Phosphorus Overpressure Research Articles

Diffusion and activation of Zn implanted into InP:S

The aim of this work is to compare carrier concentration, diffusion coefficient and shape of the zinc atomic profile in InP layers treated under different annealing conditions. Zn implanted InP:S (4×10 18 at/cm 3—active dopant) bulk samples were investigated. Samples were implanted to a fluence of 1×10 16 cm −2 at an energy of 150 keV and then coated with a 100 nm thick AlN polycrystalline film, using sputter deposition followed by annealing at temperatures in the range 500–900 °C for various different conditions. After removal of the AlN films the extent of diffusion of the implanted species was characterized using the SIMS (Secondary Ion Mass Spectrometry) technique whilst carrier concentration was evaluated by C– V profiling. The depth profiles of in-diffused zinc strongly indicated that the diffusion coefficient D was concentration-dependent. The diffusivity of Zn was found largest when InP was annealed with phosphorus overpressure combined with the AlN cap; in case of annealing at 700 °C a Zn concentration of 2×10 19 cm −3 was achieved. The diffusion range under phosphorus overpressure was almost twice that in vacuum and this is interpreted in terms of the generation of additional vacancies in the In sub-lattice. C– V measurements showed clearly that application of protective AlN together with phosphorus overpressure causes sulphur donors compensation and results in acceptor concentration of 1×10 18 cm −3 and consequently, in the formation of a well-defined p–n junction. The activation efficiency of Zn atoms can be estimated to be about 30%, without additional heat treatment. The Boltzmann–Matano analysis was employed to evaluate the concentration-dependent diffusion coefficient of Zn in InP.

Investigation of compensation defect centres in semi-insulating InP crystals

Laplace transform photoinduced transient spectroscopy (LTPITS) has been applied to study defect, centres in Fe-doped and undoped semi-insulating (SI) InP. A high resistivity (∼ 2 x 10 7 Ω cm) of the latter wins achieved by annealing at 950 °C for 40 h under a phosphorus overpressure. It is shown that shallow donors in this material have an activation energy of 10 meV and are mainly compensated with deep acceptors characterised by activation energies of 350 and 470 meV with respect to the bottom of the conduction band. In the Fe-doped material, the shallow donors are compensated with Fe-related deep acceptors having activation energies of 590 and 640 meV.

Wafer bonding with low-temperature-grown (Ga,P) as an adhesive material

Amorphous and polycrystalline (Ga,P) grown with the molecular beam epitaxy technique at low temperatures were developed as adhesive materials for wafer bonding technology. The microstructure and V/III incorporation ratio in (Ga,P) were studied with transmission electron microscopy (TEM) and Auger electron spectroscopy. The crystallinity of the low-temperature-grown (LTG) (Ga,P), amorphous or polycrystalline, is controlled by the phosphorus overpressure during growth. The n-type GaP substrates, each with LTG (Ga,P) material on top, were strongly bonded face-to-face at 600 °C for several hours. TEM shows that along the bonded interface recrystallization of the (Ga,P) layer has taken place during high-temperature annealing and hence helped the two wafers bond together. In comparison, direct wafer bonding of GaP substrates without LTG (Ga,P) requires high bonding temperatures above 750 °C. The current–voltage (I–V) characteristics of the bonded samples were also investigated. Samples bonded with Ga-rich polycrystalline (Ga,P) exhibit linear I–V dependence across the bonded interface. In contrast, samples bonded with P-rich amorphous (Ga,P) show back-to-back Schottky I–V characteristics with a breakdown of about 11–13 V. The two bonded interfaces show great similarity in microstructure except that gallium clusters are present only at the interface of the sample bonded with Ga-rich polycrystalline (Ga,P). We attribute the linear ohmic I–V dependence to the presence of Ga-clusters.

Mass transport: semiconductor microstructure fabrication by surface energy

Surface energy of an etched microstructure can induce mass transport during heat treatment and transform the structure. This phenomenon has been utilized to fabricate useful device structures for optoelectronic applications. The material process has been systematically investigated especially in the fabrication of large-numerical-aperture microlenses using deeply etched preforms. A reactor has been developed for a contamination-free process with controlled high phosphorus overpressure. Strong interactions between the phosphorus vapor and the wafer surface have been uncovered and investigated. Simple techniques have been developed for effectively maintaining the group III vapor equilibrium for complete wafer protection against erosion and defect generation. Comprehensive one- and two-dimensional models of the mass transport kinetics have been developed and have yielded useful guiding principles. These developments have led to reproducible fabrication of near perfect powerful microlenses and have opened considerable possibilities for more applications.

Annealing Behavior of the Hydrogen-Vacancy Complex in Bulk Indium Phosphide Crystals

ABSTRACTIn order to explain the effects of hydrogen on the electrical properties of bulk indium phosphide crystals, we have performed a series of high temperature annealing studies with both undoped and iron-doped indium phosphide crystals. Our samples were annealed at 900°C for 6, 36, and 72 hours, respectively, under a phosphorus overpressure of five atmospheres. Samples were characterized at 10 K by Fourier transform infrared absorption spectroscopy which allowed us to measure the concentrations of both the Fe2+ and VIn-H4 defects simultaneously. Undoped samples were further characterized by the Hall effect measurements. We find in the iron-doped samples that the [Fe2+]/[Fe3+] ratio decreases gradually with increasing annealing time, indicating a reduction in the number of donors in the samples. In the undoped samples, annealing leads to a reduction of the free electron concentration accompanied by an increase in the 77 K mobility. The increase of the sample's mobility eliminates the possibility that the reduction of the free electron concentration is due to an increase in the concentration of the compensating acceptors. Our explanation for the observed behavior in all samples is that hydrogen acts as a donor and it diffuses out of the crystal during the annealing process. Based on our experimental data, we propose a calibration equation of [VIn-H4] = 4.2×1016 cm−1 × Absorbance (cm−1) which is used to correlate the hydrogen-vacancy complex concentrations with the changes of the VIn-H4 absorption peak in both the iron-doped and the undoped samples. Our results confirm the donor nature of the hydrogen-vacancy complex and provide strong evidence regarding the reduction mechanism of free carrier concentrations in bulk indium phosphide crystals during high temperature annealing under a phosphorus atmosphere.

Desorption of ultraviolet-ozone oxides from InP under phosphorus and arsenic overpressures

The thermal desorption of ultraviolet-ozone oxide on InP substrates prepared for molecular-beam epitaxy has been performed with overpressures of P2, As2, and As4. Surface analysis using reflection high-energy electron diffraction, Auger electron spectroscopy, and x-ray photoelectron spectroscopy and thermodynamic calculations indicate that thermal desorption proceeds via a reaction between the oxide and atomic phosphorus from the substrate to produce volatile phosphorus oxides such as P2O3. The overpressure species serves to stabilize the substrate against surface dissociation once the oxide is removed. In the case of an arsenic overpressure the desorption of the final monolayer of oxide is slowed, relative to the case of phosphorus overpressure, due to the formation of InAs.

Incorporation of zinc in MOCVD growth of Ga 0.5In 0.5P

Data are presented for the Zn doping of Ga 0.5In 0.5P, showing that the hole and zinc concentrations increase almost linearly with zinc flow, and also increase with the V/III ratio and with growth rate at a fixed V/III ratio. These observations are consistent with other reports that show the incorporation of zinc to increase with V/III ratio for both GaAs and Ga 0.5In 0.5P deposition by metalorganic chemical vapor deposition (MOCVD). The growth-rate dependence of the zinc incorporation in Ga 0.5In 0.5P has not previously been reported. A model based on varying group V coverage of the step where zinc is most strongly bound is presented and compared with the data. The model predicts that the zinc incorporation increases with increasing phosphorus overpressure, but should be independent of group III overpressure in the parameter space investigated here.

Study of the influence of the phosphorus pressure on the preparation of nominally undoped semi-insulating InP wafers

The conditions for the preparation of nominally undoped semi-insulating (SI) InP wafers by annealing under controlled phosphorus pressure are described. It is demonstrated by the results of electrical profile measurements (differential Hall effect) and by photoluminescence that diffusion effects in a thin peripheral layer (≊20 μm) can be correlated to a phosphorus in- and indium out-diffusion. But diffusion of these species is correlated to a donor behavior and relatively slow. It can, therefore, not be used to explain the SI bulk property which seems not to depend on the phosphorus overpressure during annealing.

Phosphorus-overpressure rapid thermal annealing of indium phosphide

A rapid thermal annealing technique for InP is described in which a controllable phosphorus overpressure, generated by heating red phosphorus, suppresses the dissociation of InP at required annealing temperatures. Two annealing configurations were used to independently study the effects of phosphorus overpressures, anneal temperatures and gas flow rates on the post-anneal electrical and morphological properties of low dose Si-implanted InP:Fe. The advantage of phosphorus overpressure annealing over close-contact annealing is shown, and comparison is made with Si3N4 encapsulated annealing. Gas flow velocities close to the sample are found to significantly affect the surface morphology, and a static layer immediately above the sample is found to be beneficial.

Electronical properties of metal-insulator-semiconductor devices prepared on thermally treated InP in phosphorus overpressure

The effect of excess phosphorus vapor during the process of MIS structures on InP is analyzed in detail. It is shown that the density of interface state distribution is considerably modified after in situ thermal precleaning of the substrate in phosphorus overpressure (suppression of the Fermi-level pinning and significant reduction of the density of surface states near midgap). Phosphorus overpressure also provides an efficient protection of the surface during the insulator deposition. Fermi-level pinning in the upper part of the gap is assigned to the uncontrolled native oxide left on the surface after a stay of InP substrate in ambient atmosphere.

Influence of growth conditions on undoped and sulfur-doped InP grown by molecular-beam epitaxy

Undoped and sulfur-doped InP have been grown by molecular-beam epitaxy. Undoped InP is n type and contains residual sulfur incorporated from the phosphorus source material. Increasing substrate temperatures during growth cause a decrease in the residual doping level to 2–3×1015 cm−3 but are associated with an increase in the compensation ratio and an increase in the concentration of epilayer defects, up to 105 cm−2. Doping with sulfur produced from an electrochemical sulfur cell has been studied in the range 5×1016–6×1019 cm−3. Highly doped layers show greatly improved surface morphology compared to low and undoped InP layers. Sulfur is shown not to diffuse to any measurable extent during MBE growth; however, at high growth temperatures (530 °C) there is loss of sulfur as a volatile indium sulfide. The removal of the surface oxide from the InP substrate before growth has been studied as a function of substrate temperature and phosphorus overpressure. The oxide layer can be removed at a low temperature (∼480 °C) by using a low pressure of P4 rather than P2 as is used during growth. The low-temperature oxide removal leads to improved epilayer morphology for undoped layers. The removal of the surface oxide and the loss of sulfur during growth as a volatile sulfide are discussed in terms of the free energy of formation of the various possible products.

Capless annealing of InP for metal-insulator-semiconductor field-effect transistor applications

A simple capless annealing process for post-implantation annealing of InP material is described. The technique incorporates a simple boat design and uses InP+Sn as the source of phosphorus overpressure. Using this process, Si-implanted (4×1012 cm−2 dose) InP layers show mobilities at room temperature and 77 K in the range of 3200 and ∼10 000 cm2/Vs, respectively. Dopant depth profiles with peak donor densities of 2×1017 cm−3 and minimal redistribution of impurities were obtained. Depletion-mode InP metal-insulator-semiconductor field-effect transistors fabricated with 4-μm gate lengths using ion-implanted channel and source-drain regions, show channel mobilities of 2500 cm2/Vs.

Growth of Single‐Crystalline Epitaxial Group II Fluoride Films on InP ( 001 ) by Molecular‐Beam Epitaxy

Twin‐free, single‐crystalline, lattice‐mismatched as well as lattice‐matched, epitaxial dielectric films of group II cubic fluorides (, , and ) have been grown on substrates by molecular‐beam epitaxy. The surface was cleaned in vacuum by heating under phosphorus overpressure until a well‐ordered and stoichiometric surface was obtained. The film growth of ,where M is Ba, Sr, Ca, or , was followed by reflection high energy electron diffraction (RHEED). The diffraction patterns indicate a parallel epitaxial relationship: and. At low growth temperature (∼250°C) and low growth rate we could obtain twin‐free single‐crystalline films. Transmission electron microscopy show the absence of any grains. At higher growth temperature (∼350°C) we could obtain twin‐free single‐crystalline , , and films. The latter can be lattice‐matched to at room temperature.

Growth of InP and InGaAsP (E g≥1.15 eV) layers by liquid phase epitaxy under phosphorus overpressure

A method is presented for the liquid phase epitaxial growth of stoichiometric InP and InGaAsP (Eg⩾1.15 eV) epitaxial layers by creating a phosphorus overpressure in the growth ambient. The effectiveness of the method and the stoichiometry of InP layers can be monitored with the photoluminescent spectra of a phosphorus vacancy-impurity complex.

Application of the phase diagram to the diffusion of Zn in GaP

A series of experiments is described in which radio-tracer zinc was diffused into GaP at 900 degrees C in a closed ampoule. Various weights of zinc and phosphorus were used in the ampoule for the different experiments. Diffusion profiles were plotted. The profiles are not simple, i.e. they cannot be described by a single value of diffusion constant. It is shown that the effect of increasing the phosphorus over-pressure is to decrease the overall penetration of the profile. This result, taken together with the shape of the profiles, is consistent with the proposal of previous workers that zinc diffuses in GaP by a substitutional-interstitial mechanism. The previously determined Ga/P/Zn phase diagram for 900 degrees C is discussed, and it is shown that if it is used in conjunction with the available vapour-pressure data, each of the diffusions can be placed on the phase diagram. The vapour pressures of zinc and phosphorus in the ampoule can then be calculated for each of the diffusions. It is shown that the results suggest that most of the zinc occupies the gallium site as an ionized acceptor. The results also allow an estimate to be made for the form of the GaP/Zn solidus line at 900 degrees C.

Zinc Diffusion into GaAs1-xPx by Ga-P-Zn Ternary Alloy Source

In order to form the p-n junction in Te-doped GaAs1-xPx (x=0.4) grown by a Ga-As-PCl3-H2 epitaxial system, zinc diffusion by a Ga-P-Zn ternary alloy source is found to be useful. Diffusion is performed by an ampoule diffusion method, and composition of the diffusion source is calculated from the Ga-P-Zn ternary phase diagram. This diffusion source provides high phosphorus overpressure and leads to smaller diffusion coefficients and more planar p-n junctions. Diffusion depth and junction planarity might be less dependent on the GaAs1-xPx surface morphology and the degree of crystal perfection. Junction profile is a mixed abrupt-graded type from Capacitance-Voltage measurement. In the result of X-ray micro-probe analysis, a “window effect” is simultaneously revealed by the Ga-P-Zn system diffusion.

Zinc Diffusion into Gallium Phosphide under High and Low Phosphorus Overpressure

The acceptor zinc has been diffused into n‐type single crystal gallium phosphide at temperatures between 700° and 1000°C from diffusion sources which lie in two different regions of the Ga‐P‐Zn phase diagram. One source (pure zinc) provides low phosphorus overpressure and leads to p‐n junctions which are distinctly nonplanar. The other (zinc plus excess phosphorus) provides high phosphorus overpressure and leads to smaller diffusion coefficients and considerably more planar p‐n junctions. These effects are consistent with the interstitial‐substitutional model for diffusion of zinc in . With either source, diffusion induced dislocations are formed throughout the diffused layer after an incubation time. The p‐n junction depth corresponding to the first appearance of the induced dislocations is found to be greater when the high phosphorus pressure source is used.