The radiative recombination coefficient B(T) of intrinsic crystalline silicon is determined as a function of temperature over the temperature range 77-300 K. We observe that B(T) decreases as a function of temperature and we compare our results to previously published contradictory data from the literature. The radiative recombination coefficient is calculated from the absorption coefficient for band-to-band transitions, which we determine at different temperatures from photoluminescence spectra measured on planar high resistivity float zone silicon wafers. Photoluminescence spectra could be detected over a large range of more than five orders of magnitude, which allowed us to determine extremely low values of the absorption coefficient in the spectral range where absorption processes are accompanied by the simultaneous absorption of up to four phonons.

Significant performance increase for silicon solar cells is reported. This has been achieved by a combination of several mechanisms. One is the reduction of recombination at cell surfaces using atomic hydrogen passivation of silicon/silicon dioxide interfaces. Joule resistive losses in the cell have been reduced by a process which allows different thickness for fine and coarse features in the top cell metallization. Finally, reflection losses have been reduced by the use of a double layer antireflection coating. For successful incorporation, this required the development of techniques for growing the surface passivating oxide very thin, without reducing its passivation qualities. The cells display a monochromatic light energy conversion efficiency of 46.3% for 1.04 μm wavelength light, also the highest ever for a silicon devices.

Current issues of numerical modeling of crystalline silicon solar cells are reviewed. Numerical modeling has been applied to Si solar cells since the early days of computer modeling and has recently become widely used in the photovoltaics (PV) industry. Simulations are used to analyze fabricated cells and to predict effects due to device changes. Hence, they may accelerate cell optimization and provide quantitative data e.g. of potentially possible improvements, which may form a base for the decision making on development strategies. However, to achieve sufficiently high prediction capabilities, several models had to be refined specifically to PV demands, such as the intrinsic carrier density, minority carrier mobility, recombination at passivated surfaces, and optical models. Currently, the most unresolved issue is the modeling of the emitter layer on textured surfaces, the hole minority carrier mobility, Auger recombination at low dopant densities and intermediate injection levels, and fine-tuned band parameters as a function of temperature. Also, it is recommended that the widely used software in the PV community, PC1D should be extended to FermiDirac statistics.

No abstract

No abstract