Realization of atomically smooth defect free InGaAs/InAlAs superlattice on InP(111) substrate by molecular beam epitaxy

Loading...
Thumbnail Image

Date

2021-07-15

Authors

Sadeghi, Ida

Advisor

Wasilewski, Zbigniew

Journal Title

Journal ISSN

Volume Title

Publisher

University of Waterloo

Abstract

InGaAs and InAlAs epilayers and superlattices were grown on rounded edge InP(111)A and InP(111)B substrates as well as on 0.45°, 1° and 2° misoriented InP(111)B substrates by Molecular Beam Epitaxy (MBE). Rounded edge wafers exposed a broad spectrum of vicinal surfaces with varying misorientation angles. The structures grown were characterized in-situ using Reflection High Energy Electron Diffraction (RHEED) and ex-situ using Nomarski differential interference contrast (DIC) microscopy, Atomic Force Microscopy (AFM), High Resolution X-Ray Diffraction (HRXRD) and scanning transmission electron microscopy (STEM). Optimum misorientation angle for growth on InP(111)B substrate was found. Conventional MBE growth at many different growth conditions did not result in a smooth surface morphology at the center of the rounded edge wafers, i.e. when growth was done on singular InP(111)A and InP(111)B substrates. However, a smooth surface morphology was observed at the rounded edge for both InP(111)B and InP(111)A substrates, which was more evident for InP(111)B substrate. It was shown that the optimum misorientation angle for the growth of InGaAs and InAlAs on InP(111)B substrate is different; it is larger for the growth of InAlAs. This is indicative of different migration length of Ga and Al adatoms on the surface. Density Functional Theory (DFT) calculations showed that the adsorption energy of Al atom is larger than that of Ga and In atoms leading to a stronger bond of Al to the surface and consequently a slower diffusion rate on the surface. Therefore, a slower growth rate is needed for the growth of InAlAs layer compared to InGaAs layer. This entails separate optimization of growth condition for the two different layers to eliminate morphological (hillocks) and microstructural (twining and stacking faults) defects. Morphological defects originate from a low migration length of the adatoms on the surface. Growth on vicinal surfaces with narrowed terrace width to promote the step-flow growth mode is an effective way to avoid hillocks. Scanning Transmission Electron Microscopy (STEM) results revealed the presence of V-type twinning and stacking faults just under the hillocks when growth is performed on singular (111) substrates. At a moderate growth temperature of 460°C, growth of atomically smooth defect free nominally lattice-matched InGaAs/InAlAs superlattice on InP(111)B was achieved for the first time through a systematic optimization of the substrate misorientation angle as well as growth conditions. STEM analysis revealed that in fact the supperlattice is free of any defect. The presence of strain in the structure introduces defects to account for the mismatch or relieve the strain. It was seen that i) misfit dislocations that are dissociated into stacking faults bounded by partial dislocations at both ends, i) phase separation that is the formation of regions of rich in In and poor in Ga or Al and vice versa, and iii) rotation of the crystal lattice at some regions within the microstructure are the main mechanisms for strain relief. The effect of substrate annealing temperature on the surface reconstructions of InP(111)A and InP(111)B was studied using Low energy electron diffraction (LEED). It was seen that InP(111) substrate preserves a (1×1) unreconstructed surface at all the annealing temperatures studied from 250°C to 500°C, therefore, it is thermally stable. On the other hand, InP(111)A changes its reconstruction from (2×2) at low annealing temperatures to a mixture of (2×2) and (3×3) at medium annealing temperatures and eventually to a (3×3) reconstruction at high annealing temperatures, therefore, it is not thermally stable. Since the surface reconstruction plays an important role in the growth quality as was evidenced by the very different growth morphologies achieved on InP(111)A and InP(111)B substrates, different growth temperatures for growth on InP(111)A could result in a very different growth quality, while growth on InP(111)B is less affected by the growth temperature.

Description

Keywords

Molecular beam epitaxy, (111) substrate, InGaAs/InAlAs superlattice

LC Keywords

Citation