N-type GaAs Crystals Research Articles

MeV ion beam processing of III–V compound semiconductors

MeV ion beam processing provides a very promising technology for 3-dimensional device fabrication and property modification of deeply buried interfaces; its application to III–V compound semiconductors will be presented. Using MeV oxygen ion implantation one can produce deeply buried insulating layers in n-type GaAs crystals, and induce compositional disordering in GaAs/GaAlAs superlattices. It has been employed, as a simple and promising technique, for fabricating high efficiency single quantum well GRINSCH GaAs/AlGaAs lasers. Formation of deeply buried high resistivity layers in n-type InP has also been investigated. A comprehensive study of MeV-ion-implanted InP by a variety of analytical techniques has provided a coherent picture of implanted distribution, structural transition, crystalline damage, and lattice strain in the implanted InP crystals, and has led to a good understanding of the physical processes involved in MeV ion implantation and subsequent thermal annealing. Application of MeV nitrogen ion implanted InP to laser devices will also be discussed.

Evidence for the electron traps at dislocations in GaAs crystals

A new electron trap, labeled ED1, with an activation energy of 0.68 eV has been found to appear in the deep-level transient spectroscopy spectrum of n-type GaAs crystals as a result of their plastic deformation at 400 °C. The trap has been systematically investigated taking into account broadening of its energy level. The trap concentration was proportional to the dislocation density and not affected by the post-deformation annealing at 500 °C. The concentration of electrons captured at the trap was found to depend logarithmically on the duration time of the filling pulse and an acceptor character of the trap was established. The results are analyzed in a model involving barrier-limited capture rate, and it is concluded that the ED1 traps are most probably located in the dislocation core. Possible types of dislocations responsible for the traps are discussed.

Advantages of metallic-amorphous-Silicon-gate FET's in GaAs LSI applications

A Schottky barrier as high as 1 V is obtained for contact between a ternary amorphous film, a-Si-Ge-B, and an n-type GaAs crystal. A metallic-amorphous-silicon-gate FET (MASFET) was made using the amorphous film as a gate contact. GaAs MASFET characteristics are superior to GaAs MESFET characteristics in application to LSI's with a DCFL configuration because the DCFL circuits with the GaAs MASFET's provide a logic level as high as 0.94 V and widen the circuit operation margin. Full operation is obtained from a 1 Kword × 2 bit SRAM with GaAs MASFET's, which is considered to be mainly due to the wide operation margin. The measured propagation delay time of the DCFL inverter is 34 ps at supply voltage V_{DD} = 1.5 V and power consumption of 1.9 mW/gate.

A study of microdefects in n-type doped gaAs crystals using Cathodoluminescence and x-ray techniques

Microdefects were studied in n-type Te, S and Si doped ( n ⩾ 10 18 cm -3) GaAs single crystals. X-ray topography, scanning electron microscopy in the transmition cathodoluminescence mode and X-ray diffuse scattering experiments performed near Bragg peaks by double crystal diffractometry were employed. Evidence for microdefects was obtained in Te doped ( n = 0.9 x 10 18 cm -3), S doped ( n = 3 × 10 18 cm −3) and Si doped ( n = 6 × 10 18 cm −3) crystals. On the other hand, no microdefects were revealed in more lightly Si doped ( n = 3 × 10 18 cm −3) samples, thus evidencing the possibility of obtaining crystals with a high carrier concentration, but free of microdefects, in this latter case.

Effect of 2.2 MeV Electron Irradiation and Annealing on the Non-Copper-Induced 1.35 eV Emission Band in n-Type GaAs

The effect is studied of 2.2 MeV electron irradiation and subsequent annealing on the non-copper-induced extrinsic emission band peaked at hvm (77 K) near 1.35 eV (occurring due to free electronic transitions to (VGaTeAs) VAs pairs) in tellurium doped n-type GaAs. An analysis of the results obtained shows three processes following one after the other: a) radiation-induced decrease in the concentration of 1.35 eV Radiative centres (resulting from the interaction of radiation-induced Ga and As interstitials with (VGaTeAs) VAs pairs), b) annealing-induced (at moderate heating temperatures) increase in the amount of 1.35 eV Radiative centres (resulting from the interaction of radiation-induced Ga and As vacancies with isolated tellurium atoms), and c) annealing-induced (at elevated heating temperatures) decrease in the concentration of 1.35 eV Radiative centres (resulting from the radiation-induced decrease in the thermal stability of (VGaTeAs) VAs pairs). These can well account for the non-trivial features of 1.35 eV emission intensity variations observed in electron irradiated and subsequently annealed n-type GaAs(Te) crystals. [Russian Text Ignored].

EL2 distributions in doped and undoped liquid encapsulated Czochralski GaAs

We compare the longitudinal and radial distributions of EL2 in undoped semi-insulating and intentionally doped n-type GaAs crystals grown by the liquid encapsulated Czochralski technique. Longitudinal profiles in undoped crystals are controlled by changes in melt stoichiometry as the crystal is pulled from the melt. EL2 profiles along crystals doped above about 1×1017 cm−3, on the other hand, are controlled primarily by the carrier concentration as a result of the suppression of EL2 by free electrons. Radial EL2 profiles are typically W shaped and M shaped in undoped and doped (above threshold) crystals, respectively. The origin of these radial profiles is discussed in terms of residual stress, melt stoichiometry, and the suppression of EL2 by electrons. Our results are also discussed in the light of the antisite model for EL2.

GaAs P-N Junction Formation by Carbon Ion Implantation

Carbon ion implantation accompanied with boron ion implantation into n-type GaAs crystals has proved to markedly improve the carbon activation ratio from 4% to 20%. P-n junction diodes fabricated by this technique showed good rectifying characteristics and well-defined profile.

Photoluminesence of Copper-Doped n-Type GaAs after Irradiation by 2.2 MeV Electrons

The 77 K photoluminescence of electron-irradiated (E = 2.2 MeV, Φ = 1016 electrons/cm2) and annealed (T ≦ 650 °C) copper-doped n-type GaAs crystals is studied. A non-trivial increase in the intensities of copper-induced extrinsic emission bands peaked at hv (77 K) near 1.0 and 1.28 eV is found. An analysis of the results obtained shows that the radiation and annealing induced changes in the concentrations of copper-induced 1.0 and 1.28 eV radiative centres (CuGaVAs and CuGaTeAs pairs), resulting from a high-temperature interaction of interstitial copper atoms with radiation-induced defects (gallium vacancies), could well account for the effects observed. [Russian Text Ignored].

Photoluminescence of copper-compensated p-type GaAs after irradiation by 2.2 MeV electrons

The 77 K photoluminescence of electron-irradiated (E = 2.2 MeV, Φ = 1016 electrons/cm2) and annealed (T ≦ 650 °C) copper-doped n-type GaAs crystals is studied. A non-trivial increase in the intensities of copper-induced extrinsic emission bands peaked at hv (77 K) near 1.0 and 1.28 eV is found. An analysis of the results obtained shows that the radiation and annealing induced changes in the concentrations of copper-induced 1.0 and 1.28 eV radiative centres (CuGaVAs and CuGaTeAs pairs), resulting from a high-temperature interaction of interstitial copper atoms with radiation-induced defects (gallium vacancies), could well account for the effects observed. [Russian Text Ignored].

Interaction of Copper Atoms with Radiation Defects in Gallium Arsenide

A study is made of the effect of electron irradiation on the photoluminescence spectra of copperdoped n-type GaAs crystals. It is shown that electron irradiation (Φe = 1016 electrons/cm2) effectively eliminates the copper-induced emission bands peaked at hνm(77 K) ≈ 1.04 and 1.28 eV and restores the emission bands peaked at hνm(77 K) ≈ 0.98 and 1.18 eV, observed in the starting (before the copper diffusion) material, i.e. the copper complexes CuGaVAs and CuGaTeAs are transformed into VGaVAs and VGaTeAs complexes, accordingly, as a result of an electron irradiation. The data presented are interpreted as arising due to an effective ejection of substitutional copper atoms CuGa by radiation-induced gallium interstitials Gai from their original position next to the arsenic vacancy VAs or to the tellurium atom TeAs into interstitial sites (di-interstitials GaiCui, consisting of Ga and Cu interstitials). [Russian Text Ignored.]

Surface photovoltage spectroscopy with cleaved GaAs (110) surfaces: Spectroscopy of Cr2+

Surface photovoltage spectra of clean (110) surfaces cleaved from n-type GaAs crystals doped with Te or Si exhibit two narrow peaks at 0.72 and 0.89 eV superimposed on a broad shoulder extending from h/ω≊0.7 eV to Eg. Both peaks are strongly increased by chemisorption of oxygen and by increasing the band bending. The peak at 0.89 eV is attributed to intra-ion 3d transitions of deep Cr2+ impurities, with the ground state 0.72 eV below the conduction band edge. The presence of chromium in the samples was proven by AES after anneals at 800 K, causing an enrichment of chromium near to the surface. The growth of the 0.89 eV peak proportional to the square of the surface potential is explained by an increase of the net escape probability from the excited state into the conduction band via tunneling.

Quantum efficiency and radiative lifetime in degenerate n-type GaAs

The hole lifetime, the radiative lifetime, the non-radiative lifetime, and the internal quantum efficiency in degenerate n-type GaAs crystals have been investigated with a simplified model of degenerate semiconductors, in which the recombination constant B is approximately proportional to the −13 8 power of the electron density. In n-type GaAs at 77 K, the radiative lifetime reaches a minimum equal to 4 × 10 −9 sec. at 6 × 10 17 cm −3, and the internal quantum efficiency exhibits a maximum equal to 50% at 8 × 10 17cm −3, in good agreement with the theoretical prediction of Dumke and the experiments of Cusano. At high impurity concentrations, the polytropy effect is taken into account in the case of GaAs crystals doped with tellurium and selenium. Finally, it is suggested that, for a given high concentration, the internal quantum efficiency increases with increasing temperature, in accordance with the observed results.

Change in the probability of non-radiative (Auger) electronic transitions in deep radiative centres caused by heat treatment, γ-irradiation, and deformation of GaAs

en

Resonant and localised modes due to boron in gallium arsenide

Silicon-doped n-type GaAs crystals pulled from a melt encapsulated with boric oxide exhibit an IR absorption band at 123 cm −1 as well as a number of localised-mode bands associated with boron and silicon. Evidence is presented which leads us to attribute the band at 123 cm −1 to a resonant mode associated with isolated B atoms on Ga sites. Other crystals containing high concentrations of silicon or phosphorous did not show such an absorption band. The results are discussed briefly in terms of the simple mass defect model for impurity vibrations.

Study of Non-Linear Extrinsic Luminescence in GaAs III. Sublinear Emission Due to Auger Transitions in Radiative Centres

The excitation intensity dependence of the 0.94 and 1.0 eV extrinsic emission in n-type GaAs crystals (n0 ≈ 1017 cm−3) is studied at L ⪅ 1018 quanta/cm2 s and T = 77 K. Linear and sublinear 0.94 and 1.0 eV extrinsic emission is observed depending on the strength of the non-radiative electronic transitions in the radiative centres studied. The main features of the non-linear extrinsic emission (the sublinear emission increase with excitation intensity, a decrease in the internal quantum efficiency of the extrinsic emission and in the luminescence decay time as exciting power is increased) are satisfactorily explained by a model which takes into account an increase in the rate of non-radiative electronic transitions in 0.94 and 1.0 eV radiative centres as the exciting light considerably changes the hole occupancy of defects associated with the centres studied. [Russian Text Ignored.]

Study of non-linear extrinsic luminescence in GaAs. II. The role of auger recombination

The excitation intensity dependence of the 0.99, 1.03, 1.2, and 1.3 eV extrinsic emission in heavily doped n-type GaAs crystals (n0 = 5 × 1017 to 2 × 1018 cm−3) is studied at 77 to 300 K. An analysis is given in which a functional intensity dependence of the extrinsic emission on a rate of simultaneous radiative and non-radiative electronic transitions in radiative centres is considered. Linear and superlinear 0.99, 1.03, 1.2, and 1.3 eV extrinsic emission is observed at low temperatures depending on the strength of non-radiative electronic transitions in radiative centres studied. The main features of non-linear extrinsic emission (superlinear emission increase with excitation intensity, a shift of emission quenching curves to lower temperatures as excitation intensity is increased, a change from the superlinear to a linear excitation dependence of emission intensity as temperature is raised) are satisfactorily explained by a model which includes strong non-linear, in intensity, dependence, radiationless transitions in radiative centres. The model discussed takes into account a decrease in the rate of non-radiative electronic transitions in radiative centres as the exciting light considerably changes the electron occupancy of defects associated with the centres studied; as a result the internal quantum efficiency and emission decay time for the luminescence bands are increased as excitation intensity is raised. [Russian text Ignored]

Study of non-linear extrinsic luminescence in GaAs

The excitation intensity dependence of the 0.94, 1.0, 1.2, and 1.3 eV extrinsic emission in low-resistivity n-type GaAs crystals (n0 = 1017 to 2 × 1018 cm−3) is studied at 77 to 500 K. An analysis is given in which the functional non-linear intensity dependence of the extrinsic emission on the concentration of radiative and non-radiative centres, the free hole excess concentration (the free hole lifetime), and the rate of hole thermal release from radiative centres is considered. The main features for non-linear extrinsic emission observed in experiment (sublinear emission increase with excitation intensity, no luminescence saturation, a shift of emission quenching curves to higher temperatures as the excitation intensity is increased, a change from the sublinear to linear excitation dependence of emission intensity as the temperature is raised, and others) are satisfactorily explained by the accepted model of 0.94, 1.0, 1.2, and 1.3 eV radiative centres. The methods used could well be adopted for the analysis of non-linear luminescence in other semiconductors. [Russian Text Ignored].

Effect of the 0.94, 1.0, 1.2, and 1.3 ev radiative centres on the intrinsic luminescence intensity in n-GaAs

en

Electrical measurements on electron irradiated n-and p-type GaAs

Abstract Hall effect and resistivity measurements have been made on undoped n-type and Cd-doped p-type GaAs crystals irradiated with electrons at 287°K. For 1 MeV irradiations, the dose dependence of carrier removal and mobility in n-type suggests that a defect energy level at E c-0.1±0.015 eV was introduced by irradiation and was the upper level of a doubly charged acceptor. A donor level was also found in n-type at approximately E c − 0.20 eV, and deeper acceptor levels were also introduced. In p-type a similar study located a radiation introduced level at E v +0.1 ± 0.015 eV, which could be a singly or doubly ionized donor. Measurements of the carrier removal rate in p-type as a function of beam energy were in good agreement with those previously reported for n-type, thus confirming the displacement energy of 17–18 eV. It is suggested that the majority of displacement events gave rise to electrically active defects.

Calculation of distribution equilibrium of amphoteric silicon in gallium arsenide

The distribution of an amphoteric impurity silicon over the sublattices of GaAs in equilibrium with the gallium-rich solution is calculated. The change in free energy is estimated for the transfer of silicon atoms from the arsenic sublattice to the gallium sublattice via the external liquid phase. The heat of solid solution for each sublattice is evaluated from the binding energy of the silicon atom in the host GaAs and from the strain energy introduced. The thermodynamic model proposed can well explain the region where the p-type silicon-doped GaAs grows indicated in the gallium-rich corner of the gallium-arsenic-silicon ternary phase diagram. The comparison is made with the distribution equilibria calculated for other amphoteric impurities, germanium and tin, in GaAs. It is found that the silicon is only amphoteric impurity by which both p- and n-type GaAs crystals can be grown at the conventional temperatures for the liquid phase epitaxy. Expressions for the equilibrium concentrations of imperfections in the compensated silicon-doped GaAs are derived as functions of temperature and of silicon activity in the liquid phase. Consistent agreement with the experimental data is obtained.