Raman spectra were measured on arsenic-implanted silicon with micro-Raman spectroscopy in the backscattering mode and with macro-Raman spectroscopy. A peak is observed between 505 and 510 cm−1 with 488 and 514.5 nm excitation. This peak and a related peak from the substrate at about 520 cm−1 are seen in selected regions of the implanted samples when the implant dose is above 2 × 1014 As/cm2. These features may be due to a long room temperature anneal, as they are absent in recently prepared samples. Possible explanations for the features are presented.

We have measured micro-photoluminescence (PL) and micro-Raman spectra on the cross section of porous silicon multilayers to sample different layer depths. We find noticeable differences in the spectra of layers with different porosity, as expected from the quantum confinement of electrons and phonons in silicon nanocrystals with different average sizes. The PL emission band gets stronger, blue shifts, and narrows at the high porosity layers. The average size can be estimated from the shift. The Raman phonon band at 520 cm−1 weakens and broadens asymmetrically towards the low energy side. The line shape can be related quantitatively with the average size by the phonon confinement model. To get a good agreement with the model we add a band at around 480 cm−1, which has been attributed to amorphous silicon. We also have to leave as free parameters the bulk silicon phonon frequency and its line width, which depend on temperature and stress. We reduced laser power to eliminate heating effects. Then we use the change of frequency with depth to monitor the stress. At the interface with the substrate we find a compressive stress in excess of 10 kbar, which agrees with the reported lattice mismatch. Finally, average sizes are larger than those estimated from PL.

Porous layers and free-standing membranes were fabricated by anodic etching of n-GaP substrates in a sulphuric acid solution. Micro-Raman analysis of the interaction between the longitudinal optical phonons and plasmons in porous membranes allowed us to obtain specific information about the electro-optical properties of microstructured GaP. In particular, apart from the carrier exhausted areas surrounding the pores, the existence of conductive regions was demonstrated. A comparative analysis of the secondary electron and panchromatic cathodoluminescence (CL) images evidenced an increase in the emission efficiency caused by porosity. Data concerning the spectral distribution of CL in bulk and porous samples are presented.

Raman scattering studies have been recently used to relate the strain in the semiconductor layer structure with the line shape of allowed modes. It can yield important information about the nature of the solid on a scale of the order of a few lattice constants. It can also provide an evaluation on the carrier concentration in the channel layer of high electron mobility transistors (HEMTs). In this investigation, Raman scattering was used to study the effect of varying the In mole fraction (x) from 0.53 to 0.81 in the InxGa1−xAs channel layer of InGaAs/InP heterostructures. The effect of varying the doping concentration in the donor layer from 6×1017/cm3 to 2.5×1018/cm3, and the effect of varying the In0.75Ga0.25As channel thickness from 140 Å to 260 Å are also reported.

A two-mode Raman characteristic for all InxGa1−xAs/InP HEMTs is clearly seen, with the two LO modes (InAs-like LO and GaAs-like LO) located at 229cm−1 and 268.6cm−1, respectively. At a Raman frequency of 347cm−1, a small peak is observed due to InP LO mode. As the In composition increases from 0.53 to 0.81, the InAs-like LO mode peak intensity increases while that of GaAs-like LO mode decreases. The peak intensity ratio of InAs-like LO mode and GaAs-like LO mode increases from 0.78 to 1.10. By increasing the doping concentration in the donor layer (ND), there is also an increase in the carrier concentration in the InGaAs channel assuming that the donors are fully ionised. The coupled mode between the InGaAs longitudinal optical phonons and electrons in the InGaAs channel shifts continuously to a low wave number with the increasing ND in the InP donor layer. The increase in the lnGaAs channel thickness from 140Å to 260Å causes the InAs-like LO mode peak to shift to a lower wave number from 235.5 to 228.5cm−1. There is no change in the GaAs-like LO peak position located at 268.4cm−1.

We present optical and microstructural characterization of nanocrystalline silicon superlattices (nc-Si SLs). Our samples have better than 5 % Si nanocrystal size distribution and a long range order along the direction of growth provided by periodically alternating layers of Si nanocrystals and SiO2. Flat and chemically abrupt nc-Si/SiO2 interfaces with a roughness of < 4Å are confirmed by transmission electron microscopy (TEM), Auger elemental microanalysis, X-ray small angle reflection, and low-frequency Raman scattering. Photoluminescence (PL) in our structures has been studied in details including time-resolved and steady-state PL spectroscopy in a wide range of temperature, excitation wavelength and power. Resonantly excited PL spectra show phonon steps proving that the PL originates in Si nanocrystals. Electrical measurements show signature of phonon-assisted tunneling proving low defect density nc-Si/SiO2 interface.

This paper presents some of the widely used characterization tools for antimonide based III-V compounds, focusing on characterization of defects that limit device operation. The effect of microstructure on the electro-optical properties of GaSb and GaInAsSb important in devices is emphasized.

SnxGe1−x alloys are grown coherent on Ge(001) substrates by conventional molecular beam epitaxy. The lattice mismatch between SnxGe1−x and Ge produces a biaxial compression of the alloy in the plane of the substrate and a uniaxial elongation in the growth direction. The change of Eg resulting from the biaxial compression is modeled with deformation potential theory for coherent SnxGe1−x on Ge(001) and coherent SnxGe1−x on Ge(111). For SnxGe1−x on Ge(001), the decrease in the energy band gap due to strain is small, 16 meV for x = 0.10, and Γ7 and L6 shift uniformly. In the case of SnxGe1−x on Ge(111), Γ7 and L6 do not shift uniformly; a large uniaxial splitting occurs at L6, 200 meV for x = 0.10. Another interesting result of a uniaxial strain along [111] is the absence of a direct to indirect energy band gap transition for x < 0.30. Fourier transform infrared spectroscopy transmission measurements of coherent SnxGe1−x on Ge(001) confirm the strain-induced change in the energy band gap is a small effect.

The evolution of the cathodoluminescent emission bands of porous silicon at 420–480 nm, 540 nm and 640 nm under oxidizing and inert atmosphere annealings has been investigated and correlated to the structural reconstruction of the porous silicon surface during the treatments. Dry oxidation at low temperature of 450°C determines a less defected bonding interface structure of the growing oxide matrix comparatively with an oxidation treatment at 1000°C. The optical absorption band at 883 cm−1 observed in the 1000°C oxidized porous silicon is correlated to the presence of the interface defect centers. The quenching of the 540 nm and 640 nm emission bands in the high temperature oxidized porous silicon results from nonradiative transitions in the associated defect levels.

We present the preliminary results of our investigation of producing nanoclusters of gold and tin in 6H-SiC. In this work we study the changes in the linear optical properties as a function of ion fluence and post-implantation heat treatment. The optical absorption is complemented by micro-Raman spectroscopy to investigate lattice damage and recovery. We implanted 1.0 MeV Au and 120 and 160 keV Sn into the Si face of SiC at 30 or 500°C followed by annealing at various temperatures. Using optical absorption spectrophotometry, we determined the location of the absorption band for each metal nanocluster in SiC. Elevated temperature implantation greatly reduced optical absorption due to ion implantation induced defects. Using Mie theory, we determined the index of refraction in the implanted volume and estimated the size of the Au nanoclusters as 5 nm in diameter. Rutherford backscattering analysis of the oxide layer grown during annealing showed an increase of 20 percent in on the implanted side. It was found that the Au clusters occurred only in the oxide layer.

In this work, two optical techniques, Raman spectroscopy and ultraviolet reflectance, have been used to characterize the solid phase crystallization kinetics and the microstructure of SiGe films deposited by LPCVD on oxidized Si wafers. The results have been compared to those obtained by X-ray diffractometry. The Ge fraction of the films (x) was in the 0−0.38 interval. The samples were crystallized at temperatures ranging from 525 to 600 °C. The crystallization kinetics follows Avrami's model. Two different behaviours have been observed depending on the Ge fraction of the films and the crystallization temperature: a) Either the three experimental techniques yield similar results, or b) the crystallization process, as monitored by UV reflectance and Raman spectroscopy, exhibits a greater incubation time than the one obtained if X-ray diffractometry is used. The results are discussed in terms of the identification of the nucleation sites, taking into account the probe depth of the different techniques and the preferred orientations of the grains. These techniques have also been used to characterize the presence of defects, the overall crystallinity and the surface roughness of the fully crystallized films. The results are correlated to the grain morphology and grain size, obtained by means of transmission electron microscopy.

Raman spectroscopy has often been used to study the damage to semiconductors induced by ion implantation. Off-axis, macro-Raman spectra reveal extensive damage to the silicon lattice, consistent with many literature reports. However, when the same samples were analyzed in the backscattering mode by micro-Raman spectroscopy, evidence was found for orientational dependent lattice damage and an unusual defect structure. P/O micro-Raman spectra reveal the spatially-varying appearance of a band between 505 and 510 cm−1 always accompanied by that of the silicon optical mode at 520 cm−1.

The large visible optical penetration depth makes it difficult to isolate the Raman signals of SiC epilayers from those of the substrates such as SiC, sapphire and Si when visible laser lines are used. In this work, confocal micro-Raman was used to characterize 3C, 4H and 6H SiC layers on different substrates with enhanced lateral resolution (˜ 0.8 μm) and depth resolution (˜ 2 μm). Both lateral and depth variation of the free electron concentration and scattering time were measured from n− SiC epi layers on n+ SiC substrates and from H+ implanted SiC. A defect mode induced by oxidation process was also analyzed as function of the depth and the lateral position.

GaSb p-n junctions formed by Zn diffusion in Te-doped n-GaSb single crystalline wafers have been characterized by cathodoluminescence (CL) microscopy and by scanning tunneling spectroscopy. CL plane-view observations of the Zn diffused side enable to study the homogeneity of the diffusion treatment. Spectra recorded by current imaging tunneling spectroscopy (CITS) in the n and p sides of the junction, clearly show the respective conductive behaviour and provide information on the local surface band gaps. Results are related to the diffusion profile measured by secondary ion mass spectrometry (SIMS).

InxGa1−xAs epilayers grown on GaAs(100) were studied by cathodoluminescence (CL) spectroscopy and imaging technique with 0.8 nm spectral resolution, using a transmission electron microscope. Linear features appear in the monochromatic CL image taken by the emission from the InxGa1−x.As layers, and do not appear in those from the GaAs layers. There is no direct correlation between the dark-line contrast in the panchromatic CL image (due to misfit dislocations) and the strong line contrast in the monochromatic CL images of the InxGa1−xAs layers. A peak wavelength shift in the CL spectrum was observed as the electron probe was moved across the linear features. The linear features also appear in a thin sample where the misfit dislocations are removed by ion milling, which clearly reveals that the strong line contrast is not directly due to the misfit dislocation. From those results the linear features in the monochromatic CL image are considered to be due to compositional fluctuations of the In concentration in the InxGa1−xAs layer.

Photoreflectance spectroscopy has been peformed on a series of Cd1−xMnxTe/Cd1−yMnyTe superlattices. Samples were grown on (001) GaAs substrates by molecular-beam epitaxy with different barriers (x=0.3 to 0.8) and wells (y=0 to 0.01). After taking into consideration the strain-induced and quantum confinement effects, the exciton transition energies of the heavy and light holes can be determined using envelope-function calculations. The calculations are in good agreement with the photoreflectance measurement results. These results show that photoreflectance is a powerful probe for the study of quantized state structures in superlattices.

InP/InGaAs heterojunction bipolar transistor (HBT) wafers grown by metal-organic vapor phase epitaxy (MOVPE) have been characterized by photoreflectance (PR) spectroscopy. The intensity of PR signals from the InP emitter and InGaAs collector layers of the HBT wafers decreased with increasing emitter growth temperature and showed a linear positive correlation with the HBT current gain. On the other hand, the intensity of PR signals from the n-InP single layers scarcely changed with increasing InP growth temperature. The change in the PR intensity of the emitter and collector layers is expected to reflect the crystal quality of the adjacent InGaAs:C base layer, which determines the HBT current gain. Thus, the present PR method is eminently suitable for the nondestructive analysis of the current gain of InP/InGaAs HBT wafers.

A number of research efforts have been focused on self-assembled germanium quantum dots in which indirect optical transitions take place across the band gap. However, many questions regarding the confined electronic state transitions of Ge quantum dots still remain unanswered. In the present report, we have deposited ten alternating layers of crystalline Ge quantum dots embedded in an Al2O3 or an AIN matrix on sapphire substrates by pulsed laser deposition. The average dot sizes (73 Å to 260 Å) were controlled by the laser energy density, deposition time and substrate temperature. The spectral positions of both the E1 and the E2 transitions in the absorption spectra at room temperature and 77 K shift toward higher energy (ΔE1=1.19 eV, ΔE2 =0.57 eV) as the Ge dot size decreases (73 Å). Structural analysis using transmission electron microscopy and atomic force microscopy and the interpretation of optical absorption measurements in terms of quantum confinement of carriers in both transitions are presented.

In this paper, we showed the significant reduction of the energy spacing between ground state and excited state emissions from InAs/GaAs quantum dots due to interface interdiffusion induced by thermal treatment. In addition, the strong narrowing of the luminescence linewidth of the ground state and excited state emissions from the InAs dot layers for the annealed samples indicates an improvement of the size-distribution of the QDs. Large blue-shift of the energy positions of both emissions was also observed. High resolution X-Ray diffraction experiments give strong evidence of the interface atom interdiffusion in the annealed samples. This work shows ability to tune the wavelength for applications like infrared detectors and lasers based on intrasubband transitions of self-assembled QDs.

Combined TSC and PICTS measurements have been used to determine the activation energies (Et) and capture cross-sections (σ) of the trap levels inside the bandgap of CVD diamond in the energy range 0.4−0.7 eV. High temperature TSC analysis has been performed to determine the trap parameters in the energy range from 0.9 to 1.3 eV. A fieldmap in the Et−σ plane has been obtained from the combination of the TSC and PICTS data depicting the regions corresponding to two isolated trap levels and to a continuous distribution of states. The concentrations of defects have been calculated from the TSC signals and the measurement of the charge collection efficiency of the diamond samples.

Small angle light scattering has been used to quantitatively study microvoids in polycrystalline CVD diamond. Guinier's approximation was used to calculate the average radius of gyration of these defects for diamond films made by both DC arc-jet and microwave CVD assuming a spherical defect geometry. Values of the radius of gyration varied between approximately 1 and 5 μm and were found to correlate with the thickness, relative transmission and thermal conductivity measured for the films. Some inconsistencies were observed between microwave and DC arc-jet materials which may be related to fundamental differences in the growth processes. This represents the first quantitative analysis of such defects in polycrystalline films and holds great promise for improving our understanding of the diamond CVD processing and properties.