Search

The effect of the capping process on the morphology of InAs/GaAs quantum dots (QDs) by using different GaAs-based capping layers (CLs), ranging from strain reduction layers to strain compensating layers, has been studied by transmission microscopic techniques. For this, we have measured simultaneously the height and diameter in buried and uncapped QDs covering populations of hundreds of QDs that are statistically reliable. First, the uncapped QD population evolves in all cases from a pyramidal shape into a more homogenous distribution of buried QDs with a spherical-dome shape, despite the different mechanisms implicated in the QD capping. Second, the shape of the buried QDs depends only on the final QD size, where the radius of curvature is function of the base diameter independently of the CL composition and growth conditions. An asymmetric evolution of the QDs’ morphology takes place, in which the QD height and base diameter are modified in the amount required to adopt a similar stable shape characterized by a averaged aspect ratio of 0.21. Our results contradict the traditional model of QD material redistribution from the apex to the base and point to a different universal behavior of the overgrowth processes in self-organized InAs QDs. [ABSTRACT FROM AUTHOR]

Type II emission optoelectronic devices using GaAsSb strain reduction layers (SRL) over InAs quantum dots (QDs) have aroused great interest. Recent studies have demonstrated an extraordinary increase in photoluminescence (PL) intensity maintaining type II emission after a rapid thermal anneal (RTA), but with an undesirable blueshift. In this work, we have characterized the effect of RTA on InAs/GaAs QDs embedded in a SRL of GaAsSb by transmission electron microscopy (TEM) and finite element simulations. We find that annealing alters both the distribution of Sb in the SRL as well as the exchange of cations (In and Ga) between the QDs and the SRL. First, annealing causes modifications in the capping layer, reducing its thickness but maintaining the maximum Sb content and improving its homogeneity. In addition, the formation of Sb-rich clusters with loop dislocations is noticed, which seems not to be an impediment for an increased PL intensity. Second, RTA produces flatter QDs with larger base diameter and induces a more homogeneous QD height distribution. The Sb is accumulated over the QDs and the RTA enlarges the Sb-rich region, but the Sb contents are very similar. This fact leaves the type II alignment without major changes. Atomic-scale strain analysis of the nanostructures reveal a strong intermixing of In/Ga between the QDs and the capping layer, which is the main responsible mechanism of the PL blueshift. The improvement of the crystalline quality of the capping layer together with higher homogeneity QD sizes could be the origin of the enhancement of the PL emission. [ABSTRACT FROM AUTHOR]

InSb/InAs sub-monolayer (SML) nanostructures such as SML quantum dots offer sharper emission spectra, a better modal gain and a larger modulation bandwidth compared to its Stranski–Krastanov counterpart. In this work, the Sb distribution of SML InSb layers grown by migration enhanced epitaxy has been analyzed by transmission electron microscopy (TEM) techniques. The analysis of the material by diffraction contrast in 002 dark field conditions and by atomic column resolved high angle annular dark field-scanning TEM reveal the presence of a low Sb content InSbAs continuous layer with scarce Sb-rich InSbAs agglomerates. The intensity profiles obtained by both techniques point to Sb segregation during growth. This segregation has been quantified using the Muraki segregation model obtaining a high segregation coefficient R of 0.81 towards the growth direction. The formation of a continuous InSbAs wetting layer as a result of a SML deposition of Sb on the InAs surface is discussed. [ABSTRACT FROM AUTHOR]

A procedure to quantitatively analyse the relationship between the wetting layer (WL) and the quantum dots (QDs) as a whole in a statistical way is proposed. As we will show in the manuscript, it allows determining, not only the proportion of deposited InAs held in the WL, but also the average In content inside the QDs. First, the amount of InAs deposited is measured for calibration in three different WL structures without QDs by two methodologies: strain mappings in high-resolution transmission electron microscopy images and compositional mappings with ChemiSTEM x-ray energy spectrometry. The area under the average profiles obtained by both methodologies emerges as the best parameter to quantify the amount of InAs in the WL, in agreement with high-resolution x-ray diffraction results. Second, the effect of three different GaAs capping layer (CL) growth rates on the decomposition of the QDs is evaluated. The CL growth rate has a strong influence on the QD volume as well as the WL characteristics. Slower CL growth rates produce an In enrichment of the WL if compared to faster ones, together with a diminution of the QD height. In addition, assuming that the QD density does not change with the different CL growth rates, an estimation of the average In content inside the QDs is given. The high Ga/In intermixing during the decomposition of buried QDs does not only trigger a reduction of the QD height, but above all, a higher impoverishment of the In content inside the QDs, therefore modifying the two most important parameters that determine the optical properties of these structures. [ABSTRACT FROM AUTHOR]

The optical and structural properties of GaAsBi bulk and quantum well (QW) samples grown under various conditions were studied by photoluminescence (PL), high resolution x-ray diffraction (HR-XRD) and transmission electron microscopy (TEM). At 10 K, the 90 nm bulk sample shows two PL peaks at 1.18 and 1.3 eV. The temperature and power dependent PL data suggest that both PL peaks originate from the GaAsBi layer which consists of two regions with different Bi concentrations. The TEM images verify that the Bi concentration decreases monotonically across the layer, showing a high Bi concentration (∼0.053) close to the bottom interface which then reduces to ∼0.02 for thicknesses >25 nm. Besides, the high Bi content region cannot be detected by HR-XRD due to a broad and weak diffraction intensity. For multiple QW samples, a similar Bi profile was also observed in which the first well has a significantly higher Bi content compared to the other wells. The energy separation between the PL peaks is 0.12 eV and is consistent with the energy difference observed for the bulk sample. However, two PL peaks were not observed in the other GaAsBi bulk sample which was grown under different conditions, showing the importance of growth optimizations. [ABSTRACT FROM AUTHOR]