Asymmetric behavior of monolayer holes after growth in GaAs molecular beam epitaxy revealed by in situ scanning electron microscopy (original) (raw)

Molecular Beam Epitaxial Growth of GaAs on (631) Oriented Substrates

Japanese Journal of Applied Physics, 2005

The homoepitaxy of GaAs on (631)-oriented substrates has been studied as a function of the growth temperature. We observed the spontaneous formation of a high density of large scale features on the surface. The hilly like features are elongated towards the ½À5; 9; 3 direction. When the growth temperature was varied from 490 to 580 C the hillocks length exponentially increases from 1.8 to 4.3 mm, their height linearly increases from 35 to 50 nm, and the density exponentially decreases from 2:8  10 6 to 3  10 5 /cm 2 . The hillocks formation is discussed in terms of adatoms diffusion anisotropy, sticking properties at step edges, and Ehrlich-Schwoebel diffusion barriers.

Study of the homoepitaxial growth of GaAs on (631) oriented substrates

Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures, 2006

We have studied the GaAs growth on ͑631͒ oriented substrates by molecular beam epitaxy ͑MBE͒. Different samples were prepared by varying the growth temperature and the III/V equivalent pressure ratio. We observed by atomic force microscopy a high density of hilly like features elongated towards the ͓−5,9,3͔ direction formed during the MBE growth. The growth temperature dependence of the hillock length and width follows an Arrhenius-type behavior with activation energies of 1.4 and 0.5 eV, respectively. The hillock formation is discussed in terms of adatom diffusion anisotropy and diffusion barriers. Employing photoreflectance spectroscopy we found a splitting of the GaAs band gap energy transition that increases with the hillock density.

Surface morphologies in GaAs homoepitaxy: Mound formation and evolution

Physical Review B, 1998

Atomic force microscopy has been used to observe surface morphologies during growth of GaAs films on GaAs͑001͒ by chemical beam epitaxy. Mound formation is observed at the beginning of GaAs growth as a function of the surface prior to deposition. GaAs substrates exhibit a large density of pits and cracks after usual thermal treatment employed for oxide desorption. On this kind of surface mounds form and coalesce as film thickness increases; surface planarization is eventually achieved-at this point, morphologies are typically those expected from two-dimensional growth. In this sense we observe that monolayer island size distribution is determined by the kinetic conditions used for the growth; nucleation sites and island spatial distribution, however, are strongly influenced by the topography of the initial surface where the film is deposited even for films thousands of monolayers thick. The final morphologies present wide terraces and few monolayer islands on top of them independent of growth conditions. This picture agrees with theoretical results where negligible step edge barriers are considered. ͓S0163-1829͑98͒02328-5͔

Dynamical faceting and nanoscale lateral growth of GaAs by molecular beam epitaxy

Journal of Crystal Growth, 2002

Dynamical faceting during homoepitaxial growth of GaAs on nanoscale-patterned surfaces by molecular beam epitaxy is examined. Selective deposition on open GaAs(1 0 0) surfaces with lateral dimensions ranging from 130 to 250 nm, separated by 15-80 nm-wide (25-nm-thick) SiO 2 stripes aligned along the ½0 % 1 1 direction results in facet formation and lateral growth over the SiO 2 mask. At the early stage of growth, (3 1 1) facets appear on sidewalls near the boundary between an open GaAs surface and SiO 2 mask, these are replaced by (1 1 1) facets starting from SiO 2 boundaries as growth continues. After complete replacement, growth proceeds laterally in the direction perpendicular to ½0 % 1 1 retaining the (1 1 1) facets until coalescence occurs between adjacent triangular cross-sectioned GaAs stripes. Nanoscale fabrication nonuniformity results in dynamical formation and retention of multiple (3 1 1) facets even for growth thicknesses much greater than the thickness of the SiO 2 mask stripes. This dynamical faceting is interpreted by minimization of total surface free energy based on equilibrium crystal shape, in qualitative agreement with our experimental results. r

Mechanisms of GaAsN growth: Surface and step-edge diffusion

Journal of Applied Physics, 2007

We have investigated the mechanisms of GaAs 1−x N x film growth by plasma-assisted molecular-beam epitaxy. A comparison of in situ reflection high-energy electron diffraction and scanning tunneling microscopy ͑STM͒, with ex situ atomic force microscopy, reveals a temperature-dependent interplay between surface and step-edge diffusion. At low temperatures, layer-by-layer growth is observed, presumably due to limited adatom surface mobility. As the temperature increases, the interplay between surface and step-edge diffusion leads to multilayer growth. For sufficiently high temperatures, adatoms overcome the step-edge diffusion barrier, resulting in layer-by-layer growth once again. The temperature range for multilayer growth is influenced by the Ga flux and may be narrowed by using As 2. Using in situ STM, we quantified the activation energies for Ga surface diffusion, E d , and step-edge diffusion, E e , during layer-by-layer GaAsN growth. We estimate E d = 0.75 and 0.96 eV for growth using As 4 and As 2 , respectively. Thus, the narrowing of the multilayer growth temperature range is likely due to the decrease in Ga surface diffusion length through the use of As 2 in lieu of As 4. Furthermore, we estimate E e = 80 meV, larger than what has been reported for GaAs growth.

Surface diffusion length of Ga adatoms in molecular-beam epitaxy on GaAs(100)–(110) facet structures

Journal of Crystal Growth, 1995

By the molecular-beam epitaxial (MBE) growth of GaAs on [OOll-mesa stripes patterned on GaAs(100) substrates, (110) facets were formed on the mesa edges defining (loo)-(110) facet structures. The surface diffusion length of Ga adatoms along the 10101 direction on the mesa stripes was obtained for a variety of growth conditions by in-situ scanning microprobe reflection high-energy electron diffraction &-RHEED). Using these values and the corresponding growth rate on the GaAs(l10) facets, the diffusion length on the (110) plane was estimated. We found that the Ga diffusion length on the (110) plane is longer than that on the (100) and (111)B planes. The long diffusion length on the (110) plane is discussed in terms of the particular surface reconstruction on this plane.

Investigation of the interface region produced by molecular beam epitaxial regrowth

Journal of Electronic Materials, 1989

The interface region generated by molecular beam epitaxial regrowth has been studied in detail. Regrowth was carried out on epitaxial GaAs after a variety of realistic device processing steps. Combinations of wet chemical etching and ion milling with and without annealing were used with the objective of establishing the best procedure for integrated technologies during regrowth. Capacitance voltage measurements showed perturbations in the carrier profile corresponding to depletion and accumulation regions at the interface which are directly related to interface states at and around the regrowth interface. The measured concentration of the interface states are in the range 1.2 x 101~ to 7.05 x 1011 cm -2. The former is one of the lowest reported till date. The concentration of deep traps in the regrown layer and interface, observed by deep level transient spectroscopy, is much lower than the interface state density. Their contribution to carrier perturbation is insignificant, except in one case where an electron trap has a rather high concentration. Results of secondary ion mass spectroscopy indicate that the presence of carbon at the regrown interface is not principally responsible for creating the high resistivity interface region. Our data favor the concept of a disordered region created at the interface during regrowth. Interface state density and trap densities are much larger in the wet chemically etched samples, which is further supported by the results of temporal photoresponse measurements on junction photodiodes. The overall characteristics of the dry etched regrowth interfaces seem to be much more promising than the wet chemical etched ones.

Superior molecular beam epitaxy (MBE) growth on (< i> N 11) A GaAs

1999

The (311) A and (511) A planes of GaAs were used for the growth of high-quality two-dimensional hole gas (2DHG) and electron gas (2DEG) structures, respectively. A back-gated, inverted interface, AlGaAs/GaAs structure in which a 2DHG or a 2DEG was embedded was studied. This particular structure enabled the two-dimensional carrier concentration to be varied over two orders of magnitude in a single device, as well as the measurement of extremely low carrier densities in the mid 109cm− 2 range.

Effect of the starting surface on the morphology of MBE-grown GaAs

Materials Science and Engineering: B, 2000

In this paper, we study the homoepitaxial growth of GaAs by molecular beam epitaxy on substrates that have different pre-growth roughness due to the method of removing the native oxide. The evolution of the surface roughness of 1 mm thick GaAs films grown at 553°C was monitored in real time using ultraviolet light scattering, and compared with ex situ atomic force microscopy measurements of the power spectral density (PSD) of the surface morphology. The PSD at a spatial frequency of 2 mm − 1 , is approximately three orders of magnitude larger for films grown on thermally cleaned substrates than for films grown on substrates cleaned with atomic hydrogen. No mounding indicative of unstable growth was observed in the films cleaned with atomic hydrogen.