The nucleation and epitaxial growth of deposits at both low and high temperature substrate conditions and the resulting two and three dimensional growth modes have been investigated using electron diffraction and transmission electron microscopy. At high temperatures, the growth of small particles such as Au, Pt and Ge grown under low rate conditions similar to MBE are revealed by lattice imaging. Using an intermediate bufler layer Ag film, surface growth of (100) and (310) Au thin film layers are revealed by diffraction contrast to be in a step ledge growth regime. While the above investigations were performed ex situ to the microscope vacuum environment, an in situ evaporator has been also been installed on a high resolution electron microscope. Observations can be made in real time of cluster growth from atomic diameters on up. The resulting atomic motion around cluster edges is similar in appearance to surface melting.

Several TiO2 (110) surfaces, reduced via different mechanisms, were investigated in ultra-high vacuum using Scanning Tunneling Microscopy (STM) and Spectroscopy (STS) to determine the effect of reduction on the surface atomic and electronic structure. Defect chemistry and its relation to the change of property and structure are discussed.

Atomic force microscopy (AFM) and low energy electron diffraction (LEED) have been used to study the faceting behavior on (001) and (100) surfaces of a TiO2 single crystal. On the TiO2 (001) surface, LEED patterns characteristic of {011} facet planes develop after annealing below 900 °C in agreement with previous studies, but AFM shows a complex surface morphology consisting of a large distribution of facet sizes and orientations. After annealing at 1300 °C, facets do not form but rather a network of 5 – 30 nm high ridges develops over the entire surface. These ridges may be the result of surface defects produced by changes in bulk stoichiometry during annealing. On the TiO2 (100) surface, facets are also observed after annealing below 900 °C. However, these facets have extremely small height to width aspect ratios and are not discernible with LEED.

This paper reviews the problems that are generic to the direct epitaxy of quantum wire superlattices on {100} and {110} vicinal surfaces using molecular beam epitaxy.

A study of the molecular beam epitaxial (MBE) growth on singular and vicinal (110) surfaces of GaAs is presented. Quantum well structures and tilted superlattices (TSL) were grown on substrates misoriented 0.5°-2° towards the nearest [010] and [111]A azimuths at growth temperatures ranging from 450° C to 600° C under different growth conditions. The structures were characterized by Nomarski optical microscopy, transmission electron microscopy (TEM) and photoluminescence (PL) spectroscopy.

Two types of faceting were observed on the surfaces. The structures grown at temperatures above 540°C and As beam fluxes below l×10-5 torr showed V-shaped facets pointing in the [001] direction and are attributed to As deficient island growth. Lower temperatures and higher As beam fluxes lead to surfaces with microfacets that are elongated along the respective step directions on the vicinal surface and are due to step bunching during growth. Their density and height decrease with decreasing vicinal angle and they disappear on the singular (110) surface. The photoluminescence of the GaAs quantum wells grown on these samples is redshifted with respect to that of the quantum wells grown on the flat surface. This is being ascribed to the fact that on the vicinal surface, the recombination takes place at the facets where the quantum wells are wider.

The contrast in the TEM images of the TSL show for the samples misoriented towards [010] that the lateral segregation to the step edges on this surface is appreciable. The TSL spacing and the tilt however show that during growth the vicinal surfaces tend towards a surface with smaller miscut.

Recent progress in low energy ion-surface interactions, and the early stages of ion-assisted epitaxy of semiconductor thin films is described. Advances in three areas are discussed: dynamics of displacements and defect incorporation, nucleation mechanisms, and the use of ion bombardment to modify epitaxial growth kinetics in a truly surface-selective manner.

The stability of CoSi2/Si interfaces was examined in this study using columnar suicide structures grown on (111) Si substrates. In the first set of experiments, Co and Si were co-deposited (1:7 flux ratio) using molecular beam epitaxy at 800°C and the resulting columnar suicide layer was capped by epitaxial Si. Deposition of Co on the surface of the Si capping layer at 800°C results in the growth of the buried suicide columns. The buried columns grow by subsurface diffusion of the deposited Co, suppressing the formation of surface islands of CoSi2. The columns' sidewalls appear to be less stable than the top and bottom interfaces, resulting in preferential lateral growth and ultimately in the coalescence of the columns to form a continuous buried CoSi2 layer.

In the second set of experiments, annealing of a 250nm-thick buried columnar layer at 1000°C under a 100nm-thick Si capping layer results in the formation of a surface layer of CoSi2 with a reduction in the sizes of the CoSi2 columns. For a sample having a thicker (500nm) Si capping layer the annealing leads to Ostwald ripening producing buried equiaxed columns. The' high CoSi2/Si interfacial strain could provide the driving force for the observed behavior of the buried columns under high-temperature annealing.

Epitaxial single-domain CoSi2(100) layers were grown on Si(100) by use of a template technique. In-situ scanning tunneling microscopy (STM) and reflection high energy electron diffraction (RHEED) were used for a detailed surface study. The (√2×√2)R45 reconstruction of the Co-rich “C-surface” and the (3√2×√2)R45 as well as a newly discovered (√2×√2)R45 of the Si-rich “S-surface” were resolved in real space and are discussed in detail. The transition from the C- to the S-surface above 500 °C is related to a (2×2) reconstruction.

Control of epitaxial relationship of CaF2 films grown on Si(111) substrates was considered to be important to improve surface morphology and crystallini ty of GaAs films on CaF2/Si(111) structures. We successfully grew CaF2 films with the “type-A” epitaxial relationship on Si(111) substrates, that is, the crystallographic orientation of the CaF2 films were aligned in the same direction as that of the Si(111) substrates. These “type-A” CaF2 films were grown by a two step growth method. It was found that surface morphology of GaAs films on the CaF2/Si(111) structures was drastically improved by growth of the “type-A” CaF2 films.

High quality single-crystal Au(100) buffers have been grown on GaAs(100) or MgO(100) substrates via a thin bcc Fe(100) nucleation layer. Using RHEED, LEED and Auger spectroscopy the growth of Ni overlayers at room temperature was observed to follow an epitaxial layer-by-layer mode for 2 monolayers, after which islanding occured. Detailed analysis of the Auger signal reveals that atomic mixing appears within the first completed layer. Grazing incidence diffraction has been used to investigate the structure of 5–10 nm films. A mixture of (100) and (110) grains have been characterized; the latter orientation shows a large number of stacking faults along the [111] direction in agreement with the Au[001] direction.

We report the results of a reflection high energy electron diffraction (RHEED) study on the order-disorder transitions at the (111) and (110) surface of Cu3Au. Our Cu3Au films were epitaxially grown on sapphire substrates with bcc (Nb or Mo-Cr) buffer layers. The (111) and (110) surfaces of Cu3Au are atomically flat as indicated by the RHEED patterns. We use the normalized integrated intensity of the superstructure streaks to measure the long range order parameter. We observed that the surface layers of both (110) and (111) Cu3Au exhibit an apparently continuous order-disorder transition, close to that predicted by mean field theory [5]. The kinetics of surface ordering also were monitored by RHEED after initial deposition of disordered layers, 2 MLs thick, on the ordered Cu3Au templates. In studies of the surface reordering process we find that the RHEED pattern is almost entirely sensitive to the surface layer alone.

RHEED and AES studies showed deposition-concurrent surface segregation (DCSS) consists of two stages: an initial and a continuous segregation stage. The dominant terms of the enthalpies of DCSS were shown to be the difference between the surface energies of the film and the substrate material.

We report on growth of crystalline superlattices of aluminum nitride/zirconium nitride on aluminum nitride buffered Si(111) substrates. We observed increased buffer layer surface roughness as compared with the buffer/Substrate interface. This roughness in turn influences the quality of the superlattice. We report results of surface morphology examination as a function of growth rate and substrate temperature.

Epitaxial growth of Bi2Sr2CuO6 thin films on a SrTiO3, a misoriented SrTiO3 and a Bi2Sr2CaCu2O8 single crystal have been performed using computer-controlled laser molecular beam epitaxy. On the SrTiO3 substrate, the 2201 films consist of two types of domains whose b-axes are perpendicular to each other. In contrast, the 2201 film on the 2212 single crystal has a complete single type of domains. Layer-by-layer growth with a half unit cell thickness of 2201 films is observed.

The surface microstructure of epitaxial YBa2Cu3O7−x films grown by pulsed laser ablation on (001) MgO and SrTiO3 substrates has been studied at various growth stages, ranging in thickness from eight c-axis perpendicular unit cells to ∼220 nm. On MgO (lattice mismatch ∼9%) even the thinnest films grow unit cell-by-unit cell by an island growth mechanism. However, on SrTi03 (mismatch ∼1%), a transition from a layer-like growth mode to island growth is observed as the film thickness increases. Islands with clear spiral growth structures are observed in even the thinnest films on MgO, but for films grown on SrTiO3 the spiral growth features are found only for film thicknesses slightly greater than the critical thickness for the switch to an island growth mode. The islands consist of stacks of atomically flat terraces whose step heights are multiples of the c-axis lattice parameter. The island density decreases significantly with increasing film thickness, while their diameters range from 50–400 nm, increasing with growth temperature. The terraced island grain morphology causes a surface roughness of from 10 to 30 nm (depending on growth temperature) in films ∼200 nm thick.

Ultrathin superconducting YBa2Cu3O7, films were grown on (100) YSZ (yttria-stabilized-zirconia) substrates by off-axis if magnetron sputtering at a relatively high deposition rate. The structure, orientation, and morphology of the films were examined by x-ray diffraction, reflection high-energy electron diffraction (RHEED), and scanning electron microscopy. X-ray diffraction patterns of films deposited on YSZ substrates exhibited strong c-axis alignment with the YBa2Cu3O7 peaks sharpening as the film thickness was increased. The degree of epitaxy apparent in RHEED photographs was found to increase dramatically as the film thickness was increased from 12 nm to 108 nm. This behavior is attributed to a nucleation and growth process in which epitaxy develops as a result of a 3 stage progression from a random to an oriented film.

The films were in-situ superconducting, exhibiting superconducting transition temperatures, Tc(0)'s, of 80 K for a 12 nm film and 88 K for a 280 nm film. However, the relatively low critical current densities (Jc < 1 × 106 A/cm2) at 77 K are probably due to a lack of in-plane epitaxy.

This paper presents the results of silicon epitaxial growth on silicon windows surrounded with oxide walls by electron-beam evaporation in an ultra-high vacuum system with a load-lock chamber. The wafer surface was in-situ cleaned in the growth chamber to remove native oxide by thermal desorption at about 840 °C and a base pressure of better than 2 × 10-9 Torr. The growth temperature was 200°C or higher. The pre-epitaxial silicon surface structure was inspected by reflection high energy electron diffraction (RHEED). The influence of the thermal desorption on the quality of the epi/substrate interface and epitaxial layers was studied. In addtion, the deposition parameters which control the epitaxial quality were investigated. The epitaxial films were characterized by cross-sectional trasmission electron microscopy (XTEM) and secondary ion mass spectroscopy (SIMS).

High-crystalline-quality B-doped SiGe films are epitaxial-ly grown on Si(100) substrates using the Si-GeH4-B2H6 MBE system. Both Ge and B profiles in the B-doped SiGe layer are readily controlled using the present system. This technique should be very promising in the bandgap engineering of SiGe base HBTs.

The selective LPEE (liquid phase electro-epitaxy) growth of GaAs on GaAs-coated Si substrate prepared by MOCVD has been performed. Several attempts, the use of the insulator melt cap, lowering of the melt height and optimization of the stripe direction, have been tried to increase the lateral growth rate. Very wide (60 μm) lateral overgrowth on the SiO2 mask is obtained by making the stripe window in <010> and equivalent directions and by increasing the lateral current component in the LPEE. The GaAs layer laterally grown on the SiO2 has much less EPD (etch pit density) showing that the defect in the MOCVD-GaAs is blocked by the SiO2. The growth mechanism and the method of increasing the lateral growth rate are discussed.

Role of in-situ heavy As doping in Si selective growth is studied using the SiH2Cl2/H2/HCl/AsH3 gas system. By increasing AsH3 flow rate, the growth rate decreases in proportion to In (AsH3 flow rate). This result is explained well using a Frumkin-Temkin adsorption model. The As concentrations in grown films depend on surface orientation. Moreover, the maximum As concentration is about ten times lower than the solubility limit at a given temperature. These results are able to be explained qualitatively, if equilibrium concentrations of impurities at the Si(100), the Si(111), and the bulk are assumed to be different from one another, and that the concentration at the surface is frozen in the grown film.