DLTS and low frequency capacitance measurements were made on Schottky diodes formed on Ion-beam etched (IBE) GaAs surfaces. Ion bombardment causes strong low frequency capacitance dispersion, increasing with ion energy. This is attributed to a thin, heavily damaged region, with a trap density in excess of 1017 per cm3. Such dispersion disappears under weak reverse bias and at reduced temperatures. These and other measurements are consistent with a lumped R-C model having components representing the top damaged layer, the underlying space-charge region, and shunt resistance to the ohmic contact.

A systematic study of the effect of the chemical species, implanted into silicon, on the stability of the residual damage has been performed. Plan-view and cross-sectional transmission electron microscopy (TEM) studies show that the stability of the end of range damage (category II) defects upon annealing depends dramatically upon the implant species. This is exemplified by the a comparison of 69Ga and 72Ge implants in which a decrease in the dislocation density by over four orders is noted for 69Ga implants compared to 72Ge implants after identical annealing cycles. Additional comparisons of species with similar atomic masses indicate that this destabilizing influence on the dislocation loops by the implant species is related to exceeding the solid solubility of the implanted species. As a result of this dislocation loop destabilization effect complete elimination of the dislocation loops can be realized after relatively short thermal cycling. Evidence is presented indicating that the precipitates which form upon exceeding the solid solubility (category V defects) are dissolving during this enhanced defect dissolution process.

Rapid thermal annealing (RTA) of shallow Sb-implanted Si has been studied with Rutherford Backscattering Spectrometry (RBS). Single crystal Si wafers were implanted with Sb at energies of 16, 32 and 48 keV and doses of 5×1014 and 1×1015/cm2. RTA and reference furnace anneals in a nitrogen atmosphere were done to activate the dopant and remove implantation damage. Glancing-angle RBS measurements were used to determine the Sb depth distributions. Dopant profiles obtained with RBS analysis were compared with Secondary Ion Mass Spectrometry results and TRIM code calculations. RBS measurements of the projected range and range straggle did not differ significantly from TRIM code calculations. Following annealing, significant Sb diffusion from the as-implanted peak was apparent. Sb accumulation at the substrate surface was pronounced, especially for furnace-annealed samples.

The promotion of silicide reactions at the interface between silicon and a metal overlayer is described, the reactions being initiated by scanned ion beams. The relative effects of low and high energy Si+ and Si2+ beams are discussed and the results of subsequent annealing are compared with those seen when using low energy (5keV) argon ion beams. The implications for the writing of metallisation lines are also noted.

Transmission electron microscopy experiments have been performed to investigate the lattice damage created by heavy-ion bombardments in GaAs. These experiments have been performed in situ by using the HVEN - Ion Accelerator Facility at Argonne National Laboratory. The ion bcorbardments (50 keV Ar+ and Kr+) and the microscopy have been carried out at temperatures rangrin from 30 to 300 K. Ion fluences ranged from 2 × 1011 to 5 × 1013 ions cm−2.

Direct-inpact amorphization is observed to occur in both n-type and semi-insulating GaAs irradiated to low ion doses at 30 K and room temperature. The probability of forming a visible defect is higher for low temperature irradiations than for room temperature irradiations. The amorphous zones formed at low temperature are stable to temperatures above 250 K. Post implantation annealing is seen to occur at room temperature for all samples irradiated to low doses until eventually all visible damage disappears.

The effect of large dose boron implantation in single crystal CdTe has been investigated by Rutherford Backscattering Spectrometry with channeling (RBS), double crystal x-ray diffraction (DCD) and photoreflectance spectroscopy (PR). Comparisons are made with the results of identical B implantations of Si and GaAs crystals. Multiple energy implantations were performed at room temperature and liquid nitrogen temperature with total doses up to 1.5×1016 B+ ions/cm2. The implanted B distribution was measured with neutron depth profiling (NDP) and found to agree well with Monte-Carlo ion range calculations. The RBS results showed that the CdTe crystals had not been rendered completely amorphous even for the highest dose implantation, unlike GaAs and Si. Furthermore, the DCD results showed little implantation induced structure in the rocking curves from the implanted CdTe crystals, in contrast to GaAs. The consequences of annealing at 500°C in an attempt to regrow the crystal structure are also discussed.

The annealing effects of a high energy beam of Cu ions on implanted CdTe crystals are studied. Single crystals of CdTe have been implanted with Eu (energy 60 keV, fluence 1 × 1016 cm−2) at substrate temperatures of 25°C, and 400°C. Lattice damage introduced by the implantation process was measured by Rutherford backscattering. The samples were then implanted with high energy Cu ions (energy 3.5 MeV, fluence 0.5 × 1016 cm−2) at substrate temperatures of 25°C and 200°C. Channeling spectra from these samples indicate a reduction in the near-surface lattice damage as a result of the Cu implantation that can be unambiguously separated from the external heating of the substrate.

Ion implantation of Ti into LINbO3 has been shown to be an effective means of producing optical waveguides, while maintaining better control over the resulting concentration profile of the dopant than can be achieved by in-diffusion. While undoped, amorphous LiNbO3 can be regrown by solid-phase epitaxy at 400°C with a regrowth velocity of 250 Å/min, the higher concentrations of Ti required to form a waveguide (∼10%) slow the regrowth considerably, so that temperatures approaching 800°C are used. Complete removal of residual damage requires annealing temperatures of 1000°C, not significantly lower than those used with in-diffusion. Solid phase epitaxy of Agimplanted LiNbO3, however, occurs at much lower temperatures. The regrowth is completed at 400°C, and annealing of all residual damage occurs at or below 800°C. Furthermore, the regrowth rate is independent of Ag concentration up to the highest dose implanted to date, 1 × 1017 Ag/cm2. The usefulness of Ag implantation for the formation of optical waveguides is limited, however, by the higher mobility of Ag at the annealing temperature, compared to Ti.

Ion channeling of 3 MeV He ions has been employed to investigate the lattice location of Ti in Ti implanted optical waveguides in LiNbO3 after Solid Phase Epitaxy. Particle-Induced X-rays (PIXE) from Ti at 4.5 keY (K∝) and Nb at 2.2 keV (L∝β) and 16.6 keV (K∝) have been detected and analyzed simultaneously.

All scans yield similar behaviour of the Ti and the Nb signals. This provides clear evidence, that within well annealed implanted waveguides the Ti4+ and the Nb5+ ions occupy equivalent lattice positions.

Photoluminescence (PL) measurements of thin rapidly annealed As implants have been made. The experimental conditions required to enhance the PL signal from the implant region have been established. Our data suggest that residual defects within the implanted volume can completely quench the PL signal. For the range of anneal times investigated (0.3s to 10s) we find that temperatures above 1050°C are needed to remove these defects. Anneal temperatures of 1350°C remove the defects sufficiently to allow us to observe strong PL from the implanted volume..

C+ ion-implantation was carried out for extremely pure GaAs gown by moleplar Ieam epitaxy. The dose was very widely varied from 1×1015 to 1×1020 cm−3. Photoluminescence and Raman scattering measurements were performed at 2 K and room temperature, respectively, both as functions of dose and annealing temperature. The results revealed that for a dose less than 5×1017 cm−3, where the dominant damages are point defects, annealing at 200°C is sufficient to attain a damage-free lattice. When the dose is above that value, where the principal damage is a highly disordered lattice, annealing, at least above 550°C, is required. Photoluminescence spectra showed that the well-defined near-band-gap emission, [g-g], (exclusively inherent to acceptors), was a dominant emission in the above mentioned low dose region, but was strongly suppressed in the higher region. These observations indicate that ion-implanted C atoms in GaAs behave as amphoteric impurities when the dose exceeds a critical value, of around 5×1017 cm−3.

The effects of ion implantation on the electrical and structural properties of poly(dimethylsilylene-co-methylphenylsilylene), DMIOPS, thin films have been investigated. Ionic species of krypton, arsenic, fluorine, chlorine, and sulfur were implfgted at energies ranging from 35 to 200 key and with doses of up to 1 × 1016 ion/cm2. The conductivity of the polymer increased upon implantation reaching a maximum value of 9.6 × 10−6 (Ω cm)−1 for the case of arsenic ion at a dose of 1 × 1016 ion/cm2 and energy of 100 keV. The results showed that ion implantation induced conduction in DUMPS was primarily due to structural modifications of the material brought about by the energetic ions. Infrared analysis and Auger electron spectroscopy showed evidence for the formation of a silicon carbide-like structure upon implantation.

Ion bombardment of polycrystalline Ge, Si, and Au films leads to rates of grain boundary motion that greatly exceed rates of thermally-induced motion at the same temperature and which exhibit a weak temperature dependence. The enhanced migration rate is proportional to the rate of energy deposition in nuclear collisions at or very near the grain boundary. Experimental work is reviewed, and a transition state model is presented which accounts for the observed kinetics of grain boundary migration during bombardment. This model suggests that the rate limiting step in grain boundary motion may be thermally-induced migration of a bombardment-generated defect across the boundary. Also, the ratio of atomic jumps at grain boundaries to the local collision-induced Frenkel defect generation rate is shown to be characteristic of each material, but independent of ion mass and ion flux. The model is extended to the motion of an interface between two phases, and applications to crystallization during ion bombardment are discussed.

The dependence of ion irradiation induced grain growth on collision and material properties were studied in metal and alloy thin films. The uniform grain size increased with ion dose in the low dose regime; and the growth saturated in the high dose regime. The saturated grain size was determined 'by the cascade dimension as well as the atomic mobility. The growth was further influenced by the temperature, composition and impurity.

Fixed fluence ion irradiation of silicon is shown to produce either defected crystal or amorphous silicon depending on the ion flux employed. The amorphous threshold flux, defined as the minimum flux required to generate a continuous amorphous layer for a fixed fluence irradiation, is measured as a function of irradiation temperature. This critical flux for amorphization is shown to satisfy an Arrhenius expression with a unique activation energy of ∼1.2eV, which corresponds to the migration/dissociation energy of the silicon divacancy. These observations lead to the conclusion that the stability of the silicon divacancy controls the competition between defect production and dynamic defect annealing, and hence the crystalline to amorphous phase transformation.

The ion-bombardment-induced movement of an amorphous/crystalline interface in silicon has been studied under pulsed beam conditions. Irradiations were performed with a 1.5 MeV Xe beam at temperatures of 200–300°C which induced a planar motion of the interface, either to epitaxially crystallize the amorphous material or to planar amorphize the crystalline material. It is found that at a fixed peak and average beam current the result of pulsed irradiation can vary from amorphization at low pulse repetition frequency to crystallization at high pulse repetition frequency. The frequency which characterizes this change in behavior is virtually temperature independent but strongly dependent on peak beam flux.

Solid-phase epitaxy of implanted Si is observed at room temperature during in situ electron irradiation in a Transmission Electron Microscope. Results obtained from irradiation of cross sections of samples containing different doping species show that: i) the basic mechanism of the process is the migration and recombination at the amorphous-crystalline interface of radiation defects coming both from the amorphous and crystalline side; ii) the diffusion length of such defects is of the order of 40 nm; iii) the regrowth rate is impurity dependent: a factor two exists between the faste

Thin layers of Si were chemical vapor deposited onto as - received p-type <100> Si wafers and implanted with 80 KeV of As or Ge to a fluence of 1 × 1015 /cm2. Irradiation at 450°C with 600 KeV Kr++ ions causes the epitaxial growth of the entire deposited and amorphized Si layer. At lower irradiation temperatures the regrowth rate of the deposited layers is substantially reduced with respect that of the implanted amorphous layers. The presence of As enhances the regrowth rate of a factor 2.5. The results are explained qualitatively in terms of a dynamical bond breaking of SiO2, and of a dopant influence on the migration energy of the defects responsible for the growth.

The mechanism that underlies the recrystallization of amorphous silicon tas not been established. It is generally argued, however, that the rearrangement of the network occurs through the breaking of bonds or the introduction of vacancies and that this step is responsible for the observed activation energy (∼2.5 eV). It is suggested here that the rearrangement of the network is accomplished through the migration of intrinsic overcoordination defects (“floating bonds”) and that this process has a small activation energy (∼0.4 eV). The observed large activation energy is actually due to a reaction that inhibits recrystallization. This reaction may be the elimination of preexisting dangling bonds.

It has been observed that doping produces an enhancement in the recrystallization growth rate of silicon made amorphous by ionimplantation. This enhancement has been attributed to a shift of the Fermi level with doping. Evidence supporting this is based on the compensating effect of implantation of n- and p-type dopants together. We have previously proposed a model of the recrystallization growth process based on the diffusion of dangling bonds. We suggested that the rate enhancement is due to band bending at the amorphous-crystalline interface produced by doping. We have calculated the change in activation energy for the recrystallization growth velocity for a number of doping concentrations as a function of temperature. The major contribution to the apparent lowering of the activation energy with doping in an Arrhenius plot of the growth velocity versus I/kT is due to the temperature dependence of the Fermi level. Experimental data are compared with the calculated results. In addition differences in the measured growth rates in thermal and laser annealed samples are discussed, with primary emphasis on the lack of a change in the activation energy with doping in the laser annealed case.