The thermal resistivity, Seebeck coefficient, electrical resistivity, and Hall mobility of Ge-Si alloys have been measured throughout the Ge-Si alloy system as functions of impurity concentration in the range 2×1018−4×1020cm−3, and of temperature in the range 300°–1300°K. A qualitative interpretation of these properties is given. For power conversion, boron and phosphorus were found to be useful p-type and n-type impurities, respectively, because of their high solid solubilities. At 1200°K, the maximum values of the dimensionless figure of merit zT were 0.8 for p-type Ge0.15-Si0.85 alloy doped to 2.1×1020cm−3 holes, and 1.0 for n-type Ge0.15-Si0.85 alloy doped to 2.7×1020cm−3 electrons. The maximum over-all efficiency of a stable generator operating between 300°–1200°K, using the best p-type and n-type materials was computed to be 10%.

The electronic structure of nanocrystalline Si which shows visible photoluminescence is calculated using the density-functional approach for finite structures. Except for geometry this is the same theory as for first-principles band structures of semiconductors and other solids. Our results for clusters ranging up to 706 Si atoms suggest that the band gap scales linearly with I. , where I. is the cluster diameter.For such clusters it is found that dipole transitions across the gap are symmetry allowed. The finite structures thus show a direct band gap which is considerably larger than the one of bulk silicon. For larger clusters we find a strong decrease of oscillator strength, consistent with the occurrence of the indirect gap in the bulk limit.The recent discovery of visible photoluminescence from small silicon structures in anodically etched porous silicon' has greatly increased interest in these particular kinds of fine structures, as well as in small semiconductor particles ' where the same effects have been observed.An exciting perspective of this discovery is that lightemitting devices based on this effect appear feasible within the well-established silicon technology. To understand this effect, knowledge about the band structure of fine silicon structures is required. In particular, questions to be addressed are whether silicon can become intrinsically a "direct" semiconductor when in nanocrystalline clusters (or "porous") and if quantum confinement can modify the energy gap such that visible light is produced as experimentally observed. Our previous work ' has dealt with quantum confinement in small silicon particles on the basis of density-functional studies. The principal result has been that an energy gap in the visible range should occur intrinsically for small particles of a few nm in diameter. Our present work shall give a more detailed account of the theoretical findings, with due attention paid to the role of the self-energy correction to the gap and to the oscillator strength as a function of particle size. In the Ineantime, other work based on firstprinciples calculations for the quantum-wire geometry has appeared, ' where the conclusions concerning the intrinsic gap widening are consistent with our results. Although a controversy has arisen whether the observed effect is due to quantum confinement, as originally proposed, or whether visible photoluminescence is due to an extrinsic effect because of the unbiquitous presence of siloxenelike compounds, the focus of interest in the present work is on the fundamental question whether quantum confinement in fine silicon structures can lead to symmetry-allowed optical transitions across the gap with an energy in the visible range. As the content of this work definitely shows, three-dimensionally confined silicon particles have symmetry-allowed optical transitions across the gap. In the case of periodic crystal structures, this property is termed direct gap. Moreover, this work shows that the gap widens such that particles in the experimentally report...

The size dependence of the energy band gap for hydrogen saturated silicon clusters, wires and slabs are calculated using all electron density functional theory. The hydrogen saturation is considered as a model for a wider band gap insulator enclosing the silicon structures. With this perspective in mind, an effective mass model with finite barriers for both valence and conduction band is found to semiquantatively account for the numerical findings.

No abstract

The theory developed by Ziman for the scattering of phonons by electrons is extended to high temperatures using the formalism of Klemens and Callaway. The theory is applied to the experimental results recently published by Dismukes et al. on the effect of doping on the thermal conductivity of Ge-Si alloys. After subtracting the electronic contribution from the measured thermal conductivity, the resulting lattice conductivity is analyzed. In describing the effect of doping, the deformation potential is the only free parameter; its value is adjusted for each sample to obtain agreement with the experimental data at 500°K, which is close to the Debye temperature. The theory then predicts the correct temperature dependence of the lattice thermal conductivity. The deformation potentials, derived in this manner, are found to be higher for w-type than for p-type material, and to increase with carrier concentration. For lightly doped />-type and w-type material, values of 1.2 and 1.6 eV were obtained, respectively, which compare well with the available literature data.