetics are controlled by competing processes such as nucleation and glide of dislocations, as well as dislocation interaction. multiplication and pinning processes. To fully understand the kinetics of dislocation formation detailed investigations of the various mechanisms are required. Traditionally, misfit dislocations have been studied by transmission electron microscopy, x-ray topography and cathodo-luminescence. However. recent work has demonstrated that scanning tunnelling microscopy is a powerful tool for imaging of misfit dislocations. It is based on the sub-Angstrom resolution of STM that allows the direct real space imaging of the local atomic displacements on the surface induced by dislocations at the buried interfaces. In this research project we will explore and extend the capabilities of in situ scanning tunnelling microscopy for investigation of the subsurface defect structure in lattice-mismatched heterostructures to gain insights in the kinetic processes of dislocation formation. We will focus on molecular beam epitaxy of IV-VI semiconductors and their alloys with rare earth chalcogenides, which have been used for fabrication of infrared optoelectronic devices such as detectors and tuneable diode lasers. High temperature scanning tunnelling microscopy will be used for time resolved STM studies of nucleation, glide and climb of dislocations, and of dislocation reaction and pinning processes. These processes are particularly important in strain relaxation but are currently not well understood. We will identify the kinetic processes that determine the course of strain relaxation and develop a theoretical model for full description of the dislocation kinetics in IV-VI semiconductors. In addition, we will study the mechanisms for the threading dislocation density reduction in IV-VI layers and compare the dislocation formation processes for (111) and the (100) growth orientation. Other important issues are the investigation of the relaxation processes during thermal cycling and during the growth of self-assembled quantum dots. The final goal will be to determine the influence of the localized strain fields of misfit dislocations on surface step structure and on the evolution of surface morphology through stress induced surface diffusion. and we will explore how these effects can be utilized as a novel method for control and fabrication of self-assembled semiconductor nanostructures.

Semiconductor heterostructures are of great importance for advanced microelectronic devices as well as for fundamental research on low dimensional systems. While most of the early applications were restricted to material systems with well-matched lattice constants, lattice-mismatched heterostructures drastically widened the choice of materials that can be combined in such devices, allows much greater freedom in design and fabrication of such structures. However, the existence of a lattice-mismatch between the epitaxial layers and the underlying substrate poses a number of serious problems for heteroepitaxial growth because as a result the layers become unstable against the formation of misfit dislocations, which disrupt the perfect lattice registry between the layer and substrate atoms. Traditionally, misfit dislocations have been studied mainly by transmission electron microscopy, x-ray topography and cathodo-luminescence. In this project, we have developed scanning tunneling microscopy as powerful new tool for imaging of subsurface misfit dislocations. This is based on the sub-atomic resolution of STM, which allows a direct real space imaging of the local atomic displacements on the surface induced by strain fields of misfit dislocations far below the layer surface. The studies have focused on the investigation of the kinetics and energetics of the dislocation formation process in narrow band-gap IV-VI semiconductor heterostructures that are of interest for mid-infrared optoelectronic device applications. In particular, it was found that for certain growth conditions quasi-periodic networks of misfit dislocations can be formed that can be utilized as patterned templates for synthesis of novel self-organized semiconductor nanostructures. For a description of the strain relaxation process, new theoretical models were proposed that allow a much-improved prediction of the critical layer thickness as well as evolution of the layer strain as a function of layer thickness.