Library

Notes: We have studied the electrical properties of GaAs/Al0.4Ga0.6As double-barrier resonant tunneling structures incorporating finite superlattices in the contact regions. The superlattices effectively act as energy filters defining the initial and final tunneling states. We have investigated an asymmetric device with one (emitter) superlattice and a symmetric device with two (emitter and collector) superlattices. These show significantly improved J(V) properties compared to other double-barrier structures and the superlattice tunnel diode.

Notes: Arrays of GaAs pyramids with square (001) bases of length 1–5 μm have been fabricated by molecular beam epitaxy regrowth on pre-patterned GaAs (001) substrates. The optical properties of the pyramid faces have been studied by microreflection and microtransmission imaging measurements with light (λ=900–1000 nm) incident through the pyramid base. Digitized charge coupled device images indicate that total internal reflection occurs at the {110} pyramid facets and that their reflectivities are greater than 80%, provided overgrowth of the facets does not occur. These properties suggest that such structures may be suitable as the top mirror in novel micron-scale vertical microcavity devices. © 2001 American Institute of Physics.

Notes: A range of experimental techniques has been used to measure point defect concentrations in GaAs layers grown at low temperatures (250 °C) by molecular-beam epitaxy (LT-GaAs). The effects of doping on these concentrations has been investigated by studying samples containing shallow acceptors (Be) or shallow donors (Si) in concentrations of ∼1019 cm−3. Material grown under As-rich conditions and doped with Be was completely compensated and the simultaneous detection of As0Ga by near-band-edge infrared absorption and As+Ga by electron paramagnetic resonance confirmed that the Fermi level was near the midgap position and that compensation was partly related to AsGa defects. There was no evidence for the incorporation of VGa in this layer from positron annihilation measurements. For LT-GaAs grown under As-rich conditions and doped with Si, more than 80% of the donors were compensated and the detection of SiGa–VGa pairs by infrared localized vibrational mode (LVM) spectroscopy indicated that compensating VGa defects were at least partly responsible. The presence of vacancy defects was confirmed by positron annihilation measurements. Increasing the Si doping level suppressed the incorporation of AsGa. Exposure of the Be-doped layer to a radio-frequency hydrogen plasma, generated a LVM at 1997 cm−1 and it is proposed that this line is a stretch mode of a AsGa–H–VAs defect complex. For the Si-doped layer, two stretch modes at 1764 and 1773 cm−1 and a wag mode at 779 cm−1 relating to a H-defect complex were detected and we argue that the complex could be a passivated As antisite. The detection of characteristic hydrogen-native defect LVMs may provide a new method for the identification of intrinsic defects. © 1995 American Institute of Physics.

Notes: InxGa1−xAs layers (0≤x≤0.37) doped with carbon (〉1020 cm−3) were grown on semi-insulating GaAs substrates by chemical beam epitaxy using carbon tetrabromide (CBr4) as the dopant source. Hall measurements imply that all of the carbon was present as CAs for values of x up to 0.15. The C acceptors were passivated by exposing samples to a radio frequency hydrogen plasma for periods of up to 6 h. The nearest-neighbor bonding configurations of CAs were investigated by studying the nondegenerate antisymmetric hydrogen stretch mode (A−1 symmetry) and the symmetric XH mode (A+1 symmetry) of the H–CAs pairs using IR absorption and Raman scattering, respectively. Observed modes at 2635 and 450 cm−1 had been assigned to passivated Ga4CAs clusters. New modes at 2550 and 430 cm−1 increased in strength with increasing values of x and are assigned to passivated InGa3CAs clusters. These results were compared with ab initio local density functional theory. Modes due to AlInGaCAs clusters were detected in samples containing grown in Al and In. These results demonstrate that for InGaAs, CBr4 is an efficient C doping source since both In–CAs bonds as well as Ga–CAs bonds are formed, whereas there is no evidence for the formation of In–CAs bonds in samples doped with C derived from trimethylgallium or solid sources. © 1996 American Institute of Physics.

Notes: The local environment of CAs acceptors in InxGa1−xAs has been determined from the localized vibrational modes (LVMs) of both isolated CAs impurities and H–CAs pairs using infrared (IR) absorption and Raman scattering techniques. In as-grown layers, a single LVM due to CAs was observed which broadened and shifted to lower energies with increasing x. The introduction of hydrogen led to the formation of H–CAs pairs and a single antisymmetric A1− mode (stretch) and a single symmetric A+1 mode (XH) were observed for all samples. All the LVMs were identified with carbon in CAsGa4 cluster configurations implying that less than 5% of the detectable carbon atoms are present in clusters incorporating one or more CAs–In bonds. © 1995 American Institute of Physics.

Notes: Abstract This paper considers the problem of minimizing a functionalI which depends on the statex(t), the controlu(t), and the parameter π. Here,I is a scalar,x ann-vector,u anm-vector, and π ap-vector. At the initial point, the state is prescribed. At the final point, the statex and the parameter π are required to satisfyq scalar relations. Along the interval of integration, the state, the control, and the parameter are required to satisfyn scalar differential equations. Asequential algorithm composed of the alternate succession of gradient phases and restoration phases is presented. This sequential algorithm is contructed in such a way that the differential equations and boundary conditions are satisfied at the end of each iteration, that is, at the end of a complete gradient-restoration phase; hence, the value of the functional at the end of one iteration is comparable with the value of the functional at the end of any other iteration. In thegradient phase, nominal functionsx(t),u(t), π satisfying all the differential equations and boundary conditions are assumed. Variations Δx(t), Δu(t), Δπ leading to varied functions $$\tilde x$$ (t),ũ(t), $$\tilde \pi$$ are determined so that the value of the functional is decreased. These variations are obtained by minimizing the first-order change of the functional subject to the linearized differential equations, the linearized boundary conditions, and a quadratic constraint on the variations of the control and the parameter. Since the constraints are satisfied only to first order during the gradient phase, the functions $$\tilde x$$ (t),ũ(t), $$\tilde \pi$$ may violate the differential equations and/or the boundary conditions. This being the case, a restoration phase is needed prior to starting the next gradient phase. In thisrestoration phase, the functions $$\tilde x$$ (t),ũ(t), $$\tilde \pi$$ are assumed to be the nominal functions. Variations Δ $$\tilde x$$ (t), Δũ(t), Δ $$\tilde \pi$$ leading to varied functions $$\hat x$$ (t),û(t), $$\hat \pi$$ consistent with all the differential equations and boundary conditions are determined. These variations are obtained by requiring the least-square change of the control and the parameter subject to the linearized differential equations and the linearized boundary conditions. Of course, the restoration phase must be performed iteratively until the cumulative error in the differential equations and boundary conditions becomes smaller than some preselected value. If the gradient stepsize is α, an order-of-magnitude analysis shows that the gradient corrections are Δx=O(α), Δu=O(α), Δπ=O(α), while the restoration corrections are $$\Delta \tilde x = O(\alpha ^2 ), \Delta \tilde u = O(\alpha ^2 ), \Delta \hat \pi = O(\alpha ^2 )$$ . Hence, for α sufficiently small, the restoration phase preserves the descent property of the gradient phase: the functionalI decreases between any two successive iterations. Methods to determine the gradient stepsize in an optimal fashion are discussed. Examples are presented for both the fixed-final-time case and the free-final-time case. The numerical results show the rapid convergence characteristics of the sequential gradient-restoration algorithm.