Alloy Band Research Articles

Influence of hydrogen on vibrational and optical properties of a-Si1-xHx alloys.

The effect of hydrogen on the optical gap and network short-range order has been obtained in a-${\mathrm{Si}}_{1\mathrm{\ensuremath{-}}\mathrm{x}}$${\mathrm{H}}_{\mathrm{x}}$ alloys by optical-absorption and Raman-scattering measurements. Distinct differences are observed in the role of hydrogen for homogeneous chemical-vapor-deposited (HMCVD) and rf-diode-sputtered (RFS) films. A comparison of the correlation between the Raman TO peak width and the optical gap for films with and without H allows a separation of the roles of H-induced bond-angle ordering and alloy energy-band changes. While the primary effect of H in the HMCVD films is alloy band changes for xg0.05, RFS films indicate comparable structural ordering effects. The effect of hydrogen on the optical gap is found to be \ensuremath{\sim}3--4 times greater than that in a-${\mathrm{Ge}}_{1\mathrm{\ensuremath{-}}\mathrm{x}}$${\mathrm{H}}_{\mathrm{x}}$ alloys. In addition to the ordering of bond angles about Si atoms, the Raman spectra indicate changes in the TA-TO and LA-TO ratios due to hydrogen incorporation. These are interpreted in terms of the relative reductions in the number of bond-bending modes as the concentration of triads of Si atoms decreases with hydrogen.

Raman scattering of Ni and Cr amorphous disilicides

The Raman spectra and reflectivity of r.f. sputtered amorphous semimetallic a-NiSi 2 and a-CrSi 2 thin films are reported. Substantial differences between the two amorphous disilicide Raman spectra are associated with changes in the phonon density-of-states as a result of variations in Si( p)-transition metal ( d) interactions. The greater width of the bands in the Cr alloy suggests a less ordered local environment. Neither spectrum indicates distinct a-Si-like bands, demonstrating that significant short range order changes have occured.

Electrical measurements on MBE grown Si/Si 1− xGe x heterojunctions

Electrical measurements were made on Si/Si 1−l x Ge x , n-n heterojunctions grown by MBE on Si 〈100〉 substrates. The carrier concentration profile derived from capacitance-voltage measurements gives a determination of the conduction band discontinuity. We found for a strained Si layer on a relaxed Si 0.75Ge 0.25 layer that the Si conduction band is lower by 0.12 eV than the alloy band. In the case where the alloy layer is strained, the conduction band is still lower on the Si side but the magnitude of the discontinuity is small. C- V measurements on a partially relaxed layer show deep levels appearing at the interface. Cross-sectional TEM revealed defects at the interface which are correlated with the electrical states.

Reduced symmetry and the band structure of semiconductor alloys.

It is suggested that the neglect of the reduced local symmetry at some sites in alloys can be a serious shortcoming in the calculation of alloy band structures. A phenomenological scheme which takes account of the reduced local symmetry while preserving long-range order through the virtual-crystal approximation is described for calculating the band structure of alloys. It is shown that the method leads to an overall band structure similar to the corresponding virtual-crystal approximation but gives better agreement with experiment for the splitting of the valence bands at \ensuremath{\Gamma}.

Recent developments in the strained layer epitaxy of germanium–silicon alloys

This paper provides a brief review of recent work on strained layer GexSi1−x including modeling of critical thicknesses for single layers and superlattices; calculations and measurements on strain induced alteration of alloy band gap and heterostructure band alignment; and application to modulation-doped field effect transistors (MODFET) and 1.3 μm photodetector devices.

Surface states of ternary semiconductor alloys: Effect of alloy fluctuations in one-dimensional models with realistic atoms.

Calculations of the local electronic surface densities of states (DOS) of semi-infinite one-dimensional chain models for ternary semiconductor alloys are presented to illustrate the effects of alloy fluctuations on semiconductor surface states. The atoms are modeled with a basis set of s and p levels which produces the correct level mixing and energy-band dispersion.Alloys with randomly occupied cation sites are considered. The embedded-cluster approach is employed to incorporate the alloy effects on states localized near the chain end. The cluster calculations are performed by including the contributions from all possible configurations of a cluster at the end of the semi-infinite chain. These results are compared with results obtained using the coherent-potential approximation (CPA) and the virtual-crystal approximation (VCA). The CPA is implemented with one self-energy appropriate for all cation sites in the bulk and different self-energies for each cation site near the surface. In practice the most accurate CPA results can be obtained by treating only the four cation sites nearest the surface differently from the bulk.Examples which are typical of the weak and strong alloy limits are considered. In the first limit the alloy constituents are similar direct-gap semiconductors. In the strong alloy limit one constituent has an inverted band structure, as HgTe does, and the other is a normal semiconductor such as CdTe. The cluster calculation for the surface DOS is sensitive to fluctuations in the environment local to the chain end. In the weak alloy limit, the surface DOS in the band gap has a bimodal distribution with the peaks located near the energies of the surface states of the constituents. In the strong alloy limit, the HgTe-like constituent of the HgCdTe-like alloy has no surface states while the surface states of the CdTe-like constituent lie above the alloy band gap.As a consequence, the alloy surface DOS has a resonance structure above the band edge and band tailing below the edge. The CPA reproduces qualitatively the structure of the cluster calculations. The VCA predicts well-defined surface-state peaks in the band gap. These peaks are related to the average of the bimodal distribution in the weak alloy limit but cannot be consistently identified with any resonance structure or band tailing in the strong alloy limit.

The structure of adiabatic shear bands in a titanium alloy

A study has been made of the structure of adiabatic shear zones in Ti-6% Al-4% V with different parent microstructures, resulting from ballistic impact of steel spheres. Metallographic examination of well-developed shear bands showed that they consisted of zones of intense shearing distortion of the original microstructure, modified by the effects of elevated temperature. An analogy is made between their structure and that of the white-etching shear zones observed in steels. Unlike steels, however, there was no clear evidence in this alloy to suggest that the shear bands in the α + β microstructures had undergone a martensitic phase transformation. The structure of the shear zones in an α' martensite parent alloy appeared to be a tempered form of the original microstructure.

Calculation of electronic density of states for amorphous ZrCu and ZrNi alloys

The author has recently developed a simple scheme based on the recursion method which allows the relative position of the d bands in an amorphous transition-metal alloy to be determined if the average charge transfer is known. To find the hopping integrals the author uses tabulated values of the overlap parameters (dd sigma , dd pi , and dd delta ) obtained from a fit to the bands of the pure metals. This scheme allows a parameterized Hamiltonian to be built to represent the transition-metal alloys, avoiding the need for a detailed fitting of LCAO bands to the bands of a corresponding crystalline compound. In the procedure the overlap parameters are tabulated for the pure metal and charge transfer is either known or neglected. Therefore the results obtained for the density of states of the alloy are well defined and parameter free. Here the author uses the above procedure to obtain the electronic structure of amorphous ZrCu ( alpha -ZrCu) and alpha -ZrNi alloys for two different compositions. The results are compared with those obtained by other methods and with experiment. The calculations are in good agreement with existing UPS and XPS results for these alloys. The values obtained for the density of states at the Fermi level, N(EF), for alpha -ZrCu are also compatible with those inferred from experimental data for melt-spun samples.

Positron study of the Fermi surface of aNb50Mo50disordered alloy

The Fermi surface of a ${\mathrm{Nb}}_{50}$${\mathrm{Mo}}_{50}$ disordered alloy has been studied by two-dimensional angular correlation of positron-annihilation radiation. The measured Fermi surface is well defined and several dimensions of the ellipsoidal hole pocket at $N$ have been determined. The results are compared with the predictions of various alloy band calculations.

Occurrence of shear bands in the order-disordered FeCo-2V alloy

Shear bands have been observed at about 35° to the rolling plane when the alloy was cold rolled to about 5 to 10% reduction. Although this alloy consists essentially of a bcc lattice in the disordered state, brought about by quenching, and of a CsCl type of lattice in the ordered state, the bands can similarly be obtained for both states. From TEM observations, the planes on which the bands may be formed seemed to be {011} and {112}, corresponding to the slip and twinning planes, respectively, of bcc metals and alloys. An attempt has been made to interpret the formation of the shear bands: since cross slip may be difficult to obtain in the alloy because of the existence of the superlattice, sites of stress concentrations are produced relatively easily, and these are relaxed by the localized shearing and formation of shear bands.

A study of the disordered alloy Ga1-xInxAs1-yPyby the variational multiple scattering method

The authors apply, for the first time, the variational multiple scattering method (VMS) for the band calculation of a tridimensional disordered alloy. They calculated for the quaternary system Ga1-xInxAs1-yPy, with 0.605<or=x<or=0.917 and 0.094<or=y<or=0.845, lattice matched to InP. For all the alloy compositions, a unique potential for each element was used, chosen so that reasonable band structures for the pure semiconductors GaAs and InP result. No other adjustment was made. The energy density of states and the alloy bands were calculated for six compositions. The calculation is very fast when compared with similar calculations by the average t-matrix approximation method (ATA). The VMS method predicts, with the above potential choice, two separate inner peaks for the energy density of states, corresponding to the s As and s P bands, which is amenable to experimental verification. The importance of the self-consistent process in determining the degree of P and As p-band mixing is clearly shown.

EMPIRICALLY ADJUSTED ZONE-VARIATIONAL METHOD FOR BAND CALCULATIONS

An empirically adjusted zone-variational method for energy band calculations is presented in this paper. The potential outside the sphere is adjusted empirically to make the resulted energy band closer to experimental results. This method has been used to compute the energy bands of Si and Ge. The results agree with the experimental ones very well. In accordance with the special feature of using differen methods to solve the Schrodinger equation in different potential regions, a new computational model for alloy band is also suggested. By using the same adjusting parameters, the energy band of Si-Ge alloy is computed, which is superior to the OPW result. Thus it can be seen that the zone-variational method is an effective and prospective computational method in band theory and has considerable wide application.

Efficient visible luminescence from hydrogenated amorphous silicon

We report on optical and structural properties of a new class of efficient visible light emitting semiconductors: low-defect-density a-SiH x alloys. Films are prepared by HOMOCVD from silane and RF plasma from disilane. Optical gaps increase monotonically with H incorporation and decreasing substrate temperatures (200-25°C), while spin densities remain low. New broadband luminescence develops in the alloy band tails and shifts to higher energies with the gap. At ∼ 40% H, gaps reach 2.55 eV and yellow-orange PL peaks near 2.05 eV. Efficient PL persists at room temperature. Emission is attributed to band-tail excitons and phonon cooperation.

Efficient visible photoluminescence in the binary a-Si:Hx alloy system

We report the photoluminescence (PL) and structural properties of a new class of efficient visible-light-emitting semiconductors: low defect density a-Si:Hx alloys. For films prepared by the (thermal) homogeneous chemical vapor deposition (HOMOCVD) method, new broadband PL develops for x&amp;gt;0.3, reaching a peak emission energy of 2.05 eV for a hydrogen content x=0.66 (40 at. % H). We attribute the wide gaps to the influence of Si–H bonding on the density of states near the valence band edge. We ascribe the new PL process to band-to-band recombination from within the alloy band tails. This emission persists at room temperature with an integrated intensity comparable to conventional light-emitting diode (LED) materials. Qualitatively similar results are obtained for low-temperature-deposited rf plasma films prepared from Si2H6, but not from SiH4. We show that a low Si dangling bond concentration is the key factor, for all the different film types, to achieving efficient luminescence.

CPA band calculation for (Hg, Cd) Te

A calculation of the electronic structure of (Hg,Cd)Te alloys in the coherent‐potential approximation (CPA) is presented. The problem is divided into three tasks: determining accurate band structures including relativistic terms for the constituent compounds (HgTe and CdTe), characterizing the alloy disorder, and executing the alloy calculation. These tasks are facilitated through the use of a set of symmetrized, orthonormal, local orbitals. We shall introduce the calculational procedure and discuss the results obtained to date. These include the calculated band structures for HgTe and CdTe, the local alloy disorder parameters, the alloy band energies, and density of states calculated in the virtual crystal approximation and in a simplified version of CPA.

Calculation of the electronic properties of pseudo-binary semiconductor alloys

A coherent-potential approximation calculation is presented for the concentration variation of the pseudo-binary semiconductor alloy band structures and scattering rates. The procedure yields accurate results that compare favorably with experiments for ${\mathrm{Al}}_{x}{\mathrm{Ga}}_{1\ensuremath{-}x}\mathrm{As}$, $\mathrm{Ga}{\mathrm{P}}_{x}{\mathrm{As}}_{1\ensuremath{-}x}$, and ${\mathrm{In}}_{x}{\mathrm{Ga}}_{1\ensuremath{-}x}\mathrm{P}$. A theory for the scattering energy in the Brooks mobility formula is also given.

Electronic structure of pseudobinary semiconductor alloysAlxGa1−xAs,GaPxAs1−x, andGaxIn1−xP

A method for calculating detailed electronic properties of the pseudobinary III-V compound semiconductor alloys is presented. The technique begins with realistic band structures obtained for the constituent compounds by fitting the band-gap symmetry-point energies and effective masses to experimental data, where they are available, and to more sophisticated theoretical results. Then the coherent-potential approximation is used to calculate the alloy band structures and scattering rates. Detailed comparisons between the theoretical predictions and experimental data for three alloys ${\mathrm{Al}}_{x}{\mathrm{Ga}}_{1\ensuremath{-}x}\mathrm{As}$, $\mathrm{Ga}{\mathrm{P}}_{x}{\mathrm{As}}_{1\ensuremath{-}x}$, and ${\mathrm{Ga}}_{x}{\mathrm{In}}_{1\ensuremath{-}x}\mathrm{P}$ demonstrate the quantitative nature of the method. Bowing parameters for the $\ensuremath{\Gamma}$, $X$, and $L$ gaps and the direct-to-indirect band-gap crossover concentrations are all predicted to within the present degree of experimental certainty.

A model for hydrogen-assisted crack initiation on planar shear bands in near-alpha titanium alloys

A model was proposed which explains subsurface cleavage facet formation in large grained, planar slip titanium alloys by the buildup of large hydrostatic stresses at the tip of long, blocked shear bands. These shear bands can result from creep deformation. Hydrogen is attracted to the hydrostatic stress field and can initiate a cleavage facet possibly by the locally increased hydrogen concentration or more likely by actual hydride formation and subsequent cracking of the hydride. The model is qualitatively consistent with several experimental results and implies that the evaluation of dwell debits in alloys must be conducted at stresses where some creep deformation is possible.

The effect of grain size on the fatigue crack propagation behavior of age-hardened alloys in inert and corrosive environment

The influence of grain size on the fatigue crack propagation behavior of age-hardened Al-Zn-Mg-Cu and Ti-Al alloys has been investigated in vacuum and 3.5% NaCl solution. It was found that an increase in grain size reduced the crack propagation rates in vacuum considerably for both alloys. Crack propagation in vacuum occurred along slip bands in both alloys. In NaCl solution crack propagation rates increased by at least one order of magnitude for both alloys as compared to the results in vacuum. In NaCl solution the grain size dependence of crack propagation rates almost disappeared for the Al-alloy whereas the differences in propagation rates between different grain sizes still existed for the Ti-alloy. Crack propagation in NaCl solution occurred along slip bands for the Ti-alloy similar as in vacuum, while the Al-alloy showed fracture along grain boundaries at low ΔK-values with a transition to slip band fracture at higher ΔK-values. This transition in the Al-alloy occurred at ΔK-values which coincided approximately with the observed threshold values for the corresponding grain size found in vacuum.

On the behaviour of the Cd d bands in Au, Ag and Hg alloys

The valence band densities of states of selected HgCd alloys have been studied using 40.81 eV HeII photoelectron spectroscopy. The binding energies, splittings and widths of the Hg 5d5/2, 3/2 and Cd 4d5/2, 3/2 peaks have been studied for various alloy concentrations and are discussed in conjunction with previous work on AuCd and AgCd alloys and with the results of self-consistent bandstructure calculations. It has been found that the increase in full width at half maximum of the valence d bands with increasing concentration varies approximately as Z/R2, where Z is the like-atom coordination number and R is the like-atom nearest-neighbour distance.