Library

Notes: Pressure-tuned resonance Raman scattering (RRS) experiments in Ga1−xInxAs/Ga1−yAlyAs strained-layer quantum-well samples with x=0.15 and y=0.15, on a GaAs substrate, are reported. The photoluminescence (PL), 2LO-, and LO-phonon Raman-scattered intensities were followed as a function of pressure using the diamond cell. In backscattering geometry and with (100) samples, three sharply defined 2LO-phonon resonances are observed, at 2.1, 3.8, and 5.2 GPa, with 647.1-nm excitation ((h-dash-bar)ωL=1.916 eV), and these are identified to be from the AlGaAs, GaAs, and InGaAs layers, respectively. At the above pressures, the 2LO peak of the layer falls right on top of the PL peak of the corresponding layer, revealing that the resonance occurs when E0=(h-dash-bar)ωS (2LO). For the LO-phonon intensity, maxima are observed at 2.8 and 4.7 GPa, respectively, from the AlGaAs layer and GaAs substrate. While the LO-resonance profile (pressure versus phonon intensity) for the AlGaAs layer is narrow, the profile of the GaAs substrate layer is highly asymmetric. The InGaAs LO-resonance is totally masked by the strong GaAs LO phonon. The pressure dependence of the E0 gap, determined from PL data, reveals that the resonance condition for 2LO is E0=(h-dash-bar)ωS (2LO) and, for LO, E0=(h-dash-bar)ωL. Results with samples of different Al content and (h-dash-bar)ωL are consistent with the above conditions. A brief discussion of the theory of the RRS cross section is given, and it is suggested that the observed 2LO-resonance scattering is dominated by Frölich interaction, and that it could be a double resonance. The asymmetry in the LO resonance of the GaAs substrate, we believe, is due to the pressure-induced transparency, and this is a problem when dealing with pressure-tuned RRS, in multiple-quantum-well structure with GaAs or AlGaAs substrates. The usefulness of the 2LO resonance in locating the E0 gap is indicated, and the indistinguishability between luminescence and Raman scattering at the center of the 2LO resonance is pointed out. The need to study photoluminescence and Raman scattering in pressure-tuned RRS experiments is emphasized.

Notes: A simple three-dimensional vapor phase model is used to interpret and clarify the selective area growth process. The model predicts both normal and anomalous profiles of thickness and composition, including long range effects. These are verified by an extensive set of experiments. © 1999 American Institute of Physics.

Notes: We report growth of ultrapure GaAs by molecular beam epitaxy (MBE) and show that the arsenic source has a dramatic effect on the purity. With every thing else the same, by changing the arsenic source from 6N grade double refined chunks to a 7N grade slug which closely fits the 35 cm3 crucible, the deep electron trap and the residual acceptor densities were reduced by nearly two orders of magnitude. When intentionally doped with Si to an electron density of 3×1013 cm−3, the measured mobility at 77 K was 205 600 cm2 V−1 s−1, which increased to 294 700 cm2 V−1 s−1 at 42 K. These mobilities are the highest ever observed in intentionally n-doped MBE GaAs. Low-temperature photoluminescence studies of both undoped and Si-doped GaAs layers grown with the As slug show a marked decrease in acceptor- and defect-related luminescence over that observed when the As chunks are used. Our study conclusively proves that the major source of residual acceptor impurities and typical MBE GaAs traps observed by deep level transient spectroscopy is the chemical impurities present in the arsenic source, although the exact nature of these impurities is yet to be identified. These impurities or their complexes with the native crystal defects, and not the native defects alone, are responsible for the deep traps.

Notes: Thresholdlike behavior of photoluminescence (PL) in heterostructure wafers is studied. It is shown that strictly speaking there is no PL threshold, and the thresholdlike dependence of PL on the pump power results from the combination of three factors: the PL spreading along the wafer surface, the change in the radiative fraction of electron-hole recombination, and the restricted aperture of the PL detector. The first two factors were found to be dominating in the wafers studied. © 1997 American Institute of Physics.

Notes: High-quality GaAs/AlGaAs quantum well and modulation-doped heterostructures have been grown by low-pressure organometallic vapor phase epitaxy (OMVPE) using trimethylamine alane (TMAA) as a new aluminum source. TMAA is an alternative to the conventional organometallic precursors and offers the advantage of substantially reduced oxygen and carbon incorporation in AlGaAs. Intense photoluminescence (PL) with narrow linewidths at 2 K was observed from multiple quantum well samples with well widths of 1.5–10 nm. Transmission electron microscopy of a fifty period superlattice (4 nm GaAs/44 nm Al0.18Ga0.82As) revealed abrupt interfaces and excellent well-to-well thickness uniformity. Selectively doped heterostructure transistors (SDHTs) fabricated on the modulation-doped structures exhibited a maximum extrinsic transconductance of 339 mS/mm for a 1-μm-gate length at 300-K, the highest reported for OMVPE grown devices. A unity current gain cutoff frequency, ft, of 16 GHz and a maximum frequency of oscillation, fmax, of 23 GHz were obtained for these SDHTs.

Notes: Performance characteristics of a pseudomorphic p-type, normal incidence, Ge0.25Si0.75/Si strained-layer quantum well infrared photodetector on (001) Si is described for 20≤T≤77 K. The device shows broadband photoresponse (8–14 μm) which is attributed to strain and quantum confinement induced mixing of heavy, light, and split-off hole bands. Typical device responsivity at λ=10.8 μm is ∼0.04 A/W over the 20–77 K temperature range. A detectivity D*λ=3.3×109 cm (square root of)Hz/W was measured at a bias of −2.4 V for a temperature of 77 K at λ=10.8 μm and no cold shield. Room temperature FTIR measurements yield a quantum efficiency η≈3.1% at λp≈8 μm at 300 K.