P. C. Taylor

Evidence for spatially indirect recombination in Ga0.52In0.48P

In previous work we have identified a near‐gap photoluminescence in Ga0.52In0.48P which exhibits a strong dependence of emission energy on excitation intensity (‘‘moving emission’’) and correlated its presence and strength to conditions of growth. In this work we extend our investigations to the rise and decay lifetimes associated with the moving and nonmoving components of the emission. The two processes proceed simultaneously at the same energy. For the moving emission, the time constants scale approximately linearly with excitation intensity. Decaying luminescence can, in most cases, be well fitted with one or two exponentials with time constants as long as milliseconds. The rising luminescence is typically slower and in some cases has a nonmonotonic first time derivative. These results are discussed in terms of existing models of the microstructure of ordered Ga0.52In0.48P.

A general structural model for semiconducting glasses

Abstract A general structural model is proposed for multi-component, covalent amorphous and glassy systems. The model demonstrates the importance of a generalized “8 - n ” rule and of the average number of valence electrons in amorphous systems. This model explains glass forming tendencies in a wide range of multi-component systems.

Excitation intensity dependence of photoluminescence in Ga0.52In0.48P

The excitation intensity dependence of the photoluminescence (PL) from Ga0.52In0.48P grown by organometallic vapor phase epitaxy on GaAs substrates has been investigated as a function of epitaxial layer growth temperature and substrate orientation. It is well known that the degree of ordering and the band‐gap energy of this material are functions of growth conditions. We report here on a PL emission which shifts rapidly with excitation intensity. The rate of emission shift is also a function of growth conditions including substrate orientation. There is, however, no significant correlation between the band‐gap energy and the rate of emission shift in Ga0.52In0.48P. This PL shift is explained in terms of band filling of potential fluctuations that are associated with a microstructure consisting of ordered domains within a disordered matrix.

Photoluminescence, photoluminescence excitation, and resonant Raman spectroscopy of disordered and ordered Ga0.52In0.48P

Ordering in the CuPt structure is known to significantly reduce the band gap of Ga0.52In0.48P as well as induce a number of unusual details in its optical properties, including long, excitation‐intensity‐dependent lifetimes and an excitation‐intensity‐dependent emission energy. We report photoluminescence (PL), photoluminescence excitation (PLE), and resonant Raman measurements performed on ordered and disordered Ga0.52In0.48P. The dominant high energy emission process at low temperature in disordered Ga0.52In0.48P is established to be excitonic, but the exciton trapping energy is not unique. PLE from ordered Ga0.52In0.48P shows significant tailing of electronic states into the band gap and a ‘‘band edge’’ which depends on detection energy. The dominant radiative process in ordered Ga0.52In0.48P is not excitonic. A large increase in the Stokes shift between the absorption edge (band gap) and PL emission peak occurs when the material orders. Hence, low temperature PL is determined to be a particularly poor...

The formal valence shell model for structure of amorphous semiconductors

Abstract We have proposed a structural model, the Formal Valence Shell (FVS) model, for amorphous semiconductors. This model is based on assumptions of non-doping and the domination of covalency. A concept of formal valence is used. The model states that the coordination number of each constituent atom is eight minus the formal valenve for the atom and that the average coordination number of a given composition is eight minus the average valence of the composition. The model is applicable to systems which contain non-metal (4 ⩽ N ⩽7) as well as metal atoms (1 ⩽ N ⩽3), where N is the valence or column number in the periodic table. The model therefore generalizes the conventional 8 - N rule which is applicable only to non-metal atoms. Sublattice assignments in ternaries are proposed. Given a set of constituent elements and their sublattice assignments, this model can provide coordination numbers, stoichiometric compositions, and trend of glass formation. I-V-VI ternaries are used for illustration, and other systems (binaries and ternaries) are addressed briefly. The calculated results generally agree with experimental data.

Blue, green and red emission from Ce3+, Tb3+ and Eu3+ ions in amorphous GaN and AlN thin films

Abstract We report strong blue, green and red emission from Ce3+, Tb3+ and Eu3+ ions, respectively, at room temperature in amorphous GaN and AlN thin films prepared by DC magnetron co-sputtering. We observe sharp characteristic emission peaks of intra-4f-shell transitions of the Tb3+ ions ( 5 D 4 → 7 F 3,4,5,6 transitions) and Eu3+ ions ( 5 D 0 → 7 F 1,2,3,4 transitions) and a strong but broad peak of 5d–4f emission from Ce3+ ions over the temperature range 2–300 K. The photoluminescence decay time from the excited 5 D 4 state of Tb3+ ions in a-AlN is 1.1 ms, which is similar to the radiative lifetime of this state in glasses. The broad photoluminescence peak centered at ∼400 nm in a-GaN shows a weak temperature dependence and may be attributed to the intrinsic tail-to-tail transitions in the amorphous matrix. The possible origin of the deep states, close to midgap in a-AlN, is briefly discussed.

Photoluminescence in strained InGaAs-GaAs heterostructures

Photoluminescence in strained InxGa1−xAs‐GaAs single heterostructures, grown by molecular‐beam epitaxy, is studied, and the critical layer thickness is determined for a range of In mole fractions. The critical thicknesses are compared with similar values measured on the same layers by double‐crystal x‐ray diffraction. Both techniques give essentially the same results. The photoluminescence line shapes are discussed for layer thicknesses below and above the critical thickness.Photoluminescence in strained InxGa1−xAs‐GaAs single heterostructures, grown by molecular‐beam epitaxy, is studied, and the critical layer thickness is determined for a range of In mole fractions. The critical thicknesses are compared with similar values measured on the same layers by double‐crystal x‐ray diffraction. Both techniques give essentially the same results. The photoluminescence line shapes are discussed for layer thicknesses below and above the critical thickness.

Doping superlattices in organometallic vapor phase epitaxial InP

Doping superlattices (nipi structures) have been grown in InP using organometallic vapor phase epitaxy in an atmospheric pressure reactor using trimethylindium and phosphine in a hydrogen ambient. The n‐type and p‐type dopants were diethyltellurium and dimethylzinc, respectively. The 4‐K photoluminescence spectra at various excitation intensities are presented for a structure consisting of six 200‐A layers with doping levels of 1×1018 and 2×1018 cm−3 for the n and p layers. The luminescence peak is found to occur at energies considerably less than the band gap of InP and to move to higher energies with increased excitation intensity, as expected for doping superlattices where the band gap, which is indirect in real space, increases with increasing excited carrier concentration. The total photoluminescence signal decays in several steps, each exponential, with time constants ranging from 6×10−8 to 7×10−4 s at 4 K, typical of these spatially indirect band‐gap materials.

Persistent photoconductivity in a-Si1−xSx:H at low sulfur concentration

Abstract We present the observation in homogeneous, unlayered a-Si 1−x S x :H alloys of a light-induced increase of the dark conductivity which is similar to that found in compensated and doping-modulated a-Si:H. This effect occurs for small sulfur concentrations. The excess conductivity can be annealed at about 200°C independent of the hydrogen concentration in the films.

Potential p-type doping in amorphous chalcogenide films

Sputtered amorphous films of stoichiometric (Cu2/3Se1/3)x(As2/5Se3/5)1-x were produced with compositions where the Se atoms should be tetrahedrally coordinated and the majority of the intrinsic defects transformed from negative to positive effective electron-electron correlation energies, Ueff. The incorporation of oxygen appears to produce, in large densities, acceptors that are responsible for a three orders of magnitude increase in conductivity with a corresponding decrease in electrical activation energy while maintaining a constant optical bandgap. The conduction process is a thermally activated, extended state mechanism and is determined to be p-type. We speculate that adding highly ionic oxygen atoms produces Cu “vacancies” (Se-Se nearest neighbor) by pulling the Cu atoms from their regular cation sites into interstitial sites near the oxygen atoms. These arguments are roughly analgous to a known doping process in II-VI semiconductors.