This review provides numerical and graphical information about many (but by no means all) of the physical and electronic properties of GaAs that are useful to those engaged in experimental research and development on this material. The emphasis is on properties of GaAs itself, and the host of effects associated with the presence of specific impurities and defects is excluded from coverage. The geometry of the sphalerite lattice and of the first Brillouin zone of reciprocal space are used to pave the way for material concerning elastic moduli, speeds of sound, and phonon dispersion curves. A section on thermal properties includes material on the phase diagram and liquidus curve, thermal expansion coefficient as a function of temperature, specific heat and equivalent Debye temperature behavior, and thermal conduction. The discussion of optical properties focusses on dispersion of the dielectric constant from low frequencies [κ0(300)=12.85] through the reststrahlen range to the intrinsic edge, and on the ass...

Semiconducting and other major properties of gallium arsenide

Ellipsometric measurements of the dielectric constant of undoped Ge were performed between 1.25 and 5.6 eV in the temperature range of 100 to 850 K. The dependence of the ${E}_{1}$, ${E}_{1}+{\ensuremath{\Delta}}_{1}$, ${E}_{0}^{\ensuremath{'}}$, and ${E}_{2}$ critical energies on temperature was obtained. It can be represented either with Varshni's empirical formula or with an expression proportional to the Bose-Einstein statistical factor of an average phonon. Broadening parameters, amplitudes, and phase angles for the corresponding critical points were also obtained. A decrease of the excitonic interaction with increasing temperature was found. The results are discussed in the light of recent calculations of the effect of temperature on the band structure of Ge containing Debye-Waller and self-energy contributions.

Temperature dependence of the dielectric function of germanium

Using the tight-binding method, the valence band structures and densities of states for C, Si, Ge, GaAs, and ZnSe are calculated. Very good agreement is obtained with other calculations when all nearest- and one second-nearest-neighbor interactions are included. The effects of the various interactions on the density of states are discussed.



Mit der „tight-binding”-Methode werden die Valenzbandstrukturen und Zustandsdichten fur C, Si, Ge, GaAs und ZnSe berechnet. Sehr gute Ubereinstimmung mit anderen Rechnungen wird erreicht, wenn alle Wechselwirkungen mit nachsten und nur eine Wechselwirkung mit zweitnachsten Nachbarn berucksichtigt werden. Die Einflusse verschiedener Wechselwirkungen auf die Zustandsdichte werden diskutiert.

/pdf/tight-binding-calculations-of-the-valence-bands-of-diamond-4zl41qznaz.pdf

Tight‐binding calculations of the valence bands of diamond and zincblende crystals

The band gaps, band structure, and excited‐state (exciton) energies of CdS, GaAs, and GaP semiconductor clusters are calculated using pseudopotentials. In addition, the sensitivity of the exciton energies to the size, shape, crystal structure, and lattice constant of the unit cell are investigated. The calculated exciton energies of CdS clusters are in excellent agreement with experiment over a wide range of cluster sizes. Also, the exciton states of small CdS clusters are sensitive to whether their crystal structure is zinc blende or hexagonal. Such a sensitivity is absent in large CdS clusters. Furthermore, small GaAs clusters are shown to exhibit anomalous redshift of their absorption spectra, in sharp contrast to CdS and large GaAs clusters whose spectra always shift to blue with decreasing cluster size. Finally, the lowest‐energy non‐Franck–Condon transition in GaP clusters always shifts to blue with decreasing cluster size, whereas the higher‐energy Franck–Condon transition in small clusters exhibits the anomalous redshift. These novel findings reveal that (1) the optical spectroscopy of semiconductor clusters is strongly material and crystal structure dependent; (2) the spectroscopy of small clusters is dramatically different from those of large clusters and bulk; and (3) these effects cannot be explained, even qualitatively, using the effective‐mass approximation.

Quantum confinement effects in semiconductor clusters

Photoemission spectroscopy is commonly applied to study the band structure of solids by measuring the kinetic energy versus angular distribution of the photoemitted electrons. Here, we apply this experimental technique to characterize discrete orbitals of large π-conjugated molecules. By measuring the photoemission intensity from a constant initial-state energy over a hemispherical region, we generate reciprocal space maps of the emitting orbital density. We demonstrate that the real-space electron distribution of molecular orbitals in both a crystalline pentacene film and a chemisorbed p-sexiphenyl monolayer can be obtained from a simple Fourier transform of the measurement data. The results are in good agreement with density functional calculations.

https://physik.uni-graz.at/~pep/Publications/Puschnig2009a.pdf

Reconstruction of molecular orbital densities from photoemission data.

We determine bandwidths and critical-point positions with respect to the valence-band edge for the valence bands of several extensively studied semiconductors (with an accuracy of about 0.3 eV) by using photoemission densities of states derived from photoemission spectra obtained in the 24-78-eV photon energy range. These photoemission spectra were obtained using synchrotron radiation from an electron storage ring; a double-pass, electrostatic, electron energy analyzer; and samples cleaved and measured in situ in ultrahigh vacuum. We give a detailed description of the data-reduction techniques by which electronic state densities and certain valence-band feature positions are determined from photoemission spectra. This description includes a discussion of the effect of various phenomena such as optical-transition-probability variations and Auger-emission peaks. Tables are presented which compare our valence-band-position values with the results of both empirical calculations fit to optical data as well as ab initio calculations. We find systematic and significant differences between experiment and calculations fit only to optical data, these differences increasing with crystal ionicity, while ab initio calculations generally give a better fit to experiment.

Photoemission spectroscopy using synchrotron radiation. I. Overviews of valence-band structure for Ge, GaAs, GaP, InSb, ZnSe, CdTe, and Agl

The valence-band structure of the mixed silicate ${\mathrm{Si}}_{x}{\mathrm{Ge}}_{1\ensuremath{-}x}{\mathrm{O}}_{2}$ was investigated for a range of compositions by x-ray and uv photoemission spectroscopy Structure in the valence band, which is derived from the oxygen nonbonding orbitals and from the Si-O bonding orbital, is seen to move continuously in initial energy in going from Si${\mathrm{O}}_{2}$ to Ge${\mathrm{O}}_{2}$ The width of the nonbonding bands at the top of the valence band decreases from about 33 eV in Si${\mathrm{O}}_{2}$ to 20 eV in Ge${\mathrm{O}}_{2}$ in which separate peaks are no longer resolved The decrease in width of the nonbonding bands is correlated with an increase in the average oxygen-oxygen separation from 262 to 285 \AA{} The results indicate that the width of the nonbonding bands is largely due to oxygen wave-function overlap The valence bands measured by photoemission from the mixed oxide are not a superposition of Si${\mathrm{O}}_{2}$ and Ge${\mathrm{O}}_{2}$ valence bands On the other hand, band-gap excitations, as determined by energy-loss satellites on the photoemission spectra, were found to be a superposition of those for Si${\mathrm{O}}_{2}$ and Ge${\mathrm{O}}_{2}$, indicating local conduction-band states centered about the Ge and Si sites

Electronic structure of Si O 2 , Si x Ge 1 − x O 2 , and Ge O 2 from photoemission spectroscopy

This paper considers several implications of the approximation that the total amplitude for photoemission from an oriented molecule is the sum of the amplitude of coherent emission from spherically symmetric regions at the atomic centers. We show that this approximation, which is increasingly valid at high energies (accessible, for example, to synchrotron-radiation sources) is readily calculable using tabulated initial-state atomic functions and simple solutions of the radial Schr\"odinger equation. We illustrate the fact that in special but important cases (e.g., oriented benzine), one can obtain the angular distribution of photoemission at fixed final energy without any recourse to atomic-photoemission-amplitude calculations. Finally, we emphasize that fact that the orthogonalized-plane-wave approach, which is a special case of the independent-atomic-center approximation, fails in two regards---neglect of atomic phase shifts, and the neglect of initial-state core-region wave-function behavior, which is crucial at high energies. This latter point is illustrated by comparing orthogonalized-plane-wave calculations using Slater atomic functions (which emphasize the bonding-region wave-function behavior), and hydrogenic wave functions, which can have radial nodes and core-region behavior which can more closely approximate the behavior of the true initial-state wave functions.

Angle-resolved photoemission from molecules in the independent-atomic-center approximation

We analyze photoemission spectra for Ge obtained for photon energies $6.5\ensuremath{\le}h\ensuremath{\nu}\ensuremath{\le}25$ eV and determine the position of energy bands at symmetry points for both filled and empty bands within 1 Ry of the gap. This experimental band structure is obtained using a recently developed anisotropic direct-transition model of photoemission applicable to cleaved single-crystal semiconductors. For the band-structure determination we also employ a direct transition analysis of optical spectra obtained by others. We fit nonlocal-pseudopotential calculations to the experimentally-determined band positions, and thereby determine the importance of both energy and $l=2$, angular momentum nonlocality in the pseudopotential. Our results for the position of highlying conduction-band states suggest 0 to + 10% self-energy (exchange-correlation) corrections to the energy of electrons excited into the conduction bands. Energy bands which provide a good fit to the experimental band positions are used, along with pseudo-wave-function matrix elements, to calculate various physical properties (photoemission spectra, optical-response functions, and one-electron state densities), and the results of these calculations are compared with experiment. The quality of the fits obtained indicates that the electronic excited-state (and ground-state) properties of Ge for excitations far from the gap are described well by a one-electron model.

Warren David Grobman

Papers

Photoemission spectroscopy using synchrotron radiation. I. Overviews of valence-band structure for Ge, GaAs, GaP, InSb, ZnSe, CdTe, and Agl

Electronic structure of Si O 2 , Si x Ge 1 − x O 2 , and Ge O 2 from photoemission spectroscopy

Angle-resolved photoemission from molecules in the independent-atomic-center approximation

Photoemission spectroscopy using synchrotron radiation. II. The electronic structure of germanium

Spectroscopically Observed Valence Mixing in SmS and Related Compounds