N-rich Growth Conditions Research Articles

Carbon doping of GaN with CBr4 in radio-frequency plasma-assisted molecular beam epitaxy

Carbon tetrabromide (CBr4) was studied as an intentional dopant during rf plasma molecular beam epitaxy of GaN. Secondary ion mass spectroscopy was used to quantify incorporation behavior. Carbon was found to readily incorporate under Ga-rich and N-rich growth conditions with no detectable bromine incorporation. The carbon incorporation [C] was found to be linearly related to the incident CBr4 flux. Reflection high-energy electron diffraction, atomic force microscopy and x-ray diffraction were used to characterize the structural quality of the film’s postgrowth. No deterioration of structural quality was observed for [C] from mid 1017 to ∼1019 cm−3. The growth rate was also unaffected by carbon doping with CBr4. The electrical and optical behavior of carbon doping was studied by co-doping carbon with silicon. Carbon was found to compensate the silicon although an exact compensation factor was difficult to extract from the data. Photoluminescence was performed to examine the optical performance of the films. Carbon doping was seen to monotonically decrease the band edge emission. Properties of carbon-doped GaN are interpreted to be consistent with recent theoretical work describing incorporation of carbon as function of Fermi level conditions during growth.

Strong influence of Ga/N flux ratio on Mn incorporation into Ga1−xMnxN epilayers grown by plasma-assisted molecular beam epitaxy

We report the growth of Mn-doped wurtzite GaN epilayers by nitrogen plasma-assisted molecular beam epitaxy, with a systematic attention to the dependence on the growth conditions. The addition of Mn modifies the growth diagram related to the Ga/N flux ratio. In particular, the stable Ga-bilayer coverage on the growth surface for the Ga-rich condition is destabilized in the presence of Mn. Mn incorporation in the epilayers is found to strongly depend on the Ga/N flux ratio: it varies by two orders of magnitude between the Ga-rich and the N-rich growth conditions. X-ray diffraction measurements on epilayers grown in the stoichiometric condition reveal a clear contrast between the precipitation of perovskite GaMn3N clusters at Mn compositions higher than 1.7%, and the single phase of wurtzite Ga1−xMnxN at lower Mn compositions.

Theoretical study of native defects in BN nanotubes

Antisites, carbon substitutionals, and vacancy defects in BN nanotubes have been investigated by first-principles total energy calculations, based on a density functional spin-polarized method. All defects introduce localized energy levels inside the band gap. For some open-shell systems (carbon substitutionals and vacancy defects) the levels in the gap region present an exchange splitting greater than 0.5 eV. The ${\mathrm{C}}_{\mathrm{B}}$ and ${\mathrm{N}}_{\mathrm{B}}$ defects, in N-rich growth condition, and ${\mathrm{C}}_{\mathrm{N}}$ in B-rich growth condition, are the ones that present the lower formation energies.

Ga flux dependence of Er-doped GaN luminescent thin films

Er-doped GaN thin films have been grown on (111) Si substrates with various Ga fluxes in a radio frequency plasma molecular beam epitaxy system. Visible photoluminescence (PL) and electroluminescence (EL) emission at 537/558 nm and infrared (IR) PL emission at 1.5 μm from GaN:Er films exhibited strong dependence on the Ga flux. Both visible and IR PL and visible EL increase with the Ga flux up to the stoichiometric growth condition, as determined by growth rate saturation. Beyond this condition, all luminescence levels abruptly dropped to the detection limit with increasing Ga flux. The Er concentration, measured by secondary ion mass spectroscopy and Rutherford backscattering, decreases with increasing Ga flux under N-rich growth conditions and remains constant above the stoichiometric growth condition. X-ray diffraction indicated that the crystalline quality of the GaN:Er film was improved with increasing Ga flux up to stoichiometric growth condition and then saturated. Er ions in the films grown under N-rich conditions appear much more optically active than those in the films grown under Ga-rich conditions.

Initial growth of hexagonal GaN grown on an Si(1 1 1) substrate coated with an ultra-thin SiC buffer layer

The initial growth of hexagonal GaN films grown on Si(1 1 1) substrates coated with an ultra-thin SiC buffer layer was studied. It was found that a 2.5-nm-thick SiC layer is an effective buffer layer for GaN growth on an Si(1 1 1) substrate. Under Ga-rich growth conditions, Ga adatoms, in comparison to those under N-rich growth conditions, were highly mobile. Consequently, the GaN films had a flat surface and an almost stacking-fault-free microstructure. The initial GaN nucleations quickly coalesced laterally to submicron-sized grains. Under N-rich growth conditions, the initial GaN nucleations saturated at a diameter of about 50 nm (measured at the film surface). The grown GaN films showed statistical roughening of the surface and a characteristic columnar structure. The yellow-band luminescence (YL) was sensitive to the microstructures of the GaN films prepared under almost the same growth conditions, suggesting that the Ga-vacancy is not the sole source of YL.

Electronic structure of1×1GaN(0001) andGaN(0001¯)surfaces

The electronic structure of clean and Ga- or N-covered $1\ifmmode\times\else\texttimes\fi{}1$ GaN(0001) and $\mathrm{GaN}(0001\ifmmode\bar\else\textasciimacron\fi{})$ surfaces is studied using the local-density approximation of density-functional theory employing ab initio pseudopotentials together with Gaussian orbital basis sets. We use both standard and self-interaction- and relaxation-corrected pseudopotentials. The latter allow for a most accurate description of the electronic structure of these surfaces. Comparing the formation energies for the clean and adatom-covered $1\ifmmode\times\else\texttimes\fi{}1$ configurations, we determine optimal surface structures for various growth conditions. For the GaN(0001) surface in the Ga-rich case, we find a structural model consisting of Ga adatoms adsorbed in ${T}_{4}$ positions above the substrate surface to be most favorable. In the N-rich case, the clean GaN(0001) surface is the most stable $1\ifmmode\times\else\texttimes\fi{}1$ configuration. For the $\mathrm{GaN}(0001\ifmmode\bar\else\textasciimacron\fi{})$ surface, our results for both Ga- and N-rich growth conditions indicate that a full monolayer of Ga adatoms adsorbed in on top positions is the most stable configuration. Our theoretical results allow for a comparison of full calculations of the surface electronic structure for a number of optimized structural models with most recent angle-resolved photoemission spectroscopy data.

Surface morphology of GaN grown by molecular beam epitaxy

The surface morphology of GaN(0001) grown on Si(111) by molecular beam epitaxy using ammonia has been studied by near-field microscopy techniques. Two distinct morphologies are observed, depending on the growth kinetics. When using Ga-rich growth conditions, the roughness is independent of the epitaxial layer thickness, while it increases with growth time under N-rich growth conditions. Also, the morphological pattern is different for the two regimes of growth and is not related to the crystallographic subgrains delimited by edge dislocations. However, under Ga-rich growth conditions, screw-type dislocations give rise to a spiral growth mode that imposes both the morphological pattern and the surface roughness.

The effect of the first GaN monolayer on the nitridation damage of molecular beam epitaxy grown GaN on GaAs(001)

The initial molecular beam epitaxy growth of GaN on GaAs(001) was studied by real-time monitoring of the (3×3) surface reconstruction and its transition to an unreconstructed (1×1). Various growth conditions were established by variation of the V/III ratio, i.e., the Ga flux. We characterized the effect of the first two strained GaN monolayers: a N-terminated GaN (3×3) monolayer and a second unreconstructed (1×1) monolayer. A series of samples were grown under N-rich, Ga-rich, and near-stoichiometric growth conditions. The resulting morphology of the interface region was analyzed by high-resolution scanning electron microscopy, Auger-electron spectroscopy, and double crystal x-ray diffractometry. The N-rich and Ga-rich conditions resulted in extensive defect formation due to the nitridation damage of the GaAs substrate. The extent of this was found to be determined by the properties of the first GaN monolayer. The surface roughness under optimum growth conditions could be as low as ∼20 nm, defined by nanocrystalline grains, showing no observable nitridation damage.

Study of the correlation between GaN material properties and the growth conditions of radio frequency plasma-assisted molecular beam epitaxy

The material properties of GaN thin films grown by radio frequency (RF) nitrogen plasma source molecular beam epitaxy (MBE) on (0001) Al 2O 3 substrates have been correlated to the V/III flux ratio during GaN growth and to the type and thickness of the buffer layer. The most remarkable observation is the change in the sign of the residual strain, from tensile to compressive as the V/III ratio alters from N-rich to stoichiometric (or slightly Ga-rich) conditions for GaN layers with a 17 nm AlN buffer layer. The residual strain was significantly reduced for a thinner 5 nm AlN buffer and it was zero for a 20 nm GaN buffer. A reduction of the rms surface roughness from 20 to 3 nm was achieved by decreasing the V/III ratio. Finally, stacking faults were observed only for significantly N-rich growth conditions.

Growth of hexagonal GaN on Si(111) coated with a thin flat SiC buffer layer

Hexagonal GaN films were grown on Si(111) covered with a thin flat SiC buffer layer under both N- and Ga-rich growth conditions. A flat 2.5-nm-thick SiC layer was an effective buffer layer for GaN growth. The growth mode and microstructure of GaN depended strongly on the Ga/N flux ratios. Under N-rich growth conditions, the growth mode was three dimensional; GaN showed statistical roughening of the surface and a characteristic columnar structure. Under Ga-rich conditions, the GaN growth mode was two dimensional; GaN films with a flat surface and an almost stacking-fault-free microstructure were obtained. The two-dimensional growth mode was facilitated by strong wetting between Ga and SiC(111) at the first Ga-layer deposition on SiC.

Mixing Mechanism of h-GaN in c-GaN Growth on GaAs (001) Substrates

Systematic investigation of the V/III ratio dependence in the mixing of the h-GaN phase to the c-GaN layers grown on GaAs (001) substrates has been performed and a drastical change of the mixing nature in the initial process is reported. The h-GaN mixing nature on the {111} facets during growth strongly depends on the V/III ratio in the initial buffer layer formation. The h-GaN grain-mixing only occurs on {111}A facets under a Ga-rich formation condition, but the h-GaN mixing grains on the {111}B facets gradually appear together with the vanishing of h-GaN grains on the {111}A facets during the annealing process under rf-nitrogen beam irradiation. The mixing processes under N-rich and Ga-rich growth conditions are also discussed. The results indicate that it is possible to grow high quality c-GaN layers by control of the mixing at the initial growth stage through the effective V/III ratio dependence.

Mixing Mechanism of h-GaN in c-GaN Growth on GaAs (001) Substrates

Systematic investigation of the V/III ratio dependence in the mixing of the h-GaN phase to the c-GaN layers grown on GaAs (001) substrates has been performed and a drastical change of the mixing nature in the initial process is reported. The h-GaN mixing nature on the {111} facets during growth strongly depends on the V/III ratio in the initial buffer layer formation. The h-GaN grain-mixing only occurs on {111}A facets under a Ga-rich formation condition, but the h-GaN mixing grains on the {111}B facets gradually appear together with the vanishing of h-GaN grains on the {111}A facets during the annealing process under rf-nitrogen beam irradiation. The mixing processes under N-rich and Ga-rich growth conditions are also discussed. The results indicate that it is possible to grow high quality c-GaN layers by control of the mixing at the initial growth stage through the effective V/III ratio dependence.

Growth Kinetics of GaN in Ammonia Atmosphere

The kinetics of GaN growth by MBE using ammonia as the reactive nitrogen source is studied both under Ga-rich and N-rich conditions. It is shown that adsorption site blocking by Ga and N atoms as well as by surface NHx complexes is a crucial factor to control the growth rate of the crystal. Use of N-rich growth conditions easily achieved in ammonia MBE allows one to increase the GaN growth temperature at least by ≈︂80 to 90 K compared to plasma-enhanced MBE, that is favorable to improve optical properties of the grown material.

Mg doping studies of electron cyclotron resonance molecular beam epitaxy of GaN thin films

A modified electron cyclotron resonance plasma source, allowing optically active GaN thin films at growth rates up to 1 μm per h, was utilized in a molecular beam epitaxy environment for Mg doping studies of GaN. The effect of the growth parameters on the Mg incorporation was studied using secondary ion mass spectroscopy and Hall measurements. We found that Mg does not incorporate when GaN is grown under Ga-rich conditions, independently of the Mg flux, in the temperature range of 700–800 °C. Only N-rich growth conditions allow measurable Mg incorporation. This incorporation increases with decreasing Ga cell temperature. Furthermore, under N-rich growth conditions, the Mg content is more sensitive to the lowering of the substrate temperature than to the increase of the Mg cell temperature.

The growth rate evolution versus substrate temperature and V/III ratio during GaN MBE using ammonia

The growth rate evolution versus V/III ratio and substrate temperature was studied by means of optical reflectivity during MBE of GaN layers using NH3 as nitrogen source. The GaN desorption becomes observable at temperatures above 800°C and causes the reduction of growth rate accompanied with the surface roughening at temperatures above 850-870°C. Unlike GaAs, which evaporates in accordance with the action mass law, the desorption rate of GaN is found to be almost independent of V/III ratio within the N-rich growth conditions. The activation energy for GaN desorption during the growth is found to be (3.2±0.1)eV. This value is very close to the activation energy for free evaporation. At V/III ratio values exceeding 200 the GaN growth rate reduction caused by violation of the molecular flow regime is observed. The Mg-doped samples grown under these extreme conditions tend to have improved acceptor activation and thus p-type conductivity.

Amorphous Domains in GaN Layers Grown on 6H-SiC by MBE

AbstractUsing N-rich growth conditions, amorphous domains were formed in the initial stage of a cyclotron assisted MBE of GaN grown over 6H-SiC (0001) without any buffer layer. These domains are facetted along {1010} prismatic planes. Their density is about 1010cm-2 and they are mainly located in a 50 nm layer close to the substrate surface.

Initial Growth Prismatic Domains in Cyclotron Assisted MBE of GaN/SiC

In N-rich growth conditions, prismatic domains were formed in the initial stage of a cyclotron assisted MBE of GaN over 6H-SiC (0001). They exhibit {10 0} facets and are either voids or amorphous phase. Their density is of a few 109 cm−2 and they are located in a 50 nm layer closest to the substrate surface.