Although generally ascribed to the presence of defects, an ultimate assignment of the different contributions to the emission spectrum in terms of surface states and deep levels in ZnO nanostructures is still lacking. In this work we unambiguously give first evidence that zinc vacancies at the (1010) nonpolar surfaces are responsible for the green luminescence of ZnO nanostructures. The result is obtained by performing an exhaustive comparison between spatially resolved cathodoluminescence spectroscopy and imaging and ab initio simulations. Our findings are crucial to control undesired recombinations in nanostructured devices.

We report the Au-assisted chemical beam epitaxy growth of defect-free zincblende InSb nanowires. The grown InSb segments are the upper sections of InAs/InSb heterostructures on InAs(111)B substrates. We show, through HRTEM analysis, that zincblende InSb can be grown without any crystal defects such as stacking faults or twinning planes. Strain-map analysis demonstrates that the InSb segment is nearly relaxed within a few nanometers from the interface. By post-growth studies we have found that the catalyst particle composition is AuIn(2), and it can be varied to a AuIn alloy by cooling down the samples under TDMASb flux.

In_x Ga_1−x As/GaAs single heterostructures have been grown by molecular‐beam epitaxy with different growing rates and In molar fractions. Indium composition, layer thickness, and residual strain have been measured mainly by Rutherford backscattering/channeling spectrometry and the results on selected samples compared with the results of other techniques like Auger electron spectroscopy and single‐ and double‐crystal x‐ray diffraction. Cathodoluminescence, x‐ray topography, transmission electron microscopy, and ion dechanneling have been employed to observe dislocations and to characterize their nature and density. While the onset of misfit dislocations has been found to agree with the predictions of the equilibrium theory, the strain release has been found to be much lower than predicted and the results are compared with the available metastability or nucleation models. Present results are in best agreement with nucleation models. Moreover, annealing experiments show that these heterostructures are at (or very close to) thermodynamic equilibrium.

A model to compute the strain relaxation rate in InxGa1−xAs/GaAs single layers has been tested on several compositionally graded buffer layers. The existence of a critical elastic energy has been assumed as a criterion for the generation of new misfit dislocations. The surface strain accuracy results are within 2.5×10−4. The influence of different grading laws and growth conditions on residual strain, threading dislocation density, misfit dislocation confinement, and surface morphology has been studied. The probability of dislocation interaction and work hardening has been shown to strongly influence the mobility and the generation rate of the dislocations. Optimization of the growth conditions removes residual strain asymmetries and smoothes the surface roughness.