Theoretical aspects of the luminescence of porous silicon

The luminescence in the visible range of porous silicon is analyzed in the hypothesis of quantum confinement. We calculate the electronic and optical properties of silicon crystallites and wires with sizes between 0 and 4.5 nm. The band-gap energies of such confined systems are in agreement with the photon energies observed in luminescence. We calculate the radiative recombination times of the confined excitons. We conclude that experimental nonradiative processes in porous silicon are more efficient than calculated radiative ones at T=300 K. The high photoluminescence efficiency of porous silicon is due to the small probability of finding a nonradiative recombination center in silicon nanocrystallites. Recently, it has been proposed that the low-temperature dependence of the experimental radiative decay time of the luminescence of porous silicon could be explained by the exchange splitting in the fundamental exciton. We show that the influence of the valley-orbit splitting cannot be excluded. The sharp optical-absorption edge above 3.0 eV is not proof of the molecular origin of the properties of porous silicon because silicon nanostructures present a similar absorption spectrum. We calculate the nonradiative capture of electrons or holes on silicon dangling bonds and show that it is very dependent on the confinement. We find that the presence of one dangling bond at the surface of a crystallite in porous silicon must destroy its luminescent properties above 1.1 eV but can produce a luminescence below 1.1 eV due to a radiative capture on the dangling bond.