Osamu Ryuzan

Residual Donors in High Purity Gallium Arsenide Epitaxially Grown from Vapor Phase

Residual impurities involved in a GaAs layer epitaxially grown from vapor phase are identified by far-infrared photoconductivity measurement. Silicon, selenium, sulfur and germanium are readily incorporated as shallow donors into GaAs epitaxial layers. The dominant residual donor is sulfur for epitaxial layers grown by Ga-AsCl3-N2 system and it is silicon or sulfur for epitaxial layers by usual Ga-AsCl3-H2 system. From spectral peaks of the photoconductivity in a magnetic field, the donor ionization energies for silicon, selenium, sulfur and germanium are estimated to be 5.759 meV, 5.812 meV, 5.845 meV and 5.949 meV respectively.

Photoluminescence Study of Carbon Doped Gallium Arsenide

The photoluminescence studies of the carbon doped epitaxial gallium arsenide revealed that a pair of emission bands at 1.493 eV (B-A) and 1.490 eV (D-A) were attributed to the carbon acceptor on arsenic site. The sharp doublet emission lines at 1.5127 eV and 1.4937 eV were identified with the exciton recombination bound to a neutral carbon acceptor and the two-hole transition in which the neutral carbon acceptor is left in an excited state, respectively. From the observation of these emissions, the 1S3/2 ground and 2S3/2 excited state energies of carbon acceptor were found to be 26.5 meV and 7.5 meV above the valence band edge, respectively. The results were supported by the effective mass arguments and by the dependence of the binding energy on the atomic number of impurities.

Direct observation of defects in Si-doped and Ge-doped Ga0.9Al0.1As epitaxial layers by transmission electron microscopy

Abstract Defects in Si-doped and Ge-doped Ga 0.9 Al 0.1 As epitaxial layers are observed and compared by transmission electron microscopy (TEM). Dislocations and stacking faults which originate from the interface are observed in both kinds of epitaxial layers. Faulted loops are locally observed in Ge-doped layers but not observed in Si-doped layers. These loops are extrinsic in character and have Burgers vectors of the type a /3〈111〉. The density of the faulted loops tends to increase with increasing carrier concentration in a range of 1 × 10 17 − 9 × 10 18 cm −3 .

Characterization of the interface region in VPE GaAs

“Absolute” electron mobility values in the substrate interface region of n-type GaAs epitaxial layers were investigated. Space-charge scattering was found to be significant near the interface but could be minimized by the sequential growth of layers in one step. Our experimental results suggest that local deviations from stoichiometry may be responsible for the space-charge scattering.

Hot Electron Effect in Short n^+nn^ GaAs

Current-voltage characteristics are measured by applying pulse voltages in short n+nn+ GaAs with a channel length of 1 µm. The observed non-ohmic behavior is interpreted in terms of the hot electron effect, the analysis being made by taking into account the non-uniform electric field distribution due to the boundary conditions of the short semiconductors. The results reveal that the dependence of the electron mobility on the electric field is not simple as in n-Si, but is complicated by the transfer of hot electrons to the higher-lying conduction valleys. The analysis gives the spatial distribution of the electron drift velocity, indicating the existence of drift velocity overshoot. The observed maximum drift velocity is about 3.2×107cm/s.

GaP Liquid Phase Epitaxial Growth Using a Vertical Furnace System

GaP liquid phase epitaxial growth using a vertical furnace system, in which hydrogen sulphide and zinc vapor are used as doping sources, has been studied. The Henrys law constant of sulphur to gallium in the temperature range of 1000–1080°C is about 9.3, and zinc to gallium in the temperature range of 950–1030°C is about 3.6×10-2. n- and p-type epitaxial layers are grown consecutively in the same gallium solution by using this method and GaP light emitting diodes are fabricated.

Magneto-optical studies on excitons bound to deep donors with strong axial symmetry in gallium arsenide

Abstract New emission line due to a recombination of exciton bound to a deep impurity state was found at 1.5037 eV in epitaxial GaAs. Strong dependence of the Zeeman spectral lines on the magnetic field direction and temperature was observed and also polarizations of some emission lines were detected. These results suggest that the center is a deep donor and has strong axial symmetry. An excellent theoretical fit with the experimental results is obtained. Relevant splitting factors are g e = -0.44 (electron) and g eff = 0.80 (hole).

Nitrogen doping of GaP during LPE growth—A kinetic study

The emission spectra of GaP light-emitting diodes have been measured in order to study the rate at which the N concentration in GaP grown by liquid-phase epitaxy changes after a change in the NH 3 content of the ambient H 2 gas stream. The response rate was fairly rapid for the growth conditions used. A steady state was reached about 20 min after NH 3 was introduced and about 5 min after the NH 3 flow was terminated.

New developments in III–V transistor technology

Abstract Described in this paper is the main technology breakthrough brought in these years which has caused a rapid advance in III–V compound, especially GaAs, transistor and IC. It is promising that they will acquire the territory where they play an important role in semiconductor industry.

Effect of boundary conditions on hot electron transport in short semiconductors

Abstract Importance of the boundary conditions in short semiconductors is demonstrated by solving current continuity and Poissons equations iteratively. A new method to obtain drift velocity overshoot experimentally in short n+n n+ GaAs samples is described, which is based on the measurements and analysis of the current-voltage characteristics by taking into account of the boundary conditions. The observed overshoot velocity is about 7×107 cm/s at 3V application to GaAs sample with 1μm channel length. The excess electrons near the n+ contacts were found to result in J ∞ V 3 2 relation in very short samples assuming drift motion for electrons. Therefore an observation of J ∞ V 3 2 is not a good measure for the criterion of the ballistic transport.