Search

The morphological evolution of Ge layers growing on the SiO2/Si(100) substrate by photo-excited chemical vapor deposition was traced through an analysis of pseudodielectric functions measured by real-time spectroscopic ellipsometry. Simulation and fitting were carried out on multiple samples with various Ge film thicknesses as well as on sequential optical spectra from a sample with an incremental buildup of Ge atoms on one substrate. Single- and two-layer models involving crystalline Ge (c-Ge), amorphous Ge (a-Ge), and void components were employed under the Bruggeman effective medium approximation to represent wetting of the SiO2 surface, nucleation of Ge seeds for the subsequent dot/island formation, and steady-state dot/island growth. A combination of c-Ge and a-Ge represents intermediate crystallinity, and void represents vacant space between dots/islands. A single-layer model with a mixture of c-Ge, a-Ge, and void components was used for crude estimation of the composition from which the time evolution of the volume fraction of the components was derived. However, fitting in the early growth stage resulted in an unrealistic structure, indicating that the dielectric function of the thin hydrogenated Ge network layer was very different from those of c-Ge and a-Ge. The optical spectra of dots/islands at the intermediate growth stage could be reproduced by a two-layer model consisting of a (a-Ge + void) layer overlaid on a (c-Ge + void) base layer. The real-time Ψ–Δ trajectories of ellipsometric angles monitored at a photon energy of 3.4 eV consisted of three branches. They could be reproduced by assuming the growth of an outer layer with an appropriate composition. After wetting on SiO2 (branch 1), the Ge seed layer nucleates while the volume fraction of Ge rapidly decreases from 70% to 25% with proceeding growth (branch 2). Then, the volume fraction of Ge continuously increases up to 65%, eventually reaching steady-state dots/island growth (branch 3) [ABSTRACT FROM AUTHOR]

We investigated the photoluminescence of Eu-doped ZnO and ZnGaxO1 + 1.5x (x = 1.8–2.8) films upon bandgap excitation to identify the occupation sites of Eu3+ ions through varying the host crystal material, substrate material, and reactant gas. Eu3+ emissions from ZnGaxO1 + 1.5x:Eu films deposited on Si and SiO2 substrates had similar spectral shapes, but their intensities were different, correlated with different defect emission levels. When depositions were carried out in H2O gas, the Eu3+ emissions from the c-axis oriented ZnO:Eu films consisted of a very sharp main peak (P1) at 612 nm and a sub-peak (P2) at 619 nm for 5D0 → 7F2 transition, whereas those from randomly oriented polycrystalline ZnGaxO1 + 1.5x:Eu films were broader and could be divided into four components (P1, P2, P3, and P4), of which the wavelengths of P1 and P2 matched those of the ZnO:Eu films. The availability of both Ga3+ and Zn2+ sites in the ZnGa2O4 lattice expands the variance of chemically distinct sites that can be occupied with Eu3+ compared with only divalent Zn2+ sites in the ZnO crystal. Films deposited with O2 exhibited lower emission intensities with a narrower spectrum width than those deposited with H2O. This behavior was correlated with the X-ray diffraction peak width of ZnGa2O4(311). High-quality crystal lattices formed with O2 allowed only a few sites to accommodate Eu3+ ions. In contrast, OH− and H+ derived from H2O will modify crystallinity to generate various emission-active sites, resulting in intense and broad emissions. Possible sites for the four components are discussed. [ABSTRACT FROM AUTHOR]

The kinetics of roughening and smoothing in Ge(001) surface morphology during synchrotron-radiation-excited Ge[H.sub.4]-source homoepitaxy are examined. Roughening is terminated when all dangling bonds are tied up with H atoms and in smoothing process, valleys are preferentially embedded and a two-dimensional (2D) flat film is eventually recovered with no defects in the Ge film.

A simple method is presented to determine the linear electro-optic coefficient of thin films with spectroscopic ellipsometry. The technique allows noncontact and nondestructive measurement of the as-grown states of films without the need to fabricate a waveguide.

SiH2CI2 was used as a source gas to investigate the contribution of certain dangling-bond regeneration channels during synchrotron-radiation excited epitaxial growth of Si.

Er3+-doped Bi2O3 films were sputter deposited on Si(100) substrates at room temperature with H2O vapor as an oxygen source gas. Crystal phases appearing after postannealing in an O2 atmosphere included single-phases of α-Bi2O3, γ-Bi2O3, and δ-Bi2O3, as well as a mixed phase of α-Bi2O3 and γ-Bi2O3. Selection of the crystal phase was possible in terms of H2O pressure and postannealing temperature. Photoluminescence spectra from Er3+ ions excited at a laser wavelength of 532 nm revealed distinct spectral features specific to the crystal phases. A clear crystal-field splitting feature consisting of eight emission lines was observed in PL spectra from sufficiently oxidized α-Bi2O3:Er films, indicating that Er3+ ions occupied low-symmetry C2v sites of Bi3+. The optimum annealing temperature for optical activation of Er3+ ions was between 400 and 450 °C. The emission intensity of α-Bi2O3:Er deposited with H2O was seven times higher than that of α-Bi2O3:Er deposited with O2 probably because larger numbers of Er3+ could substitute Bi3+ sites under reduced condition. Increasing deposition temperature lowered the emission intensity due to the loss of OH and H species from the as-deposited films. The emission spectra of γ-Bi2O3:Er were featureless and its emission intensity was one order of magnitude lower than that of α-Bi2O3:Er. The low-temperature phase of δ-Bi2O3:Er turned out to be entirely emission inactive. Reactions at the interface between the Bi2O3 film and the Si substrate were promoted upon postannealing at 500 °C. The resulting Bi2SiO5:Er exhibited a weak emission spectrum with four emission lines, which reflected occupation at tetragonal Bi3+ sites in the (Bi2O2)2− layers. [ABSTRACT FROM AUTHOR]

We studied the role of hydrogen-containing species (OH and/or H) in promoting photoluminescence (PL) from 1 at. % Eu3+ ions doped in ZnO thin films. The hydrogen concentration in the films was systematically changed by varying the substrate temperature and the vapor pressure of H2O gas supplied during sputter deposition. The correlation between the PL spectra via bandgap excitation and the degree of oxidization/hydrogenation was investigated by x-ray diffraction and Fourier-transform infrared spectroscopy. Films deposited at room temperature under H2O partial pressures higher than 10−2 Pa were sufficiently hydroxylated, as confirmed by the appearance of diffractions peaks from Zn(OH)2 coexisting with ZnO(002). Eu3+ emissions were observed after post-annealing in a vacuum or O2 atmosphere. When the H2O pressure was lower than 10−2 Pa, the ZnO:Eu films were so oxygen-deficient as to exhibit a metallic character, which deactivated the Eu3+ emission. Deposition at temperatures above 200 °C reduced the OH and/or H species incorporated in the ZnO films and only a faint Eu3+ emission was observed. The H2O pressure under which a sharp Eu3+ emission could be obtained was between 1.0 and 2.5 × 10−2 Pa if subsequent post-annealing was done in a vacuum. For more oxidized films deposited at 3.5 × 10−2 Pa, reduction by post-annealing in an H2 atmosphere was effective to generate a sharp and intense Eu3+ emission signal through reduction and hydrogenation, confirming that a moderate oxidization/hydrogenation state is a necessary condition. Codoped hydrogen species will facilitate substituting Zn2+ sites with emission-active Eu3+ ions and stabilize them. [ABSTRACT FROM AUTHOR]

Morphological structures of two-dimensional (2D) Ge thin films and three-dimensional (3D) Ge dots/islands grown on SiO2 substrates were analyzed with UV–visible spectroscopic ellipsometry. The pseudo-dielectric functions (〈ε〉 = 〈ε1〉 + i〈ε2〉) were calculated under the Bruggeman effective medium approximation. The 〈ε〉 spectra of 2D films were well fitted with a simulation assuming a single-layer or two-layer model. The single-layer model has a mixture of c-Ge, a-Ge, and void components. The two-layer model has a base-layer consisting of 100% c-Ge and a surface-roughness layer consisting of a mixture of c-Ge, a-Ge, and voids. In the case of 3D films, the overall shape of the 〈ε〉 spectra could only be reproduced by using a two-layer model: a base-layer consisting of mixtures of c-Ge and a-Ge with a surface-roughness layer consisting of c-Ge, a-Ge, and voids. The degree of disagreement correlated with the root mean square surface roughness measured by atomic force microscopy. [ABSTRACT FROM AUTHOR]