The effects of annealing temperatures on the reduction of the defects induced by the combination of arsenic ion implantation and thin film stresses were studied. Cross-sectional transmission electron microscopy (XTEM), preferential etching, optical microscopy and scanning electron microscopy (SEM) were used to characterize the defects. The results indicated that a low temperature annealing step right after As+ ion implantation is critical to anneal out the implantation damage and prevent the damage acts as dislocation sources to generate dislocations. Without proper annealing, the generated dislocations can propagate into the active device regions resulting in excessive leakage current.

Extreme variations in surface reflectivity/topography have been observed on silicon wafers with evaporated aluminum metallization following thermal annealing processes. Such topographic variations negatively impacted the performance of wire bonder pattern recognition systems. It was proposed that this variation was attributed to differences in deposition rates during evaporation of the aluminum front metallization. SEM analysis revealed that the topographically rough sites, deemed as normal for this processing sequence, had large numbers of hillocks, while the smooth sites exhibited limited numbers of hillocks after heat treatment. SIMS depth profiling analyses did not indicate that any elemental difference existed between the smooth and the rough sites, although interfacial differences were detected. Specifically, the Al-Si interface for the rough surface was more graded in nature while the Al-Si interface for the smooth surface was more abrupt. TEM analysis of the rough sites revealed that a three-dimensional island growth mechanism prevailed during film growth, resulting in uneven surface development, whereas analysis of the smooth sites showed a columnar grain structure.

Atomic resolution images of Ba0.6K0.4BiO3 have been obtained at electron-dose levels of 102 e-/Å2 using a slow-scan CCD camera on a high-resolution electron microscope; at this dose level electron-beam damage can be avoided. The superconducting material is a clean perovskite without evidence for any distortions such as charge-density waves. At higher dose levels modulations appear which, from measurements of the desorbing species under low-energy electron radiation, are due to ionization damage and loss of oxygen.

Experimental data on the phase transformation kinetics in vanadium pentoxide due to surface oxygen loss are analyzed theoretically. A model for the process as a one-dimensional problem with oxygen loss from the surface and coupled interface and diffusion controlled growth modes is described. This model appears to match well the experimental data with reasonable numbers for the surface loss rate and diffusion constant. In particular, the model reproduces changes in the number of phase fronts as a function of electron beam flux. In addition, the analysis confirms that the effective diffusion constant is electron beam flux dependent.

A model to describe experimental results on phase transitions in vanadium oxides was suggested recently by Marks, Volpert and Ai (1992). This work is devoted to a more detailed analysis of the model. An analytical approach to the problem based on a quasistationary approximation is used to find the distribution of the concentration and the interface velocity. A multiplicity of modes of the interface propagation is shown to be possible depending on the kinetics of the phase transformation. The stability of the interface to small perturbations is studied. Two-interface propagation is also considered.

The intensities of the diffraction spots for the boron induced √3×√3R30° Si(111) reconstruction in a bulk electron microscope sample examined in ultrahigh vacuum are compared with the results of multislice simulations. The intensities of the spots support the relaxed S5 model. We rule out the existence of any subsurface structure such as the stacking faults present in the Si(111) 7×7 surface.

Effects of electron irradiation on ReO3 supported on different substrates (carbon and silicon monoxide) in both non-UHV and UHV environments were investigated using high resolution electron microscopy. It was observed that clean ReO3 was stable under electron irradiation in a carbon free environment. When carbon was present, electron irradiation induced surface reactions of ReO3 with carbon to form a metastable ReC phase. ReC has a NaCl fcc structure and a lattice parameter of 4.00 ± 0.05 (Å). The kinetics of the reaction were found to be linear with time.

GaAs was grown by molecular beam epitaxy (MBE) and ion-assisted MBE on Si(100) substrates. Three-dimensional (3D) island nucleation, observed during MBE growth, was eliminated during ion-assisted MBE when the ion energyE was >25 eV and the product ofE and the current densityJ was ≈6-12 eV mA/cm2. IncreasingEJ to ≈15 eV mA/ cm2 resulted in excessive ion damage. Decreasing the substrate temperature from 280 to 580° C during ion-assisted MBE yielded a slight decrease in surface roughness, and flatter surfaces were obtained for lower As4/Ga flux ratios. The suppression of 3D island nucleation led to an improvement in the crystalline perfection of thicker GaAs films. For example, the x-ray diffraction rocking-curve full-width-at-half-maximum values for 0.5 μm thick films grown at 380° C decreased from 1700 arcsec to 1350 arcsec when ion irradiation was used during nucleation. IAMBE allowed nucleation of thin, relatively flat-surfaced GaAs films even at 580° C, resulting in FWHM values of 1850 arcsec for 0.14 /μm thick films.

Much of the study of reconstructed surfaces has been done with surface sensitive techniques that rarely allow the investigator to probe both the bulk and surface properties. As a result, not much has been written about the possiblity of bulk defects influencing the shape and position of domains on a reconstructed surface. With the recent advent of UHV microscopes with attendent surface science chambers connected, it has become possible to examine both the bulk and surface of a given sample and obtain a more complete picture of the surface reconstruction process. We have found in several investigations of Au (001) and Si (111) surfaces that bulk defects such as dislocations and stacking faults near the surface can act as boundaries for domains of a reconstructed surface.

It has been known for a long time that electron irradiation induces damage in maximal valence transition metal oxides such as TiO2, V2O5, and WO3, of which transition metal ions have an empty d-shell. This type of damage is excited by electronic transition and can be explained by the Knoteck-Feibelman mechanism (K-F mechanism). Although the K-F mechanism predicts that no damage should occur in transition metal oxides of which the transition metal ions have a partially filled d-shell, namely submaximal valence transition metal oxides, our recent study on ReO3 shows that submaximal valence transition metal oxides undergo damage during electron irradiation.ReO3 has a nearly cubic structure and contains a single unit in its cell: a = 3.73 Å, and α = 89°34'. TEM specimens were prepared by depositing dry powders onto a holey carbon film supported on a copper grid. Specimens were examined in Hitachi H-9000 and UHV H-9000 electron microscopes both operated at 300 keV accelerating voltage. The electron beam flux was maintained at about 10 A/cm2 during the observation.

Results are reported using UHV electron microscopy to determine the role of background gases in influencing surface damage experiments and on the gold (001) surface prepared by ion-beam cleaning/thinning and annealing. In maximum valence oxides the end product is a higher-symmetry oxide or metal in UHV, but in a non-UHV environment secondary reactions take place. No evidence is found for electronic damage of non-maximal valence oxides, only sputtering and electron-stimulated reactions. For the gold (001) surface we have reproduced conventional surface-science preparations using ion-beam cleaning and annealing to produce the known reconstructed hexagonal monolayer on the surface.

It has previously been suggested, based upon multislice calculations. that electron diffraction from a surface reconstruction can be interpreted kinematically with a pseudo-kinematical condition setup for the underlying crystal. These calculations were carried out under conditions where the numerical multislice integration would not have fully converged. The intention of this note is to provide direct experimental evidence from the boron-doped Si(111) square-root 3 X square-root 3 R30-degrees surface that the kinematical assertion is valid.

The recent development of ultra high vacuum high resolution electron microscopes has made it possible to use an electron microscope to study surface structures with atomic resolution. Although surface images of Au(110) 2x11 and Si(111) 7x72 reconstructions have been obtained, no standard TEM sample preparation technique for surface imaging has been developed for routine uses. In conventional surface science, the common method of producing an UHV clean sample is a combination of ion sputtering and annealing; can this process be used to produce TEM samples for surface imaging. Our studies show that clean, well order TEM samples can be achieved by this approach.

A variety of techniques including LEED, STM and RHEED have been used to study surface reconstructions on the silicon <111> surface. Additionally, ultra high vacuum-transmission electron microscopy (UHV-TEM) has been used for a limited number of studies most notably on the 7x7 reconstructed surface. The limiting factor in these studies has been the availability of microscopes capable of in-situ sample preparation and imaging in a UHV environment. The Hitachi UHV-H9000 located at Northwestern University has recently been used to observe several surface reconstructions on a single crystal silicon <111> thin film. Transmission electron diffraction (TED) patterns were obtained for 7x7, and 5x1 surface reconstructions.

It is shown that ion-beam thinning and cleaning of bulk single-crystal samples coupled with annealing can produce surfaces suitable for conventional plan-view imaging including HREM. The key elements are to ensure the absolute cleanliness of the sample preparation system, care in choosing proper ion beam energies, and choosing the appropriate annealing conditions so as to minimize coarsening by bulk diffusion. The presence of surface reconstructions can be readily detected in off-zone diffraction patterns and in on-zone two-beam bright- and dark-field images.

Very-low-energy (almost-equal-to 28 eV), high-flux (almost-equal-to 0.4 mA/cm2) Ar-ion irradiation during molecular-beam epitaxy changed the nucleation or GaAs on Si(100) from Stranski-Krastanov to a mechanism approaching layer-by-layer growth. While three-dimensional island nucleation was eliminated, the growth surface exhibited low-amplitude undulations. The results are explained by ion-induced removal of atoms from stable 3D islands, which suppressed 3D island nucleation.

Suppression of three-dimensional (3D) island nucleation during growth of InAs on Si (100), achieved by using very low energy, high-flux Ar ion irradiation, reduced planar defect densities. For 13 eV ion irradiation, 3D islands nucleated after ∼2 monolayers (ML) of deposition, similar to conventional molecular beam epitaxy. High-resolution transmission electron microscopy studies of nominally 18-ML-thick films showed 3D InAs islands with {111} facets. A high density of {111} twins and stacking faults was observed adjacent to many of the {111} facets. Most of these defects propagated into the film upon further growth. When nucleation was carried out with 28 eV ion irradiation, flat InAs films were observed for thicknesses up to ∼10 ML. The 3D islands that nucleated at higher thicknesses were flatter with less faceting than in the 13 eV case. The density of planar defects in the initial nucleation layer and in thicker InAs films was reduced when 3D island nucleation was suppressed. These results indicate that planar defects formed directly on the {111} facets of the 3D islands.

III-V semiconductor films nucleate by the Stranski-Krastanov (SK) mechanism on Si substrates. Many of the extended defects present in the films are believed to result from the island formation and coalescence stage of SK growth. We have recently shown that low (-30 eV) energy, high flux (4 ions per deposited atom), Ar ion irradiation during nucleation of III-V semiconductors on Si substrates prolongs the 1ayer-by-layer stage of SK nucleation, leading to a decrease in extended defect densities. Furthermore, the epitaxial temperature was reduced by >100°C due to ion irradiation. The effect of ion bombardment on the nucleation mechanism was explained as being due to ion-induced dissociation of three-dimensional islands and ion-enhanced surface diffusion.For the case of InAs grown at 380°C on Si(100) (11% lattice mismatch), where island formation is expected after ≤ 1 monolayer (ML) during molecular beam epitaxy (MBE), in-situ reflection high-energy electron diffraction (RHEED) showed that 28 eV Ar ion irradiation prolonged the layer-by-layer stage of SK nucleation up to 10 ML. Otherion energies maintained layer-by-layer growth to lesser thicknesses. The ion-induced change in nucleation mechanism resulted in smoother surfaces and improved the crystalline perfection of thicker films as shown by transmission electron microscopy and X-ray rocking curve studies.

Radiation damage in V2Os induced by electrons in the energy range 0-3 keV has been studied by high resolution electron microscopy, X-ray photoelectron spectroscopy and mass spectrometry. Different phase transformations were observed when V205 was irradiated by electrons of different energies. The damage kinetics and the role of electron radiation enhanced diffusion were investigated. It was found that nucleation and growth of lower oxides was driven by electron radiation enhanced diffusion.

In order to examine surfaces of materials, a prerequisite is a microscope which combines ultra-high vacuum (UHV) with surface science cleaning and characterization techniques such as ion beam sputtering, annealing, and Auger spectroscopy. In order to achieve this, we have mounted onto the side of a UHV-H9000 microscope LEED/Auger, an ion gun, and optical heating; in the transfer chamber specimens can be cleaned at a base pressure of 2×10-10 torr and transferred into the microscope which operates at pressures better than 2×10-10 torr. With this marriage, it is relatively simple to prepare and characterize clean surfaces.As an example, thin gold film specimens, textured with the [111] normal to the film, were made in a standard vacuum evaporator and floated onto a gold grid. The transfer chamber was then baked-out at 250°C for about 12 hours to achieve UHV conditions. Figure 1 shows an image taken from the gold film after bakeout.