Exploring the Implementation of GaAsBi Alloys as Strain-Reducing Layers in InAs/GaAs Quantum Dots.

This paper investigates the effect of GaAsBi strain reduction layers (SRLs) on InAs QDs with different Bi fluxes to achieve nanostructures with improved temperature stability. The SRLs are grown at a lower temperature (370 °C) than the usual capping temperature for InAs QDs (510 °C). The study finds that GaAs capping at low temperatures reduces QD decomposition and leads to larger pyramidal dots but also increases the threading dislocation (TD) density. When adding Bi to the capping layer, a significant reduction in TD density is observed, but unexpected structural changes also occur. Increasing the Bi flux does not increase the Bi content but rather the layer thickness. The maximum Bi content for all layers is 2.4%. A higher Bi flux causes earlier Bi incorporation, along with the formation of an additional InGaAs layer above the GaAsBi layer due to In segregation from QD erosion. Additionally, the implementation of GaAsBi SRLs results in smaller dots due to enhanced QD decomposition, which is contrary to the expected function of an SRL. No droplets were detected on the surface of any sample, but we did observe regions of horizontal nanowires within the epilayers for the Bi-rich samples, indicating nanoparticle formation.

Identification of the Segregation Kinetics of Ultrathin GaAsSb/GaAs Films Using AlAs Markers.

For optoelectronic devices from the near to the far infrared, the advantages of using ultrathin III-Sb layers as quantum wells or in superlattices are well known. However, these alloys suffer from severe surface segregation problems, so that the actual profiles are very different from the nominal ones. Here, by inserting AlAs markers within the structure, state-of-the-art transmission electron microscopy techniques were used to precisely monitor the incorporation/segregation of Sb in ultrathin GaAsSb films (from 1 to 20 monolayers (MLs)). Our rigorous analysis allows us to apply the most successful model for describing the segregation of III-Sb alloys (three-layer kinetic model) in an unprecedented way, limiting the number of parameters to be fitted. The simulation results show that the segregation energy is not constant throughout the growth (which is not considered in any segregation model) but has an exponential decay from 0.18 eV to converge asymptotically towards 0.05 eV. This explains why the Sb profiles follow a sigmoidal growth model curve with an initial lag in Sb incorporation of 5 MLs and would be consistent with a progressive change in surface reconstruction as the floating layer is enriched.

Tailoring of AlAs/InAs/GaAs QDs Nanostructures via Capping Growth Rate.

The use of thin AlA capping layers (CLs) on InAs quantum dots (QDs) has recently received considerable attention due to improved photovoltaic performance in QD solar cells. However, there is little data on the structural changes that occur during capping and their relation to different growth conditions. In this work, we studied the effect of AlA capping growth rate (CGR) on the structural features of InAs QDs in terms of shape, size, density, and average content. As will be shown, there are notable differences in the characteristics of the QDs upon changing CGR. The Al distribution analysis in the CL around the QDs was revealed to be the key. On the one hand, for the lowest CGR, Al has a homogeneous distribution over the entire surface, but there is a large thickening of the CL on the sides of the QD. As a result, the QDs are lower, lenticular in shape, but richer in In. On the other hand, for the higher CGRs, Al accumulates preferentially around the QD but with a more uniform thickness, resulting in taller QDs, which progressively adopt a truncated pyramidal shape. Surprisingly, intermediate CGRs do not improve either of these behaviors, resulting in less enriched QDs.

Suppressing the Effect of the Wetting Layer through AlAs Capping in InAs/GaAs QD Structures for Solar Cells Applications.

Recently, thin AlAs capping layers (CLs) on InAs quantum dot solar cells (QDSCs) have been shown to yield better photovoltaic efficiency compared to traditional QDSCs. Although it has been proposed that this improvement is due to the suppression of the capture of photogenerated carriers through the wetting layer (WL) states by a de-wetting process, the mechanisms that operate during this process are not clear. In this work, a structural analysis of the WL characteristics in the AlAs/InAs QD system with different CL-thickness has been made by scanning transmission electron microscopy techniques. First, an exponential decline of the amount of InAs in the WL with the CL thickness increase has been found, far from a complete elimination of the WL. Instead, this reduction is linked to a higher shield effect against QD decomposition. Second, there is no compositional separation between the WL and CL, but rather single layer with a variable content of InAlGaAs. Both effects, the high intermixing and WL reduction cause a drastic change in electronic levels, with the CL making up of 1-2 monolayers being the most effective configuration to reduce the radiative-recombination and minimize the potential barriers for carrier transport.

Control of Nitrogen Inhomogeneities in Type-I and Type-II GaAsSbN Superlattices for Solar Cell Devices.

Superlattice structures (SLs) with type-II (GaAsSb/GaAsN) and -I (GaAsSbN/GaAs) band alignments have received a great deal of attention for multijunction solar cell (MJSC) applications, as they present a strongly intensified luminescence and a significant external quantum efficiency (EQE), with respect to the GaAsSbN bulk layers. Despite the difficulties in characterizing the distribution of N in dilute III-V nitride alloys, in this work we have obtained N-compositional mappings before and after rapid thermal annealing (RTA) in both types of structures, by using a recent methodology based on the treatment of different scanning transmission electron microscopy (STEM) imaging configurations. Texture analysis by gray level co-occurrence matrixes (GLCM) and the measurement of the degree of clustering are used to compare and evaluate the compositional inhomogeneities of N. Comparison with the Sb maps shows that there is no spatial correlation between the N and Sb distributions. Our results reveal that a better homogeneity of N is obtained in type-I SLs, but at the expense of a higher tendency of Sb agglomeration, and the opposite occurs in type-II SLs. The RTA treatments improve the uniformity of N and Sb in both designs, with the annealed sample of type-II SLs being the most balanced structure for MJSCs.