The kinetics of hydrogen diffusion and complex formation in crystalline silicon are investigated using a model previously developed to explain the influence of dopant type and concentration on observed deuterium profiles in silicon. The predictions of the model have been shown to be in close agreement with recent SIMS profiles of deuterated FZ-Si crystals, with the in-diffusion process dominated by trapping at impurity sites and by the formation of immobile hydrogen molecules.

Previous studies have treated the surface concentration of hydrogen and the capture radii of hydrogen at complexes as free kinetic parameters. We present an analytic relationship between the diffusion coefficient D of neutral hydrogen and both the hydrogen surface concentration and the capture radius for molecule formation. The consequences of fixing D at the known high-temperature value for the diffusion coefficient in the model are determined. The existence of this analytic relation reduces the number of free parameters in the kinetic model and leads to an improved understanding of hydrogen reactions in silicon.

We apply quantitative chemical mapping techniques to study thermal interdiffusion and ion-implantation induced intermixing at single heterointerfaces at the atomic level. Our results show thermal interdiffusion to be strongly depth dependent. This is related to the need for the presence of native point defects (interstitials and vacancies) to bring about interdiffusion. Since their initial concentration in the bulk is negligible, the point defects must be injected at the surface and transported to the interface for interdiffusion to occur. In the case of ion-implanted samples, we find the passage of a single energetic ion through a sample at 77 K causes significant intermixing, even when the sample receives no subsequent thermal treatment.

We report on the low temperature growth of GaAs on Ge substrates using Atomic Layer Epitaxy. Low temperature deposition has resulted in substantial reduction of the outdiffusion of Ge into the GaAs epilayer as being indicated from SIMS. The I-V characteristics of the GaAs/Ge heterojunction were thyristor like or near abrupt depending on the growth temperature. We also report on the use of the Atomic Layer Epitaxy self-limiting adsorption process of TMGa to control the diffusion of Ga into Ge substrates at the monolayer level.

Annealed Mg implants into GaAs show three diffusion regions: 1) rapid uphill diffusion in the peak of the implant; 2) rapid concentration-independent diffusion in the tail; and 3) slow concentration-dependent diffusion in between. Implanted Be, in contrast, exhibits only concentration-dependent diffusion. Constant Fermi level experiments show that this diffusion is actually hole-dependent. Uphill diffusion can be induced in Be implants by co-implanting with a heavier element such as Ar. Paradoxically, this retards the concentration-dependent diffusion. This behavior can be explained with the Substitutional-Interstitial-Diffusion (SID) mechanism and an understanding of the defect chemistry after implantation. In the region of uphill diffusion, the dopants are seen to getter from areas of excess Ga interstitials toward areas of excess Ga vacancies. The magnitude of the Ga interstitial gradient with respect to the dopant concentration is shown to be critical for the uphill diffusion. The reduction in concentration-dependent diffusion with co-implants is thought to be caused by implant damage allowing dopant atoms to shift from interstitial to substitutional sites.

The microstructure induced by the Zn diffusion at 1170 K into doped and undoped semi-insulating GaAs single crystals was characterized for various diffusion times t < 1740 min by analytical electron microscopy. The results were compared with Zn concentration profiles obtained by spreading resistance measurements (SRM) on the same samples. At the diffusion front the formation of prismatic interstitial dislocation loops, dislocation networks, and of cavities partly filled with Ga was observed. Closer to the surface facetted voids and, for the undoped samples, vacancy-type dislocation loops formed. The near surface region of highest Zn-concentration showed a high density of Zn-rich precipitates. A model is presented which accounts .for these observations. It is based on fast interstitial Zn diffusion and the kick-out mechanism for interstitial-substituional exchange.

Deposition of Mn at room temperature onto atomically clean CdTe(110) surfaces yields atomic interdiffusion for metal coverages <3 angstroms with Mn atoms occupying cation sites within the surface and near-surface layers of the semiconductor. Synchrotron radiation photoemission studies with variable photoelectron escape depth indicate the formation of a relatively homogeneous semiconductor surface alloy. The highest Mn concentration observed in the alloy exceeds those obtainable with bulk crystal growth methods.

An examination of the three available quantitative models of Si diffusion in GaAs has led to the conclusion that the Fermi-level effect mechanism plays the most essential role. In some experimental results a point defect concentration transient is involved which should be incoorporated in future models.

Using the Si-pair diffusion formalism of Greiner and Gibbons, a new Si diffusion model is developed based on the dominant diffusion species being SiGa+ - VGa- pairs, where VGa is the Ga vacancy. In the model, the unknown parameters are the pair diffusion coefficient (Dp) and the equilibrium constant, which are fitted to the experimental data. Dp is also derived to be equal to one-half the Ga vacancy diffusivity. To verify this relation, an experiment to determine the VGa diffusivity, Dv, is performed. A relation for Dv of the form 0.962 exp(-2.72 eV/kT) cm2/s is obtained, and it is shown that the fitted Dp values are indeed approximately equal to 0.5 Dv.

The anomalous shape of the Zn diffusion profile in GaAs has been quantitatively explained. The interchange between interstitial Zn and substitutional Zn is assumed to occur via the Ga vacancies. These vacancies are proposed to be either neutral or singly ionized, depending on the position of the Fermi level. In addition, two physical phenomena are proposed. Substitutional Zn thermally generates interstitial Zn-Ga vacancy pairs and there is pairing between the donor, interstitial Zn and the acceptor, substitutional Zn. This pairing leads to the interstitial Zn diffusivity being a decreasing function of the substitutional Zn concentration. The model is found to be in good agreement with the experimental data.

The diffusion of ion-implanted tin in gallium arsenide was investigated as a function of temperature, dose and background doping. The chemical depth profiles were determined using SIMS and the carrier profiles were determined by CV Etch Profiling. The data was fit using a numerical process simulator, SUPREM 3.5. Sn diffusivity was found to depend on the square of the electron concentration. Sn and Ge were found to have relatively high diffusivities when implanted, while Si diffused very little.

A multilayered Si doped MBE structure was used to study the effective diffusivity of Si and the results are modeled with an (n/ni)2 dependence over the temperature range of 750°C – 950°C. An activation energy, Ea′, of 4.0 eV is obtained which is higher than normally reported in the literature. This higher Ea′ value results from appropriate accounting of the temperature dependence of ni, which is often neglected in the expression for Deff. Si diffusion at a buried n+/n++junction of a MOCVD grown structure also follows an (n/ni)2 dependence. These results support a Fermi-level model of Si diffusion in GaAs and suggest that the local point defect chemistry of the GaAs, through Si doping, is responsible for this diffusion behavior, regardless of the proximity to the surface.

The purpose of this work is to give an overview of the current phenomenological understanding of impurity-induced layer disordering (IILD). First, we identify key experimental findings such as the influence of the crystal surface-ambient interaction, the Fermi-level effect, and the impurity concentration on Al-Ga interdiffusion. Second, we review the strengths and weaknesses of existing IILD models in consideration of the above mentioned experimental data. Finally, we discuss the pitfalls involved in generalizing the results of individual Al-Ga interdiffusion experiments in order to explain a broader collection of IILD data.

During the course of investigation of the mixing of highly silicon-doped GaAs/AlAs superlattices, defects such as dislocation loops and Si-rich precipitates were found to form in the specimens. These defects formed at particular doping levels upon annealing of the samples. The presence of the defects can be related to changes in mixing behavior. In the present study, transmission electron microscopy has been used to characterize the defects. Superlattices with varying silicon doping levels were annealed at different temperatures for varying time-periods, to observe the temperature-time behavior of the dislocation loops. The defects aggregate preferentially in the GaAs as opposed to the AlAs in the superlattice. A number of the dislocation-loops were investigated using high-resolution TEM. All the loops observed were interstitial in nature.

The mechanisms of Ga self-diffusion can be derived from interdiffusion experiments in intrinsic and doped GaAs-based superlattices. These experiments allow to conclude that Ga self-diffusion in intrinsic and n-doped GaAs is carried by triply negatively charged gallium vacancies whereas in p-doped GaAs positively charged gallium self-interstitials dominate Ga self-diffusion. The diffusion mechanisms of Zn and Be are discussed with special emphasis on the difference between their in- and out-diffusion behavior which is due to diffusion-induced non-equilibrium point defects.

We review three important experimental results which suggest that electron capture and emission by the DX center in AlxGa1-xAs proceeds via an excited intermediate state: the very different dependencies of the thermal capture and emission rates on alloy composition, the exponential dependence of the thermal capture rate on the quasi-equilibrium Fermi energy, and the thermal activation of the hot electron capture rate. None of these results is readily explained by a conventional lattice relaxation model, in which an electron is captured directly from the lowest lying band edge, but each can be simply explained if the dominant channel for multiphonon capture is via a transition state which lies well above the band edge. This picture is consistent with recent pseudopotential calculations which predict that the lattice relaxed state (the DX state) is stabilized by capture of more than one electron, since such a model naturally admits the possibility of an intermediate one-electron state.

The effects of specific local environments on DX centers were investigated for two different substitutional-site-donors of Si and Te in selectively-doped ordered-alloy samples. Capacitance transient technique under hydrostatic pressure was utilized for the determination of the DX energy-level structure. Downward movement of the energy levels for both Si- and Te-DX centers was observed when Al occupies the 2nd or 1st nearest neighbor site, respectively. A very large shift of the electron-emission activation-energy with pressure was found for the Te-DX centers surrounded by specific Al and Ga mixed-environments, whereas negligible shift for the Si-DX centers was observed regardless of the environment. These effects were discussed in relation to the lattice relaxation of the relevant DX centers.

We present data which show that electron emission from the DX center is sensitive to the Poole-Frenkel effect. We demonstrate that this result implies that the DX center is an L effective-mass state of the donor impurity accompanied by a small lattice relaxation. We show that all the observations so far obtained on this center are in agreement with this model.

The influence of chemically different donor species on the nature of shallow donor states in Al0.6Ga0.4As/GaAs heterostructures has been investigated by optically detected magnetic resonance (ODMR). Previous theoretical work by Morgan predicts a triplet state for group IV donors and a singlet state for group VI donors. ODMR experiments were performed on as-grown and implanted Si-, Se-, and S-doped epitaxial layers of Al0.6Ga0.4As grown on (001) GaAs substrates. The effective-mass states are modified by the heter-oepitaxial strain in these layers. The Si donors are characterized as quasi-independent valley states. The Se and S donors have valley-orbit splitting energies (i.e. chemical shifts) of 19-20 meV . The results indicate that Si, Se, and S donors are on the lattice sites in the metastable state of DX.

Extremely heavily doped n-GaAs was produced by pulsed-laser annealing of Si implanted GaAs, achieving carrier concentrations exceeding 3.2×1019/cm3. Our photoluminescence (PL) spectra indicate a bandgap narrowing due to heavy n-type doping with a functional form of Δεg(eV) = - 6.3 × 10cm-8[(cm-2)]1/3 . At the highest carrier concentration, the bandgap shrinkage reaches -200 meV and the electron Fermi energy is -410 meV. These large values indicate that there exists a considerable conduction band “stretch” between the Γ and L-valley of GaAs for very high n-type concentrations.

A theoretical model for DX centers which explains their unusual electronic properties in terms of two distinct bonding configurations for donor impurities in AlxGa1-x As alloys is examined. The results of our ab initio self-consistent pseudo-potential calculations show that for x > ≃20%, the normal fourfold coordinated substitutional site becomes unstable with respect to a large lattice distortion. The model explains the large difference between the thermal and optical ionization energies of DX centers.