Low dose (5x109-5x1011 cm−2) ion implantations of several ions including C, Be, Mg, and Mn have been performed into high purity epitaxial and bulk InP samples. A comparison of the low temperature (1.7-20 K) photoluminescence spectra of these deliberately doped samples was made to similar spectra of undoped high purity InP grown by LPE, PH3-VPE, and LEC techniques in order to identify the residual acceptors in the undoped samples. Ionization energies obtained for the C, Be, and Mg acceptors (assuming Eg=1.4237 eV) were 44.6 ± 0.3, 41.3 ± 0.3, and 41.0 ± 0.3 megV, respectively. These valu were compared to the ionization energy of the dominant residual acceptor in LPE InP, which also occurs in LEC, polycrystalline, and PH3-VPE material. This acceptor level was first reported in LPE InP by Hesset al. in 1974, who denoted it A1; it has widely been assumed to be C. The value we measure for the ionization energy of A1, which is 41.2 ± 0.3 meV, however, matches that of Mg and Be andnot C. We conclude that C is almost never present as a residual acceptor in undoped InP, and that the residual level in LPE and other material is Mg, or possibly Be. Similar measurements on Si and Sn doped samples showed no evidence of any Si or Sn acceptor level. The Mn acceptor peak was observed to occur at 1.19 eV.

The ionization energies of C, Be, and Mg acceptors in InP have been determined by means of low-temperature (1.7–20 K) photoluminescence measurements on high purity epitaxial and bulk samples which have been implanted with low (5×109–5×1011 cm−2) doses of those impurities. The measured values are 44.6±0.3, 41.3±0.3, and 41.0±0.3 meV, respectively. A comparison with the ionization energies of the residual acceptors in liquid phase epitaxial (LPE), PH3-vapor phase epitaxial, liquid encapsulated Czochralski, and polycrystalline samples indicates that C is almost never present as a residual acceptor in undoped InP. The ionization energy of the main residual acceptor in the LPE samples matches that of both Mg and Be.

Methods for protecting GaAs and InP surfaces against degradation during annealing are reviewed. Various thin film encapsulants are discussed, including SiO2, Si3N4 and doped glasses. A discussion of deposition methods and thin film evaluation techniques is presented.

Using Auger electron spectroscopy (AES) and secondary ion mass spectrometry (SIMS) we examined the InP encapsulation properties of chemically vapor-deposited SiO2, chemically vapor-deposited phosphosilicate glass (PSG) and r.f.-plasma-deposited Si3N4. Following 60 min 750°C anneals, indium is detected by both AES and SIMS on the surfaces of Si3N4 encapsulants. SIMS results also suggest that indium may out-diffuse through PSG caps during similar anneals. Silicon in-diffusion from the cap to the underlying InP is found to be significant in annealed SiO2-and Si3N4-capped samples. Little such silicon contamination of InP is observed after annealing with PSG encapsulation.

For light mass projectile implantations of InP that are highly dominated by electronic stopping, the standard Gaussian approximation provides a poor fit to the observed experimental profile. Using four moments based on theoretical considerations we have constructed profiles using a terminated Edgeworth and Pearson I distribution. It is found that Edgeworth and Pearson I distributions constructed with four moments derived from theoretical values are not significant improvements over the standard LSS Gaussian. The moments of the experimental profiles have therefore been empirically determined. These moments together with the Person IV distribution are shown to fit experimental profiles over at least two orders of magnitude.

Low-temperature (5 °K) photoluminescence data on the annealing characteristics of both unimplanted and implanted (4He, 9Be) semi-insulating, Fe-doped (100) InP are presented. A rf plasma-deposited Si3N4 was used as the encapsulant during anneals. Annealing with this cap results in the appearance of a new spectral feature at 1.378 eV. For all samples examined here, high temperature (T?750 °C) anneals result in band-edge integrated intensities greater than in virgin material. This phenomenon is correlated with the appearance of the 1.378-eV peak. In samples implanted with 100 keV 9Be the band-edge peak is observed to shift to lower energies with increasing dose. Also, a new emission peak at 1.382 eV is shown to be associated with Be acceptors. The ionization energy of Be is estimated to be 41.3±3 meV. For 30-min anneals, maximum Be activation appears to occur for all doses at ∼750 °C. At this temperature, Be activation appears to saturate between 30 and 60 min.

The migration of implanted 9Be in (100) semi‐insulating during thermal annealing has been studied using secondary ion mass spectrometry. Under typical annealing conditions for (, ) we have observed that implanted 9Be is a rapid diffusant in semi‐insulating for fluences as low as . Redistribution of the compensating impurity (Fe or Cr) has also been observed. In several respects Cr redistribution differs from that of Fe. Twin‐peaked structures appear in the impurity profiles of 550°C anneals of 100 keV, 1015 cm−2 9Be implanted materials. Models for this phenomenon are discussed. Correlations are noted between 9Be and compensating impurity profiles in annealed, high fluence implanted samples. Flat tails of 9Be extending over several microns are observed in semi‐insulating .

Using secondary ion mass spectrometry (SIMS) the annealing characteristics of amorphizing implants of Mg and Si in semi‐insulating have been examined. Substantial redistribution of Mg and Fe occurs during 30 min anneals of implanted with 1015 cm−2, 250 keV Mg. Atomic profiles indicate that Mg and Fe are gettered out of the amorphous zone into an implant damaged, bulk region. Examination of electron channelling patterns (ECP) from samples annealed for 30 min at 550° and 650°C show no discernible patterns at the surface or at depths near the theoretical damage profile peak. Weak channelling patterns are visible at the calculated peak damage depth, however, in samples annealed at 750°C for 30 and 60 min. Flat tails of Mg extending over a distance of about 1 μm are seen in the bulk. Little redistribution of Si occurs during 30 min, 750°C anneals of , 240 keV Si implants. Redistribution of Fe is observed in these Si implanted samples, however, resulting in an accumulation region near the implanted Si peak.

Using low temperature photoluminescence (PL) we examined samples annealed in controlled atmospheres (vacuum, phosphorus vapor or indium vapor) and also samples annealed with SiO2, Si3N4 or phosphosilicate glass (PSG) encapsulants. PL spectra from samples annealed in the controlled environments indicate that a spectral line at 1.393 eV results from phosphorus-vacancy-related defect luminescence. PSG and Si3N4 caps appear to be comparable in their ability to suppress the formation of substantial 1.393 eV luminescence, but in the spectra of SiO2-capped and annealed samples this feature is more pronounced. High temperature anneals (T ≈ 750 °C) with Si3N4 result in the emergence of a new peak at 1.378 eV and integrated band intensities greater than those observed in the virgin material. From results discussed in this article and those of the preceding article, it appears that, of the three encapsulants, the PSG cap best preserves the characteristics of the encapsulated InP following furnace anneals.

The redistribution of the compensating dopants, iron or chromium, in semi‐insulating indium phosphide has been studied using secondary ion mass spectrometry. Annealing with silicon nitride encapsulation results in impurity accumulation within the first 1000Å of the surface followed by depletion extending to a depth of ∼1 μm. Profiles resulting from the implantation of "neutral" elements (He, B) exhibit accumulation at the surface and also accumulation at the projected range peak. The profiles are explained in terms of gettering of the compensating dopant to defect‐rich regions.

We have examined polycrystalline and single crystal, liquid encapsulated Czochralski (LEC) grown InP for evidence of boron contamination using secondary-ion mass spectrometry (SIMS) and photoluminescence. Precipitates of boron or a boron compound have been found in InP grown by the LEC method with boric oxide (B2O3) encapsulation and pyrolytic boron nitride (pBN) crucibles. The density of precipatates appears to increase towards the last-to-freeze end of the boules. Neither B2O3 or pBN appears to be solely responsible for the observed phenomenon. At present, the source of boron contamination most consistent with our observations is the result of interaction between B2O3 and pBN. A diffusion coefficient of D(750°C) < × 10-14 cm2 s-1 for boron in InP has been determined in this work.

We observed cracking in chemically vapor-deposited SiO 2 encapsulating layers on InP when these samples were annealed at temperatures above 650°C. Using optical microscopy, scanning electron microscopy and Auger electron spectroscopy we studied the detailed nature of this form of film failure. Thermally induced stress between the film and the InP does not fully account for the phenomena reported here. The observed cracking apparently results from an interaction between mechanical stress, chemical effects and possibly defects in the SiO 2 layers.

An electron trap with a thermal activation energy of 0.83 eV from the conduction band is common in the deep level transient spectroscopy (DLTS) spectra of vapor phase epitaxial (VPE) n-GaAs, but is not observed in the DLTS spectra of as-grown molecular beam epitaxial (MBE) n-GaAs. We show here that this trap is created during high temperature annealing of MBE samples with a Si3N4, encapsulant. The trap concentration is correlated with the annealing temperature and time, suggesting the outdiffusion of a constituent atom resulting in the formation of a vacancy or vacancy-complex. Other electron traps observed in the DLTS spectra of asgrown MBE n-GaAs are annealed out for temperatures at or above 800° C.

We have observed a radiative transition (1.306 eV) in the photoluminescence spectra of InP which we show is due to near surface states created by surface damage.

The effect of different annealing temperatures in the 450–800 °C range on the photoluminescence of Be-implanted InGaAsP has been studied. The results of these measurements indicate that the annealing temperature should be above 700 °C for optimum lattice recovery. Avalanche photodetectors with leakage currents as low as 1 μA at 100 V and with gains ≳100 at 116 V have been fabricated. The quantum efficiency for these devices is about 65% throughout the 1.00–1.30-μm wavelength range.

In recent years, the SIMS technique has approached significantly closer to the desired role as a universal microanalytical technique with part-per-million sensitivity. Initial successes with boron and other p-dopants in silicon demonstrated the analytical power of the technique for depth profiling of dopant distributions in semiconductors. With the introduction of the cesium ion source [1,2], electronegative elements became accessible to analysis. Recent work has been directed towards extending the applicability of the technique to those elemental species which are neither strongly electronegative or electropositive and exploiting the capability of the technique to perform multi-element analysis in a given material. This paper will review recent progress in two such areas involving III-V compound semiconductors: analysis of nitrogen in gallium arsenide and gallium arsenide phosphide and studies of the redistribution of both the implanted dopant and the bulk compensating dopant during annealing of Be+ ion-implanted, Fe-doped indium phosphide.