Search

It is generally known that dry etching induces surface damage, but the structure of etched surfaces has not been studied in detail. Nor has it been established how or whether a clean, ordered surface can be recovered after etching damage. Generally, etching damages the surface by introducing structural disorder and impurities to the surface region. Such damage may be so severe that a clean, ordered surface is not recoverable even after heating up to ∼1400 K in UHV. We subjected Si (1 1 1) surfaces to different reactive ion etching conditions and/or post-etch treatments and examined their effect on the surface. Low temperature STM revealed that a clean, ordered surface as evidenced by the appearance of large area 7 × 7 reconstructions can be obtained only under certain etching/post-etch recipes. On the other hand, LEED showing 7 × 7 diffraction spots but with high intensity background and Auger electron spectroscopy (AES) showing no impurity signal cannot be used as evidence that an ordered surface has been obtained.

A high quantity of tungsten oxide nanosheets were synthesized by oxidizing tungsten plates with potassium hydrate as the catalyst and tungsten plate as the substrate. The structural and geometrical properties were characterized by various techniques. It was found that the crystalline nanosheets have a WO3 structure with thicknesses of 30–50 nm and widths up to tens of micrometers. There exist two characteristic acute angles of about 37° or 51° on the nanosheet plane. The formation of these angles and the growth mechanism were discussed.

Tungsten oxide comblike nanostructures were synthesized using a two-step thermal evaporation method. The first step involving high reactor pressure and temperature was to synthesize the cores of the comb structures, upon which the teeth of the comb were grown in the second step using low operation pressures and temperatures. The teeth of the comb structure are well aligned and vertical to the side surfaces of the cores. The effects of growth parameters were examined, and the growth mechanism was investigated.

Uniform tungsten oxide nanowires with small diameter have been fabricated with potassium hydroxide as catalyst on a tungsten plate and by the method of thermal evaporation via a unique two-step heating process. Typical temperatures of this two-step process are 390 and 610 °C, respectively. The first-step heating was found crucial for the growth of small and uniform nanowires because it made the potassium hydroxide solution more uniformly distributed on the tungsten plate surface. The structure and composition of the grown nanowires were characterized by various methods. The diameter of the nanowires ranges from 30–200 nm and the length is up to several tens of micrometers. The structure of the nanowires is found to be orthorhombic tungsten oxide (W3O8), with cell parameters a=1.035 nm, b=1.399 nm, c=0.378 nm.

The surface structure of InN film heteroepitaxially grown on a GaN buffer layer by MBE is followed by low energy electron diffraction (LEED). The metallic surfactant layers on top of the InN surfaces show an incommensurate structure rather than being disordered. The metal in the incommensurate structure induces additional diffraction spots in the LEED. Based on the Auger experiments, not only In atoms but also Ga are present on the surface of the InN films.

We report the synthesis of potassium tungstate (K 2 W 3 O 10 ) micro-walls by a simple method of thermal evaporation. Using tungsten powder, tungsten oxide powder and potassium iodide as the source materials and tungsten plate as the substrate, K 2 W 3 O 10 micro-walls with thicknesses ranging from 400 nm to 2 μm, lengths up to hundreds of micrometers and heights up to tens of micrometers are obtained on a large scale. The possible mechanism for the formation of these tungstate micro-walls is discussed.

A new reconstruction of √3 × √3-R30° has been observed on a GaN film grown on a 6H–SiC (0 0 0 1)-√3 × √3 surface using RHEED and LEED experimental techniques. The experimental LEED PF shows that the GaN film is Ga-terminated hexagonal. The surface is a mixture of two structures with a single bilayer height difference between them. One is a √3 × √3-R30° reconstruction with Ga-adatoms occupying the T4 sites. Another is a Ga-terminated 1 × 1 with no extra Ga on top. The area ratio of the √3 × √3 part to the 1 × 1 part is slightly larger than 1. The first principle total energy calculations and Tensor-LEED I–V curves simulations further confirm this structure model.

During submonolayer deposition of GaN by molecular-beam epitaxy (MBE), novel `ghost' islands are observed by scanning tunneling microscopy (STM). The island contrast depends on gap voltage of the STM, suggesting a different atomic structure of the islands than that of GaN terrace or normal islands. Formation of `ghost' islands is related to the presence of excess gallium atoms on surface during MBE. Island size distribution show no peak but a monotonically decaying curve, indicating a surfactant mediated process for nucleation of such islands. A model is put forward explaining the origin and the structure of such islands.

Ab initio total energy calculations are performed to determine the interface structure of GaN films grown on the 6H - SiC (0001)[Formula: see text] substrate. The results show that the GaN film is of the wurtzite structure and has the Ga-polarity. It is also shown that stacking mismatch boundaries (SMBs) caused by the coalescence of GaN islands grown on stepped terraces of the 6H - SiC (0001) surface may be removed by stacking faults as the film grows. The types of SMBs on a stepped 6H - SiC (0001) surface are discussed.

The k derivative spectra (KDS) transform is used for construction of the three-dimensional atomic structure of the C 2 H 4/ Si (100)-(2×1) system from photoelectron diffraction data. The image function obtained by the KDS transform clearly observes the second-layer Si atoms and the C emitters apart from the first-layer Si atoms. The observations of the second-layer Si atoms and the C emitters make it easy to measure the C–C bond length correctly. Then a conclusive adsorption model — the di-σ model — for the C 2 H 4/ Si (100)-(2×1) system is established. In comparison with the KDS transform, the normal small-cone transform hardly measures the C–C bond length. The ability to observe more scatterers of a photoelectron emitter by the KDS transform expands the applicability of holographic imaging.

A quantitative structural determination of the Ga-polar 1×1 (0001) surface of GaN is performed by quantitative low energy electron diffraction (LEED). The global best-fit structure is obtained by a new frozen LEED approach connected to a simulated annealing algorithm. The global minimization frozen (GMF) LEED search finds that the ordered structure consists of 1 ML of Ga adatoms at atop sites above Ga-terminated bilayers. The Ga adatoms are bonded with a Ga–Ga bond length of 2.51 Å. The spacings within surface bilayers show a weak oscillatory trend, with the outmost bilayer thickness expanding to 0.72 Å and the next bilayer thickness contracting to 0.64 Å, compared to the bulk thickness of 0.65 Å. The interlayer spacing between the first and second bilayers is 1.89 Å, while the next interlayer spacing is 1.94 Å, compared to the bulk value of 1.95 Å. These results are compared with data from other theoretical and experimental studies.

We review the symmetrized data acquisition method for producing artifact-free Patterson functions from low-energy electron diffraction?(LEED) IV spectra. The differences between LEED Patterson functions and LEED holograms are discussed. This article was originally intended to be part of the special issue on holographic and other direct methods for surface structures using electrons and photons which was published as issue 47, volume 13.

New understanding has been reached on competing forces acting to stabilize the polar surfaces of intrinsic ZnO. To compensate an accumulating dipole moment normal to ionic planes, the Madelung electrostatic force and the bonding ability of undercoordinated Zn ions compete to deplete more Zn atoms from the (0001) face and more O atoms from the $(\mathbf{000}\overline{\mathbf{1}})$ face. In this competition, the former mechanism wins because it provides very low energy binding sites for O ions at face-centered-cubic registries on both surfaces. On the Zn-face, a distorted tetrahedral structure is formed, while on the O-face, a vertical Y structure is formed. In both structures, O ions form the topmost atomic plane. The reconstructed polar surfaces containing these novel structures have cleavage energy of 2.36 J/${\mathrm{m}}^{2}$, comparable to that of nonpolar surfaces and in agreement with experimental observation. An earlier structure found on the Zn-face annealed at below 1000 K is stabilized mainly by the Madelung electrostatic force and is a metastable structure of that surface.

The atomic structure of the Si (111)-[Formula: see text]R30°-Ga surface has been studied by comparing measured low-energy electron diffraction (LEED) intensity (IV) curves with calculated IV spectra using the method of automated tensor LEED. The experimental LEED IV curves used in this work contain many beams and a wide energy range. The results show that the Ga atoms occupy T4 sites, at 2.62 Å above the second-atomic-layer Si atoms. The Ga–Si vertical spacing is 1.44 Å and the bond length between the Ga atom and the first-layer Si atom is 2.52 Å. Large bucklings are found in the first and second Si bilayers below the adatom layer.

Reflection high-energy electron diffraction (RHEED) intensity oscillations have been observed during radio-frequency plasma-assisted molecular beam epitaxy of GaN on its (0001) A face. The starting A (Ga) face was prepared by growing a micrometer-thick GaN layer directly on a low-index 6H–SiC(0001) substrate at 650°C. RHEED intensity oscillations are measured with substrate temperatures less than 550°C in both Ga-limited and N-limited growth conditions. In the N-limited condition, an initial transient high-frequency oscillation is observed before it reaches a steady-state frequency. If the Ga flux is subsequently stopped while keeping the N flux unchanged, a few extra oscillations are recorded. Scanning tunneling microscopy images of surfaces quenched during growth show triangular-shaped islands, verifying a two-dimensional growth mode. At substrate temperatures greater than 550°C, neither island nucleation nor intensity oscillation is observed, suggesting a step-flow growth mode.

Anisotropic growth is observed for GaN(0001) during molecular beam epitaxy for both the step-flow growth mode and two-dimensional (2D) nucleation growth mode. Using scanning tunneling microscopy, we find that in the step-flow growth mode, growth anisotropy strongly influences the shape of terrace edges, making them strikingly different between hexagonal and cubic films. In the 2D nucleation growth mode, anisotropic growth results in triangularly shaped islands. The importance of understanding growth anisotropy to achieve better grown GaN films is discussed.

We examine the differences between low-energy electron-diffraction patterns (holograms) and optical holograms. We show that electron-diffraction patterns in solids are not analogous to optical holograms because of strong dynamical factors. We also show that low-energy electron holograms can be inverted by a large-wave-number small-angle integral transformation. The grid sizes in wave number and angular spaces used in the transformation are derived.

During molecular-beam epitaxy of InN films on GaN(0001) surface, three-dimensional (3D) islands are observed following an initial wetting layer formation. Depending on deposition condition, the 3D islands take different shapes. Pyramidal islands form when excess nitrogen fluxes are used, whereas pillar-shaped islands are obtained when excess indium fluxes are employed. The pillar-shaped islands are identified to represent the equilibrium shape, whereas the pyramidal ones are limited by kinetics. As the size of islands increases, their aspect ratio shows a decreasing trend, which is attributed to a gradual relaxation of strain in the layer by defects.

Island nucleation of GaN on its (0001) surface is studied by scanning tunneling microscopy. A comparison is made between surfaces with and without excess Ga and among surfaces with different excess Ga coverages. Evidence is provided for the change of step characteristics of GaN(0001) by excess Ga adlayers, where the Ehrlich-Schwoebel effect is seen to be mediated by excess Ga coverage. For single Ga adlayer covered GaN(0001) surfaces, nucleation island densities are evaluated, which are used to derive the kinetic barriers of adatom diffusion on a terrace. A barrier of less than $1\phantom{\rule{0.3em}{0ex}}\mathrm{eV}$ is obtained for the system, and the Ga adlayers make GaN growth surfactant mediated.

We observe three different kinds of islands, namely the "bare," "ghost," and "normal" islands, by scanning tunneling microscopy of GaN(0001) prepared under Ga-rich conditions with different excess Ga coverage. They correlate to three different growth regimes. The various morphologies of the islands are suggested to be due to different states of wetting of the surface by excess Ga. The dynamical behavior of the wetting process is followed, and an island size dependence of Ga wetting is noted.

GaN homoepitaxial growth by molecular-beam epitaxy under both excess gallium (Ga) and excess nitrogen (N) conditions is investigated. Based on two-dimensional island shape and surface step structures, we suggest the growth is kinetic-limited under the excess-Ga condition but diffusion-limited in the excess-N regime. The triangular GaN islands and double step bunching seen on surfaces prepared under excess-Ga are attributed to a difference in adatom attachment and/or site exchange rates between $A$ and $B$ steps, which is induced by surfactant Ga adlayers on GaN(0001).

A low-energy electron diffraction (LEED) Patterson function (PF) with multiple incident angles is used to determine the structure of the indium-induced $\mathrm{Si}(111)\text{\ensuremath{-}}4\ifmmode\times\else\texttimes\fi{}1\text{\ensuremath{-}}\mathrm{In}$ reconstructed surface. The experimental PF maps were compared with ones calculated by both single scattering and the tensor-LEED method according to various theoretical models so that one can pick out the right one. It was found that the model proposed by Bunk et al. is the best appropriate one, from which the induced PF maps are similar to the experimental ones. Further analysis of the PF spots obtained from the experimental PF maps directly generates the same model without any presupposition. It indicates that the multiple-incidence Patterson function is an effective method to determine the surface structure. Moreover, detailed atomic positions on surface were deduced from first-principles calculations and tensor-LEED $I\text{\ensuremath{-}}V$ curve simulations. We also compared the results with those of previous investigations. Good agreement was found between them.

Indium nitride (InN) exists in two different structural phases, the equilibrium wurtzite (w) and the metastable zinc-blende (zb) phases. It is of scientific interest and practical relevance to examine the crystal structure of the epifilms during growth. In this paper, we use Patterson function inversion of low-energy electron diffraction I-V curves to reveal the preferential formation of zinc-blende InN wetting layer during the Stranski-Krastanov growth on GaN(0001). For three-dimensional islands nucleated afterwards on top of the wetting layer and for thick InN films, the equilibrium wurtzite structure is observed instead. This in situ revelation of the InN lattice structure is confirmed by ex situ transmission electron microscopy studies. Finally, the formation of zb-InN layer on w-GaN is explained in terms of the strain in the system.

Molecular-beam epitaxy of In{sub x}Ga{sub 1-x}N alloy on GaN(0001) is investigated by scanning tunneling microscopy. The Stranski-Krastanov mode of growth of the alloy is followed, where the newly nucleated three-dimensional islands are initially coherent to the underlying GaN and the wetting layer, but then become dislocated when grown bigger than about 20 nm in the lateral dimension. Two types of islands show different shapes, where the coherent ones are cone shaped and the dislocated ones are pillar like, having flat-tops. Within a certain range of material coverage, the surface contains both coherent and dislocated islands, showing an overall bimodal island-size distribution. The continued deposition on such surfaces leads to the pronounced growth of dislocated islands, whereas the sizes of the coherent islands change very little.

The scaling property of three-dimensional InN islands nucleated on GaN(0001) surface during molecular-beam epitaxy (MBE) is investigated. Due to the large lattice mismatch between InN and GaN (similar to10%), the islands formed from the Stranski-Krastanow growth mode are dislocated. Despite the variations in (residual) strain and the shape, both the island size and pair separation distributions show the scaling behavior. Further, the size distribution resembles that for submonolayer homoepitaxy with the critical island size i=1, suggesting that detachment of atoms is not significant. The above results also indicate strain is insignificant in determining the nucleation and growth of dislocated islands during heteroepitaxy by MBE.

The surface morphology of GaN films grown by molecular beam epitaxy (MBE) is investigated by scanning tunneling microscopy (STM). A comparison is made between flat and vicinal surfaces. The wurtzite structure of GaN leads to special morphological features such as step pairing and triangularly shaped islands. Spiral mounds due to growth at screw threading dislocations are dominant on flat surfaces, whereas for vicinal GaN, the surfaces show no spiral mound but evenly spaced steps. This observation suggests an effective suppression of screw threading dislocations in the vicinal films. This finding is confirmed by transmission electron microscopy (TEM) studies. Continued growth of the vicinal surface leads to step bunching that is attributed to the effect of electromigration.

For heteroepitaxy of InN on GaN(0001) by molecular-beam epitaxy, the lattice misfit strain is relieved by misfit dislocations (MDs) formed at the interface between InN and GaN. Imaging by scanning tunneling microscopy (STM) of the surfaces of thin InN epifilms reveals line feature parallel to ⟨112¯0⟩. Their contrast becomes less apparent for thicker epifilms. From the interline spacing as well as a comparison with transmission electron microscopy studies, it is suggested that they correspond to the MDs beneath the surface. The STM contrast originates from both the surface distortion caused by the local strain at MDs and the electronic states of the defects.

crystal’s electronic band states relative to the 4f N or 4f N-1 5d 1 states responsible for the ion’s optical transitions is important for understanding the properties and performance of each material since energy and electron transfer between these states influences the material’s efficiency and stability. 1 Little is known about the relationships between these states, but there is growing motivation to explore these properties for developing ultraviolet laser materials, phosphors for applications including plasma displays and mercury-free lamps, scintillator materials for medical imaging, and optical data processing and storage technologies based on photorefractivity or photon-gated photoionization holeburning. Continued advances in optical technologies require knowledge of the systematic trends and behavior of rare-earth energies relative to crystal band states so that the properties of current materials may be fully understood and new materials may be logically developed. We have recently initiated a systematic study of the relative energies of the rare-earth ions’ electronic states and the host band states in optical materials using resonant electron photoemission spectroscopy (RPES). 2,3 RPES directly determines the energies of all occupied electronic states relative to a common energy reference and can unambiguously separate and assign spectral features to a particular electronic state. 4 Figure 1 presents results for yttrium aluminum garnet (YAG), the most important host crystal for solidstate lasers. Circles represent measured binding energies of the rare-earth 4f N ground state relative to the valence band maximum (the host’s highest energy occupied state). These results have led to an empirical model that successfully describes the rareearth binding energies in optical materials with two parameters: one describes a constant shift experienced by all rare-earth ions and the second describes a smaller dependence on the rare earth’s ionic radius. These empirical parameters may be determined from measurements on just two different rare-earth ions, or, in certain cases, simply from measurements on the host crystal itself. With parameters determined from our measured

Raman Scattering Extrema Method (RSEM), originally proposed by author, to determine the orientation of a thin layer with diamond structure, is extented to zincblende structure. The Raman scattering intensities of LO and TO phonons as functions of both the orientation of this layer and the polarization direction of the incident light have been derived for a zincblende-structured thin layer. Then the orientation of the layer is determined by means of the extrema of these functions. The obtained results for GaP wafers using this method is compared with that determined by the X-ray diffraction method.