A semi-analytical model is constructed for single- and multi-junction solar cells. This model incorporates the key performance aspects of practical devices, including nonradiative recombination, photon recycling within a given junction, spontaneous emission coupling between junctions, and non-step-like absorptance and emittance with below-bandgap tail absorption. Four typical planar structures with the combinations of a smooth/textured top surface and an absorbing/reflecting substrate (or backside surface) are investigated, through which the extracted power and four types of fundamental loss mechanisms, transmission, thermalization, spatial-relaxation, and recombination loss are analyzed for both single- and multi-junction solar cells. The below-bandgap tail absorption increases the short-circuit current but decreases the output and open-circuit voltage. Using a straightforward formulism this model provides the initial design parameters and the achievable efficiencies for both single- and multiple-junction solar cells over a wide range of material quality. The achievable efficiency limits calculated using the best reported materials and AM1.5 G one sun for GaAs and Si single-junction solar cells are, respectively, 27.4 and 21.1% for semiconductor slabs with a flat surface and a non-reflecting index-matched absorbing substrate, and 30.8 and 26.4% for semiconductor slabs with a textured surface and an ideal 100% reflecting backside surface. Two important design rules for both single- and multi-junction solar cells are established: i) the optimal junction thickness decreases and the optimal bandgap energy increases when nonradiative recombination increases; and ii) the optimal junction thickness increases and the optimal bandgap energy decreases for higher solar concentrations.

The limitation in sample size and the choice in parameters for the application of the micro-crystal (MC) model have been tested and discussed. The testing was performed through the fitting of calculated Raman spectra by using the MC model with the observed Raman spectra for nearly uniform-sized nanocrystalline (NC)-Si of smaller than 2 nm. The results showed that the good fittings could not be obtained, if the phonon dispersion curves !.q/ and Raman linewidths 0 of bulk-crystals (BCs) were used, i.e. if the bulk-like approximation (BLA) was used in the fitting. Moreover, an analysis of the fitting results indicates that when the size of NC-Si is smaller than about 4 nm, the MC model based on -BLA will be invalid.

It was found in ZnO nanoparticles of nearly uniform size that the Raman frequencies of both the polar A1(LO) and nonpolar E2(H) modes do not shift with particle sizes. This observation is in perfect agreement with previous theoretical prediction. The observed result is consistent with the prediction that the size insensitivity of optic vibration frequency in polar nanosemiconductors originates from the Fröhlich interaction, which exists only in polar nanosemiconductors.

Four‐junction solar cells are designed using lattice‐matched II–VI (ZnCdSeTe) and III–V (AlGaAsSb) semiconductors grown on GaSb substrates. These materials have a zinc blende crystal structure, similar thermal expansion coefficients, and bandgaps that cover the entire solar spectrum. Numerical simulations of the energy conversion efficiencies of various designs for both the AM0 and AM1.5D spectra are performed using published material parameters. These results indicate that the achievable 1 sun AM0 efficiency is 43% for an optimal design and 40% for a more practical design; for comparison the ideal limit provided by Henry's model is 49%. While for the AM1.5D spectrum an optimal design can reach 46% under 1 sun and 55% under 1000 suns while a more practical design can reach 44 and 54%, respectively; for comparison Henry's model gives 51 and 62%, respectively. Copyright © 2010 John Wiley & Sons, Ltd.

A lack of dependence of the Raman frequency of optical vibrational modes on excitation wavelength in polar nanosemiconductors was observed. This is in contrast to the earlier observed dependence in nonpolar nanomaterials: carbon nanotubes and Si nanowires. This difference has been ascribed to the different crystallographic natures of their Raman spectra: crystalline for nonpolar and amorphous for polar nanosemiconductors. The result has been explored theoretically to the Raman spectra being insensitive to sample sizes and thus indicates that the size confinement effect, a basic effect in nanomaterials, does not exhibit in the optical vibrational modes of polar nanosemiconductors.